opinions, perspectives & reviews ISSN 1948-6596 thesis abstract Environmental heterogeneity–species richness relationships from a global perspective Anke Stein Biodiversity, Macroecology & Conservation Biogeography Group, University of Göttingen, Göttingen, Germany; [email protected] Abstract. Spatial environmental heterogeneity (EH) is considered one of the most important factors pro- moting species richness, but no general consent about the EH–richness relationship exists so far. This is because research methods and study settings vary widely, and because non-significant and negative as- sociations have also been reported. My thesis provides a comprehensive review of the different meas- urements and terminologies of EH used in the literature, and presents strong quantitative evidence of a generally positive relationship between biotic and abiotic EH and species richness of terrestrial plants and animals from landscape to global extents. In a meta-analysis and a subsequent case study compar- ing multiple EH measures and their association with mammal species richness worldwide, I furthermore reveal that the outcome of EH–richness studies depends strongly on study design, including both the EH measure chosen and spatial scale. My research contributes to a better understanding of the EH–richness relationship, while identifying future research needs. Keywords. habitat diversity, heterogeneity measures, meta-regression, spatial grain, species diversity, terminology, topography, vegetation structure Introduction and Methods cies richness have been attributed to ecological, Spatial environmental heterogeneity (EH) has fas- historical and evolutionary mechanisms. First, in- cinated researchers from ecology, biogeography, creased EH leads to an increase in diversity of re- conservation biology, and evolutionary biology for sources, structural complexity and other environ- decades, and is considered one of the most impor- mental conditions, which should increase avail- tant factors determining species richness (Tews et able niche space and thereby promote species’ al. 2004, Field et al. 2009). Early research revealed coexistence (Hutchinson 1959). Second, increasing positive relationships between vegetation struc- EH is expected to enhance species’ persistence ture and bird and lizard species diversity through the provision of shelter and of refuges (MacArthur and MacArthur 1961, Pianka 1967), from long-lasting adverse environmental condi- and close positive associations between area, EH tions, such as glaciations (e.g. Svenning and Skov and species richness were also recognized long 2007). Finally, EH should increase the probability ago (Hamilton et al. 1963, Williams 1964). Since of species diversification through isolation or ad- then, many different studies have investigated the aptation to diverse environmental conditions (e.g. effect of EH on species richness. However, in addi- Hughes and Eastwood 2006). Negative and hump- tion to positive EH–richness relationships, nega- shaped EH–richness relationships are mainly tive, hump-shaped and non-significant relation- thought to result from unfavourable fragmenta- ships have also been found (e.g. Tamme et al. tion effects (Tamme et al. 2010, Fahrig et al. 2011) 2010, Allouche et al. 2012). Therefore, even and an area–heterogeneity trade-off, i.e. a reduc- though ecological theory predicts a positive effect tion in the area of individual habitat types associ- of EH on species richness (reviewed in Stein and ated with increasing EH (Allouche et al. 2012, but Kreft 2014), the generality of positive EH–richness see, e.g., Hortal et al. 2013). Considering this relationships in nature is still debated. range of possible mechanisms and factors in- volved, it is unsurprising that the relationship be- Positive relationships between EH and spe- frontiers of biogeography 7.4, 2015 — © 2015 the authors; journal compilation © 2015 The International Biogeography Society 168 Anke Stein — Environmental heterogeneity and species richness tween EH and species richness is difficult to quan- taxon and spatial scale. The coherent framework tify and comprehend. developed in this chapter then allowed me to con- The difficulty of understanding EH–richness duct a meta-analysis to examine the strength and relationships also arises from the considerable direction of EH–richness relationships across ter- variability in taxa, ecosystems and spatial scales restrial study systems worldwide (chapter 2; Stein addressed in different studies, and the fact that et al. 2014). In this study, I tested whether the EH is measured in numerous different ways. For relationship is positive overall and whether it dif- instance, EH has been quantified with regard to fers between EH measures, study taxa, habitat vegetation structure, plant diversity, topographi- types and spatial scales, using mixed-effects meta- cal complexity and habitat diversity, and with regression. I used Fisher's z as a measure of effect many different measures based on indices, ranges size and applied robust variance estimation to al- and other calculation methods. Moreover, the low the combination of multiple, dependent effect terminology regarding EH is diverse and discor- size estimates per study in a single analysis dant, making it difficult to comprehend exactly (Hedges et al. 2010). what is being studied by individual studies and Based on the insights from the literature whether different authors refer to the same con- review, I then studied how different EH measures cept. Thus, the variability in EH–richness research relate to each other. To this end, I computed 51 hampers attempts to find and compare studies EH measures for all land areas worldwide, using and limits our understanding of the general EH– various environmental datasets and a range of richness relationship. calculation methods (chapter 3; Stein et al. 2015). The overall aims of my thesis (Stein 2014) I investigated the variability in EH measures using were to examine the concept and role of EH correlation and principal components analysis. within a broad, global framework, and to synthe- Furthermore, I analysed how different measures sise the current state of EH–richness research, vary in their relationship with species richness of including abiotic and biotic EH and a wide range of terrestrial mammals (derived from IUCN 2013) taxonomic groups. I thereby aimed to gain a more using simultaneous autoregressive models. I com- fundamental and general understanding of the EH pared single-predictor models (each with one EH –richness relationship. To this end, I conducted an measure) with multi-predictor models that addi- extensive, systematic review of the EH–richness tionally accounted for current climate, bio- literature, covering observational studies that ana- geographic region and human influence. I com- lysed the relationship between EH and species puted conditional inference trees (Hothorn et al. richness of terrestrial plants or animals at land- 2006) to examine whether model support de- scape to global extents. Based on 192 studies in- pended more on the subject area or calculation cluding 1148 data points, I first scrutinised the method of EH measures. Based on the strong methodology and terminology used in EH– scale-dependence of EH–richness relationships richness research (chapter 1; Stein and Kreft found in chapter 2, I kept the area of study units 2014). Specifically, I investigated how EH has been constant and conducted the analyses across three quantified and termed, and I classified the various different grain sizes, which are commonly used in EH measures by subject area, such as vegetation macroecological analyses (12,364 km², or topography, and calculation method, such as 49,457 km², and 197,829 km²; approximately 110 range or standard deviation. I used this classifica- km × 110 km, 220 km × 220 km, and 440 km × 440 tion, combined with information on study taxon, km, respectively). location, habitat type and spatial scale, to identify trends and gaps in research. Moreover, I reviewed Results and discussion the postulated mechanisms underlying positive EH My review is the first, to my knowledge, that sys- –richness relationships and linked them to the EH tematically quantifies the terms for EH combined subject areas and other study characteristics like with the measures used and mechanisms dis- frontiers of biogeography 7.4, 2015 — © 2015 the authors; journal compilation © 2015 The International Biogeography Society 169 Anke Stein — Environmental heterogeneity and species richness cussed in the literature. I revealed how heteroge- effects of EH. neous and ambiguous the quantification and ter- I detected multiple mechanisms that were minology of EH have been in past research: I iden- used in the literature to explain positive effects of tified 165 different EH measures, with even more EH on species richness. The majority of studies measure variants, related to biotic EH in land referred to niche theory, i.e. more diverse re- cover and vegetation, and abiotic EH in climate, sources and increased niche space allowing more soil and topography. These measures were de- species to coexist. These studies mostly investi- noted by more than 350 measure names; for in- gated how vegetation EH affects animal richness. stance, elevation range was also called altitude, Studies addressing evolutionary mechanisms altitudinal range, elevation variability, relief, and mostly related richness to topographic EH, assum- topography. I also detected more than 100 terms ing higher topographic