Shortcomings of reusing species interaction networks created by different sets of researchers.

Given the requisite cost associated with observing species interactions, ecologists often reuse species interaction networks created by different sets of researchers to test their hypotheses regarding how ecological processes drive network topology. Yet, topological properties identified across these networks may not be sufficiently attributable to ecological processes alone as often assumed. Instead, much of the totality of topological differences between networks-topological heterogeneity-could be due to variations in research designs and approaches that different researchers use to create each species interaction network. To evaluate the degree to which this topological heterogeneity is present in available ecological networks, we first compared the amount of topological heterogeneity across 723 species interaction networks created by different sets of researchers with the amount quantified from non-ecological networks known to be constructed following more consistent approaches. Then, to further test whether the topological heterogeneity was due to differences in study designs, and not only to inherent variation within ecological networks, we compared the amount of topological heterogeneity between species interaction networks created by the same sets of researchers (i.e., networks from the same publication) with the amount quantified between networks that were each from a unique publication source. We found that species interaction networks are highly topologically heterogeneous: while species interaction networks from the same publication are much more topologically similar to each other than interaction networks that are from a unique publication, they still show at least twice as much heterogeneity as any category of non-ecological networks that we tested. Altogether, our findings suggest that extra care is necessary to effectively analyze species interaction networks created by different researchers, perhaps by controlling for the publication source of each network.

Rapid monitoring of ecological persistence.

Effective conservation of ecological communities requires accurate and up-to-date information about whether species are persisting or declining to extinction. The persistence of an ecological community is supported by its underlying network of species interactions. While the persistence of the network supporting the whole community is the most relevant scale for conservation, in practice, only small subsets of these networks can be monitored. There is therefore an urgent need to establish links between the small snapshots of data conservationists can collect, and the "big picture" conclusions about ecosystem health demanded by policymakers, scientists, and societies. Here, we show that the persistence of small subnetworks (motifs) in isolation-that is, their persistence when considered separately from the larger network of which they are a part-is a reliable probabilistic indicator of the persistence of the network as a whole. Our methods show that it is easier to detect if an ecological community is not persistent than if it is persistent, allowing for rapid detection of extinction risk in endangered systems. Our results also justify the common practice of predicting ecological persistence from incomplete surveys by simulating the population dynamics of sampled subnetworks. Empirically, we show that our theoretical predictions are supported by data on invaded networks in restored and unrestored areas, even in the presence of environmental variability. Our work suggests that coordinated action to aggregate information from incomplete sampling can provide a means to rapidly assess the persistence of entire ecological networks and the expected success of restoration strategies.

Scale of population synchrony confirms macroecological estimates of minimum viable range size.

Global ecosystems are facing a deepening biodiversity crisis, necessitating robust approaches to quantifying species extinction risk. The lower limit of the macroecological relationship between species range and body size has long been hypothesized as an estimate of the relationship between the minimum viable range size (MVRS) needed for species persistence and the organismal traits that affect space and resource requirements. Here, we perform the first explicit test of this assumption by confronting the MVRS predicted by the range-body size relationship with an independent estimate based on the scale of synchrony in abundance among spatially separated populations of riverine fish. We provide clear evidence of a positive relationship between the scale of synchrony and species body size, and strong support for the MVRS set by the lower limit of the range-body size macroecological relationship. This MVRS may help prioritize first evaluations for unassessed or data-deficient taxa in global conservation assessments.

Generalism drives abundance: A computational causal discovery approach.

A ubiquitous pattern in ecological systems is that more abundant species tend to be more generalist; that is, they interact with more species or can occur in wider range of habitats. However, there is no consensus on whether generalism drives abundance (a selection process) or abundance drives generalism (a drift process). As it is difficult to conduct direct experiments to solve this chicken-and-egg dilemma, previous studies have used a causal discovery method based on formal logic and have found that abundance drives generalism. Here, we refine this method by correcting its bias regarding skewed distributions, and employ two other independent causal discovery methods based on nonparametric regression and on information theory, respectively. Contrary to previous work, all three independent methods strongly indicate that generalism drives abundance when applied to datasets on plant-hummingbird communities and reef fishes. Furthermore, we find that selection processes are more important than drift processes in structuring multispecies systems when the environment is variable. Our results showcase the power of the computational causal discovery approach to aid ecological research.

How network size strongly determines trophic specialisation: A technical comment on Luna et al. (2022).

