Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico

FACULTADE DE BIOLOXÍA Departamento de Edafoloxía e Química Agrícola Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage Patricia Sanmartín Sánchez 2012 Este trabajo de investigación doctoral fue financiado por el Ministerio de Ciencia e Innovación (MICINN) dentro del programa de Formación de Personal Investigador (FPI-MICINN). Referencia: BES-2007-16996, en el marco del proyecto de investigación titulado: Desarrollo de nuevas técnicas no destructivas de cuantificación y estudio de la evolución de la colonización biológica sobre construcciones graníticas. Referencia: BIA2006-02233. Investigador Principal: Beatriz Prieto Lamas. This doctoral research was funded by the Spanish Ministry of Science and Innovation (MICINN) as part of the Spanish Scientific program FPI (Formación de Personal Investigador)-MICINN. Fellowship reference code: BES-2007-16996. The study was carried out within the framework of a research project entitled Desarrollo de nuevas técnicas no destructivas de cuantificación y estudio de la evolución de la colonización biológica sobre construcciones graníticas. Project reference code: BIA2006-02233. Principal Investigator: Beatriz Prieto Lamas. Departamento de Edafoloxía e Química Agrícola Tesis Doctoral PhD Thesis Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage Patricia Sanmartín Sánchez 2012 Departamento de Edafoloxía e Química Agrícola Dña. Beatriz Prieto Lamas y Dña. Benita Silva Hermo, Profesoras del departamento de Edafoloxía e química agrícola de la Universidade de Santiago de Compostela. CERTIFICAN: Que la presente memoria titulada Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico (Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage) presentada por Dña. Patricia Sanmartín Sánchez para optar al grado de Doctora por la Universidade de Santiago de Compostela, fue realizada bajo nuestra dirección y supervisión, y consideramos que reúne todos los requisitos y condiciones necesarias para ser presentada como trabajo de Tesis doctoral. Para que así conste, expedimos la presente certificación en Santiago de Compostela el día veintiséis de abril de dos mil doce. Fdo. Dra. Beatriz Prieto Lamas Fdo. Dra. Benita Silva Hermo Contenidos Contents Resumen ........................................................................................................................9 Summary ...................................................................................................................... 19 Introducción General .................................................................................................... 29 I. Del modo planctónico a la formación y crecimiento del biofilm ....................................................... 29 II. Interacción biofilm-sustrato: colonización biológica de superficies expuestas al exterior................ 33 III. Detección y cuantificación de la colonización biológica y desarrollo del biofilm. Medida instrumental del color y sistema CIELAB ........................................................................................ 41 General Introduction ..................................................................................................... 51 I. From a planktonic lifestyle to formation and growth of biofilms ....................................................... 51 II. Biofilm-substrate interaction: biological colonization of surfaces exposed to outdoor conditions........................................................................................................................................ 54 III. Detection and quantification of biological colonization and biofilm development. Instrumental color measurements and the CIELAB color system ................................................... 62 Investigación / Research 1ª línea de trabajo Desarrollo de la metodología de medida y caracterización del color de las rocas graníticas 1st line of research Fine-tuning of the methodology for measuring and characterizing the color of granite surfaces Chapter 1. Measuring the color of granite rocks: A proposed procedure ........................................ 73 PRIETO, B.; SANMARTÍN, P.; SILVA, B.; MARTÍNEZ-VERDÚ, F. Chapter 2. Effect of surface finish on roughness, color, and gloss of ornamental granites ............. 83 SANMARTÍN, P.; SILVA, B.; PRIETO, B. 2ª línea de trabajo Desarrollo de la metodología de medida y caracterización del color de las cianobacterias 2nd line of research Fine-tuning of the methodology for measuring and characterizing the color of cyanobacteria Chapter 3. Color of cyanobacteria: some methodological aspects.................................................. 95 PRIETO, B.; SANMARTÍN, P.; AIRA, N.; SILVA, B. Chapter 4. Relationship between color and pigment production in two stone biofilm-forming cyanobacteria (Nostoc sp. PCC 9104 and Nostoc sp. PCC 9025).............................. 105 SANMARTÍN, P.; AIRA, N.; DEVESA-REY, R.; SILVA, B.; PRIETO, B. Chapter 5. Color measurements as a reliable method for estimating chlorophyll degradation to phaeopigments............................................................................................................ 119 SANMARTÍN, P.; VILLA, F.; SILVA, B.; CAPPITELLI, F.; PRIETO, B. 3ª línea de trabajo Detección y cuantificación de los organismos sobre las construcciones 3rd line of research Detection and quantification of microorganisms colonizing the surface of buildings and monuments Chapter 6 .Quantification of phototrophic biomass on rocks: optimization of chlorophyll-a extraction by response surface methodology.............................................................. 131 FERNÁNDEZ-SILVA, I.; SANMARTÍN, P.; SILVA, B.; MOLDES, A.; PRIETO, B. Chapter 7. Spectrophotometric color measurement for early detection and monitoring of greening on granite buildings ...................................................................................... 143 SANMARTÍN, P.; VÁZQUEZ-NION, D.; SILVA, B; PRIETO, B. 4ª línea de trabajo Aplicación práctica de la medida del color a otros sustratos inorgánicos naturales 4th line of research Practical application of color measurement on other naturally occurring inorganic substrates Chapter 8. Color characterization of roofing slates from the Iberian Peninsula for restoration purposes ..................................................................................................................... 155 PRIETO, B.; FERRER, P.; SANMARTÍN, P.; CÁRDENES, V.; SILVA, B. Chapter 9. El color como indicador de la intensidad de los incendios en suelos de Galicia: Resultados preliminares / Color as an indicator of fire intensity on soils of Galicia, NW Spain: Preliminary results..................................................................................... 169 SANMARTÍN, P.; CANCELO-GONZÁLEZ, J.; RIAL, M.E.; SILVA, B.; DÍAZ-FIERROS, F.; PRIETO, B. Chapter 10. Nondestructive assessment of phytopigments in riverbed sediments by the use of instrumental color measurements ........................................................................... 177 SANMARTÍN, P.; DEVESA-REY, R.; PRIETO, B.; BARRAL, M.T. Discusión General .......................................................................................................191 General Discussion......................................................................................................207 Conclusiones ...............................................................................................................223 Conclusions .................................................................................................................229 Referencias / References ............................................................................................235 Agradecimientos / Acknowledgements........................................................................251 Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 14 – 3ª. Interacción superficies graníticas-cianobacterias: detección y cuantificación de los organismos sobre las construcciones En las dos primeras líneas de trabajo se trataron los aspectos colorimétricos de las rocas graníticas y las cianobacterias por separado, trabajando siempre en condiciones de laboratorio. Sin embargo, para cubrir el objetivo principal del proyecto de tesis es necesario probar la metodología desarrollada en casos reales de colonización biológica sobre superficies graníticas. Para tal fin, y de forma previa al estudio de la detección y monitorización del crecimiento de biofilms sobre fachadas graníticas empleando la medida del color, fue necesario resolver ciertos problemas metodológicos en el estudio de la interacción roca-microorganismos. Aunque la cuantificación de la clorofila-a es el método de referencia por excelencia para la cuantificación de biomasa fototrófica, cuando se trata de estimar la cantidad de clorofila-a presente en un material pétreo colonizado, este método presenta algunos problemas ya que la extracción de este pigmento de la roca da lugar a una subestimación en la medida, tal como apuntaban Prieto et al. (2004). Estos autores advertían sobre la necesidad de optimizar el método de extracción de la clorofila-a en sustratos rocosos y señalaban que su recuperación no era completa debido a que la propia roca dificultaba la extracción, resultando el límite de detección de este pigmento demasiado alto para la cuantificación de la colonización temprana. La investigación en la mejora del método de extracción de la clorofila-a de biofilms fototróficos sobre sustratos pétreos (Capítulo 6) fue necesaria para poder comparar este método con el método basado en la medida del color puesto a punto en esta tesis. Con este objetivo, se llevó a cabo una optimización del método de extracción de clorofila-a sobre roca granítica usando el método de superficie de respuesta de un diseño factorial incompleto de 3er orden, que predice las condiciones experimentales óptimas para la máxima extracción de clorofila-a (Fernández-Silva et al., 2011). Se evaluaron tres pretratamientos mecánicos para determinar el mejor procedimiento en la extracción de clorofila-a con dimetil sulfóxido (empleado como solvente) en bloques de granito inoculados con un cultivo mixto de tres cepas cianobacterianas que forman biofilms sobre rocas graníticas, Nostoc sp. PCC 9104, Nostoc sp. PCC 9025 y Scytonema sp. CCC 9801. Dos de ellos, A y B, incluían la trituración de la probeta de granito seguida, en el caso del pretratamiento B de una sonicación de los fragmentos de roca y extracción en baño de ultrasonidos. En el tercer pretratamiento, C, se sustituía la trituración de la probeta por la aplicación de una sonda de ultrasonidos (sonicador) al extractante que baña los bloques enteros. Para cada pretratamiento se utilizó un diseño de Box-Behnken para 3 factores experimentales o variables de optimización: (1) relación volumen de extractante/volumen de muestra, (2) temperatura durante la extracción y (3) tiempo de extracción, siendo la clorofila-a extraída la Resumen – 15 – variable dependiente o de respuesta. Tras la incubación en DMSO con las correspondientes condiciones experimentales, se determinó la concentración de clorofila-a por espectrofotometría UV-Vis. El análisis de los resultados determinó que: (i) la aplicación de ultrasonidos mejora la extracción de clorofila-a de los biofilms desarrollados sobre sustratos rocosos; (ii) la precisión de la medida usando el método de espectrofotometría UV-Vis para el cálculo de la concentración de clorofila-a disminuye cuando las muestras están altamente diluidas, por lo que son más convenientes los métodos que requieren un menor volumen de extractante, y, (iii) la temperatura de extracción es el factor experimental más importante (comparado con los otros dos analizados) en la extracción de clorofila-a de los biofilms desarrollados sobre sustratos rocosos. Con todo ello se propone un método que implica la aplicación de ultrasonidos directamente a la muestra intacta sin necesidad de triturarla (con lo que el método de extracción de clorofila-a deja de ser un método destructivo y pasa a ser un método meramente invasivo) seguido de una incubación en 0,43 ml de DMSO/cm2 de muestra, a 63ºC durante 40 minutos. Esta mejora fue empleada en un experimento de confirmación en el que se consiguió la recuperación del 90% de la cantidad real (i.e. inoculada) de clorofila-a, un incremento sustancial frente al 68% que era el máximo de recuperación en trabajos anteriores. El Capítulo 7 se centra en la detección y monitorización del crecimiento de biofilms fototróficos y epilíticos en la fachada de granito de un edificio (Sanmartín et al., 2012). Se trata de un caso real orientado a la conservación preventiva, que proporciona una base para el establecimiento de criterios de detección precoz del verdín y monitorización de su desarrollo en los edificios graníticos, utilizando las variaciones de color registradas mediante un espectrofotómetro portátil y representadas en el espacio de color CIELAB teniendo en cuenta el grupo de coordenadas escalares (L*a*b*). Con este trabajo se establecen por primera vez los límites de percepción del enverdecimiento (greening) sobre una superficie de granito en términos de incremento de los parámetros de color CIELAB L* (ΔL*), a* (Δa*) y b* (Δb*), cantidad de organismos (en peso seco) y cantidad de clorofila-a extraída. Asimismo, se analiza la evolución de los valores de L*, a* y b* en un caso real de limpieza y recolonización natural de una fachada granítica, con el objetivo de determinar qué parámetro/s CIELAB refleja/n mejor el comienzo del enverdecimiento. Los resultados indican que es el parámetro b* (que define el componente amarillo-azul de un color) el que proporciona mayor información, ya que es el que detecta la colonización más temprana y el que varía en mayor magnitud con el tiempo por lo que condiciona en mayor medida el cambio de color total (ΔE*ab, ΔE*ab = [(ΔL*)2 + (Δa*)2 + (Δb*)2]1/2). Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 16 – Además, los resultados expuestos en este Capítulo 7 presentan utilidad práctica en la gestión y mantenimiento de edificaciones con problemas de biodeterioro, al permitir detectar la colonización cuando ésta es aún imperceptible al ojo humano e incluir indicaciones relacionadas con el momento de limpieza y aplicación de tratamientos para detener o revertir la colonización fototrófica en las fachadas. 4ª. Aplicación práctica de la medida del color a otros sustratos inorgánicos naturales: rocas pizarrosas, suelos y sedimentos La metodología planteada en el Capítulo 1 para el análisis de los factores instrumentales y de superficie que afectan a la determinación del color de cualquier tipo de material heterogéneo en color y textura, fue empleada para el estudio del color de rocas pizarrosas en el Capítulo 8. Partiendo de la premisa de que los criterios para la selección de rocas para reemplazar las existentes en edificios históricos deben estar basados en parámetros geológicos, geotécnicos y estéticos, entre los cuales se encuentra el color, en el Capítulo 8 se realizó un exhaustivo y útil estudio colorimétrico de las pizarras de techar de la Península Ibérica (Prieto et al., 2011). Así, se determinó el color de cincuenta variedades comerciales de pizarra de techar, extraídas de doce distritos mineros repartidos entre España y Portugal, con un espectrofotómetro portátil, considerando el espacio de color CIELAB y teniendo en cuenta el grupo de coordenadas escalares (L*a*b*) y el grupo de coordenadas cilíndricas o polares (L*C*abhab). El análisis de las condiciones de medida indicó que para la determinación del color de este tipo de rocas, el número de medidas mínimo necesario es sustancialmente menor que en las rocas graníticas (Sanmartín et al., 2010b) lo cual es, sin duda, debido a la mayor homogeneidad de color y textura de las rocas pizarrosas con respecto a las graníticas. Asimismo, se definió la zona tridimensional del espacio de color CIELAB en la que se encuadran las pizarras de techar de la Península Ibérica. La delimitación de esta zona ofrece las mismas aplicaciones prácticas que las señaladas en el Capítulo 2 para las rocas graníticas ornamentales. Por su parte, el tono angular (hab) resultó el parámetro de color CIELAB más útil en cuanto a la formación de grupos de pizarras para cubierta colorimétricamente similares y el componente especular excluido (SCE) es el modo más sensible para detectar diferencias de color (ΔE*ab) entre dos variedades de pizarra. Además, se establecieron las similitudes y diferencias en el color y la microestructura de las diferentes variedades comerciales, así como el grado de aceptación para la sustitución de un tipo de pizarra por otra. Así, cinco variedades de pizarra pueden reemplazar a y ser reemplazadas por prácticamente el resto de las cincuenta variedades y una única variedad no puede reemplazar ni ser reemplazada por ninguna otra pizarra. Resumen – 17 – El Capítulo 9 se planteó como una primera aproximación al estudio de la relación existente entre el color que un suelo adquiere tras un incendio y la temperatura que alcanza, con el fin de valorar el uso del color del suelo como indicador de la intensidad de un incendio (Sanmartín et al., 2010c). A pesar de que han sido muchos los casos en los que se han descrito variaciones apreciables en el color del suelo tras un incendio (como el ennegrecimiento debido a carbonización parcial o total de la materia orgánica y el enrojecimiento debido a la deshidratación de los óxidos de hierro), la variación de los parámetros de color y su relación con la intensidad del fuego no habían sido hasta ahora determinados en el contexto de los suelos gallegos. Por este motivo, en este estudio se realizó una quema controlada en el laboratorio de muestras inalteradas del horizonte superficial de un regosol úmbrico empleando lámparas de infrarrojos para alcanzar temperaturas de 200ºC y 400ºC a 1 cm de la superficie. Sobre la superficie de cada una de las muestras quemadas y de las muestras sin quemar (que fueron empleadas como control) se determinó el color en 220 puntos con un colorímetro portátil, considerando los espacios de color CIELAB y Munsell. Los resultados obtenidos mostraron un descenso de la claridad en las muestras quemadas con respecto a las sin quemar, reflejado en los valores de Value Munsell y L* CIELAB, así como un descenso del croma, marcado por el Chroma Munsell y C*ab CIELAB, probablemente debidos a la carbonización de la materia orgánica. Además, se produjo un empardecimiento-enrojecimiento del suelo reflejado en los valores de Hue Munsell, el descenso de hab CIELAB, el aumento de a* CIELAB y el descenso de b* CIELAB, que debe atribuirse simplemente a la desaparición de los componentes orgánicos del suelo que enmascaraban el color de los componentes inorgánicos coloreados, principalmente óxidos de hierro. La comparación entre los valores de los suelos quemados a 200ºC y a 400ºC, temperaturas que se corresponden con incendios de intensidad ligera y moderada respectivamente, muestra que los únicos parámetros que establecen diferencias significativas son aquellos que se refieren al Hue Munsell y hab CIELAB. En este caso, las diferencias pueden ser asignadas principalmente a variaciones en el estado de deshidratación de los compuestos de hierro como el cambio de goethita, de un color anaranjado, a hematite, de color rojo sangre. Esto parece indicar que será el tono el parámetro a tener en cuenta en el uso del color como indicador de la intensidad de incendios en suelos de Galicia. Para finalizar esta cuarta línea orientada a la aplicación práctica de la medida del color a otros sustratos inorgánicos naturales y concluir así el trabajo de tesis, el último capítulo presenta el uso de las medidas de color en la estimación de los fitopigmentos contenidos en sedimentos de río. La cuantificación de fitopigmentos en los sedimentos de río recibe la atención de los especialistas porque proporciona información sobre el efecto sinérgico que provocan los nutrientes, en especial nitrógeno y fósforo, en el crecimiento de las algas y por tanto sobre la Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 18 – eutrofización del río, un descriptor válido del estado trófico y de la calidad de las aguas y sedimentos. El desarrollo de una metodología no-destructiva, rápida, sencilla y económica para la cuantificación del contenido de fitopigmentos en los sedimentos de río, evitando la extracción y análisis químico de los mismos, sería de gran utilidad en este campo. El Capítulo 10 muestra la validez de las medidas de color realizadas con un espectrofotómetro portátil y representadas en el espacio de color CIELAB en la estimación precisa de los fitopigmentos en sedimentos de río (Sanmartín et al., 2011c). Se recogieron muestras de sedimentos con cilindros metálicos en tres zonas de baja pendiente, favorables para la deposición de sedimentos, a lo largo del cauce del río Anllóns (Galicia). En ellas se realizaron mediciones de color representadas en el espacio de color CIELAB teniendo en cuenta el grupo de coordenadas escalares (L*a*b*) y el grupo de coordenadas cilíndricas o polares (L*C*abhab), para posteriormente extraer con dimetil sulfóxido y cuantificar por espectrofotometría UV-Vis su contenido en fitopigmentos (clorofila-a, clorofilab, feopigmentos y carotenoides totales). Los resultados mostraron que los parámetros de color CIELAB se correlacionaban significativamente con el contenido en fitopigmentos permitiendo el desarrollo de ecuaciones de predicción para el contenido en clorofila-a, clorofila-b, carotenoides totales y contenido total en fitopigmentos a partir del parámetro a* y el parámetro C*ab. El mejor ajuste se alcanzó en la predicción de clorofila-a (Chl-a): Chl-a = 46,4 – 70,6 log10(a*) y Chl-a = –15,7 + 238,1 (1/C*ab) con un valor ajustado de R2 de 0,84. – 19 – Summary In the conservation of buildings and structures, particularly those of historical and artistic importance, it is essential to mitigate the effects of biological colonization, since this supposes not only an aesthetic problem, but can also cause mineralogical, chemical and physical damage to the materials, with the consequent loss of value. Biological colonization and biofilm formation are unavoidable because they occur as a result of interactions between the building material and the environment. However, early detection of biological colonization, and better knowledge of the developmental stage would suppose a great advantage as regards management of the maintenance and rehabilitation of the buildings, as this would tackle the problem at the first stages, even before the colonization could be assessed visually. Moreover, in the case of heavily colonized structures, early detection would be extremely useful for monitoring the efficacy of the treatments used to eliminate microorganisms and for detecting any recolonization. Therefore, the overall aim of this doctoral research project was to develop and implement a novel methodology for the early detection of biofilm formation on granite stone in buildings and monuments, based on color measurements. This novel method satisfies various requirements that validate its application to heritage structures: it is non-destructive (allowing further analysis on single samples); it can be applied on site (avoiding sampling and allowing results to be obtained immediately); it is relatively inexpensive and easily applied, so that any unskilled operator with minimal training can perform the measurements. The hypothesis that led to the approach used in this project is based on the idea that biological colonization on structures can be quantified by determining the color generated by the microorganisms involved. This hypothesis is supported by the findings of previous studies. One of these studies demonstrated the existence of a direct relationship between the quantity and physiological state of phototrophic organisms and the color change generated when they are deposited on a surface (Prieto et al., 2002; Spain Patent No. P200002956, 2000). Another study, which compared different methods of quantifying the phototrophic biomass on rocky substrates, revealed a close correlation between the results obtained with the most appropriate methods (i.e. those based on the quantification of chlorophyll-a) and those results obtained from the color quantification (Prieto et al., 2004). Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 20 – The present research focused on the pioneering microorganisms in the colonization of rocky substrates, specifically cyanobacteria, because these are the main constituents of the phototrophic biofilms that form on building façades. The study specifically evaluated biological colonization on granite stonework, which is a complex case because of the spatially heterogeneous color formed by the different colors of the constituent minerals. Three different aspects were considered in order to achieve the overall aim of the study: 1. Fine-tuning of the methodology for measuring and characterizing the color of granite surfaces (Prieto et al., 2010a; Sanmartín et al., 2011a). 2. Fine-tuning of the methodology for measuring and characterizing the color of cyanobacteria (Prieto et al., 2010b; Sanmartín et al., 2010a, 2011b). 3. Interactions between granite surfaces and cyanobacteria: detection and quantification of microorganisms colonizing the surface of buildings and monuments (Fernandez-Silva et al., 2011; Sanmartín et al., 2012). These three lines eventually converged in a 4th line of research that addressed the practical application of color measurement in other natural inorganic substrates of interest in the field of environmental studies. Some examples of practical applications include color characterization of roofing slates as a decision-making tool for restoration work (Sanmartín et al., 2010b; Prieto et al., 2011), color characterization of burned soils as an indicator of fire intensity (Sanmartín et al., 2010c) and color measurements in riverbed sediments as an indicator of phytopigment content (Sanmartín et al., 2011c). 1. Fine-tuning of the methodology for measuring and characterizing the color of granite surfaces The first research line (Chapter 1) involves the fine-tuning of a measurement protocol for the color characterization of ornamental granite, originally developed by Prieto et al. (2010a). The aim of this study was to determine the minimum area and the minimum number of measurements required for objective characterization of the color of granite rocks. A spectrophotometer and a tristimulus colorimeter were used to measure the color of granite samples, and the reflectance measurements were analyzed with the CIELAB color parameters, using Cartesian (L*a*b*) coordinates. Three parameters were considered as variable factors: the type of rock (color and texture), the surface finish (polished, honed, sawn, and flamed), and the target area of the measurement head (circular apertures of diameter 5, 8, 10, and 50 mm). The results revealed that, although all of the factors considered affected the minimal area and the number of measurements required, the different circular apertures of both instruments can be disregarded if the number of measurements and area recommended in this study are used. Summary – 21 – One of the most important findings of this research is that it provides a methodology for determining the conditions for color measurements on heterogeneous surfaces of different textures, and can be used in different materials. The methodology was successfully used with roofing slates (Prieto et al., 2011) and in the 2nd research line in this doctoral research project, in which the methodology for measuring and characterizing the color of cyanobacteria was finetuned. All ornamental stone in built structures has a certain type of surface finish, and Chapter 2 reports how the effects of the type of commercial finish on the color, gloss and roughness of ornamental granite were studied and a relationship between these three aesthetic parameters was established (Sanmartín et al., 2011a). The results demonstrated that different surface finishes (polished, honed, sawn, and flamed or bush hammered) almost always produce notable differences in color, especially in the lightness parameter (L*); the magnitude of these differences depends on the color of the ornamental granite and is greatest in dark colored rocks. However, the variation in the color parameters (L*a*b*C*abhab) with the different surface finishes did not depend on roughness, and no general conclusions could be drawn regarding the influence of the roughness on the color of ornamental granite. Gloss values were affected by the color, but in a different way for smooth and rough surfaces. Variations in gloss also depended on the mineral composition of the rock. Gloss and roughness were inversely related, but only within the range of low roughness values. The influence of the inclusion or exclusion of the specular component in the measurement mode is also discussed in Chapter 2. It was found that exclusion of the specular component (SCE mode) magnifies the differences in color produced by the different surface finishes, so that this mode is preferred over the specular component included (SCI) mode. The results obtained also enable identification of the three-dimensional color area of the CIELAB space in which the color of the ornamental granites under study is defined. The information obtained in this study should be taken into account when aesthetic criteria are used to select ornamental granites for building and construction materials, and also to characterize digital cameras for use as colorimeters in studies of cultural heritage monuments. 2. Fine-tuning of the methodology for measuring and characterizing the color of cyanobacteria Following the methodology developed for measuring color in heterogeneous surfaces (Chapter 1), the second research line in this doctoral research project established a methodology for measuring the color of biofilms on solid surfaces (Chapter 3). For this purpose, samples of a mixed culture of three types of cyanobacteria, Nostoc sp. PCC 9104, Nostoc sp. PCC 9025 and Scytonema sp. CCC 9801, which form biofilms on granite Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 22 – surfaces, were filtered under vacuum through nitrocellulose filter discs, to simulate the composition and characteristics of a biofilm. The color of the areas coated by the microorganisms was measured, as described in Chapter 1, with two contact-type color measuring devices (spectrophotometer and colorimeter), and the reflectance measurements were analyzed with the CIELAB color parameters, using Cartesian (L*a*b*) coordinates, in order to analyze the effect of the coverage by organisms and their moisture content on their color. The results showed that the minimum number of measurements required increased with increasing heterogeneity of the color of the measured area, which depended on the concentration of the microorganisms, and decreased with the diameter of the measuring head. To control for the influence of the heterogeneity of the color of the area measured, the color of cyanobacteria should be measured on filters that are completely covered by the microorganisms, so that the color of the filter is totally concealed. The L* values were greatly affected by the moisture content of the samples. Thus, the values of L* decreased exponentially with moisture contents of up to 50%, which can be considered as the point from which a stable asypmtote is reached on the corresponding graph. The color of cyanobacteria should therefore be determined in samples with moisture contents of more than 50%. Considering all of the above, at least 10 consecutive measurements/9.62cm2 should be made at different randomly selected points on the surface of filters completely covered by films of cyanobacteria in which the moisture contents are higher than 50%. Chapter 4 describes how the relationship between color and composition and/or abundance of photosynthetic pigments and the efficiency of color measurements was used to evaluate the extent to pigment production is affected by environmental parameters such as light intensity, combined nitrogen and nutrient availability. This was tested with two cyanobacteria, Nostoc sp. strains PCC 9104 and PCC 9025, which form biofilms on stone surfaces. Both strains were acclimated, in aerated batch cultures for 2 weeks, to three different culture media, BG-11, BG110, and BG-110/10, at either high or low light intensity. The chlorophyll-a, total carotenoid, and phycocyanin contents were measured throughout the experiment, together with variations in the color of the cyanobacteria, which were represented in the CIELAB color space, using Cartesian (L*a*b*) and cylindrical (L*C*abhab) coordinates. The results confirmed that the CIELAB color parameters are correlated with pigment content in such a way that variations in the latter are reflected as variations in color. However, not all CIELAB color parameters are equally sensitive to changes in pigment concentrations. Thus, L*, and to a lesser extent a*, is the most informative CIELAB color parameter, and hab, which refers to the dominant wavelength, represents the major color perception attribute. The color data recorded in Chapter 4 indicate the three-dimensional color area of the CIELAB space in which both strains define their color. This is the first time that objective color values of Summary – 23 – different species of cyanobacteria growing in different environmental conditions have been obtained, and is a preliminary step in defining the color gamut of cyanobacteria, which is essential information for the development of any color-based methodology for monitoring cyanobacterial growth. The second research line (Chapter 5) involves the color characterization of phototrophic microorganisms (cyanobacteria) for analysis of their physiological state after treatment with biocides (Sanmartín et al., 2011b). The effectiveness of biocides against phototrophic organisms can be tested by the degree of chlorophyll degradation. When chlorophyll degrades, it forms a series of degradation products or phaeopigments, and the relation between phaeopigments and chlorophyll can be established by the phaeophytination indexes A435/415 and A665/665a. Chapter 5 discusses the suitability the CIELAB color measurements as a reliable method for monitoring the effectiveness of the chemical biocide Biotin T® (which is commercially available and widely used in the field of stone cultural heritage materials) against a cyanobacterial strain (Nostoc sp. PCC 9104) in both planktonic and biofilm mode of growth. Determination of the phaeophytination indexes (A435/415 and A665/665a), which have proved useful for describing the degradation of chlorophyll-a to phaeopigments in Nostoc sp. PCC 9104 planktonic and biofilm biocide susceptibility assays, could be substituted by color measurements as there is a close relation between these and the color parameters. This confirms that the effectiveness of a biocide and the physiological state of a microorganism can be determined by color characterization. Of the CIELAB parameters, L* appeared to be the best for describing the biocidal activity of Biotin T® against Nostoc sp. in both planktonic and biofilm mode of growth. Moreover, a* and C*ab are correlated with the A665/665a index in both modes of bacterial growth. 