Experimental evidence that spate-induced downstream dispersal drives spatial sorting of visual–cognitive traits in a stream fish

Convergent evolution, the evolution of similar phenotypes among distantly related lineages, is often attributed to adaptation in response to similar selective pressures. Here, we assess the prevalence and degree of convergence in functional traits of stream fishes at the microhabitat scale in five zoogeographical regions across the world. We categorized species by microhabitat, water velocity and preference for substrate complexity and calculated the prevalence of convergence, degree of convergence and functional diversity for each category. Among species occupying similar microhabitats of small, low-gradient streams, 34% had combinations of convergent traits. Convergence occurred at higher rates than expected by chance alone, implying that adaptation to similar environmental conditions often resulted in similar evolutionary patterns along multiple niche dimensions. Two of the microhabitat groupings had significantly convergent species represented in all zoogeographical regions. Fishes occupying microhabitats with high water velocity and low structural complexity generally occupied a restricted morphospace and exhibited greater prevalence and higher degrees of convergence. This suggests that water velocity and habitat structural complexity interact, selecting a restricted distribution of trait distributions and higher degrees of convergence in stream fish assemblages. Furthermore, these results suggest that microhabitat features in streams select for fish trait distributions in a fairly predictable and deterministic manner worldwide.

Community convergence across biogeographically distinct regions suggests the existence of key, repeated, evolutionary mechanisms relating community characteristics to the environment. However, convergence studies at the community level often involve only qualitative comparisons of the environment and may fail to identify which environmental variables drive community structure. We tested the hypothesis that the biological traits of fish communities on two continents (Europe and North America) are similarly related to environmental conditions. Specifically, from observations of individual fish made at the microhabitat scale (a few square meters) within French streams, we generated habitat preference models linking traits of fish species to local scale hydraulic conditions (Froude number). Using this information, we then predicted how hydraulics and geomorphology at the larger scale of stream reaches (several pool–riffle sequences) should quantitatively influence the trait composition of fish communities. Trait composition for fishes in stream reaches with low Froude number at low flow or high proportion of pools was predicted as nonbenthic, large, fecund, long-lived, nonstreamlined, and weak swimmers. We tested our predictions in contrasting stream reaches in France (n = 11) and Virginia, USA (n = 76), using analyses of covariance to quantify the relative influence of continent vs. physical habitat variables on fish traits The reach-scale convergence analysis indicated that trait proportions in the communities differed between continents (up to 55% of the variance in each trait was explained by “continent”), partly due to distinct evolutionary histories. However, within continents, trait proportions were comparably related to the hydraulic and geomorphic variables (up to 54% of the variance within continents explained). In particular, a synthetic measure of fish traits in reaches was well explained (50% of its variance) by the Froude number independently of the continent. The effect of physical variables did not differ across continents for most traits, confirming our predictions qualitatively and quantitatively. Therefore, despite phylogenetic and historical differences between continents, fish communities of France and Virginia exhibit convergence in biological traits related to hydraulics and geomorphology. This convergence reflects morphological and behavioral adaptations to physical stress in streams. This study supports the existence of a habitat template for ecological strategies. Some key quantitative variables that define this habitat template can be identified by characterizing how individual organisms use their physical environment, and by using dimensionless physical variables that reveal common energetic properties in different systems. Overall, quantitative tests of community convergence are efficient tools to demonstrate that some community traits are predictable from environmental features.

Understanding the ecological dimensions of drought is critical for predicting how humans and nature will be affected by the expected increased prevalence of drought in the future. We tested life‐history‐based predictions for fish assemblage responses to drought using retrospective analysis of long‐term (1986–2003) fish surveys from two streams in the Appalachian Mountains of North America. We hypothesised that (1) fish assemblage composition would correlate with wet and dry hydrologic conditions as assemblages fluctuated within a loose equilibrium and (2) life‐history traits of fishes would correlate with dry versus wet periods such that opportunistic life‐history strategists would dominate during drought. Results showed fish assemblage changes in Little River and Cataloochee Creek correlated with drought severity measured one year prior to fish surveys. Fish assemblages at all three sampling sites in Little River and two sites in Cataloochee Creek fluctuated within a loose equilibrium, while the remaining two sites in Cataloochee Creek indicated directional change. Life‐history traits for fishes in Cataloochee Creek correlated with one‐year time lag fluctuations in drought caused by opportunistic species being dominant during drought and periodic/equilibrium species dominant during wet periods. Time series plots of fish abundances aggregated by life‐history strategy revealed dominance of opportunistic species emerged at the onset of a multi‐year drought spanning 1998–2004, particularly for the two sites undergoing directional change. Our work provides empirical evidence for theoretical linkages between life history and environmental fluctuations and can ultimately be used to predict stream fish community response to future drought regimes.

