Spatial context strongly affects community composition of both passively and actively dispersing pool invertebrates in a highly heterogeneous landscape
Abstract
The spatial distribution of suitable habitats and dispersal abilities of the constituent taxa jointly affect the structure of metacommunities in standing freshwaters. Most studies exploring spatial effects on aquatic metacommunities, however, focus on at most a few taxonomic groups.
Within two consecutive seasons, we studied spatial patterns in the species richness and composition of three passively dispersing and three actively flying freshwater invertebrate groups (rotifers, microcrustaceans and molluscs vs. hemipterans, aquatic beetles and odonates) in a metacommunity system consisting of 42 newly or recently created fishless pools in a highly heterogeneous Central European sandstone landscape consisting of deep valleys and steep ridges. We hypothesized that the extent to which these dispersal barriers affect invertebrate groups depends on their dispersal mode, and that the ability of each group to colonize new habitats is affected by the landscape morphology.
Moreover, we predicted that the history and age of the pools would play a major role in structuring of invertebrate communities. Following the classical island biogeography pattern, habitat size (measured as pool surface area or depth) was the key characteristic influencing species richness for each of the six studied groups (range of explained variation: 10%-58.7%).
The number of nearby aquatic habitats (i.e., potential colonization sources) was also an important determinant of species richness for molluscs (18.8%), crustaceans (36.4%) and aquatic beetles (27.2%). After pool size, the most important factor influencing species richness was the presence and functional composition of aquatic macrophytes in the pools, which affected the species richness of odonates (25.2%), aquatic beetles (12.2%), rotifers (11.1%), and crustaceans (8.3%).
Valley distances between localities, defined as the shortest distance that avoids crossing steep ridges, explained consistently slightly more variation in species composition (2.6%-12.6%) than did Euclidean distances (1.0%-10.1%) for all six groups. Spatial variables (the valley distance matrix, position of pools within clusters in the landscape, and the number of nearby aquatic habitats) explained more variation in species composition (3.4%-25.4%) than local pool characteristics (2.8%-9.4%) or temporal variation (0%-7.6%) in all taxa except hemipterans, whose species composition was almost equally affected by local (3.3%) and spatial factors (3.4%).
We conclude that landscape-level spatial structure in our study area affects the dispersal and metacommunity assembly of both actively and passively dispersing invertebrates more than studied pool characteristics or temporal variation. The observed congruence between groups with different dispersal modes is likely because flying insects follow similar dispersal routes as the key animal vectors of passive dispersers.
Our study highlights the importance of including relevant topography features in studies of aquatic metacommunities in complex and heterogeneous landscapes, even for taxa considered to be efficient dispersers.