Search

Environmental conditions can create spatial and temporal variability in growth and distribution processes, yet contemporary stock assessment methods often do not explicitly address the consequences of these patterns. For example, stock assessments often assume that body weight-at-age (i.e. size) is constant across the stocks’ range, and may thereby miss important spatio-temporal patterns. This is becoming increasingly relevant given climate-driven distributional shifts, because samples for estimating size-at-age can be spatially unbalanced and lead to biases when extrapolating into unsampled areas. Here, we jointly analysed data on the local abundance and size of walleye pollock (Gadus chalcogrammus) in the Bering Sea, to demonstrate a tractable first step in expanding spatially unbalanced size-at-age samples, while incorporating fine-scale spatial and temporal variation for inclusion in stock assessments. The data come from NOAA’s bottom trawl survey data and were evaluated using a multivariate spatio-temporal statistical model. We found extensive variation in size-at-age at fine spatial scales, though specific patterns differed between age classes. In addition to persistent spatial patterns, we also documented year-to-year differences in the spatial patterning of size-at-age. Intra-annual variation in the population-level size-at-age (used to generate the size-at-age matrix in the stock assessment) was largely driven by localized changes in fish size, while shifts in species distribution had a smaller effect. The spatio-temporal size-at-age matrix led to marginal improvement in the stock assessment fit to the survey biomass index. Results from our case study suggest that accounting for spatially unbalanced sampling improved stock assessment consistency. Additionally, it improved our understanding on the dynamics of how local and population-level demographic processes interact. As climate change affects fish distribution and growth, integrating spatiotemporally explicit size-at-age processes with anticipated environmental conditions may improve stock-assessment forecasts used to set annual harvest limits.

Ecologists and fisheries scientists are faced with forecasting the ecological responses of non-stationary processes resulting from climate change and other drivers. While much is known about temporal change, and resulting responses vis-à-vis species distributional shifts, less is known about how spatial variability in population structure changes through time in response to temporal trends in drivers. A population experiencing decreasing spatial variability would be expected to be more evenly spatially distributed over time, and an increasing trend would correspond to greater extremes or patchiness. We implement a new approach for modelling this spatiotemporal variability in the R package sdmTMB. As a real-world application, we focus on a long-term groundfish monitoring dataset, from the west coast of the USA. Focusing on the 36 species with the highest population densities, we compare our model with dynamic spatiotemporal variance to a model with constant spatiotemporal variance. Of the 36 species examined, 13 had evidence to support increasing patchiness, including darkblotched rockfish, lingcod, and petrale sole. Species appearing to be more uniformly spatially distributed over time included: Dover sole, Pacific ocean perch, and Dungeness crab. Letting spatiotemporal variation change through time generally results in small differences in population trend estimates, but larger estimated differences in precision.

Long-term data allow ecologists to assess trajectories of population abundance. Without this context, it is impossible to know whether a taxon is thriving or declining to extinction. For parasites of wildlife, there are few long-term data—a gap that creates an impediment to managing parasite biodiversity and infectious threats in a changing world. We produced a century-scale time series of metazoan parasite abundance and used it to test whether parasitism is changing in Puget Sound, United States, and, if so, why. We performed parasitological dissection of fluid-preserved specimens held in natural history collections for eight fish species collected between 1880 and 2019. We found that parasite taxa using three or more obligately required host species—a group that comprised 52% of the parasite taxa we detected—declined in abundance at a rate of 10.9% per decade, whereas no change in abundance was detected for parasites using one or two obligately required host species. We tested several potential mechanisms for the decline in 3+-host parasites and found that parasite abundance was negatively correlated with sea surface temperature, diminishing at a rate of 38% for every 1 °C increase. Although the temperature effect was strong, it did not explain all variability in parasite burden, suggesting that other factors may also have contributed to the long-term declines we observed. These data document one century of climate-associated parasite decline in Puget Sound—a massive loss of biodiversity, undetected until now.

Identifying changes in fish growth is important for accurate scientific advice used for fisheries management, because environmental change is affecting fish growth and size-at-age is a critical component of contemporary stock assessment methods. Growth-increment biochronologies are time series of growth-increments derived from hard parts of marine organisms that may reveal dynamics of somatic fish growth. Here we use time series of otolith increments of two fish stocks to fit and compare a biologically derived growth model and a generalized statistical model. Both models produced similar trajectories of annual growth trends, but the biologically based one was more precise and predicted smaller interannual fluctuations than the statistical model. The biologically based model strongly indicated covariance between anabolic and catabolic rates among individuals. Otolith size-at-age did not closely match fish length-at-age, and consequently the growth model could not accurately hindcast observed fish length-at-age. For these reasons, fitted growth dynamics from otolith biochronologies may best suited to identify growth rate fluctuations, understand past drivers of growth dynamics, and improve ecological forecast in the face of rapid environmental change.