3. Interactions between granite surfaces and cyanobacteria: detection and quantification of the microorganisms on the surfaces of buildings and monuments In the first two research lines of this doctoral study, the colorimetric aspects of granite and cyanobacteria were dealt with separately, always working under laboratory conditions. However, to meet the main objective of the study, the methodology must be tested in real cases of biological colonization on granite surfaces. For this purpose, and prior to the study of the detection and monitoring of the growth of biofilms on granite façades using color measurements, it was necessary to resolve certain methodological problems in the study of granite-microorganism interactions. Although quantification of chlorophyll-a is considered the reference method of excellence for quantifying phototrophic biomass, in the case of estimating the amount of chlorophyll-a present in colonized rocky substrata, a relatively high value is obtained as the lower limit of detection; this prevents early detection of stone biocolonization below this lower limit, as pointed out by Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 30 – La definición actual de biofilm o biopelícula describe un concepto mucho más complejo según el cual se trata de “comunidades universales de microorganismos (bacterias, hongos, cianobacterias, algas, protozoos), complejas e interdependientes, asociadas a superficies húmedas (vivas o inertes) o a interfases (columnas de agua sujetas a gradientes), mediante una matriz polimérica (EPS) de glicoproteínas extracelulares, de consistencia viscosa, generada ex profeso y recorrida por canales de agua, los cuales permiten un intercambio eficiente de agua, nutrientes y gases entre las poblaciones constitutivas y el ambiente exterior” (Costerton, 2007). Así, la capacidad de formación del biofilm no está restringida a ningún grupo específico de microorganismos, considerándose que, bajo condiciones ambientales adecuadas, todos los microorganismos son capaces de formar biofilms. A este respecto, se ha demostrado que muchas células necesitan estar sujetas a las superficies (formando parte de un biofilm) para sobrevivir y proliferar; este requisito de supervivencia es conocido como “dependencia del anclaje” (anchorage dependence, Pierres et al., 2002). La principal característica que distingue las biopelículas del crecimiento planctónico es la existencia de una matriz de sustancia polimérica extracelular (EPS) embebiendo el total de los componentes del biofilm y adhiriendo éste a la superficie. La matriz constituye un medio complejo esencialmente compuesto por agua (hasta un 97%), que contiene además polímeros extracelulares, polisacáridos, proteínas, ácidos nucleicos, lípidos/fosfolípidos, nutrientes absorbidos y metabolitos. Pese a que la matriz extracelular es la característica definitoria de la biopelícula, su papel aún no es totalmente comprendido en la actualidad. A este respecto, la mayoría de los autores defienden que la función primaria de la matriz extracelular es la de proteger la biopelícula de la acción de agentes antimicrobianos, a través de la neutralización química o la creación de una barrera de difusión que impida a los microorganismos ser alcanzados por estos (Lindsay & von Holy, 2006), y de los cambios ambientales adversos en términos de humedad, temperatura, presión osmótica y pH. Así por ejemplo, se sabe que las cianobacterias incrementan la producción de polímeros extracelulares frente a la limitación de luz o de nutrientes (Albertano et al., 2003). Otros autores sin embargo, consideran que el papel más relevante de los productos poliméricos extracelulares (EPS) ocurre durante las etapas iniciales de la formación del biofilm al facilitar la unión de las células al sustrato (Decho, 2000; Barranguet et al., 2005), dar cohesión al biofilm incipiente y facilitar la interacción entre sus distintos componentes (Flemming & Wingender, 2001). Los microorganismos que forman un biofilm también se diferencian de sus homólogos en suspensión en que presentan comportamientos fenotípicos particulares que los diferencian de sus formas libres en su modo de ubicación, agrupación y organización colonial, movilidad, adaptación metabólica y ambiental, y en su forma y tasa de crecimiento (Donlan, 2002). Durante el complejo proceso de adhesión, los microorganismos alteran sus características fenotípicas como respuesta a la proximidad a una superficie. Además, durante los estados iniciales de la formación de la biopelícula, los microorganismos sésiles se encuentran en una Introducción General – 31 – yuxtaposición estable con microorganismos de la misma especie y de otras especies, formando de esta manera microcolonias. La interrelación que se produce en estas microcolonias hace posible el intercambio de material genético, que resulta en un cambio global en el fenotipo de la biopelícula (Stolz, 2000), así como la creación de ambientes heterogéneos desde el punto de vista físico-químico en los cuales los microorganismos asociados a la biopelícula presentan características fisiológicas distintas de las planctónicas. Muchos microorganismos alteran sus procesos fisiológicos para adaptarse a estos nichos particulares, por ejemplo creciendo anaeróbicamente (Stoodley et al., 2002). Además, los diferentes microorganismos contenidos en la biopelícula responden a las condiciones de sus microambientes específicos, presentando diferentes patrones de crecimiento. La cooperación fisiológica es el factor principal que ayuda a conformar la estructura y establecer la eventual yuxtaposición, haciendo de las biopelículas maduras, adheridas a las superficies, comunidades microbianas muy eficientes (Costerton et al., 1994). En un contexto amplio, la formación de un biofilm (Figura 1) implica que los microorganismos constituyentes, al dejar su condición de células móviles (planctónicas), conforman poblaciones no estratificadas de microorganismos, con una estructura y una interacción multiespecie, así como un metabolismo concreto con tasa de crecimiento controlada y una ubicación relativa concreta. En la formación del biofilm se distinguen varios pasos o fases (O'Toole et al., 2000; Boonaert et al., 2001). En primer lugar, macromoléculas y moléculas orgánicas de distintos orígenes se fijan al sustrato por adsorción, acondicionándolo para una eventual colonización (Paso 1, Figura 1). A continuación, los microorganismos se aproximan al sustrato mediante movimientos propios y brownianos, y entonces se producen interacciones físico-químicas con las moléculas adsorbidas que conducen a una adhesión primaria de las células (Paso 2, Figura 1). Esta adherencia al sustrato, como paso fundamental de la formación del biofilm, es un proceso complejo, regulado por el medio circundante, la superficie del sustrato y los microorganismos (Ophir & Gutnick, 1994; O'Toole et al., 2000; Donlan, 2002). Tras ella, el anclaje de los microorganismos se provoca mediante la excreción de productos poliméricos extracelulares (EPS) (Paso 3, Figura 1). A partir de este punto la multiplicación celular lleva a la proliferación de las comunidades microbianas y a su adhesión permanente al sustrato. La maduración de la biopelícula corresponde al crecimiento tridimensional e involucra la generación de una arquitectura específica y compleja del biofilm con canales y poros, así como una redistribución de los microorganismos lejos del sustrato (Paso 4, Figura 1). La “arquitectura del biofilm” (biofilm architecture, Lawrence et al., 1991) y en particular la disposición espacial de las microcolonias (grupos de microorganismos) en relación con otras, tiene profundas implicaciones para la función de estas complejas comunidades (Donlan, 2002). La arquitectura de la matriz del biofilm no es sólida y presenta canales que permiten el flujo de agua, nutrientes y oxígeno incluso hasta las zonas más profundas del biofilm. La existencia de estos canales no evita sin embargo, que dentro del biofilm se encuentren ambientes con diferentes Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 32 – concentraciones de nutrientes, pH y oxígeno (Pasos 5 y 6, Figura 1). Tras la maduración del biofilm, la última fase es de retorno de los microorganismos del biofilm a un modo de crecimiento planctónico, provocado por una reducción en la producción de exopolímeros (Paso 7, Figura 1. O'Toole et al., 2000). Por consiguiente, los microorganismos que abandonan la biopelícula cierran el ciclo de vida de la misma (Costerton, 2007). Figura 1. Representación esquemática de la formación de un biofilm. Adaptado de Cuzman (2009). (1) Establecimiento de las macromoléculas y moléculas orgánicas sobre la superficie a colonizar, (2) adhesión reversible de los colonizadores primarios, (3) transición a la adhesión irreversible, multiplicación y comienzo de la producción de EPS (sustancia polimérica extracelular), (4) comienzo del desarrollo tridimensional de la estructura del biofilm, (5) adhesión de colonizadores secundarios y desarrollo continuo del biofilm, (6) maduración del biofilm con una estructura expresa (arquitectura del biofilm) formando un micro-ecosistema específico, (7) fase/estado de homeostasis que mantiene en equilibrio al biofilm, con un crecimiento continuo de la estructura dotada de canales (Ch “channel”), y a los microorganismos que se desprenden y retornan a un modo de crecimiento planctónico. El reconocimiento de la importancia de los biofilms constituye un fenómeno reciente y aunque sólo hace un par de decenios que se viene investigando la fisiología de estas comunidades, se ha puesto de manifiesto la existencia de importantes mecanismos de comunicación microbiana que afectan a la estabilidad de la biopelícula permitiendo la organización/desorganización de la actividad y el asentamiento/desprendimiento de su soporte. Así, existe un mecanismo de comunicación bacteriana de efecto positivo (Quorum Sensing) y otro negativo (Quorum Quenching). Las proteínas responsables de estos mecanismos de comunicación bacteriana o Quorum Sensing (QS) y de bloqueo de comunicación bacteriana o Quorum Quenching (QQ) fueron descubiertas a comienzos de los noventa (Fuqua et al., 1994). Según los trabajos publicados por Otero et al. (2004) y Otero & Romero (2010), la detección del quórum o Quorum Introducción General – 33 – Sensing (QS) describe la capacidad de un microorganismo para percibir y responder a la densidad poblacional mediante la regulación de la expresión génica, siendo así capaz de desarrollar un comportamiento social coordinado mediante la producción de pequeñas moléculas señal. Si estas moléculas señal son eliminadas del medio, es decir, si cortamos la comunicación entre los organismos se produce el proceso contrario denominado interceptación del quórum o Quorum Quenching (QQ). II. Interacción biofilm-sustrato: colonización biológica de superficies expuestas al exterior La formación de un biofilm requiere la presencia de un sustrato y son pocos los sustratos que no se ven afectados por biofilms, ocupando los biofilms subaéreos la mayor parte de la superficie terrestre (Krumbein et al., 2003). En ciertos casos su presencia resulta beneficiosa, ya que estabilizan y fertilizan suelos (Acea et al., 2001, 2003), sedimentos (Liess & Francoeur, 2010) y rocas (Prieto et al., 2005, 2006). En otros casos su aparición, que suele producirse desde las primeras etapas de exposición de prácticamente cualquier superficie a la intemperie (Figura 2, Silva et al., 1997; Gaylarde & Morton, 1999; Warscheid & Braams, 2000), provoca no sólo un cambio estético derivado de la presencia de pigmentos biológicos (Urzì et al., 1992) sino también un problema evidente en términos de biodeterioro (“cualquier cambio indeseable en las propiedades de un material, originado por la actividad vital de los organismos”, Hueck, 1965) y envejecimiento acelerado (Warscheid & Braams, 2000). El desarrollo de los biofilms degrada activamente los materiales al producir condiciones ácidas/alcalinas, retener la humedad y absorber el calor de forma diferencial en función de la coloración que presenten (Krumbein, 1988; Garty, 1990; Warscheid et al., 1991, 1996), por lo que su desarrollo resulta especialmente dañino al producirse sobre superficies valiosas desde un punto de vista artístico e histórico (Figura 2, Dornieden et al., 2000). El sustrato representa para los microorganismos formadores del biofilm un espacio o superficie donde desarrollar su actividad biológica, así como la fuente de energía para su desarrollo (Gorbushina & Krumbein, 2000). Por esta razón, en el caso de superficies inertes es necesario un proceso de colonización secuencial en el que el sustrato en cuestión se va convirtiendo poco a poco en el ecosistema adecuado para los colonizadores implicados. Así, la secuencia de colonización no es caprichosa sino que responde a la propia estrategia de los organismos. Los colonizadores primarios del sustrato son por lo general los microorganismos autotróficos tales como diatomeas, algas, bacterias fotoautotróficas sensu lato y cianobacterias, que sólo necesitan agua y una mínima cantidad de sales minerales para asentarse y poder desarrollarse sobre el sustrato (Saiz-Jimenez & Ariño, 1995). Su crecimiento por lo general provoca el denominado verdín (greening) que raramente es uniforme formando manchas de tonalidad verdosa en áreas húmedas y con una presencia mayor en los lugares sombreados que en los Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 34 – expuestos, debido a que los últimos se secan más rápidamente (Barberousse et al., 2006; Smith et al., 2011). Es la duración del período de humedad, más que la frecuencia, lo que resulta crucial para predisponer un sustrato a su colonización por este tipo de organismos, que resulta favorecida además, cuando existe vegetación adyacente (Saiz-Jimenez & Ariño, 1995; Smith et al., 2011). En el caso concreto de las cianobacterias (bacterias oxigénicas fotoautotróficas), éstas poseen una serie de capacidades como: tolerancia a la desecación y al estrés hídrico, a altos niveles de sales, resistencia a altas temperaturas y uso eficiente de radiaciones de baja intensidad de la luz solar, que explican su presencia generalizada en ciertos sustratos soportando las condiciones ambientales más extremas (e.g. Friedmann, 1980). La presencia de los organismos pioneros da paso a la formación evolutiva de consorcios microbianos más complejos al permitir el asentamiento de otras especies (microorganismos heterótrofos) con el consecuente crecimiento del biofilm, en el que se incluyen ahora líquenes, bacterias heterótrofas, hongos, briófitos sensu lato e incluso plantas superiores (Ortega-Calvo et al., 1991a; Saiz-Jimenez, 1992; Tiano, 1993; Tomaselli et al., 2000; Crispim et al., 2003; Sterflinger, 2010). Estos nuevos organismos son más deteriorantes para el sustrato, ya que la mayor parte (bacterias heterótrofas, hongos y líquenes) metabolizan la materia orgánica producida por los fotoautótrofos y secretan ácidos orgánicos que degradan químicamente el sustrato (Warscheid & Braams, 2000). Además, provocan un daño físico causado por la penetración de las estructuras filamentosas de anclaje, particularmente las hifas de los hongos, dentro de las cavidades porosas de la superficie (Hirsch et al., 1995; Prieto et al., 2000; Silva & Prieto, 2004), que al humectarse se expanden ejerciendo presión en el interior del poro y abriendo fisuras que pueden derivar en la formación de ampollas y el desprendimiento del material (detachment/spalling). También, la superficie mucosa y por lo general cargada negativamente (Costerton et al., 1978) del biofilm favorece la adsorción de cationes y de moléculas orgánicas, y la adherencia de partículas transportadas por el viento (polvo, polen, partículas de carbón elemental, cenizas volantes, etc.) dando lugar a una pátina oscura más consistente y difícil de eliminar, que recibe el nombre de pátina negra (black crust) (Camuffo, 1995). Estas pátinas además del origen biogénico en ambientes contaminados, pueden presentar otros orígenes, como el biogénico en ambientes no contaminados (Aira, 2007; Gaylarde et al., 2007; Prieto et al., 2007; Silva et al., 2009) y el no-biogénico (Del Monte et al., 2001; Aira, 2007; Prieto et al., 2007; Silva et al., 2009). Debido a que provoca un cambio estético de los elementos sobre los que se asienta, el aspecto del biofilm debe ser tenido muy en cuenta. En lo que se refiere a su coloración, ésta varía según el tipo de organismos y partículas que lo conforman. El color de los organismos está determinado por el tipo y abundancia de pigmentos biológicos que poseen, que a su vez varían en función de las condiciones ambientales, tales como cambios en nutrientes, luz, temperatura, radiación ultravioleta y pH (e.g. Tandeu de Marsac, 1977; Collier & Grossman, 1992; Soltani et Introducción General – 35 – al., 2006). De modo general, los pigmentos biológicos presentan tonalidades: rojizas (carotenoides y ficobiliproteínas ficoeritrinas), amarillentas (carotenoides), verdosas (clorofilas y Figura 2. Diversidad de biofilms subaéreos (terrestres). (a) “La casa de esquí”, también llamada “La casa del futuro” o “Futuro”, diseñada por el arquitecto finlandés Matti Suuronen en 1965 en Múnich (Baviera, Alemania). Su exterior, compuesto por plástico y fibra de vidrio reforzada con poliéster, poliuretano-poliéster y polimetacrilato de metilo (plexiglás), presenta una visible colonización biológica. Se trata de un ejemplo de la presencia de biofilms en objetos de arte contemporáneo. (Imagen extraída de Cappitelli et al., 2006). (b) Detalle del exterior de “Futuro”, mostrando una zona de crecimiento del biofilm formado esencialmente por cianobacterias y arqueas (Cappitelli et al., 2006). (c) Detalle de la fachada este (donde se encuentra ubicada la Puerta Santa) de la Catedral de Santiago de Compostela (A Coruña, España), construida en su totalidad en roca granítica. La colonización biológica en esta zona es variada con presencia de briófitos, líquenes, algas y cianobacterias. (d) Detalle de la pared con un abundante verdín en la parte superior del claustro del monasterio de San Martín Pinario situado en Santiago de Compostela (A Coruña, España). (e) Detalle de pared y suelo con visible colonización biológica de una de las construcciones de la villa romana de Chedworth, una de las mayores villas romanas en Gran Bretaña, situada en Gloucestershire (Inglaterra). ficobiliproteínas aloficocianinas) y azuladas (ficobiliproteínas ficocianinas), así como pardasoscuras-negras debidas a las melaninas y scytoneminas (García-Pichel & Castenholz, 1991; Proteau et al., 1993). Los pigmentos biológicos no son comunes a todos los microorganismos, las ficobiliproteínas por ejemplo son exclusivas de cianobacterias y algas rojas (Glazer, 1977) y las scytoneminas sólo aparecen en cianobacterias (García-Pichel & Castenholz, 1991), por lo Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 36 – que en ciertas ocasiones el color del biofilm puede orientar acerca del tipo de microorganismo que lo forma. Sin embargo, es difícil determinar visualmente a qué tipo de pigmento y microorganismo es debida la coloración de la biopelícula; así por ejemplo, una biopelícula rosada cubriendo una de las superficies de piedra caliza de las ruínas mayas de Edzná (Campeche, México) debía su color a una concentración alta de carotenoides pertenecientes al alga Trentepohlia umbrina (Gaylarde et al., 2006), mientras que una coloración similar presente en los frescos realizados por Luca Signorelli en la Capilla de San Brizio (Catedral de Orvieto, Umbría, Italia) era debida a las ficobiliproteínas rojizas (ficoeritrinas) pertenecientes a organismos cianobacterianos (Cappitelli et al., 2009). En otros casos, sobre todo cuando el biofilm es de escaso grosor y su coloración no es la más habitual, como es el caso de los colores azulados y rosados, resulta difícil a simple vista determinar si el color es propio del sustrato o se debe a la colonización biológica (Figura 3, Gaylarde et al., 2006). Existen algunos trabajos en los que el color del biofilm (determinado cualitativamente) se relaciona con el tipo de pigmento y/o microorganismo; algunos ejemplos son las biopelículas de color naranja debidas a algas y bacterias y las películas de color grisáceo producidas principalmente por hongos que se encontraban en la roca calcárea de la torre barroca de Noto en Siracusa (Sicilia, Italia) (Urzì & Realini, 1998) y las biopelículas de color negro de las piedras del exterior del templo de Bayón (Angkor, Camboya) formadas por especies del género Dematiáceous (ascomicetos) que producían un alto contenido en melanina (Sterflinger, 2000). A este respecto, sin embargo, cabe señalar que el conocimiento de la magnitud de la diversidad microbiana que se encuentra en los biofilms subaéreos está lejos de ser completo, ya que mediante las técnicas tradicionales de cultivo se aíslan, y por tanto identifican, menos del 1% de la diversidad total (Ward et al., 1990). En los últimos años, se han desarrollado diversos métodos moleculares para permitir la identificación de microorganismos en muestras ambientales (Amann et al., 1995). Técnicas de biología molecular como la electroforesis en gel con gradiente desnaturalizante (EGGD), el análisis de polimorfismo conformacional de un solo filamento o de cadena sencilla (PCCS) y la hibridación fluorescente in situ (HISF) apuntan la posibilidad de que bacterias halófilas o halotolerantes, alcalófilas y también arqueas formen parte de los biofilms subaéreos que se encuentran en superficies del Patrimonio histórico artístico (Saiz-Jiménez & Laiz, 2000; McNamara et al., 2003; Ortega-Morales et al., 2004; Cappitelli et al., 2007; Piñar et al. 2009). Por lo general, los materiales sintéticos (Figura 2a, b) se consideran más resistentes a los daños químicos, físicos y biológicos producidos por la presencia del biofilm que los materiales naturales (Figura 2c, d, e) (Cappitelli et al., 2006). En cualquier caso, la disponibilidad de agua y nutrientes en el sustrato es el principal factor limitante de la colonización biológica y del consiguiente desarrollo del biofilm, y las condiciones ambientales junto a las características del sustrato determinan el tipo de biofilm así como su velocidad de crecimiento (Silva et al., 1997). En este sentido, el conjunto de las propiedades de un sustrato que determina su capacidad Introducción General – 37 – para ser colonizado por microorganismos se ha denominado biorreceptividad (Guillitte, 1995). Características como una alta porosidad y rugosidad superficial favorecen a priori la colonización biológica y formación del biofilm, al facilitar la adhesión de los microorganismos al sustrato, aumentando por tanto su biorreceptividad (Caneva et al., 1991; Silva et al., 1997; Warscheid & Braams, 2000; Prieto & Silva, 2005; Miller et al., 2006). Sin embargo, la influencia exacta de las propiedades del sustrato en el crecimiento microbiano y resultante biorreceptividad no está aún completamente clara (Miller et al., 2012). Figura 3. Columnas de mármol en el interior de una iglesia de la costa mediterránea. La columna de la derecha presenta una coloración rojo anaranjada que a simple vista parece propia del material y en realidad es debida a una biopelícula. Ciñéndonos al ámbito litológico, la presencia de biofilms también supone uno de los principales problemas que atañen a la conservación del parque de edificios, tanto a aquellos que pertenecen al Patrimonio histórico-artístico como a los de reciente construcción (Herrera & Videla, 2004; Crispim & Gaylarde, 2005; Gorbushina, 2007; Scheerer et al., 2009). La mayor parte de los monumentos elaborados por la humanidad están construidos con piedra natural o la incluyen como elemento auxiliar. Entre las rocas más utilizadas para este fin se encuentran las de mayor durabilidad, es decir, de alta resistencia a la meteorización como son los mármoles y las rocas graníticas. La aparición de biofilms sobre este tipo de construcciones suele ser epilítica (sobre la superficie) debido en parte a la baja porosidad de este tipo de rocas (Prieto et al., 1999, 2000; Silva et al., 1999). Rocas más porosas como calizas y areniscas presentan en muchos casos un crecimiento endolítico (en el interior de la roca) del biofilm Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 38 – (Saiz-Jimenez et al., 1990; Miller et al., 2010). Los biofilms endolíticos han sido clasificados en más detalle por Golubic et al. (1981) en función de su presencia en grietas (chasmoendolíticos), en poros (criptoendolíticos), o si muestran una verdadera capacidad de perforación de la matriz de piedra (euendolíticos). El clima de la zona también influye en este aspecto, Warscheid et al. (1996) consideran que mientras en los climas templados los microorganismos tienden a colonizar la superficie de las rocas, en los climas tropicales y subtropicales prefieren penetrar en el interior del sustrato con el fin de protegerse de la luz del sol y la desecación. Matthes-Sears et al. (1997), sin embargo, sugieren que los microorganismos se ven obligados a convertirse en endolíticos no por protección frente a las condiciones ambientales adversas sino en la búsqueda de mayores nutrientes y un mayor espacio. Según la mayoría de los autores, las poblaciones endolíticas son extremadamente difíciles de eliminar y altamente deteriorantes, al contribuir a la disminución de la cohesión entre los granos minerales a profundidades de hasta varios milímetros (e.g. Ascaso et al., 1998; Miller et al., 2010), aunque Hoppert et al. (2004) a este respecto apuntan que, dado que sus tasas de crecimiento en el interior de la roca son bajas, las poblaciones endolíticas rara vez están asociadas a procesos de biodeterioro. En el caso de biofilms epilíticos, el biodeterioro está provocado en primera instancia por la decoloración a la que dan lugar en la superficie, que ocasiona no sólo un problema estético sino también un problema físico ya que en las áreas oscuramente coloreadas decrece el albedo, absorbiendo más radiación solar, lo que repercute en un aumento de los ciclos frío/calor y humectación/secado, que termina causando un estrés físico en el biofilm y un estrés mecánico en el interior de la roca (Warke et al., 1996; Warscheid & Braams, 2000; Sand et al., 2002). Las temperaturas que se alcanzan en las superficies pétreas colonizadas por organismos oscuros fueron registradas por Garty en 1990, mostrando una diferencia en más de 8ºC con respecto a las áreas sin colonizar. Una variación muy similar de temperatura fue obtenida por Carter & Viles (2004) en bloques de roca calcárea con y sin cobertura liquénica, mostrando además los mayores gradientes de temperatura justo por debajo del biofilm liquénico. Diversos estudios han demostrado que las cianobacterias constituyen la biomasa mayoritaria que se desarrolla sobre las superficies de piedra natural (Ortega-Morales et al., 2000; Gaylarde et al., 2001). Existen dos teorías sobre el efecto concreto de biodeterioro que provocan este tipo de organismos, junto con los organismos fotoautotróficos en general, sobre las superficies pétreas. Mientras algunos autores consideran despreciable la acción agresiva de las cianobacterias y algas sobre el sustrato, otros reportan datos que apuntan a un mecanismo de deterioro directo en el que en primer lugar se produce un deterioro físico llevado a cabo por el biofilm, que emplea la masa de filamentos entrecruzados reunidos en la matriz mucosa de las vainas de cianobacterias y algas para extraer los granos minerales de la roca y las partículas de la superficie (Ortega-Calvo et al., 1991a, 1991b). En contra de lo que pueda parecer muchas cianobacterias, no necesariamente filamentosas, presentan esta dañina habilidad (Scheerer et Introducción General – 39 – al., 2009). A partir de ahí mediante contracciones y expansiones, la biopelícula hace que se desprendan los granos minerales y partículas, contribuyendo a la desagregación gradual del material (Ortega-Calvo et al., 1991a, 1991b). Además, la formación de biofilms de cianobacterias y algas produce una mayor retención de agua en la superficie del material, incrementando los procesos de hidrólisis y otros mecanismos de meteorización. Una conclusión común a todos los estudios relativos al biodeterioro de los materiales pétreos provocado por la aparición de biopelículas, es la necesidad de invertir esfuerzos en la detección precoz de la colonización biológica, es decir cuando el biofilm aún se encuentra en las primeras etapas de su formación, ya que en ese momento la intervención sería mínima y se enmarcaría en el campo de la conservación preventiva, que se define como “todas aquellas medidas y acciones que tienen como objetivo evitar o minimizar futuros deterioros o pérdidas” (ICOM-CC, 2008). Una intervención mayor para la eliminación de biofilms ya formados debe ser cuidadosamente evaluada, ya que en muchos casos supone un deterioro importante de la superficie tras la eliminación de la biopelícula (Figura 4). Además, en el caso de aplicar a continuación algún tratamiento, incluido la simple limpieza con agua, éste puede exacerbar una futura colonización biológica (Young, 1997). A finales de los años noventa Young (1997) apuntaba que tras la limpieza de las fachadas de edificios en Escocia (Reino Unido), la reaparición de partículas inorgánicas en superficie tardaba varios años en producirse, sin embargo, el crecimiento de las algas verdes (greening) ocurría en pocos meses. Dos ejemplos de la aplicación de tratamientos de restauración que han llevado a un aumento de la colonización microbiana, son el empleo de un consolidante a base de extractos de plantas locales utilizado en las ruinas mayas de Joya de Cerén (El Salvador), el cual favoreció el crecimiento de hongos y actinomicetos (Caneva & Nugari, 2005), y los consolidantes sintéticos de polilaurilmetacrilato y poliisobutilmetacrilato empleados en la catedral de Milán (Italia) en los años setenta que favorecieron también el crecimiento de hongos (Cappitelli et al., 2007). Las acciones contra la colonización biológica y el desarrollo de biofilms se pueden dividir en cuatro grandes categorías (Scheerer et al., 2009): (1) control indirecto mediante la modificación de las condiciones ambientales, generando condiciones que resulten adversas para los microorganismos, (2) eliminación mecánica, (3) empleo de productos químicos (biocidas), y, (4) métodos físicos de erradicación, como el uso de un campo electromagnético o el uso de ultrasonidos. La primera de las acciones resulta la más recomendable, sin embargo en pocas ocasiones se puede aplicar debido a la gran adaptabilidad de los microorganismos, en especial bacterias y cianobacterias, a los cambios externos, no pudiendo aplicarse nunca por razones obvias en casos de superficies expuestas a la intemperie. Según las dimensiones de la superficie a tratar, en la eliminación mecánica se puede emplear el bisturí o chorros de aire, agua o diversos abrasivos a presión. El principal inconveniente de este tipo de técnicas es que pueden dañar el sustrato por microabrasión debido al contacto directo entre los agentes Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 46 – b* definen las componentes de color rojo-verde y amarillo-azul, respectivamente. Así, un valor negativo de a* define un color más verde que rojo, mientras que un valor positivo de b* define un color más amarillo que azul. Además de este grupo de coordenadas escalares (L*a*b*) el espacio CIELAB puede ser representado mediante otro sistema de coordenadas cilíndricas o polares que sustituye a* y b* por, croma 22 *** baC ab += y, tono angular *)/*(tangente-arco abhab =. En la percepción humana del color se distinguen tres características principales que corresponden a los tres parámetros básicos del color: claridad, croma y tono. La consideración de las coordenadas cilíndricas (L*C*abhab) en lugar de las escalares (L*a*b*) permite cuantificar el color del mismo modo en que es percibido, es decir teniendo en cuenta estos tres atributos de apreciación visual: claridad (su análogo en CIELAB es L*), croma (su análogo en CIELAB es C*ab) y tono (su análogo en CIELAB es hab). Una de las ventajas del espacio de color CIELAB frente a otros es que permite cuantificar la diferencia de color entre dos muestras mediante la combinación de las diferencias de sus tres coordenadas, tanto escalares como cilíndricas, en un espacio euclidiano. Este es el fundamento de la fórmula clásica de 1976 de diferencia de color CIELAB (Wyszecki & Stiles, 1982), que asume la uniformidad del espacio de color: ()()() 222 **** baLE ab Δ+Δ+Δ=Δ ó ()( )( ) 222 **** ababab HCLE Δ+Δ+Δ=Δ Esta fórmula de distancia euclídea para medir la diferencia de color ha sido progresivamente revisada y mejorada debido al reconocimiento de la falta de uniformidad de este espacio. La mayoría de las fórmulas de diferencia de color utilizadas en la actualidad parten de las coordenadas del espacio CIELAB, introduciendo factores de ponderación (weighting functions), indicados por S (en otros textos, W), que consiguen una más estrecha relación entre las diferencias de color calculadas y percibidas respecto a CIELAB; y los números k, factores paramétricos (parametric factors) que representan variables dependientes de las condiciones de observación. Estos factores de ponderación (S) y factores paramétricos (k) se introducen sobre las diferencias CIELAB de claridad, croma y tono para corregir la falta de uniformidad perceptual del espacio, y así surgen las fórmulas de diferencia de color CIE94 (CIE Publication 116-1995, 1995): 2 2 2 94 ** * )::(*        Δ +        Δ +        Δ =Δ HH ab CC ab LL HCL Sk H Sk C Sk L kkkE Introducción General – 47 – y CIEDE2000 (CIE Publication 142-2001, 2001), que incluye además un factor denominado término de rotación (RT) y una transformación específica de la coordenada a*, que afecta principalmente a los colores con bajo croma (colores neutros): 2/1 2 2 2 00 ''·''' )::(                  ΔΔ +        Δ +        Δ +        Δ =Δ HHCC T HHCCLL HCL SkSk HC R Sk H Sk C Sk L kkkE Em ambas ecuaciones los factores paramétricos: los valores kL, kC, kH sirven para ajustar las contribuciones relativas de claridad, croma y tono, y así, adoptando los valores kL=1, kC=1, kH=1, las fórmulas que se obtienen son CIE94(1:1:1) y CIEDE2000(1:1:1); mientras que aumentando la contribución relativa a la diferencia de