Abundance estimation of Adélie penguins at the Esperanza/Hope Bay mega colony

ANTARCTIC krill (Euphausia superba) is the dominant component of the diet of many whales, seals, birds, fish and squid, and their survival could be affected by a reduction in krill abundance due to fishing1. Commercial fishing takes nearly half a million tonnes of krill annually2 and accurate estimates of abundance are needed for rational management of this resource. Indirect estimates of abundance based on predator consumption rates give a roughly estimated total annual production of several hundred million tonnes3. The life-span of krill is at least two and maybe five years, so the standing stock would need to be of at least the same magnitude as the annual production. But direct estimates of abundance using nets and acoustics have indicated biomass figures lower by an order of magnitude3. Acoustic estimates are sensitive to the scaling factor or target strength (TS) used to convert echo energy to absolute abundance. Previous published values for TS (ref. 4), when applied to survey data, gave estimates of krill abundance that were much too low to account for local bird and seal predation rates near South Georgia5, and were also lower than expected when compared with density estimates from net hauls6. We therefore sought to determine TS using direct measurements developed for fish studies7, and also by applying models developed for other crustacean zooplankton8. Our results show that krill TS is much lower than previously thought, and consequently that acoustic estimates of krill abundance are likely to have been gross underestimates.

Deer (Cervidae) are key components of many ecosystems and estimating deer abundance or density is important to understanding these roles. Many field methods have been used to estimate deer abundance and density, but the factors determining where, when, and why a method was used, and its usefulness, have not been investigated. We systematically reviewed journal articles published during 2004–2018 to evaluate spatio‐temporal trends in study objectives, methodologies, and deer abundance and density estimates, and determine how they varied with biophysical and anthropogenic attributes. We also reviewed the precision and bias of deer abundance estimation methods. We found 3,870 deer abundance and density estimates. Most estimates (58%) were for white‐tailed deer (Odocoileus virginianus), red deer (Cervus elaphus), and roe deer (Capreolus capreolus). The 6 key methods used to estimate abundance and density were pedestrian sign (track or fecal) counts, pedestrian direct counts, vehicular direct counts, aerial direct counts, motion‐sensitive cameras, and harvest data. There were regional differences in the use of these methods, but a general pattern was a temporal shift from using harvest data, pedestrian direct counts, and aerial direct counts to using pedestrian sign counts and motion‐sensitive cameras. Only 32% of estimates were accompanied by a measure of precision. The most precise estimates were from vehicular spotlight counts and from capture–recapture analysis of images from motion‐sensitive cameras. For aerial direct counts, capture–recapture methods provided the most precise estimates. Bias was robustly assessed in only 16 studies. Most abundance estimates were negatively biased, but capture–recapture methods were the least biased. The usefulness of deer abundance and density estimates would be substantially improved by 1) reporting key methodological details, 2) robustly assessing bias, 3) reporting the precision of estimates, 4) using methods that increase and estimate detection probability, and 5) staying up to date on new methods. The automation of image analysis using machine learning should increase the accuracy and precision of abundance estimates from direct aerial counts (visible and thermal infrared, including from unmanned aerial vehicles [drones]) and motion‐sensitive cameras, and substantially reduce the time and cost burdens of manual image analysis.

Summary1. Models have been devised previously that allow the estimation of abundance from detection data of unmarked individuals while accounting for imperfect detection, but these are restricted to models for discrete sampling protocols, i.e. replicated detection/non‐detection or count data. Furthermore, these models assume that the detections from each individual are independent; however, there are cases in which this assumption is likely to be violated. For example, in surveys along transects, clustering in the signs left by each individual could be expected.2. Here, we propose models to estimate abundance from species‐detection data collected continuously along transects considering two cases: (i) independent detections and (ii) clustering within the detections of each individual. We account for clustering by describing the detection process as a Markov‐modulated Poisson process. We study the properties of the estimators via simulation, assessing the impact of unmodelled detection clustering.3. We show that bias may be induced in the estimator of abundance if clustering in individual detections is not accounted for and how an estimator with better coverage properties is obtained if clustering is modelled. We demonstrate that both abundance and the clustering pattern can be well estimated simultaneously, given enough data.4. To illustrate our approach, we fit the models to tiger pugmark detection data from transect surveys in Kerinci Seblat National Park in Sumatra. The analysis suggested strong abundance‐induced heterogeneity in detections when clustering was disregarded, but the evidence reduced drastically when clustering was accounted for. This example illustrates how unmodelled clustering can affect the estimation of abundance.5. Estimates of abundance need to be reliable to ensure that conservation and management interventions are not misguided. Provided certain model assumptions are met, abundance can be estimated from detection data of unmarked individuals. This requires an adequate description of the detection process, or otherwise, bias may be induced in the abundance estimator. The models and discussion provided here deal with the issue of clustering within the detections of individuals and are of relevance for ecologists interested in methodological developments for the estimation of animal abundance.