There is a paucity of age data for chondrichthyan fishes owing, in large part, to limitations in traditional age estimation methods. Fourier transform near-infrared (FT-NIR) spectroscopy has shown promise as an alternative, more efficient method for acquiring age data from chondrichthyans. However, studies are limited to sharks in the southern hemisphere. We explored FT-NIR spectroscopy to predict age for a batoid species in the northern hemisphere. The longnose skate (Raja rhina) is one of a small number of batoids for which annual band periodicity in vertebral centra has been validated, allowing for traditional age estimation and making it an ideal candidate for this study. We fit a multivariate partial least-square predictive model between FT-NIR spectra collected from vertebral centra and traditional age estimates, and tested model predictive skill by using external validation. Using FT-NIR spectroscopy, we were able to predict age for longnose skates between the ages of 1 and 14 years with precision and bias near equal to those of traditional methods in less than a quarter of the time. These results support potential for FT-NIR spectroscopy to increase the amount of age data available for assessments used to inform the conservation and management of this sensitive group of species.

A major challenge in sustainability science is identifying targets that maximize ecosystem benefits to humanity while minimizing the risk of crossing critical system thresholds. One critical threshold is the biomass at which populations become so depleted that their population growth rates become negative—depensation. Here, we evaluate how the value of monitoring information increases as a natural resource spends more time near the critical threshold. This benefit emerges because higher monitoring precision promotes higher yield and a greater capacity to recover from overharvest. We show that precautionary buffers that trigger increased monitoring precision as resource levels decline may offer a way to minimize monitoring costs and maximize profits. In a world of finite resources, improving our understanding of the trade-off between precision in estimates of population status and the costs of mismanagement will benefit stakeholders that shoulder the burden of these economic and social costs.

Diet-based annual biomass indices can potentially use predator stomach contents to provide information about prey biomass and may be particularly useful for species that are otherwise poorly sampled, including ecologically important forage fishes. However, diet-based biomass indices may be sensitive to underlying ecological dynamics between predators and prey, such as predator functional responses and changes in overlap in space and time. To evaluate these factors, we fit spatio-temporal models to stomach contents of five Atlantic herring (Clupea harengus) predators and survey catch data for predators and Atlantic herring. We identified drivers of variation in stomach contents, evaluated spatial patterns in stomach content data, and produced predator-specific indices of seasonal Atlantic herring biomass. After controlling for spatio-temporal processes and predator length, diet-based indices of biomass shared similar decadal trends but varied substantially between predators and seasons on shorter time scales. Diet-based indices reflected prey biomass more than prey availability, but weak correlations indicated that not all biological processes were controlled for. Results provide potential guidance for developing diet-based biomass indices and contribute to a body of evidence demonstrating the utility of predator diet data to provide information about relative prey biomass.

We examined historical and contemporary trends for a suite of groundfish species in Puget Sound, Washington, USA, to ask how the groundfish community has responded following shifts in fishing regulations, climate, food web, and a growing human population in the surrounding watershed. We used contemporary data (1990-2017) from a standardized annual bottom trawl survey with historical logbook information (1948-1977) from a research vessel. We standardized data to account for spatial and temporal effects on catch rate by first fitting a Bayesian model to the contemporary data, and then using the posterior distributions of the covariates as prior distributions when fitting models to the historical data. We found that most of the 15 species chosen for analysis had highly variable population dynamics within both time periods, and that patterns of variability were similar in the contemporary and historical time period. Surprisingly, there was little evidence of community-wide recovery following regulations that first limited and then banned commercial bottom trawling in the late 1980s to mid-1990s. Spotted ratfish and English sole dominated catches in both time periods, and species that were common in the historical time period were also common in the contemporary time period. The absence of coherent community changes in response to fisheries regulations and other major social and ecological changes may signal the complex dynamics of an urban estuary that is subject to multiple external drivers, and speaks to the need to consider long-term dynamic behavior of populations and communities when establishing ecological indicators.