Luna et al. (2022) concluded that the environment contributes to explaining specialisation in open plant-pollinator networks. When reproducing their study, we instead found that network size alone largely explained the variation in their specialisation metrics. Thus, we question whether empirical network specialisation is driven by the environment.

Metapopulation persistence can be inferred from incomplete surveys.

Habitat destruction and fragmentation are principal causes of species loss. While a local population might go extinct, a metapopulation-populations inhabiting habitat patches connected by dispersal-can persist regionally by recolonizing empty patches. To assess metapopulation persistence, two widely adopted indicators in conservation management are metapopulation capacity and patch importance. However, we face a fundamental limitation in that assessing metapopulation persistence requires that we survey or sample all the patches in a landscape: often these surveys are logistically challenging to conduct and repeat, which raises the question whether we can learn enough about the metapopulation persistence from an incomplete survey. Here, we provide a robust statistical approach to infer metapopulation capacity and patch importance by sampling a portion of all patches. We provided analytic arguments on why the metapopulation capacity and patch importance can be well predicted from sub-samples of habitat patches. Full-factorial simulations with more complex models corroborate our analytic predictions. We applied our model to an empirical metapopulation of mangrove hummingbirds (Amazilia boucardi). On the basis of our statistical framework, we provide some sampling suggestion for monitoring metapopulation persistence. Our approach allows for rapid and effective inference of metapopulation persistence from incomplete patch surveys.

Managing for the unexpected: Building resilient forest landscapes to cope with global change.

Natural disturbances exacerbated by novel climate regimes are increasing worldwide, threatening the ability of forest ecosystems to mitigate global warming through carbon sequestration and to provide other key ecosystem services. One way to cope with unknown disturbance events is to promote the ecological resilience of the forest by increasing both functional trait and structural diversity and by fostering functional connectivity of the landscape to ensure a rapid and efficient self-reorganization of the system. We investigated how expected and unexpected variations in climate and biotic disturbances affect ecological resilience and carbon storage in a forested region in southeastern Canada. Using a process-based forest landscape model (LANDIS-II), we simulated ecosystem responses to climate change and insect outbreaks under different forest policy scenarios-including a novel approach based on functional diversification and network analysis-and tested how the potentially most damaging insect pests interact with changes in forest composition and structure due to changing climate and management. We found that climate warming, lengthening the vegetation season, will increase forest productivity and carbon storage, but unexpected impacts of drought and insect outbreaks will drastically reduce such variables. Generalist, non-native insects feeding on hardwood are the most damaging biotic agents for our region, and their monitoring and early detection should be a priority for forest authorities. Higher forest diversity driven by climate-smart management and fostered by climate change that promotes warm-adapted species, might increase disturbance severity. However, alternative forest policy scenarios led to a higher functional and structural diversity as well as functional connectivity-and thus to higher ecological resilience-than conventional management. Our results demonstrate that adopting a landscape-scale perspective by planning interventions strategically in space and adopting a functional trait approach to diversify forests is promising for enhancing ecological resilience under unexpected global change stressors.

The geography of metapopulation synchrony in dendritic river networks.

Dendritic habitats, such as river ecosystems, promote the persistence of species by favouring spatial asynchronous dynamics among branches. Yet, our understanding of how network topology influences metapopulation synchrony in these ecosystems remains limited. Here, we introduce the concept of fluvial synchrogram to formulate and test expectations regarding the geography of metapopulation synchrony across watersheds. By combining theoretical simulations and an extensive fish population time-series dataset across Europe, we provide evidence that fish metapopulations can be buffered against synchronous dynamics as a direct consequence of network connectivity and branching complexity. Synchrony was higher between populations connected by direct water flow and decayed faster with distance over the Euclidean than the watercourse dimension. Likewise, synchrony decayed faster with distance in headwater than mainstem populations of the same basin. As network topology and flow directionality generate fundamental spatial patterns of synchrony in fish metapopulations, empirical synchrograms can aid knowledge advancement and inform conservation strategies in complex habitats.

Network ecology in dynamic landscapes.

Network ecology is an emerging field that allows researchers to conceptualize and analyse ecological networks and their dynamics. Here, we focus on the dynamics of ecological networks in response to environmental changes. Specifically, we formalize how network topologies constrain the dynamics of ecological systems into a unifying framework in network ecology that we refer to as the 'ecological network dynamics framework'. This framework stresses that the interplay between species interaction networks and the spatial layout of habitat patches is key to identifying which network properties (number and weights of nodes and links) and trade-offs among them are needed to maintain species interactions in dynamic landscapes. We conclude that to be functional, ecological networks should be scaled according to species dispersal abilities in response to landscape heterogeneity. Determining how such effective ecological networks change through space and time can help reveal their complex dynamics in a changing world.

