INTRODUCTION
A variety of disturbances from human activities are causing changes in
riverine ecosystems worldwide (Reid et al. 2018), with potential effects
on the capacity of these ecosystems to maintain functions and provide
services. Empirical assessments show that increasing human pressure
causes declines in the diversity of multiple organismal groups of
varying trophic levels in riverine systems (Moi et al. 2022). Such
declines, in turn, break-down the relationship between diversity and
ecosystem functioning (BEF), negatively influencing ecosystem functions,
such as nutrient cycling and biomass production (Soliveres et al. 2016).
It has been proposed that ecosystem functions are products of biological
processes controlled by multitrophic interactions. For instance,
different trophic levels can combine to have strongest impacts on
ecosystem functioning (Lefcheck et al. 2015; Thompson et al. 2012).
However, current evidence supporting human-induced impacts on diversity
and ecosystem functioning, as well as on BEF relationships, is mostly
from studies limited to single trophic levels (Hector et al. 1999; van
der Plas 2019).
Food web approaches offer the opportunity to integrate inter-specific
interactions across multiple trophic levels to understand the functional
consequences of a human-induced decline in diversity to ecosystem
functioning (Eisenhauer et al. 2019). Quantifying energy flux in food
webs has emerged as a powerful measure that describe the functioning of
multitrophic systems (Barnes et al. 2018; Gauzens et al. 2018). Energy
flux calculations include community attributes (abundance and body
mass), network topology (who feeds on whom), metabolic demand and
assimilation efficiency (Thompson et al. 2012; Barnes et al. 2018).
Empirical evidences suggest that diversity and energy flux positively
correlate (Barnes et al. 2014). Consequently, changes in diversity can
reduce the ability of trophic guilds to process energy, ultimately
decoupling energy flux through food webs (Barnes et al. 2014; Polazzo et
al. 2022).
For centuries, human activities
have caused profound changes in fish communities (Hoffmann et al. 1996)
through species loss and declining abundance in fish trophic guilds
(Lefcheck et al. 2021; Moi et al. 2022). Human activities are likely
altering the topology of fish food webs (Kortsch et al. 2021), which
could modify the strength of trophic interactions (i.e., the magnitude
of energy flux). It has been proposed a decrease in energy flux per
species loss in human-dominated ecosystems (Barnes et al. 2014). It is
also expected that human pressure impacts are stronger on higher trophic
levels, as top-carnivores are more sensitive to human-induced
environmental changes (Estes et al. 2011). To data, there is no evidence
of the aforementioned effects on freshwater food webs.
In addition to human pressure, empirical assessments have reported
several other factors driving the diversity and functioning of riverine
food webs, such as precipitation, nitrogen:phosphorus (N:P) ratios,
turbidity, and water discharge. Precipitation is a major driver fish
diversity, recruitment, and productivity
(Oliveira et al. 2015;
González-Bergonzoni et al. 2019). The N:P ratios also influence fish
diversity and productivity by influencing the basal resources that fuel
fish, such as composition and productivity of primary producers (Elser
et al. 2007; Glibert et al. 2012; Pineda et al 2020). Turbidity and
water discharge influence fish diversity and productivity by increasing
or decreasing reproductive success and prey encounter rates (Ortega et
al. 2020). How these drivers interact with human pressure to modulate
diversity and energy flux through fish food webs remains unknown.
Here we investigated the impacts of increasing human pressure on fish
species richness and food web functioning (i.e., energy flux) in a
neotropical river, the Río Uruguay, which is undergoing major changes
induced by anthropic activities such as urbanization and intensive
cropland (Figs. S1 and S2). We used a unique long-term dataset from fish
communities comprising more than 20,000 individuals of 117 species. Our
dataset includes species richness, total abundance and body mass of four
fish trophic guilds, namely top-carnivore, mesocarnivore, detritivores
and omnivores, as well as information on the local human pressure
(quantified by using human footprint index ; Venter et al. 2016)
and local abiotic conditions, including precipitation, N:P ratio,
turbidity and water discharge. We quantified the annual energy flux by
each trophic compartment and at the whole-network level (Fig. 1) to
estimate how ecosystem functions changed over time. We first analyzed
whether species richness, abundance, and energy flux in single trophic
guilds and in the whole-fish communities have changed over 17-years.
Accounting for the influence of precipitation, N:P ratio, turbidity, and
water discharge, we used structural equation modeling to assess whether
and how human pressure affects species richness, and energy flux of each
trophic compartment.