Kevin T. Shoemaker

Imputation of missing data in life‐history trait datasets: which approach performs the best?

1. Despite efforts in data collection, missing values are commonplace in life-history trait databases. Because these values typically are not missing randomly, the common practice of removingmissing data not only reduces sample size, but also introduces bias that can lead to incorrect conclusions. Imputingmissing values is a potential solution to this problem. Here, we evaluate the performance of four approaches for estimating missing values in trait databases (K-nearest neighbour (kNN), multivariate imputation by chained equations (mice), missForest and Phylopars), and testwhether imputed datasets retain underlying allometric relationships among traits. 2. Starting with a nearly complete trait dataset on the mammalian order Carnivora (using four traits), we artificially removed values so that the percent ofmissing values ranged from 10% to 80%. Using the original values as a reference, we assessed imputation performance using normalized root mean squared error.We also evaluated whether including phylogenetic information improved imputation performance in kNN, mice, and missForest (it is a required input in Phylopars). Finally, we evaluated the extent to which the allometric relationship between two traits (body mass and longevity) was conserved for imputed datasets by looking at the difference (bias) between the slope of the original and the imputed datasets or datasets with missing values removed. 3. Three of the tested approaches (mice, missForest and Phylopars), resulted in qualitatively equivalent imputation performance, and all had significantly lower errors than kNN. Adding phylogenetic information into the imputation algorithms improved estimation of missing values for all tested traits. The allometric relationship between body mass and longevity was conserved when up to 60% of data were missing, either with or without phylogenetic information, depending on the approach. This relationship was less biased in imputed datasets compared to datasets withmissing values removed, especially whenmore than 30%of values weremissing. 4. Imputations provide valuable alternatives to removing missing observations in trait databases as they produce low errors and retain relationships among traits. Although we must continue to prioritize data collection on species traits, imputations can provide a valuable solution for conducting macroecological and evolutionary studies using life-history trait databases.

Reexamining the minimum viable population concept for long-lived species.

For decades conservation biologists have proposed general rules of thumb for minimum viable population size (MVP); typically, they range from hundreds to thousands of individuals. These rules have shifted conservation resources away from small and fragmented populations. We examined whether iteroparous, long-lived species might constitute an exception to general MVP guidelines. On the basis of results from a 10-year capture-recapture study in eastern New York (U.S.A.), we developed a comprehensive demographic model for the globally threatened bog turtle (Glyptemys muhlenbergii), which is designated as endangered by the IUCN in 2011. We assessed population viability across a wide range of initial abundances and carrying capacities. Not accounting for inbreeding, our results suggest that bog turtle colonies with as few as 15 breeding females have >90% probability of persisting for >100 years, provided vital rates and environmental variance remain at currently estimated levels. On the basis of our results, we suggest that MVP thresholds may be 1-2 orders of magnitude too high for many long-lived organisms. Consequently, protection of small and fragmented populations may constitute a viable conservation option for such species, especially in a regional or metapopulation context.

Demographic Outcomes and Ecosystem Implications of Giant Tortoise Reintroduction to Española Island, Galapagos

Restoration of extirpated species via captive breeding has typically relied on population viability as the primary criterion for evaluating success. This criterion is inadequate when species reintroduction is undertaken to restore ecological functions and interactions. Herein we report on the demographic and ecological outcomes of a five-decade-long population restoration program for a critically endangered species of “ecosystem engineer”: the endemic Española giant Galapagos tortoise (Chelonoidis hoodensis). Our analysis of complementary datasets on tortoise demography and movement, tortoise-plant interactions and Española Island’s vegetation history indicated that the repatriated tortoise population is secure from a strictly demographic perspective: about half of tortoises released on the island since 1975 were still alive in 2007, in situ reproduction is now significant, and future extinction risk is low with or without continued repatriation. Declining survival rates, somatic growth rates, and body condition of repatriates suggests, however, that resources for continued population growth are increasingly limited. Soil stable carbon isotope analyses indicated a pronounced shift toward woody plants in the recent history of the island’s plant community, likely a legacy of changes in competitive relations between woody and herbaceous plants induced by now-eradicated feral goats and prolonged absence of tortoises. Woody plants are of concern because they block tortoise movement and hinder recruitment of cactus–a critical resource for tortoises. Tortoises restrict themselves to remnant cactus patches and areas of low woody plant density in the center of the island despite an apparent capacity to colonize a far greater range, likely because of a lack of cactus elsewhere on the island. We conclude that ecosystem-level criteria for success of species reintroduction efforts take much longer to achieve than population-level criteria; moreover, reinstatement of endangered species as fully functioning ecosystem engineers may often require large-scale habitat restoration efforts in concert with population restoration.

How interactions between animal movement and landscape processes modify local range dynamics and extinction risk.

