Quantitative ecology and resource management

Permanent URI for this collectionhttps://digital.lib.washington.edu/handle/1773/4961

Browse

Now showing 1 - 20 of 51

Modeling population status and demographic rates of baleen whales using historical whaling data
(2026-04-20) Rand, Zoe; Branch, Trevor A
Commercial whaling in the 20th century decimated the populations of many baleen whale species. Blue whales (Balaenoptera musculus) were especially impacted, with multiple populations whaled to near extinction, resulting in their current listing globally as Endangered by the IUCN. During whaling, extensive biological data were collected that, when combined with contemporary statistical methods, can be used to answer long-standing questions about baleen whale demography and population status. In this dissertation, I use historical data from whaling and contemporary Bayesian statistical methods to model baleen whale population dynamics and demography, with a particular focus on the highly exploited blue whale. In Chapter 1, I used a multi-state mark-recovery model to investigate population structure in Antarctic blue whales using historical mark-recovery data and found that they move frequently in the Southern Ocean, suggesting they are a single well-mixed circumpolar population. In Chapter 2, using the extensive fetal sex ratio data collected during whaling, I investigated ecological theories about adaptive sex ratio behavior, finding that longer rorqual (family Balaenopteridae) whale mothers have more female offspring. This suggests there is an advantage to being large that larger mothers pass on their daughters, likely stemming from the high costs of gestation and lactation for female baleen whales. In Chapter 3, I combined historical catch data and contemporary abundance estimates to build a population assessment model for Antarctic blue whales, finding that at the end of whaling they were at just 0.2% of pre-whaling levels, although their population size is currently increasing. Despite increasing, they are currently at less than 2% of pre-whaling levels so they still have many decades before they recover from whaling. In Chapter 4, using a global compilation of blue whale aging and reproductive data, I modeled age-length relationships and estimated reproductive rates for Antarctic, pygmy, and eastern North Pacific blue whales, finding that asymptotic lengths were longer for females than males across subspecies, and that Antarctic blue whales were the largest while pygmy blue whales were the smallest. In addition, I estimated pregnancy rates and age of sexual maturity for female eastern North Pacific and pygmy blue whales and, in Chapter 5, estimated natural survival for pygmy blue whales. These projects provide a deeper understanding of blue whale population dynamics and demography and lay the groundwork for age-structured stock assessments of blue whales.
Using Species Distribution Modeling to Assess Illegal Trade Risk: A Case Study of US Imports of Endangered Mahogany (Swietenia spp.)
(2024-10-16) Pollack, Sarah Erali; Ganguly, Indroneil
Illegal logging threatens biodiversity and the global conservation of endangered species. To protect tree species from this threat, governments have enacted legislation against importing illegally sourced plant products. These laws, including the US Lacey Act Amendment of 2008, involve collecting information from importers on the scientific name and country of harvest of their plant products. For many species, a lack of consistent, comprehensive global distribution data makes validating these declarations difficult. I aimed to assist in this analysis by developing species distribution models for the American mahogany species (genus Swietenia). American mahoganies are high-value, high-demand tropical timber species restricted in international trade and known to be illegally logged and traded. I combined presence-absence species occurrence data from government forest surveys and presence-only species occurrence data from citizen science programs and herbarium collections for each Swietenia species. Both data types were modeled as dependent on climate variables, with the presence-only data also modeled as dependent on sampling bias predictors, such as distances to roads or cities. I combined the range maps generated by the models with previously compiled range country lists to create a global distribution for each species. I compared my range map predictions and the range country lists to a set of US import declarations of Swietenia species. Each declared species-country pair was classified as confirmed or unconfirmed based on the range country lists, and suitable or unsuitable based on the model output. By examining the unconfirmed, but climatically suitable, species-country pairs, I identified new range countries for Swietenia species that were poorly documented in the literature. I also found that many unconfirmed but suitable country-species pairs were from countries that did not contain Swietenia but did contain other, closely related species. Many of these species are endangered and/or regulated and produce wood that can be indistinguishable from Swietenia. These results show how species distribution modeling can expand knowledge of a species’ global range and help identify patterns of potential illegal trade. While this study focused on US imports of Swietenia, these methods could be applied to any species or regulatory framework where knowing the country of harvest is necessary.
Fungi, forests, and fish: the role of fungi in forest biogeochemical cycling
(2024-10-16) Polyakov, Anne Y; Berdahl, Andrew M.