Median timing of reproduction in salmonid populations is generally consistent among years, reflecting long-term patterns of natural selection from characteristics of the local environment. However, altered selection from factors related to climate change or human intervention might shift timing over generations, with implications for the population’s persistence. To study these processes, we modeled median timing of redd (nest) counts as an index of spawning timing by natural-origin Chinook salmon (Oncorhynchus tshawytscha) in the Skagit River system in Washington State, USA. Over the last 2–6 decades, natural-origin salmon have been spawning later by 0.03–0.52 days·year–1, while a naturally spawning group that is influenced by strays from a hatchery has been spawning earlier by 0.19 days·year–1. Trends in the spawning timing of hatchery-origin strays may reflect opposing selection from the hatchery, where egg take for propagation has become earlier by 0.58 days·year–1. As mean August river temperatures have risen over the period of record, hatchery timing trends may be moving in the opposite direction from the plastic or adaptive patterns expressed by natural-origin fish.

Interest is growing in developing conservation strategies to restore and maintain coral reef ecosystems in the face of mounting anthropogenic stressors, particularly climate warming and associated mass bleaching events. One such approach is to propagate coral colonies ex situ and transplant them to degraded reef areas to augment habitat for reef-dependent fauna, prevent colonization from spatial competitors, and enhance coral reproductive output. In addition to such "demographic restoration" efforts, manipulating the thermal tolerance of outplanted colonies through assisted relocation, selective breeding, or genetic engineering is being considered for enhancing rates of evolutionary adaptation to warming. Although research into such "assisted evolution" strategies has been growing, their expected performance remains unclear. We evaluated the potential outcomes of demographic restoration and assisted evolution in climate change scenarios using an eco-evolutionary simulation model. We found that supplementing reefs with pre-existing genotypes (demographic restoration) offers little climate resilience benefits unless input levels are large and maintained for centuries. Supplementation with thermally resistant colonies was successful at improving coral cover at lower input levels, but only if maintained for at least a century. Overall, we found that, although demographic restoration and assisted evolution have the potential to improve long-term coral cover, both approaches had a limited impact in preventing severe declines under climate change scenarios. Conversely, with sufficient natural genetic variance and time, corals could readily adapt to warming temperatures, suggesting that restoration approaches focused on building genetic variance may outperform those based solely on introducing heat-tolerant genotypes.

Structured, systematic processes for decision-making can facilitate implementation of ecosystem-based fisheries management (EBFM). In US fisheries management, existing fishery ecosystem plans (FEPs) are primarily descriptive documents — not action-oriented planning processes. “Next-generation” FEPs extend existing FEPs by translating ecosystem principles into action through a structured process, including identifying and prioritizing objectives and evaluating trade-offs while assessing alternative management strategies for meeting objectives. We illustrate the potential for implementing a structured decision-making process for EBFM by reviewing fisheries management case studies through the lens of the next-generation FEP process, highlighting two perspectives. First, across case studies almost all steps occur, many occurring in multiple regions, indicating scientific and fisheries management capacity exists to conduct structured process components. Second, adjustments would be needed to transition to next-generation FEPs, as existing activity is rarely conducted within a fully structured, integrated process and examples of certain steps are scarce, but existing examples can guide future management. Implementing ongoing activity within next-generation FEPs would likely streamline fisheries management activity, saving time and resources while improving outcomes for stakeholders and ecosystems.

The chapter “Model Selection” provides a brief overview of alternative ways of thinking about what is (are) the best model(s) and shows the core motivation behind the Akaike information criterion as a measure of the predictive ability of a model fitted via maximum likelihood. It then gives stepwise practical guidance for using information theory as a basis of model selection, including nested versus non-nested models, goodness of fit, and overdispersion. Advanced topics cover some of the philosophy of information theory, other types of information theory criteria, and other ways of evaluating the predictive ability of models. As an example, the chapter examines the case of the Western monarch butterfly (Danaus plexippus plexippus), which, over the past two decades, has experienced a 97% decline from its historical average abundance, declining by 86% from 2017 to 2018 alone. Undoubtedly, there is more than one cause—indeed, overwintering habitat loss and pesticide use are both believed to be important contributors to the decline. Adopting a hypothesis-evaluation framework makes it possible to consider multiple alternative hypotheses simultaneously and measure degrees of support for alternative hypotheses.

The chapter “Random Variables and Probability” serves as both a review and a reference on probability. The random variable is the core concept in understanding probability, parameter estimation, and model selection. This chapter reviews the basic idea of a random variable and discusses the two main kinds of random variables: discrete random variables and continuous random variables. It covers the distinction between discrete and continuous random variables and outlines the most common probability mass or density functions used in ecology. Advanced sections cover distributions such as the gamma distribution, Student’s t-distribution, the beta distribution, the beta-binomial distribution, and zero-inflated models.