Forecasts of range dynamics now incorporate many of the mechanisms and interactions that drive species distributions. However, connectivity continues to be simulated using overly simple distance-based dispersal models with little consideration of how the individual behaviour of dispersing organisms interacts with landscape structure (functional connectivity). Here, we link an individual-based model to a niche-population model to test the implications of this omission. We apply this novel approach to a turtle species inhabiting wetlands which are patchily distributed across a tropical savannah, and whose persistence is threatened by two important synergistic drivers of global change: predation by invasive species and overexploitation. We show that projections of local range dynamics in this study system change substantially when functional connectivity is modelled explicitly. Accounting for functional connectivity in model simulations causes the estimate of extinction risk to increase, and predictions of range contraction to slow. We conclude that models of range dynamics that simulate functional connectivity can reduce an important source of bias in predictions of shifts in species distributions and abundances, especially for organisms whose dispersal behaviours are strongly affected by landscape structure.

Evaluating Basking-Habitat Deficiency in the Threatened Eastern Massasauga Rattlesnake

Abstract Woody plant succession is hypothesized to threaten many reptile populations by reducing the amount of solar energy available for thermoregulation. Mitigation via vegetation management is often recommended; however, the need for such management practices rarely has been evaluated. We examined the need for basking-site enhancement for the eastern massasauga rattlesnake (Sistrurus c. catenatus; hereafter EMR) in New York, USA, where only 2 populations remain: one at an open-canopy site and another at a closed-canopy site. Microhabitat temperatures were substantially lower at the closed-canopy site, where EMRs selected the warmest available basking sites. Eastern massasauga rattlesnakes in the open-canopy population selected basking sites that afforded greater cryptic cover. We recommend experimental reduction of shrub cover to improve EMR basking habitat at the closed-canopy site. More generally, we caution that management efforts to reduce shrub cover for basking EMRs should maintain adequate cryptic cover.

Effects of prey metapopulation structure on the viability of black-footed ferrets in plague-impacted landscapes: a metamodelling approach

Summary 1. Species interactions have been largely ignored in extinction risk assessment. However, the black-footed ferret Mustela nigripes exemplifies a class of endangered species for which strong species interactions cannot be ignored. This species is an obligate predator of prairie dogs Cynomys spp., and sylvatic plague Yersinia pestis epizootics threaten to undermine recovery efforts by functionally eliminating the prey base. Multispecies ‘metamodelling’ techniques offer new opportunities for exploring population dynamics under strong species interdependencies and disease. 2. To investigate ferret extinction risk in plague-affected landscapes, we simultaneously modelled plague epidemiological processes, prairie dog metapopulation dynamics and ferret demographic responses. Ferret population dynamics were investigated at an important release site (Conata Basin in South Dakota) and for 500 artificial prey landscapes spanning a wide range of realistic colony configurations (e.g. total area, # colonies, spatial clustering) and demographic characteristics. 3. Our simulation models indicate that ferrets are unlikely to persist through episodes of plague at Conata Basin unless they can access prey resources from a wider region or unless management actions can otherwise substantially reduce plague transmission. 4. We show that large, diffuse prairie dog metapopulations (those with colonies spread over a region >2500 km 2 ) are most likely to support ferret populations in plague-affected landscapes. Our results also highlight the potential importance of metapopulation connectivity in fuelling plague epizootics and thereby imperilling black-footed ferret conservation efforts. 5. We describe a cycle (c. 5- to 25-year period) of plague-driven population crashes that is an emergent property of our models, and which can destabilize ferret populations. 6. Synthesis and applications. On the basis of our models, we conclude that few North American prairie dog complexes cover sufficient land area to sustain black-footed ferret populations through plague-driven crashes in prey abundance. Consequently, our results underscore the importance of working with many constituents to conserve large prairie dog landscapes in addition to continued development of plague mitigation tools. In addition, the strong relationship between plague-induced oscillatory prey cycles and predator population persistence highlights the potential conservation benefits of imposing strategic barriers to connectivity in areas over which plague outbreak cycles are strongly synchronous.

Inferring the nature of anthropogenic threats from long‐term abundance records

Diagnosing the processes that threaten species persistence is critical for recovery planning and risk forecasting. Dominant threats are typically inferred by experts on the basis of a patchwork of informal methods. Transparent, quantitative diagnostic tools would contribute much-needed consistency, objectivity, and rigor to the process of diagnosing anthropogenic threats. Long-term census records, available for an increasingly large and diverse set of taxa, may exhibit characteristic signatures of specific threatening processes and thereby provide information for threat diagnosis. We developed a flexible Bayesian framework for diagnosing threats on the basis of long-term census records and diverse ancillary sources of information. We tested this framework with simulated data from artificial populations subjected to varying degrees of exploitation and habitat loss and several real-world abundance time series for which threatening processes are relatively well understood: bluefin tuna (Thunnus maccoyii) and Atlantic cod (Gadus morhua) (exploitation) and Red Grouse (Lagopus lagopus scotica) and Eurasian Skylark (Alauda arvensis) (habitat loss). Our method correctly identified the process driving population decline for over 90% of time series simulated under moderate to severe threat scenarios. Successful identification of threats approached 100% for severe exploitation and habitat loss scenarios. Our method identified threats less successfully when threatening processes were weak and when populations were simultaneously affected by multiple threats. Our method selected the presumed true threat model for all real-world case studies, although results were somewhat ambiguous in the case of the Eurasian Skylark. In the latter case, incorporation of an ancillary source of information (records of land-use change) increased the weight assigned to the presumed true model from 70% to 92%, illustrating the value of the proposed framework in bringing diverse sources of information into a common rigorous framework. Ultimately, our framework may greatly assist conservation organizations in documenting threatening processes and planning species recovery.