Forests are threatened by a multitude of stressors and assessing how forests will respond to these stressors requires a comprehensive understanding of net primary productivity (Npp). However, one parameter of significant uncertainty is belowground Npp (bNpp), or total photosynthetic carbon allocated belowground. I evaluated the effect of climatic and soil variables on bNpp using a global productivity dataset and found that soil nutrient flux rates were the main drivers of bNpp across forest biomes. Within biomes, environmental constraints on bNpp varied between forests with low versus high bNpp, indicating that environmental drivers are site-specific and the development of within-biome, site-scale classifications for forest ecosystems could be useful. Finally, I found that soil variables caused abrupt and large decreases in bNpp for boreal, but not cold temperate forests, indicating that boreal forests might have lower adaptive capacity and higher sensitivity to disturbances than cold temperate forests. I then examined nutrient cycling in an Alaskan boreal riparian forest and evaluated the role of fungi in the flow of salmon-derived nitrogen (SDN). Pacific salmon subsidize terrestrial systems with SDN, which can have ecological significance for animals, plants, and soils. However, despite the crucial role of fungi in decomposition and nutrient cycling, the importance of fungal mycelium as an SDN sink in the soil has been largely overlooked. I found that SDN was present in fungal sporocarps, soils, and foliage, but only significantly increased N concentrations in fungi, suggesting that SDN was retained in fungal hyphae. Furthermore, SDN from recent salmon carcasses increased the relative abundance and diversity of medium-distance and long-distance ectomycorrhizal fungal types. In systems such as salmon streams where nutrients are available but patchily distributed and occur in pulses, these fungi might be preferred by plant hosts due to their strong ability to maintain large network structure, mine nutrient hotspots, and mobilize organically bound N and P. These results illustrate that fungi are essential to consider to accurately assess the flow and significance of SDN in riparian systems.
Quantitative modeling tools for invasive species management decisions
(2024-09-09) Thompson, Brielle Kwarta; Converse, Sarah J; Olden, Julian D
Invasive species threaten biodiversity and cause harm to the environment, economies, and human health. Natural resource managers tasked with determining management plans for controlling or eradicating invasive species often grapple with challenges such as ecosystem complexity, uncertainty about the effectiveness of management actions, limited budgets, and conflict with the public regarding management practices. Quantitative population models applied to invasive species management provide a cost-effective tool for evaluating management outcomes in a virtual environment before management is implemented. In particular, simulation models provide insight into the performance of alternatives under varying ecological states and management assumptions prior to substantial time investment or expensive on-the-ground experiments. Here, I demonstrate how quantitative models can be harnessed to effectively inform invasive species management decisions. First, I provide an extensive review of mechanistic models that are used for invasive species management to address the gap between those who build models and those who are tasked with actual management implementation (Chapter 2). Second, I provide a simulation study to assess different spatial strategies for invasive rusty crayfish (Faxonius rusticus) removal in the complex riverine environment of the John Day River, USA (Chapter 3). The model indicated that to minimize overall population abundance, crayfish should be removed in locations where their abundance is highest, and removal at the most downstream extent of their invasion is key for preventing invasion to new areas, i.e., the Columbia River, USA. Third, I provide an adaptive management framework for invasive flowering rush (Butomus umbellatus) management in the Columbia River to support decisions regarding allocation of resources towards monitoring and control under two invasion conditions (Chapter 4). The model revealed that for an established invasion, it was beneficial to conduct monitoring and removal at spatially fixed areas, whereas for an emerging invasion, effort can be more effectively allocated in highly invaded areas. The model also indicated that for an emerging invasion, managers may benefit by integrating community science data into their monitoring to help track the emerging invasion. Finally, I examined how to identify optimal invasive species management actions involving multiple decision makers, each with their own management objectives (Chapter 5). To do so, I compared multiple-criteria decision analysis (MCDA), used for decisions involving multiple objectives, and game theory, used in circumstances with multiple decision makers. I showed that MCDA sometimes failed to reveal invasive species harvest actions that were identified in game-theoretic analyses as providing improved outcomes, but MCDA provided better insight into the preferred actions of each individual decision maker. Overall, my research demonstrates ways in which quantitative models can be used to help decision makers identify promising solutions to invasive species management. Broadly, my research demonstrates ways in which quantitative modeling tools can be used to help inform decision making in natural resource management.
