Colonisation, emigration and equilibrium of stream invertebrates in patchy habitats

Abstract

Understanding of the ecology of stream assemblages is enhanced by quantifying colonisation and dispersal rates, but these are difficult to measure at small scales. Using field‐located artificial streams that mirrored natural patches of habitat, we investigated immigration, emigration and stability of invertebrates on leaf‐litter packs over 38 days. We fitted equilibrium and “drift‐out” models to immigration and emigration data. Overall colonisation by individuals (plateau at ˜Day 12) and taxa (total 50; plateau ˜21 taxa at ˜Day 24) was rapid and curvilinear towards a stable asymptote, suggesting saturation and equilibrium. Abundant early colonists, mainly several chironomid species, dictated the shape of the colonisation curve, but the patterns varied among species. Thus, the plateaux were due to a successional shift in composition rather than a simple equilibrium. The equilibrium model fitted total numbers and taxa, and individual taxa that were gradual colonists, but not taxa lacking constant immigration. High turnover was due mainly to rarer taxa. The assemblage was maintained by drift‐mediated immigration, compensating for individuals lost through emigration, emergence or mortality. The masking effect of a few taxa and substantial turnover rather than simple accumulation of individuals are important considerations regarding assemblage composition over time. Rapid turnover of individuals provides for rapid colonisation of denuded substrata downstream.

Do constrained immigration rates and high β diversity explain contrasting productivity–diversity patterns measured at different scales?

“…The artificial stream channels were colonised by a diverse assemblage of macroinvertebrates, as previously (Connolly and Pearson 2018). Despite the low number of channels, strong patterns in invertebrate abundance and diversity were evident.…”

“…The 40-day duration of the experiment allowed for adequate colonisation by invertebrates (Connolly and Pearson 2018), but was not long enough for leaf packs to degrade (Pearson and Connolly 2000;Connolly and Pearson 2013). We previously showed that accumulation of abundance and richness on leaf packs in the artificial channels plateaued at around 24 days and then remained stable up to the 38-day duration of the experiments as a result of the equilibrium between local immigration and emigration (whether active or passive-i.e., local colonisation and extinction) (Connolly and Pearson 2018). At the end of the present experiment, all cobbles and cobble/leaf-pack combinations were removed in sequence from downstream to upstream, with cobbles and leaf packs, when present, placed in separate plastic containers, and labelled with the channel number, treatment and position within the channel (1-30).…”

“…We examined the relationship in a largely detrital food web, dependent mainly on riparian leaf litter input and the associated biomass of microbes and deposited fine organic material (Fisher and Likens 1972;Rosemond et al 2001;Cheshire et al 2005;Connolly and Pearson 2013). As well as providing food sources, leaf litter provides complex microhabitat, and both types of resource contribute to productivity, as judged by, for example, invertebrate abundance (Richardson 1992;Dudgeon and Wu 1999;Pearson and Connolly 2000;Connolly and Pearson 2018). We established density gradients of terrestrial leaf litter in streamside artificial stream channels to represent a gradient in productivity and measured the response of the invertebrate assemblage.…”

“…We established density gradients of terrestrial leaf litter in streamside artificial stream channels to represent a gradient in productivity and measured the response of the invertebrate assemblage. We had previously shown that colonisation by drifting invertebrates was rapid and that there was high turnover of individuals, especially of rarer species, and that loss of individuals by drift was compensated for by continuous supply from upstream (Connolly and Pearson 2018). We, therefore, investigated the relationship between species diversity and the resource gradient at two scales-on individual leaf packs within channels, and across channels with different leaf-pack densities.…”

“…We follow the lead of previous reports (see above) regarding different patterns at different scales by applying essentially a productivity-diversity model. We did not pursue a metacommunity model (e.g., Economo and Keitt 2008) because our litter patches were all highly connected by invertebrate drift, as is the case within riffles naturally (e.g., Benson and Pearson 1987;Bunn and Hughes 1997;Connolly and Pearson 2018), and because metacommunity studies of streams generally use the whole riffle or stream as the base unit (e.g., Tonkin et al 2016;Heino et al 2017;Crabot et al 2020).…”