Genetic Connectivity among Populations of the Threatened Bog Turtle (Glyptemys muhlenbergii) and the Need for a Regional Approach to Turtle Conservation

The threatened Bog Turtle (Glyptemys muhlenbergii) is considered among the most sedentary of turtles, yet with population sizes generally below 50 individuals, gene flow among populations is clearly necessary to maintain healthy levels of genetic diversity. Therefore, designing effective reserve networks for this species will require clarification of the rates of among-population gene flow over several spatial scales. We obtained genetic samples from a complex of 11 Bog Turtle populations within the Berkshire-Taconic region of Massachusetts and New York, and all individuals (n  =  234) were genotyped across 15 microsatellite loci. Average multi-generation dispersal rates were inferred from population-level differences in allele frequencies using an approximate-likelihood approach, and recent dispersal rates were inferred using genetic assignment algorithms. Over small geographic distances (average inter-fen distance of ca. 1 km), among-population dispersal rates historically averaged between 0.25 and 0.5 effective migrants per population per year (ca. 1% of each population dispersing each year), and these dispersal rates appear to have persisted in recent decades. Over larger geographic distances (≥10 km), we infer that Bog Turtle populations in the Berkshire-Taconic region have experienced low rates of gene flow among populations according to a “stepping-stone” model. We conclude that (1) Bog Turtle populations with nearest-neighbor distances of <2 km should be managed as inter-connected demographic units, (2) dispersal movements among adjacent populations may enhance regional population stability, and (3) gene flow over larger spatial and temporal scales probably requires dispersal among “stepping stone” habitats that may not harbor viable populations. Regional conservation planning for these and other small-bodied, endangered turtles should focus on establishing and maintaining networks of loosely connected population complexes to mimic historical connectivity patterns.

Climatic and geographic predictors of life history variation in Eastern Massasauga (Sistrurus catenatus): A range-wide synthesis

Elucidating how life history traits vary geographically is important to understanding variation in population dynamics. Because many aspects of ectotherm life history are climate-dependent, geographic variation in climate is expected to have a large impact on population dynamics through effects on annual survival, body size, growth rate, age at first reproduction, size–fecundity relationship, and reproductive frequency. The Eastern Massasauga (Sistrurus catenatus) is a small, imperiled North American rattlesnake with a distribution centered on the Great Lakes region, where lake effects strongly influence local conditions. To address Eastern Massasauga life history data gaps, we compiled data from 47 study sites representing 38 counties across the range. We used multimodel inference and general linear models with geographic coordinates and annual climate normals as explanatory variables to clarify patterns of variation in life history traits. We found strong evidence for geographic variation in six of nine life history variables. Adult female snout-vent length and neonate mass increased with increasing mean annual precipitation. Litter size decreased with increasing mean temperature, and the size–fecundity relationship and growth prior to first hibernation both increased with increasing latitude. The proportion of gravid females also increased with increasing latitude, but this relationship may be the result of geographically varying detection bias. Our results provide insights into ectotherm life history variation and fill critical data gaps, which will inform Eastern Massasauga conservation efforts by improving biological realism for models of population viability and climate change.

Pan-Antarctic analysis aggregating spatial estimates of Adélie penguin abundance reveals robust dynamics despite stochastic noise

Colonially-breeding seabirds have long served as indicator species for the health of the oceans on which they depend. Abundance and breeding data are repeatedly collected at fixed study sites in the hopes that changes in abundance and productivity may be useful for adaptive management of marine resources, but their suitability for this purpose is often unknown. To address this, we fit a Bayesian population dynamics model that includes process and observation error to all known Adélie penguin abundance data (1982–2015) in the Antarctic, covering >95% of their population globally. We find that process error exceeds observation error in this system, and that continent-wide “year effects” strongly influence population growth rates. Our findings have important implications for the use of Adélie penguins in Southern Ocean feedback management, and suggest that aggregating abundance across space provides the fastest reliable signal of true population change for species whose dynamics are driven by stochastic processes.Adélie penguins are a key Antarctic indicator species, but data patchiness has challenged efforts to link population dynamics to key drivers. Che-Castaldo et al. resolve this issue using a pan-Antarctic Bayesian model to infer missing data, and show that spatial aggregation leads to more robust inference regarding dynamics.