Impacts of Environmental and Social Factors on Fish Movement and Behavior Over Three Timescales
(2024-09-09) Kuruvilla, Maria; Berdahl, Andrew
The ability to move influences essential behaviors in fish, including foraging, predator escape, migration, mating, and breeding. These behaviors occur over various timescales, from the immediate response to predators to migrations over a longer timescale. Environmental factors like temperature impact both instantaneous and long-term behaviors. Additionally, social factors, such as the presence and behavior of other fish, also play a significant role in these behaviors. This dissertation explores the impacts of environmental and social factors on fish movement and behavior across three distinct timescales. In the first chapter, I examine how temperature affects the collective response of fish to a predation threat in the timescale of seconds. Using lab experiments I found that while fish can move faster in higher temperatures, during a predation threat, fish at lower temperatures are able to compensate for the lower speeds by increasing their probability to startle. In the second chapter, I explore the influence of social cues on the juvenile migration of salmon on a daily timescale in three rivers in Washington state. I use hatchery releases as a way to test the ‘pied-piper’ hypothesis that the large number of hatchery salmon moving downstream motivates wild salmon to migrate along with them. Our findings support the ‘pied-piper’ hypothesis, demonstrating that both environmental and social cues stimulate downstream migration in salmon. In the third chapter, I investigate the influence of temperature, cumulative flow, and hatchery releases on the timing of peak migration and the duration of migration of juvenile salmon at the seasonal timescale. The results show that increased temperatures have a negative effect on the timing of migration of sub-yearling Chinook salmon and sub-yearling chum salmon. Similarly, increased flow has a negative effect on the timing of migration of coho salmon. Larger and earlier hatchery releases have a negative effect on the duration of migration of coho salmon. Broadly, this research shows us both environmental factors and social factors influence the movement and behavior of fish in nuanced ways. By understanding these relationships across various timescales, we can make better management strategies to reduce the impact of various anthropogenic stressors on fish populations.
Characterizing and forecasting fish recruitment in a changing world
(2024-02-12) Sellinger, Emily; Punt, Andre E
Recruitment, the entry of young fish into a stock, is an important process in population dynamics models, which form the basis for many stock assessments. As such, recruitment has been a focus of research among fishery scientists for over a hundred years. Recruitment is often assumed to be a function of the spawning biomass of a stock. However, quantifying this relationship is challenging. The first chapter of this thesis examines the prevalence of a detectable influence of spawning biomass on recruitment for the stocks included in the RAM Legacy Stock Assessment database. We found most stocks (57%) did not have a detectable relationship between spawning biomass and recruitment over the observed ranges of spawning biomass. Environmental factors appear to play a larger role in recruitment variation. Furthermore, many of the stocks we examined had evidence of a regime shift (46%). Our results highlight the need to develop effective methods to model and forecast large variations in recruitment over time. The second chapter of this thesis evaluates the forecast performance of six methods on groundfish recruitment. The methods varied in the level of parameterization and operational use. We found that forecast performance depends on the time period, the performance metric, and the characteristics of the time series. Our results indicate an advantage in using non-parametric forecast methods, especially for mid-term projections.
From Mark-Resight to Management: Bayesian Hierarchical Models for Endangered Bird Populations
(2024-02-12) Bratt, Abby Elizabeth; Converse, Sarah
Producing reliable estimates of demographic rates is critical to our understanding of wildlife population dynamics and can provide valuable information for prioritizing conservation and management efforts. Precise and unbiased estimates are challenging to obtain when monitoring data are sparse, knowledge gaps are pervasive, or model assumptions are violated. This is often the case for species of conservation concern, which may be poorly understood and difficult to monitor. Bayesian hierarchical models are particularly useful for estimating demographic rates because they separate imperfect observation processes from the underlying biological processes, especially when combined in an integrated framework that leverages multiple data sources for increased precision and parameter identifiability. Here I present three case studies using Bayesian hierarchical models to better understand the demography of threatened birds, with particular contributions to mark-resight and integrated population modeling. In Chapter 2, I addressed a common but poorly understood problem in mark-resight studies of open populations: partial mark loss and degradation. I present a novel approach to sampling latent states in a Markov Chain Monte Carlo framework using a backtracking algorithm, and I apply this approach in the context of a multi-event model to the Oregon Vesper Sparrow (Pooecetes gramineus affinis) in South Puget Sound, Washington, USA. The results from this model constitute some of the first estimates of age-specific survival and dispersal rates for this species of conservation concern. In Chapter 3, I developed a novel multi-site integrated population model (IPM) to better understand the population dynamics of Streaked Horned Larks (Eremophila alpestris strigata) in South Puget Sound, Washington. These estimates will inform future habitat management and a planned reintroduction effort, and the multi-site framework addresses a critical gap in modeling small populations monitored over fragmented landscapes. In Chapter 4, I developed an IPM to examine the impact of a cryptic threat, bycatch in commercial fisheries, on the population dynamics of Atlantic Yellow-nosed Albatross (Thalassarche chlororhynchos). Results from this model will motivate ongoing monitoring of Atlantic Yellow-nosed Albatross and seabird bycatch in the South Atlantic and inform fisheries regulation decisions. Broadly, the work I present here makes contributions to the development of complex demographic models with the goal of supporting conservation and management decisions by quantifying and reducing key uncertainties in the population dynamics of threatened species.
