Summary1 Predicting and explaining the distribution and density of species is one of the oldest concerns in ecology. Species distributions can be estimated using geostatistical methods, which estimate a latent spatial variable explaining observed variation in densities, but geostatistical methods may be imprecise for species with low densities or few observations. Additionally, simple geostatistical methods fail to account for correlations in distribution among species and generally estimate such cross-correlations as a post hoc exercise. 2 We therefore present spatial factor analysis (SFA), a spatial model for estimating a low-rank approximation to multivariate data, and use it to jointly estimate the distribution of multiple species simultaneously. We also derive an analytic estimate of cross-correlations among species from SFA parameters. 3 As a first example, we show that distributions for 10 bird species in the breeding bird survey in 2012 can be parsimoniously represented using only five spatial factors. As a second case study, we show that forward prediction of catches for 20 rockfishes (Sebastes spp.) off the U.S. West Coast is more accurate using SFA than analysing each species individually. Finally, we show that single-species models give a different picture of cross-correlations than joint estimation using SFA. 4 Spatial factor analysis complements a growing list of tools for jointly modelling the distribution of multiple species and provides a parsimonious summary of cross-correlation without requiring explicit declaration of habitat variables. We conclude by proposing future research that would model species cross-correlations using dissimilarity of species' traits, and the development of spatial dynamic factor analysis for a low-rank approximation to spatial time-series data.

Abstract. The study of population dynamics requires unbiased, precise estimates of abundance and vital rates that account for the demographic structure inherent in all wildlife and plant populations. Traditionally, these estimates have only been available through approaches that rely on intensive mark-recapture data. We extended recently developed Nmixture models to demonstrate how demographic parameters and abundance can be estimated for structured populations using only stage-structured count data. Our modeling framework can be used to make reliable inferences on abundance as well as recruitment, immigration, stage-specific survival, and detection rates during sampling. We present a range of simulations to illustrate the data requirements, including the number of years and locations necessary for accurate and precise parameter estimates. We apply our modeling framework to a population of northern dusky salamanders (Desmognathus fuscus) in the mid-Atlantic region (USA) and find that the population is unexpectedly declining. Our approach represents a valuable advance in the estimation of population dynamics using multistate data from unmarked individuals and should additionally be useful in the development of integrated models that combine data from intensive (e.g., mark-recapture) and extensive (e.g., counts) data sources.

2021. Estimating carrying capacity for juvenile salmon using quantile random forest models.

Anthropogenic impacts on riverine systems have, in part, led to management concerns regarding the population status of species using these systems. In an effort to assess the efficacy of restoration actions, and in order to improve monitoring of species of concern, managers have turned to PIT (passive integrated transponder) tag studies with in-stream detectors to monitor movements of tagged individuals throughout river networks. However, quantifying movements in a river network using PIT tag data with incomplete coverage and imperfect detections presents a challenge. We propose a flexible Bayesian analytic framework that models the imperfectly detected movements of tagged individuals in a nested PIT tag array river network. This model structure provides probabilistic estimates of upstream migration routes for each tagged individual based on a set of underlying nested state variables. These movement estimates can be converted into abundance estimates when an estimate of abundance is available for a location within the river network. We apply the model framework to data from steelhead (Oncorhynchus mykiss) in the Upper Columbia River basin and evaluate model performance (precision/variance of simulated population sizes) as a function of population tagging rates and PIT tag array detection probability densities within the river system using a simulation framework. This simulation framework provides both model validation (precision) and the ability to evaluate expected performance improvements (variance) due to changes in tagging rates or PIT receiver array configuration. We also investigate the impact of different network configurations on model estimates. Results from such investigations can help inform decisions regarding future monitoring and management.

This study examined how a suite of habitat and environmental variables relate to the ability of a stream surveyor to identify (observer efficiency) and distinguish (observer accuracy) steelhead (Oncorhynchus mykiss) redds from other stream features. Two existing spawning survey protocols that included one or two redd observers were used to develop models to estimate redd observer error. In most cases, steelhead redd abundances using raw redd counts were underestimated. Mean annual rates of observer efficiency ranged from 0.44 to 0.57, and observer accuracy ranged from 0.67 to 0.83. Regardless of the observer error model used, adjusted annual redd abundance estimates were generally unbiased (range 1.6–0.6 redds). A Gaussian area-under-the-curve methodology that incorporates redd count data and observer error rates was used to generate unbiased estimates of steelhead redd abundance in the Wenatchee (170 redds, coefficient of variation (CV) = 44%) and Methow (106 redds, CV = 41%) rivers. Unbiased estimates of redd abundance will help inform new population viability analyses to better prioritize those populations with the greatest conservation need.

The primary conservation prioritization tool for spring Chinook Salmon (Oncorhynchus tshawytscha), population viability analysis, is often conducted with biased spawner abundance data with no associated statistical uncertainty or error. This study estimated observation error of surveyors counting redds in two spring Chinook Salmon populations where hatchery supplementation is implemented as a conservation tool. Habitat complexity, redd density and the amount of observer experience were important in estimating error rates. Increases in both habitat complexity and experience reduced net error rates. Conversely, net error rates increased as redd density increased. Unbiased estimates of redd abundance were generated and converted to spawner abundance using population specific redd expansion factors. The precision (i.e., coefficient of variation [CV]) of spawner abundance estimates were similar in the Wenatchee (natural CV = 5%; hatchery CV = 6%) and Methow (natural CV = 5%; hatchery CV = 2%) watersheds because average net error rates were similar (Wenatchee = −0.1512; Methow = −0.1748). This study addresses a criticism of population viability analysis (i.e., parameter uncertainty) that should result in more scientifically defensible conservation priorities and recommendations that can be implemented with greater certainty.

The European green crab, Carcinus maenas, was first documented in San Francisco Bay in 1989, and has since spread north along the west coast of North America. The spread of this invasion has not been a smooth expansion, which has raised questions about the underlying causes of variation in recruitment. We modeled larval development and transport along the West Coast by employing an individual-based model that incorporated oceanographic model output of water temperature and ocean currents at fine spatial and temporal scales. The distance that larvae were advected depended primarily on the timing of larval release. However, the effect of seasonal ocean currents varied across latitude and years. Our results imply that the furthest northern transport from California occurs when larvae are released from Humboldt Bay during the fall of an El Niño year, making this a particularly risky time for invasion to Oregon and Washington estuaries. To precisely predict future spread and potential impacts of green crab, we recommend further empirical research to determine the precise timing of larval release and seasonal abundance of green crab larvae from North American west coast populations.

Forecasting the risk of population decline is crucial in the realm of biological conservation and figures prominently in population viability analyses (PVA). A common form of available data for a PVA is population counts through time. Previous research has suggested that improving estimates of population trends and risk from count data depends on longer observation periods, but that is often impractical or undesirable. Making multiple observations within a single time step is an alternative way to gather more data without extending the observation period. In this paper, we examine the trade-off between the length of the time period over which observations of the population have been taken and the total number of observations or samples that have been recorded through an analysis of simulated data. We found that when the ratio of process error to measurement error variance is high, more precise estimates of quasi-extinction risks can be obtained if replicated observations are taken at each time step, but when the ratio is low, replicated observations add little benefit in improving precision. These results can be used to efficiently design effective monitoring schemes for species of conservation concern.