Network analysis can guide resilience-based management in forest landscapes under global change.

Forests are projected to undergo dramatic compositional and structural shifts prompted by global changes, such as climatic changes and intensifying natural disturbance regimes. Future uncertainty makes planning for forest management exceptionally difficult, demanding novel approaches to maintain or improve the ability of forest ecosystems to respond and rapidly reorganize after disturbance events. Adopting a landscape perspective in forest management is particularly important in fragmented forest landscapes where both diversity and connectivity play key roles in determining resilience to global change. In this context, network analysis and functional traits combined with ecological dynamic modeling can help evaluate changes in functional response diversity and connectivity within and among forest stands in fragmented landscapes. Here, we coupled ecological dynamic modeling with functional traits analysis and network theory to analyze forested landscapes as an interconnected network of forest patches. We simulated future forest landscape dynamics in a large landscape in southern Quebec, Canada, under a combination of climate, disturbance, and management scenarios. We depicted the landscape as a functional network, assessed changes in future resilience using indicators at multiple spatial scales, and evaluated if current management practices are suitable for maintaining resilience to simulated changes in regimes. Our results show that climate change would promote forest productivity and favor heat-adapted deciduous species. Changes in natural disturbances will likely have negative impacts on native conifers and will drive changes in forest type composition. Climate change negatively impacted all resilience indicators and triggered losses of functional response diversity and connectivity across the landscape with undesirable consequences on the capacity of these forests to adapt to global change. Also, current management strategies failed to promote resilience at different spatial levels, highlighting the need for a more active and thoughtful approach to forest management under global change. Our study demonstrates the usefulness of combining dynamic landscape-scale simulation modeling with network analyses to evaluate the possible impacts of climate change as well as human and natural disturbances on forest resilience under global change.

Incorporating putatively neutral and adaptive genomic data into marine conservation planning.

The availability of genomic data for an increasing number of species makes it possible to incorporate evolutionary processes into conservation plans. Recent studies show how genetic data can inform spatial conservation prioritization (SCP), but they focus on metrics of diversity and distinctness derived primarily from neutral genetic data sets. Identifying adaptive genetic markers can provide important information regarding the capacity for populations to adapt to environmental change. Yet, the effect of including metrics based on adaptive genomic data into SCP in comparison to more widely used neutral genetic metrics has not been explored. We used existing genomic data on a commercially exploited species, the giant California sea cucumber (Parastichopus californicus), to perform SCP for the coastal region of British Columbia (BC), Canada. Using a RAD-seq data set for 717 P. californicus individuals across 24 sampling locations, we identified putatively adaptive (i.e., candidate) single nucleotide polymorphisms (SNPs) based on genotype-environment associations with seafloor temperature. We calculated various metrics for both neutral and candidate SNPs and compared SCP outcomes with independent metrics and combinations of metrics. Priority areas varied depending on whether neutral or candidate SNPs were used and on the specific metric used. For example, targeting sites with a high frequency of warm-temperature-associated alleles to support persistence under future warming prioritized areas in the southern coastal region. In contrast, targeting sites with high expected heterozygosity at candidate loci to support persistence under future environmental uncertainty prioritized areas in the north. When combining metrics, all scenarios generated intermediate solutions, protecting sites that span latitudinal and thermal gradients. Our results demonstrate that distinguishing between neutral and adaptive markers can affect conservation solutions and emphasize the importance of defining objectives when choosing among various genomic metrics for SCP.