claridad, kL=2, kC=1, kH=1, las fórmulas son CIE94(2:1:1) y CIEDE2000(2:1:1). En cuanto a la importancia relativa de estas correcciones (CIE94 y CIEDE2000), los autores destacan que la mejora de CIE94 sobre CIELAB es notablemente superior a la mejora de CIEDE2000 sobre CIE94 (Melgosa et al., 2004). El espacio de color CIELAB intenta linealizar las diferencias de color perceptibles por el ojo humano, usando la métrica de diferencias de color descritas por las elipses de MacAdam (MacAdam, 1942). Las relaciones no lineales para L*, a* y b* pretenden emular la respuesta no lineal del ojo humano. Además, los cambios uniformes de los componentes en el espacio de color L*a*b* tienen por objeto corresponder a cambios uniformes en el color percibido, por lo que las diferencias relativas de percepción entre dos colores en el espacio L*a*b* se pueden aproximar tratando cada color como un punto en un espacio tridimensional, con tres componentes: L*, a*, b* y tomando la distancia euclídea entre ellos. Son varias las causas que podrían justificar la aceptación generalizada de CIELAB en los estudios científicos, así como en las aplicaciones industriales. Entre esas causas cabría citar el carácter oponente de sus coordenadas L*a*b*, la asociación de su otra terna de coordenadas L*C*abhab con los tres atributos clásicos del color (claridad, croma y tono) o incluso su analogía respecto al espacio de color Munsell (Munsell, 1905), con el que mucha gente está bien familiarizada. La instrumentación para la medida del color disponible en la actualidad muestra fundamentos similares, diferenciándose normalmente en el grado de sofisticación y en la configuración que presenta el aparato, la cual depende de la aplicación concreta a la que se destina (Capilla et al., 2002). Las medidas instrumentales de color están sujetas a las condiciones de medida del espectrofotómetro o del colorímetro como son la configuración óptica y geometría de medida o de iluminación-observación (0/45, 45/0, d/0, …), el iluminante (A, C, D65, F11, …) y el observador (observadores estándar CIE 1931 - 2º y CIE 1964 - 10º), pudiendo estos dos últimos, iluminante y observador, prefijarse para generalizar el proceso de medida obteniendo Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 48 – así medidas fiables y reproducibles que resultan comparables entre sí. La configuración óptica y geometría de medida de un colorímetro o un espectrofotómetro es propia del aparato y viene determinada por dos valores, X/Y, donde X es el valor que indica la dirección del haz que ilumina la muestra, e Y, la dirección de medición del haz reflejado. Según cómo se ilumine la muestra (X) y cómo se coloque el detector de luz reflejada con respecto a ésta (Y), los resultados pueden mostrar alguna discrepancia, por lo que, cuando se vaya a emplear más de un instrumento en la medida es recomendable comparar los resultados de las mediciones para poner de manifiesto esas diferencias instrumentales (e.g. Campos Acosta et al., 2004). Entre las geometrías de medida más comunes están 0/45 y 45/0, que corresponden respectivamente a la situación en la que la muestra y el blanco de referencia son iluminados normalmente, es decir con 0º de inclinación, mientras que el detector está colocado de forma que registra la radiación reflejada a 45º y viceversa. También existen aparatos cuya configuración óptica o geometría es de esfera integradora, en estos casos la geometría se indica con una “d” que indica o que la muestra se ilumina difusamente, esto es con luz proveniente de todos lados (d/Y), o que en el detector se mide la componente total difusa (X/d). Así, si la iluminación es difusa y el detector se encuentra colocado normalmente a la muestra, la geometría es d/0, si se encuentra desplazado 8º, d/8, si se encuentra desplazado sólo 6º, d/6, etc. Si la iluminación está colocada en la normal a la muestra y al detector llega la componente total difusa, la geometría es 0/d, si la iluminación se encuentra desplazada 8º, 8/d, si se encuentra desplazada sólo 6º, 6/d, etc. Figura 5. Espacio de color CIELAB. (1) Representación clásica del espacio de color CIELAB en una esfera, donde aparecen reflejados los dos grupos de coordenadas, escalares (L*a*b*) y polares (L*C*abhab); (2) Representación real del espacio de color CIELAB, que no es ni una esfera, ni un cilindro, ni un cubo, sino un cuerpo sólido de límites irregulares. (Adaptado de: http://www.brucelindbloom.com/index.html?ColorCheckerRGB.html); (3) Planos de color a*– b* con un valor de claridad (L*) constante. Introducción General – 49 – Cuando se mide con las geometrías 0/45 y 45/0, se excluye automáticamente la componente especular de la reflectancia, es decir “aquélla que viene de una dirección concreta y se refleja con un angulo igual de salida”. Por esta razón en materiales lisos y brillantes éstas son unas geometrías recomendables. Cuando la geometría es de esfera integradora con luz difusa (X/d ó d/Y), cuyo uso es recomendable para los materiales heterogéneos en textura, es preciso ajustar en el aparato la inclusión (SCI) o exclusión (SCE) de la reflectancia especular, lo cual se consigue mediante lo que se llama una trampa óptica (tapón blanco). En el modo SCE, con componente especular excluida, la reflectancia especular está excluida de la medición y sólo se mide la reflectancia difusa. Cuando se utiliza el modo SCI ambas reflectancias están incluidas. Precisamente dada la necesidad de excluir la componente especular de las mediciones con la esfera integradora, muchos instrumentos emplean una geometría 8/d ó 6/d en lugar de 0/d. Figura 6. Izquierda: Instrumentos de medida del color por contacto. Arriba: Colorímetro portátil triestímulo Minolta con procesador de datos DP-310 y cabezal de lectura CR-300. Medio: Espectrofotómetro portátil GretagMacbeth CE-XTH. Abajo: Espectrofotómetro portátil Konica Minolta CM-700d. Derecha: Realización de la medida instrumental de color on site. Teniendo en cuenta tales premisas antes de comenzar la medición, las medidas obtenidas por un espectrofotómetro o por un colorímetro triestímulo (Figura 6) serían completamente equivalentes aunque ambos aparatos no son instrumentalmente iguales. La diferencia fundamental entre un espectrofotómetro y un colorímetro consiste en que el colorímetro trabaja Cuantificación del color en el estudio de la formación de biofilms en rocas graníticas del Patrimonio histórico artístico – 50 – únicamente en el espectro de luz visible y selecciona una longitud de onda determinada mediante filtros fijos, mientras que un espectrofotómetro es capaz de trabajar en otras regiones del espectro electromagnético (ultravioleta e infrarroja) y posee un monocromador para seleccionar la longitud de onda deseada. Además, el colorímetro funciona como una cámara digital RGB, cuyos canales de color se ajustan espectralmente a las funciones de igualación CIE combinando las sensibilidades espectrales del sensor, la lente y los filtros coloreados RGB. Por esta razón los colorímetros dan valores triestímulo relativos mientras que los espectrofotómetros proporcionan mediciones espectrales. Con todo lo expuesto hasta el momento, parece posible en principio medir el color en cualquier situación que se presente. Sin embargo, las aplicaciones prácticas de la medida del color son tan extensas y variadas que en algunas situaciones es necesario definir una serie de conceptos particulares asociados al objeto concreto de estudio. En ciertas áreas como la alimentaria, dental y textil, la medida instrumental del color posee una amplia tradición, por lo que el desarrollo e implementación de los métodos de medida se encuentran muy avanzados, existiendo incluso en ciertos casos, como el vino y el aceite, métodos oficiales para la determinación de su color. Sin embargo en otras áreas, como la que recoge el objetivo principal de esta tesis, así como áreas afines relacionadas con sustratos inorgánicos naturales, tales como rocas, suelos y sedimentos, la medida instrumental del color apenas ha sido aplicada, lo que lleva a la necesidad de determinar un procedimiento de medida previo a la aplicación de la técnica (Artigas et al., 2002). – 51 – General Introduction I. From a planktonic lifestyle to formation and growth of biofilms Historically, microorganisms were considered as free-living unicellular life forms and characterized on the basis of their growth in different culture media (Donlan, 2002). However, advances in microscopic and molecular techniques have made possible the direct observation of a wide variety of natural habitats, and it has been established that most microorganisms live attached to surfaces within a structured biofilm and not as free-floating organisms (Costerton et al., 1987). The discovery of microbial biofilms, in the late seventeenth century, can be attributed to the Dutch microscopist Antonie van Leeuwenhoek (1632-1723), who, after observing the “animalcules” in plaque from his own teeth, affirmed that bacteria could adhere and grow universally on exposed surfaces. In the nineteenth century, the Russian microbiologist and soil scientist Sergei N. Winogradsky (1856-1953) developed a system of simulating natural ecosystems (the Winogradsky column), and thus managed to isolate associations of microorganisms (biofilms) on glass slides. In 1940, Heukelekian & Heller observed the “bottle effect” in marine microorganisms, whereby bacterial growth and activity were substantially enhanced by the incorporation of a surface to which these organisms were able to attach. In 1943, Zobell observed that the number of bacteria on surfaces was much higher than in the surrounding medium (in this case, seawater). On the basis of observations of dental plaque and sessile communities in mountain streams, Costerton et al. (1978) proposed a theory of biofilms that explained the mechanisms whereby microorganisms adhere to living and nonliving materials, as well as the benefits afforded by this ecologic niche. The term biofilm was coined by Costerton in 1978, in an article in Scientific American, in which he stated that, “in nature (but not in laboratory cultures) bacteria are covered by a "glycocalyx" of fibres that adhere to surfaces and to other cells”. This supposed a paradigm shift in the way microbiologists viewed the growth and ecology of microorganisms. The current definition of a biofilm describes a more complex concept: "a universal community of microorganisms (bacteria, fungi, cyanobacteria, algae, protozoa), complex and interdependent, linked to wet surfaces (biotic or abiotic) or interfaces (subject to water column gradients) by a polymeric matrix (EPS) of extracellular glycoproteins with viscous texture, generated ex profeso Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 52 – and travelled by water channels, which allow an efficient water, nutrients and gas exchange between constituent populations and the outside environment "(Costerton, 2007). The capacity for biofilm formation is not restricted to any specific group of microorganisms; under suitable environmental conditions all microorganisms can form biofilms. In this regard, it has been shown that many cells must attach to solid surfaces (forming part of a biofilm) to survive and proliferate, a requirement for survival known as "anchorage dependence" (Pierres et al., 2002). The main feature that distinguishes sessile biofilm cells from their motile or free-floating planktonic counterparts is the existence of a matrix of extracellular polymeric substance (EPS) that embeds all components of the biofilm and enables it to adhere to surfaces. The biofilm matrix is a complex milieu, comprising mainly water (up to 97%) and a mixture of EPS such as polysaccharides, proteins, nucleic acids, lipids/phospholipids, absorbed nutrients and metabolites. Although the extracellular matrix is the defining characteristic of the biofilm, its role is not yet fully understood. Most authors believe that the primary function of the extracellular matrix is to protect the biofilm from antimicrobial action and from adverse changes in humidity, temperature, osmotic pressure and pH. The extracellular matrix acts via chemical neutralization or the creation of a diffusion barrier that prevents antimicrobial agents reaching the microorganisms (Lindsay & von Holy, 2006). It is also known that cyanobacteria increase their production of extracellular polymers under lightand nutrient-limiting conditions (Albertano et al., 2003). However, other authors consider that the most important role of the matrix occurs during the initial stages of biofilm formation when it facilitates cell attachment to the substrate (Decho, 2000; Barranguet et al., 2005); this gives cohesion to the nascent biofilm and facilitates interaction between its various components (Flemming & Wingender, 2001). It has also been established that sessile cells are physiologically distinct from planktonic microorganisms and that biofilms exhibit a distinct phenotype from their free-floating counterparts. The main phenotypic differences are related to location, grouping and colonial organization, mobility, metabolic and environmental adaptation, and form and growth rate (Donlan, 2002). During the complex process of adhesion, microorganisms alter their phenotypic characteristics in response to their proximity to a surface. During the initial stages of biofilm formation, sessile microorganisms form microcolonies within which they are in stable juxtaposition with microorganisms of the same and other species. Interactions within these microcolonies enable exchange of genetic material, which results in an overall change in the phenotype of the biofilm (Stolz, 2000). This also creates physico-chemically heterogeneous environments in which microorganisms associated with biofilm have different physiological characteristics than the planktonic cells. Many microorganisms alter their physiological processes to adapt to these particular niches (e.g. by growing anaerobically) (Stoodley et al., 2002). In addition, the different micro-organisms in the biofilm respond to the conditions of their specific micro-environments by presenting different growth patterns. Physiological cooperation General Introduction – 53 – is the main factor that helps form the structure and establish equilibria, so that mature biofilms (i.e. attached to surfaces) are highly efficient microbial communities (Costerton et al., 1994). During biofilm formation, the constituent microorganisms usually leave their status of motile cells (planktonic) to form unstratified populations of microorganisms within which several species of microorganisms interact; each species has its own metabolic rate, with a controlled growth rate and specific, relative location within the biofilm (Figure 1). Several stages of biofilm formation can be distinguished (O'Toole et al., 2000; Boonaert et al., 2001). In the first stage, macromolecules and organic molecules from different sources are adsorbed onto the substrate and thus prepare it for possible colonization (Step 1, Figure 1). The microorganisms then approach the substrate by self-propulsion and Brownian motion; physicochemical interactions with the adsorbed molecules lead to initial adhesion of cells (Step 2, Figure 1). Adherence to the substrate, which is a key step in biofilm formation, is a complex process regulated by the surrounding medium, the surface of the substrate and the microorganisms themselves (Ophir & Gutnick, 1994; O'Toole et al., 2000; Donlan, 2002). The subsequent attachment of the microorganisms is caused by excretion of EPS (Step 3, Figure 1). From this moment, cell division leads to the proliferation of microbial communities and their continued adherence to the substrate. Maturation of the biofilm involves three-dimensional growth via the generation of a specific and complex 3D structure of channels and pores, often referred to as biofilm architecture (Lawrence et al., 1991) and redistribution of microorganisms away from the substrate (Step 4, Figure 1). The architecture of the biofilm, in particular the spatial arrangement of microcolonies (clusters of microorganisms) in relation to others, has important implications for the function of these complex communities of microorganisms attached to surfaces or associated with interfaces (Donlan, 2002). The architecture of the biofilm matrix is strongly substrate dependent; it is not solid, but has channels that allow the flow of water, nutrients and oxygen to even the deepest parts of the biofilm. However, the existence of these channels does not preclude the existence of environments in which the pH and concentrations of nutrients and oxygen will differ within the biofilm (Steps 5 and 6, Figure 1). Once the biofilm is mature, some of the microorganisms will return to a planktonic mode of growth, as a result of a decrease in exopolymer production (Step 7, Figure 1. O'Toole et al., 2000). The microorganisms leave the biofilm, thus completing the life cycle (Costerton, 2007). Recognition of the importance of the biofilm phenomenon is recent, and therefore a concerted effort to study microbial biofilms began only two decades ago. However, study of these communities has already revealed the existence of important mechanisms of microbial communication that affect the stability of the biofilm and enable organization/disorganization and settlement/release. Thus, some bacterial communication mechanisms have positive effects (Quorum Sensing), and others have negative effects (Quorum Quenching). The proteins responsible for bacterial communication (Quorum Sensing) and blockage of communication (Quorum Quenching) were discovered in the early 1990s (Fuqua et al., 1994). According to Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 54 – studies published by Otero et al. (2004) and Otero & Romero (2010), Quorum sensing (QS) refers to the ability of microorganisms to regulate their gene expression according to population density and thus develop a type of social behaviour coordinated by producing small signal molecules. If these signal molecules are removed from the environment, the communication between organisms will break down; this is known as Quorum Quenching (QQ). Figure 1. Schematic model of the phases involved in biofilm formation on a solid surface. Adapted from Cuzman (2009). (1) Establishment of macromolecules and organic molecules on the surface being colonized; (2) Reversible adhesion of primary colonizers; (3) Transition to irreversible adhesion, followed by multiplication and start of production of extracellular polymeric substance (EPS); (4) Start of three-dimensional development of the biofilm structure; (5) Adhesion of secondary colonizers and continuous development of the biofilm; (6) Mature biofilm with an expressed structure (biofilm architecture) and forming a specific micro-ecosystem; (7) Biofilm homeostasis during which continuous growth of the structure with channels (Ch "channel") takes place and some microorganisms break off and return to a planktonic mode of growth. II. Biofilm-substrate interaction: biological colonization of surfaces exposed to outdoor conditions Biofilm formation requires the presence of a substrate, and as very few substrates are not affected by biofilms, all of the earth’s surface is covered by subaerial biofilms (Krumbein et al., 2003). Moreover, biofilms usually appear from the early stages of exposure of almost all outdoor surfaces under all types of environmental conditions (Figure 2, Silva et al., 1997; Gaylarde & Morton, 1999; Warscheid & Braams, 2000). In some cases, the presence of biofilms is beneficial as it stabilizes and fertilizes soils (Acea et al., 2001, 2003), sediments (Liess & Francoeur, 2010) and rocks (Prieto et al., 2005, 2006). However, the presence of biofilms General Introduction – 55 – sometimes leads to aesthetic deterioration of the surface, which is associated with staining by biogenic pigments (Urzì et al., 1992), and is an important problem in terms of biodeterioration (“the undesirable change in the properties of a material caused by the activities of organisms”; Hueck, 1965) and accelerated aging (Warscheid & Braams, 2000). The presence of biofilms actively degrades materials because the films produce acidic/alkaline conditions, retain moisture and absorb heat differently depending on the coloration (Krumbein, 1988; Garty, 1990; Warscheid et al., 1991, 1996). From the viewpoint of cultural heritage, the development of biofilms on valuable surfaces is particularly detrimental (Figure 2, Dornieden et al., 2000). For biofilm-forming microorganisms, substrates provide a surface for development of biological activity and also supply the energy required for development (Gorbushina & Krumbein, 2000). Therefore, for inert surfaces, a sequential process of colonization is required in which the substrate in question gradually becomes suitable for the colonizers. Thus, the sequence of colonization is not random, but responds to the microorganisms’ own strategies. The primary colonizers of the substrate are generally autotrophic microorganisms such as diatoms, algae, photoautotrophic bacteria sensu lato and cyanobacteria (e.g. Grant, 1982), which only need water and a minimal supply of mineral salts to settle and grow on the substrate (Saiz-Jimenez & Ariño, 1995). Growth of these microorganisms often leads to the formation of greenish patches in damp areas, in a phenomenon often referred to as greening; this is more apparent in shaded areas than in exposed sites because the latter dry out more readily (Barberousse et al., 2006; Smith et al., 2011). The duration of the dampness (rather than the frequency of wetting) is crucial in predisposing a substrate to colonization by these types of microorganisms, and colonization is more rapid in sites with adjacent or overhanging vegetation (Saiz-Jimenez & Ariño, 1995; Smith et al., 2011). Cyanobacteria (oxygenic photoautotrophic bacteria) have several characteristics that enable them to withstand extreme environmental conditions, such as tolerance to desiccation and water stress, tolerance to high levels of salts, high temperature resistance and efficient use of lowintensity sunlight radiation; these features explain the widespread presence of cyanobacteria on certain substrates (e.g. Friedmann, 1980). The presence of photoautotrophic microorganisms, cyanobacteria and algae (primary colonizers) leads to the formation of more complex microbial associations, which in turn enables settlement of other species (heterotrophic microorganisms) and the concomitant growth of a biofilm, which now includes lichens, heterotrophic bacteria, fungi, bryophytes sensu lato and sometimes even higher plants (Ortega-Calvo et al., 1991a; SaizJimenez, 1992; Tiano, 1993; Tomaselli et al., 2000; Crispim et al., 2003; Sterflinger, 2010). Heterotrophic microorganisms (bacteria, fungi and lichens) may have more severe effects on substrates because most of them metabolize organic matter produced by photoautotrophs, and they secrete organic acids that chemically degrade the substrate (Warscheid & Braams, 2000). Heterotrophic microorganisms also cause physical damage because their filaments (particularly the hyphae of fungi) penetrate the porous cavities on the substrate surface (Hirsch et al., 1995; Prieto et al., 2000; Silva & Prieto, 2004). When wet, the filaments expand and exert pressure Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 62 – III. Detection and quantification of biological colonization and biofilm development. Instrumental color measurements and the CIELAB color system It can be assumed that all stone surfaces will be colonized at some stage by living organisms (leading to biofilm formation) and that the settlement involves not only aesthetic damage, but also physical and chemical damage, which increase as the colonization spreads. Therefore, the early detection of biofilm formation would constitute significant progress as regards preventive conservation of stone buildings, monuments and sculptures. Identification of autotrophic microorganisms is a key step in the early detection of the biological colonization processes as these organisms are often pioneers in the colonization of stone surfaces (e.g. Grant, 1982). For this reason, research in this field has mainly focused on the detection and quantification of these types of microorganisms. Martín (1990) reviewed the traditional methods used to study biological damage in stoneworks of cultural heritage and classified the methods used to quantify the presence of microorganisms as follows: morphological, microbiological, histochemical and biochemical, and chemical and physicochemical methods. A slightly different classification is considered here, and three types of methods are distinguished: structural and morphological, microbiological and extractive methods. Structural and morphological methods are the most commonly used to identify and classify pioneer species. These methods involve macroscopic observations of stone material in the field (Urzì et al., 1992; Krumbein, 1993) and/or microscopic observations in the laboratory. Use of the stereomicroscope enables recognition of specific physiological structures of microorganisms (Prieto et al., 1995; Shirakawa et al., 2003); use of the optical microscope (OM) enables recognition of specific cellular morphology and cellular structure of each group of microorganisms with a very simple sample preparation (Ariño, 1996). Fluorescent probes (vital dyes and fluorescent proteins) can also be used with the fluorescence microscope (Tayler & May, 1991) and bioluminescence techniques (Ranalli et al., 2000) to enhance the fluorescent signal. Each of these techniques provides a particular type of information, which is often complementary to that offered by the other techniques. Direct observations can also be made by conventional scanning electron microscopy (SEM) (Ortega-Calvo et al., 1991a; Prieto et al., 1997), SEM in backscattered electron mode (SEM-BSE), low temperature SEM (LTSEM), confocal laser scanning microscopy (CLSM) and transmission electron microscopy (TEM) (De los Ríos et al., 2008), all of which allow three-dimensional observations of the microorganism at 10,000x magnification. Microbiological methods for biofilm quantification are based on counting colonies (number of colony forming units per gram of sample, CFU/g) after a long incubation period in specific media and suitable conditions of temperature and light (ICR-CNR, 1988). If the biofilm is incubated in General Introduction – 63 – liquid medium, the statistical approach originally suggested by McCrady (Hurley & Roscoe, 1983) is used to determine the most probable number of organisms per gram of sample (MPN/g). The studies carried out by Ortega-Calvo et al. (1993) and Hyvärinen et al. (2002) are examples of the application of these methods in samples taken from stone monuments. In extractive methods, the biomass is estimated from the amount of various cell components present in microorganisms (Bartolini & Monte, 2000), such as certain proteins (Schwenzfeier et al., 2011), ATP (Tiano et al., 1989), DNA (Saiz-Jimenez & Laiz, 2000; Tomaselli et al., 2000), lipids and polysaccharides (Di Pippo et al., 2009). In the case of phototrophic microorganisms, the content of photosynthetic pigments in general, and the content of chlorophyll-a in particular (Bell & Sommerfeld, 1987; Nagarkar & Williams, 1997; Prieto et al., 2004; Schumann et al., 2005; Eggert et al., 2006), is a good indicator of the presence and amount of microorganisms, and is probably the most prolific and one of the most reliable methods. Physiological methods, in which the biomass is estimated from the initial increase in respiratory rate (Bartolini & Monte, 2000), are often used in combination with extractive methods. Furthermore, microbial activity can be estimated by measuring enzymatic activities, such as dehydrogenase activity and total enzyme activity, via the hydrolysis of fluorescein diacetate (FDA). Tayler & May (2000) measured dehydrogenase activity on sandstone from ancient monuments, and Prieto et al. (2004) measured FDA in laboratory experiments carried out with granitic rock. Although the above techniques have certain advantages and their use is widespread in research, they also have some disadvantages. The main disadvantages are the associated costs and the length of time they take. Moreover, most of the techniques are destructive and require prior sample preparation and acquisition, which is always difficult or even impossible task in the case of cultural heritage monuments. Techniques based on optical methods can be used to overcome these problems. Because of the small size of the organisms in question, the methods must use a high magnification and resolving power. In this respect two of the physical characteristics of phototrophic microorganisms can be made use of: the fluorescence emitted by their chlorophyll, and the color that they form when they accumulate on solid surfaces. Each of these characteristics has enabled development of a technique for detecting and monitoring the growth of biofilm on surfaces: (i) measurement of the fluorescence emitted by chlorophyll in vivo and, (ii) instrumental color measurement with a spectrophotometer or a tristimulus colorimeter. Large surfaces such as building façades, can even be monitored as both techniques are nondestructive (non-invasive) and they are applied on site, so that sampling is not required. Several studies have investigated the use of chlorophyll-a fluorescence in vivo for direct quantification and monitoring of phototrophic colonization of lithotypes (Guillite & Dreesen, 1995; Brechet et al., 1996, 1997; McStay et al., 2001; Miller et al., 2006, 2010). However, fluorescence measurements have some disadvantages relative to color measurement: the fluorescence signal occurs within a short time, depending on the physiological conditions of the Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 64 – organisms (not taking into account the dead cells), and moreover, it is very difficult to analyze spectra with unknown characteristics. In other words, the technique requires basic knowledge of the spectral features of the parameters investigated so that the bands are chosen correctly to enable contributions from factors other than the microorganisms, such as the fluorescence of certain minerals within in the same spectral region (Lognoli et al., 2003). Calcium carbonate, fluorite, scheelite and uranium ores are some of the mineral components that may emit fluorescence, which usually consists of a broad emission band in the blue region of the spectrum. Carbonate calcium, for example, has an emission maximum at about 420 nm (Cecchi et al., 2000). In relation to the quantification of biological colonization on building façades by instrumental color measurements, Newby et al. (1991) were the first authors to relate measurements of reflectance (which can be obtained with a spectrophotometer or tristimulus colorimeter) to the color change caused by deposition of particles (from natural sources as well as pollutants) on a solid surface. On the basis of a previous study by Ball (1989), Newby et al. (1991) defined soiling of building materials as “an optical effect, a darkening of the surface that can be measured as a change in light reflectance, and is generally related to the deposition of airborne particulate matter onto the building surface”. Several years later and taking this definition into account, Gorbushina & Krumbein (2000) stated that many of the spectral changes are related to the growth of massively pigmented biofilms. Krumbein (1995) considered that airborne particles (dust, pollen, particulate elemental carbon (PEC), fly ash, etc.) and microorganisms often act concurrently to fouling (soiling) of the building, and noted that as early as 1853, Ehrenberg had came to the conclusion that microorganisms can trap airborne particles, in their slimy extracellular products, more efficiently than the rock surface itself. The use of portable tristimulus colorimeters and portable spectrophotometers to measure the color generated by phototrophic microorganisms developed epilithically on stony substrates came into common use in the late 1990s. As part of a larger research programme funded by Historic Scotland (http://www.historic-scotland.gov.uk/), researchers at Robert Gordon University (Aberdeen, Scotland. United Kingdom) conducted a series of experiments on sandstone buildings in Scotland. These researchers used a portable tristimulus colorimeter to monitor the color change brought about on building façades by biological growth (algae) after application of various treatments biocides (Urquhart et al., 1995; Young et al., 1995; Young, 1997). Around the same time, Urzì & Realini (1998) published a study in which the color of orange and grey patinas developed on the calcarenite stone from which the Baroque town of Noto is built (Syracuse, Sicily, Italy) was determined with a portable tristimulus colorimeter and correlated with the associated microflora (respectively algae, and bacteria and fungi). As part of the British National Materials Exposure Programme (NMEP), Viles and colleagues monitored (between 1987 and 1995) microbial colonization, soiling and decay of limestone tablets located over 20 sites around the UK for periods of 1, 2, 4 and 8 years (depending on the climate and General Introduction – 65 – pollution conditions of the location), using a portable spectrophotometer and characterization techniques such as scanning electron microscopy (SEM) (Viles et al., 2002). The study subsequently focused on limestone buildings located along roads, in Oxford (England, UK), with different levels of traffic, and the buildings were monitored for a period of three years (Viles & Gorbushina, 2003). In addition, after access to vehicles (particularly private vehicles) was restricted in the centre of the historic city of Oxford, in 1999, as part of the Oxford Transport Strategy (OTS), Thornbush & Viles (2004, 2006) recorded the changing patterns changes of microbial growth and soiling on building stone in Oxford. Furthermore, the reliability of color measurements for estimating the biomass of phototrophic organisms in biofilms on stone surfaces was demonstrated by Prieto et al. (2004) in a comparative study of the most common methods of quantifying phototrophic biomass, such as as the in vitro measurement of chlorophyll-a and fluorescein diacetate (FDA) hydrolysis. Moreover, Prieto et al. (2002) carried out laboratory experiments that demonstrated a direct relationship between the amount of phototrophic organisms deposited on a surface and the color generated. The same authors subsequently conducted a field study in which they used a portable spectrophotometer to monitor growth of a biofilm that was induced on quartz surfaces to mask the brilliant white color and thus mitigate the visual impact (Prieto et al., 2005). The latter