The chapter “Sensitivity Analysis” reviews why sensitivity analysis is a critical component of mathematical modeling, and the different ways of approaching it. A sensitivity analysis is an attempt to identify the parts of the model (i.e. structure, parameter values) that are most important for governing the output. It is an important part of modeling because it is used to quantify the degree of uncertainty in the model prediction and, in many cases, is the main goal of the model (i.e. the model was developed to identify the most important ecological processes). The chapter covers the idea of “local” versus “global” sensitivity analysis via individual parameter perturbation, and how interactive effects of parameters can be revealed via Monte Carlo analysis. Structural versus parameter uncertainty is also explained and explored.

“Why Fit Models to Data?” is a brief introductory chapter that sets the stage for the forthcoming chapters and serves as a link between Part 1, which contains the introductory chapters on mathematical ecology, and Part 2, which contains the statistical analyses of the models presented in Part 1: Part 1 illustrated how models can be used to make inferences about the real world, to help clarify ecological understanding, aid decision-making, and evaluate risk. However, if models are presented as hypotheses written in mathematical form, then it becomes possible to use statistical methods to determine which hypotheses have the greatest support. Part 2 will focus on developing statistical tools so that the reader will be able to express hypotheses as mathematical models, fit the models to data, and assess the degree of support for each. The chapter also illustrates the limitations of null-hypothesis testing in decision-making in high-dimensional, multicausal systems.

The chapter “Competition and Predation Models” considers models with two or more interacting species. What needs to happen for there to be “stable equilibria” that contain all possible members of a system? This is where simple models can be useful: these interactions can be represented by mathematical equations, and then solved for conditions that allow species to coexist. This chapter shows three techniques that make it possible to take a model system and determine whether the system has a stable equilibrium with all members present. The basic principles of model stability are presented, as well as how mathematical models can be used to address basic ecological questions in competition and predator-prey systems. Isocline analysis and analytical stability analysis are explained as ways to predict model behavior and are then used to draw inferences about the processes acting in the real world.

The chapter “Bayesian Statistics” gives a brief overview of the Bayesian approach to statistical analysis. It starts off by examining the difference between frequentist statistics and Bayesian statistics. Next, it introduces Bayes’ theorem and explains how the theorem is used in statistics and model selection, with the prosecutor’s fallacy given as a practice example. The chapter then goes on to discuss priors and Bayesian parameter estimation. It concludes with some final thoughts on Bayesian approaches. The chapter does not answer the question “Should ecologists become Bayesian?” However, to the extent that alternative models can be posed as alternative values of parameters, Bayesian parameter estimation can help assign probabilities to those hypotheses.

The chapter “Putting It Together: Fitting a Dynamic Model” provides a synthesis of the material presented in the book, by presenting a worked example that embraces concepts of density dependence, complex model dynamics, parameter estimation, and model selection using the Akaike information criterion. The example chosen is the recovery of gray wolf (Canis lupus) population in Washington State since 2008. The chapter begins by explaining how to fit an observation error model. Next, it examines how to fit a process error model. It then discusses parameter estimates and model selection. The chapter concludes with discussion of how to determine whether the population given as an example exhibits complex population dynamics.

The chapter “Modeling in Spreadsheets” introduces some of the basic skills needed to implement the models reviewed in this book. It explains the difference between absolute and relative cell references, how to write formulas with both types of cell references, and basic spreadsheet functions such as sum, average, count, and so on. It also shows how to make simple plots (e.g. x-y scatter plots). In addition, it covers topics such as how to structure a spreadsheet for clear model analysis, naming cells, useful commands, and the data table function. Finally, it provides a brief introduction to Visual Basic, which can be used to extend to the power of spreadsheets.

The chapter “Skills for Dynamic Models” provides worked examples of the dynamic models presented in Part 1, both in spreadsheets and in R. It also covers some of the mathematical steps used in model analysis. In most cases, instructions are given for both spreadsheets and R. However, when some activities are far easier to do in a programming environment than in spreadsheets, only the instructions for R are shown. The chapter starts out by discussing the skills needed for structured population models, including setting up age structure and creating cobweb plots. Next, it reviews the skills needed for multivariable models, including calculating isoclines and Jacobian matrices. Finally, it introduces the concept of Monte Carlo methods and provides guidance on how to implement them.