Incorporación de Datos Genómicos Putativamente Neutros y Adaptativos dentro de la Planeación de la Conservación Marina Resumen La disponibilidad de los datos genómicos para un número creciente de especies posibilita la incorporación de los procesos evolutivos dentro de los planes de conservación. Los estudios recientes muestran cómo los datos genéticos pueden informar a la priorización de la conservación espacial (PCE) pero tienden a enfocarse más en las medidas de la diversidad y la distinción derivadas principalmente de los conjuntos de datos genéticos neutrales. La identificación de los marcadores genéticos adaptativos puede proporcionar información importante con respecto a la capacidad de las poblaciones para adaptarse al cambio ambiental. Aun así, no se ha explorado el efecto de la inclusión de las medidas basadas en los datos genéticos adaptativos dentro de la PCE y cómo se comparan con las medidas genéticas neutrales de uso más amplio. Usamos datos genómicos existentes sobre una especie de explotación comercial, el pepino de mar gigante de California (Parastichopus californicus), para realizar la PCE para la región costera de la Columbia Británica (BC) en Canadá. Usamos un conjunto de datos RAD-seq para 717 individuos de la especie P. californicus en 24 localidades de muestreo para identificar los polimorfismos de un solo nucleótido (PSNs) putativamente adaptativos (es decir, candidatos) con base en las asociaciones genotipo-ambiente manifestadas con la temperatura del fondo marino. Calculamos varias medidas para los PSNs neutrales y los PSNs candidatos y comparamos los resultados de la PCE con medidas independientes y con combinaciones de medidas. Las áreas prioritarias variaron dependiendo de si se usaron los SNP neutrales o los candidatos y de la medida específica que se utilizó. Por ejemplo, enfocarse en sitios con una frecuencia alta de alelos asociados con agua cálida para fortalecer la persistencia frente al futuro calentamiento prioriza las áreas en la región del sur de la costa. Al contrario, enfocarse en sitios con una alta heterocigosidad esperada en los loci de los candidatos para fortalecer la persistencia frente a la incertidumbre ambiental prioriza las áreas en la parte norte de la costa. Cuando combinamos las medidas, todos los escenarios generaron soluciones intermedias, protegiendo así los sitios que abarcan gradientes latitudinales y gradientes térmicos. Nuestros resultados demuestran que la distinción entre los marcadores neutrales y los adaptativos puede afectar las soluciones de conservación y también enfatizan la importancia de la definición de los objetivos cuando se elige entre varias medidas genómicas para la PCE.

Accounting for stochasticity in demographic compensation along the elevational range of an alpine plant.

Demographic compensation arises when vital rates change in opposite directions across populations, buffering the variation in population growth rates, and is a mechanism often invoked to explain the stability of species geographic ranges. However, studies on demographic compensation have disregarded the effects of temporal variation in vital rates and their temporal correlations, despite theoretical evidence that stochastic dynamics can affect population persistence in temporally varying environments. We carried out a seven-year-long demographic study on the perennial plant Arabis alpina (L.) across six populations encompassing most of its elevational range. We discovered demographic compensation in the form of negative correlations between the means of plant vital rates, but also between their temporal coefficients of variation, correlations and elasticities. Even if their contribution to demographic compensation was small, this highlights a previously overlooked, but potentially important, role of stochastic processes in stabilising population dynamics at range margins.

Moderate disturbances accelerate forest transition dynamics under climate change in the temperate-boreal ecotone of eastern North America.

Several temperate tree species are expected to migrate northward and colonize boreal forests in response to climate change. Tree migrations could lead to transitions in forest types, but these could be influenced by several non-climatic factors, such as disturbances and soil conditions. We analysed over 10,000 forest inventory plots, sampled from 1970 to 2018 in meridional Québec, Canada, to identify what environmental conditions promote or prevent regional-scale forest transitions. We used a continuous-time multi-state Markov model to quantify the probabilities of transitions between forest states (temperate, boreal, mixed, pioneer) as a function of climate (mean temperature and climate moisture index during the growing season), soil conditions (pH and drainage) and disturbances (severity levels of natural disturbances and logging). We further investigate how different disturbance types and severities impact forests' short-term transient dynamics and long-term equilibrium using properties of Markov transition matrices. The most common transitions observed during the study period were from mixed to temperate states, as well as from pioneer to boreal forests. In our study, transitions were mainly driven by natural and anthropogenic disturbances and secondarily by climate, whereas soil characteristics exerted relatively minor constraints. While major disturbances only promoted transitions to the pioneer state, moderate disturbances increased the probability of transition from mixed to temperate states. Long-term projections of our model under the current environmental conditions indicate that moderate disturbances would promote a northward shift of the temperate forest. Moreover, disturbances reduced turnover and convergence time for all transitions, thereby accelerating forest dynamics. Contrary to our expectation, mixed to temperate transitions were not driven by temperate tree recruitment but by mortality and growth. Overall, our results suggest that moderate disturbances could catalyse rapid forest transitions and accelerate broad-scale biome shifts.

Evaluating forest resilience to global threats using functional response traits and network properties.