Food availability often drives consumer population dynamics. However, food availability may also influence capture probability, which if not accounted for may create bias in estimating consumer abundance and confound the effects of food availability on consumer population dynamics. This study compared two commonly used abundance indices (minimum number alive (MNA) and number of animals captured per night per grid) with an abundance estimator based on robust design model as applied to the white-footed mouse ( Peromyscus leucopus (Rafinesque, 1818)) in food supplementation experiments. MNA consistently generated abundance estimates similar to the robust design model, regardless of food supplementation. The number of animals captured per night per grid, however, consistently generated lower abundance estimates compared with MNA and the robust design model. Nevertheless, the correlations between abundance estimates from MNA, number of animals captured, and robust design model were not influenced by food supplementation. This study demonstrated that food supplementation is not likely to create bias among these different measures of abundance. Therefore, there is a great potential for conducting meta-analysis of food supplementation effect on consumer population dynamics (particularly in small mammals) across studies using different abundance indices and estimators.

Conducting surveys from blinds when supplemental feed (bait) has been provided has not been evaluated for estimating parameters of ungulate populations. We conducted blind count surveys of white-tailed deer (Odocoileus virginianus) in a 214-ha enclosure in central Texas, USA, in 2007 and 2008 to address 2 main objectives: 1) to evaluate a blind count survey protocol developed for use on small parcels of land, and 2) to use data collected from blind count surveys to conduct simulations to evaluate the reliability of abundance and sex ratio estimates obtained from Bowden's estimator. In each year population abundance (2007: 60; 2008: 48) and sex ratio (M:F, 2007: 0.58; 2008: 0.71) were known as were sighting frequencies of every animal. The enclosure had 5 blinds and we baited each blind with corn. We encountered many deer during surveys because there were only 2 deer in 2007 and 1 deer in 2008 that we did not view from blinds ≥1 time. To evaluate bias and precision of abundance and sex ratio estimates we conducted 10,000 bootstrap simulations. We evaluated both parameters in relation to the percentage of each population marked, number of surveys conducted from blinds, and whether surveys were conducted in the morning, evening, or both morning and evening. Also, we evaluated abundance in relation to whether we identified animals with unique marks to individual, and we evaluated sex ratio in relation to intersexual distribution of marks. Abundance estimates were less biased and more precise when we uniquely identified all marked animals and 40–70% of the population was marked. Sex ratio estimates were less biased when 40–70% of the population was marked and surveys were conducted in the morning and evening. Sex ratio estimates, however, were less precise than abundance estimates. Unbiased estimates of white-tailed deer population parameters can be obtained from blind count surveys conducted on small parcels of enclosed land and when animals are baited.

Abstract: Conducting surveys from blinds when supplemental feed (bait) has been provided has not been evaluated for estimating parameters of ungulate populations. We conducted blind count surveys of white‐tailed deer (Odocoileus virginianus) in a 214‐ha enclosure in central Texas, USA, in 2007 and 2008 to address 2 main objectives: 1) to evaluate a blind count survey protocol developed for use on small parcels of land, and 2) to use data collected from blind count surveys to conduct simulations to evaluate the reliability of abundance and sex ratio estimates obtained from Bowden's estimator. In each year population abundance (2007: 60; 2008: 48) and sex ratio (M:F, 2007: 0.58; 2008: 0.71) were known as were sighting frequencies of every animal. The enclosure had 5 blinds and we baited each blind with corn. We encountered many deer during surveys because there were only 2 deer in 2007 and 1 deer in 2008 that we did not view from blinds ≥1 time. To evaluate bias and precision of abundance and sex ratio estimates we conducted 10,000 bootstrap simulations. We evaluated both parameters in relation to the percentage of each population marked, number of surveys conducted from blinds, and whether surveys were conducted in the morning, evening, or both morning and evening. Also, we evaluated abundance in relation to whether we identified animals with unique marks to individual, and we evaluated sex ratio in relation to intersexual distribution of marks. Abundance estimates were less biased and more precise when we uniquely identified all marked animals and 40–70% of the population was marked. Sex ratio estimates were less biased when 40–70% of the population was marked and surveys were conducted in the morning and evening. Sex ratio estimates, however, were less precise than abundance estimates. Unbiased estimates of white‐tailed deer population parameters can be obtained from blind count surveys conducted on small parcels of enclosed land and when animals are baited.