studies concluded that one of the major advantages of instrumental color measurement for the quantification of phototrophic biomass on surfaces that are largely homogeneous in color is that it enables early detection of biological colonization, even before it is visible to the human eye (Prieto et al., 2002, 2005). This is an important advantage for the management and maintenance of surfaces recently exposed to outdoor conditions. It is also advantageous in treated surfaces in which the biofilm has been removed, as careful examination has shown that, despite cleaning efforts, some biological structures remain between the grains of the stone, which facilitates the re-establishment of microorganisms (Prieto et al., 1995). The CIELAB color space (CIE Publication 15-2, CIE, 1986) is used in most studies involving analysis of the instrumental color measurement, to represent and analyze the color values obtained. CIELAB or CIE 1976 L*a*b* attempts to produce a color space, that is perceptually more linear or uniform than other color spaces, such as RGB or CMYK. The term perceptually linear or uniform is used to indicate that similar changes in the amount of a color value should produce change of similar visual importance. This CIELAB color space was specifically developed for this purpose by the International Commission on Illumination (usually abbreviated to CIE for its French name, Commission Internationale d'Eclairage). The CIELAB or CIE 1976 L*a*b* is based on the CIE 1931 color space XYZ (CIE, 1932) and is strongly influenced by the Munsell color space (Munsell, 1905). Color depends on the interaction between three factors: the light source, the illuminated object and the observer who captures the image, and therefore the CIE established the CIELAB space on the basis of four key points: standard light source, Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 66 – exact conditions of observation, appropriate mathematical units and standard observer curves (Wyszecki & Stiles, 1982). The CIELAB color space (Figure 5), which can be visualized inside a sphere, a cylinder or a cube (although this does not conform to reality), uses three values to determine each color (L*a*b* or L*C*abhab) and is based on the opponent-colors theory of color vision, which states that one color cannot be both green and red at the same time, nor blue and yellow at the same time. As a result, single values can be used to describe the red/green and the yellow/blue attributes. Thus, one color can be expressed with Cartesian coordinates L*a*b*, where L* defines lightness ranging from 100 (absolute white) to 0 (absolute black), a* denotes the red/green value and b* the yellow/blue value. A movement of color measurement in a positive a* direction depicts a shift towards red. Along the b* axis, a positive b* movement represents a shift towards yellow. The CIELAB space can also be represented by another coordinate system, by using polar coordinates (L*C*abhab) that replace a* and b* with chroma 22 *** baC ab += and, hue angle *)/*( tangentarc abhab =. The human perception of color appears to correspond to three attributes: hue, chroma and value (lightness) (Wyszecki & Stiles, 1982). The use of polar coordinates (L*C*abhab) rather than Cartesian coordinates (L*a*b*) leads to quantification of a color’s similarity to it is perceived color, i.e. by taking into account the three attributes of visual assessment: value or lightness (L*), chroma (C*ab) and hue (hab). One advantage of the CIELAB color space is that allows the color difference between two samples to be quantified by combining the differences in their three coordinates, both Cartesian and polar, in a Euclidean space. This is the basis of the classical color difference formula of 1976 CIELAB (Wyszecki & Stiles, 1982), which assumes a uniform color space: ()()() 222 **** baLE ab Δ+Δ+Δ=Δ or ()( )( ) 222 **** ababab HCLE Δ+Δ+Δ=Δ The CIELAB color difference has been progressively revised and improved since the lack of uniformity of the CIELAB color space was recognized. Currently, most color difference formulae are based on the classical CIELAB color formula. They are therefore considered as CIELABbased formulae and have a common structure. The lightness, chroma, and hue differences are always computed from CIELAB, and each is weighted by two different types of variables: the weighting functions (designated by the letter S, also W in other texts), which are intended to improve the perceptual uniformity of CIELAB, and the parametric factors (designated by the General Introduction – 67 – letter k), which are meant to account for the influence of specific experimental conditions on perceived color differences. This leads to the CIELAB-based color-difference formulae, CIE94 (CIE Publication 116-1995, 1995): 2 2 2 94 ** * )::(*        Δ +        Δ +        Δ =Δ HH ab CC ab LL HCL Sk H Sk C Sk L kkkE and CIEDE2000 (CIE Publication 142-2001, 2001), which also incorporates a rotation term (RT) that accounts for the interaction between chroma and hue differences in the blue region and a modification of the a* coordinate of CIELAB, which mainly affects colors with low chroma (neutral colors): 2/1 2 2 2 00 ''''' )::(                  ΔΔ +        Δ +        Δ +        Δ =Δ HHCC T HHCCLL HCL SkSk HC R Sk H Sk C Sk L kkkE The parametric factors can be set to kL=1, kC=1 and kH=1 for CIE94(1:1:1) and CIEDE2000(1:1:1), and to kL=2, kC=1, kH=1 (to increase the relative contribution of the lightness term) for CIE94(2:1:1) and CIEDE2000(2:1:1). As for the relative importance of CIELAB-based formulae, Melgosa et al., (2004) found that the improvement of CIE94 over CIELAB was considerably greater than that of CIEDE2000 over CIE94. The CIELAB color space attempts to linearize the perceptibility of color differences, by use of the color difference metric described by the MacAdam ellipses (MacAdam, 1942). The nonlinear relations for L*, a*, and b* are intended to mimic the nonlinear response of the human eye. Furthermore, uniform changes in components of the L*a*b* color space should correspond to uniform changes in perceived color, so that the relative perceptual differences between any two colors in L*a*b* can be approximated by treating each color as a point in a three dimensional space (with three components: L*, a*, b*) and measuring the Euclidean distance between them. There are several reasons for the widespread acceptance of CIELAB in scientific and industrial applications, including the opposing character of the coordinates L*a*b*, the association of another set of coordinates, L*C*abhab, with the three color attributes (lightness, chroma and hue), and even its analogy relative to the Munsell color space (Munsell, 1905), with which many people are familiar. The instrumentation for color measurement available nowadays has a similar basis, but usually differs in the degree of sophistication and configuration of the device, which depends on the particular application for which it is intended (Capilla et al., 2002). Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 68 – Figure 5. CIELAB color space. (1) Classical representation inside a sphere of CIELAB color space in which the two groups of coordinates are represented, Cartesian (L*a*b*) and polar (L*C*abhab), (2) Real representation of the CIELAB color space, which is neither a sphere, cylindar nor a cube, but a solid body of irregular boundaries (adapted from http://www.brucelindbloom.com/index.html?ColorCheckerRGB.html), (3) a* - b* color planes with a constant value of lightness (L*). Instrumental color measurements are subject to the conditions of measurement of the spectrophotometer or colorimeter, such as the optical configuration, measuring geometry or illumination/observation geometry (0/45, 45/0, d/0, …), the illuminant (A, C, D65, F11, …), and the observer (standard observers CIE 1931 - 2º and CIE 1964 - 10º). Illuminant and observer can be preset in order to generalize the measurement process and thus obtain reliable, reproducible and comparable measures. The illumination/observation geometry of a colorimeter or spectrophotometer is characteristic of the measuring device and it is determined by two values, X/Y, where X is the value that indicates the direction in which the specimen surface is illuminated, and Y is the value that indicates the direction of the reflected beam, where the detector is located. The results may differ depending on how the sample is illuminated (X) and on how the detector is positioned relative to the sample (Y). Therefore, when more than one color measuring instrument is used, it is advisable to compare the results of the measurements to highlight these instrumental differences (e.g. Campos Acosta et al., 2004). The most common measuring geometries are 0/45 and 45/0. With 0/45 geometry, the specimen surface is illuminated from the normal line direction (0 degrees) and the light is received at an angle of 45 degrees from the normal line. With 45/0 geometry, the specimen surface is illuminated from an angle of 45 degrees to the normal line and the light is received in the normal direction (0 General Introduction – 69 – degrees). Some devices also use an integrating sphere for illuminating or viewing a sample uniformly from all directions. An integrating sphere is a spherical device in which the internal surfaces are coated with a white material such as barium sulphate, so that the light is uniformly diffused. In such cases, the geometry is indicated by a "d", which indicates a diffuse illumination integrating sphere system. An instrument with X/d optical geometry illuminates the sample in the X degrees direction and collects the light reflected in all directions. An instrument with d/Y optical geometry illuminates the sample diffusely and detects the light in the Y degrees direction. Thus, if the sample is diffusely illuminated and the detector is placed at the normal direction relative to the sample, the optical geometry is d/0, if the detector is displaced 8 degrees, the optical geometry is d/8; if it is displaced only 6 degrees, d/6, etc. On the contrary, if the sample is illuminated at the normal angle (0 degrees) and the reflected light collected in all directions the optical geometry is 0/d, if the lighting beam is displaced 8 degrees, the optical geometry is 8/d, if it is displaced only 6 degrees, 6/d, etc. When the measuring geometries are 0/45 and 45/0, the specular component of reflectance, i.e. the reflected light from the surface such that the angle of reflection equals the angle of incidence, is automatically excluded. These geometries are therefore recommended in smooth and shiny materials. For integrating sphere-based geometries (X/d or d/Y), the use of which is recommended for heterogeneous materials, the specular component of reflectance can be included or excluded mechanically by opening and closing an optical trap provided inside the integrating sphere. If the specular reflectance is included in the color measurement, by completing the sphere with a specular plug, it is referred to as Specular Component Included (SCI). In Specular Component Excluded (SCE) mode, the specular reflectance is excluded from the measurement and only the diffuse reflectance is measured. Because of the need to exclude the specular component, many devices with integrating sphere based geometries use 8/d or 6/d geometry instead of 0/d geometry. Taking into account the previous premises, the measurements obtained with a spectrophotometer and with a tristimulus colorimeter (Figure 6) should be equivalent, although both devices are not instrumentally equal. The fundamental difference between a spectrophotometer and a colorimeter is that a colorimeter only works in the visible light spectrum and at a wavelength selected by fixed filters, while a spectrophotometer works in other regions of the electromagnetic spectrum (ultraviolet and infrared) and has a monochromator to select the required wavelength. In addition, the colorimeter works as a digital camera RGB, whose color channels spectrally fit the CIE matching functions by combining the spectral sensitivities of the sensor, lens and RGB colored filters. For this reason, colorimeters provide tristimulus relative values, while spectrophotometers provide spectral measurements. Color quantification in the study of biofilm formation on granite stone in historical and artistic heritage – 70 – Figure 6. Left: Contact color-measuring instruments. Top: Minolta portable tristimulus colorimeter with CR-300 measuring head. Middle: GretagMacbeth (now XRite) portable spectrophotometer CE-XTH. Bottom: Konica Minolta's CM-700d spectrophotometer. Right: On site instrumental color measurement. From all of the above, it appears possible a priori to measure the color in any situation. However, the practical applications of color measurement are so extensive and varied that in some situations it is necessary to define a particular set of concepts associated with the particular object of study. Instrumental color measurement has a long tradition in certain fields of study such as those related to food, dentistry and textiles, so that the development and implementation of measurement methods are highly advanced; some official methods for determination of color have even been developed, e.g. for wine and oil. However, in other areas, such as that including the overall scientific aim of this doctoral thesis and those related to natural inorganic substrates such as rocks, soils and sediments, instrumental color measurement has scarcely been applied, so that a measurement procedure must be established prior to application of the technique (Artigas et al., 2002). Investigación / Research Once the minimum number of measurements required to characterize the color of granite rocks was established, the minimum area of measurement was determined. For this, five specimens of each type of granite and surface finish were prepared with different areas (36 cm 2 ,54cm 2 , 72 cm 2 , and 90 cm 2 ), on which the corresponding number of measurements were made for each area and measurement head, according to the results obtained in the previous paragraph and by using the four different measurement heads (5, 8, 10, and 50 mm). To determine whether the color varied with the measurement area, the data were subjected to MANOVA for each type of rock and measurement head (Table III), with the mean values of L*,a*, and b* for each specimen as dependent variables and the measurement area (36, 54, 72, and 90 cm 2 ) as independent variables. As shown in Tables III and IV, there were statistically significant differences between the color obtained with the 50-mm diameter measurement head for an area of 36 cm 2 and the other areas. According to these results, to characterize the color of granite, a measurement area of 36 cm 2 and 14 measurements with random replacement is sufficient only if measurement heads of diameter 10 mm are used. To determine the minimum area of measurement required with a 50 mm diameter measurement head, the statistical analysis was repeated with only the values of L*,a*, and b* obtained for areas greater than 36 cm 2 , i.e., with those corresponding to areas of 54, 72, and 90 cm 2 and with those obtained for areas greater than 54 cm 2 , i.e., with areas of 72 and 90 cm 2 . Comparison of the values obtained for areas larger than 36 cm 2 revealed statistically significant differences for Grissal, Blanco Cristal, and Labrador Claro between the values obtained in an area of 54 cm 2 and those obtained in the other two areas (72 and 90 cm 2 ). However, there were no statistically significant differences between the values obtained for areas of 72 and 90 cm 2 (Table V), and therefore it can be concluded that to characterize the color of granite when working with a 50 mm diameter measurement head, a minimum surface area of 72 cm 2 is required. Once the number and area of measurement were established in relation to the dimensions of the measurement head, the color determined by each head was compared to establish whether the results obtained are comparable. Considering the previous results, only the data obtained for an area of 72 cm 2 were used, and the number of measurements required was 12, 28, 28, and 34 for measurement heads of 50 mm, 10 mm, 8 mm, and 5 mm, respectively. The results of a MANOVA, with L*, a*, and b* as dependent variables and the diameter of the measurement head as the independent variable revealed statistically significant differences between the values obtained with the different measurement heads (Wilks’ lambda: 0.949; F: 44 739; df: 9; significance: 0.000) with the 50 mm measurement head producing the greatest differences (Table VI). TABLE I. Three-way multivariate analysis of variance (MANOVA) for the number of measurements. Source Wilk’s lambda F-value df Significance Type of rock (granite) 0.625 13.663 9 0.000 Surface finish 0.678 11.093 9 0.000 Diameter of the measurement head 0.123 87.715 9 0.000 Type of rock 3 surface finish 0.594 5.645 24 0.000 Type of rock 3 diameter of the measurement head 0.637 4.282 27 0.000 Surface finish 3 diameter of the measurement head 0.729 2.933 27 0.000 Type of rock 3 surface finish 3 diameter of the measurement head 0.487 2.681 72 0.000 Dependent: L*, a*, and b*. TABLE II. Minimum number of measurements in relation to the measurement head, surface finish, and type of granite. Type of rock (granite) Surface finish Diameter of the measurement head (mm) 50 10 8 5 Grissal Flamed 4 11 12 14 Sawn 1 10 11 15 Honed 2 14 14 17 Polished 1 6 12 17 Rosa Porrin˜o Flamed 3 13 14 15 Sawn 2 9 13 16 Honed 6 13 9 10 Polished 3 13 13 17 Blanco crystal Flamed 1 14 14 17 Sawn 5 11 12 15 Honed 1 12 13 17 Polished 4 13 14 17 Labrador claro Sawn 4 8 12 14 Honed 4 11 13 17 Polished 5 13 11 15 Volume 35, Number 5, October 2010 371 With the aim of analyzing the real significance, rather than the statistical significance, of the differences in color obtained with the 5, 8, 10, and 50 mm diameter measurement heads, the total color differences (DE) for pairs of measurement heads (Fig. 4) were calculated with the classical CIELAB formula (DE ab) and the CIELAB-based color-difference formulae (DE 94 and DE 00 ) and they were expressed in CIELAB units 13,14 and the more recent CIE94 and CIEDE2000 units, 10 respectively. The differences in color were calculated with the Granite Training group, to obtain generalized results. The results obtained (Fig. 4) show that the differences in color expressed in CIEDE2000 and CIE94 units, which clearly improve on the CIELAB units 15–17 and are more appropriate for evaluating samples with low chromaticity values, 18 were the same as or lower in magnitude than the CIELAB color differences. It was also found that there was no equivalence of scale factor among the values produced with the three formulae considered. The visual color difference threshold or just noticeable difference (jnd), which constitutes the lower limit of perception in an individual with normal color vision 19–21 and that it can be established as 0.73 CIELAB units, 22,23 is also shown in the Fig. 4. Below this value, the differences in color are inappreciable and therefore the colors determined with the 8 and 10 mm measurement heads would be considered identical by an observer. However, if the suprathreshold color-difference, which is approximately 1.75 CIELAB units, 15 is taken into account, all of the values of the total color differences (DE) for pairs of measurement heads are below this value and are only outshined in the case of the color differences for the 8 and the 50 mm measurement heads, which were measured with the same device (the Minolta colorimeter). These two measurement heads provided the greatest differences in the total color differences (DE) from the three formulae used: DE ab,DE 94, and DE 00 . On the other hand, if we take into account the color tolerances or units of color differences applied in the industrial field, where larger color differences are used, all of the values of DEobtained for the different measurement head were lower than 3 CIELAB units, the value considered as the upper limit of rigorous color tolerance. 14,24 With the aim of standardizing the results obtained with the granites included in the Granite Training group, a test TABLE III. Three-way multivariate analysis of variance (MANOVA) for the measurement area in relation to the size of the measurement head. Type of rock (granite) Diameter of the measurement head (mm) Wilk’s lambda F-value df Significance Grissal 5 0.996 0.032 9 1.000 8 1.000 0.003 9 1.000 10 1.000 0.004 9 1.000 50 0.004 172.529 9 0.000 Rosa Porrin˜o 5 0.984 0.133 9 0.999 8 0.998 0.013 9 1.000 10 0.998 0.013 9 1.000 50 0.009 118.037 9 0.000 Blanco Cristal 5 0.984 0.130 9 0.999 8 1.000 0.003 9 1.000 10 0.999 0.006 9 1.000 50 0.427 8.378 9 0.000 Labrador Claro 5 0.999 0.006 9 1.000 8 0.998 0.011 9 1.000 10 0.999 0.007 9 1.000 50 0.473 5.265 9 0.000 Dependent: L*, a*, and b*. 372 COLOR research and application was performed to determine whether these were applicable to other granites. For this, the color of one of the granites most commonly used at present in the building industry for cladding and for constructing monuments—a brownish gold granite denominated Silvestre—was determined. The values obtained for this granite (for an area of 72 cm 2 and number of measurements depending on the measurement head used: 34/5 mm head, 28/8 and 10 mm head, 12/50 mm head) were subjected to MANOVA, with L*, a*, and b* as dependent variables and the diameter of the measurement head (5, 8, 10, and 50 mm) as the fixed factor. The results were similar to those obtained with the Granite Training group; there were statistically significant differences in relation to the measurement head (Wilks’ lambda: 0.865, F: 33.843; df: 9; significance: 0.000) for each of the parameters considered (Table VII). The total color differences (DE) were calculated for the different measurement heads (Fig. 5) and a rigorous color tolerance (i.e., \3 CIELAB units) was obtained. Again, the greatest difference was between the color measurements made with the measurement heads of 50 mm and 8 mm diameter. The total color difference DE ab comparing small heads (5–10 and 5–8 mm) decreased by 0.5 CIELAB units in relation to the results obtained with the Granite Training group, whereas any comparison (DE ab,DE 94, and DE 00 ) between the 50 mm and the 5 mm measurement heads caused an increase in the difference of 0.5 units and of 1 unit when the comparison was between the 50 mm and the 8 or 10 mm measurement heads. It must be stressed that the chromatic appearance of heterogeneous surfaces, and in particular their chromatic discrimination, should be based not only on conventional colorimetric measures but on visual correlations with real observers. In this regard, although there are some recent contributions combining both the methods, 25–28 all of them are related to textured materials having well delimited contours which makes dubious their application to other kind of materials. In the case of granite, it has been TABLE IV. Tukey-b test for the 50-mm-diameter measurement head. Type of rock (granite) Area (cm 2 )L*a*b* Grissal 36 57.02 a 3.88 a 28.49 a 54 64.07 b 20.75 b 0.86 b 72 64.08 b 20.54 c 0.79 b 90 63.48 b 20.54 c 0.78 b Rosa Porrin˜o 36 57.45 a 5.91 a 23.67 a 54 63.89 b 2.20 b 5.63 b 72 63.91 b 2.33 b 5.68 b 90 63.93 b 2.34 b 5.70 b Blanco Cristal 36 68.05 a 2.75 a 23.42 a 54 72.38 b 20.28 b 3.13 b 72 71.96 b 20.11 b 2.72 b 90 71.34 b 20.10 b 2.72 b Labrador Claro 36 48.28 a 20.49 a 21.32 a 54 48.05 a,b 20.40 a,b 21.34 a 72 47.75 b 20.28 b 21.42 a 90 47.74 b 20.28 b 21.42 a Superscript letters a–c in each column for each granite indicate significant differences (a: 0.05). TABLE V. Three-way multivariate analysis of variance (MANOVA) comparing the different areas measured. Type of rock (granite) Areas (cm 2 ) Wilk’s lambda F-value df Significance Grissal 54–72–90 0.511 7.313 6 0.000 72–90 0.981 0.227 3 0.877 Rosa Porrin˜ o 54–72–90 0.958 0.396 6 0.880 Blanco Cristal 54–72–90 0.581 5.712 6 0.000 72–90 0.969 0.388 3 0.762 Labrador Claro 54–72–90 0.615 3.662 6 0.003 72–90 0.999 0.002 3 0.999 Dependent: L*, a*, b*. TABLE VI. Tukey-b test for the four measurement heads. Diameter of the measurement head (mm) L*a*b* 50 62.87 a 0.39 a 2.17 a 10 64.45 b 0.08 b 2.84 b 8 64.32 b 0.06 b 3.43 c 5 62.76 a 0.18 b 3.17 c Superscript letters a–c in each column indicate significant differences (a: 0.05). FIG. 4. Total color differences for granite training group, in CIELAB units (DE ab), CIE94 units (DE 94), and CIEDE2000 units (DE 00 ), obtained with different pairs of measurement heads. Volume 35, Number 5, October 2010 373 one attempt to combine colorimetric measures and visual assessment to analyze changes in the appearance of the surface of granite rocks when they are colonized by microorganisms 29 and it has been established that the minimum increment in color that observers can appreciate is DE ab ¼3.17, which is bigger than the tolerance above commented. However, owing to that value has been determined in a very specific experiment it cannot be considered as a definitive value and future research in this field would be focussed on performing visual experiments using psychophysical techniques with the aim of defining the granitic rocks threshold. Therefore, in a sense and awaiting progress in granite visual assessment, the reader should be careful with the interpretation of differences in total color data as two granite samples could have identical average CIELAB color and could be judged as different images by an observer. CONCLUSIONS This work provides an adaptable and affordable methodology of study supported by statistical analysis, which has been successful in examination of the factors affecting the color measurements on granite rocks and demonstration that color may be affected by both material and instruments properties. In this way, the methodology used enabled the determination of the minimum number of measurements and the minimum measuring area required to characterize the representative color of the whole nonhomogeneous surface of granite rocks. However, it should be noted that although the application of the present color-difference formulae to heterogeneous surfaces has not yet been validated, they have been employed in this study as a tool for comparison assuming a correlation between spatial averaging of CIELAB and visual judgment of granite. In relation to the minimum number of measurements required to characterize the color of granite rocks, it was found that it is affected by the type of rock, surface finish, and independently of the rock, by the diameter of the measurement head of the device. By use of a wide variety of granite rocks that differ in terms of color, texture, and surface finish, it was determined that the minimum number of measurements required to characterize the color of granitic rocks is higher the smaller measurement head. Thus, the number of measurements required is 6/36 cm 2 surface area for a 50 mm diameter measurement head, 14/36 cm 2 for 8 and 10 mm diameter measurement heads, and 17/36 cm 2 for a 5 mm measurement head. As regards the minimum measuring area, the results showed that an area of 36 cm 2 is sufficient when measurement heads of diameter 10 mm are used. A surface area equal to or larger than 72 cm 2 is required for measurement heads of larger diameters. The validity of the above results was tested with other ornamental granite (Granite Test group) allowing to conclude that if and when the number of measurements considered corresponds to those established in this study for the measurement head diameter and the measurement area, any results thus obtained will be comparable. As granite rocks can be considered as a complicated case of surface heterogeneity due to the complexity of their mineral distribution, it could be hypothesized that the methodology here proposed can be extrapolated to other types of rocks. 1. Capilla P, Artigas JM, Pujol J. Fundamentos de Colorimetrı ´a. Valencia: Universitat de Vale `ncia; 2002. 2. Calvo B, Menduina J, Parra JL. Ensayo metodolo ´gico para el estudio de las propiedades de granitos meteorizados. Boletı ´n Geolo ´gico y Minero 1991;102:295–307. 3. Benavente D, Martı ´nez-Verdu ´F, Bernabeu A, Viqueira V, Fort R, Garcı ´a del Cura MA, Illueca C, Ordo ´n ˜ez S. Influence of surface roughness on color changes in building stones. Color Res Appl 2003;28:343–351. 4. I ´n ˜igo AC, Vicente-Tavera S, Rives V, Vicente MA. Color changes in the surface of granitic materials by consolidated and/or water repellent treatments. Color Res Appl 1997;22:133–141. 5. Rivas T, Silva B, Prieto B. Medida de la eficacia de dos hidrofugantes aplicados a rocas granı ´ticas. Materiales de Construccio ´n 1998;249:5–21. 6. Sanmartı ´n P, De los Santos DM, Rivas T, Prieto B, Mosquera MJ, Silva B. Different behaviour of TEOS-based consolidant applied to granite and biocalcareous sandstone. Proceedings of the International Symposium on Stone Consolidation in Cultural Heritage Research and Practice, Lisbon, 2008. 7. Grossi CM, Esbert RM, Dı ´az-Pache F, Alonso FJ. Soiling of buildings stones in urban environments. Build Environ 2003;38:147– 159. 8. Grossi C, Brimblecombe P, Esbert RM, Alonso FJ. Color changes in architectural limestones from pollution and cleaning. Color Res Appl 2007;32:320–331. TABLE VII. Tukey-b test for the four measurement heads and Silvestre granite. Diameter of the measurement head (mm) L*a*b* 50 70.02 a 20.04 a 3.48 a 10 72.56 b 20.11 b 4.37 b 8 72.56 b 20.24 c 4.68 5 71.41 c 0.02 d 4.41 b Superscript letters a–d in each column indicate significant differences (a: 0.05). FIG. 5. Total color differences for granite test group (Silvestre), in CIELAB units (DE ab), CIE94 units (DE 94 ), and CIEDE2000 units (DE 00 ), obtained with different pairs of measurement heads. 374 COLOR research and application 9. Prieto B, Rivas T, Silva B. Rapid quantification of phototrophic microorganisms and their physiological state through their color. Biofouling 2002;18:229–236. 10. CIE Publ. 15:2004. Colorimetry, 3rd edition. Vienna: CIE Central Bureau; 2004. 11. Montag ED, Berns RS. Lightness dependencies and the effect of texture on suprathreshold lightness tolerances. Color Res Appl 2000;25: 241–249. 12. Chorro E, Perales E, de Fez D, Luque MJ, Martı ´nez-Verdu ´FM. Application of the S-CIELAB color model to processed and calibrated images with a colorimetric dithering method. Opt Express 2007;15:7810–7817. 13. Wyszecki G, Stiles WS. Color Science. Concepts and Methods, Quantitative Data and Formulae. New York: Wiley; 1982. 14. Berns RS. Billmeyer and Saltzman’s Principles of Color Technology, 3rd edition. New York: Wiley; 2000. 15. Melgosa M, Hita E, Poza AJ, Alman DH, Berns RS. Suprathreshold color-difference ellipsoids for surface colors. Col Res Appl 1997;22:148–155. 16. Melgosa M, Huertas A, Berns RS. Relative significance of the terms in the CIEDE2000 and CIE94 color-difference formulas. J Opt Soc Am A 2004;21:2269–2275. 17. Kim D-H, Cho EK, Kim JP. Evaluation of CIELAB-based colourdifference formulae using a new dataset. Color Res Appl 2001;26: 369–375. 18. Johnson G, Fairchild MD. A top down description of S-CIELAB and CIEDE2000. Color Res Appl 2003;28:425–435. 19. Brown WRJ, MacAdam DL. Visual sensitivities to combined chromaticity and luminance differences. J Opt Soc Am 1949;39:808– 834. 20. Brown WRJ. Color discrimination of twelve observers. J Opt Soc Am 1957;47:137–143. 21. Wyszecki G, Fielder G. New color matching ellipses. J Opt Soc Am 1971;61:1135–1152. 22. MacAdam DL. Visual sensitivities to color differences in daylight. J Opt Soc Am 1942;32:247–274. 23. Melgosa M, Hita E, Pe ´rez MM, El Moraghi A. Sensitivity differences in chroma, hue, and lightness from several classical threshold datasets. Color Res Appl 1995;20:220–225. 24. Vo ¨lz HG. Industrial Color Testing. Weinheim: Wiley–VCH; 2001. 25. Day EA, Berns RS, Taplin LA, Imai FH. A psychophysical experiment evaluating the color and spatial image quality of several multispectral image capture techniques. J Imaging Sci Technol 2004;48: 93–104. 26. Kuriki I. Testing the possibility of average-color perception from multicolored patterns. Opt Rev 2004;11:249–257. 27. Xin JH, Shen HL, Lam CC. Investigation of texture effect on visual colour difference evaluation. Color Res Appl 2005;30:341–347. 28. Arin ˜o I, Johansson S, Kleist U, Liljenstro ¨m-Leander E, Rigdahl M. The effect of texture on the pass/fail colour tolerances of injectionmolded plastics. Color Res Appl 2007;32:47–54. 