Automated conservation assessment of the orchid family with deep learning

International Union for Conservation of Nature (IUCN) Red List assessments are essential for prioritizing conservation needs but are resource intensive and therefore available only for a fraction of global species richness. Automated conservation assessments based on digitally available geographic occurrence records can be a rapid alternative, but it is unclear how reliable these assessments are. We conducted automated conservation assessments for 13,910 species (47.3% of the known species in the family) of the diverse and globally distributed orchid family (Orchidaceae), for which most species (13,049) were previously unassessed by IUCN. We used a novel method based on a deep neural network (IUC-NN). We identified 4,342 orchid species (31.2% of the evaluated species) as possibly threatened with extinction (equivalent to IUCN categories critically endangered [CR], endangered [EN], or vulnerable [VU]) and Madagascar, East Africa, Southeast Asia, and several oceanic islands as priority areas for orchid conservation. Orchidaceae provided a model with which to test the sensitivity of automated assessment methods to problems with data availability, data quality, and geographic sampling bias. The IUC-NN identified possibly threatened species with an accuracy of 84.3%, with significantly lower geographic evaluation bias relative to the IUCN Red List and was robust even when data availability was low and there were geographic errors in the input data. Overall, our results demonstrate that automated assessments have an important role to play in identifying species at the greatest risk of extinction.     

Determinants of population persistence and abundance of terrestrial and arboreal vertebrates stranded in tropical forest land-bridge islands

Megadams are among the key modern drivers of habitat and biodiversity loss in emerging economies. The Balbina Hydroelectric Dam of Central Brazilian Amazonia inundated 312,900 ha of primary forests and created approximately 3500 variable-sized islands that still harbor vertebrate populations after nearly 3 decades after isolation. We estimated the species richness, abundance, biomass, composition, and group size of medium- to large-bodied forest vertebrates in response to patch, landscape, and habitat-quality metrics across 37 islands and 3 continuous forest sites throughout the Balbina archipelago. We conducted 1168 km of diurnal censuses and had 12,420 camera-trapping days along 81 transects with 207 camera stations. We determined the number of individuals (or groups) detected per 10 km walked and the number of independent photographs per 10 camera-trapping days, respectively, for each species. We recorded 34 species, and patch area was the most significant predictor of vertebrate population relative abundance and aggregate biomass. The maximum group size of several group-living species was consistently larger on large islands and in continuous patches than on small islands. Most vertebrate populations were extirpated after inundation. Remaining populations are unlikely to survive further ecological disruptions. If all vertebrate species were once widely distributed before inundation, we estimated that approximately 75% of all individual vertebrates were lost from all 3546 islands and 7.4% of the animals in all persisting insular populations are highly likely to be extirpated. Our results demonstrate that population abundance estimates should be factored into predictions of community disassembly on small islands to robustly predict biodiversity outcomes. Given the rapidly escalating hydropower infrastructure projects in developing counties, we suggest that faunal abundance and biomass estimates be considered in environmental impact assessments and large strictly protected reserves be established to minimize detrimental effects of dams on biodiversity. Conserving large tracts of continuous forests represents the most critical conservation measure to ensure that animal populations can persist at natural densities in Amazonian forests.     

Global protected-area coverage and human pressure on tidal flats

Tidal flats are a globally distributed coastal ecosystem important for supporting biodiversity and ecosystem services. Local to continental-scale studies have documented rapid loss of tidal habitat driven by human impacts, but assessments of progress in their conservation are lacking. With an internally consistent estimate of distribution and change, based on Landsat satellite imagery, now available for the world's tidal flats, we examined tidal flat representation in protected areas (PAs) and human pressure on tidal flats. We determined tidal flat representation and its net change in PAs by spatially overlaying tidal flat maps with the World Database of Protected Areas. Similarly, we overlaid the most recent distribution map of tidal flats (2014–2016) with the human modification map (HM_c) (range from 0, no human pressure, to 1, very high human pressure) to estimate the human pressure exerted on this ecosystem. Sixty-eight percent of the current extent of tidal flats is subject to moderate to very high human pressure (HM_c > 0.1), but 31% of tidal flat extent occurred in PAs, far exceeding PA coverage of the marine (6%) and terrestrial (13%) realms. Net change of tidal flat extent inside PAs was similar to tidal flat net change outside PAs from 1999 to 2016. Substantial shortfalls in protection of tidal flats occurred across Asia, where large intertidal extents coincided with high to very high human pressure (HM_c > 0.4–1.0) and net tidal flat losses up to 86.4 km² (95% CI 83.9–89.0) occurred inside individual PAs in the study period. Taken together, our results show substantial progress in PA designation for tidal flats globally, but that PA status alone does not prevent all habitat loss. Safeguarding the world's tidal flats will thus require deeper understanding of the factors that govern their dynamics and effective policy that promotes holistic coastal and catchment management strategies.     