We compared the life-history traits of native and invasive fish species from Catalan streams in order to identify the characters of successful invasive fish species. Most of the exotic fish species were characterized by large size, long longevity, late maturity, high fecundity, few spawnings per year, and short reproductive span, whereas Iberian native species exhibited predominantly the opposite suite of traits. Species native to the southeastern Pyrenees watershed were also significantly different from species native to the rest of the Iberian Peninsula but not native to this watershed. Iberian exotic species come predominantly from large river basins, whereas Catalan streams (and other small, coastal river basins) correspond to basins and streams of a smaller size and different hydrology, with differences in species composition and life-history traits of fish. The occurrence and spread of invasive species was not significantly related to life-history traits but to introduction date. The successful predi ction of future invasive species is limited due to small differences in life-history and ecological traits between native and exotic species. Fecundity, age at maturity, water quality flexibility, tolerance to pollution and habitat seem the most discriminating life-history variables.

Biodiversity is a multifaceted concept for which the functional component, that influences many aspects of ecosystem functioning, remains difficult to characterise. In several taxa, many studies rely on morphological traits as proxies of species ecological attributes. However, the extent to which morphology provides a good surrogate of ecological traits has rarely been acknowledged.Here, we used morphological and ecological trait databases to test whether morphological traits are linked to ecological traits and can be used to assess various biodiversity facets in 254 stream fish communities across France.We firstly computed co‐inertia analyses to test the association between morphological and ecological traits related to either habitat use, feeding or life‐history strategies. We then computed a suite of diversity indices to investigate the relationship between morphological and ecological indices.We found a strong association between morphological and ecological traits, and significant relationships between diversity indices computed on morphological and ecological traits. However, these relationships varied according to the index and type of ecological traits, with the strongest relationship observed with traits related to habitat use.Although these results highlight that the use of morphological data is a promising way to understand ecological diversity, they also reveal that the choice of diversity index and the type of ecological traits targeted are particularly important and need to be taken into consideration.

Prairie stream fish assemblages experience seasonal shifts in habitat conditions, often coupled with powerful disturbance regimes, where systems fluctuate widely between drought and deluge. These abiotic drivers, more often than biotic factors, structure fish assemblages. Scouring floods may reset stream communities to early successional seres, and droughts may progressively eliminate species from communities. Therefore, fish assemblage structure resilience depends on recolonisation and reproductive success following a disturbance. Many plains fishes have life history traits that facilitate the wide dispersal and immediate contribution of offspring following disturbance events. However, descriptions of temporal and spatial patterns in larval assemblage structure in prairie streams are scarce, and few studies evaluate the effect of disturbances on reproduction. The objectives of this research were to (i) study the timing and contribution of native prairie stream fishes to drifting larval fish assemblages in three eastern South Dakota basins (Vermillion, James and Big Sioux) and (ii) describe larval fish responses to floods in the James River. We sampled drifting larval fish diversity during the spawning season from 2010 to 2012 throughout each basin. We also examined the effect of flooding on larval fish assemblage structure using data collected near the confluence of the James and Missouri rivers during baseflow (2003, 2004 and 2012) and flood conditions (2010). Subtle variability in spawning traits of prairie fishes and an apparent seasonal progression of larval fish taxa that secondarily respond to disturbance regimes suggests reproductive efforts are adapted to maximise the potential for successful recruitment in these severe systems.

<em>Abstract</em>.—We develop a conceptual model of habitat factors, operating as filters at multiple spatial and temporal scales that structure local fish assemblages in flowing waters of Arctic North America. Following the habitat template approach, we classify streams into six types based on environmental gradients, including ice formation, flow and temperature regimes, sediment dynamics, and edaphic factors. We discuss likely mechanisms influencing Arctic stream fish assemblages among the six stream types and consider this habitat template in the context of regional and local filters. There is a general increase in species richness from the eastern Arctic to Alaska that is inversely proportional to the distance from the Beringian glacial refugium. The fish fauna of Arctic streams comprises nine different families, although 14 of 24 species (58%) are members of Salmonidae. Certain life history traits are predicted to be advantageous in a periglacial environment known for cold, impoverished environments with highly variable yet generally predictable ice conditions. Examples of adaptive traits displayed by Arctic fishes include fall spawning, relatively large egg size, longevity, habitat and diet generalists, and efficient metabolism. Life history theory predicts that periodic strategists should dominate in the Arctic environment. Nevertheless, in the face of a rapidly changing climate and increasing resource development, there is much need for additional research on relationships between the traits of Arctic stream fishes and their habitat.