29. Prieto B, Silva B, Aira N, A ´lvarez L. Toward a definition of a bioreceptivity index for granitic rocks: Perception of the change in appearance of the rock. Int Biodeter Biodeg 2006;58:150–154. Volume 35, Number 5, October 2010 375 Chapter 2*. Effect of surface finish on roughness, color, and gloss of ornamental granites Sanmartín, P.; Silva, B.; Prieto, B. Journal of Materials in Civil Engineering 23 (8): 1239-1248 (2011) JCR index (IF) 2010 = 0.677 (19/53, 36 percentile in Construction & Building Technology; 54/115, 47 percentile in Engineering Civil) Total number of times cited: 3 * Selected as Research Highlight by the journal during the months of August, September, October, November and December 2011. Effect of Surface Finish on Roughness, Color, and Gloss of Ornamental Granites P. Sanmartín1; B. Silva2; and B. Prieto3 Abstract: The effects of four of the most common types of surface finish on the appearance of five varieties of ornamental granite, all widely used in building construction and selected for their different colors, were analyzed by means of roughness, color, and gloss measurements. The results demonstrated that different surface finishes produce differences in color, especially in the lightness parameter (L), and that the magnitude of these differences depends on the color of the ornamental granite and is greatest in dark colored rocks. However, the variation in the color parameters with the different surface finishes did not depend on roughness, and no general conclusions could be drawn regarding the influence of the roughness on the color of ornamental granite. Gloss values were affected by the color of the ornamental granite, but in a different way for smooth and rough surfaces. Variation in gloss also depended on the mineral composition of the rock. Gloss and roughness were inversely related, but only within the range of low roughness values. In addition, the color gamut of the studied ornamental granites was defined within the CIELAB color space. These results will contribute to providing a standardized, objective method of characterizing the color of granite that will be useful for different workers in the field of building construction. DOI: 10.1061/(ASCE)MT.1943-5533.0000285. © 2011 American Society of Civil Engineers. CE Database subject headings: Surface roughness; Stones; Color; Coating. Author keywords: Color assessment; Gloss; Ornamental granite; Roughness; Surface finishes. Introduction Many of the most impressive monuments and buildings in the world are built from granite because this is one of the oldest, most durable, and most respected of building materials. The popularity of granite in the building, construction, monument, and tombstone industries is attributable not only to its resistance to weathering but also to its appearance. Selection of ornamental granites for new constructions is primarily based on their aesthetic properties because these define the architectural harmony of the construction with the surroundings and enhance the visual perception of the building. Several types of surface finish, such as polished, flamed, and sawn, have been developed to produce different appearances and thus increase the decorative potential of the granite. Surfaces finishes differ in surface roughness and affect slab aesthetics because they induce variations in the perceived texture, color, and gloss. Therefore, when appearance is the main selection criterion, aesthetic properties such as texture, color, and gloss must be taken into account. Texture is defined as the visual characteristic and tactile quality of the surface of a material, although the term is also used in geology to refer to the degree of crystallinity, grain size, and fabric (geometrical relationships) of the constituents of a rock. Two terms were therefore used in the present study: surface texture and petrographic texture. Obviously, the petrographic texture affects the surface texture. However, in the case of ornamental rocks, other surface properties such as the surface finish of the rock, which is not related to the petrographic texture because it is induced by humans, must be considered because it is very important from an aesthetic point of view. Surface roughness is a measure of the surface texture, and the parameter most frequently used to describe roughness is the average roughness (Ra or ΔRa), defined as the integral of the absolute value of the roughness profile over an evaluation length and measured with a profilometer (ISO 1984). Although color is the aesthetic property most often used in selecting building material, objective colorimetric characterization of different granitic rocks is not usually carried out in the granite industry, and operators must trust in their skill to differentiate colors. This leads to misunderstandings between architects and builders. This problem is reflected in the colorimetric terms used to classify granite. Terms such as “champagne cream”and “golden yellow”may suggest different colors to different people. One clear example, granites commercialized as Rosa Porriño and Rosabel are classified as pink granites in some catalogs and as red granites and gray granites, respectively, in others. Likewise, despite the great acceptance of contact colormeasurement devices for objective measurement of the color of granite (Grossi et al. 2007a;Iñigo et al. 2004,1997) and other more homogeneous rocks (Durán-Suárez et al. 1995;García-Talegón et al. 1998;Grossi et al. 2007b), a standardized protocol enabling comparison of the results obtained by different authors and instruments has only recently been published (Prieto et al. 2010). In that work, which took into account the high heterogeneity in color and texture of this plutonic rock, a protocol for characterization of color and a methodology for the analysis of instrumental and surface factors that may affect the determination of color were proposed by using statistical tools and colorimetric criteria. 1Ph.D. Student, Departamento Edafología y Química Agrícola, Facultad Farmacia, Universidad de Santiago de Compostela, 15782-Santiago de Compostela, Spain. E-mail: [email protected] 2Full Professor, Departamento Edafología y Química Agrícola, Facultad Farmacia, Universidad de Santiago de Compostela, 15782-Santiago de Compostela, Spain. E-mail: [email protected] 3Full Professor, Departamento Edafología y Química Agrícola, Facultad Farmacia, Universidad de Santiago de Compostela, 15782-Santiago de Compostela, Spain (corresponding author). E-mail: [email protected] Note. This manuscript was submitted on September 23, 2010; approved on January 27, 2011; published online on January 29, 2011. Discussion period open until January 1, 2012; separate discussions must be submitted for individual papers. This paper is part of the Journal of Materials in Civil Engineering, Vol. 23, No. 8, August 1, 2011. ©ASCE, ISSN 0899-1561/ 2011/8-1239–1248/$25.00. JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / AUGUST 2011 / 1239 Gloss is an optical phenomenon related to the appearance of a surface and represents the capacity of a surface to reflect directed light (ASTM 1995). According to Hunter (1937), six different visual criteria relate to the perception of gloss: (1) specular gloss, (2) sheen, (3) contrast gloss or luster, (4) absence-of-bloom gloss, (5) distinctness-of-image gloss, and (6) surface-uniformity gloss. In the particular case of ornamental rocks, the specular gloss is often used to monitor surface quality because it is related to the surface roughness (Huang et al. 2002;ASTM 1997). Several studies analyzed the relationships among these three aesthetic properties (roughness, color, and gloss) in different materials (Ignell et al. 2009;Shih et al. 2008;Briones et al. 2006;Ariño et al. 2005;Eliades et al. 2004;Keyf and Etikan 2004;Simonot and Elias 2003;Dalal and Natale-Hoffman 1999; Thomas 1999;Barnett 1973). Thus, it has been demonstrated that (1) an inverse relationship exists between roughness and gloss; (2) lightness (LCIELAB coordinate) and chroma (C ab CIELAB coordinate) of the color vary with roughness and gloss; (3) lightness (LCIELAB coordinate) affects gloss; and (4) the magnitude of the change in color with superficial texture is governed by the color of the material. However, all these studies were carried out with homogeneous surfaces from the point of view of color and composition rather than from arbitrary heterogeneous samples such as granite rocks. In this sense, none of the aforementioned studies included ornamental rocks. To date, only one study examining the influence of the induced surface roughness on the color change of ornamental rocks has been published (Benavente et al. 2003). These authors demonstrated that the color variations caused by acid attack on building stones basically depend on variations in surface roughness and the type of rock. In the present study, the effect of four of the most common surface finishes on the appearance of five varieties of ornamental granite, all widely used in building construction and selected for their different colors, was analyzed by means of roughness, color, and gloss measurements to provide an objective standardized method of aesthetic characterization useful for the different workers in the fields of construction and building materials. Experimental Rocks and Mechanical Surface Finishes Five varieties of ornamental rocks of very different color were selected: Grissal, a gray coarse-grained granite; Rosa Porriño, a pinkish coarse-grained granite; Blanco Cristal, a white mediumgrained granite; Silvestre, a white medium-grained with some ochre spots owing to biotite weathering; and Labrador Claro, a bluishblack coarse-grained rock although classified petrographically as a diorite-gabbro, was included in this study because its dark (almost black) color increased the color gamut of the rocks under study. Labrador Claro is one of the most commercialized of the darkest rocks and is also referred to as granite in the ornamental stone industry (Fig. 1). Each variety was petrographically and Fig. 1. Macroscopic appearance of the studied ornamental granites 1240 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / AUGUST 2011 mineralogically characterized by optical microscopy. Information relating to petrographic classification, petrographic texture, mineral composition, and mineral size grain of the studied rocks is shown in Table 1. Five square specimens (36 cm2) of each granitic rock were prepared with different types of surface finish. Four types of finish were applied to four of the granites (in order of increasing roughness): polished, honed, sawn, and flamed (López Jiménez 1996). Flamed and polished finishes cannot be applied to the Silvestre variety of granite, which is more weathered than the others, and these finishes were therefore substituted by bush hammered and polished without glow, respectively. Polishing is a surface treatment often used in granite stones, because it highlights all the colors and petrographic textures of the material. The polishing process is carried out with particles of different grain size, i.e., successively finer particles. The process can close the rock pores because during the polished process, the voids are filled with crushed material from the rock itself or from the abrasive materials used; the process makes the material more resistant to external aggressions because it reduces its water absorbing capacity (Rojo et al. 2003). The resulting surface is flat and glossy with a perceptually darker tone than achieved with other treatments. Honing is very similar to polishing, but the stone does not acquire the characteristic gloss. The honing process is similar to polishing, but only coarse-grained grinding particles are used, and a matte finish is achieved. Polishing without glow, an intermediate finish between polishing and honed, was also used. Sawing is a finish resulting from cutting the granite with steel sheets or diamond disks. This is usually carried out before applying another finish. The resulting appearance is even, matte, and slightly rough, sometimes with small undulations caused by the cut. Flaming is carried out by heating the granite surface with a blowtorch. The thermal shock causes some mineral grains to be shed from the surface, particularly those fragmented by the saw. The result is a rough surface with a glassy appearance. This finish cannot be achieved in brownish, moderate, or highly weathered granite (e.g., Silvestre) because it would modify the aspect of the surface unevenly as a result of chemical alterations. Bush hammering is one of the most commonly used finishes; in earlier times, it was carried out by beating the stone manually with a bush hammer. Nowadays, the system is mechanized so that it produces a more homogeneous rough finish. Roughness Measurements Roughness measurements were performed to quantify the surface texture achieved on the granite specimens with the different surface finishes. Roughness was characterized with a noncontact laser profilometer (UBM microfocus measurement system, UBM Messtechnik GmbH, Ettlingen, Germany) with a measurement range of 500 μm. The noncontact laser profilometer was equipped with a 780 nm wavelength laser triangulation sensor with a spot diameter of 1 μm and an axis and sensor resolution of 0.06 μm. One specimen of each studied rock and surface finish (except flamed and bush hammered finishes, which produce roughness above the threshold of 500 μm) was analyzed. To take into account the surface anisotropy of the ornamental granites, the laser profilometer was driven in a horizontal (XY) plane by a twodimensional (2D) positioning system with two linear motor actuators. A total of six lines of 17.50 mm long, three on longitudinal axis (X) and three on transverse axis (Y), were scanned. The average roughness (Ra) is defined as the integral of the absolute value of the roughness profile over an evaluation length (12.50 mm) and was determined from the six measurements with a resolution or point density of 150 points=mm. Color Measurements The color was measured with a GretagMacbeth (now XRite; Eibar, Spain) portable spectrophotometer CE-XTH equipped with OptiviewSilver/i QC Basic software. The measuring conditions fixed in the device were: diameter viewing aperture of 10 mm, illuminant D65, and observer 2° (CIE 1932) with a d=8°illumination viewing geometry (Prieto et al. 2010). Because the surface of granite samples is not totally reflective or matte, inclusion or exclusion of the specular component may be important for color measurements. Thus, the measurements were made in both specular component Table 1. Petrographic Classification, Petrographic Texture, Mineral Composition and Mineral Grain Size of the Studied Ornamental Granites Mineral % Mineral size (mm) Mean Maximum Minimum Blanco cristal (biotite adamellitic granite; texture: heterogranular-panallotriomorphic of medium grain) Quartz 26 1:71:0 5.4 0.3 Feldspar-K 29 3:21:7 5.5 0.6 Plagioclases 27.5 1:60:9 3.7 0.7 Biotite 9 0:90:3 1.6 0.6 Moscovite 2 0:40:1 0.5 0.2 Chlorite 4.5 0:50:2 0.4 0.3 Grissal (alcaline granite; texture: porphyritic-panallotriomorphic of coarse grain) Quartz 30.5 2:60:9 4.2 1.1 Feldspar-K 34.5 7:06:0 17.6 1.3 Plagioclases 20.75 2:01:4 4.2 0.3 Biotite 8.5 0:60:3 1.0 0.3 Moscovite 0.5 0:20:1 0.3 0.1 Chlorite 3.5 0:70:4 1.3 0.3 Rosa porriño (biotite adamellitic granite; texture: porphyritic-panallotriomorphic of coarse grain) Quartz 30 3:62:3 8.2 0.8 Feldspar-K 33 5:75:2 14.4 1.3 Plagioclases 21 3:32:5 8.0 0.6 Biotite 9 1:40:9 3.2 0.3 Chlorite 3.5 0:80:5 1.6 0.3 Labrador claro (diorite-gabbro with labradorite feldspar; texture: phaneritic-hipidiomorphic of coarse grain) Feldspar 48 8:74:1 15.3 3.5 Pyroxene 16.5 3:93:0 9.6 1.0 Olivine 6.5 0:80:5 1.4 0.4 Biotite 3.5 1:61:7 4.8 0.3 Amphiboles 5.5 1:81:5 4.6 0.5 Opaques 15 0:50:3 1.0 0.3 Apatite 4.0 0:50:2 0.6 0.2 Silvestre (two mica adamellitic granite; texture: equigranular-panallotriomorphic of medium grain) Quartz 29 2:00:6 2.9 0.5 Feldspar-K 26 2:21:0 3.8 1.0 Plagioclases 24 1:90:8 3.2 1.0 Biotite 8 1:00:3 1.6 0.4 Moscovite 8 1:20:6 2.0 0.3 Chlorite 3.5 0:50:2 1.0 0.3 JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / AUGUST 2011 / 1241 Commission International de l’Eclairage (CIE). (1986) “Colorimetry.”Pub. 15-2, CIE Central Bureau, Vienna, Austria. Commission International de l’Eclairage (CIE). (1976). “Recommendations on uniform colour spaces, colour-difference equations, psychometric colour terms.”Proc., 18th Session CIE, Bureau Central de la CIE, Paris. Dalal, E. N., and Natale-Hoffman, K. M. (1999). “The effect of gloss on color.”Color Res. Appl., 24(5), 369–376. Díaz Mauriño, C. (1976). Iniciación práctica a la mineralogía, S. A. Alhambra, Madrid, Spain (in Spanish). Durán-Suárez, J., García-Beltrán, A., and Rodríguez-Gordillo, J. (1995). “Colorimetric cataloguing of stone materials (biocalcarenite) and evaluation of the chromatic effects of different restoring agents.”Sci. Total Environ., 167(1-3), 171–180. Eliades, T., Gioka, E., Heim, M., Eliades, G., and Makou, M. (2004). “Color stability of orthodontic adhesive resins.”Angle Orthod., 74(3), 391–393. García-Talegón, J., Vicente, M. A., Vicente-Tavera, S., and MolinaBallesteros, E. (1998). “Assessment of chromatic changes due to artificial ageing and/or conservation treatments of sandstones.”Color Res. Appl., 23(1), 46–51. Grossi, C. M., Alonso, F. J., Esbert, R. M., and Rojo, A. (2007a). “Effect of laser cleaning on granite color.”Color Res. Appl., 32(2), 152–159. Grossi, C. M., Brimblecombe, P., Esbert, R. M., and Alonso, F. J. (2007b). “Color changes in architectural limestones from pollution and cleaning.”Color Res.Appl., 32(4), 320–331. Hardeberg, J. Y. (1999). “Acquisition and reproduction of color images: Colorimetric and multispectral approaches.”Ph.D. dissertation, Ecole Nationale Superieure des Telecommunications, Paris. Huang, H., Li, Y., Shen, J. Y., Zhu, H. M., and Xu, X. P. (2002). “Microstructure detection of a glossy granite surface machined by the grinding process.”J. Mater. Process. Technol., 129(1–3), 403–407. Huertas, R., Melgosa, M., and Hita, E. (2006). “Influence of random-dot textures on perception of suprathreshold color differences.”J. Opt. Soc. Am. A, 23(9), 2067–2076. Hunter, R. S. (1937). “Methods of determining gloss.”J. Res. Natl. Bur. Stand., 18(77), 281. Hunter, R. S., and Harold, R. W. (1987). The measurement of appearance, Wiley, New York. Ignell, S., Kleist, U., and Rigdahl, M. (2009). “On the relations between color, gloss, and surface texture in injection-molded plastics.”Color Res.Appl., 34(4), 291–298. Iñigo, A. C., Vicente-Tavera, S., and Rives, V. (2004). “MANOVA-Biplot statistical analysis of the effect of artificial ageing (freezing/thawing) on the colour of treated granite stones.”Color Res. Appl., 29(2), 115–120. Iñigo, A. C., Vicente-Tavera, S., Rives, V., and Vicente, M. A. (1997). “Color changes in the surface of granitic materials by consolidated and/or water repellent treatments.”Color Res.Appl., 22(2), 133–141. International Commission on Illumination (CIE). (2004). Colorimetry, 015:2004, 3rd Ed., Vienna. ISO. (1984). “Surface roughness—terminology: Part 1. Surface and its parameters.”ISO 4287-1:1984, Geneva. Keyf, F., and Etikan, I. (2004). “Evaluation of gloss changes of two denture acrylic resin materials in four different beverages.”Dent. Mater., 20(3), 244–251. Kim, I. J., Lee, Y. K., Lim, B. S., and Kim, C. W. (2003). “Effect of surface topography on the color of dental porcelain.”J. Mater. Sci. Mater. Med., 14(5), 405–409. López-Arce, P., Varas-Muriel, M. J., Fernández-Revuelta, B., Álvarez de Buergo, M., Fort, R., and Pérez-Soba, C. (2010). “Artificial weathering of Spanish granites subjected to salt crystallization tests: Surface roughness quantification.”Catena, 83(2-3), 170–185. López Jiménez, C., ed. (1996). Manual de rocas ornamentales, S. L. Entorno Gráfico, Madrid (in Spanish). Prieto, B., Sanmartín, P., Silva, B., and Martínez-Verdú, F. (2010). “Measuring the color of granite rocks: A proposed procedure.”Color Res. Appl., 35(5), 368–375. Prieto, B., Silva, B., Aira, N., and Álvarez, L. (2006). “Toward a definition of a bioreceptivity index for granitic rocks: Perception of the change in appearance of the rock.”Int. Biodeterior. Biodegrad., 58(3-4), 150–154. Rodriguez, J. M., Curtis, R. V., and Bartlett, D. W. (2009). “Surface roughness of impression materials and dental stones scanned by non-contacting laser profilometry.”Dent. Mater., 25(4), 500–505. Rojo, A., Alonso, F. J., and Esbert, R. M. (2003). “Hydric properties of some Iberian ornamental granites with different superficial finishes: A petrophysical interpretation.”Mater. Construcc., 53(269), 61–72. Shih, T.-S., Wei, P.-S., and Wu, C.-L. (2008). “Effect of abrasives on the glossiness and reflectance of anodized aluminum alloys.”J. Mater. Sci., 43(6), 1851–1858. Simonot, L., and Elias, M. (2003). “Color change due to surface state modification.”Color Res. Appl., 28(1), 45–49. Taboada, T., and Garcia, C. (1999). “Smectite formation produced by weathering in a coarse granite saprolite in Galicia (NW Spain)”.Catena, 35(2-4), 281–290. Thomas, T. R. (1999). Rough surfaces, 2nd Ed., Imperial College Press, London. Wang, L., Huang, T., Kamal, M. R., and Rey, A. D. (2000). “The surface topography and gloss of polyolefin blown films.”Polym. Eng. Sci., 40(3), 747–760. Wyszecki, G., and Stiles, W. S. (1982). Color science: Concepts and methods, quantitative data and formulae, Wiley, New York. 1248 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / AUGUST 2011 2ª línea de trabajo Desarrollo de la metodología de medida y caracterización del color de las cianobacterias 2nd line of research Fine-tuning of the methodology for measuring and characterizing the color of cyanobacteria Chapter 3*. Color of cyanobacteria: some methodological aspects Prieto, B.; Sanmartín, P.; Aira, N.; Silva, B. Applied Optics 49 (11): 2022-2029 (2010) JCR index (IF) 2010 = 1.703 (23/78, 29 percentile in Optics) Total number of times cited: 4 * Selected for publication in Virtual Journal for Biomedical Optics. Sec. Microbiology, Vol. 5(8). This article was ranked one of the top 20 articles published on the same topic (domain of article 20390000) since its publication (2010). BioMedLib, "Who Is Publishing in My Domain?". April 14, 2012 Color of cyanobacteria: some methodological aspects Beatriz Prieto,* Patricia Sanmartín, Noelia Aira, and Benita Silva Dpto. Edafología y Química agrícola, Fac. Farmacia, University Santiago de Compostela, 15782-Santiago de Compostela, Spain *Corresponding author: [email protected] Received 8 December 2009; revised 5 March 2010; accepted 9 March 2010; posted 10 March 2010 (Doc. ID 120994); published 1 April 2010 Although the color of cyanobacteria is a very informative characteristic, no standardized protocol has, so far, been established for defining the color in an objective way, and, therefore, direct comparison of experimental results obtained by different research groups is not possible. In the present study, we used colorimetric measurements and conventional statistical tools to determine the effects on the measurement of the color of cyanobacteria, of the concentration of the microorganisms and their moisture content, as well as of the size of the target area and the minimum number of measurements. It was concluded that the color measurement is affected by every factor studied, but that this can be controlled for by making at least 10 consecutive measurements=9:62 cm2at different randomly selected points on the surface of filters completely covered by films of cyanobacteria in which the moisture contents are higher than 50%. © 2010 Optical Society of America OCIS codes: 120.3940, 120.4290, 330.1710. 1. Introduction Although color is one of the defining characteristics of cyanobacteria and provides the name of blue-green algae, it has, so far, not been defined in an objective way. Many studies have explored the color changes that cyanobacteria undergo in response to alterations in the energy distribution in the light spectrum, but none of these studies provide colorimetric data expressed in any of the standard International Commission on Illumination (CIE) color spaces: CIEXYZ, CIELUV, and CIELAB [1,2]. Likewise, in the scientific literature on cyanobacteria, objective terms are not used to refer to the color of the microorganisms. Thus, for instance, bleaching or chlorosis is defined as color change of the cells from “blue-green”to “yellow-green”by phycobilisome degradation, and nitrogen chlorosis is defined as the “yellowing”of cyanobacterial cells following the onset of nitrogen starvation. In some cases the language used is more precise and more specific names of colors, such as red, green, blue, and yellow, are given. However, as communication of the color has an important cultural component, sometimes it is difficult to know the real meaning of the words. The characteristic blue-green color of cyanobacteria is due to their photosynthetic pigment composition of chlorophyll a, which is a greenish pigment that makes photosynthesis possible by passing on charged electrons to other molecules to manufacture energy, and phycobiliproteins, which capture light energy and are exclusive of cyanobacteria and red algae. Between phycobiliproteins, phycocyanin is responsible for the blue color. Other colored photosynthetic pigments in cyanobacteria are phycoerythrin, pink, phycoerythrocyanin, purple, and allophycocyanin, greenish-blue, which are phycobiliproteins, and the yellow-orange pigments called carotenoids. As pigment content is closely associated with environmental conditions, such as nitrogen source, light intensity, light quality, and nutrient availability, among others parameters [3–6], variation in environmental conditions gives rise to an appreciable change in the color of cyanobacteria cultures. Thus, for example, in some cyanobacteria, the light quality, 0003-6935/10/112022-08$15.00/0 © 2010 Optical Society of America 2022 APPLIED OPTICS / Vol. 49, No. 11 / 10 April 2010 i.e., the relative number of photons of blue, green, red, far red, and other portions of the light spectrum emitted from a light source, influences the composition of phycobilisomes. In green light, the cells accumulate more phycoerythrin, whereas in red light, they produce more phycocyanin, and, hence, the bacteria appear green in red light and red in green light. This process is known as complementary chromatic adaptation and is a way for the cells to maximize the use of available light for photosynthesis. Taking into account relationships between pigment content and color, and pigment content and environmental conditions, the health of the cultures or the influence of environmental changes could be assessed by their color. To date, no objective color data have been reported in relation to cultures because there is no standardized protocol for the measurement of cyanobacterial color, even though it is a common practice in microbiological laboratories to make a first assessment of the health of the cultures by visual inspection. Thus, the objective measurement of color is of great importance, not only for a common understanding among researchers, but also for carrying out scientific studies on the ecology and physiology of cyanobacteria. In this sense, it is important to bear in mind that the objective measurement of color is affected by the color measuring devices used and the protocol applied, and comparable results should be obtained by different operators and instruments. Regarding the devices, it must be taken into account that the discrepancies in the color obtained by several contact-type color measuring devices are due to integration of the field of view, which is circular and of a size determined by the diameter of the measuring head (3−60 mm), and only highly homogeneous colors will be unaffected by this fact. Regarding the protocol, objective colorimetric characterization of cyanobacteria has been carried out in only two studies [7,8]. In the first study [7], organisms were previously deposited on a 47 mm diameter acetate filter by means of vacuum filtration of 5ml of culture, five times (i.e., 1 ml each time) to obtain a homogeneous distribution of the cells in a circular area (35 mm diameter); the color of the deposits was then measured with a contact-type color measuring device. To obtain a representative color of the cyanobacteria culture, 10 sequential circular measurements of 8mm diameter were made and color was expressed as the average of the 10 measurements, in the CIELAB color space. The procedure yielded good results and was used in a subsequent study [8]. However, the 10 measurements used in the former studies cannot be considered as standard since the diameter of the circular measurement head differs from one device to another, and as demonstrated with other types of surface [9], the minimum number of measurements required to characterize the color of a surface depends on the diameter of the measuring head, among other factors. Thus, to standardize the measurement of the color of cyanobacteria in order to obtain reproducible and comparable results, independently of the dimensions of the measuring head of the device, the number of measurements required in relation to the surface area must first be determined. Furthermore, taking into account that, in the first study [7], the moisture content of the cyanobacteria was found to affect the colorimetric characterization, relationships between moisture content and colorimetric parameters must be studied to obtain sufficient information to enable selection of the most appropriate conditions for standardization of the measurement of cyanobacterial color. In view of the above, and with the aims of encouraging communication between researchers and of obtaining a standardized protocol for the measurement of cyanobacterial color, the number of measurements, in relation to the surface area, required to characterize their color independently of the dimensions of the measuring head of the device, the amount of cyanobacteria, and the proper moisture content of the microorganisms to obtain reproducible and comparable results were established by a combination of colorimetric and statistical methodology. 2. Materials and Methods Experiments were carried out on a mixed culture (713 μgml −1) of three filamentous N2−fixing heterocyst-forming cyanobacteria: Nostoc sp. strain PCC 9104, Nostoc sp. strain PCC 9025 and Scytonema sp. CCC 9801. All strains were grown in BG −110 medium [10], a diazotrophic culture medium commonly used for nitrogen-fixing strains, composed of K2PO4:3H2Oð0:04 gl −1ÞþMgSO4:7H2Oð0:075 gl −1Þþ CaCl2:2H2Oð0:036 gl −1ÞþCitric acidð0:006 gl −1Þþ Ferric ammonium citrate ð0:006 gl −1ÞþEDTA ð0:001 gl −1ÞþNa2CO3ð0:02 gl −1Þþ Trace metal mix A5ð1ml l−1Þ:H3BO3ð2:86 gl −1ÞþMnCl2·4H2O ð1:81 gl −1ÞþZnSO4·7H2Oð0:222 gl −1Þþ Na2MoO4·2H2Oð0:39 gl −1ÞþCuSO4·5H2O ð0:079 gl −1ÞþCoðNO3Þ2·6H2Oð0:0494 gl −1Þþ distilled water. Depending on the experiment, five replicates of different aliquots of the culture were filtered under vacuum through nitrocellulose filter discs (0:45 μm and 47 mm diameter) to achieve 9:62 cm2of coated area (35 mm diameter). The color of the coated areas of the filters in Fig. 1was then measured with one or two contact-type color measuring devices. To determine the appropriate range of humidity of the cyanobacteria for characterization of their color, a larger volume (20 ml) of the culture was filtered to achieve complete coating of the filter. In this case, the filters were first dried in an oven (105 °C) to constant weight. The filters with the cyanobacterial deposits were then immersed in water until saturated, then air dried, and color and weight were measured at different times (10, 20, 30, 45, 60, 90, 120, and 150 min) until the filters appeared completely dry. Finally, the filters were dried in an oven (at 105 °C) to 10 April 2010 / Vol. 49, No. 11 / APPLIED OPTICS 2023 constant weight to enable determination of the moisture content of the cyanobacteria. The color was measured with a Gretag Macbeth portable spectrophotometer (CE-XTH) with two diameter viewing apertures of 5 and 10 mm. The following measuring conditions were selected: 5mm viewing aperture, illuminant D65, which represents a typical phase of daylight with a correlated color temperature of approximately 6500 K, and a 2° observer (CIE 1931). The measurements were made by spectral reflectance, by use of diffuse illumination geometry, with an integration sphere covered with a white material so that the light was uniformly diffuse in all directions illuminating the sample and was observed with the specular component included. A total of 20 measurements were made consecutively at different randomly selected points on the surface of the filters, in order to obtain sufficient data to characterize the color. The results were expressed as the average of only 10 measurements since this was established as the minimum number of measurements required for each 9:62 cm2of surface (see Subsection 3.B). Color measurements were pointed in the CIELAB color space, the most perceptually uniform of the color spaces [11–13]. The CIELAB system has been widely used (since 1976) for calculating color differences for most practical applications. Use of the CIELAB system enables estimation of three color parameters: L,a, and b, where Lrepresents lightness (a value of 100 indicates white, and a value of 0, black), ais associated with changes in rednessgreenness (positive ais red, and negative ais green), and bis associated with changes in yellowness-blueness (positive bis yellow, and negative b is blue). The three parameters are plotted on three orthogonal axes in a Cartesian coordinate system. In addition, the classic CIELAB formula (ΔEab)[2] and three CIELAB-based color-difference formulas {ΔE94(1:1:1) [14], CMC(2:1) [15], and ΔE00(1:1:1) [11]} were applied. In order to determine the minimum number of measurements required to characterize the color of cyanobacteria, independently of the dimensions of the measuring head of the device and the concentration of the microorganisms, three dilutions (1=5; 1=2, and 1=1) of the culture were prepared (to provide relative concentrations of 20%, 50%, and 100%), and 5ml aliquots were filtered. As a control, a 5ml aliquot of the culture medium was also filtered (0% concentration). In this case, the color was measured with a Gretag Macbeth portable spectrophotometer (CE-XTH) with two diameter viewing apertures of 5 and 10 mm, and a Minolta colorimeter, with one measuring head (CR-300) with an 8mm diameter viewing area. Thus, the areas measured with the two devices were circular areas of 5, 8, and 10 mm diameters. The same measuring conditions as above were fixed in both devices so that the measurements were comparable. A total of 20 measurements were made consecutively, with each of the three measuring heads, of 5, 8,and10 mmdiameters,atdifferentrandomlyselected points on the surface of the filters within the appropriate range of moisture contents (see Subsection 3.A). All the data were subjected to multivariate analysis of variance (MANOVA) and the Tukey-B multiple comparison test, by use of SPSS (version 15.0). 3. Results and Discussion A. Range of Moisture Contents The variations in the L,a, and bvalues for the cyanobacteria deposits with moisture content (expressed as a percentage) are shown in Fig. 2. The values of the Lparameter were relatively low in 28.07–20.00 CIELAB units. The values of the a chromatic parameter were negative, ranging from −1:21 to −2:51 CIELAB units, whereas the values of the bchromatic parameter were positive, ranging from 10.46 to 4.25 CIELAB units, so that the color of the cyanobacteria was within the zone of the yellowFig. 1. Appearance of the filters after depositing the aliquots of cultures containing 0%, 20%, 50%, and 100% cyanobacteria. Fig. 2. Variation in L,a, and bvalues of the filters containing cyanobacteria deposits with moisture content (%). 