The bean method as a tool to measure sensitive behavior

Conservationists need to measure human behavior to guide decisions and evaluate their impact. However, activities can be misreported and reporting accuracy may change following conservation interventions, making it hard to verify any apparent changes. Techniques for asking sensitive questions are increasingly integrated into survey designs to improve data quality, but some can be costly or hard for nonexperts to implement. We demonstrate a straightforward, low-cost approach, the bean method in which respondents give anonymous answers by adding a colored bean to a jar to denote a yes or no response. We applied the bean method to measure wild-meat hunting and trading over 2 years at a conservation-project (hunting reduction) site in Gola Forest, Liberia. We extended the technique to accommodate questions about hunting and meat-selling frequency. We compared responses given using the bean method and direct questioning for groups that did and did not participate in conservation interventions. Results from the bean method corresponded to those from direct questioning, and there was no indication of change in question sensitivity following conservation interventions. Estimates from both methods indicated that wild-meat trading decreased in project and nonproject households (from 36% to 20%) and that hunting decreased in 1 project group (38–28%). Where inconsistent answers were given (2–6% of respondents), differences were in both directions and were most likely attributable to measurement error. The bean method was quick and straightforward to administer in a low-literacy setting. We showed how it can be modified for answers of more than 2 categories and consider it a valuable tool that could be adapted for a wide range of conservation settings.     

Bridging gaps in demographic analysis with phylogenetic imputation

Phylogenetically informed imputation methods have rarely been applied to estimate missing values in demographic data but may be a powerful tool for reconstructing vital rates of survival, maturation, and fecundity for species of conservation concern. Imputed vital rates could be used to parameterize demographic models to explore how populations respond when vital rates are perturbed. We used standardized vital rate estimates for 50 bird species to assess the use of phylogenetic imputation to fill gaps in demographic data. We calculated imputation accuracy for vital rates of focal species excluded from the data set either singly or in combination and with and without phylogeny, body mass, and life-history trait data. We used imputed vital rates to calculate demographic metrics, including generation time, to validate the use of imputation in demographic analyses. Covariance among vital rates and other trait data provided a strong basis to guide imputation of missing vital rates in birds, even in the absence of phylogenetic information. Mean NRMSE for null and phylogenetic models differed by <0.01 except when no vital rates were available or for vital rates with high phylogenetic signal (Pagel's λ > 0.8). In these cases, including body mass and life-history trait data compensated for lack of phylogenetic information: mean normalized root mean square error (NRMSE) for null and phylogenetic models differed by <0.01 for adult survival and <0.04 for maturation rate. Estimates of demographic metrics were sensitive to the accuracy of imputed vital rates. For example, mean error in generation time doubled in response to inaccurate estimates of maturation time. Accurate demographic data and metrics, such as generation time, are needed to inform conservation planning processes, for example through International Union for Conservation of Nature Red List assessments and population viability analysis. Imputed vital rates could be useful in this context but, as for any estimated model parameters, awareness of the sensitivities of demographic model outputs to the imputed vital rates is essential.     

Addressing behavior in pollinator conservation policies to combat the implementation gap

Solutions for conserving biodiversity lie in changing people's behavior. Ambitious international and national conservation policies frequently fail to effectively mitigate biodiversity loss because they rarely apply behavior-change theories. We conducted a gap analysis of conservation behavior-change interventions advocated in national conservation strategies with the Behavior Change Wheel (BCW), a comprehensive framework for systematically characterizing and designing behavior-change interventions. Using pollinator conservation as a case study, we classified the conservation actions listed in national pollinator initiatives in relation to intervention functions and policy categories of the BCW. We included all national-level policy documents from the European Union available in March 2019 that focused on conservation of pollinators (n = 8). A total of 610 pollinator conservation actions were coded using in-depth directed content analysis, resulting in the identification of 787 intervention function and 766 policy category codes. Overall, these initiatives did not employ the entire breadth of behavioral interventions. Intervention functions most frequently identified were education (23%) and environmental restructuring (19%). Least frequently identified intervention functions were incentivization (3%), and restriction (2%) and coercion were completely absent (0%). Importantly, 41% of all pollinator conservation actions failed to identify whose behavior was to be changed. Building on these analyses, we suggest that reasons for the serious implementation gap in national and international conservation policies is founded in insufficient understanding of which behavioral interventions to employ for most beneficial impacts on biodiversity and how to clearly specify the intervention targets. We recommend that policy advisors engage with behavior-change theory to design effective behavior-change interventions that underpin successful conservation policies.     

