Niche position drives interspecific variation in occupancy and abundance in a highly-connected lake system

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY-NC-ND 4.0 https://creativecommons.org/licenses/by-nc-nd/4.0/ Niche position drives interspecific variation in occupancy and abundance in a highlyconnected lake system © 2018 Elsevier Ltd. Accepted version (Final draft) Vilmi, Annika; Tolonen, Kimmo; Karjalainen, Satu Maaria; Heino, Jani Vilmi, A., Tolonen, K., Karjalainen, S. M., & Heino, J. (2019). Niche position drives interspecific variation in occupancy and abundance in a highly-connected lake system. Ecological Indicators, 99, 159-166. https://doi.org/10.1016/j.ecolind.2018.12.029 2019 Accepted to Ecological Indicators 1 2 Niche position drives interspecific variation in occupancy and abundance in a highly3 connected lake system 4 5 Annika Vilmi*,1,2, Kimmo T. Tolonen3, Satu Maaria Karjalainen2, Jani Heino4 6 7 1State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and 8 Limnology, Chinese Academy of Sciences, 18 North Linshan Road, Qixia, Nanjing 210046, 9 China 10 2Freshwater Centre, Finnish Environment Institute 11 3Department of Biological and Environmental Science, University of Jyväskylä, Finland 12 4Biodiversity Centre, Finnish Environment Institute 13 14 *Corresponding author, e-mail: [email protected], phone: +86 18351006316 15 16 Author Contributions: AV and JH designed the study. AV and KT performed the species 17 identifications. AV and JH analyzed the data. AV, KT, SK and JH wrote the manuscript. 18 Abstract 19 We examined how niche position, niche breadth, biological traits and taxonomic relatedness 20 affect interspecific variation in occupancy and abundance of two commonly-used indicator 21 groups, i.e. diatoms and macroinvertebrates. We studied 327 diatom and 117 macroinvertebrate 22 species that occupied the littoral zones of a large (305 km2) highly-connected freshwater 23 system. We collated information on the biological traits and taxonomic relatedness of each 24 species. Using principal coordinates analysis, we formed biological trait and taxonomic vectors 25 describing distances between species and used the resulting vectors as predictor variables. As 26 environmental data, we had site-specific physico-chemical variables, which were used in 27 outlying mean index analyses to determine the niche position and niche breadth of a species. 28 We used linear models to study if and how these two niche parameters and biological traits as 29 well as taxonomic relatedness affected occupancy and abundance. We observed positive 30 occupancy-abundance relationships for both diatoms and macroinvertebrates. We further found 31 that, for both groups, occupancy was better explained by the predictor variables compared with 32 abundance. We also observed that niche parameters, especially niche position, were the main 33 determinants of variation in occupancy and abundance for both diatoms and 34 macroinvertebrates. Local abundances of diatom and macroinvertebrate species were also, to a 35 small degree, affected by biological traits or taxonomic relatedness. We further saw that the 36 relationship between niche position and occupancy was negative, indicating that the more 37 marginal the niche position, the rarer a species is. Our findings provide support for the use of 38 diatoms and macroinvertebrates as ecological indicators as their occupancies and abundances 39 were affected by niche parameters, which is not necessarily always clear in challenging study 40 systems with high connectivity (i.e. high movement of material and species) among sites. These 41 findings also suggest that indices using information on species’ occupancy, abundance and 42 niche requirements are useful in environmental assessment. 