Modern practice of ecology, conservation, and resource management demands unprecedented levels of quantitative proficiency in mathematical modeling and statistics. This text provides foundational training in the concepts and methods of mathematical and statistical modeling used in ecology, for readers with all levels of quantitative proficiency and confidence. The first chapter presents a generalized approach to develop ecological models and introduces the “describe, explain, and interpret” framework for linking the model world to the real world. Detailed treatment of population models illustrates the myriad ways in which one can develop a model, shows how modeling choices are informed by the ecological question at hand, and emphasizes the epistemology of quantitative techniques. The second part of the book illustrates how to estimate parameters of models from data, and how to use mathematical models combined with statistics to test hypotheses. The third part of the book is devoted to an in-depth development of technical skills to implement models in two common platforms: spreadsheets and the R programming language. The book concludes by demonstrating a quantitative approach to addressing a question that spans density-dependent versus density-independent population models, fitting models to data, evaluating the strength for density dependence using model selection, and evaluating the types of dynamic behaviors that the population might exhibit.

The chapter “Mathematics Refresher” provides a brief reminder of operations with logarithms, matrices, and calculus, for student reference. It starts off by reviewing the differences between regular logarithms and natural logarithms and provides some examples of common operations with logarithms. It then introduces derivatives and integrals (although it is never necessary to compute an integral in this book, it is still useful to know what an integral is) and explains the sum rule, the product rule, the quotient rule, and the chain rule. Next, it provides a brief overview of matrices and matrix operations, including matrix dimensions, and addition and multiplication of matrices. It concludes with a discussion of the identity matrix.

The chapter “Stochastic Population Models” introduces the concept of stochasticity, why it is sometimes incorporated into models, the consequences of stochasticity for population models, and how these types of models are used to evaluate extinction risk. Ecological systems are (seemingly) governed by randomness, or “stochasticity.” A stochastic model is one that explicitly includes randomness in the prediction of state variable dynamics. Because these models have a random component, each model run will be unique and will rarely look like a deterministic simulation. In this chapter, simple unstructured and density-dependent models are presented to show core concepts, and extensions to structured and density-dependent models are given.

The chapter “Likelihood and Its Applications” introduces the likelihood concept and the concept of maximum likelihood estimation of model parameters. Likelihood is the link between data and models. It is used to estimate model parameters, judge the degree of precision of parameter estimates, and weight support for alternative models. Likelihood is therefore a crucial concept that underlies the ability to test multiple models. The chapter contains several worked examples that progress the reader through increasingly complex problems, ending at likelihood profiles for models with multiple parameters. Importantly, it illustrates how one can take any dynamic model and data and use likelihood to link the data (random variables) to a probability function that depends on the dynamic model.

The final chapter of the book, “Next Steps,” reviews the material covered in the previous chapters and suggests pathways that readers might choose to further expand their knowledge. The chapter makes the three following points. First, virtually no one becomes expert in all the areas listed in the chapter. Second, the purpose of the book was to show the types of models that might be seen in more advanced textbooks and scientific literature, together with the opportunities they present to expand skills and knowledge. Third, while this book probably will not make readers skilled at every quantitative problem that will come their way, it will provide them with the foundational knowledge of concepts and basic skills that will enable them to broaden their understanding and further develop their quantitative competence and confidence.

The chapter “Why Do We Model?” addresses the question of why it is important to use models. It is not possible to represent all of reality with a series of mathematical or statistical expressions. Luckily, modelers do not intend to do this. Rather, modelers simplify reality on purpose, so that they can better understand it. However, models must be faithful to reality. In addition, they must be purposeful, that is, they must guide experience in very specific ways. This chapter covers the epistemological basis of ecological models, introduces the core concept of “describe, explain, and interpret” as the core steps of learning from models, and then walks the reader through the process of model development.

The chapter “Modeling in R” covers the basic interface of R and R Studio, types of objects, core programming language tools, common and useful commands, and creating functions. This chapter is not a substitute for a dedicated book or training program on R; rather, it gives the basics needed to get the reader started. It starts off with an overview of the basics of coding in R and provides some of the statistical functions that might be used in modeling. Next, it discusses basic plotting, data input and output, and the concept of looping. It then works though a logistic population growth model to see how a dynamic model might be implemented in R. The chapter concludes with an explanation of how to create functions in R.

The chapter “Skills for Fitting Models to Data” provides worked examples of the model parameter estimation and model-selection examples presented in Part 2, both in spreadsheets and in R. This chapter presumes that the reader is reasonably comfortable setting up spreadsheets and R code and applying the modeling skills presented in Chapter 15. It begins with maximum likelihood estimation, presenting first a direct method and then numerical methods that yield more precise (usually) estimates of maximum likelihood parameter estimates. It then examines how to estimate parameters that do not appear in probability functions (e.g. a model in which survivorship rate is density dependent). The chapter concludes by discussing likelihood profiles.

