Response of Tuber melanosporum fruiting to canopy opening in a PinusQuercus forest

Document downloaded from: This paper must be cited as: The final publication is available at Copyright Additional Information http://doi.org/10.1016/j.ecoleng.2012.12.006 http://hdl.handle.net/10251/81283 Elsevier Garcia Barreda, S.; Reyna Domenech, S. (2013). Response of Tuber melanosporum fruiting to canopy opening in a PinusQuercus forest. Ecological Engineering. 53:54-60. doi:10.1016/j.ecoleng.2012.12.006. 1 This is the authors’ accepted manuscript of the article: Garcia-Barreda S, Reyna S, 2013. Response of Tuber melanosporum fruiting to canopy opening in a Pinus-Quercus forest. Ecological Engineering 53: 54-60 The final publication is available at: http://www.sciencedirect.com/science/article/pii/S0925857412003606 2 Response of Tuber melanosporum fruiting to canopy opening in a Pinus-Quercus forest Sergi Garcia-Barredaa, Santiago Reynaa,b a Fundación Centro de Estudios Ambientales del Mediterráneo, C/ Charles Darwin 14 Parque Tecnológico, 46980 Paterna, Spain. b ETS Ingeniería Agronómica y del Medio Natural, Universidad Politécnica de Valencia, Camino de Vera s/n, 46021 Valencia, Spain. Abstract The wild production of the highly appreciated fungus Tuber melanosporum is negatively affected by canopy closure in the stand. Habitat improvement has been proposed as a tool to recover the production in close forests, but evaluations based on scientific monitoring are still lacking. This study analyses the short-term effect of a pilot project on improvement of T. melanosporum reproduction habitat. The results support the project hypothesis that the canopy closure was hampering truffle fruiting in the larger brûlés. The silvicultural treatment alone has not triggered a clear positive response in all the truffières, suggesting that complementary actions are necessary to ensure their sustainability. Weather conditions provoke a year-to-year variation in the fruiting and determine the responsiveness of the truffières to the treatment. Keywords: Tuber melanosporum; non-wood forest products; canopy closure; Quercus ilex; silviculture 1. Introduction Wild edible mushrooms are an important forest product worldwide (Boa, 2004). Communities of ectomycorrhizal fungi are frequently linked to particular forest types and structures (Pilz et al., 1996). Detailed studies about the habitat requirements of these fungi are crucial to developing ecosystem management criteria aimed to enhance fungal diversity and productivity, and ultimately supporting the sustainability of ecosystems. The European black truffle (Tuber melanosporum Vitt.) is a commercially harvested mushroom responsible for a multi-million euro industry in France, Italy and Spain (Reyna and Garcia-Barreda, 2009). T. melanosporum is ectomycorrhizal and grows wild in open forests of Quercus (Olivier et al., 2002; Reyna et al., 2004). The ground where it fruits is affected by the phytotoxic activity of the fungus (Splivallo, 2008), and this causes the appearance of the so-called brûlé (truffle burn). Fruiting is typically associated to particular host trees and brûlés from year to year (the so-called truffières). The sporocarp production of a truffière has been found related to the characteristics of the brûlé and the host tree (Sourzat et al., 2004; Oliach et al., 2005; Garcia-Montero et al., 2007b), although highly variable from year to year. This variability has been found closely linked to weather conditions (Ricard et al., 2003). In Spain, wild T. melanosporum production has suffered an important decline in the last decades, due to both natural and human causes (Reyna and Garcia-Barreda, 2009). This 3 occurs in the framework of Mediterranean forests, with low economic profitability (Domínguez-Torres and Plana, 2002) and the loss of involvement of the rural communities in the protection and management of forests. Some experts propose to improve T. melanosporum reproduction habitat by opening the canopy around the truffière (Reyna et al., 2004; Diette and Lauriac, 2005), with the ultimate aims of conserving local wild populations of the fungus (and their gene pool), and enhancing ecosystem provisioning services. But for the moment the application of this technique in wild truffières has not been evaluated with large-sample studies. The full development of rehabilitation techniques requires a periodic evaluation of the success and the eventual adjustment of the management criteria (Vallauri et al., 2005). In one of these habitat enhancement projects, in El Toro (Spain), an evaluation programme involving the monitoring of the fruiting, the ectomycorrhizal community and the tree vegetation was designed (Reyna et al., 2004). This paper analyses the short-term (9 years) advances in the process of improvement of T. melanosporum reproduction habitat in El Toro pilot site. The effectiveness of the project is evaluated through two criteria: (1) the influence of the canopy cover on the response of the treated truffières, and (2) the time trend of fruiting after treatment. The former is used to test the project hypothesis that canopy closure hampered T. melanosporum fruiting. The latter is used as a criterion of self-sustainability: following Lugo (1992), it is considered that a truffière with a positive trend is approaching the recovery goal. A secondary aim is to analyse if the technique has succeeded in all the range of truffière attributes in which it has been applied, with the purpose of “fine-tuning” the technique. 