Andrew M. Edwards

Climate change is impacting marine ecosystems in increasingly rapid and unpredictable ways. The spatial extent of these impacts on marine fisheries and endangered and threatened species often spans regional and international boundaries. The transboundary nature of changing ocean conditions can challenge detection and anticipation of changes, and delay coordinated actions at the scales required to mitigate and respond. Here, we identify challenges and opportunities for transboundary science in the Northeast Pacific Ocean. Specifically, we aim to support the management of fisheries and protected species in response to a changing climate, while highlighting successful efforts. Challenges include (1) limited coordination of monitoring surveys; (2) institutional, cultural, and technological barriers to the sharing of data and analytical tools; (3) incomplete understanding of relationships among oceanographic conditions, trophic interactions, population dynamics, and species’ distributions; (4) limited availability of high-resolution Earth system model projections that can be linked to ecosystem and fishery responses; and (5) differing prioritization of ecosystem information and limited communication among nations. Three opportunities to overcome these challenges include (1) coordinated monitoring and sharing of data at a transboundary scale to detect and understand marine ecosystem responses to climate change; (2) common assessment frameworks and modeling approaches to improve understanding and projections of ecological responses to climate change; and (3) increased communication of ecosystem information to support management needs across jurisdictions, enhance the use of existing science products, and strengthen pathways for science to inform management of marine resources. We focus on examples of these opportunities drawn from our collective experience as government scientists working on Northeast Pacific Ocean ecosystems. We suggest that by strategically focusing on these opportunities, transboundary science worldwide can improve predictions of ecosystem responses to climate change and better support regional and international management of shared resources.

Empirical dynamic modelling (EDM) is becoming an increasingly popular method for understanding the dynamics of ecosystems. It has been applied to laboratory, terrestrial, freshwater and marine systems, used to forecast natural populations and has addressed fundamental ecological questions. Despite its increasing use, we have not found full explanations of EDM in the ecological literature, limiting understanding and reproducibility. Here we expand upon existing work by providing a detailed introduction to EDM. We use three progressively more complex approaches. A short verbal explanation of EDM is then explicitly demonstrated by graphically working through a simple example. We then introduce a full mathematical description of the steps involved. Conceptually, EDM translates a time series of data into a path through a multi‐dimensional space, whose axes are lagged values of the time series. A time step is chosen from which to make a prediction. The state of the system at that time step corresponds to a ‘focal point’ in the multi‐dimensional space. The set (called the library) of candidate nearest neighbours to the focal point is constructed, to determine the nearest neighbours that are then used to make the prediction. Our mathematical explanation explicitly documents which points in the multi‐dimensional space should not be considered as focal points. We suggest a new option for excluding points from the library that may be useful for short‐term time series that are often found in ecology. We focus on the core simplex and S‐map algorithms of EDM. Our new R package, pbsEDM, enhances understanding (by outputting intermediate calculations), reproduces our results and can be applied to new data. Our work improves the clarity of the inner workings of EDM, a prerequisite for EDM to reach its full potential in ecology and have wide uptake in the provision of advice to managers of natural resources.

Fishery-independent longline surveys provide valuable data to monitor fish populations. However, competition for bait on the finite number of hooks leads to biased estimates of relative abundance when using simple catch-per-unit-effort methods. Numerous bias-correcting instantaneous catch rate methods have been proposed, modelling the bait removal times as independent random variables. However, experiments have cast doubts on the many assumptions required for these to accurately infer relative abundance. We develop a new approach by treating some observations as right-censored, acknowledging that observed catch counts are lower bounds of what they would have been in the absence of hook competition. Through simulation experiments we confirm that our approach consistently outperforms previous methods. We demonstrate performance of all methods on longline survey data of 11 species. Accounting for hook competition leads to large differences in relative indices (often −50% to +100%), with effects of hook competition varying among species (unlike other methods). Our method can be applied using existing statistical packages and can include environmental influences, making it a general and reliable method for analyzing longline survey data.