Spatio-temporal patterns of forest disturbance in western North America: implications for forest resilience
(2023-08-14) Buonanduci, Michele Susan; Harvey, Brian J.
Globally, forest disturbance activity is changing in response to changing climate. As disturbance regimes change, concerns have been raised that the mechanisms of forest resilience (i.e., the capacity of forests to tolerate disturbance) may begin to break down. To successfully monitor, forecast, and manage for forest resilience in the context of changing disturbance regimes, quantifying indicators of forest resilience across spatial and temporal scales is critical. In this dissertation, I quantified facets of forest resilience to biotic disturbance (i.e., insects and plant pathogens) and wildfire in western North America. First, I evaluated compensatory responses of forests following severe bark beetle outbreak. I found that compensatory growth responses are strongly shaped by both the characteristics and spatial arrangement of surviving trees, and that increased post-disturbance growth acts as a key mechanism of forest resilience by providing continuity in forest function. Next, I characterized the patterns and drivers of biotic disturbance hotspots, an emerging phenomenon in the western United States (US) in which two or more distinct biotic disturbance agents co-occur in space and time. I found that while biotic disturbance hotspots are driven by forest composition and regionally important bioclimatic factors, they are also stochastic processes that cannot be predicted solely from deterministic landscape characteristics or other known drivers. Interactions among multiple disturbances such as these are important to understand, as they have the potential to erode compensatory responses and therefore mechanisms of forest resilience. Finally, I quantified the range of variation in burn severity patch structure characterizing Northwest US fire regimes. Despite changes in climate and fire activity in recent decades, I found that the range of variation in high-severity burn patches, conditional on fire size, has remained remarkably stationary in recent decades. Stationarity in the relationship between burn severity patterns and fire size offers a simple yet powerful means to anticipate the range of ecological effects of future fire activity at regional scales. Building on this finding, I conducted a simulation study demonstrating that shifts in fire size distributions towards larger fire events will lead to increasingly large high-severity burn patches with interior areas that are increasingly far from unburned seed sources following fire. Large high-severity patches directly affect rates of tree regeneration and forest recovery following fire, along with the potential for forests to transition to non-forest ecosystems. Collectively, this work provides insights into a range of mechanisms of forest resilience in western North America and has important implications for managing forests in the face of continued climate change and increasing disturbance activity.
The Age of Infection: A Semi-Markov Framework for Developing Mechanistic Models of Malaria Epidemiology
(2023-08-14) Henry, John M; Smith, David L
Malaria is an epidemiologically complex disease which poses a significant burden onhumanity, contributing an estimated 643,000 deaths in 2019 alone [1]. Infection with one cohort of parasites does not prevent concurrent infection with others [2], making prevalence alone an incomplete measure of infection in a population. Further, immunity is slow to develop in response to exposure and is only partially protective [3]. Therefore, it acts to both suppress disease in individuals and to mask infections from detection and treatment, which allows for longer periods of transmission to mosquitoes. In this way past exposure modifies the bias in the observed prevalence and incidence data collected from each cohort of hosts. Natural variability in the system further obfuscates the connection between data and the process which generates it. Determining the transmission and burden of malaria becomes incredibly difficult without the use of theoretical models to connect the observed data to the latent states and parameters which guide our basic understanding and policy decisions, such as the distribution of the number of infections in each host and the detectable fraction of infections. Currently existing mechanistic models typically fall into one of two categories: simple and transparent, but deterministic and not descriptive enough to connect to available data [4]; or incredibly detailed individual-based simulation frameworks which are realistic but difficult to implement or calibrate [5,6]. Here, we propose a different approach of intermediate complexity which embraces the probabilistic structure of the system. In chapter 1, we start by demonstrating how the distribution of the multiplicity of infection modeled above is impacted by access to treatment, a major factor for the state of the system. In chapter 2, using data from controlled human malaria infections, we show how infection age is statistically predictive of parasitemia, and therefore detection, fever rates, and transmission rates. These relationships imply that infection age, a relatively simple quantity to model, can help us build predictive models for the latent and highly variable parasite densities. Chapter 3 explores this through simulation, and develops a theoretical framework for tracking the probability densities using simple ODEs. Chapter 4 expands on this, and by treating the outcomes of interest as generalized linear models of our semi-Markov model of infection age, we obtain simple ODE models for these probability density function tracking variables, with closed form equations relating them to observables such as theoretical true prevalence, detection, and fever-prompted treatment (using the results from chapter 1 along the way), all while incorporating immunity as a covariate. Finally in chapter 5 we turn our attention to fitting these models to time series data of observables, and ultimately develop an algorithm for fitting without any probabilistic simulation of the underlying stochastic process. It is our hope that this series of papers inspires others to build upon and push the new tools developed here, and use them to better improve future efforts toward malaria eradication everywhere.