Ecosystem functions provided by forests are threatened by direct and indirect effects of global change drivers such as climate warming land-use change, biological invasions, and shifting natural disturbance regimes. To develop resilience-based forest management, new tools and methods are needed to quantitatively estimate forest resilience to management and future natural disturbances. We propose a multidimensional evaluation of ecological resilience based on species functional response traits (e.g., functional response diversity and functional redundancy) and network properties of forested patches (e.g., connectivity, modularity, and centrality). Using a fragmented rural landscape in temperate south-eastern Canada as a reference landscape, we apply our multidimensional approach to evaluate two alternative management strategies at three levels of intensity: (1) functional enrichment of current forest patches and (2) multi-species plantations in previously non-forested patches. Within each management strategy, planted species are selected to maximize functional diversity, drought tolerance, or pest resistance. We further compare how ecological resilience under these alternative management strategies responds to three simulated disturbances: drought, pest outbreak, and timber harvesting. We found that both management strategies enhance resilience at the landscape scale by increasing functional response diversity and connectivity. Specifically, when the less functionally diverse patches are prioritized for management, functional enrichment is more effective than the establishment of new multi-species plantations in increasing resilience. In addition, randomly allocated multi-species plantations increased connectivity more than those allocated in riparian areas. Our results show that across various management strategies, planting species to enhance biodiversity led to the highest increase in functional response diversity while planting pest-resistant species led to the highest increase in landscape connectivity. Planting biodiversity-enhancing species (i.e., species that maximize functional diversity) mitigated drought effects equally well as planting with drought-tolerant species. Our multidimensional approach facilitates the characterization at the landscape scale of forest resilience to disturbances using both functional diversity and network properties while accounting for the importance of response traits to future disturbances. The simulation approach we used can be applied to forest landscapes across different biomes for the evaluation and comparison of forest management initiatives to enhance resilience.

The mechanisms generating community phylogenetic patterns change with spatial scale.

Studies that test community assembly hypotheses in observational communities frequently evaluate patterns for plots or entire communities, yet studies that examine assembly patterns across spatial scales show that they are greatly influenced by scale. Here, we test the spatial dependency of patterns of relatedness and plant height for all individual herbaceous plants along five 40-m old-field transects (Southern Ontario, Canada). We identified each individual plant and measured its distance along the transect and its height, and we constructed a molecular phylogeny for all observed species. To uncover the scale at which community phylogenetic and trait similarities shift, we used partial Mantel correlograms and distance-based Moran Eigenvector Maps (dbMEMs). We found that communities shift from significantly overdispersed at relatively smaller scales (i.e., < 15 m) to spatially clustered at larger scales, showing that assembly mechanism influence depends on scale of observation. This pattern was observed for both phylogeny and height, but was the strongest when considering phylogeny only. These results reveal the importance of spatial scale when examining community phylogenetic or trait patterns, where finding support for one assembly mechanism at a single scale does not necessarily mean that other mechanisms are also not important for structuring community composition and diversity.

Putatively adaptive genetic variation in the giant California sea cucumber (Parastichopus californicus) as revealed by environmental association analysis of restriction-site associated DNA sequencing data.

Understanding the spatial scale of local adaptation and the factors associated with adaptive diversity are important objectives for ecology and evolutionary biology, and have significant implications for effective conservation and management of wild populations and natural resources. In this study, we used an environmental association analysis to identify important bioclimatic variables correlated with putatively adaptive genetic variation in a benthic marine invertebrate-the giant California sea cucumber (Parastichopus californicus)-spanning coastal British Columbia and southeastern Alaska. We used a redundancy analysis (RDA) with 3,699 single nucleotide polymorphisms (SNPs) obtained using RAD sequencing to detect candidate markers associated with 11 bioclimatic variables, including sea bottom and surface conditions, across two spatial scales (entire study area and within subregions). At the broadest scale, RDA revealed 59 candidate SNPs, 86% of which were associated with mean bottom temperature. Similar patterns were identified when population structure was accounted for. Additive polygenic scores, which provide a measure of the cumulative signal across all candidate SNPs, were strongly correlated with mean bottom temperature, consistent with spatially varying selection across a thermal gradient. At a finer scale, 23 candidate SNPs were detected, primarily associated with surface salinity (26%) and bottom current velocity (17%). Our findings suggest that environmental variables may play a role as drivers of spatially varying selection for P. californicus. These results provide context for future studies to evaluate the genetic basis of local adaptation in P. californicus and help inform the relevant scales and environmental variables for in situ field studies of putative adaptive variation in marine invertebrates.