<em>Abstract</em>.—Species traits are used to study the functional organization of fish communities for a range of reasons, from simply reducing data dimensionality to providing mechanistic explanations for observed variation in communities. Ecological and life history traits have been used to understand the basic ecology of fishes and predict (1) species and community responses to habitat and climate alteration, and (2) species extinction, species invasion, and community homogenization. Many approaches in this arena have been developed during the past three decades, but they often have not been integrated with related ecological concepts or subdisciplines, which has led to confusion in terminology. We review 102 studies of species traits and then summarize patterns in traits being used and questions being addressed with trait-based approaches. Overall, studies of fish–habitat relationships that apply habitat templates and hierarchical filters dominate our sample; the most frequently used traits are related to feeding. We define and show the relationships among key terms such as fundamental and realized niches; functional traits, performance, and fitness; tactic, trait-state, syndromes, and strategies; and guilds and functional groups. We propose accelerating research to (1) quantify trait plasticity, (2) identify traits useful for testing ecological hypotheses, (3) model habitat and biotic interactions in communities while explicitly accounting for phylogenetic relationships, (4) explore how traits control community assembly, and (5) document the importance of traits in fish– community responses to anthropogenic change and in delivering ecosystem services. Further synthesis of these topics is still needed to develop concepts, models, and principles that can unify the disparate approaches taken in trait-based analysis of fish communities, link fish community ecology to general community ecology, and inform sustainable management of ecosystems.

Biological invasions are of special conservation concern in the Iberian Peninsula and other regions with high levels of endemism. Environmental variability, such as the seasonal fluctuations of Mediterranean streams, is a key factor that affects the spread of aquatic species in novel habitats. Fish parasites have a great potential to reflect such changes in the habitat features of freshwater ecosystems. The aim of this study consisted of seasonally analysing the health status and parasitological traits of non-native fish in Iberian waters. In particular, a strongly invasive population of Languedoc minnow Phoxinus septimaniae (leuciscid species native to south-east France) was assessed in Tordera Stream (north-eastern Iberian Peninsula, Mediterranean conditions). Fish were sampled in April, July, and October 2023 by electrofishing. Health status (external/internal organs) was significantly better in autumn (HAI = 28.8) than spring (HAI = 35.6). Life-cycle complexity was higher in spring (LCI = 1.98), whereas parasite abundance and Shannon diversity were significantly lower in autumn (TA = 19.6 and H' = 2.15, respectively). In October (more 'benign' environmental conditions in Iberian streams), minnows could display elevated foraging activity, with fish increasing their health condition and level of parasite resistance/tolerance. Overall results showed a particular seasonal profile of health and parasite infra-communities that allow this minnow species to thrive under highly fluctuating habitat conditions. This information could help environmental managers to control non-native fish in Mediterranean streams.

Life-history traits of invasive fish in small Mediterranean streams

Water depth and physical structure are important components of habitat complexity for stream fishes. Experiments in a semi-natural stream, containing four depth and structure treatments, quantified the effect of these two habitat components on the distribution, growth, and survival of five fish species common to coastal streams of Washington State. When fishes were permitted to move freely among the various habitat types, most species and age-classes were underrepresented in shallow pools lacking structure. In some cases water depth or physical structure alone appeared to explain these distributions. However, assemblage- and species-level responses were strongly influenced by the combined effects of depth and structure. A subsequent 30-d experiment revealed that mortality (likely due to bird predation) of water-column species using the simplest habitat type was as much as 50% greater than in the other treatments. However, there were no differences in survival of the benthic species, coastrange sculpin (Cottus aleuticus), among the different treatments, nor were there significant differences in growth among treatments for any species. These results indicate that predation risk may help to explain the importance of both deep water and physical structure to fishes in coastal streams in Washington. The results of our habitat-selection experiment also support the growing view that a community-level approach may be more effective than the single-species approach in evaluating the effects of human activities on stream fishes.