Mechanistic Statistical Models of the Environment
(2023-04-17) Okasaki, Connie; Berdahl, Andrew M
Statistical models are often abstract in nature. However, in environmental contexts, data are often limited and important insight can be gained by applying knowledge of real-world mechanisms. In this dissertation, I present three mechanistic statistical models, applied to the environment. In my second chapter, I model the effect of sociality on the migration of adult Pacific salmon past large-scale dams in the Columbia River Basin, USA. I explicitly break down and model the process by which a salmon passes a dam. In all three steps of this process, I evaluate the effect of the density of conspecifics, to determine whether sociality plays a role in dam passage. In my third chapter, I present a method for inferring the source of a signal which has been deformed by well-understood linear dynamics. I use as an example the case of a pollutant which, upon entering the environment, is subject to advective-diffusive transport. I show how to incorporate a mechanistic linear partial differential equation (PDE) model into the classic stochastic PDE (SPDE) method from spatial statistics, and how to invert the transport dynamics within a statistical model. In my fourth chapter, I present a mixed integer linear program (MILP) model for constructing optimal sampling design under complex logistical or budgetary constraints. I use as an example the case of the US Forest Service (USFS) Forest Inventory and Analysis (FIA) program in Tanana, Alaska. I compare solutions of this model with three randomized, design-based benchmarks based on MSE and feasibility.
Hydrology, temperature, and water source dynamics across river basins of western North America
(2022-09-23) McGill, Lillian Marie; Holtgrieve, Gordon; Steel, E. Ashley
Climate change is altering temperature and precipitation regimes across the globe, resulting in often extreme modifications to river dynamics. Such impacts are particularly pronounced in western North America, a region with both water surplus and scarcity and therefore a long history of water resource challenges. To preserve riverine ecosystems, it is essential to improve our understanding of fundamental processes governing river dynamics, how river systems have and continue to change with anthropogenic forcing, and what tools and management actions may best facilitate conservation. Despite the breadth of studies examining western river hydrology and temperature, a persistent need remains for fundamental understanding and predictions on a management-relevant scale and the generation of insights that considers both temporal and spatial variation simultaneously. My dissertation research answers questions relating to how we describe and model patterns in riverine ecosystems and consists of a portfolio of projects aimed at improving our ability to understand fundamental drivers of, and predict and mitigate anthropogenic induced changes to, river hydrology, temperature, and water source.
Fish in Space: Estimating groundfish population distribution in the Gulf of Alaska for management apportionment by subregion
(2022-07-14) Mistry, Kelly; Scheurell, Mark
Management of fisheries relies on information about biomass of stocks in order to determine how many fish can be sustainably harvested in a given year. In addition to predicting total biomass, it is frequently important to accurately predict the distribution of stock biomass through space in order to avoid local depletions of stock and to more evenly distribute harvest among many stakeholders. Towards that end, this study compares the current modeling approach to predict the geographic apportionment of stock biomass for groundfish in the Gulf of Alaska (GOA), a random walk model, with a delta-GLMM spatiotemporal model implemented using the VAST package in R. These stocks are managed using subregional catch allocation, whereby the GOA is divided into 3 management areas: western, central, and eastern GOA. This analysis uses bottom trawl survey data collected by the Alaska Fisheries Science Center (AFSC) of the National Oceanic Atmospheric Administration National Marine Fisheries Service (NOAA NMFS) for two species of groundfish, Pacific Ocean Perch (Sebastes alutus) and Northern Rockfish (Sebastes polyspinis). Model performance was evaluated using the accuracy of the model estimate from jackknife resampled results for population proportion by subregion compared to survey design-based proportions, and on the precision of the model jackknife estimates. In terms of accuracy, the models performed similarly well, with the mean absolute difference between the model jackknife estimates and the design-based estimates for the random walk results being smaller by 0.086 or less than the delta-GLMM results, with significant overlap in the jackknife absolute difference values. Precision was measured by the CV calculated with the model jackknife results, and the delta-GLMM results had smaller mean CV values by at least 0.105, and very little overlap in the jackknife CV values. However, the precision of the delta-GLMM is small enough that it may lead to significant over- or underestimation compared to the survey design-based proportions and therefore may be a riskier option than the random walk model for estimating subregional catch apportionment for these stocks in the GOA.