Responses of New Zealand forest birds to management of introduced mammals

Over the past 1000 years New Zealand has lost 40–50% of its bird species, and over half of these extinctions are attributable to predation by introduced mammals. Populations of many extant forest bird species continue to be depredated by mammals, especially rats, possums, and mustelids. The management history of New Zealand's forests over the past 50 years presents a unique opportunity because a varied program of mammalian predator control has created a replicated management experiment. We conducted a meta-analysis of population-level responses of forest birds to different levels of mammal control recorded across New Zealand. We collected data from 32 uniquely treated sites and 20 extant bird species representing a total of 247 population responses to 3 intensities of invasive mammal control (zero, low, and high). The treatments varied from eradication of invasive mammals via ground-based techniques to periodic suppression of mammals via aerially sown toxin. We modeled population-level responses of birds according to key life history attributes to determine the biological processes that influence species’ responses to management. Large endemic species, such as the Kaka (Nestor meridionalis) and New Zealand Pigeon (Hemiphaga novaeseelandiae), responded positively at the population level to mammal control in 61 of 77 cases for species ≥20 g compared with 31 positive responses from 78 cases for species <20 g. The Fantail (Rhipidura fuliginosa) and Grey Warbler (Gerygone igata), both shallow endemic species, and 4 nonendemic species (Blackbird [Turdus merula], Chaffinch [Fringilla coelebs], Dunnock [Prunella modularis], and Silvereye [Zosterops lateralis]) that arrived in New Zealand in the last 200 years tended to have slight negative or neutral responses to mammal control (59 of 77 cases). Our results suggest that large, deeply endemic forest birds, especially cavity nesters, are most at risk of further decline in the absence of mammal control and, conversely suggest that 6 species apparently tolerate the presence of invasive mammals and may be sensitive to competition from larger endemic birds.     

Ethics and governance for internet-based conservation science research

Internet-based research is increasingly important for conservation science and has wide-ranging applications and contexts, including culturomics, illegal wildlife trade, and citizen science. However, online research methods pose a range of ethical and legal challenges. Online data may be protected by copyright, database rights, or contract law. Privacy rights may also restrict the use and access of data, as well as ethical requirements from institutions. Online data have real-world meaning, and the ethical treatment of individuals and communities must not be marginalized when conducting internet-based research. As ethics frameworks originally developed for biomedical applications are inadequate for these methods, we propose that research activities involving the analysis of preexisting online data be treated analogous to offline social science methods, in particular, nondeceptive covert observation. By treating internet users and their data with respect and due consideration, conservationists can uphold the public trust needed to effectively address real-world issues.     

Conducting conservation social science surveys online

The coronavirus (COVID-19) pandemic is affecting the environment and conservation research in fundamental ways. Many conservation social scientists are now administering survey questionnaires online, but they must do so while ensuring rigor in data collection. Further, they must address a suite of unique challenges, such as the increasing use of mobile devices by participants and avoiding bots or other survey fraud. We reviewed recent literature on online survey methods to examine the state of the field related to online data collection and dissemination. We illustrate the review with examples of key methodological decisions made during a recent national study of people who feed wild birds, in which survey respondents were recruited through an online panel and a sample generated via a project participant list. Conducting surveys online affords new opportunities for participant recruitment, design, and pilot testing. For instance, online survey panels can provide quick access to large and diverse samples of people. Based on the literature review and our own experiences, we suggest that to ensure high-quality online surveys one should account for potential sampling and nonresponse error, design survey instruments for use on multiple devices, test the instrument, and use multiple protocols to identify data quality problems. We also suggest that research funders, journal editors, and policy makers can all play a role in ensuring high-quality survey data are used to inform effective conservation programs and policies.     

Improving species status assessments under the U.S. Endangered Species Act and implications for multispecies conservation challenges worldwide

Despite its successes, the U.S. Endangered Species Act (ESA) has proven challenging to implement due to funding limitations, workload backlog, and other problems. As threats to species survival intensify and as more species come under threat, the need for the ESA and similar conservation laws and policies in other countries to function efficiently has grown. Attempts by the U.S. Fish and Wildlife Service (USFWS) to streamline ESA decisions include multispecies recovery plans and habitat conservation plans. We address species status assessment (SSA), a USFWS process to inform ESA decisions from listing to recovery, within the context of multispecies and ecosystem planning. Although existing SSAs have a single-species focus, ecosystem-based research can efficiently inform multiple SSAs within a region and provide a foundation for transition to multispecies SSAs in the future. We considered at-risk grassland species and ecosystems within the southeastern United States, where a disproportionate number of rare and endemic species are associated with grasslands. To initiate our ecosystem-based approach, we used a combined literature-based and structured World Café workshop format to identify science needs for SSAs. Discussions concentrated on 5 categories of threats to grassland species and ecosystems, consistent with recommendations to make shared threats a focus of planning under the ESA: (1) habitat loss, fragmentation, and disruption of functional connectivity; (2) climate change; (3) altered disturbance regimes; (4) invasive species; and (5) localized impacts. For each threat, workshop participants identified science and information needs, including database availability, research priorities, and modeling and mapping needs. Grouping species by habitat and shared threats can make the SSA process and other planning processes for conservation of at-risk species worldwide more efficient and useful. We found a combination of literature review and structured discussion effective for identifying the scientific information and analysis needed to support the development of multiple SSAs. Article impact statement: Species status assessments can be improved by an ecosystem-based approach that groups imperiled species by shared habitats and threats.