43 Keywords 44 Diatoms, macroinvertebrates, niche parameters, biological traits, taxonomic relationships 45 46 1. Introduction 47 Species occupancy, abundance and their relationships are amongst the most widely-studied 48 topics in macroecology and biogeography (Brown 1984; Gaston et al. 2000), and researchers 49 have described the occupancy-abundance patterns in a variety of ecosystems across the globe 50 (Venier and Fahrig 1996; Soininen and Heino 2005; Foggo et al. 2007; Roney et al. 2015; 51 Tonkin et al. 2016). The message has been relatively consistent: occupancy and abundance 52 tend to be positively and often strongly associated with each other. Species with low local 53 abundances tend to show limited distributions, whereas locally abundant species usually are 54 widespread in a region (Gaston and Blackburn 2000; Gaston et al. 2000). These ideas are 55 important for meaningful biodiversity conservation (Gaston et al. 2000), especially in an era 56 when human impacts are greater than before (Lewis and Maslin 2015; Waters et al. 2016). As 57 freshwaters are one of the most threatened ecosystems in the world (Heino et al. 2009; 58 Vörösmarty et al. 2010; Vilmi et al. 2017), they also deserve special attention in the context of 59 the interspecific occupancy-abundance relationship (Gaston et al. 2000; Heino and Tolonen 60 2018). 61 An increasing number of attempts have been made to detect various factors that account 62 for variation in species occupancy and abundance. In recent years, the niche breadth (Brown 63 1984) and niche position (Hanski et al. 1993; Venier and Fahrig 1996) hypotheses have been 64 employed to investigate how species’ niche characteristics (i.e. their relationships to the 65 environment) account for their regional occupancy and local abundance (Tales et al. 2004; 66 Faulks et al. 2015; Tonkin et al. 2016). Ultimately, species optima are located at different parts 67 of a continuum of environmental conditions (Fig. 1). Some species have the same niche 68 positions, while niche positions of other species are located at different parts of the continuum. 69 Some niches may be positioned at the very end of the environmental range, while other niches 70 may be positioned in average environmental conditions. Although species may have similar 71 niche positions, their niche breadths can strongly differ from each other at the same time. 72 Species with large niche breadths can tolerate a wide variety of environmental conditions, 73 while species with small niche breadths are very specialized to certain environmental 74 conditions (Fig. 1, Brown 1984; Venier and Fahrig 1996; Heino and de Mendoza 2016). 75 Previous research has provided strong evidence that the two niche parameters, niche 76 position and niche breadth, strongly affect occupancy and abundance of different types of 77 organisms (Passy 2012; Heino and Tolonen 2018). For instance, when studying stream 78 macroinvertebrates in a tropical region, Tonkin et al. (2016) found that the two niche 79 parameters explained well variation in occupancy, but it was not correlated with mean local 80 abundance. A study on fish species across boreal lakes suggested that intraspecific niche 81 variation and a positive abundance-occupancy relationship are connected to each other (Faulks 82 et al. 2015). Tales et al. (2004) found support for the niche position hypothesis as the largest 83 contributor to regional occupancy and local abundance in temperate river fishes. Recent studies 84 on subarctic stream diatoms and insects (Rocha et al. 2018) as well as boreal lake 85 macroinvertebrates (Heino and Tolonen 2018) have also shown that especially niche position 86 has strong effects on regional occupancy and local abundance. 87 Species characteristics other than niche parameters may also affect occupancy, 88 abundance and their relationships (Tales et al. 2004; Heino and de Mendoza 2016; Heino and 89 Tolonen 2018; Rocha et al. 2018). Body size measures and resource use features are typical 90 examples of species characteristics, i.e. biological traits (e.g. Passy 2012). The importance of 91 biological traits in affecting occupancy, abundance and their relationships has been reported, 92 for instance, for riverine fishes (Tales et al. 2004), for aquatic macroinvertebrates across lentic 93 waterbodies (Verberk et al. 2010) and for diatoms across streams (Rocha et al. 2018). In 94 addition to biological traits of species, also taxonomic relatedness as a proxy for evolutionary 95 aspects has been considered in the same context. Heino and Tolonen (2018) did not find strong 96 influences of taxonomic relatedness nor traits similarity on occupancy and abundance of 97 macroinvertebrates across a set of lakes, while the effects of niche parameters, particularly that 98 of niche position, were clear. 