2. Materials and methods 2.1. Study site and habitat improvement strategy El Toro pilot site is located in the Valencian Community, eastern Spain (39º 58’ to 40º 2’ N, 0º 44’ to 0º 47’ W, 990-1110 m a.s.l.). The climate is Continental Mediterranean, with a mean annual rainfall of 500-550 mm and a mean annual temperature of 11.9-12.4ºC. The soils are calcixerepts developed on a calcareous glacis with less than 5% slope. Dominating texture is sandy clay loam, pH ranges from 7.7 to 8.3, and organic matter content ranges from 2 to 3%. Until 1958, patches of Q. ilex L - Quercus faginea Lam. coppices and cereal cultures dominated the landscape. In the latter, isolated Quercus were also present in field boundaries. About 1955, the local population began to harvest truffles. Since then, T. melanosporum is intensively exploited as a common-pool resource and no management practices have been carried out. From 1958 to 1969, the forest administration reforested the site with Pinus nigra 4 Arnold. By the 1990s, the reforestation had formed a close canopy (density: 900-2500 trees ha-1, canopy cover: 60-95%) and the Quercus survived as suppressed trees. In 1997 the forest administration initiated a programme to support the recovery of T. melanosporum reproduction habitat and the economic revenues provided by wild truffières to the local community. El Toro was selected as pilot site: with the aid of local harvesters, the truffières that still remained productive were located and measured. According to harvesters, T. melanosporum production had sharply decreased from the 1970s, and the major causes of degradation were the canopy closure around the truffières (mainly due to the pine plantation) and the intensive harvesting. Consequently, a silvicultural model that adapted the forest structure to the habitat requirements of T. melanosporum was developed and implemented from 2000 to 2001 (Reyna et al., 2004). The pines within a radius of 14-50 m from the brûlé (depending on the height of the host tree and the surrounding pines) were systematically cut down, since Pinus species do not encourage T. melanosporum fruiting (García-Montero et al., 2007a). All shrubs within the radius were removed, whereas Quercus were preserved and slightly pruned (less than 10% of the live crown was removed). This resulted in the opening of circular gaps with reduced canopy cover around the brûlé (Table 1). During the silvicultural operations, the brûlés were completely encircled with marking tape, in such a way that trampling, tractor passage and incorporation of logging residues were avoided. 2.2. Data collection and statistical analysis The 74 treated truffières selected for the monitoring meet the following requirements: (1) they produced sporocarps of T. melanosporum until (at least) the fruiting season 1997-1998, (2) they do not produce sporocarps of other Tuber species, (3) before being treated, they were located inside the pine plantation, and (4) the canopy cover within a radius of 15 m from the brûlé was reduced below 50% by the treatment. Other eleven brûlés are located in nonreforested plots. They have not been considered treated truffières but reference areas, and they have also been intensively exploited. Ten truffières located inside the pine plantation were either not treated or the canopy cover (within the radius of 15 m) was not reduced below 50%; they have been considered as non-treated. Reference and non-treated areas have not been statistically compared to treated truffières, because they are scarce and the range of truffière attributes is narrower. Instead, the effect of truffière attributes on reference and non-treated areas has been analysed independently. The exploitation as a common-pool resource has made it impossible to measure sporocarp number or biomass, which would be a direct measure of the provisioning services delivered 5 by the truffières. Although not ideal, we have measured instead the number of digs made by harvesters as an estimator of the occurrence of successful fruiting events. The occurrence of fruiting events is relevant in biophysical terms because it can reflect the annual suitability of environmental conditions for a species to reproduce sexually and therefore site functionality as a reproduction habitat. Thus, this approach focuses on the occurrence of successful fruiting events and not on sporocarp productivity, which could be influenced by variations in either the number of fruiting events or in sporocarp size. In the study area raking the brûlés is forbidden, sporocarps must be located with trained dogs, which only mark mature sporocarps, and there were few reports of poaching, so that digs in which no sporocarp was extracted were likely very scarce. Truffle sporocarps are harvested with a small spade, and that makes the diggings clearly different from animal excavations. In a previous fruiting season, sporocarp biomass was monitored in eight truffières, and a mean of 56 g per dig (standard deviation: 16) was obtained. In one of these truffières, sporocarp biomass was monitored during four fruiting seasons, with a mean of 55 g per dig (standard deviation: 30). The abundance of harvester digs has been monitored from the fruiting season 2001-2002 to 2008-2009 (i. e. until 9 years after treatment) to evaluate the temporal trend. The