Individual differences in cognition are widespread across animal taxa, and despite progress in identifying possible mechanisms shaping evolution and maintenance of this variation, significant gaps in our understanding persist. A crucial but often overlooked factor is the covariation between cognitive traits and traits from other domains, which can impose evolutionary constraints and drive indirect selection, shaping cognitive variance. To explore this factor, the present study investigated the covariation between cognitive, behavioral, physiological, and life-history traits in a wild-caught fish, the tench (Tinca tinca). We focused on five cognitive traits: motor lateralization, visual lateralization, spatial learning, cognitive flexibility, and memory. These cognitive traits were interrelated, with for instance, faster learners exhibiting lower cognitive flexibility. Additionally, cognitive traits correlated with several behavioral and physiological traits. Using a network analysis, we demonstrated the relevance of cognitive variance in the relationship among traits. Visual lateralization was central in the network, connecting subnetworks including behavioral and physiological traits. Memory was strongly linked to physiological traits. Last, spatial learning, cognitive flexibility, and visual lateralization were the traits with the highest impact in the whole network. Overall, our study not only highlights the complex associations between cognitive, behavioral and physiological, but also underscores the critical role of cognition in these relationships. This finding supports the idea that the evolution and maintenance of cognitive variance are both influenced by, and exert remarkably influence on, the variation in other traits.

<em>Abstract</em>.—Ecological communities are structured by a combination of stochastic and deterministic processes, the latter including both abiotic factors and biotic interactions such as predation. Many studies, mostly in relatively stable ecosystems such as lakes, have demonstrated top-down effects on community structure and function. Communities or species in dynamic nonequilibrium ecosystems such as streams may also respond strongly to predation pressure. In this chapter, we review experimental research on effects of predation on fish assemblages in lotic systems, focusing on developments in the decades since Matthews and Heins (1987). Direct experimental evidence indicates that predators strongly affect lotic fish assemblages via direct and indirect pathways of lethal and nonlethal interactions. Across studies, predators consistently reduced prey density, caused changes in prey habitat use, and decreased prey activity levels. Predators may also affect aspects of prey life history and reproduction in streams, and the presence of multiple predator species may result in prey risk enhancement. Our review identified five areas needing additional research that may lead to further advances in stream fish community ecology: (1) linking predation experiments with theoretical models of fish assemblage structure and function, (2) quantifying functional traits of predators and prey, (3) manipulating whole assemblages and testing multispecies interactions, (4) understanding the role of predation in human-modified ecosystems, and (5) application of analytical approaches that facilitate integration among these areas of research as well as with observational field studies.

Rapid phenotypic diversification during biological invasions can either arise by adaptation to alternative environments or by adaptive phenotypic plasticity. Where experimental evidence for adaptive plasticity is common, support for evolutionary diversification is rare. Here, we performed a controlled laboratory experiment using full-sib crosses between ecologically divergent threespine stickleback populations to test for a genetic basis of adaptation. Our populations are from two very different habitats, lake and stream, of a recently invaded range in Switzerland and differ in ecologically relevant morphological traits. We found that in a lake-like food treatment lake fish grow faster than stream fish, resembling the difference among wild type individuals. In contrast, in a stream-like food treatment individuals from both populations grow similarly. Our experimental data suggest that genetically determined diversification has occurred within less than 140 years after the arrival of stickleback in our studied region.

In this study, we assessed the effects of multiple land uses and local habitat variables on the composition of fish functional trophic groups (FTG’s) and on the ecomorphological traits of fish in Amazonian streams. We evaluated land use types and local habitat variables in 26 streams distributed within a land use gradient. Land use and habitat variables affected the composition of FTG’s, as evidenced by the increased abundance of diurnal channel drift feeders in areas with high dissolved oxygen and deeper thalweg. At the same time, diurnal surface pickers, as well as diggers, and ambush and stalker predators were more abundant in streams with higher canopy density. Only habitat variables affected the ecomorphological characteristics of the species. Fish with higher values of relative caudal peduncle length were positively associated with high canopy density, while fish with greater relative mouth width were negatively associated with the variables impact in the riparian zone and cover of fish shelter. The stream fish functional structure was mainly affected by the impacts caused to the local habitat resulting from different land uses. Thus, preserving forest remnants, as well as recovering degraded areas, is essential for the maintenance of aquatic biodiversity in the region.