99 Previous research centering on the topic of occupancy and abundance and the factors 100 determining them have covered different types of freshwater ecosystems, ranging from across101 streams (Rocha et al. 2018) to across-lakes (Heino and Tolonen 2018) and to across102 waterbodies (Verberk et al. 2010) studies. However, there is a knowledge gap for information 103 from systems with high connectivity, such as large lake systems. We previously showed that 104 diatom and macroinvertebrate communities exhibited pure spatial patterns (e.g. dispersal105 related processes) and that local environmental conditions only comparatively little affected 106 community structures in such a highly-connected system (Vilmi et al. 2016; Tolonen et al. 107 2017). These findings were not in line with main assumptions of current bioassessment 108 methods that assume a high importance of local environment on species compositions (Heino 109 2013). The high connectivity of a system is thus a characteristic which can enhance the effects 110 of dispersal and other spatial processes (Foggo et al. 2007). Thus, in that sense, it is worth 111 studying how interspecific occupancy and abundance, the building blocks of ecological 112 indicators, are formed in these sorts of open, large freshwater systems. 113 Here, we investigated if and how niche parameters, biological traits and taxonomic 114 relatedness affect occupancy and abundance of freshwater diatoms and macroinvertebrates in 115 a large lake system which contains no apparent barriers for dispersal. For each species, we 116 calculated niche position and niche breadth (based on an extensive dataset of local 117 environmental variables) and determined biological traits and taxonomic relatedness. Using 118 species as data points, we asked the following questions: (1) What are the relationships between 119 occupancy and abundance of freshwater diatom and macroinvertebrate species within a large 120 lake system? (2) Which factors (niche parameters, traits, and taxonomy) best predict 121 occupancy, abundance and their relationship of diatom and macroinvertebrate species in such 122 a study system? (3) Are the findings similar for the two distinct groups of organisms? (4) Are 123 the findings similar to studies examining patterns across waterbodies? 124 125 2. Material and Methods 126 2.1. Field sampling and laboratory analyses 127 We used data on diatom and macroinvertebrate taxa to explore our research questions. The 128 biological data were collected in early autumn 2013 from a large (305 km2) lake system of 129 Lake Kitkajärvi. The originally oligotrophic lake system is located in north-eastern Finland. 130 Due to land use and inflow of purified municipal waste water, some parts of the lake system 131 have shown signs of eutrophication (e.g. Vilmi et al. 2015). We collected the diatom and 132 macroinvertebrate samples from 81 similar, stony littoral sites around the lake system (see map 133 of study area in Fig. A.1). In the laboratory, for diatoms, we identified approximately 500 134 valves from each site, and for macroinvertebrates, we identified all individuals that were 135 captured in a site-specific kick-net sampling. We identified the diatoms and macroinvertebrates 136 to the lowest taxonomic level possible, which was in most cases species level, although some 137 valves or individuals were assigned to genus level. Thus, from now on, we refer to the studied 138 taxa here as ‘species’. 139 We gathered a broad set of site-specific local environmental variables. In the field, we 140 visually assessed the particle sizes of the benthic substratum and measured the slope of bottom. 141 We also collected water samples, which were analyzed in the laboratory. We used fetch, 142 calculated with the Wind Fetch Model (Rohweder et al. 2008), as a proxy for wave disturbance 143 at each site. Further details on the sampling and laboratory methods are thoroughly presented 144 in our earlier publications on the effects of local environmental variables on diatom and 145 macroinvertebrate community structures (Vilmi et al. 2016; Tolonen et al. 2017). 