Spatial modeling, parameter uncertainty, and precision of density estimates from line-transect surveys: a case study with Western Arctic bowhead whales
(2022-04-19) Ferguson, Megan Caton; Essington, Timothy
Spatially-explicit models of animal density, such as density surface models (DSMs), are diverse, flexible, and powerful tools for investigating spatial patterns in animal density, examining associations between animal density and environmental covariates, and estimating abundance. Advances in spatial modeling methods and subsequent incorporation into widely accessible software allow the non-specialist to add these tools to their analytical toolbox. However, limitations in some software may prevent a thorough treatment of uncertainty. I expanded the functionality of tools for constructing DSMs from line-transect survey data to derive a population abundance estimate that honestly accounts for multiple sources of detection bias and associated uncertainty. As an illustrative case study, I used data collected during an aerial line-transect survey for Western Arctic bowhead whales (\textit{Balaena mysticetus}) over their summering grounds in the Beaufort Sea and Amundsen Gulf during August 2019. Using spatially explicit hierarchical generalized additive models that incorporated correction factors and associated uncertainty for perception and availability bias, I estimated the abundance of the Western Arctic bowhead whale population to be 17,175 whales (CV($\hat{N}$)= 0.237; 95\% confidence interval = [10,793, 27,330]). This model-based abundance estimate is similar in magnitude to the two most recent estimates for this population based on data from ice-based surveys in 2011 and 2019. Additionally, my abundance estimate is sufficiently precise to inform management decisions for this protected species. The enhanced precision of my abundance estimate over the estimate derived using design-based analytical methods applied to the same data is due to explicit modeling of the spatial correlation in whale density. Applying the power of DSMs to the aerial line-transect survey data made this survey methodology a viable alternative to ice-based surveys, which are facing obstacles due to climate change, for updating abundance estimates for Western Arctic bowhead whales in the future. My analytical developments can easily be applied to other line-transect datasets with similar and common challenges due to multiple survey platforms, spatial heterogeneity in animal density and environmental conditions, and habitat partitioning among groups (e.g., defined by age, sex, activity state) in the target population.
Best Practices for Constructing Size-Structured Population Dynamics Models used for Stock Assessments
(2022-01-26) Cronin-Fine, Lee; Punt, Andre E
All models are predicated on assumptions which makes them simplified versions of reality. An important goal of stock assessment scientists is to expand the capability of stock assessment models to improve their ability to estimate the size of fished populations and how they will change over time. Several types of population dynamics models can be used for stock assessment, with age-structured models, which track the population by age, being the most popular. Unfortunately, there are valuable fished species such as crabs and lobsters that are difficult to age. Size-structured models that track the population by size are good alternatives. The objective of this dissertation is to improve the performance of size-structured models used for stock assessments. Three aspects of size-structured models are explored: growth, selectivity, and natural mortality. Size-transition matrices define growth in size-structured models. They are constructed from an underlying growth curve, typically the von Bertalanfy growth curve, defined by three parameters: the growth rate (k), the asymptotic height (L∞), and the variation in the size increment. Most assessments assume individuals follow a single growth curve with process error, which is unrealistic. A new size-transition matrix construction method that allows L∞ and k to vary among individuals through numerical integration was developed and compared with methods that allow individuals to follow (a) a single growth curve with process error, or (b) one of three growth curves, each with process error. The number of size-classes in the size-transition matrix and how the data are generated heavily dictate performance. Not accounting for temporal variation in selectivity can lead to biased estimates of abundance and mortality. Simulations suggest that discrete time blocking of selectivity can adequately capture time-varying selectivity as this could reduce the number of estimated parameters and hence the variance of estimated quantities. As for likelihood functions for size-composition data, the results reveal that multinomial, Dirichlet-multinomial and multivariate normal are all valid options. Natural mortality (M) has a strong influence on stock assessment model outputs including estimates of spawning stock biomass, MSY and fishing mortality. Estimating M is difficult since it is confounded with several factors including catchability, recruitment, and growth. Simulation shows that terminal molt does not affect the ability to estimate M. However, estimating growth simultaneously with M has a negative impact on the ability to estimate M but a positive effect on the quality of the estimates for spawning stock biomass.