The chapter “Introduction to Population Models” introduces unstructured population models and shows how model decisions can change model behavior, the different ways that feedbacks can be represented, and how one evaluates the consequences of those feedbacks. The goal here is to show how modeling a single entity, population density, can be done in many different ways, depending on the purpose of the model. Understanding the dynamics of populations remains one of the fundamental goals of ecology. Not surprisingly, many models have contributed to the theory of population dynamics and regulation. The models vary considerably in terms of depth, breadth, intended uses (e.g. prediction vs. generality), and structure. This chapter will largely focus on the behavior of simple models, to see how intrinsic factors can dictate variability in population size. Density-independent and density-dependent models are covered, as well as methods used to understand model behaviors.

Global environmental change is challenging species with novel conditions, such that demographic and evolutionary trajectories of populations are often shaped by the exchange of organisms and alleles across landscapes. Current ecological theory predicts that random networks with dispersal shortcuts connecting distant sites can promote persistence when there is no capacity for evolution. Here, we show with an eco‐evolutionary model that dispersal shortcuts across environmental gradients instead hinder persistence for populations that can evolve because long‐distance migrants bring extreme trait values that are often maladaptive, short‐circuiting the adaptive response of populations to directional change. Our results demonstrate that incorporating evolution and environmental heterogeneity fundamentally alters theoretical predictions regarding persistence in ecological networks.

Corals are experiencing unprecedented decline from climate change‐induced mass bleaching events. Dispersal not only contributes to coral reef persistence through demographic rescue but can also hinder or facilitate evolutionary adaptation. Locations of reefs that are likely to survive future warming therefore remain largely unknown, particularly within the context of both ecological and evolutionary processes across complex seascapes that differ in temperature range, strength of connectivity, network size, and other characteristics. Here, we used eco‐evolutionary simulations to examine coral adaptation to warming across reef networks in the Caribbean, the Southwest Pacific, and the Coral Triangle. We assessed the factors associated with coral persistence in multiple reef systems to understand which results are general and which are sensitive to particular geographic contexts. We found that evolution can be critical in preventing extinction and facilitating the long‐term recovery of coral communities in all regions. Furthermore, the strength of immigration to a reef (destination strength) and current sea surface temperature robustly predicted reef persistence across all reef networks and across temperature projections. However, we found higher initial coral cover, slower recovery, and more evolutionary lag in the Coral Triangle, which has a greater number of reefs and more larval settlement than the other regions. We also found the lowest projected future coral cover in the Caribbean. These findings suggest that coral reef persistence depends on ecology, evolution, and habitat network characteristics, and that, under an emissions stabilization scenario (RCP 4.5), recovery may be possible over multiple centuries., We used an eco‐evolutionary metapopulation model to simulate coral adaptation to warming across reef networks in the (a) Caribbean, (b) the Southwest Pacific, and (c) the Coral Triangle. We found that evolution is critical in preventing extinction and facilitating recovery in all regions. Furthermore, we found that the strength of immigration to a reef and current sea surface temperature robustly predicted reef persistence.

Contemporary stock assessment models used by fisheries management often assume that natural mortality rates are constant over time for exploited fish stocks. This assumption results in biased estimates of fishing mortality and reference points when mortality changes over time. However, it is difficult to distinguish changes in natural mortality from changes in fishing mortality, selectivity, and recruitment. Because changes in size structure can be indicate changes in mortality, one potential solution is to use population size-structure and fisheries catch data to simultaneously estimate time-varying natural and fishing mortality. Here we test that hypothesis, using a simulation experiment to test performance for four alternative estimation models that estimate natural and fishing mortality from size structure and catch data. We show that it is possible to estimate time-varying natural mortality in a size-based model, even when fishing mortality, recruitment, and selectivity are changing over time. Finally, we apply the model to North Sea sprat, and show that estimates of recruitment and natural mortality are similar to estimates from an alternative multispecies population model fitted to additional data sources. We recommend exploring potential trends in natural mortality in forage fish assessments using tools such as the one presented here.