2024 APPLIED OPTICS / Vol. 49, No. 11 / 10 April 2010 greenish colors, regardless of the moisture content (see Fig. 3). On the other hand, Linitially decreased rapidly with increasing moisture content in Fig. 2. However, it remained practically constant for moisture contents higher than 50%, since the variation in that part of the curve is only 2 Lunits, which is below the perceptibility threshold [16,17]. This was not observed with the aand bparameters, for which differences between ranges of moisture content were not noticeable, since there was no variation in these parameters with moisture content. Moreover, the results of the MANOVA, with the CIELAB color parameters L,a, and bas dependent variables, and the moisture content as the independent variable confirmed statistically significant differences only among the Lvalues obtained at the different moisture contents (Wilks lambda, 0.024; F-value, 26.294; degrees of freedom, 69; significance, 0.000). Therefore, Lis the only CIELAB parameter related to moisture content, and decreased exponentially with increasing humidity, so that moisture content clearly affected the lightness values, as it caused a darkening of the color. A similar result was obtained in a previous study [7], in which variation in the Lparameter with moisture content was observed during the 15 days during which samples were left to dry, but not thereafter. The above results clearly suggest that, in order to obtain comparable results, color measurements must be made on test samples with a moisture content of more than 50%, as in that range, the L,a, and b values remain fairly constant. To determine the color of liquid cultures deposited on filters, the measurements must be made immediately after deposition of cyanobacteria, while the microorganisms are still moist; when determining the color of cyanobacteria colonizing surfaces, such as building facades, the measurement protocol should ensure that the moisture content of the surface is above 50%. B. Minimum Number of Measurements In order to determine the minimum number of measurements required to characterize the color of cyanobacteria, the cumulative averages of the CIELAB color parameters (L,a, and b) were plotted for each of the filters. The general shape of the graphs obtained was an inverted exponential decay model with a horizontal asymptote, with the stable section corresponding to the number of measurements after which the mean become constant; consequently, the first point of this section of the curve represents the minimum number of measurements required to characterize the L,a, and bcoordinates for each case. By way of example, the plots obtained for one of the filters are shown in Fig. 4. In each of the three graphs, the number of measurements after which the horizontal asymptote is reached and the mean Fig. 3. (Color online) a−bdiagram representing the yellowgreenish area corresponding to the filters containing cyanobacteria in relation to range of moisture contents. Fig. 4. Example of the graphs used to determine the minimum number of measurements required. Cumulative averages of the CIELAB color parameters: L,a, and bin relation to the number of measurements, corresponding to filters containing 20% cyanobacteria, filter b, and 10 mm measurement head. 10 April 2010 / Vol. 49, No. 11 / APPLIED OPTICS 2025 becomes constant, is indicated by the marked segment. The minimum number of measurements required for characterization of each color coordinates, for the coated area of each filter and for each measuring head, is shown in Table 1. The number of measurements required was different for each of the coordinates (L,a, and b). The data were subjected to MANOVA, with the minimum number of measurements determined for L,a, and bas dependent variables and concentration of the microorganisms and diameter of the measuring head as independent variables. The results of the analysis revealed that both the concentration of microorganisms and the diameter of the measuring head produced significantly different results in terms of the minimum number of measurements required to characterize L,a, and b. The results, including the Wilks lambda, the F-factor, the level of significance, and the number of degrees of freedom are shown in Table 2, and the significant differences between the average value of number of measurements required to define L,a,andbfor each organisms concentration and measurement head are indicated in Table 3. There were significant differences in the minimum number of measurements required for color characterization of the filters with concentrations of 0% or 100% microorganisms and those with 20% or 50% microorganisms, for all three CIELAB color parameters (L,a, and b), with fewer measurements required for the former than the latter. This may be attributed to the greater heterogeneity of the color of the measured area corresponding to the filters with intermediate concentrations of microorganisms (20% and 50%): as the deposits did not cover the whole area [see Fig. 1], this gave rise to an irregular layer of microorganisms in which two very different colors—the yellow-greenish, fairly dark color of the cyanobacteria, and the pure, bright white color of the filters—appeared together. Thus, depending on the concentration of filtered organisms, these chromatic patches on an achromatic support provided different scales of heterogeneity, which did not exist in the filters without cyanobacteria (0%) and was almost absent in the filters with 100% microorganisms, as a rather homogeneous layer of microorganisms was obtained. The minimum number of measurements required also increased with the heterogeneity of the color of the filters (see Table 1). Regarding the measuring head, the minimum number of measurements required to characterize bwas independent of this factor, while the significant differences for Land awere determined by the 5mm measuring head, for which the number of measurements required was greater than for the 8 and 10 mm measuring heads (see Table 3). Since the color value of each of the five filters with the same concentration of microorganisms must be approximately equal, irrespective of the number of measurements, the greatest differences in partial and total color between filters with the same concenTable 1. Minimum Number of Measurements Determined for Characterization of Each CIELAB Color Parameter, for Each Filter with Different Concentrations of Cyanobacteria, and for Different Measuring Heads Dilution (%) Filter Diameter of the Measuring Head (mm) 10 8 5 LabLabLab 0 a 234222233 b 222222222 c 222222222 d 222222332 e 222222222 20 a 353333653 b 554833874 c 665444333 d 563888887 e 465276355 50 a 444323775 b 6 5 4 2 2 10 10 9 9 c 885595883 d88854510103 e 4 4 3 3 3 3 9 10 10 100 a 222443352 b 222334553 c 422433443 d 322433443 e 433333553 Table 2. Three-Way MANOVA of the Number of Measurementsa Source Wilks Lambda FValue Degrees of Freedom Significance Organism concentration 0.327 6.315 9 0.000 Diameter of the measuring head 0.676 2.882 6 0.014 aDependent: L,a, and b. 2026 APPLIED OPTICS / Vol. 49, No. 11 / 10 April 2010 tration of microorganisms were calculated. The number of measurements required to characterize the color of each filter (see Table 4) was the highest number of measurements from among those for L,a, and b(see Table 1) for each filter and each measuring head. Assuming a large tolerance of five CIELAB units [16–18], partial (ΔL,Δa, and Δb) and total [ΔEab,ΔE94ð1:1:1Þ, CMCð2:1Þ, and ΔE00ð1:1:1Þ] color differences between filters with the same concentration of microorganisms were not perceptible (see Table 5), and were far from the six CIELAB units considered as a perceptible but acceptable difference in color [19]. In addition, taking into account previous studies in which three CIELAB units are considered as the upper limit of perceptibility of the color [8,20,21], all values of the total color differences obtained, except those obtained with the 8 and 5mm measuring heads for filters with a covering of 20% microorganisms, were below this threshold. The total color differences were minimized by use of the newer and improved color formulas, i.e., ΔE94ð1:1:1Þ, CMCð2:1Þand ΔE00ð1:1:1Þ. It was also found that there was no equivalence of the scale factor among the results obtained with the three formulas considered. Moreover, the changes in total color were mainly produced by the partial differences of the lightness ΔL, except for filters with 0% microorganisms, in which the changes were mainly due to Δb, owing to the tendency to yellowing of nitrocellulose filters when they are exposed to ultraviolet light. The greatest effect of ΔLon the filters with concentrations of microorganisms that gave rise to an irregular layer (20%, 50%, and 100%) was due to the textured nature of the samples, where Lis the color parameter that varied most between the white and the green patches. A reduction in the total color differences would be unfeasible, even if more measurements were made, as the values of L,a, and bwould not change [see Fig. 1]. Since the minimum number of measurements required varied with the color parameter, the concentration of microorganisms and the diameter of the measuring head, the minimum number required to characterize the color of cyanobacteria, irrespective of the concentration of microorganisms, would be the largest number obtained for all the Table 3. Tukey-B Test for the Number of Measurements in Relation to Different Concentrations of Cyanobacteria and Diameter of the Measuring Heada Dilution (%) Lab 02:07a2:20a2:20a 20 4:86b5:28b4:23b 50 6b6:07b5:36b 100 3:36a3:18a2:73a Diameter of the measuring head (mm) Lab 10 3:67a3:83a3:33a 83:53a3:53a3:84a 55:12b5:35b3:94a aDifferent superscript letters indicate significant differences (α:0:05). Table 4. Minimum Number of Measurements Required to Characterize Each Filter for Each Measuring Head Dilution (%) Filter Diameter of the Measuring Head (mm) 10 8 5 0a 4 2 3 b2 2 2 c2 2 2 d2 2 3 e2 2 2 20 a 5 3 6 b5 8 6 c6 4 5 d6 8 8 e6 7 5 50 a 4 3 7 b 6 10 10 c8 9 8 d8 5 10 e4 3 10 100 a 2 4 5 b2 4 5 c4 4 4 d3 4 4 e4 3 5 Table 5. Maximum Partial (ΔL,Δa, and Δb) and Total [ΔEab,ΔE94ð1:1:1Þ,ΔE00ð1:1:1Þ, and CMCð2:1Þ] Color Differences for Each Concentration of Cyanobacteria Deposited on the Filters Diameter of the Measuring Head (mm) Dilution (%) ΔLΔaΔbΔEab ΔE94ð1:1:1ÞCMCð2:1ÞΔE00ð1:1:1Þ 10 0 0.15 0.04 0.23 0.28 0.28 0.25 0.25 20 2.31 0.47 0.66 2.45 2.37 1.28 1.56 50 1.63 0.80 0.24 1.83 1.71 0.96 1.25 100 2.84 0.57 0.77 2.99 2.88 1.50 2.47 8 0 0.06 0.12 0.29 0.32 0.29 0.29 0.29 20 3.49 1.39 1.82 4.17 3.78 2.26 2.62 50 2.37 0.86 −0:85 2.66 2.45 1.34 1.74 100 1.92 0.64 0.47 2.08 1.97 1.05 1.55 5 0 0.16 0.06 0.34 0.38 0.37 0.35 0.35 20 2.76 1.30 1.93 3.61 3.12 2.02 2.27 50 2.62 0.94 0.91 2.93 2.70 1.46 1.91 100 2.33 0.69 0.11 2.43 2.36 1.23 1.87 10 April 2010 / Vol. 49, No. 11 / APPLIED OPTICS 2027 concentrations tested for each measuring head. Thus, 10 measurements=9:62 cm2are required for measuring heads of 8 and 5mm diameter and 8measurements=9:62 cm2for a measuring head of 10 mm diameter. C. Influence of Target Area Diameter Once the minimum number of measurements was established in relation to the dimensions of the measuring head, the color values obtained with each measuring head were compared in order to establish whether the results were equivalent. The result of a MANOVA with values of L,a, and bas dependent variables and the diameter of the measurement head as the independent variable revealed no statistically significant differences (p>0:05), among the values obtained with the different measuring heads. Therefore, the color characteristics obtained with the different measuring heads are comparable when the established minimum number of measurements is applied (see Subsection 3.B). From a practical point of view, and taking into account that in heterogeneous surfaces an increase in the field of view of the device reduces the effect of the different colors in the target area, with the consequent similarity in sequential measurements, the use of a 10 mm measuring head is more convenient, as fewer measurements are required (8=9:62 cm2) and the error derived from heterogeneity in the filter color is also reduced. 4. Conclusions The results of the study demonstrate the influence of both the microorganisms and the measuring instrument on the characterization of the color of cyanobacterial biofilms. Instrument properties, such as the diameter of the measuring head, affect the minimum number of measurements required to characterize the color, and properties of the microorganisms, such as concentration and moisture content, affect both the minimum number of measurements required and the color obtained. The minimum number of measurements required increased with increasing heterogeneity of the color of the area measured, which depended on the concentration of the microorganisms, and decreased with the diameter of the measuring head. To control for the influence of the heterogeneity of the color of the area measured, the color of cyanobacteria should be measured on filters that are completely covered by the microorganisms, so that the white color of the filter is hidden. The influence of the size of the measuring head is more difficult to control for, as researchers usually only have one color measuring device available. However, if the number of measurements corresponds to those established in this study in relation to the measuring head diameter, results thus obtained will be comparable. It was established that a total of 10 measurements=9:62 cm2are required for measuring heads of 8 and 5mm diameter and 8measurements=9:62 cm2for a measuring head of 10 mm diameter. We therefore suggest that in order to standardize measurement of the color of cyanobacteria, 10 measurements=9:62 cm2are sufficient for characterization of the color, regardless of the dimensions of the measuring head of the device. However, we also recommend that, when possible, a10 mm measuring head should be used, as fewer measurements are required (8=9:62 cm2) and the error derived from heterogeneity in the color of the filter is reduced. The Lvalues were greatly affected by the moisture content of the samples, whereas the chromatic parameters aand bwere unaffected. Since the values of Ldecreased exponentially with moisture contents up to 50%, which can be considered as the point from which a stable asypmtote is reached on the corresponding graph, characterization of the cyanobacterial color should be made with samples with moisture contents of more than 50%. Otherwise, the results will not be comparable. The methodology proposed here allows objective measure of the color of cyanobacteria, which would be very useful in those studies where knowledge of pigment content variations leads to conclusions. Moreover, it provides some advantages on traditional methodology as it is nondestructive and saves time and materials. The present study was financed by the Science and Education Ministry of Spain (MEC, BIA2006-02233/ BES-2007-16996). References 1. http://www.cie.co.at 2. G. Wyszecki and W. S. Stiles, Color Science: Concepts and Methods, Quantitative Data and Formulae (Wiley, 1982). 3. L. G. Erokhina, “Spectral effects of the chromatic adaptation of nitrogen-fixing cyanobacteria grown on different nitrogen sources,”Mikrobiologiya 61, 960–967 (1992). 4. S. Liotenberg, D. Campbell, R. Rippka, J. Houmard, and N. Tandeau de Marsac, “Effect of the nitrogen source on phycobiliprotein synthesis and cell reserves in a chromatically adapting filamentous cyanobacterium,”Microbiology 142, 611–622 (1996). 5. S. R. Miller, M. Martin, J. Touchton, and R. W. Castenholz, “Effects of nitrogen availability on pigmentation and carbon assimilation in the cyanobacterium Synechococcus sp. strain SH-94-5,”Arch. Microbiol. 177, 392–400 (2002). 6. E. Miśkiewicz, A. G. Ivanov, J. P. Williams, M. U. Khan, S. Falk, and N. P. A. Huner, “Photosynthetic acclimation of the filamentous cyanobacterium, Plectonema boryanum UTEX 485, to temperature and light,”Plant Cell Physiol. 41, 767– 775 (2000). 7. B. Prieto, T. Rivas, and B. Silva, “Rapid quantification of phototrophic microorganisms and their physiological state through their colour,”GBIF 18, 237–245 (2002). 8. B. Prieto, B. Silva, N. Aira, and L. Laiz, “Induction of biofilms on quartz surfaces as a means of reducing the visual impact of quartz quarries,”GBIF 21, 237–246 (2005). 9. B. Prieto, P. Sanmartín, B. Silva, and F. Martínez-Verdú, “Measuring the color of granite rocks. a proposed procedure,” Color Res. Appl., article in press, doi:10.1002/col.20579 (2010). 10. R. Rippka, J. Deruelles, J. B. Waterbury, M. Herdman, and R. Y. Stanier, “Generic assignments, strain histories and 2028 APPLIED OPTICS / Vol. 49, No. 11 / 10 April 2010 and PCs were higher in cultures grown on nitrate containing media than in the cultures grown on media without nitrate, from the 12th day of culture onward. However, from the 10th day of culture, the concentration of carotenoids was affected by different factors, depending on the strain: the availability of nutrients was the main factor affecting carotenoid concentration for Nostoc sp. strain PCC 9104, whereas the presence of nitrates together with the high light intensity affected Nostoc sp. strain PCC 9025 to a greater extent. Statistical comparison of chl a, PC, and carotenoid concentrations relative to culture age was performed for both Nostoc strains (Table 1). For all six experimental cultures, ie with two different light intensities and three different culture media, the mean concentration of each pigment increased significantly during the culture period in both strains. At the end of the experiment, the concentration of chl ain strain PCC 9104 was higher in non-diazotrophic medium under low light conditions (72.63 mg l 71 ) and lower in Table 1. Mean concentrations of the pigments of Nostoc sp. strains PCC 9104 and PCC 9025 for the six experimental cultures over a 14-day growth period. Day 0 2 4 6 8 10 12 14 Nostoc sp. PCC 9104 Chl a (mg l 71 ) BG11 HL 2.99 a 1.07 a 5.93 ab 16.83 abc 23.92 bc 30.96 c 56.96 d 56.17 d LL 1.52 a 0.89 a 5.44 ab 15.76 ab 22.12 ab 39.04 b 78.62 c 72.63 c BG11 0 HL 1.87 a 1.70 a 7.39 a 13.87 ab 23.17 bc 32.35 cd 44.00 d 44.24 d LL 0.59 a 1.48 a 7.49 ab 14.92 bc 21.50 c 32.58 d 45.00 e 41.58 de BG11 0 /10 HL 0.98 a 1.58 a 4.70 ab 9.78 bc 12.72 c 14.67 c 7.68 abc 11.30 bc LL 2.51 a 1.80 a 6.31 ab 13.26 c 11.71 bc 12.83 bc 14.38 c 10.76 bc PCs (10 72 mg l 71 ) BG11 HL 0.1 a 0.3 a 1.2 ab 1.3 ab 1.4 ab 2.9 b 5.0 c 6.7 c LL 0.2 a 0.2 a 1.1 ab 1.0 ab 0.7 a 2.7 b 7.8 c 6.6 c BG11 0 HL 0.2 ab 0.1 a 0.8 b 0.7 b 0.6 ab 1.8 cd 1.5 c 2.2 d LL 0.2 a 0.1 a 0.9 ab 0.7 ab 0.5 a 2.0 b 1.8 b 4.5 c BG11 0 /10 HL 0.1 a 0.1 a 0.6 a 0.5 a 0.2 a 0.2 a 0.1 a 0.8 a LL 0.1 a 0.2 a 0.7 b 0.9 a 0.2 a 0.3 a 0.1 a 0.3 a Carotenoids (mg l 71 ) BG11 HL 0.35 ab 0.15 a 0.42 ab 1.18 bc 1.34 bc 2.05 cd 2.60 d 3.83 e LL 0.31 a 0.10 a 0.23 a 0.96 a 1.21 a 2.38 b 2.60 b 4.36 c BG11 0 HL 0.26 a 0.15 a 0.54 a 0.73 a 1.80 b 2.80 cd 2.17 bc 3.63 d LL 0.36 a 0.27 a 0.59 a 0.86 ab 1.60 bc 2.28 c 2.22 c 3.19 d BG11 0 /10 HL 0.35 a 0.14 a 0.28 a 0.65 ab 1.18 bc 1.51 c 0.63 ab 1.24 bc LL 0.30 a 0.16 a 0.41 a 1.39 b 1.26 b 1.39 b 1.47 b 1.20 b Nostoc sp. PCC 9025 Chl a (mg l 71 ) BG11 HL 5.77 a 15.24 b 15.77 b 11.79 ab 20.20 b 46.27 c 46.99 c 46.02 c LL 6.84 a 6.54 a 15.52 ab 16.15 ab 11.78 ab 31.39 bc 39.33 c 40.96 c BG11 0 HL 7.99 a 8.14 a 10.76 a 11.24 a 10.64 a 33.36 b 24.00 c 31.14 b LL 4.08 a 9.73 b 12.25 b 14.11 bc 12.14 b 19.12 c 25.00 d 25.74 d BG11 0 /10 HL 11.17 a 15.38 ab 15.75 ab 18.35 ab 13.07 ab 30.73 c 14.70 ab 19.85 b LL 8.74 ab 5.91 a 18.08 cd 16.22 bcd 13.31 abc 22.20 d 12.16 abc 13.17 abc PCs (10 72 mg l 71 ) BG11 HL 0.3 a 0.4 a 0.4 a 0.2 a 0.8 a 2.4 b 3.5 bc 4.5 c LL 0.2 a 0.4 a 0.4 a 0.3 a 0.6 a 0.7 a 1.9 b 2.0 b BG11 0 HL 0.2 ab 0.1 a 0.2 ab 0.7 c 0.7 c 0.3 ab 0.8 c 0.9 c LL 0.2 ab 0.1 a 0.2 ab 0.7 c 0.3 b 0.3 b 0.6 cd 0.5 d BG11 0 /10 HL 0.2 ab 0.1 a 0.4 abc 0.4 abc 0.7 c 0.5 bc 0.6 c 0.2 ab LL 0.2 ab 0.1 a 0.6 cd 0.5 bc 0.8 d 0.3 abc 0.5 bc 0.3 abc Carotenoids (mg l 71 ) BG11 HL 0.23 a 0.80 ab 0.71 ab 0.65 ab 1.30 b 2.61 c 2.69 c 2.11 c LL 0.23 a 0.26 a 0.86 abc 1.35 bcd 0.68 ab 1.94 d 1.80 d 1.55 cd BG11 0 HL 0.30 a 0.39 a 0.73 a 0.68 a 0.68 a 1.91 b 1.71 b 1.51 b LL 0.18 a 0.61 ab 0.78 bc 0.84 bcd 0.98 bcd 1.35 cd 1.45 d 1.08 bcd BG11 0 /10 HL 0.39 a 1.01 b 0.99 b 1.21 bc 1.06 b 1.59 d 1.40 cd 1.43 cd LL 0.47 a 0.68 ab 1.04 bc 1.09 bc 1.32 c 1.52 c 1.17 bc 0.73 ab Different superscript letters in each row indicate significant differences (p50.05) between the means of three independent replicates for the pigment content of a given culture. Chl a¼chlorophyll a; PCs ¼phycocyanins; BG-11 ¼non-diazotrophic medium; BG-11 0 ¼diazotrophic medium; BG-11 0 /10 ¼modified diazotrophic medium; HL ¼grown under high light intensity; LL ¼grown under low light intensity. 502 P. Sanmartı´net al. modified diazotrophic conditions at both light intensities (11.30 and 10.76 mg l 71 ). In strain PCC 9025, the lowest chl acontent was also found in nutrientdeficient cultures, but the greatest amount of chl a corresponded to growth medium containing nitrate and a high light intensity. Similar results were observed for the concentrations of PCs, which, in non-diazotrophic media, reached values of 6.6 and 6.7 610 72 mg l 71 for strain PCC 9104 and 2.0 and 4.5 610 72 mg l 71 for strain PCC 9025 by the end of the experiment. The concentrations of carotenoids were similar in the six cultures, and ranged from 4.36 (BG11 at LL) to 1.20 mg l 71 (BG11 0 /10 at LL) in strain PCC 9104 and from 2.11 (BG11 at HL) to 0.73 mg l 71 (BG11 0 /10 at LL) in strain PCC 9025. Significant differences in the mean values of each pigment in relation to the six experimental conditions referred to above are shown in Table 2. The different light intensities (high: 170 mmol photon m 72 s 71 and low: 65 mmol photon m 72 s 71 ) did not have significant effects on the concentration of any of the pigments in Nostoc sp. PCC 9104, irrespective of the culture medium (Table 2). However, for Nostoc sp. PCC 9025 it caused significant changes for all pigments, on BG11 and BG11 0 , ie on both non nutrient-limited cultures. On the other hand, under high light intensity, the concentrations of chl aand carotenoids were significantly different in Nostoc sp. PCC 9104 grown in nutrient-deficient culture medium; in the case of PCs under high light, significant differences were observed between BG11 and both diazotrophic media. In the case of Nostoc sp. PCC 9025, although there was no difference in the concentration of carotenoids in the different media, the concentrations of chl aand PC were significantly different in this strain growing in BG11 medium (Table 2). CIELAB color parameters: L*, a*, b*, C*, h Both strains of Nostoc underwent perceptible and objectively measured variations in color throughout the entire experiment (Figure 3). Statistical tests applied to the color data revealed significant differences for each CIELAB color parameter (L*, a*, b*, C* and h) in relation to age of the culture (Table 3), culture medium (Table 4) and light intensity (Table 4). In the case of Nostoc sp. strain PCC 9104, the L* parameter decreased (the culture darkened) from day 0 to day 10, regardless of the culture medium and light intensity, with values ranging from 86.34 to 47.26 CIELAB units. However, from the 10th day of culture, nutrient-deficient cultures recovered lightness (the L* value increased up to 78.17 CIELAB units), whereas the lightness of the other cultures stabilized at the Table 2. Mean concentrations of the pigments of Nostoc sp. strains PCC 9104 and PCC 9025 grown in three different culture media and exposed to either high or low light intensity. Light intensity Chl a (mg l 71 ) PCs (10 72 mg l 71 ) Carotenoids (mg l 71 ) HL LL HL LL HL LL Nostoc sp. PCC 9104 BG11 24:36a a29:50a a2:5a a2:8a a1:49a a1:52a a BG11 0 21:08a a20:64a ab 1:1a b1:4a ab 1:51a a1:42a a BG11 0 /10 7:93a b9:19a b0:3a b0:3a b0:75a b0:96a a Nostoc sp. PCC 9025 BG11 26:01a a21:06b a1:6a a0:8b a1:39a a1:08b a BG11 0 17:16a b15:27b ab 0:5a b0:3b b0:99a a0:91b a BG11 0 /10 17:37a b13:72a b0:4a b0:4a b1:13a a1:00a a Different superscript letters indicate significant differences (p50.05) between means pigment contents in the same culture medium in relation to light intensity. Different subscript letters indicate significant differences (p50.05) between means pigment contents in the same light intensity in relation to culture medium. Chl a¼chlorophyll a; PCs ¼phycocyanins; BG-11 ¼non-diazotrophic medium; BG-11 0 ¼diazotrophic medium; BG-11 0 /10 ¼modified diazotrophic medium; HL ¼grown under high light intensity; LL ¼grown under low light intensity. Figure 3. Cultures of Nostoc sp. PCC 9104 over a 14-day growth period, showing difference in appearance consequent on the culture medium. Only cultures grown under low light intensity are shown. Biofouling 503 Table 3. Values of CIELAB parameters for Nostoc sp. strains PCC 9104 and PCC 9025 for the six experimental cultures over a 14-day growth period. Nostoc sp. PCC 9104 Nostoc sp. PCC 9025 Day 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 L* BG11 HL 84.68 a 82.96 ab 85.72 a 75.43 bc 72.77 c 68.19 d 51.12 de 47.37 e 91.82 a 86.31 a 66.22 b 57.84 b 58.57 b 44.22 c 36.85 c 43.59 c LL 82.33 ab 86.34 a 76.50 b 78.82 b 69.14 c 60.02 d 49.38 e 50.43 e 92.39 a 87.48 a 70.28 bc 64.96 cd 77.27 b 55.76 e 57.35 de 61.58 de BG11 0 HL 80.56 a 77.89 ab 71.28 cd 73.15 bc 66.25 d 47.26 e 48.75 e 41.47 f 91.06 a 90.66 a 71.78 c 71.63 c 80.02 b 56.25 d 56.39 d 60.58 d LL 83.98 a 80.58 a 66.27 b 75.96 c 67.03 b 49.12 d 52.69 d 44.33 e 90.20 a 89.33 a 74.77 b 75.67 b 78.95 b 61.59 c 59.66 c 65.55 c BG11 0 /10 HL 81.77 a 78.50 ab 70.89 c 74.75 bc 70.54 c 62.71 d 78.17 ab 75.38 bc 83.85 a 83.99 a 71.17 b 67.08 bc 70.69 b 58.39 d 52.89 d 64.77 c LL 80.17 a 77.86 ab 65.88 d 76.20 abc 72.86 c 61.11 e 67.32 d 75.12 bc 88.08 a 90.66 a 66.92 b 71.89 c 73.30 c 60.23 d 53.70 e 61.78 d a* BG11 HL 77.80 a 74.54 b 74.07 b 79.28 a 715.33 c 714.92 c 719.67 d 719.44 d 72.30 a 74.59 b 79.57 c 711.19 c 713.83 d 715.01 d 713.48 d 714.74 d LL 76.95 a 72.50 b 79.41 cd 78.13 ac 711.02 d 717.23 e 716.71 e 718.01 e 72.28 a 73.99 a 77.78 b 710.21 c 78.20 bc 712.85 d 714.65 e 713.39 de BG11 0 HL 77.49 a 76.08 b 714.20 c 711.10 d 717.07 e 721.31 f 723.28 g 720.45 f 72.42 a 72.87 a 76.02 b 79.73 c 78.69 c 713.57 d 716.83 e 714.27 d LL 77.72 a 75.78 b 718.42 c 713.44 d 719.69 c 723.21 e 723.49 e 723.92 e 72.57 a 73.51 a 74.92 b 79.47 c 79.10 c 712.31 d 717.24 e 715.67 e BG11 0 /10 HL 77.01 a 75.44 a 714.04 b 712.25 b 714.44 b 716.91 c 712.51 b 713.73 b 74.85 a 75.82 a 78.12 b 712.09 c 712.34 c 715.11 d 717.39 e 715.76 d LL 77.33 a 75.76 a 715.72 cd 712.68 b 714.49 bc 717.09 de 719.10 e 714.77 bc 73.42 a 73.21 a 79.87 b 710.80 b 712.73 c 714.95 d 718.18 e 716.16 de b* BG11 HL 9.00 a 4.03 b 4.31 b 9.45 a 17.91 c 14.93 d 21.88 e 21.81 e 2.88 a 5.76 b 7.21 bc 8.73 cd 10.68 de 10.05 de 9.61 de 10.87 e LL 7.06 a 1.30 b 9.82 cd 8.18 ac 11.12 d 15.75 e 17.44 ef 19.62 f 3.21 a 5.32 b 5.72 b 7.68 c 7.22 c 7.93 c 10.59 d 9.78 d BG11 0 HL 7.22 a 11.80 b 26.17 c 20.25 d 34.22 e 31.59 e 33.74 e 32.21 e 3.13 a 4.18 a 4.99 a 10.54 bc 9.68 b 11.74 c 15.55 d 15.53 d LL 8.53 a 10.92 a 28.79 bd 19.35 c 31.60 d 27.76 b 30.21 bd 30.89 d 3.04 a 5.37 b 3.14 a 9.65 c 10.15 c 9.26 c 14.47 d 9.66 c BG11 0 /10 HL 7.22 a 10.18 a 24.51 b 24.60 b 30.28 c 30.40 c 24.10 b 27.00 bc 5.20 a 8.33 b 8.09 b 15.49 c 17.56 d 18.35 d 23.57 e 22.69 e LL 7.76 a 9.55 a 27.57 b 25.67 b 33.09 cd 29.73 bc 35.22 d 28.02 b 4.15 a 4.44 a 10.45 b 13.79 c 17.29 d 17.87 d 22.47 e 18.38 d C* BG11 HL 11.93 a 6.16 b 5.96 b 13.27 a 23.60 c 21.13 c 29.46 d 29.23 d 3.71 a 7.37 b 12.01 c 14.23 cd 17.54 e 18.09 e 16.58 de 18.36 e LL 9.93 a 2.94 b 13.60 cd 11.57 ac 15.67 d 23.36 e 24.18 e 26.64 e 3.95 a 6.66 b 9.74 c 12.82 de 10.96 cd 15.12 ef 18.10 g 15.85 fg BG11 0 HL 10.42 a 13.29 b 29.78 c 23.12 d 38.25 e 38.11 e 41.00 e 38.16 e 3.97 a 5.08 a 7.88 b 14.37 c 13.03 c 17.95 d 22.93 e 21.09 e LL 11.52 a 12.37 a 34.19 bd 23.61 c 37.25 d 36.19 b 38.30 bd 39.16 d 4.01 a 6.42 b 5.88 b 13.53 c 13.64 c 15.41 d 22.51 e 13.58 c BG11 0 /10 HL 10.08 a 11.55 a 28.28 bc 27.51 bc 33.58 cd 34.79 d 27.17 b 30.29 bcd 7.15 a 10.17 b 11.49 b 19.67 c 21.48 c 23.78 d 29.30 e 27.64 e LL 10.79 a 11.17 a 31.74 bc 28.66 b 36.13 cd 34.31 c 40.07 d 31.68 bc 5.41 a 5.48 a 14.38 b 17.53 c 21.47 d 23.32 d 28.90 e 22.58 d h BG11 HL 131.54 a 141.61 b 131.93 a 134.16 a 130.66 a 135.20 a 132.23 a 131.92 a 128.33 a 128.24 a 143.39 b 141.73 b 141.71 b 146.23 b 144.72 b 143.87 b LL 135.73 a 158.80 b 133.51 a 134.29 a 134.57 a 137.63 a 134.37 a 132.63 a 124.89 a 127.24 a 143.99 cd 142.47 bc 138.36 b 148.49 d 143.87 cd 141.12 bc BG11 0 HL 136.36 a 118.63 b 118.51 b 118.35 b 116.51 b 124.02 c 124.65 c 122.56 c 127.76 a 124.35 ab 142.18 e 133.12 c 131.51 bc 139.21 de 137.35 d 132.51 c LL 132.52 a 118.66 b 122.64 c 124.07 c 121.96 c 129.95 ad 127.89 d 128.07 d 130.06 a 123.14 b 149.43 c 134.31 c 131.88 ac 143.25 d 139.95 d 134.60 c BG11 0 /10 HL 134.36 a 118.48 bc 120.31 b 117.13 cd 115.48 d 119.13 bc 117.37 cd 117.11 cd 132.13 a 124.65 c 135.25 b 128.20 cd 124.88 c 129.57 ad 126.42 cd 124.65 c LL 137.70 a 121.40 b 119.60 b 116.20 bc 113.60 c 119.90 b 118.43 c 117.78 c 129.15 ab 125.97 ad 133.44 c 128.52 abd 126.37 ad 129.90 b 128.98 ab 125.40 d Different superscript letters indicate significant differences (p50.05) between the means of three independent replicates (the reported color parameters for each replicate are the mean values of five measurements) for the values of CIELAB parameters for a given culture (same row). L*¼lightness; a*¼redness-greenness changes; b*¼yellowness-blueness changes; C*¼chroma or saturation; h¼hue or tone; BG-11 ¼non-diazotrophic medium; BG-11 0 ¼diazotrophic medium; BG-11 0 /10 ¼modified diazotrophic medium; HL ¼grown under high light intensity; LL ¼grown under low light intensity. 504 P. Sanmartı´net al. lowest values (Figure 4 and Table 3). In the same way, the L* values decreased throughout the experiment for Nostoc sp. strain PCC 9025, although in this case, from 8 days, the L* values of cultures growing in nitrate containing medium under high light intensity were lower than in the other cultures, and only reached 36.85 CIELAB units (Figure 4 and Table 3). Color parameters (a* and b*) associated to redness-greenness and yellowness-blueness changes respectively, ranged between 72.50 and 723.92 for a*inNostoc sp. PCC 9104, and between 72.28 and 718.18 for Nostoc sp. PCC 9025, while b* values ranged between 35.22 and 1.30 for Nostoc sp. PCC 9104, and between 23.57 and 2.88 for Nostoc sp. PCC 9025. Thus, both strains of Nostoc showed a yellowgreenish color, irrespective of the growth conditions, although Nostoc sp. PCC 9104 turned a stronger bluish-greener color. In both strains, the response of a* to the age of the culture was similar to that reported for L*, with a significant decrease up to the 10th day of culture for Nostoc sp. PCC 9104, and up to the 12th day for Nostoc sp. PCC 9025, followed by stabilization of values (Figure 4). With respect to b*, a significant and notable difference in values corresponding to nondiazotrophic cultures from the diazotrophic cultures was observed for Nostoc sp. PCC 9104 from the second day onwards, while in the case of Nostoc sp. PCC 9025, the difference was between the nutrient-deficient culture and non nutrient-deficient cultures. The b* values of the samples growing on the nitrate-containing medium ranged from 1.30 to 21.88 for Nostoc sp. PCC 9104, and between 2.88 and 10.87 for Nostoc sp. PCC 9025, indicating that such samples were bluer than those growing on nitrate-free media, which Table 4. Values of CIELAB parameters of Nostoc sp. strains PCC 9104 and PCC 9025 in three different culture media and exposed to either high or low intensity. L*a*b*C*h Light intensity HL LL HL LL HL LL HL LL HL LL Nostoc sp. PCC 9104 BG11 69:59a a69:12a a11:88a a11:25a a12:92a a11:29b a17:59a a15:99b a133:66a a137:69b a BG11 0 63:33a b64:99a b15:12a b16:96a b24:65a b23:51a b29:02a b29:08a b122:45a b125:72a b BG11 0 /10 74:09a c72:06a a12:04a a13:37a c22:29a b24:58a b25:41a b28:07b b119:92a b120:60a c Nostoc sp. PCC 9025 BG11 60:68a a70:88b a10:59a ab 9:04a a8:22a a7:18a a13:48a a11:65a a139:77a a138:80b a BG11 0 71:05a b74:46b a9:30a a8:58b a9:42a a8:09b a13:29a a11:87b a133:50a b135:83b b BG11 0 /10 67:85a b70:82a a11:43a b10:78a b14:91a b13:60a b18:84a b17:39a b128:24a c128:47a c Different superscript letters indicate significant differences (p50.05) between the means of three independent replicates (the reported color parameters for each replicate are the mean values of five measurements) for the values of CIELAB parameters in the same culture medium in relation to light intensity. Different subscript letters indicate significant differences (p50.05) between the means of three independent replicates (the reported color parameters for each replicate are the mean values of five measurements) for the values of CIELAB parameters in the same light intensity in relation to culture medium. L*¼lightness; a*¼redness-greenness changes; b*¼yellowness-blueness changes; C*¼chroma or saturation; h¼hue or tone; BG-11 ¼non-diazotrophic medium; BG-11 0 ¼diazotrophic medium; BG-11 0 /10 ¼modified diazotrophic medium; HL ¼grown under high light intensity; LL ¼grown under low light intensity. Figure 4. CIELAB color data. Values of L* (lightness), a* (redness-greenness changes), b* (yellowness-blueness changes), C* (chroma or saturation) and h(hue or tone) in aerated batch cultures of Nostoc sp. PCC 9104 (left) and Nostoc sp. PCC 9025 (right). -&-¼non-diazotrophic medium BG-11, -~-¼diazotrophic medium BG-11 0 ,-.- ¼modified diazotrophic medium BG-11 0 /10 under high light intensity (open symbols) and low light (solid symbols). Data points represent the average of three replicates; vertical bars represent the SD. Biofouling 505 exhibited the highest b* values (between 7.22 and 35.22 for Nostoc sp. PCC 9104, and between 4.15 and 23.57 for Nostoc sp. PCC 9025) (Table 3). Regarding the chroma or saturation of the samples, in both strains, the results for C* showed a similar pattern to those observed for b*, with values ranging between 2.94 and 41.00 CIELAB units for Nostoc sp. PCC 9104 and between 3.71 and 29.30 CIELAB units for Nostoc sp. PCC 9025. Hue angle (h) varied significantly in Nostoc sp. PCC 9104 at the second day of culture in all growth conditions, with increasing values in non-diazotrophic medium and decreasing values in diazotrophic and modified diazotrophic media; from the second day in the case of non-diazotrophic medium, the hvalues returned to the initial values and remained stable until the end of the experiment (Figure 4). The initial increase, which was greater in non-diazotrophic medium under low light intensity (Figure 4), appeared to reflect the sharp change in appearance (color and density) that the culture underwent on the second day (Figure 3). In Nostoc sp. PCC 9025, significant changes in non-diazotrophic cultures began from the fourth day, whereas they began on the second day in the diazotrophic cultures (Table 3). However, there was no acute change during the experiment (Figure 4). The effects of light intensity on the value of CIELAB color parameters are shown in Table 4. In the case of Nostoc sp. PCC 9104, differences in light intensity only gave rise to differences in b*, C*, and h when this strain was grown in non-diazotrophic medium and in C* when it was grown in modified diazotrophic medium. In the case of Nostoc sp. PCC 9025, differences in light intensity gave rise to significant differences in all CIELAB parameters, on diazotrophic medium, and only in L* and hon nondiazotrophic medium. Differences in the composition of the culture medium had significant effects on the CIELAB color parameters (Table 4). The changes in color of the strains in response to nutritional status differed: in Nostoc sp. strain PCC 9025, growth on modified diazotrophic medium gave rise to significant differences in most of the CIELAB color parameters, while in Nostoc sp. strain PCC 9104, the nondiazotrophic medium resulted in highly significant differences. Correlations between pigment content and CIELAB color parameters Spearman correlation coefficients were calculated to evaluate relationships between pigment contents and CIELAB color parameters (Table 5). All CIELAB parameters, except h, were correlated [(**), p50.01] with all pigment concentrations (chl a, carotenoids, PCs) in both strains, and were more closely correlated in Nostoc sp. strain PCC 9104 than in strain PCC 9025. Moreover, there was a closer correlation between carotenoid concentration and CIELAB color parameters than between either chl aor PCs, and CIELAB color parameters. The Spearman correlation coefficients revealed that the increase in pigment content led Table 5. Bivariate Spearman’s correlation matrix among pigment contents and CIELAB color parameters for Nostoc sp. strains PCC 9104 and PCC 9025. PCs Chl aCarotenoids L*a*b*C*h Nostoc sp. PCC 9104 PCs 1.000 Chl a0.686** 1.000 Carotenoids 0.631** 0.941** 1.000 L*70.689** 70.836** 70.833** 1.000 a*70.608** 70.798** 70.816** 0.891** 1.000 b* 0.262** 0.538** 0.597** 70.668** 70.819** 1.000 C* 0.360** 0.621** 0.671** 70.745** 70.897** 0.983** 1.000 h0.305** 0.079 70.006 0.022 0.146 70.626** 70.508** 1.000 Nostoc sp. PCC 9025 PCs 1.000 Chl a0.586** 1.000 Carotenoids 0.610** 0.876** 1.000 L*70.603** 70.800** 70.844** 1.000 a*70.598** 70.735** 70.825** 0.913** 1.000 b* 0.452** 0.545** 0.670** 70.687** 70.883** 1.000 C* 0.529** 0.643** 0.762** 70.811** 70.962** 0.970** 1.000 h0.383** 0.518** 0.401** 70.551** 70.315** 70.069 0.102 1.000 n¼144; significance level **p50.01. PCs ¼phycocyanin concentration; Chl a¼chlorophyll aconcentration; Carotenoids ¼carotenoid concentration; L*, a*, b*, C*, h¼CIELAB color parameters. 