Estimating demographic rates to improve monitoring of highly mobile species
(2021-10-29) Emmet, Robert; Gardner, Beth
Estimating demographic rates of wildlife species, such as survival and fecundity, is crucial for monitoring wildlife populations and informing management of these species. Monitoring highly mobile species is especially challenging, as their life histories and behaviors (e.g., migration) can affect inference on demographic rates and ultimately render monitoring less effective. Species’ movements can expose them to a variety of hazards and opportunities, creating spatial and temporal variation in demographic rates that must be accounted for in models. Furthermore, the movement behaviors of many highly mobile species can violate key assumptions of the standard statistical models used to estimate demographic rates, so that new monitoring frameworks and models need to be designed to minimize violations of model assumptions or relax those assumptions. In this dissertation, I used several case studies to demonstrate how novel models and monitoring frameworks can improve demographic rate estimation, ecological inference, and population monitoring capabilities for highly mobile species.
Physiological causes and biogeographic consequences of thermal optima in the hypoxia tolerance of marine ectotherms
(2021-08-26) Endress, Martin-Georg Alexander; Deutsch, Curtis
Recent measurements of critical O2 thresholds (‘Pcrit’) in aquatic animals have revealed thermal optima in their hypoxia tolerance. To discern the prevalence, physiological drivers, and biogeographic manifestations of such Pcrit curves, this research investigates experimental and occurrence data using a dynamic model of aquatic water breathers. The model simulates the transfer O2 from ambient water into animal tissues driven by temperature-dependent rates of metabolism, diffusion, and ventilatory and circulatory systems with O2-protein binding. Results show that thermal optima in Pcrit can arise even when all physiological rates increase steadily with temperature. This occurs when O2 supply at low temperatures is limited by a process that is more temperature sensitive than metabolism, and when O2 supply at warmer temperatures is limited by a less sensitive process. Analysis of species respiratory traits suggests this scenario is not uncommon in marine biota, with ventilation and circulation limiting supply under cold conditions and diffusion limiting supply under warm conditions. State-space habitats reveal that species with these physiological traits inhabit lowest O2 waters near the optimal temperature for hypoxia tolerance, and are restricted to higher O2 at temperatures above and below this optimum. These results imply that tolerance to low oxygen can decline under cold and warm conditions and may influence species range limits.
Mortality Associated with Extreme Heat in Washington State: The historical and projected public health burden
(2021-08-26) Arnold, Logan; Busch Isaksen, Tania
Extreme heat is one of the most important pathways illustrating the connection between climate and human health, including in temperate areas such as the Pacific Northwest. Moreover, climate change is expected to exacerbate this important public health issue. This research has two components. First, a time-stratified case-crossover analysis is used to characterize the historical (1980 – 2018) association between summertime (May - September) heat and non-traumatic mortality in Washington state. A separate analysis is conducted for each of the state’s climate divisions to produce ten distinct exposure-response curves expressing odds of mortality as a function of humidex. Stratified analyses are used to assess the impact of age, sex, race/ethnicity, and select causes of death, and the reported results are pooled across all climate divisions using fixed-effects meta-analysis. Second, the historical heat-mortality relationship is combined with climate change projections to estimate the impact of a changing climate on heat-related deaths in 2030, 2050, and 2080 under two warming scenarios. The odds ratio and 95% confidence intervals of mortality at the 99th percentile of humidex compared to the 50th percentile did not include the null value in four of the ten climate divisions, and the point estimates in all ten divisions indicated increased risk of mortality at high values of humidex. Across Washington, the odds of mortality are 8% higher (6%, 10%) on 99th percentile days compared to 50th percentile days. Risk is higher for women than men and for Blacks than Whites. Risk also increases with age and for diabetic, circulatory, cardiovascular, ischemic, cerebrovascular, and respiratory deaths. 95% confidence intervals of projected heat-attributable mortality did not overlap with zero in three of the ten climate divisions. In these three divisions, the average percent increase in heat-attributable deaths across both warming scenarios is 35%, 135%, and 603% in 2030, 2050, and 2080, respectively, over the historical period. This research is the most extensive study of heat-related mortality in Washington to date and can help inform public health initiatives which aim to improve both present and future health outcomes in the state.
