Spatial modelling of temporal dynamics in stream fish communities under anthropogenic change

Mário Mota‐Ferreira,Ana Filipa Filipe,Sara Carona,Pedro Beja,Maria Filomena Magalhães,Alana Grech

doi:10.1111/ddi.13189

Mário Mota‐Ferreira, Ana Filipa Filipe + Show 4 more

Open Access

https://doi.org/10.1111/ddi.13189

Copy DOI

Abstract

AbstractAimUnderstanding temporal changes in aquatic communities is essential to address the freshwater biodiversity crisis. In particular, it is important to understand the patterns and drivers of spatial variation in local community dynamics, generalizing temporal trends from discrete locations to entire landscapes that are the main focus of management. Here, we present a framework for producing spatially continuous views of community dynamics, focusing on stream fish affected by hydropower development.LocationRiver Sabor, NE Portugal.MethodsWe sampled stream fish at thirty sites between 2012 and 2019. Community trajectory analysis was used to quantify the directionality and velocity of community change, and the geometric resemblance of community trajectories between sites. Geostatistical models for stream networks were used to relate metrics describing community dynamics to environmental variables, while controlling for Euclidean and hydrologic spatial dependencies, and to map spatial variation in community dynamics across the watershed.ResultsTrajectories in multivariate space underlined strong temporal dynamics, with local communities deviating and returning to previous states, but without evidence for directional changes. Accordingly, directionality values were low and not consistently affected by environmental variables. The velocity of community change varied markedly across the watershed and it was strongly affected by stream order and elevation, with faster changes observed in lowland streams draining into hydroelectric reservoirs and with a high proportion of exotic species. Pairwise distances between community trajectories were strongly related to hydrologic and environmental distances between sites.Main conclusionsLocal stream fish communities were in a loose equilibrium across the watershed, but they fluctuated at a faster rate closer to a hydroelectric reservoir. Integrating community trajectory analysis and geostatistical modelling provides a relatively simple framework to understand how, where and why temporal community dynamics vary across dendritic stream networks and to visualize spatial patterns of community change over time in relation to anthropogenic impacts.

Highlights

In freshwater ecosystems, biological communities are changing worldwide due to multiple anthropogenic pressures (Albert et al, 2020; Reid et al, 2019)
Note: For each species, we indicate its status in the region (native [N] vs. exotic [E] status, the total, mean (± SD), range and coefficient of variation (CV) of number of individuals > 5 cm collected per year, the percentage of sites where it was detected (% Sites) and the percentage of years (% Years) when it was detected. aThe species is native to the Iberian Peninsula, but it was considered exotic in our study area based on Doadrio et al (2011), but see Sánchez-Hernández et al (2018). bThe species was listed as Achondrostoma arcasii by Ferreira et al (2016) but probably belongs to an undescribed species (Robalo et al, 2006)
Our study shows that combining robust descriptors of community change with state-of-the-art geostatistical modelling contributes to understanding and predicting where, how and why the dynamics of local communities vary across stream networks

Summary

Introduction

Biological communities are changing worldwide due to multiple anthropogenic pressures (Albert et al, 2020; Reid et al, 2019). Changes may follow distinct temporal patterns, being for instance gradual or saltatory, or reflecting variations around loose equilibria, shifts between alternative stable states or gradual directional transitions away from initial community structures (Collins, 2000; DeAngelis et al, 1985; Matthews & Marsh-Matthews, 2017) These changes can be quantified using relatively simple metrics, such as Kendall's coefficient of concordance to estimate constancy in species rank abundances or the coefficient of variation to estimate variability in species abundances (Grossman et al, 1990). The lengths and speed of trajectories can be used to quantify how much the community changes over time and whether changes are gradual or abrupt, while direction can quantify whether changes are directional or not (De Cáceres et al, 2019)

Objectives

Findings

Discussion

Conclusion