truffières have been repeatedly visited during each fruiting season and the location of digs of the same year has been recorded to distinguish fresh digs from the old ones. The digs from previous seasons are easy to differentiate thanks to the soil texture in the study site and to the low rainfall and snowfall during the fruiting season. A linear mixed model (LMM) is used to test the null hypothesis of no temporal trend in the number of harvester digs after treatment, and a Toeplitz covariance structure has been specified (SPSS, 2006). Data from the first season have been excluded because of the excessive number of zeros. Time from treatment is treated as repeated measures variable, and both linear and quadratic components have been tested. An estimate of the weather conditions has been included as a covariate, to deal with interannual variability. Given that no meteorological station existed in the study site, the width of latewood rings in dominant P. nigra (calculated as the mean of 20 trees) has been chosen. Lebourgeois (2000) and Martín-Benito et al. (2008) found that the width of P. nigra latewood related to abundant rainfall and mild temperatures during late spring and summer, and this is the period with the greatest incidence on T. melanosporum fruiting (Ricard et al., 2003). In our study site, latewood width shows a positive correlation with august rainfall estimated through interpolation from the four nearest meteorological stations, located 5-12 km away (Pearson’s 6 r=0.66, P=0.019, n=12), and a negative correlation with july mean temperature in the only nearby thermometric station, located 10 km away (Pearson’s r=-0.64, P=0.036, n=12). The main attributes of the truffière (measured in 1998) are included as fixed predictors in the LMM: abundance of harvester digs in the season 1997-1998 (as an indicator of the frequency of fruiting events before treatment), surface of the brûlé, canopy surface of the host tree, and surface stoniness (percentage of the brûlé surface covered by stones). Truffle production in the season 1997-1998 was extraordinary both in the study site and nationwide (not repeated since), and therefore dig abundance in that season can be considered an indicator of the fruiting potential of the truffières before treatment. The LMM also includes the characteristics of the tree vegetation around the brûlé before (1997) and after (2005) treatment: canopy cover within a radius of 15 m from the brûlé centre, and distance from the brûlé center to the untouched dense pine plantation. Model adequacy has been assessed using Akaike’s information criterion, and non-significant factors and interactions have been removed from the final model. The response variable has been log transformed to more closely meet the assumptions of normal distribution and constant variance. The interaction effects are investigated using the “pick-a-point” approach for testing simple slopes (Bauer and Curran, 2005). 3. Results The abundance of digs in the treated truffières is significantly related to several interactions involving time elapsed from treatment, width of latewood rings, brûlé surface, abundance of digs in season 1997-1998 and variation in canopy cover from 1997 to 2005 (Table 2). The quadratic component of time, the canopy surface of the host tree, the stoniness, the canopy cover in 2005 and the distance to the pine plantation have not been included in the final model because they do not show a significant effect and do not improve the adequacy of the model. The post-treatment temporal trend of the dig abundance is significant but determined by the value of annual latewood growth (Table 2, Fig. 1). According to the test of simple slopes, the regression between dig abundance and time from treatment is negative and significant at values of latewood growth less than 1.1 (latewood growth is expressed as a ratio to the mean value in the period 1997-2008). Six of the eight years of monitoring showed lower scores. The simple slope is positive and significant at values of latewood growth more than 1.4 (two of the eight years showed a higher score). From the other point of view, when time is considered the moderating variable, the test of simple slopes shows that the regression 7 between dig abundance and annual latewood growth is significantly positive only from the fourth year (P<0.001), whereas it was not significant previously. The interaction between annual latewood growth and dig abundance before treatment significantly affects post-treatment dig abundance (Table 2). According to the test of simple slopes, the regression between post-treatment dig abundance and latewood growth is significant and positive at values of pre-treatment dig abundance higher than 1 (71 of the truffières are above this threshold). The slope of this simple regression increases with increasing values of pre-treatment dig abundance (Fig. 2). The interaction between pre-treatment dig abundance and brûlé surface significantly influences post-treatment dig abundance (Table 2). According to the test of simple slopes, the regression between post-treatment dig abundance and brûlé surface is significant and positive at values of pre-treatment dig abundance higher than 10 (38 of the truffières are above this threshold). The simple slope is significant and negative at values of pre-treatment dig abundance lower than 3 (4 of the truffières are below this threshold). The