In the face of increasing extinction rates, it is vital to have estimates of relative and absolute species abundance and their relationship to important factors. For species that live in the oceans or large lakes, this can be a difficult task. Here, we present a method for estimating absolute abundance from a single binary acoustic time series. The dependence in the series is exploited to allow the estimation of abundance when some animals remain hidden, and in the face of uncertainty about the range over which sounds carry. Simulations show that the method works well, even when some assumptions are violated. The method is illustrated using data on sperm whales in the Sargasso Sea.

We estimated the abundance of humpback whales in the North Pacific by capture‐recapture methods using over 18,000 fluke identification photographs collected in 2004–2006. Our best estimate of abundance was 21,808 (CV = 0.04). We estimated the biases in this value using a simulation model. Births and deaths, which violate the assumption of a closed population, resulted in a bias of +5.2%, exclusion of calves in samples resulted in a bias of −10.5%, failure to achieve random geographic sampling resulted in a bias of −0.4%, and missed matches resulted in a bias of +9.3%. Known sex‐biased sampling favoring males in breeding areas did not add significant bias if both sexes are proportionately sampled in the feeding areas. Our best estimate of abundance was 21,063 after accounting for a net bias of +3.5%. This estimate is likely to be lower than the true abundance due to two additional sources of bias: individual heterogeneity in the probability of being sampled (unquantified) and the likely existence of an unknown and unsampled breeding area (−8.7%). Results confirm that the overall humpback whale population in the North Pacific has continued to increase and is now greater than some prior estimates of prewhaling abundance.

Many groundfish stocks on the Atlantic coast of North America are monitored by bottom trawl surveys which use a stratified random design. The resulting estimates of abundance often exhibit high variability within and between surveys, particularly on inter-annual time scales, and inconsistencies in the estimates of relative year-class strength over time. These inconsistencies, or year-effects, may reflect changes in the distribution or catchability of the fish that may be related in turn to environmental factors such as water temperature and salinity. The estimates of cod abundance derived from the research vessel trawl surveys conducted in July over the eastern Scotian Shelf in NAFO areas 4Vs and 4W from 1970–1993 are shown to exhibit these features. Analyses of these survey data also show that cod exhibit age- and area-specific associations with near-bottom water temperature and salinity ranges that are consistent with the properties of a Cold Intermediate Layer (CIL) water mass in the area and that inter-annual variability in the estimates of abundance is correlated with the area of the bottom found within the CIL. A model relating the variability in the CIL to the estimates of age-specific cod abundance obtained from the surveys was used to derive new survey indices of abundance. These new indices do not have detectable year-effects and may provide more consistent estimates of true relative year-class strength than the original time series.

Obtaining robust estimates of population abundance is a central challenge hindering the conservation and management of many threatened and exploited species. Close-kin mark-recapture (CKMR) is a genetics-based approach that has strong potential to improve the monitoring of data-limited species by enabling estimates of abundance, survival, and other parameters for populations that are challenging to assess. However, CKMR models have received limited sensitivity testing under realistic population dynamics and sampling scenarios, impeding the application of the method in population monitoring programs and stock assessments. Here, we use individual-based simulation to examine how unmodeled population dynamics and aging uncertainty affect the accuracy and precision of CKMR parameter estimates under different sampling strategies. We then present adapted models that correct the biases that arise from model misspecification. Our results demonstrate that a simple base-case CKMR model produces robust estimates of population abundance with stable populations that breed annually; however, if a population trend or non-annual breeding dynamics are present, or if year-specific estimates of abundance are desired, a more complex CKMR model must be constructed. In addition, we show that CKMR can generate reliable abundance estimates for adults from a variety of sampling strategies, including juvenile-focused sampling where adults are never directly observed (and aging error is minimal). Finally, we apply a CKMR model that has been adapted for population growth and intermittent breeding to two decades of genetic data from juvenile lemon sharks (Negaprion brevirostris) in Bimini, Bahamas, to demonstrate how application of CKMR to samples drawn solely from juveniles can contribute to monitoring efforts for highly mobile populations. Overall, this study expands our understanding of the biological factors and sampling decisions that cause bias in CKMR models, identifies key areas for future inquiry, and provides recommendations that can aid biologists in planning and implementing an effective CKMR study, particularly for long-lived data-limited species.