146 In this study, we used the following variables as local environmental variables: electrical 147 conductivity, saturation of oxygen, suspended solids, slope, fetch, mean particle size, and 148 particle size diversity, as well as concentrations of aluminium, boron, manganese, NH4-N, 149 NO2+NO3-N, oxygen, PO4-P, silicon, sodium, soluble total nitrogen, soluble total phosphorus, 150 and zinc. These variables were not highly correlated with each other and showed considerable 151 among-site variation within the data. 152 153 2.2. Niche position and niche breadth 154 We first determined niche position (OMI values) and niche breadth (Tolerance values; Tales 155 et al. 2004) of each species using the outlying mean index (OMI) analysis (Dolédec et al. 2000). 156 The analysis basically measures the marginality of species habitat distributions by the distances 157 between mean environmental conditions used by a species and average environmental 158 conditions that are available among the study sites. The ecological interpretation of the OMI 159 and Tolerance values, which the analysis produces, is as follows: species with high OMI values 160 have marginal niches and species with low OMI values have non-marginal niches. Species with 161 high Tolerance values occur in a variety of environmental conditions thus having large niche 162 breadths. Species with low Tolerance values are present only in certain environmental 163 conditions and they have smaller niches. 164 For the OMI analysis, a site-by-species abundance data matrix is needed, as well as an 165 environmental variables data matrix. We made logarithmic transformations for the 166 environmental variables and standardized them. Before proceeding to the OMI analyses, we 167 excluded species that were present only at one site. After doing so, we had a set of 327 diatom 168 and 117 macroinvertebrate species to investigate. We log-transformed (log x + 1) the species 169 abundance matrices. We used the R package ade4 (Dray et al. 2018) for conducting the OMI 170 analyses. 171 172 2.3. Biological traits 173 We collated information on biological traits for all 327 diatom and 117 macroinvertebrate 174 species. For diatoms, we followed Rimet and Bouchez’s (2012) work and gathered information 175 on sizes, ecological guilds and colonial formation of species. The sizes are reported as 176 biovolume classes: 0-99 µm3 (class S1), 100-299 µm3 (S2), 300-599 µm3 (S3), 600-1499 µm3 177 (S4), and > 1500 µm3 (S5). Diatom species were assigned to four guilds: low profile, high 178 profile, motile and planktonic guilds. The different ecological guilds differ from each other by, 179 for instance, resource use and motility (Passy 2007). As a third biological trait, we 180 distinguished colonial and single cell species. This aspect is important in terms of, for example, 181 resource use (e.g. light or nutrients) or potential grazing pressure. Although the biological trait 182 information was mainly collected from Rimet and Bouchez (2012), not all of our species were 183 in their list. We hence used OMNIDIA software (Lecointe et al. 1993) and its databases to find 184 out the missing information. In some cases, we made trait assignments based on characteristics 185 of similar species belonging to the same genus. 186 For macroinvertebrates, we also considered a measure of size as a biological trait. Here, 187 the size is actually the dry mass of each species, and it was divided to five classes: 0-0.99 mg 188 (class DM1), 1-3.49 mg (DM2), 3.5-9.99 mg (DM3), 10-34.99 mg (DM4), and > 35 mg (DM5) 189 to facilitate comparisons with the diatom data. Further, we assigned each species to functional 190 feeding groups (FFG), which were shredders, scrapers, predators, piercers, collector-gatherers 191 and filterers. The third biological trait for macroinvertebrates was their substrate-association 192 type. In our data, there were crawlers, swimmers, burrowers, sessiles and semisessiles. The 193 species was not as well explained by the explanatory factors as their occupancy (34% vs. 77% 344 explained) indicates that, for macroinvertebrates, the formation of variation in abundance is a 345 more complex process than formation of variation in occupancy, and could partly be a result 346 of other factors, such as stochasticity or random effects. The complexity was also visible in the 347 scatter plots, where the occupancy-abundance relationship was weaker for macroinvertebrates 348 than for diatoms. Previous research has also shown that abundance of macroinvertebrates 349 cannot be as clearly linked to niche parameters as their occupancy (Tonkin et al. 2016). 