506 P. Sanmartı´net al. to a significant decrease in L*anda* values and a significant increase in b* and C* values. Discussion Previous studies have shown that the color of epilithic cyanobacterial biofilms growing on rocky substrata changes as the biofilm develops and that the directions in which the color parameters move in the CIELAB space indicate whether the population is developing or aging (Prieto et al. 2002, 2005). This was considered in further detail in the present study, and the relationship between color and pigments such as chlorophyll (green), phycocyanins (blue), and carotenoids (orange) was investigated. Moreover, since growth and pigment production are affected by environmental conditions, different culture conditions were employed in order to cover a wider range of color so that differences in color could be analyzed and relationships established. The experimental design enabled significant differences in the appearance of the cultures to be observed, and their color characterized objectively. The results confirmed the original hypotheses: (a) color, as pigment content, is a changing characteristic related to environmental parameters, and (b) there is a close relationship between pigment production and color of the microorganisms. The same optimal range of growth was observed for both strains of cyanobacteria tested, in terms of the presence of nitrates and availability of nutrients. In contrast, the effect of light intensity differed: low intensity favored growth of strain PCC 9104, and high intensity favored growth of strain PCC 9025. Assessment of the optimal conditions for growth was made by assessing the concentrations of pigments. However, the results obtained showed that values of CIELAB color parameters could have been employed for the same purpose, as the concentration of each pigment was correlated with each CIELAB color parameter, except for h; thus when the pigment content increased, b* and C* increased, and L* and a* decreased (Table 5). However, not all CIELAB color parameters are equally sensitive to changes in pigment concentrations. Thus, a* moved to negative values (the culture became greener) as the culture aged (Figure 4) and pigment content increased (Figure 2), but did not respond to the differences in concentrations of chl a, PCs, and carotenoids that occurred from the 10th day among cultures (Figure 2). L* appears to be most informative parameter since, besides being the CIELAB color parameter most closely related to each pigment studied (Table 5), it reflects the changes in pigment concentrations due to the different environmental conditions considered (Figures 3 and 4). The increase in the concentration of chl a, PCs, and carotenoids throughout the experiment was reflected in the increase in parameters b* and C*. Moreover, as expected, both parameters provided important information about the amount of PCs in relation to the amount of the other pigments studied (carotenoids þ chl a). Proof of this is that the evolution of b* and C* was different on the second and fourth days for both strains growing on nitrate cultures. On these days, the b*andC* values for Nostoc sp. PCC 9025 hardly changed relative to the culture growing on the other media (Db* and DC* under 10 CIELAB units); however, in Nostoc sp. PCC 9104 there was an acute difference between nitrate-containing cultures and culture without nitrate. Thus although the values of b*andC* increased sharply (cultures became more yellow) in cultures without nitrate, they decreased (cultures became more blue) on the second and fourth days in nitrate-containing cultures under high light intensity and only on the second day in nitratecontaining cultures under low light intensity (Figure 4). These variations are reflected in the variations in the ratio between PCs and carotenoids þchl afor both strains (data not shown). It can be deduced that the increase in the concentrations of PCs, with the concomitant increase in the blue color of the organisms is revealed by the b*andC* values. Although his correlated with each pigment in Nostoc sp. PCC 9025 and with chl aconcentration in Nostoc sp. PCC 9104 (Table 5), the correlation coefficients are too low to establish any relationship between pigments and the CIELAB color parameter. However, the constant surveillance to which cultures were subjected, together with analysis of the hvalues enabled this parameter to be related to changes in the appearance of the cultures (color and density) since the most important change in appearance occurred between day 2 and day 4 in PCC 9025 and day 0 and day 2 in PCC 9104, for all conditions analyzed, and on the second day the appearance of the PCC 9104 cultures growing in media with nitrate was quite different from the cultures growing in media without nitrate (Figure 3). These changes are clearly reflected in the h values (Figure 4). This is consistent with the findings of other authors, considering hangle as the color parameter most closely related to human color perception (Wyszecky and Stiles 1982; Martı´nez-Verdu´2002). It is important to note that while differences in pigment contents were most notable from the eighth day, differences in CIELAB color parameters occurred earlier, which indicates that color quantification provides information not only on pigment production but also on changes in appearance. The color data recorded in this work indicate the area of CIELAB space in which both strains define their color, with Biofouling 507 maximum values as follows: L*: 86.34, a*: 72.50, b*: 35.22, C*: 41.00, and h: 158.808for Nostoc sp. PCC 9104 and L*: 92.39, a*: 72.28, b*: 23.57, C*: 29.30, and h: 149.438for Nostoc sp. PCC 9025, and minimum values as follows: L*: 41.47, a*: 723.92, b*: 1.30, C*: 2.94, and h: 113.608for Nostoc sp. PCC 9104 and L*: 36.85, a*: 718.18, b*: 2.88, C*: 3.71, and h: 123.148for Nostoc sp. PCC 9025. This is the first time that objective color values of different species of cyanobacteria growing in different environmental conditions have been obtained, and is a first approximation to the definition of the color gamut of cyanobacteria, which is essential information for the development of any color-based methodology for monitoring cyanobacterial growth. Conclusions The results of the present study demonstrate that Nostoc sp. strain PCC 9104 and Nostoc sp. strain PCC 9025 respond in different ways in terms of color and pigment production to environmental conditions such as light intensity, macronutrient limitation, and availability of nitrates. While PCC 9104 is more strongly affected by nutrient limitations, the presences of nitrate together with high light intensity are the factors that determine growth of strain PCC 9025. Moreover, the effect of these factors was only observed from the eighth day of culture. The color of the strains differed throughout the experiment; PCC 9104 was lighter and more chromatic than PCC 9025. CIELAB color parameters were correlated with the chl a, carotenoid and PC contents so that variations in pigment contents were reflected by variations in color. Each CIELAB color parameter adequately described a specific aspect of the development of the two Nostoc strains. L*, and to a lesser extent a*, was related to the chl a, PC, and carotenoid concentrations; b* and C* were related to the concentration of PCs, and h, which refers to the dominant wavelength, reflected the changes in appearance of the cultures, ie changes in color and density, and this parameter represents the major color perception attribute. Analysis of cyanobacterial color in relation to a set of environmental conditions is a useful method to study growth thereby enabling the relationships between color and other physiological properties to be established. It should also provide useful information needed to develop the method of quantification for early detection of epilithic biofilms on rocky substrata. However, it must be stressed that the conclusions are based on data for two samples cultured for only a short period of time. Future research in this field will focus on the study of communities forming biofilms subjected to different environmental conditions in the field. Acknowledgments The present study was financed by the Spanish Ministry of Science and Education (MEC, BIA2006-02233/BES-200716996). The authors thank Prof. Dr A. Otero for her discussion and constructive suggestions for this work. References Bartual A, Lubian LM, Galvez JA, Niell FX. 2002. Effect of irradiance on growth, photosynthesis, pigment content and nutrient consumption in dense cultures of Rhodomonas salina (Wislouch) (Cryptophyceae). Cienc Mar 28:381–392. Bell RA, Sommerfeld MR. 1987. Algal biomass and primary production within a temperature zone sandstone. Am J Bot 74:294–297. Bennet A, Bogoard L. 1973. Complementary chromatic adaptation in a filamentous blue-green alga. J Cell Biol 58:419–435. Berns RS. 2000. Billmeyer and Saltzman’s principles of color technology. 3rd ed. New York: John Wiley & Sons. Collier JL, Grossman AR. 1992. Chlorosis induced by nutrient deprivation in Synechococcus sp. strain PCC 7942: not all bleaching is the same. J Bacteriol 174:4718– 4726. Crispim CA, Gaylarde PM, Gaylarde CC, Neilan BA. 2006. Deteriogenic cyanobacteria on historic buildings in Brazil detected by culture and molecular techniques. Int Biodeterior Biodegr 57:239–243. Duke CS, Allen MM. 1990. Effects of nitrogen starvation on polypeptide composition, ribulose-1,5-bisphosphate carboxylase/oxygenase, and thylakoid carotenoprotein content of Synechocystis sp. strain PCC6308. Plant Physiol 94:752–759. Fernandez-Valiente E, Legane ´s F. 1989. Regulatory effect of pH and incident irradiance on the levels of nitrogenase activity in the cyanobacterium Nostoc UAM205. J Plant Physiol 135:623–627. Gorbushina AA. 2007. Life on the rocks. Environ Microbiol 9:1613–1631. Grondona I, Monte E, Rives V, Vicente MA. 1997. Lichenized association between Septonema tormes sp. nov., a coccoid cyanobacterium, and a green alga with an unforeseen biopreservation effect of Villamayor sandstone at Casa Lis of Salamanca, Spain. Mycol Res 101:1489–1495. Jonte L, Rosales N, Bricen ˜o B, Morales E. 2003. La irradiancia y la salinidad modulan el crecimiento de la cianobacteria Synechocystis minuscula en cultivos discontinuos. Multiciencias 3:7–14. Lonneborg A, Lind LK, Kalla SR, Gustafsson P, Oquist G. 1985. Acclimation processes in the light-harvesting system of the cyanobacterium Anacystis nidulans following a light shift from white to red light. Plant Physiol 78:110–114. Loreto C, Rosales N, Bermu´dez J, Morales E. 2003. Produccio ´n de pigmentos y proteı´nas de la cianobacteria Anabaena PCC 7120 en relacio ´n a la concentracio ´nde nitro ´geno e irradiancia. Gayana Bot 60:83–90. Martı´n-Trillo M. 1995. Afloramientos masivos (blooms) de cianobacterias en los arrozales valencianos: seguimiento de su desarrollo y caracterizacio ´n de dos estirpes formadoras [PhD Dissertation]. Universidad Auto ´noma de Madrid, Spain: Ciencias Biolo ´gicas. 508 P. Sanmartı´net al. Martı´nez-Verdu´F. 2002. Revisio ´n de los aspectos croma ´ticos sobre la captura de ima ´genes (II). O ´ptica Pura Aplic 35:61–75. McNamara CJ, Mitchell R. 2005. Microbial deterioration of historic stone. Front Ecol Environ 3:445–451. Prieto B, Rivas T, Silva B. 2002. Rapid quantification of phototrophic microorganisms and their physiological state through their colour. Biofouling 18:229–236. Prieto B, Silva B, Aira N, Laiz L. 2005. Induction of biofilms on quartz surfaces as a means of reducing the visual impact of quartz quarries. Biofouling 21:237– 246. Prieto B, Sanmartı´n P, Aira N, Silva B. 2010. Color of cyanobacteria: some methodological aspects. Appl Optics 49:2022–2029. Ramı´rez M, Herna ´ndez-Marine ´M, Novelo E, Rolda ´nM. 2010. Cyanobacteria-containing biofilms from a Mayan monument in Palenque, Mexico. Biofouling 26:399–409. Raps S, Wyman K, Siegelman HW, Falkowski PG. 1983. Adaption of the cyanobacterium Microcystis aeruginosa to light intensity. Plant Physiol 72:829–832. Raven J. 1984. A cost-benefit analysis of photon absorption by photosynthetic unicells. New Phytol 98:593– 625. Rippka R, Deruelles J, Waterbury JB, Herdman M, Stanier RY. 1979. Genetic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111:1–61. Saiz-Jimenez C, Hermosin B, Ortega-Calvo JJ, Go ´mezAlarco ´n G. 1991. Aplication of analytical pyrolysis to the study of stony cultural properties. J Anal Appl Pyrol 20:239–251. Satapathy DP, Adhikary SP. 1993. Epilithic algae from temple walls and caves at Brubaneswar, Puri and Konark. Phykos 32:17–20. Silva B, Prieto B. 2004. Deteriorative effects of lichens on granite monuments. In: Clair LL, Mark RD, editors. Biodeterioration of stone surfaces: lichens and biofilms as weathering agents of rocks and cultural heritage. Dordrecht, The Netherlands: Kluwer Academic Publishers. p. 69–77. Silva B, Rivas T, Prieto B. 1999. Effects of lichens on the geochemical weathering of granitic rocks. Chemosphere 39:379–388. Stebvens WA, Spurdon C, Onyon LJ, Stirpe F. 1981. Effect of inhibitors of protein synthesis from plants on tabacco mosaic virus infection. Experientia 37:257–259. Tandeau de Marsac N. 1977. Ocurrence and nature of chromatic adaptation in cyanobacteria. J Bacteriol 130:82–91. Thornbush MJ. 2008. Grayscale calibration of outdoor photographic surveys of historical stone walls in Oxford, England. Col Res Appl 33:61–67. Vo ¨lz HG. 2001. Industrial color testing. Weinheim: Wiley– VCH. Warscheid T, Braams J. 2000. Biodeterioration of stone: a review. Int Biodeterior Biodegr 46:343–368. Wellburn AR. 1994. The spectral determination of chlorophylls aand b, as well as total carotenoids, using various solvents with spectrophotometers of different resolution. J Plant Physiol 144:307–313. Wollenweider RA. 1969. A manual on methods for measuring primary production in aquatic environments. IBP Handbook. No. 12. Oxford, England: Blackwell Scientific Publishers. Wyman M, Fay P. 1986. Underwater light climate and the growth and pigmentation of planktonic blue-green algae (Cyanobacteria). I. Influence of light quantity. Proc R Soc Lond 227:367–380. Wyszecki G, Stiles WS. 1982. Color science. Concepts and methods, quantitative data and formulae. New York: John Wiley and Sons. Biofouling 509 Chapter 5*. Color measurements as a reliable method for estimating chlorophyll degradation to phaeopigments Sanmartín, P.; Villa, F.; Silva, B.; Cappitelli, F.; Prieto, B. Biodegradation 22 (4): 763-771 (2011) JCR index (IF) 2010 = 2.012 (77/160, 48 percentile in Biotechnology and Applied Microbiology) Total number of times cited: 5 * This article was ranked one of the top 20 articles published on the same topic (domain of article 20425659) since its publication (2011). BioMedLib, "Who Is Publishing in My Domain?". April 15, 2012. resulted in a significant color change (Table 2). The L* parameter, associated to lightness, increased significantly with the exposure time of biocide. The increase of L* was higher in planktonic assay, with values ranging from 58.21 to 85.17 CIELAB units, whereas biofilm assay had a slighter range of variation between 60.51 and 74.06 CIELAB units. Previous studies evidenced the suitability of L* parameter to estimate the cell population growth (Prieto et al. 2002; Sanmartı ´n et al. 2010), being the decrease in L* close related with the population growth, and its increase with the end of growth. Thus, in both planktonic and biofilm biocide susceptibility assays, cell activity and growth were affected, which could have been followed by a degradation of pigments. This process was more severe in planktonic cells where the reduction of L* parameter was two times higher with respect to biofilm assay (Table 2). Regarding a*, associated to greenness (-)–redness (?) changes, ranged between -20.23 and -0.88 for planktonic assay, and between -15.09 and -2.39 for biofilm assay. b* values, associated to blueness (-)– yellowness (?) changes, ranged between 27.03 and 8.49 for planktonic assay, and between 11.78 and 4.35 for biofilm assay (Table 2). Thus, application of the biocide gave rise to a decrease in the green and yellow components of the color, being higher in the planktonic culture. The latter indicates a better physiological state (Prieto et al. 2002; Sanmartı ´n et al. 2010)ofNostoc sp PCC 9104 forming biofilm at the end of the experiment. Results for chroma C* ab , showed a similar pattern to that observed for b*, with values ranging between 33.77 and 8.94 CIELAB units for planktonic assay and between 19.15 and 4.98 CIELAB units for biofilm assay. In both cases Nostoc sp. PCC 9104 cells lost chroma in their color. Hue angle (h ab )ofNostoc sp. PCC 9104 cells decreased significantly with the exposure time of biocide, varied in planktonic assay between a yellow–greenish hue (126.82°) and a yellow hue (95.04°); and in biofilm assay between a very slightly bluish green hue (142.01°) and a yellow– greenish hue (120.16°). h ab offered a clear and comprehensive vision of color changes occurred in Nostoc sp. PCC 9104 in both planktonic and biofilm biocide susceptibility assays. It described the lower chlorophyll adegradation occurred in the biofilm tests. Table 2 Values of CIELAB color parameters (L*, a*, b*, C* ab ,h ab )ofNostoc sp. PCC 9104 strain in planktonic state and forming biofilm Planktonic assay Biofilm assay Exposure time (h) 0 3 9 24 0 3 9 24 L* 58.21 ±2.92 a 71.37 ±5.77 b 79.86 ±3.38 c 85.17 ±2.77 d 60.51 ±6.32 a 60.68 ±7.33 a 62.04 ±0.42 a 74.06 ±4.78 b a* -20.23 ±1.29 a -8.14 ±1.44 b -2.73 ±0.41 c -0.88 ±0.17 d -15.09 ±1.82 a -9.45 ±0.65 b -5.80 ±0.60 c -2.39 ±0.39 d b* 27.03 ±2.03 a 8.72 ±0.96 b 8.49 ±0.87 b 9.96 ±0.29 c 11.78 ±1.44 a 6.70 ±0.91 b 11.76 ±0.15 a 4.35 ±1.41 c C* ab 33.77 ±2.39 a 12.04 ±0.29 b 8.94 ±0.74 c 10.00 ±0.29 c 19.15 ±2.23 a 11.63 ±0.37 b 13.12 ±0.35 c 4.98 ±1.40 d h ab 126.82 ±0.38 a 132.85 ±8.15 b 108.05 ±4.08 c 95.04 ±0.93 d 142.01 ±1.95 a 144.66 ±5.14 a 116.19 ±2.20 b 120.16 ±5.57 c Different superscript letters indicate significant differences (p\0.05) between the means of three independent replicates (the reported color parameters for each replicate are the mean values of five measurements) for the values of CIELAB color parameters in planktonic and biofilm assays, in relation to the exposure time to biocide 768 Biodegradation (2011) 22:763–771 123 Spearman correlation coefficients were calculated to evaluate relationships among phaeophytination indexes and CIELAB color parameters, chlorophyll aand ATP cell contents (Table 3). The statistical analysis revealed that the phaeophytination indexes A 435 nm /A 415 nm and A 665 nm /A 665a nm were related to chlorophyll acontent in both planktonic (r 435/415 : 0.615**; r 665/665a : 0.727** Pearson correlation test) and biofilms tests (r 435/415 : 0.916**; r 665/665a : 0.578* Pearson correlation test). These findings demonstrated the validity of the phaeophytinization ratios in evaluating the chlorophyll adegradation for both Nostoc sp. planktonic and biofilm lifestyle. Pearson correlation tests also showed that the L* CIELAB color parameter represents an effective tool in describing chlorophyll degradation as is correlated to both phaeophytinization ratios in both lifestyles. Moreover, a* and C* ab are correlated to the A 665 nm / A 665a nm index in both lifestyles. Conclusions The suitability of color measurements as a reliable method for estimating chlorophyll degradation to phaeopigments has been stated in this work. The determination of the phaeophytination indexes A 435 nm /A 415 nm and A 665 nm /A 665a nm , which have been proved as useful in describing degradation of chlorophyl ato phaeopigments in both Nostoc sp. PCC 9104 planktonic and biofilm biocide susceptibility assays, could be substituted by color measurements as there is a closed relation between them and the color parameters. Among the CIELAB parameters L* appeared to be the most informative parameter in describing the biocidal activity of Biotin T Ò against Nostoc sp. in both planktonic and biofilm mode of growth. The methodology tested in this study will be very useful in those fields where sampling is a critical aspect, as in the field of conservation of works of art, since it will allows to monitorizate the effectiveness of biocides as the degree of chlorophyll degradation to phaeopigment avoiding sampling. Acknowledgments The present study was financed by Xunta de Galicia (REF: 09TMT014203PR) and Science and Education Ministry of Spain (MEC) (BES-2007-16996). The authors would like to thank Dr Pamela Principi for critically reading this manuscript. Table 3 Correlation matrices showing the Pearson coefficients (r) found among the phaeophytination indexes, the chlorophyll acontent, ATP values and CIELAB color parameters Planktonic assay Biofilm assay Chl aATP L* a* b* C* ab h ab Chl aATP L* a* b* C* ab h ab A 435 nm /A 415 nm 0.615** 0.690** -0.603** -0.735** 0.088 0.467 0.541* 0.916** 0.081 -0.676* -0.408 0.292 0.204 0.317 A 665 nm /A 665a nm 0.727** 0.564* -0.547* -0.536* 0.269 0.620** 0.387 0.578* 0.534 -0.613* -0.697* 0.574 0.613* 0.368 Significance level *p\0.05; **p\0.01 Biodegradation (2011) 22:763–771 769 123 References Acea MJ, Diz-Cid N, Prieto-Fernandez A (2001) Microbial populations in heated soils inoculated with cyanobacteria. Biol Fertil Soils 33(2):118–125 Acea MJ, Prieto-Fernandez A, Diz-Cid N (2003) Cyanobacterial inoculation of heated soils: effect on microorganisms of C and N cycles and on chemical composition in soil surface. Soil Biol Biochem 35(4):513–524 Agrawal SB (1992) Effect of supplemental UV-B radiation on photosynthetic pigment, protein and glutathione contents in green algae. Environ Exp Bot 32:137–143 Anderl JN, Franklin MJ, Stewart PS (2000) Role of Antibiotic penetration limitation in Klebsiella pneumoniae biofilm resistance to ampicillin and ciprofloxacin. Antimicrob Agents Chemother 44:1818–1824 Barnes JD, Balaguer L, Manrique E, Davison AW (1992) A reappraisal of the use of DMSO for the extraction and determination of chlorophyll aand bin lichens and higher plants. Environ Exp Bot 32:85–90 Batista JF, Pereira RFC, Lopes JM, Carvalho MFM, Feio MJ, Reis MAM (2000) In situ corrosion control in industrial water systems. Biodegradation 11(6):441–448 Bell RA, Sommerfeld MR (1986) Algal biomass and primary production within a temperature zone sandstone. Am J Bot 74:294–297 Cappitelli F, Abbruscato P, Foladori P, Zanardini E, Ranalli G, Principi P, Villa F, Polo A, Sorlini C (2009) Detection and elimination of Cyanobacteria from Frescoes: The case of the St. Brizio Chapel (Orvieto Cathedral, Italy). Microb Ecol 57:633–639 Carlson RE, Simpson J (1996) A coordinator’s guide to volunteer lake monitoring methods. North American Lake Management Society, Madison Collier PJ, Ramsey AJ, Waigh RD, Douglas KT, Austin P, Gilbert P (1990) Chemical reactivity of some isothiazolone biocides. J Appl Bacteriol 69:578–584 CIE Publication 15-2. Colorimetry (1986) CIE Central Bureau, Vienna Costerton JW (2007) The biofilm primer Springer. Springer, Berlin De Muynck W, Maury Ramirez A, De Belie N, Verstraete W (2009) Evaluation of strategies to prevent algal fouling on white architectural and cellular concrete. Int Biodeterior Biodegrad 63(6):679–689 De Saravia SGG, Naranjo JDLP, Guiamet P, Arenas P, Borrego SF (2008) Biocide activity of natural extracts against microorganisms affecting archives. Boletin latinoamericano y del caribe de plantas medicinales y aromaticas 7(1):25–29 Dubosc A (2000) Etude du de ´veloppement de salissures biologiques sur les parements en be ´ton: mise au point d’essais acce ´le ´re ´s de viellissement. Laboratoire Mate ´riaux et Durabilite ´des Constructions. Toulouse (France), INSA Escadeillas G, Bertron A, Ringot E, Blanc P, Dubosc A (2009) Accelerated testing of biological stain growth on external concrete walls. Part 2: quantification of growths. Mater Struct 42(7):937–945 Fonseca AMD (2009) Avaliac¸a ˜o da efica ´cia de tratamentos convencionais e aplicac¸o ˜es alternativas para prevenir a biodeteriorac¸a ˜o em patrimo ´nio cultural. Dissertac¸a ˜ode Mestrado em Conservac¸a ˜o e Restauro. Departamento de Conservac¸a ˜o e Restauro Faculdade de Cie ˆncias e Tecnologia, Universidade Nova de Lisboa Gladis F, Eggert A, Karsten U, Schumann R (2010) Prevention of biofilm growth on man-made surfaces: evaluation of antialgal activity of two biocides and photocatalytic nanoparticles. Biofouling 26(1):89–101 Hall-Stoodley L, Costerton JW, Stoodley P (2004) Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Microbiol 2(2):95–108 Herasimenka Y, Cescutti P, Sampaio Noguera CE, Ruggiero JR, Urbani R, Impallomeni G, Zanetti F, Campidelli S, Prato M, Rizzo R (2008) Macromolecular properties of cepacian in water and in dimethylsulfoxide. Carbohydr Res 343(1):81–89 Hueck HJ (1965) The biodeterioration of materials as part of hylobiology. Material und Organismen 1(1):5–34 Lopez J, Retuerto R, Carballeira A (1997) D665/D665a index vs. frequencies as indicators of bryophyte response to physicochemical gradients. Ecology 78(1):261–271. Ecological Society of America, USA Louda JW, Li J, Liu L, Winfree MN, Baker EW (1998) Chlorophyll-adegradation during cellular senescence and death. Org Geochem 29(5–7):1233–1251 Manrique E, Redondo EL, Serin ˜a E, Izco J (1989) Estimation of chlorophyll degradation into phaeophytin in Anaptychia ciliaris as a method to detect air pollution. Lazaroa 11:141–148 Martı ´nez-Abaigar J, Nu ´n ˜ez-Olivera E (1998) Ecophysiology of photosynthetic pigments in aquatic bryophytes. In: Bates JW, Ashton NW, Duckett JG (eds) Bryology for the Twenty-first Century. Maney Publishing and the British Bryological Society, Leeds, pp 277–292 Martı ´nez-Abaigar J, Otero S, Toma ´sR,Nu ´n ˜ez-Olivera E (2008) High-level phosphate addition does not modify UV effects in two aquatic bryophytes. Bryologist 111(3): 444–454 Pisani T, Paoli L, Gaggi C, Pirintsos SA, Loppi S (2007) Effects of high temperature on epiphytic lichens: Issues for consideration in a changing climate scenario. Plant Biosyst 141(2):164–169 Prieto B, Rivas T, Silva B (2002) Rapid quantification of phototrophic microorganisms and their physiological state through their colour. Biofouling 18:229–236 Prieto B, Silva B, Aira N, Laiz L (2005) Induction of biofilms on quartz surfaces as a means of reducing the visual impact of quartz quarries. Biofouling 21(5–6):237–246 Prieto B, Sanmartı ´n P, Aira N, Silva B (2010) Color of cyanobacteria: some methodological aspects. Appl Opt 49: 2022–2029 Principi P, Villa F, Bernasconi M, Zanardini E (2006) Metal toxicity in municipal wastewater activated sludge investigated by multivariate analysis and in situ hybridization. Water Res 40(1):99–106 Ricart M, Barcelo ´D, Geiszinger A, Guasch H, de Alda ML, Romanı ´AM, Vidal G, Villagrasa M, Sabater S (2009) Effects of low concentrations of the phenylurea herbicide diuron on biofilm algae and bacteria. Chemosphere 76(10):1392–1401 770 Biodegradation (2011) 22:763–771 123 Rippka R, Deruelles J, Waterbury JB, Herdman M, Stanier RY (1979) Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111:1–61 Ronen R, Galun M (1984) Pigment extraction from lichens with dimethyl sulfoxide (DMSO) and estimation of chlorophyll degradation. Environ Exp Bot 24:239–245 Sanmartı ´n P, Aira N, Devesa-Rey R, Silva B, Prieto B (2010) Relationship between color and pigment production in two stone biofilm-forming cyanobacteria (Nostoc sp. PCC 9104 and Nostoc sp. PCC 9025). Biofouling 26(5):499–509 Silkina A, Bazes A, Vouve F, Le Tilly V, Douzenel P, Mouget JL, Bourgougnon N (2009) Antifouling activity of macroalgal extracts on Fragilaria pinnata (Bacillariophyceae): A comparison with Diuron. Aquat Toxicol 94(4):245–254 Underwood GJC, Paterson DM (1993) Recovery of intertidal benthic diatoms after biocide treatment and associated sediment dynamics. J Mar Biol Assoc UK 73(1):25–45 UNESCO (2000) Unesco to protect masterpieces of the oral and intangible heritage of humanity. http://www.unesco. org/bpi/eng/unescopress/2000/00-48e.shtml Wollenweider RA (1969) A manual on methods for measuring primary production in aquatic environments. IBP Handbook 12 Davis Co. 213 Wyszecki G, Stiles WS (1982) Color science, concepts and methods, quantitative data and formulae, 2nd edn. Wiley, New York Xu KD, McFeters GA, Stewart PS (2000) Biofilm resistance to antimicrobial agents. Microbiology 146:547–549 Young ME, Alakomi HL, Fortune I, Gorbushina AA, Krumbein WE, Maxwell I, McCullagh C, Robertson P, Saarela M, Valero J, Vendrell M (2008) Development of a biocidal treatment regime to inhibit biological growths on cultural heritage: BIODAM. Env Geol 56(3–4):631–664 Biodegradation (2011) 22:763–771 771 123 3ª línea de trabajo Detección y cuantificación de los organismos sobre las construcciones 3rd line of research Detection and quantification of microorganisms colonizing the surface of buildings and monuments Chapter 6 .Quantification of phototrophic biomass on rocks: optimization of chlorophyll-a extraction by response surface methodology Fernández-Silva, I.; Sanmartín, P.; Silva, B.; Moldes, A.; Prieto, B. Journal of Industrial Microbiology and Biotechnology 38: 179-188 (2011) JCR index (IF) 2010 = 2.416 (63/160, 39 percentile in Biotechnology and Applied Microbiology) Total number of times cited: 2 ORIGINAL PAPER Quantification of phototrophic biomass on rocks: optimization of chlorophyll-aextraction by response surface methodology I. Ferna ´ndez-Silva •P. Sanmartı ´n•B. Silva • A. Moldes •B. Prieto Received: 18 March 2010 / Accepted: 26 July 2010 / Published online: 6 September 2010 ÓSociety for Industrial Microbiology 2010 Abstract Biological colonization of rock surfaces constitutes an important problem for maintenance of buildings and monuments. In this work, we aim to establish an efficient extraction protocol for chlorophyll-aspecific for rock materials, as this is one of the most commonly used biomarkers for quantifying phototrophic biomass. For this purpose, rock samples were cut into blocks, and three different mechanical treatments were tested, prior to extraction in dimethyl sulfoxide (DMSO). To evaluate the influence of the experimental factors (1) extractant-to-sample ratio, (2) temperature, and (3) time of incubation, on chlorophyllarecovery (response variable), incomplete factorial designs of experiments were followed. Temperature of incubation was the most relevant variable for chlorophyll-aextraction. The experimental data obtained were analyzed following a response surface methodology, which allowed the development of empirical models describing the interrelationship between the considered response and experimental variables. The optimal extraction conditions for chlorophyllawere estimated, and the expected yields were calculated. Based on these results, we propose a method involving application of ultrasound directly to intact sample, followed by incubation in 0.43 ml DMSO/cm 2 sample at 63°C for 40 min. Confirmation experiments were performed at the predicted optimal conditions, allowing chlorophyll-arecovery of 84.4 ±11.6% (90% was expected), which implies a substantial improvement with respect to the expected recovery using previous methods (68%). This method will enable detection of small amounts of photosynthetic microorganisms and quantification of the extent of biocolonization of stone surfaces. Keywords Chlorophyll-aIncomplete factorial design  Response surface methodology Stone biofilms  Ultrasonic methods Introduction Stone surfaces exposed to the open air are inevitably colonized by a variety of organisms, some of which are responsible for biofilm formation. Among them, algae and cyanobacteria have great importance, since they feature an extracellular matrix composed primarily of exopolymers (EPS) which are involved in the formation of the biofilm and in the resistance of biofilms to adverse abiotic conditions [1–4]. Once established, algal-cyanobacterial biofilms are added to by heterotrophic bacteria and fungi, forming a microbial biocenosis [5]. Biofilms are responsible for apparent staining of rocks due to the biogenic pigments of phototrophic organisms [1,6] and the formation of black patinas [7,8], and may also affect the physicochemical properties of mineral materials [5]. This process is particularly important when the stone under consideration is the building material of monuments of historic and cultural interest [9–12]. Thus, determining the extent of algal and cyanobacterial colonization is crucial for the study of the deterioration of rocky works of art and for the development of methods to This article is part of the BioMicroWorld 2009 Special Issue. I. Ferna ´ndez-Silva P. Sanmartı ´nB. Silva B. Prieto (&) Dpto. Edafoloxı ´a e Quı ´mica Agrı ´cola, Fac. Farmacia, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain e-mail: [email protected] A. Moldes Dpto. Ingenierı ´a Quı ´mica. E.T.S. Ingenieros Industriales, Vigo, Spain 123 J Ind Microbiol Biotechnol (2011) 38:179–188 DOI 10.1007/s10295-010-0843-1 control their deterioration, as well as for