The appetite for ecosystem-based fisheries management (EBFM) approaches has grown, but the perception persists that implementation is slow. Here, we synthesize progress toward implementing EBFM in the United States through one potential avenue: expanding fish stock assessments to include ecosystem considerations and interactions between species, fleets, and sectors. We reviewed over 200 stock assessments and assessed how the stock assessment reports included information about system influences on the assessed stock. Our goals were to quantify whether and how assessments incorporated broader system-level considerations, and to explore factors that might contribute to the use of system-level information. Interactions among fishing fleets (technical interactions) were more commonly included than biophysical interactions (species, habitat, climate). Interactions within the physical environment (habitat, climate) were included twice as often as interactions among species (predation). Many assessment reports included ecological interactions only as background or qualitative considerations, rather than incorporating them in the assessment model. Our analyses suggested that ecosystem characteristics are more likely to be included when the species was overfished (stock status), the assessment is conducted at a science centre with a longstanding stomach contents analysis program, and/or the species life history characteristics suggest it is likely to be influenced by the physical environment, habitat, or predation mortality (short-lived species, sessile benthic species, or low trophic-level species). Regional differences in stomach contents analysis programs may limit the inclusion of predation mortality in stock assessments, and more guidance is needed on best practices for the prioritization of when and how biophysical information should be considered. However, our results demonstrate that significant progress has been made to use best available science and data to expand single-species stock assessments, particularly when a broad definition of EBFM is applied.

Ecosystem‐based fishery management requires considering the effects of actions on social, natural and economic systems. These considerations are important for forage fish fisheries, because these species provide ecosystem services as a key prey in food webs and support valuable commercial fisheries. Forage fish stocks fluctuate naturally, and fishing may make these fluctuations more pronounced, yet harvest strategies intended to ameliorate these effects might adversely affect fisheries and communities. Here, we evaluate trade‐offs among a diverse suite of management objectives by simulating outcomes from several harvest strategies on forage fish species. We demonstrate that some trade‐offs (like those between catches and minimizing collapse length) were universal among forage species and could not be eliminated by the use of different control rules. We also demonstrate that trade‐offs vary among forage fish species, with strong trade‐offs between stable, high catches and high‐biomass periods (“bonanzas”) for menhaden‐ and anchovy‐like fish, and counterintuitive trade‐offs for sardine‐like fish between shorter collapses and longer bonanzas. We find that harvest strategies designed to maintain stability in catches will result in more severe collapses. Finally, we show that the ability of assessments to detect rapid changes in population status greatly affects control rule performance and the degree and type of trade‐offs, increasing the risk and severity of collapses and reducing catches. Together, these results demonstrate that while default harvest strategies are useful in data‐poor situations, management strategy evaluations that are tailored to specific forage fish may better balance trade‐offs.

Recent observations of record low winter sea-ice coverage and warming water temperatures in the eastern Bering Sea have signaled the potential impacts of climate change on this ecosystem, which have implications for commercial fisheries production. We investigate the impacts of forecasted climate change on the eastern Bering Sea food web through the end of the century under medium- and high-emissions climate scenarios in combination with a selection of fisheries management strategies by conducting simulations using a dynamic food web model. The outputs from three global earth system models run under two greenhouse gas emission scenarios were dynamically downscaled using a regional ocean and biogeochemical model to project ecosystem dynamics at the base of the food web. Four fishing scenarios were explored: status quo, no fishing, and two scenarios that alternatively assume increased fishing emphasis on either gadids or flatfishes. Annual fishery quotas were dynamically simulated by combining harvest control rules based on model-simulated stock biomass, while incorporating social and economic tradeoffs induced by the Bering Sea’s combined groundfish harvest cap. There was little predicted difference between the status quo and no fishing scenario for most managed groundfish species biomasses at the end of the century, regardless of emission scenario. Under the status quo fishing scenario, biomass projections for most species and functional groups across trophic levels showed a slow but steady decline toward the end of the century, and most groups were near or below recent historical (1991–2017) biomass levels by 2080. The bottom–up effects of declines in biomass at lower trophic levels as forecasted by the climate-enhanced lower trophic level modeling, drove the biomass trends at higher trophic levels. By 2080, the biomass projections for species and trophic guilds showed very little difference between emission scenarios. Our method for climate-enhanced food web projections can support fisheries managers by informing strategic guidance on the long-term impacts of ecosystem productivity shifts driven by climate change on commercial species and the food web, and how those impacts may interact with different fisheries management scenarios.