Whether a species is rare or overabundant, accurate estimates of population abundance are essential for the development and assessment of conservation plans and management goals. Aerial surveys are commonly used to estimate population abundance and a variety of methods have been used to account for recognized biases associated with imperfect detection. Rarely addressed, however, is the possibility of recording duplicate observations and the influence of duplicate observations on estimates of abundance. Using data provided by Global Positioning System (GPS)‐collared bison ( Bison bison ) and helicopter survey paths conducted during 2011–2013 in the Henry Mountains, Utah, USA, we determined whether GPS‐collared bison were observed, missed, or observed multiple times. Using these data, covariates hypothesized to influence observation errors, and generalized linear models, we found the probability of detecting a GPS‐collared bison (study‐wide mean of 95%) was consistently greater for individuals in larger groups and less rugged terrain. We also found that the probability of recording a duplicate observation (study‐wide mean of 4.4%) increased with time since the prior survey observation. By incorporating both types of observation error into a modified Horvitz–Thompson estimator, more accurate estimates of bison abundance can be attained, particularly in years when animals segregate into small groups, utilize terrain that makes them difficult to see, or have ample time to move into areas not yet surveyed and be counted again. Our approach of accounting for multiple sources of observation error can easily be applied to aerial surveys for other species in other systems, and has the potential to improve the rigor of wildlife abundance estimation from aerial surveys. © 2016 The Wildlife Society.

A critical variable in both ecological and conservation field studies is determining how many individuals of a species are present within a defined sampling area. Labor intensive techniques such as capture-mark-recapture and removal sampling may provide estimates of abundance, but there are many logistical constraints to their widespread application. Many studies on terrestrial and aquatic salamanders use counts as an index of abundance, assuming that detection remains constant while sampling. If this constancy is violated, determination of detection probabilities is critical to the accurate estimation of abundance. Recently, a model was developed that provides a statistical approach that allows abundance and detection to be estimated simultaneously from spatially and temporally replicated counts. We adapted this model to estimate these parameters for salamanders sampled over a six year period in area-constrained plots in Great Smoky Mountains National Park. Estimates of salamander abundance varied among years, but annual changes in abundance did not vary uniformly among species. Except for one species, abundance estimates were not correlated with site covariates (elevation, soil and water pH, conductivity, air and water temperature). The uncertainty in the estimates was so large as to make correlations ineffectual in predicting which covariates might influence abundance. Detection probabilities also varied among species and sometimes among years for the six species examined. We found such a high degree of variation in our counts and in estimates of detection among species, sites, and years as to cast doubt upon the appropriateness of using count data to monitor population trends using a small number of area-constrained survey plots. Still, the model provided reasonable estimates of abundance that could make it useful in estimating population size from count surveys.

Accurate estimates of abundance are imperative for successful conservation and management. Classical, stratified abundance estimators provide unbiased estimates of abundance, but such estimators may be imprecise and impede assessment of population status and trend when the distribution of individuals is highly variable in space. Model-based procedures that account for important environmental covariates can improve overall precision, but frequently there is uncertainty about the contribution of particular environmental variables and a lack of information about variables that are important determinants of abundance. We develop a general semiparametric mixture model that incorporates measured habitat variables and a nonparametric smoothing term to account for unmeasured variables. We contrast this spatial habitat approach with two stratified abundance estimators and compare the three models using an intensively managed marine fish, darkblotched rockfish (Sebastes crameri). We show that the spatial habitat model yields more precise, biologically reasonable, and interpretable estimates of abundance than the classical methods. Our results suggest that while design-based estimators are unbiased, they may exaggerate temporal variability of populations and strongly influence inference about population trend. Furthermore, when such estimates are used in broader meta-analyses, such imprecision may affect the broader biological inference (e.g., the causes and consequences of the variability of populations).