the development of bioreceptivity assays of building materials. Therefore, as a starting point, it is necessary to establish a reliable method for quantifying biocolonization on stone surfaces. One of the most commonly used biomarkers for quantifying microalgal and cyanobacterial biomass is chlorophyll-a, which has been extensively used to estimate photosynthetic growth in water, liquid media, and soil, and to a minor extent for estimating algal biomass in rocky substrata [13–16]. Prieto et al. [17] compared several methods for biomass quantification on stone surfaces, and found that chlorophyll-awas a good estimator of biofilm biomass. However, they observed problems in the total extraction of chlorophyll-afrom stone and obtained a relatively high value as a lower limit of detection, which would prevent early detection of stone biocolonization below this lower limit. These authors suggested that the stone itself impedes total chlorophyll-aextraction, and concluded that optimization experiments carried out on stone samples were necessary. Traditional optimization methods examine a single factor at a time, while fixing all other variables at one level. Their major disadvantage is that these methods do not include the interactive effects among the variables studied. As a consequence, these techniques do not depict the complete effects of the parameters on the response. To avoid this problem, response surface methodology (RSM) was developed by Box and collaborators in the 1950s [18]. RSM is a collection of mathematical and statistical techniques based on the fit of a polynomial equation to the experimental data, which must describe the behavior of a data set with the objective of making statistical predictions. It can be applied when a response is influenced by several variables with the objective of simultaneously optimizing the levels of these variables to attain the best system performance [19]. Response surface designs are useful for modeling a curved quadratic surface to continuous factors. A response surface model can pinpoint a minimum or maximum response inside the factor region. Three distinct values for each factor are necessary to fit a quadratic function, so standard two-level designs cannot fit curved surfaces. Three-level full factorial designs are used, in which factors can take on three values: low, medium or center, and high. Generally, if midpoints or center points are added to a 2 k full factorial design then it will become a 3 k full factorial design, where kis the number of factors. However, the main disadvantage of this design (3 k ) is the need for a large number of experimental runs, which produces unwanted high-order interactions, and in addition can be expensive and time consuming. Therefore, designs that present a smaller number of experimental points, such as the Box–Behnken method where the number of experiments required (N) is given by N=2k(k-1) ?C 0 , where kis the number of variables and C 0 is the number of center points, are more often used. In this work, we tried various combinations of mechanical and ultrasonic methods as pretreatments to improve the extraction efficiency of chlorophyll-afrom stone samples, as this is the primary photosynthetic pigment present in organisms responsible for biofilm formation on building materials. To find the optimal conditions of the three independent variables (sample-to-extractant ratio, temperature, and extraction time) potentially influencing the efficiency of phytopigment extraction with dimethyl sulfoxide (DMSO), we performed a response surface analysis following a Box–Behnken design. Finally, we compare the results achieved by the different pretreatments and discuss the adequacy of these methods for quantification of the extent of biocolonization on rock surfaces. Materials and methods Preparation of samples Experiments were carried out with a mixed culture of the filamentous N 2 -fixing heterocyst-forming cyanobacteria strains Nostoc PCC 9025, Nostoc PCC 9104, and Scytonema CCC 9801, grown in BG11 0 medium. The mixed inoculum consisted of 0.42 g (dry weight) of each strain per liter of medium. Ten milliliters of mixed culture (equivalent to 12.7 g total dry weight and 90.5 lg chlorophyll-a) was sprayed onto the surface of 6 9691cm 3 blocks cut from a granite rock. The blocks were placed in a climatic chamber for 2 days under stationary conditions at 25°C, 95% humidity, and 12 h of light (1,600 lx) to induce biofilm formation before pigment extraction was carried out. To determine the chlorophyll-acontent from the culture mixture, five replicate aliquots of 5 ml culture were filtered, and the filters were added to 5 ml DMSO and heated to 65°C for 1 h as described in [17]. After filtration of the samples to remove filter fragments, absorbance of the extracts was measured at 649.1 and 665.1 nm wavelength (A 665.1 and A 649.1 ). The concentration of chlorophyll-a(C a ) was calculated using the equation proposed by Wellburn [20]: Ca¼12:47A665:13:62A649:1: Mechanical pretreatment of samples Three different block pretreatments were assayed to determine the best procedure for complete chlorophyllaextraction (Table 1). For the first method (pretreatment A), 15 inoculated blocks were crushed to obtain fragments no larger than 0.25 cm 3 , which were added to DMSO following the protocol described by Prieto et al. [17]. For pretreatment B, 15 inoculated blocks were 180 J Ind Microbiol Biotechnol (2011) 38:179–188 123 crushed and introduced into 250-ml glass flaks, DMSO was added, and the flasks were introduced into an ultrasonic bath (Transonic T780, Elma TM ) filled with enough water to apply ultrasound to the whole of each sample, during 30 min (the water temperature was measured during this process to ensure that no heating was taking place). For pretreatment C, 15 intact inoculated blocks (without crushing) were placed onto Petri plates containing DMSO and sonicated by inserting the narrow tip of an ultrasonic generator (UP200S; Dr Hielscher GmbH) into the extractant. Sonication was for 5 930 s (0.5 duty cycle, 60% amplitude), with 30 s breaks to avoid overheating. Design of experiments For each pretreatment we performed a Box–Behnken design for three-variable optimization with 13 experimental points plus 2 additional experiments at the central point (three central replicates), to study the influence of the experimental conditions on the extraction yield. (Note that this makes a total of 15 experiments, which in comparison with a 3 3 design with 27 experiments, is more economical and efficient.) The independent variables used in this study and their variation limits are listed in Table 1.ES corresponds to the extractant/sample ratio, expressed as volume of DMSO/ volume of the sample blocks (v/v) for the crushed samples (pretreatments A and B) and as volume of DMSO/sample surface (cm 2 ) for the intact blocks (pretreatment C); Tcorresponds to the temperature of extraction (°C); and t, to the extraction time (min). The levels of the variable were coded, namely each studied real value was transformed into coordinates inside a scale with dimensionless values proportional to its location in the experimental space. Codification enables the investigation of variables of different orders of magnitude without the greater influencing the evaluation of the lesser [19]. The standardized (coded) dimensionless independent variables employed, having variations limits (-1, 1), were defined as x 1 (coded extractant/sample ratio), x 2 (coded temperature), and x 3 (coded extraction time). The correspondence between coded and uncoded variables was established by linear equations deduced from their respective variation limits (Table 1). The dependent variable considered was chlorophyll-a, measured by the variable y 1 . After incubation in DMSO at their corresponding conditions, samples were filtered, and the concentration of chlorophyll-awas determined, as described before. Table 1 Independent and dependent variables employed in this study Sample pretreatment Nomenclature Experiment Block crushing A Block crushing and sonication in DMSO (ultrasonic bathing 30 min) B Whole blocks sonication in DMSO (tip ultrasonic generator 2 min 30 s) C Variable Nomenclature Units Variation range Independent variables Extractant/sample (v/v) ratio a,b or extractant volume/sample surface c ES ml/cm 3a,b or ml/cm 2c 1.39–1.94 a ; 1.67–2.22 b ; 0.28–0.56 c Temperature T°C 30–80 Time tmin 30–90 Variable Nomenclature Definition Variation range Dimensionless, coded independent variables Dimensionless extractant/sample ratio x 1 (ES-1.67)/1.39 a ;(ES-1.94)/1.67 b ; (ES-0.42)/0.28 c (-1, 1) Dimensionless temperature x 2 (T-55)/40 (-1, 1) Dimensionless time x 3 (t-60)/30 (-1, 1) Variable Nomenclature Units Dependent variables Chlorophyll-ay 1 lg a Pretreatment A b Pretreatment B c Pretreatment C J Ind Microbiol Biotechnol (2011) 38:179–188 181 123 microbial biomass, and sediment stability in biofilms of intertidal sediments. Microb Ecol 39(2):116–127 17. Prieto B, Silva B, Lantes O (2004) Biofilm quantification on stone surfaces: comparison of various methods. Sci Total Environ 333:1–7 18. Box GEP, Wilson KB (1954) The exploration and exploitation of response surfaces: some general considerations and examples. Biometrics 10:16–60 19. Bezerra MA, Santelli RE, Oliveira EP, Villar LS, Escaleira LA (2008) Response surface methodology (RSM) as a tool for optimization in analytical chemistry. Talanta 76:965–977 20. Wellburn AR (1994) The spectral determination of chlorophylls a and b, as well as total carotenoids, using various solvents with spectrophotometers of different resolution. J Plant Physiol 144:307–313 21. Haaland PD (1989) Experimental design in biotechnology. Marcel Dekker, New York, NY 22. Devesa R, Moldes A, Dı ´az-Fierros F, Barral MT (2007) Extraction study of algal pigments in river bed sediments by applying factorial designs. Talanta 72:1546–1551 23. Arin ˜o, X (1996) Estudio de la colonizacio ´n, distribucio ´n e interaccio ´ndelı ´quenes, algas y cianobacterias con materiales pe ´treos de los conjuntos arqueolo ´gicos de Baelo Claudia y Carmona. Ph.D. thesis. Universidad de Barcelona and IRNA de Sevilla 188 J Ind Microbiol Biotechnol (2011) 38:179–188 123 Chapter 7. Spectrophotometric color measurement for early detection and monitoring of greening on granite buildings Sanmartín, P.; Vázquez-Nion, D.; Silva, B; Prieto, B. Biofouling: The Journal of Bioadhesion and Biofilm Research 28 (3): 329-338 (2012) JCR index (IF) 2010 = 3.333 (5/92, 5 percentile in Marine and Freshwater Biology; 38/160, 23 percentile in Biotechnology and Applied Microbiology) Total number of times cited: 0 Spectrophotometric color measurement for early detection and monitoring of greening on granite buildings P. Sanmartı´n*, D. Va ´zquez-Nion, B. Silva and B. Prieto Departamento Edafologı´a y Quı´mica Agrı´cola, Facultad Farmacia, Universidad de Santiago de Compostela, 15782-Santiago de Compostela, Spain (Received 12 January 2012; final version received 1 March 2012) This paper addresses the detection and monitoring of the development of epilithic phototrophic biofilms on the granite fac¸ade of an institutional building in Santiago de Compostela (NW Spain), and reports a case study of preventive conservation. The results provide a basis for establishing criteria for the early detection of phototrophic colonization (greening) and for monitoring its development on granite buildings by the use of color changes recorded with a portable spectrophotometer and represented in the CIELAB color space. The results show that parameter b* (associated with changes of yellowness-blueness) provides the earliest indication of colonization and varies most over time, so that it is most important in determining the total color change. The limit of perception of the greening on a granite surface was also established in a psycho-physical experiment, as Db*: þ0.59 CIELAB units that correspond, in the present study, to 6.3 mg of biomass dry weight cm 72 and (8.43 +0.24) 610 73 mgof extracted chlorophyll acm 72 . Keywords: preventive conservation; CIELAB color system; biofouling; greening; monitoring; phototrophic colonization Introduction The fouling of stone surfaces by abiotic substances and organisms poses serious problems for the maintenance of all types of buildings. Fouling by microorganisms, known as biofouling, is considered to start with phototrophic organisms (algae and cyanobacteria) conditioning the inert surfaces for subsequent growth of heterotrophic organisms (eg Grant 1982; OrtegaCalvo et al. 1993; Saiz-Jimenez and Arin ˜o 1995; Prieto et al. 2005; Miller et al. 2010). However in the absence of a primary colonizing film of phototrophs, heterotrophic bacteria and fungi can grow on building fac¸ades, using organic compounds from organic pollutants, painted surfaces or conservation treatments (eg Cappitelli et al. 2005, 2007). Unlike heterotrophic biofilms, phototrophic biofilms have received little attention until recently (Di Pippo et al. 2009, 2011). Biofilms that grow in areas exposed to light tend to contain photosynthetic organisms (Ramı´rez et al. 2010). The presence of these pioneer photosyntheticbased microbial communities generally result in the appearance of thin green films, which usually adhere to the stone substratum (ICOMOS-ISCS 2008) in a phenomenon often referred to as greening. The presence of greening depends on a variety of factors, such as a suitable combination of dampness, warmth and light on the stone surface (Tiano 2002) as well as the intrinsic characteristics of stone, such as permeability, porosity and surface roughness, which influence its bioreceptivity (Guillitte 1995; Prieto and Silva 2005; Miller et al. 2006). Factors such as the exposure site (Barberousse et al. 2006), the presence or absence of adjacent vegetation (Smith et al. 2011) and the urban or rural location (Tanaca et al. 2011) also affect the development of biofouling. The study of greening on building fac¸ades deserves considerable attention because pollution-related soiling has been widely superseded by algal growth. Since the beginning of the twentieth century, when the first data on the aspect of some important buildings were recorded, there has been an apparent decrease in the occurrence of black deposits (Newby et al. 1991; Davidson et al. 2000; Brimblecombe and Grossi 2009), partly because of the decrease in air pollution resulting from the declining use of coal and the switch to gas and electrical heating (Brimblecombe 1987; Grossi and Brimblecombe 2008). At the same time, there has been an increase in biological activity, favored by lower concentrations of sulfur dioxide and greater deposition of organic compounds and nitrogen (Grossi and Brimblecombe 2008). Higher humidity caused by climate change also favors the development of greening *Corresponding author. Email: [email protected] Biofouling Vol. 28, No. 3, March 2012, 329–338 ISSN 0892-7014 print/ISSN 1029-2454 online Ó2012 Taylor & Francis http://dx.doi.org/10.1080/08927014.2012.673220 http://www.tandfonline.com as it increases the time that stone structures remain wet and possibly the depth of penetration of moisture. This has already been observed on buildings in places such as Northern Ireland and London, as respectively, an increased incidence of algal greening (Smith et al. 2011) and a decreased incidence of blackening (Brimblecombe and Grossi 2009). Several authors have demonstrated that the discoloration of stone caused by the growth of microorganisms can be easily determined by a non-invasive (non-destructive) technique based on measurement of the reflectance by a spectrophotometer or a tristimulus colorimeter. The data are expressed in CIE-L*a*b* color system units (Wyszecki and Stiles 1982; Sanmartı´n et al. 2010). Using this technique, Urzı`and Realini (1998) correlated the orange or grey color of the patina on Noto’s calcareous sandstone with the associated microflora, Prieto et al. (2005) quantified the development of biofilm induced on the open rock faces of quartz quarries to reduce the visual impact, De Muynck et al. (2009) evaluated strategies for preventing algal fouling on two types of concrete (man-made stone), and more recently Tanaca et al. (2011) evaluated fungal colonization on three fiber cement (man-made stone) formulations exposed to urban, rural and coastal environments. Nevertheless, to the authors’ knowledge, the color measurement technique has not previously been used to study the phototrophic colonization (greening) of granite stone buildings naturally exposed to the outdoor environment. Apart from the chemical and/or physical deterioration of buildings caused by phototrophic-based microbial communities, greening must be considered as aesthetic damage that depends not only on the general conditions of the local environment, but also on the individual perception by the people involved (Smith et al. 2011). In this respect, green is more easily perceived than other colors, since the differences detected by the human eye are not of the same magnitude in the different parts of the spectrum; wavelengths close to 400 (blues) and 700 nm (reds) are less important from the point of view of perception than those around 500 nm. Five hundred and sixty nm (corresponding to the green area) is the wavelength at which the human eye is most sensitive (Gescheider 1976; McDonald 1997). Green has been used for the quantification of biomass of the green alga Ulva (syn. Enteromorpha) on test panels coated with a range of experimental formulations (Casse ´et al. 2007). Knowledge of the threshold of perception of greening by the human eye would be very useful for the effective and sustainable management of stone buildings and monuments. To date, few studies have addressed the human perception of changes in the appearance of stone due to general fouling, and most of the existing studies have investigated the changes in appearance in terms of blackening or darkening (Newby et al. 1991; Andrew 1992; Grossi and Brimblecombe 2004; Brimblecombe and Grossi 2005). As far as the authors are aware, only one laboratory-based study has examined the limits of perception of the phototrophic colonization on building stones (Prieto et al. 2006). In the latter study the qualitative terms, ranging from inappreciable to very intense, referred to the change in appearance caused by the live microorganisms on the surface of granite rocks were related to the total color change (DE* ab ). However, variations in the three color parameters, lightness-darkness DL*, redness-greenness Da* and yellowness-blueness Db*, were not analyzed. In the present study, these variations were taken into account with the aim of selecting the parameter that provides most information regarding the detection and monitoring of greening on granite fac¸ades. Thus, the overall aim of the study was to demonstrate the suitability of a spectrophotometric measurement technique for detecting phototrophic colonization (greening) and real-time monitoring of the development of the colonization in a real case of a granite building. The specific purposes of the study were as follows: (1) to establish the threshold of human perception of greening on granite rock and its value in terms of partial color differences (DL*, Da* and Db*), amount of phototrophic biomass (dry weight) and extracted chlorophyll a(chl a) content; (2) to analyze the changes in the L*, a* and b* CIELAB coordinates in a real case involving the cleaning and recolonization of a granite fac¸ade, with the aim of determining the parameter or parameters that best indicate the start of the greening process; and (3) to describe a methodology that could be used as a tool for making decisions about the required frequency of cleaning and biocide treatment in a real case. Materials and methods The perception of color depends on the incident light and variations in illumination, thus changes in the natural light affect the judgment of the observer about the color observed. For this reason, a visual sorting task used to estimate the threshold of perception of greening by the human eye was carried out in a laboratory test, under standard conditions of illumination. Detection and monitoring of greening were performed outside, since measurements made with a portable spectrophotometer are not affected by external light parameters. Psycho-physical and monitoring experiments were carried out with granite rocks of similar color characteristics and mineralogical nature in order to obtain comparable results. 330 P. Sanmartı´net al. Psycho-physical experiment A visual sorting task (psycho-physical experiment) was conducted to investigate the threshold of perception of greening on granite surfaces. Blanco Cristal, a mediumgrain, heterogranular-panallotriomorphic, biotitic adamellitic leucogranite (with dark minerals absent), and feldspar-K, plagioclases, quartz, biotite, chlorite and moscovite as major minerals was selected for the experiments. The petrographic characteristics and mineral composition of this lithotype were described in a previous study (Sanmartı´n et al. 2011). Twenty-one blocks (6 6361 cm) were cut and sterilized before starting the experiments. The upper surface of each granite block (18 cm 2 ) was inoculated in a laminar flow cabinet with a mixed culture of three isolates of subaerial stone biofilm-forming cyanobacteria grown in BG11 0 medium (Rippka et al. 1979), viz. Nostoc sp. PCC 9025, Nostoc sp. PCC 9104 and Scytonema sp. CCC 9801. The mixed inoculum consisted of 0.21 mg (dry weight) of each strain per ml of medium giving 0.63 mg (dry weight) per ml of mixed inoculum. Distinct volumes from 150 ml(the minimum volume required to fully cover the total area of 18 cm 2 of the surface block) to 540 ml mixed cyanobacterial suspension, were inoculated uniformly with the point of a pipette onto the surface blocks, in order to adjust phototrophic biomass between 0 (to exclude the possibility of an abiotic contribution to the green color) and 18.9 mg of biomass (dry weight) cm 72 of surface area. Experiments were performed in triplicate. Each inoculated block was assessed by eight different observers (five females and three males aged from 25 to 60 years) with normal color vision, ie normal trichromat observers without color blindness (see Fletcher and Voke (1985) for details of color vision examination). The task of each observer was to decide if the greening due to the presence of organisms was perceptible. The responses reported by the observers (yes/no) were recorded and analyzed. The point at which the observers began to note the green color has been coined as the just noticeable difference (jnd). The conceptualization of the greening threshold has its roots in the study of thresholds for other sensory-related stimuli (Gescheider 1976). The color of the upper surface of each granite block was measured before and after inoculation, following the methods proposed by Prieto et al. (2010a, 2010b). The CIELAB coordinates (L*, a* and b*) were measured before and after inoculation, with a portable spectrophotometer (Konica Minolta CM-700d/600d) equipped with CM-S100w (SpectraMagicTM NX) software; the measuring conditions were illuminant D65, observer 28and a 8-mm diameter viewing area. A total of 14 readings were taken at different randomly selected zones on each wet surface block, and the results expressed as the mean values. To relate the phototrophic biomass (dry weight) to color and extracted chl acontent, the concentration of chl aon each block was determined after the visual evaluation and color measurements, following the protocol for the extraction of chl afrom microorganisms colonizing rocks, recently proposed by Ferna ´ndez-Silva et al. (2011). The extracts were measured in a UV–Visible spectrophotometer (UVIKON XS, BioTek), and the equation proposed by Wellburn (1994) was used to calculate the concentration of chl a, expressed as mg chl acm 72 of surface area. In addition, the thresholds between imperceptible and perceptible greening, in terms of partial color differences (DL*, Da* and Db*), amount of phototrophic biomass (dry weight) and extracted chl a content, were estimated (Prieto et al. 2006). For this purpose, a data table was built in which for each sample (1–21), including its measured parameters, the eight answers from observers were included (1–8) (Berns 2000). The whole of cases (1–1, 1–2, (. . .), 1–8, 2–1, 2–2, (. . .), 21–8; ie sample-observer) were sorted into two sets, imperceptible and perceptible, and cumulative percentages were calculted for each group using the following equations where n p represents the number of answers that were ‘yes, ie perceptible, n i represents the number of answers that were ‘no, ie imperceptible’, and z represents answer 1, 2, . . . , n p or n i : Cumulativeperceptible;z¼100 z=np  ; Cumulativeimperceptible;z¼100 100 z=ni ðÞ The greening thresholds on granite were derived by plotting the cumulative percentages, cumulative frequency of answers for each qualitative term (imperceptible and perceptible) vs ordered partial color differences, phototrophic biomass (dry weight) and extracted chl acontent. The intersection of the two sets of data defines the jnd in greening. Monitoring experiment The progress of recolonization of the granite fac¸ade of the Supercomputing Centre of Galicia (CESGA, www.cesga.es), a centre par excellence in research and high-performance computing services built in 1993 in Santiago de Compostela (Galicia, NW Spain), was monitored fortnightly for a period of 10 months (289 days, March/2009–January/2010), using a spectrophotometer for instrumental color measurements. Three target areas (1, 2, 3) of the south facing wall, each of which was divided in two subareas, upper (A) and lower (B), were selected and cleaned mechanically with distilled water and a brush to remove the existing phototrophic colonization Biofouling 331 (Figure 1). The six study areas (1A, 1B, 2A, 2B, 3A and 3B), each of 21.5 cm 631.0 cm, did not receive direct sunlight because of the proximity of the neighboring buildings. They are also very damp due to water run-off. The weather conditions in Santiago de Compostela during the experiment (according to data from the Galician regional meteorological office: MeteoGalicia, www.meteogalicia.es) are shown in Table 1. Following the methodology proposed by Prieto et al. (2010b), a total of 234 readings were taken fortnightly at random points in the six study areas. A portable reflection spectrophotometer (CE-XTH) equipped with OptiviewSilver/i QC Basic software, was used under the following conditions: illuminant D65; observer 28and a 10-mm diameter viewing area. Color measurements were analyzed by considering the CIELAB color system (CIE Publication 15-2, 1986), Table 1. Temperature, humidity, sunshine duration and daily global radiation during the exposure period (March 2009 to January 2010), recorded at the Santiago EOAS meteorological station (Santiago de Compostela) and captured by Meteogalicia. Date Average air temperature (8C) Dew point temperature (8C) Mean relative humidity (%) Rain (l m 72 ) Sunshine duration (h) Daily global radiation (10 KJ m 72 day 71 ) March 2009 11.1 4.5 68 30.4 246.6 1617 April 2009 10.3 6.2 79 106.2 177.6 1581 May 2009 14.5 8.9 73 55 241.6 1932 June 2009 17.2 12.9 78 106.6 218.4 1942 July 2009 17.1 13.1 80 105.2 211.4 1917 August 2009 18.8 14.4 78 24.5 236.9 1917 September 2009 18.0 13.5 78 13.5 248.6 1691 October 2009 16.2 13.4 86 189 149.0 935 November 2009 11.7 9.9 92 272.2 60.1 433 December 2009 8.1 5.9 89 424.5 96.6 434 January 2010 7.6 5.6 91 200.2 85.4 489 Figure 1. Granite fac¸ade of the Supercomputing Centre of Galicia (CESGA), showing extensive phototrophic colonization (greening) developed as streaks following the path of water run-off. (a) On site measurement of the color on the fac¸ade; (b) overview of the fac¸ade with perceptible colored stains (greening) in humid zones; (c) location of the six study areas before cleaning; (d) location of the six study areas after cleaning. Six study areas: three target areas (1, 2, 3) each of which was divided in two subareas, upper (A) and lower (B). 332 P. Sanmartı´net al. taking into account the mean values of CIELAB coordinates (L*, a* and b*). Variations in the three color parameters (partial color differences), lightnessdarkness DL*, redness-greenness Da* and yellownessblueness Db*, were calculated as follows: DL¼L tL 0; Da¼a ta 0; Db¼b tb 0; where L 0;a 0;b 0¼the mean value of CIELAB color parameter at the study area after cleaning, and L t;a t;b t¼the mean value of CIELAB color parameter at the study area after a period of exposure (t). Results and discussion The laboratory-induced epilithic cyanobacterial biofilms on granite blocks served to emulate the presence of initial green fouling. Photographs of the granite blocks inoculated for the psycho-physical experiment, along with the data for the amount of phototrophic biomass (dry weight) and the concentration of chl a extracted are shown in Table 2. On the basis of the visual evaluation, the threshold of human perception between imperceptible and perceptible greening in terms of the partial color differences (DL*, Da* and Db*), biomass (dry weight) and extracted chl acontent, were calculated by plotting the cumulative frequency of answers given by each observer (Figure 2). The visual perception of the color difference differed with each observer, and the point at which individuals began to note the greening on the surface due to phototrophic biomass also differed among the different observers. From the results summarized in Figure 2, it was concluded that greening became perceptible when colonization resulted in a decrease of 4.37 CIELAB units in parameter L*, or when the values of parameters a* and b* decreased and increased, respectively, by only 0.41 and 0.59 CIELAB units. In terms of mg biomass (dry weight) and extracted chl acontent, a lower abundance of phototrophic biomass was required to reach the greening threshold or jnd in greening in terms of yellowness-blueness variations, Db* (6.3 mg biomass dry weight cm 72 and (8.43 +0.24) 610 73 mg extracted chl acm 72 ) and a larger quantity was required for the other two color parameters, lightness-darkness, DL*, and redness-greenness, Da* (12.6 mg biomass dry weight cm 72 and (20.33 +1.60) 610 73 mg extracted chl acm 72 ). This indicates that of the CIELAB color parameters, parameter b* provides the earliest indication of the presence of phototrophic microorganisms. The thresholds calculated for the partial color differences were used in the outdoor experiment, in Table 2. Blanco Cristal granite specimens inoculated with mixed cyanobacterial culture used in the psycho-physical experiment. Photograph of the granite block surface Phototrophic biomass (mg dry weight cm 72 ) 0.0 0.4 0.7 1.1 1.4 2.1 3.2 6.3 12.6 18.9 Chlorophyll a extracted (*) (10 73 mgcm 72 ) 0.0 1.65+0.12 2.49+0.21 4.04+0.08 4.49+0.02 5.74+0.02 4.95+0.91 8.43+0.24 20.33+1.60 31.50+2.47 Note: Scale bar ¼3.5 cm. *Mean value of three replicates+SD. Biofouling 333 which the rate of greening on building granite was monitored for 10 months after cleaning, by changes in L*, a* and b*. The results are shown in Figure 3 as a summary of the variations in L*, a* and b* values in the six study areas after the exposure period; the time at which the observers began to note the greening is indicated by a shaded circle in the figure. This corresponds to DL* 574.37; Da* 570.41 and Db* 4þ0.59 CIELAB units. Thus, the first bar of the histograms (BC) represents the color difference before and after cleaning the colonized surface, and was used as a reference. Subsequent bars reflect the development of the recolonization process, ie the difference between the value on the first day after cleaning and the subsequent values. In all cases, the trends were all in the same direction as the value of the difference between the colonized and the clean rock (BC) (Figure 3), thus indicating that the color change was due to recolonization. Thus, during the recolonization process a decrease in the DL* and Da* values and an increase in the Db* value were observed, indicating that the fac¸ade became darker and more yellowgreenish color. One important aspect of the results obtained is that analysis of the Da* and Db* parameters enabled identification of the point at which recolonization began, as in both cases the difference (D) remained constant for a certain period of time and then increased gradually. The moment at which the difference (D) began to increase also coincided with the moment at which the value of the increase changed direction towards the reference value (BC), thus indicating the start of colonization. Analysis of the values in Figure 3 shows that recolonization began at between 80 and 129 days (marked with an arrow in the Figure 3). However, analysis of the DL* values was not as useful for identifying the moment at which recolonization began as the variations were more erratic (Figure 3). Comparison of the day on which recolonization began, determined by analysis of the trends (direction changes towards reference value) in the Da* and Db* values, and the moment at which greening is considered perceptible in terms of the Da* and Db* values (both indicated in Figure 3), revealed that in all cases, the earliest detection of recolonization was by analysis of the trends. Thus monitoring the color of granite fac¸ades using a portable spectrophotometer enables early detection of the appearance of greening, even Figure 2. Results of the psycho-physical experiment. Top: lightness-darkness DL*, middle: redness-greenness Da*, and bottom: yellowness-blueness Db*. Partial color differences CIELAB units (on the bottom x-axis) and mg biomass (dry weight) cm 72 surface area and mg extracted chl acm 72 surface area (in italics) (on the top x-axis) plotted against the cumulative frequency of answers for perceptible and imperceptible qualitative term (%). The values of the greening threshold in terms of these parameters are shown in a box in the figures. 334 P. Sanmartı´net al. Figure 3. Results of the monitoring experiment. Color changes (top: DL*, middle: Da* and bottom: Db*) during the period of exposure (289 days) at the six study areas (1A, 1B, 2A, 2B, 3A and 3B, see Figure 1). DL* ¼L* t –L* 0 ;Da* ¼a* t – a* 0 ;Db* ¼b* t –b* 0 , where L* 0 ,a* 0 ,b* 0 ¼mean value of CIELAB color parameter at the study area after cleaning with water and a brush, and L* t ,a* t ,b* t ¼mean value of CIELAB color parameter at the study area after a period of exposure (t). For comparison, the color change before and after cleaning is shown as BC (Before Cleaning). The change in the direction of the partial color difference is indicated by an arrow, and the time when colonization becomes perceptible by a shaded circle (DL* 574.37; Da* 570.41 and Db* 4þ0.59 CIELAB units). Biofouling 335