350 The importance of niche position, and to a lesser extent that of niche breadth, was evident 351 for explaining variation in our response variables. However, that was generally not the case 352 with biological traits or taxonomic relatedness. There were small effects of the trait or 353 taxonomic vectors, but nothing evident. This basically means that the biological traits studied 354 here were not strongly related to occupancy and abundance. Although previous research on 355 streams has shown that biological traits, such as body size, are connected to species regional 356 occupancy and mean local abundance (Passy 2012; Rocha et al. 2018), we could not detect any 357 clear patterns suggesting the importance of biological traits. Perhaps the characteristics of our 358 study system, i.e. high connectivity, resulted in the lack of a clear relationship. Previous 359 research, however, has shown that biological traits (i.e. organism’s dispersal capacity) may 360 affect the elevation of the abundance-occupancy relationship in highly-connected marine 361 systems (Foggo et al. 2007). This finding was reported from a significantly larger area than 362 what we investigated here. Consequently, perhaps we did not detect clear effects of biological 363 traits because of the small spatial scale addressed (Brändle and Brandl 2001) or maybe we used 364 the wrong traits (Heino and Tolonen 2018). In addition, it is possible that niche characteristics 365 are simply just more important than biological traits for the formation of variation in occupancy 366 and abundance of aquatic organisms (e.g. Rocha et al. 2018). 367 Similarly, taxonomic relatedness did not play a role in affecting occupancy, abundance 368 and their relationships for freshwater diatom and invertebrate species. Previous research has 369 neither found support for clear effects of taxonomy to variation in occupancy and abundance 370 of freshwater organisms (Tales et al. 2004; Heino and Tolonen 2018). As biological traits are 371 products of evolution and thus portrayed by phylogeny (Harvey 1996), it is not surprising that 372 taxonomic relatedness neither appeared as a strong predictor of occupancy and abundance. On 373 the other hand, some biological traits have evolved many times and can be characteristic of 374 comparatively distant orders (Rimet and Bouchez 2012), so in that sense, biological traits may 375 not always be as closely related to phylogeny as expected (Harvey 1996). It is also possible 376 that the spatial scale investigated in this study was not sufficient to detect clear effects of 377 taxonomic relatedness on occupancy and abundance. Phylogenetic signals might have been 378 found over larger areas crossing regional species pools (Heino and Tolonen 2018). 379 A possible caveat of this study is that the species’ niche parameters were calculated based 380 on environmental variables collected during the same sampling as the biological samples. 381 Because of no other suitable data existed to calculate the species-specific niche parameters, we 382 opted to use the same dataset. There may thus be a possibility that the effects of niche 383 parameters on occupancy and abundance may have been overestimated. Previous research 384 however has shown that irrespective of the underlying data (i.e. same or different dataset) to 385 calculate niche parameters, they arise as important determinants of occupancy and abundance 386 (Heino 2005; Heino & Grönroos 2014; McCreadie and Adler 2014; Teittinen et al. 2018). 