Group size affects predation risk and foraging success in Pacific salmon at sea
(2021-07-07) Polyakov, Anne Y; Berdahl, Andrew M
Grouping (schooling, flocking, herding) is broadly distributed across taxa and environments, and is particularly common in marine fishes. A rich body of theory outlines ways in which grouping can enhance individual fitness, especially by improved predator avoidance and foraging efficiency. However, such theories are difficult to test in the wild, especially in marine environments where observing individuals is challenging, and quantifying predation risk or foraging success is often impossible. To overcome these difficulties, I analyzed a multi-decadal data set from sampling of Pacific salmon (Oncorhynchus spp.) in the ocean. Across all species, individuals in larger groups had lower risk of predator attack, inferred from wounds on surviving fish. Within groups, outliers (smaller and larger fish) were disproportionately attacked by predators, suggesting that collective predator avoidance resulted from a confusion effect. For slower growing species, individuals in larger groups had lower foraging success, indicating that schooling presented a trade-off between predator avoidance and resource consumption. In contrast, for faster growing species, individuals in larger groups had higher foraging success, indicating that this trade off may not exist and individuals in larger groups may even benefit from collective foraging. These results support long-standing theories on the benefits of group living that have rarely been tested in wild populations, and challenge earlier assertions that adult salmon do not school in the marine environment. Ultimately, these results indicate that survival and growth are group-size dependent and thus understanding the relationship between group-size distributions and population size may be critical to unraveling ecology and population dynamics for these and other marine fishes.
The Relationship Between Natural Environments and Subjective Well-being as Measured by Sentiment Expressed on Twitter
(2021-07-07) Lin, Yian; Lawler, Joshua J
There is growing evidence that time spent in nature can affect well-being. Nonetheless, assessing this relationship can be difficult. We used social media data—1,971,045 geolocated tweets sent by 81,140 users from locations throughout Seattle, Washington, USA—to advance our understanding of the relationship between subjective well-being and natural environments. Specifically, we quantified the relationships between sentiment (negative/neutral/positive) expressed in individual geolocated tweets and their surrounding environments focusing on land-cover type, tree-canopy density, and urban parks. Controlling for multiple covariates such as location types and weather conditions, we estimated three random-intercept partial proportional odds models corresponding to the three environmental indicators. Our results suggest that for a given type of land-use, tweets sent from some natural land-cover types were less likely to be negative compared to tweets sent from the urban-built land-cover type. We also found that for tweets sent in industrial zones, the association between tree-canopy coverage and sentiment polarity was positive: an increase in tree-canopy coverage was associated with a lower probability of having negative sentiments and with a higher probability of having positive sentiments; but for tweets sent in commercial/mixed zones, the association between tree-canopy coverage and sentiment polarity was negative. For parks, tweets sent from urban parks in commercial/mixed zones and residential zones were less likely to be negative compared to tweets sent from outside parks. In industrial zones, only tweets sent from large natural parks (with area >= 40,000 sf2 and impervious surface < 30%) were less likely to be negative. Surprisingly, we also found that tweets sent from large natural parks in residential zones were less likely to be positive compared to tweets sent from outside parks. Geolocated social media data allows nuanced analyses that reveal the complexity of the relationship between subjective well-being and natural environments.
Accounting for model uncertainties in statistical forecasts of wildfire parameters
(2021-03-19) Podschwit, Harry Richard; Alvarado, Ernesto C; Cullen, Alison
Gauging the magnitude of model uncertainty and incorporating model uncertainty into predictions is of critical importance when models are used to inform wildfire-related decisions, where ignoring potential risks threaten human health, property, and the environment. Although techniques exist for addressing model uncertainty, these uncertainties are commonly ignored in most analyses. In this dissertation, I will evaluate the effects of model uncertainty on statistical predictions of wildfire activity in multiple contexts and propose techniques to incorporate these uncertainties into predictions. I will determine how uncertainty in the choice of predictive model and climate model influence forecasts of very-large fire activity in the second half of the 21st century, and integrate this uncertainty using a novel Bayesian model averaging approach to produce robust predictions. I find that when these model uncertainties are accounted for, that one may conclude, across the suite of model choices, that the frequency of very-large wildfires should be expected to increase in most regions of the United States if climate changes are not mitigated. The effects of model uncertainty will also be explored in the context of predicting final wildfire size for individual fires that have no yet finished growing. Specifically, I will gauge how the choice of utility function and the inclusion of growth information that is unavailable early in the wildfire’s life alters the predictive ability of statistical models of final fire size and the stability of the model structure. I find that predictions of fire size can drastically change when new utility functions are considered, particularly in models that use growth information. I also find that the covariates used in the best model are sensitive to the choice of utility function, and that no single model is likely to optimally address the preferences of all wildfire-related decisionmakers they are intended to inform. The results of this analysis that (1) the preferred model will often change when new performance measures are considered, and (2) that the preferred model may change over time. I also present a method of integrating the model uncertainties associated with time-varying covariates and ill-defined utility functions into a single predictive distribution using Bayesian model averaging. I find that this novel model averaging approach generally improves predictive performance across a number of performance measures compared to the individual models contained within it. I discuss how the novel methods developed can be applicable to other forecasting applications and how they might allow wildfire professionals make better decisions.

Browse

Recent Submissions