Fisheries for forage fish may affect the survival and reproduction of piscivorous predators, especially seabirds. However, seabirds have evolved life history strategies to cope with natural fluctuations in prey and it is difficult to separate effects of fishing on seabirds from impacts of natural variability. To date, potential impacts of forage fisheries on seabirds have mainly been explored using ecosystem models that simplify seabird-forage-fish dynamics. We sought to explore how different forage fish harvest policies affect seabirds, accounting for structured population dynamics, life history specifics, and variation in forage fish dependencies across life stages; and how impacts vary across seabird and forage fish life histories. To explore these impacts, we developed an age-stage structured seabird model that incorporates seabird diet specialization, foraging behavior, and reproductive strategy, as well as different functional responses between prey availability and adult survival, juvenile survival, reproductive success, and breeder propensity. We parameterized this model for two contrasting seabird life histories: (1) a low fecundity, limited foraging range, diet specialist ("restricted"); and (2) a high fecundity, wide ranging, diet generalist ("flexible"). Each was paired with two different forage fish prey archetypes that were fished under various control rules. The restricted seabird population was expectedly less robust to constant fishing pressure than the flexible seabird, and this sensitivity was mainly due to functional response parameterization, rather than other life history parameters. Particularly, the restricted seabird was highly sensitive to the relationship between prey availability and adult survival but was not sensitive to the relationship between prey and reproductive success. An adaptive biomass-limit harvest rule for forage fish resulted in substantially higher seabird abundance compared to constant fishing across all scenarios, with minimal trade-offs to the fishery (depending on fishery management objectives). However, mechanisms governing the impact of the forage fish fishery on the seabird varied by forage fish type. Therefore, tailoring forage fish management strategies to forage fish life history can lead to mutually acceptable outcomes for fisheries and seabirds. If data or time are limited, an adaptive control rule is likely a safe bet for meeting seabird conservation objectives with limited impacts to fisheries.

Natural resource management is evolving toward holistic, ecosystem-based approaches to decision making. The ecosystem science underpinning these approaches needs to account for the complexity of multiple interacting components within and across coupled natural-human systems. In this research, we investigate the potential economic and ecological gains from adopting ecosystem-based approaches for the sardine and anchovy fisheries off of the coast of California, USA. Research has shown that while predators in this system are likely substituting one forage species for another, the assemblage of sardine and anchovy can be a significant driver of predator populations. Currently, the harvest control rules for sardine and anchovy fisheries align more with traditional single species framework. We ask what are the economic and ecological gains when jointly determining the harvest control rules for both forage fish stocks and their predators relative to the status quo? What are the implications of synchronous and anti-synchronous environmental recruitment variation between the anchovy and sardine stocks on optimal food-web management? To investigate these questions, we develop an economic-ecological model for sardine, anchovy, a harvested predator (halibut), and an endangered predator (Brown Pelican) that includes recruitment variability over time driven by changing environmental conditions. Utilizing large-scale numerical optimal control methods, we investigate how the multiple variants of integrated management of sardine, anchovy, and halibut impact the overall economic condition of the fisheries and Brown Pelican populations over time. We find significant gains in moving to integrated catch control rules both in terms of the economic gains of the fished stocks, and in terms of the impacts on the Brown Pelican populations. We also compare the relative performance of current stylized catch control rules to optimal single species and optimal ecosystem-based fisheries management (EBFM) across ecological and economic dimensions, where the former trade-off considerable economic value for ecological goals. More generally, we demonstrate how EBFM approaches introduce and integrate additional management levers for policymakers to achieve non-fishery objectives at lowest costs to the fishing sectors.

In coastal marine ecosystems, the depletion of dissolved oxygen can cause behavioral and distributional shifts of organisms and thereby alter ecological processes. We used the spatiotemporal variation in the onset and intensity of low dissolved oxygen in Hood Canal, Washington, USA, to investigate consequences of seasonally reduced oxygen on fish-zooplankton predator-prey interactions. By simultaneously monitoring densities of zooplankton (primarily the euphausiid; Euphausia pacifica) and zooplanktivorous fish (Pacific herring, Clupea pallasii, and Pacific hake, Mercluccius productus), and the feeding of zooplanktivorous fish, we could separate the effects of dissolved oxygen on fish-zooplankton interactions from other seasonal changes. We expected that fish predators (especially Pacific herring) would be less abundant and have lower feeding rates when oxygen levels declined below biological thresholds, and that this would result in increased zooplankton abundance in areas with lowest dissolved oxygen. However, these expectations were not borne out. Overall, there was mixed evidence for an effect of dissolved oxygen on many of our response variables, and when effects were detected, they were frequently in the opposite direction of our expectations. Specifically, the pelagic fish community became more abundant (as measured by increasing acoustic backscatter), which was particularly pronounced for Pacific herring. Zooplankton had weak evidence for a response to dissolved oxygen, but the direction was negative instead of positive. Although predator feeding composition was unrelated to dissolved oxygen, stomach fullness (an index of feeding intensity) of Pacific herring declined, as per our expectations. These unexpected findings highlight the importance of in situ measurements of multiple aspects of predator-prey linkages in response to environmental stress to enhance our ability to predict ecological consequences of declining oxygen.