Our study aimed to investigate the roles of environmental filtering and local adaptation in structuring zooplankton communities. The results indicated that environmental filtering played a more significant role in zooplankton community composition than local adaptation. Zooplankton communities developed differently in their home and away environments, however, on the individual species level, there was no consistent indication of local adaptation.
Environmental Filtering
We found a habitat effect on the zooplankton community, both on species richness and on community composition suggesting that environmental filtering is at least partly a driving factor for zooplankton communities in the study area. Comparing the species richness of the zooplankton communities from ponds 807, 892 and 2484 in their own environment and in the environment of pond 1189 revealed that all these communities lost species when they were transferred to pond 1189. A closer inspection on the species level showed that these were not the same species that got lost in the 1189 environment. This suggests that environmental filtering in concert with biotic interactions drives the zooplankton community. These findings are underlined by the analysis of the community composition using PERMANOVA. The same three ponds (807, 892 and 2484) that were relatively species-rich compared to pond 1189, exhibited different shifts in community composition when transferred to the other environments. Only the species-poor community from Pond 1189 did not shift in response to habitat transfer. One potential explanation is that the community is composed of species with broad ecological niches that can thrive in various environments. An important issue in studying environmental filtering is a clear definition of the environment. In plant studies, this is often regarded as the sum of abiotic factors. For animals, or zooplankton in particular, not only the abiotic environment is important, but also the biotic environment such as the food basis, typically phytoplankton and bacteria. Therefore, we filtered the pond water through a 30 µm mesh in order to collect the zooplankton but also to keep the edible fraction of phytoplankton in the respective pond environment. Ultimately, we cannot pinpoint the causal factors of our findings, but we can demonstrate experimentally that environmental filters act on the communities. This direct observation complements previous findings on the zooplankton community assembly in the study region that indirectly inferred from species occurrence data that environmental filtering is a relevant factor (Kiemel et al., 2022). Environmental filtering has been found in several aquatic systems as a mechanism that drives the community structure in various organismal groups. For example, Chaparro et al. (2018) found a significant habitat effect on zooplankton communities in floodplain ponds by using distance-based redundancy analysis (dbRDA). In another study, Anas et al. (2015) found a habitat effect on zooplankton in fishless ponds. They suggest that lake productivity, acid-base status and invertebrate predation were relevant environmental filters. In addition, also for prokaryotes such as the bacterial communities in 35 Belgian shallow ponds, environmental filtering by abiotic factors was identified as the main factor explaining community variation by using dbRDA (Hanashiro et al., 2022). However, biotic interactions have also been found as a structuring force of lake communities. García-Girón et al (2020) found by applying partial correlation networks that biotic interactions contribute substantially to species sorting among five organismal groups (macrophytes, phytoplankton, zooplankton macroinvertebrates and fish). But not only in ponds that are separated habitats with the landscape, also in lentic, riverine systems where sub-habitats are unidirectionally connected, environmental filtering has been identified as a driving factor. The environment was a significant determinant for benthic diatoms in rivers in south-east China that are not strongly attached to their substrate (Liu et al., 2013). Also for rivers in Finland, environmental filtering was found, however, only for three (insects, macrophytes and fish) out of six organismal groups (Heino et al., 2017). These and other studies, for example Kulkarni et al. (2019) and Cottenie (2005), demonstrated that environmental filtering is a common driver in aquatic metacommunities. Based on statistical methods such as variation partitioning the explanatory power i.e. the explained variance by the environment, though significant, is often low, suggesting other drivers contribute as well or stochastic/neutral processes also occur.
Local adaptation
Whereas environmental filtering is a mechanism explaining the presence or absence of a species, local adaptation is a mechanism acting among subspecies or evolutionary lineages within species. Local adaptation means that genotypes of the same species have a higher fitness in their home environment than away (home vs. away criterion) or individuals of a species have a higher fitness in their home habitat than genotypes of the same species that come from other habitats (Kawecki & Ebert, 2004). Our common garden experiment revealed no consistent pattern of local adaptation across species. Some species performed better at home than away, but for others, we found the opposite pattern and some responded indifferently. In principle, in cases where we did not infer local adaptation, this might have two reasons: 1) local adaptation does not play a role and 2) local adaptation does play a role, but we could not measure it.
1) Although the ponds we studied lack physical connections, potential dispersal vectors (e.g. wind, animals) in the landscape (Colangeli, 2018; Parry et al., 2023) might have facilitated frequent dispersal leading to the homogenization of habitats preventing local adaptation of species (Kisdi, 2002; Sanford & Kelly, 2011). Another reason might be that the seasonal variation in environmental factors in such small ponds is so high (Chase, 2003) that the periods during which a certain set of environmental factors occur are too short for adaptive processes.
2) Since we tested for local adaptation only at one point in time, we cannot exclude that changes in environmental conditions might lead to different results. Thus, species that appear to be maladapted might show local adaptation, when the environment changes. On a species level, local adaptation might manifest in a higher average fitness at home than away over a longer period of time.
In general, restricted dispersal might lead to species becoming adapted to their local environment, shaping the population through natural selection and improving their fitness in the environment, turning into local adaptation (Balaguer et al., 2001). This implies that local adaptation could increase the fitness of species in a specific environment (Weisse, 2008). Additionally, although there is improved fitness, it may be primarily restricted to the specific habitat and might have no effect or reduced effect in other habitats (Lenormand, 2002; Kawecki & Ebert, 2004; Leimu & Fischer, 2008; Hereford, 2009).
Our findings are similar in some respect to a previous result reported by Weithoff et al. (2019) from isolated extremely acidic mining lakes. In five rotifer isolates, they found a clear habitat effect but there was no clear indication of local adaptation of species. They however attributed this to the small population size of the rotifers and the relatively young age of the acidic lakes sampled.