387 388 5. Conclusions 389 We found that niche position was the strongest predictor of occupancy and abundance of 390 diatoms and macroinvertebrates in a freshwater system with high connectivity. This finding is 391 consistent with previous knowledge from across-waterbody systems with presumably lower 392 connectivity among sites (Tales et al. 2004; Heino and Soininen 2006; Tonkin et al. 2016; 393 Heino and Tolonen 2018; Rocha et al. 2018). The fact that our highly-connected freshwater 394 system showed similar results to comparatively weakly-connected across-waterbody systems 395 implies that these patterns in occupancy and abundance are perhaps universal and are not 396 strongly related to the connectivity of a study system. Furthermore, due to high connectivity, 397 our study setting had fairly subtle environmental ranges, indicating that niche position and, to 398 a smaller extent, niche breadth, can have strong effects on occupancy and abundance also in a 399 situation where environmental variation is comparatively small. Importantly, these findings 400 were evident even when controlling for biological traits and taxonomic relatedness. As species 401 abundances and occupancies are basically the building blocks for a number of ecological 402 indicators, the observed importance of niche parameters is alleviating regarding the use of these 403 sorts of indicators in environmental assessment. 404 The effect of niche position on the response variables was negative, indicating that the 405 more marginal the niche, the rarer the species both in terms of occupancy and abundance. In 406 other words, rare species tended to possess marginal niches within a large lake system. This 407 was evident for two very distinct groups of freshwater organisms, which represent different 408 trophic positions of the food web and contribute differently to the overall functioning of the 409 aquatic ecosystem. Thus, in order to enhance biodiversity conservation, protection of 410 regionally marginal habitats is important for protecting regionally rare species, also in systems 411 of high connectivity. 412 413 Acknowledgements 414 Annika Vilmi was supported by the Chinese Academy of Sciences President’s International 415 Fellowship Initiative (2018PS0007). We acknowledge Marja Lindholm and Mariana Perez 416 Rocha for sharing their information on biological traits of some diatom species. 417 418 Conflict of Interest: The authors declare that they have no conflict of interest. 419 420 Supplementary material 421 Appendix A. A map showing the sampling sites around a large lake system of Lake Kitkajärvi. 422 Appendix B. Examples of how to interpret vectors. 423 Appendix C. The diatom and macroinvertebrate species studied, in order of occupancy. 424 425 References 426 Brändle M, Brandl R (2001) Distribution, abundance and niche breadth of birds: scale matters. 427 Glob Ecol Biogeogr 10:173–177 428 Brown JH (1984) On the Relationship between Abundance and Distribution of Species. Am 429 Nat 124:255–279 430 Dolédec S, Chessel D, Gimarat-Carpentier C (2000) Niche separation in community analysis: 431 a new method. Ecology 81:2914–2927 432 Dray S, Dufour A-B, Thioulouse J (2018) ade4: Analysis of Ecological Data: Exploratory and 433 Euclidean Methods in Environmental Sciences. R Package version 1.7-11. URL 434 https://CRAN.R-project.org/package=ade4 435 Faulks L, Svanbäck R, Ragnarsson-Stabo H, Eklöv P, Östman Ö (2015) Intraspecific Niche 436 Variation Drives Abundance-Occupancy Relationships in Freshwater Fish Communities. 437 Am Nat 186:272–283 438 Foggo A, Bilton DT, Rundle SD (2007) Do developmental mode and dispersal shape 439 abundance–occupancy relationships in marine macroinvertebrates? Journal of Animal 440 Ecology 76:695–702 441 Gaston KJ, Blackburn TM (2000) Pattern and Process in Macroecology. Blackwell, Oxford. 442 Gaston KJ, Blackburn TM, Greenwood JJD, Gregory RD, Quinn RM, Lawton JH (2000) 443 Abundance–occupancy relationships. J Appl Ecol 37:39–59 444 Hanski I, Kouki K, Halkka A (1993) Three explanations of the positive relationship between 445 distribution and abundance of species. In: Ricklefs, R. E. and Schluter, D. (eds), Species 446 diversity in ecological communities: historical and geographical perspectives. Univ. of 447 Chicago Press, pp. 108–116. 448 Harvey PH (1996) Phylogenies for ecologists. J Anim Ecol 65:255–263. 449 Heino J (2013) The importance of metacommunity ecology for environmental assessment 450 research in the freshwater realm. Biol Rev 88:166–178. 