Species responses to climate change are often measured at broad spatiotemporal scales, which can miss the fine‐scale changes that are most relevant to conservation and fisheries management. We develop a scaleable geostatistical approach to assess how juvenile and adult fish distributions have been shaped by changes in bottom temperature and dissolved oxygen over a recent decade of warming in the northeast Pacific. Across 38 demersal fishes, biomass trends were associated negatively with warming and positively with dissolved oxygen, but when trends in both biomass and climate were converted to velocities—the speed and direction a population would have to move to maintain consistent conditions—the effect of temperature change differed depending on local conditions. In the warmest locations, warming velocities were associated with negative biotic velocities for 19 of 69 species‐maturity combinations, and yet were almost always associated with stable or positive biotic velocities in the coolest locations (64 of 69). These spatially consistent biomass declines (negative biotic velocities) in the warmest locations and increases in cooler locations suggest a redistribution of species with the potential for new ecological and fisheries interactions. After controlling for temperature, the more spatially consistent effects of dissolved oxygen were often negative, suggesting a mechanism other than hypoxia avoidance—potentially changes in primary production. Our approach identifies the species and locations that are most sensitive to observed changes in the environment at any scale, thus facilitating future vulnerability assessments.

Estimates of the basic reproduction number ( R ₀ ) for Coronavirus disease 2019 (COVID-19) are particularly variable in the context of transmission within locations such as long-term health care (LTHC) facilities. We sought to characterise the heterogeneity of R ₀ across known outbreaks within these facilities. We used a unique comprehensive dataset of all outbreaks that have occurred within LTHC facilities in British Columbia, Canada. We estimated R ₀ with a Bayesian hierarchical dynamic model of susceptible, exposed, infected, and recovered individuals, that incorporates heterogeneity of R ₀ between facilities. We further compared these estimates to those obtained with standard methods that utilize the exponential growth rate and maximum likelihood. The total size of an outbreak varied dramatically, with a range of attack rates of 2%–86%. The Bayesian analysis provides more constrained overall estimates of R ₀ = 2.19 (90% CrI [credible interval] 0.19–6.69) than standard methods, with a range within facilities of 0.48–10.08. We further estimated that intervention led to 57% (47%–66%) of all cases being averted within the LTHC facilities, or 73% (63%–78%) when using a model with multi-level intervention effect. Understanding the risks and impact of intervention are essential in planning during the ongoing global pandemic, particularly in high-risk environments such as LTHC facilities.

Following successful widespread non-pharmaceutical interventions aiming to control COVID-19, many jurisdictions are moving towards reopening economies and borders. Given that little immunity has developed in most populations, re-establishing higher contact rates within and between populations carries substantial risks. Using a Bayesian epidemiological model, we estimate the leeway to reopen in a range of national and regional jurisdictions that have experienced different COVID-19 epidemics. We estimate the risks associated with different levels of reopening and the likely burden of new cases due to introductions from other jurisdictions. We find widely varying leeway to reopen, high risks of exceeding past peak sizes, and high possible burdens per introduced case per week, up to hundreds in some jurisdictions. We recommend a cautious approach to reopening economies and borders, coupled with strong monitoring for changes in transmission.

Extensive physical distancing measures are currently the primary intervention against coronavirus disease 2019 (COVID-19) worldwide. It is therefore urgent to estimate the impact such measures are having. We introduce a Bayesian epidemiological model in which a proportion of individuals are willing and able to participate in distancing measures, with the timing of these measures informed by survey data on attitudes to distancing and COVID-19. We fit our model to reported COVID-19 cases in British Columbia, Canada, using an observation model that accounts for both underestimation and the delay between symptom onset and reporting. We estimate the impact that physical distancing (also known as social distancing) has had on the contact rate and examine the projected impact of relaxing distancing measures. We find that distancing has had a strong impact, consistent with declines in reported cases and in hospitalization and intensive care unit numbers. We estimate that approximately 0.78 (0.66–0.89 90% CI) of contacts have been removed for individuals in British Columbia practising physical distancing and that this fraction is above the threshold of 0.45 at which prevalence is expected to grow. However, relaxing distancing measures beyond this threshold re-starts rapid exponential growth. Because the extent of underestimation is unknown, the data are consistent with a wide range in the prevalence of COVID-19 in the population; changes to testing criteria over time introduce additional uncertainty. Our projections indicate that intermittent distancing measures—if sufficiently strong and robustly followed— could control COVID-19 transmission, but that if distancing measures are relaxed too much, the epidemic curve would grow to high prevalence.