We attempted to determine whether electrofishing removal estimates or single-pass snorkeling was a more reliable method for Oregon Department of Fish and Wildlife (ODFW) monitoring of juvenile coho salmon (Oncorhynchus kisutch) and steelhead (O. mykiss) abundance and occupancy trends. Based on 1997 to 2000 data, we assumed abundance estimates from the method that tracked more closely with parental abundance would better approximate true juvenile abundance. Parental abundance from spawning ground surveys and juvenile abundance metrics unique to each method were estimated from 2000 to 2004 and 2007 to 2008. Parental abundance did not explain the variation in juvenile abundance from either method (r2 < 0.22), invalidating our assumption, but results had relevance for snorkel surveys used in ODFW monitoring. For both species, correlations between density (fish m2 -1) and abundance (quantity, based on fish km-1) estimates from snorkeling were weak (r < 0.379), but correlations between abundance estimates from both methods were strong (r > 0.846), implying abundance was more appropriate than density for ODFW monitoring. Neither method could sample all habitats, and annually variable proportions of coho salmon (15 to 47%) and steelhead (0 to 24%) abundance estimates obtained by electrofishing were in pools too shallow to meet the ODFW depth criterion for snorkeling. This resulted in lowering the criterion to ≥ 20 cm in 2010. The lower criterion, relative to original, has not shown differences in trends, but 30% more pools have been sampled, resulting in 23% higher abundance estimates with 10% proportionately smaller confidence intervals. These changes improved ODFW monitoring and related management decisions.

First posted September 11, 2023 For additional information, contact: Western Ecological Research CenterU.S. Geological Survey3020 State University Drive EastSacramento, California 95819 Marbled Murrelets (Brachyramphus marmoratus) have been listed as “endangered” by the State of California and “threatened” by the U.S. Fish and Wildlife Service since 1992 in California, Oregon, and Washington. Information regarding murrelet abundance, distribution, and habitat associations is critical for risk assessment, effective management, evaluation of conservation efficacy, and ultimately, the meeting of Federal- and State-mandated recovery efforts. From 1999 to present, line-transect surveys have been performed to estimate at-sea abundance and reproductive output of Marbled Murrelets in the marine environment in U.S. Fish and Wildlife Service Conservation Zone 6 (San Francisco Bay to Point Sur in central California). Using this long-term annual time series, we developed a new and comprehensive analytical framework to estimate annual murrelet abundance and trend at sea, evaluated the effectiveness of spatial and temporal components of the monitoring study design, assessed two measures of annual murrelet reproductive output, and developed new spatial models to map murrelet at-sea density and estimate model-based annual at-sea abundances. The long-term average, design-based after-hatch-year (AHY) abundance estimate for the study area was 376 murrelets (range: 163–586 annually), and we did not detect any significant trend during the 23 years of monitoring. Spatial-model-based AHY abundance estimates were similar to design-based estimates but with smaller estimated variance. The AHY murrelets were most abundant nearshore, with little annual variation; alongshore, distribution was more annually variable, and some long-term hotspots occurred, particularly around Point Año Nuevo. The AHY murrelet densities were greatest in July and least in June and August. The long-term average hatch-year (HY) abundance estimate was 13 murrelets (range: 0–31 annually), and the long-term average HY:AHY ratio was 0.052; both metrics indicated similar interannual patterns. Evidence of a significant trend in either metric of reproductive output was not detected; although large overlap among interannual abundance and ratio estimates at the 95-percent confidence interval level made it difficult to evaluate interannual differences. Despite the apparent long-term stability in murrelet abundance in this region from 1999 to 2021, future long-term annual monitoring at sea will be critical to determine if the large-scale August 2020 CZU Santa Cruz Mountain wildfire that occurred adjacent to our study area affects local murrelet at-sea abundance and distribution. We also evaluated potential changes to survey and analytical design that could benefit this monitoring program in the future. Results indicated that eliminating the offshore stratum, focusing more effort on the nearshore stratum, and doing fewer surveys focused on a narrower timeframe could maintain or improve AHY trend estimates while preserving the ability to compare them to past years.