451 Heino J, de Mendoza G (2016) Predictability of stream insect distributions is dependent on 452 niche position, but not on biological traits or taxonomic relatedness of species. Ecography 453 39: 001-011 454 Heino J, Soininen J (2006) Regional occupancy in unicellular eukaryotes: a reflection of niche 455 breadth, habitat availability or size-related dispersal capacity? Freshw Biol 51:672–685. 456 Heino J, Tolonen KT (2018) Ecological niche features override biological traits and taxonomic 457 relatedness as predictors of occupancy and abundance in lake littoral macroinvertebrates. 458 Ecography 41, in press, doi: 10.1111/ecog.03968 459 Heino J, Virkkala R, Toivonen H (2009) Climate change and freshwater biodiversity: detected 460 patterns, future trends and adaptations in northern regions. Biol Rev 84:39–54 461 Laliberté E, Legendre P, Shipley B (2014) FD: measuring functional diversity (FD) from 462 multiple traits, and other tools for functional ecology. R package version 1.0-12. URL 463 https://CRAN.R-project.org/package=FD 464 Lange-Bertalot H, Hofmann G, Werum M, Cantonati M (2017) Freshwater Benthic Diatoms 465 of Central Europe: Over 800 Common Species Used in Ecological Assessment. Edited 466 by Cantonati M, Kelly MG, Lange-Bertalot H. Koeltz Botanical Books, Germany, 942 467 pp. 468 Lecointe C, Coste M, Prygiel J (1993) Omnidia: software for taxonomy, calcula-tion of diatom 469 indices and inventories management. Hydrobiologia 269/270:509–513 470 Legendre P, Legendre L (2012) Numerical Ecology, 3rd edn. Elsevier, Amsterdam 471 Lewis SL, Maslin MA (2015) Defining the Anthropocene. Nature 519:171–180. 472 Merritt RW, Cummins KW (1996) An introduction to the aquatic insects of North America, 473 3rd ed. Kendall/Hunt Publishing. 474 Nimon K, Oswald F, Roberts JK (2013) yhat: Interpreting Regression Effects. R package 475 version 2.0-0. URL https://CRAN.R-project.org/package=yhat 476 Oksanen J, Blanchet FG, Friendly M, Kindt Roeland, Legendre P, McGlinn D, Minchin PR, 477 O’Hara RB, Simpson GL, Solymos P, Stevens MHM, Szoecs E, Wagner H (2018) vegan: 478 Community Ecology Package. R package version 2.5-2. URL https://CRAN.R479 project.org/package=vegan 480 Passy SI (2007) Diatom ecological guilds play distinct and predictable behavior along nutrient 481 and disturbance gradients in running waters. Aquat Bot 86:171–178 482 Passy SI (2012) A hierarchical theory of macroecology. Ecol Lett 15:923–934 483 R Core Team (2018) R: A language and environment for statistical computing. R Foundation 484 for Statistical Computing, Vienna, Austria. https://www.R-project.org/. 485 Ray-Mukherjee J, Nimon K, Mukherjee S, Morris DW, Slotow R, Hamer M (2014) Using 486 commonality analysis in multiple regressions: a tool to decompose regression effects in 487 the face of multicollinearity. Methods Ecol Evol 5:320–328 488 Rimet F, Bouchez A (2012) Life-forms, cell-sizes and ecological guilds of diatoms in European 489 rivers. Know Manag Aquat Ecosyst 406:1–14 490 Roberts DW (2016) labdsv: Ordination and Multivariate Analysis for Ecology. R package 491 version 1.8-0. URL https://CRAN.R-project.org/package=labdsv 492 Rocha MP, Bini LM, Siqueira T, Hjort J, Grönroos M, Lindholm M, Karjalainen SM, Heino J 493 (2018) Predicting occupancy and abundance by niche positions, niche breadth and body 494 size in stream organisms. Oecologia 186:205–216 495 Rohweder J, Rogala JT, Johnson BL, Anderson D, Clark S, Chamberlin F, Runyon K (2008) 496 Application of wind fetch and wave models for habitat rehabilitation and enhancement 497 projects: U.S. Geological Survey, Open-File Report 2008–1200, 43 p. 498 Roney NE, Kuparinen A, Hutchings JA (2015) Comparative analysis of abundance–occupancy 499 relationships for species at risk at both broad taxonomic and spatial scales. Can J Zool 500 93:515–519 501 Schmera D, Podani J, Heino J, Erös T, Poff NLR (2015) A proposed unified terminology of 502 species traits in stream ecology. Freshwater Sci 34:823–830 503 Soininen J, Heino J (2005) Relationships between local population persistence, local 504 abundance and regional occupancy of species: distribution patterns of diatoms in boreal 505 streams. J Biogeogr 32:1971–1978 506 Tales E, Keith P, Oberdorff T (2004) Density-range size relationships in French riverine fishes. 