slope of this simple regression increases with increasing values of pre-treatment dig abundance (Fig. 3). The interaction between brûlé surface and variation in canopy cover from 1997 to 2005 significantly influences post-treatment dig abundance (Table 2, Fig. 4). According to the test of simple slopes, the regression between post-treatment dig abundance and brûlé surface is significant and positive at values of canopy cover reduction higher than 42% (39 of the truffières are above this threshold). The simple slope is significant and negative at values of canopy cover reduction lower than 18% (7 of the truffières are below this threshold). The observed number of digs in each truffière has been compared to the values inferred by the LMM for the case that canopy cover had not been reduced. The model predicts that in the larger brûlés (surface>77 m2) the mean number of digs would have decreased by 66% if no treatment had been applied, whereas in the smaller brûlés it would have increased by 58%. In the non-treated truffières, the abundance of digs does not show a significant time trend, does not relate to latewood growth and is not significantly affected by any other of the variables evaluated. Its global mean is not significantly different from zero, and 50% of the truffières did not show any dig during the monitoring period. In the reference areas, dig abundance shows a positive and significant relationship with latewood growth (P<0.001) and a positive and marginally significant relationship with canopy cover in 2005 (P=0.06). No significant time trend has been observed. 8 4. Discussion 4.1. Influence of the canopy cover In large brûlés the intensity of canopy cover reduction shows a direct relationship with frequency of fruiting events after treatment (Fig. 4), indicating that the initial levels of canopy closure in the study site were able to influence fruiting success of the fungus. Truffle fruiting in large brûlés appears to have been suppressed before treatment and to be released by the treatment, and according to the LMM if the treatment had not been executed the frequency of fruiting events would have decreased in these truffières (the large brûlés account for 57% of the digs in season 1997-1998 and for 64% from 2002 to 2009). The suppressed fruiting is supported by the fact that treated truffières have responded to propitious weather conditions only from the fourth year after the silvicultural treatment. The dependence of T. melanosporum fruiting on annual weather is widely accepted (Ricard et al., 2003). The effect of weather has been greater in the truffières showing more digs before treatment. Accordingly, the relationship between weather and dig abundance has been significant in the reference areas but not in the non-treated truffières. The truffière responsiveness to weather appears related to its productivity, and we therefore consider that the increase in responsiveness after treatment is an improvement in the functioning of the truffières. Canopy opening eliminates pine roots, reduces leaf litter accumulation and increases the insolation of the soil surface. López et al. (2003) found that thinning a Q. ilex forest increased soil water and temperature. These environmental conditions have been associated with T. melanosporum reproduction habitats (Olivier et al., 2002; Reyna et al., 2004; García-Montero et al., 2007a). Ricard et al. (2003) found that coarse fraction organic matter decreases in T. melanosporum brûlés. García-Montero et al. (2007b) pointed that solubilisation of calcium carbonate is also characteristic in T. melanosporum brûlés and that it is related to aeration and water flow through the soil. 4.2. Time trend after treatment The time trend of dig abundance in the treated truffières has not been clearly positive in the short term (years 2-9), suggesting that the occurrence of truffle fruiting events is not selfsustainable at mean weather conditions under the current management. Similarly, in the reference areas (which are also subject to intensive harvesting) no significant time trend has been found. However, in both cases the truffières are weather-responsive. Neither the canopy cover after treatment nor the distance to the pine plantation influence dig abundance in the 15 0 1 2 3 4 5 6 7 8 9 050 100 150 200 250 300 dig abundance (post-treatment) brûlésurface (m2) No. of digs pre-treatment = 4 No. of digs pre-treatment = 15 No. of digs pre-treatment = 26 Fig. 3. Predicted effect of the interaction between brûlé surface and pre-treatment dig abundance on posttreatment dig abundance (mean annual abundance from year 2 to 9). The values of pre-treatment dig abundance (in season 1997-1998) depicted are the mean in the treated truffières and the mean minus and plus one standard deviation. The remaining variables are fixed at mean values. 0 1 2 3 4 5 050 100 150 200 250 300 dig abundance (post-treatment) brûlésurface (m2) Canopy cover reduction = 24% Canopy cover reduction = 43% Canopy cover reduction = 62% Fig. 4. Predicted effect of the interaction between brûlé surface and canopy cover reduction on post-treatment dig abundance (mean annual abundance from year 2 to 9). The values of canopy cover reduction (from 1997 to 2005) depicted are the mean in the treated truffières and the mean minus and plus one standard deviation. The remaining variables are fixed at mean values. View publication statsView publication stats