Realized niche and microhabitat selection of the eastern green lizard (Lacerta viridis) at the core and periphery of its distribution range.

Ana M Prieto‐Ramirez,Dennis Rödder,Guy Pe’Er,Klaus Henle

doi:10.1002/ece3.4612

Abstract

The available range of habitats and suitable abiotic conditions like temperature and radiation tends to be narrower toward the periphery of the distribution range of species. Peripheral populations of generalist species could then be more specialized and have a smaller and differentiated realized niche (habitat niche in our study) compared to populations at the core. Likewise, patterns of microhabitat selection can differ between periphery and core. In our study, we compared niche size and microhabitat selection among core (Bulgaria) and northern peripheral (Germany, Czech Republic) populations of Lacerta viridis and estimated niche differentiation among regions. We collected data on vegetation structure and abiotic parameters at the microhabitat scale in each region. In order to compare niche size among regions and estimate niche differentiation, we built multidimensional niche hypervolumes. We applied generalized linear mixed models and model averaging, accounting for spatial autocorrelation when necessary, to analyze microhabitat differences among regions and microhabitat selection in each region. Peripheral populations were more specialized, having a smaller niche than core ones, and their niche differed from that in the core (Sørensen overlap in all comparisons <0.3). Microhabitats at the periphery had lower radiation and soil compaction and less structured vegetation. Microhabitat selection at the core depended solely on abiotic parameters, while at the periphery it was defined by only vegetation structure (Czech Republic) or a combination of both, vegetation structure, and abiotic factors (Germany). Thus, peripheral populations seem to compensate for overall harsher climatic conditions by responding to different parameters of the microhabitat compared to core populations. We suggest specific conservation measures for L. virids in each studied region and point out the general implications of a higher specialization degree of peripheral populations in relation to climate change and habitat fragmentation.

Highlights

We expected to find (a) smaller realized niches in northern edge pop‐ ulations compared to the core, with a niche differentiation pres‐ ent in populations located around Prague but not in those in Passau; (b) higher preference of L. viridis in the periphery for specific vegetation structures at the microhabitat scale, like low and open vegetation, as compensation for overall suboptimal climatic conditions; and (c) higher influence of vegetation structure in the microhabitat selec‐ tion in the northern periphery, where the availability of suitable hab‐ itats for the species is a limiting factor, while in the core, where the available range of habitats is broader, abiotic parameters will have a higher influence in the microhabitat selection
The inclusion of random intercepts can enormously improve the explanatory capacity of models, and a high conditional R2 value is a very common output in GLMM that in‐ tend to find the best set of variables to explain the data (Nakagawa & Schielzeth, 2013) (Supporting Information Appendix S4, Table S4.1 for model selection and model averaging separately for vegetation structure and abiotic parameters)
The most important variables affecting microhabitat selection in the periphery‐Pa were a combination of vegetation structure and abiotic parameters: Branches, S‐N aspect, W‐E aspects, and tem‐ perature (RVI = 1) Lacerta viridis in the periphery‐Pa avoided loca‐ tions with high coverage of branches and selected places with an eastern and southern aspect where temperatures are higher

Summary

Introduction

We expected to find (a) smaller realized niches in northern edge pop‐ ulations compared to the core, with a niche differentiation pres‐ ent in populations located around Prague (relicts) but not in those in Passau (which are part of the continuous distribution range); (b) higher preference of L. viridis in the periphery for specific vegetation structures at the microhabitat scale, like low and open vegetation, as compensation for overall suboptimal climatic conditions; and (c) higher influence of vegetation structure in the microhabitat selec‐ tion in the northern periphery, where the availability of suitable hab‐ itats for the species is a limiting factor, while in the core, where the available range of habitats is broader, abiotic parameters will have a higher influence in the microhabitat selection.

Results

Conclusion