507 Oecologia 138:360–370 508 Tolonen KT, Hämälainen H, Holopainen IJ, Mikkonen K, Karjalainen J (2003) Body size and 509 substrate association of littoral insects in relation to vegetation structure. Hydrobiologia 510 499:179–190 511 Tolonen K, Vilmi A, Karjalainen SM, Hellsten S, Sutela T, Heino J (2017) Ignoring spatial 512 effects results in inadequate models for variation in littoral macroinvertebrate diversity. 513 Oikos 126:852–862 514 Tonkin JD, Arimoro FO, Haase P (2016) Exploring stream communities in a tropical 515 biodiversity hotspot: biodiversity, regional occupancy, niche characteristics and 516 environmental correlates. Biodivers Conserv 25:975–993 517 Venier LA, Fahrig L (1996) Habitat availability causes the species abundance-distribution 518 relationship. OIKOS 76:564–570 519 Verberk WCEP, van der Velde G, Esselink H (2010) Explaining abundance–occupancy 520 relationships in specialists and generalists: a case study on aquatic macroinvertebrates in 521 standing waters. J Anim Ecol 79:589–601 522 Verberk WCEP, van Noordwijk CGE, Hildrew AG (2013) Delivering on a promise: integrating 523 species traits to transform descriptive community ecology into a predictive science. 524 Freshwater Sci 32:531–547 525 Vilmi A, Karjalainen SM, Hellsten S, Heino J (2016) Bioassessment in a metacommunity 526 context: are diatom communities structured solely by species sorting? Ecol Indic 62:86– 527 94 528 Vilmi A, Karjalainen SM, Landeiro VL, Heino J (2015) Freshwater diatoms as environmental 529 indicators: evaluating the effects of eutrophication using species morphology and 530 biological indices. Environ Monit Assess 187:243 531 Vilmi A, Alahuhta J, Hjort J, Kärnä O-M, Leinonen K, Rocha MP, Tolonen KE, Tolonen KT, 532 Heino J (2017) Geography of global change and species richness in the North. Environ 533 Rev 25:184–192 534 Vörösmarty CJ, McIntyre PB, Gessner MO, Dudgeon D, Prusevich A, Green P, Glidden S, 535 Bunn SE, Sullivan CA, Liermann CR, Davies PM (2010) Global threats to human water 536 security and river biodiversity. Nature 467:555–561 537 Waters CN, Zalasiewicz J, Summerhayes C, Barnosky AD, Poirier C, Galuszka A, Cearreta A, 538 Edgeworth M, Ellis EC, Ellis M, Jeandel C, Leinfelder R, McNeill JR, Richter D de B, 539 Steffen W, Syvitski J, Vidas D, Wagreich M, Williams M, Zhisheng A, Grinewald J, 540 Odada E, Oreskes N, Wolfe AP (2016) The Anthropocene is functionally and 541 stratigraphically distinct from the Holocene. Science 351: aad2622. 542 doi:10.1126/science.aad2622. 543 Wickham H (2016) ggplot2: Elegant Graphics for Data Analysis. R package version 3.0.0. 544 Springer-Verlag New York. https://CRAN.R-project.org/package=ggplot2 545 546 Tables 547 548 Table 1. Linear regression model statistics presenting the relationship of logit-transformed 549 occupancy and log-transformed mean abundance at occupied sites for diatom and 550 macroinvertebrate species. Bolded p values indicate statistically significant results (p ≤ 0.05). 551 The explanatory power of the linear models, as indicated by R2 values, were 0.568 for diatoms 552 (p < 0.001) and 0.384 for macroinvertebrates (p < 0.001). 553 Estimate SE t p Diatoms (Intercept) -4.0807 0.12281 -33.23 <0.001 Abundance 1.80665 0.08748 20.65 <0.001 Invertebrates (Intercept) -3.7342 0.2589 -14.43 <0.001 Abundance 1.4106 0.1667 8.46 <0.001 Figure captions 574 Fig. 1. A schematic figure illustrating three species, A, B and C, and their niche positions and 575 niche breadths across an environmental range. Species A has a marginal niche position, while 576 species B and C have non-marginal niche positions (i.e. their niches are located close to the 577 mean environmental conditions). Species A and C have small niche breadths, while species B 578 has a large niche breadth (i.e. it is able to live in a broader range of environmental conditions 579 compared to species A and C). It is noteworthy that two species can have the same niche 580 position although the niche breadth differs (species B vs. species C). Thus, although niche 581 position of species C is non-marginal, it is still a specialist species for those non-marginal 582 conditions. 583 Fig. 2. Scatter plots describing the relationships between logit-transformed occupancy and log584 transformed mean abundance at occupied site of diatoms (A) and macroinvertebrates (B). 585 Figures 586 587 Fig. 1. 588 589 590 Fig. 2. 591