Asymmetric oceanographic processes mediate connectivity and population genetic structure, as revealed by RADseq, in a highly dispersive marine invertebrate (Parastichopus californicus).

Marine populations are typically characterized by weak genetic differentiation due to the potential for long-distance dispersal favouring high levels of gene flow. However, strong directional advection of water masses or retentive hydrodynamic forces can influence the degree of genetic exchange among marine populations. To determine the oceanographic drivers of genetic structure in a highly dispersive marine invertebrate, the giant California sea cucumber (Parastichopus californicus), we first tested for the presence of genetic discontinuities along the coast of North America in the northeastern Pacific Ocean. Then, we tested two hypotheses regarding spatial processes influencing population structure: (i) isolation by distance (IBD: genetic structure is explained by geographic distance) and (ii) isolation by resistance (IBR: genetic structure is driven by ocean circulation). Using RADseq, we genotyped 717 individuals from 24 sampling locations across 2,719 neutral SNPs to assess the degree of population differentiation and integrated estimates of genetic variation with inferred connectivity probabilities from a biophysical model of larval dispersal mediated by ocean currents. We identified two clusters separating north and south regions, as well as significant, albeit weak, substructure within regions (FST  = 0.002, p = .001). After modelling the asymmetric nature of ocean currents, we demonstrated that local oceanography (IBR) was a better predictor of genetic variation (R2  = .49) than geographic distance (IBD) (R2  = .18), and directional processes played an important role in shaping fine-scale structure. Our study contributes to the growing body of literature identifying significant population structure in marine systems and has important implications for the spatial management of P. californicus and other exploited marine species.

Optimizing the choice of a spatial weighting matrix in eigenvector-based methods.

Eigenvector-mapping methods such as Moran's eigenvector maps (MEM) are derived from a spatial weighting matrix (SWM) that describes the relations among a set of sampled sites. The specification of the SWM is a crucial step, but the SWM is generally chosen arbitrarily, regardless of the sampling design characteristics. Here, we compare the statistical performances of different types of SWMs (distance-based or graph-based) in contrasted realistic simulation scenarios. Then, we present an optimization method and evaluate its performances compared to the arbitrary choice of the most-widely used distance-based SWM. Results showed that the distance-based SWMs generally had lower power and accuracy than other specifications, and strongly underestimated spatial signals. The optimization method, using a correction procedure for multiple tests, had a correct type I error rate, and had higher power and accuracy than an arbitrary choice of the SWM. Nevertheless, the power decreased when too many SWMs were compared, resulting in a trade-off between the gain of accuracy and the loss of power. We advocate that future studies should optimize the choice of the SWM using a small set of appropriate candidates. R functions to implement the optimization are available in the adespatial package and are detailed in a tutorial.

Human activities as a driver of spatial variation in the trophic structure of fish communities on Pacific coral reefs.

Anthropogenic activities such as land-use change, pollution and fishing impact the trophic structure of coral reef fishes, which can influence ecosystem health and function. Although these impacts may be ubiquitous, they are not consistent across the tropical Pacific Ocean. Using an extensive database of fish biomass sampled using underwater visual transects on coral reefs, we modelled the impact of human activities on food webs at Pacific-wide and regional (1,000s-10,000s km) scales. We found significantly lower biomass of sharks and carnivores, where there were higher densities of human populations (hereafter referred to as human activity); however, these patterns were not spatially consistent as there were significant differences in the trophic structures of fishes among biogeographic regions. Additionally, we found significant changes in the benthic structure of reef environments, notably a decline in coral cover where there was more human activity. Direct human impacts were the strongest in the upper part of the food web, where we found that in a majority of the Pacific, the biomass of reef sharks and carnivores were significantly and negatively associated with human activity. Finally, although human-induced stressors varied in strength and significance throughout the coral reef food web across the Pacific, socioeconomic variables explained more variation in reef fish trophic structure than habitat variables in a majority of the biogeographic regions. Notably, economic development (measured as GDP per capita) did not guarantee healthy reef ecosystems (high coral cover and greater fish biomass). Our results indicate that human activities are significantly shaping patterns of trophic structure of reef fishes in a spatially nonuniform manner across the Pacific Ocean, by altering processes that organize communities in both "top-down" (fishing of predators) and "bottom-up" (degradation of benthic communities) contexts.