Search USGS⌕ Search

Geology topics

David R. Smith

Publications and source records attributed to David R. Smith.

At least 19 recordsLinked to original sources
USGS2026

Decision analysis of long-term research and monitoring investments for the Upper Mississippi River Restoration program

Congressional authorization of the Water Resource Development Act of 2020 increased funding for the Upper Mississippi River Restoration (UMRR) program and its Long Term Resource Monitoring (LTRM) element. With adequate resources there is an opportunity for LTRM to expand its understanding of the Upper Mississippi River System, address unmet partner information needs, and better inform restoration and management of the Upper Mississippi River System. Additional information needs will need to be identified and prioritized in a manner that accounts for potential tradeoffs among the preferences of UMRR partners for long-term commitments of effort. In this report, we describe a decision analysis process used by UMRR partners to develop a comprehensive framework for prioritizing and selecting the information needs they would address in the event of changes in available resources. The partners of the UMRR formed a “planning team” with members representing the UMRR partner agencies to work through the decision analysis process. This team included two facilitators who led the decision analysis process. The planning team began this process by articulating a consensus understanding of the specific decisions to be made and identifying a set of decision outcomes (objectives) the UMRR partners wanted to achieve. Next, they identified information needs not being addressed by ongoing monitoring and research, developed criteria for assessing the expected benefit of addressing each information need, and estimated the costs of addressing each information need. They then used an optimization approach to identify the portfolio of information needs that would produce the most benefit for a given cost if they were effectively addressed. Originally, the planning team expected this analysis to inform a project plan that would start in 2024, but in recognition of financial uncertainties, we refer to planning dates as the number of years after resources become available (for example, “year 1,” “year 2,” and so on).

Illinois, Iowa, Minnesota, Missouri, Wisconsin↗
USGS2025

Recovery of Delaware Bay horseshoe crabs following harvest reductions

Objective Horseshoe crabs Limulus polyphemus play a vital role in the Delaware Bay ecosystem. The migratory stopover of several shorebird species occurs during the horseshoe crab spawning season, and the eggs of horseshoe crabs provide an essential food source to fuel their northward migration to breeding areas. High commercial fishery use of horseshoe crabs as bait during the 1990s coincided with a decline in crabs and shorebirds, particularly the red knot Calidris canutus rufa , which has been listed as threatened under the U.S. Endangered Species Act since 2015. In response to the population decline of shorebirds, the Atlantic States Marine Fisheries Commission began reducing the harvest of horseshoe crabs in 2000 with a goal of rebuilding the population of horseshoe crabs and shorebirds that depend upon them. The objective of this analysis was to determine whether horseshoe crab harvest management in the Delaware Bay region has increased the abundance of the species in recent years. Methods We analyzed data from fisheries-independent trawl surveys of horseshoe crab relative abundance using a Bayesian hierarchical model to determine whether harvest management has resulted in the rebuilding of the horseshoe crab population to levels seen in 1990—a period before the overuse of horseshoe crabs and the decline in the population of red knots. Results Data from multiple surveys showed that the horseshoe crab population in Delaware Bay declined from the 1990s through approximately 2005, was relatively low and stable until 2010, and then increased through 2023, with a 0.38 probability of exceeding the 1990 level. Conclusions The results of this analysis support the effectiveness of management decisions related to horseshoe crabs in the Delaware Bay region. In response to harvest restrictions, the abundance of horseshoe crabs has neared levels observed in the early 1990s—a period prior to high commercial use and a decline in both horseshoe crabs and shorebirds that depend on them for food during annual migrations.

Delaware, Maryland, New Jersey↗
USGS2025

Recovering the American horseshoe crab through a commitment to collaboration

American horseshoe crab Limulus polyphemus populations are recovering because of cooperation among diverse stakeholders and data-driven collective action. Limulus polyphemus is one of four extant species facing common threats, and conservation successes and limitations hold lessons applicable to all the species. We review the advancement in management and monitoring over recent decades, discuss the current population status throughout the species’ range, and describe the potential future status and recovery based on recent assessments. In retrospect, L. polyphemus conservation has followed the frameworks promoted by the International Union for the Conservation of Nature: mobilize networks to increase assessment capacity, engage diverse stakeholders, measure impacts, and amplify successes. Data show significantly increased populations in the Delaware Bay region and improved status in the Northeast. The average abundance of adults in Delaware Bay over the recent decade (2013–2022) is more than twice that in the previous decade (2003–2012). In 2022, the abundances for adult females and males in the Delaware Bay population were estimated to be 16 million and 40 million, respectively. However, reversing persistent poor conditions in some regions and mitigating the widespread threat of habitat loss from coastal development and sea level rise will rely on collaboration among diverse stakeholders to build upon the current conservation successes. Scientists and conservationists working on the horseshoe crab species indigenous to Asia are advancing along a similar track, establishing a monitoring network and mobilizing diverse stakeholders. However, there is a need for capacity building for robust assessment of the species in Asia to measure the impact of conservation, just as that need remains unmet for L. polyphemus in the Gulf of Mexico. The importance of partnerships and collaborations is apparent in their absence. The regions where partnerships and collaborations among researchers, conservationists, and agency scientists do not exist are those where the capacity for monitoring and assessment is notably lacking.

Fisheries↗
USGS2025

Evaluating freshwater mussel sampling methodologies using a simulation model

Field surveys form the basis of many research efforts and are the foundation for estimates of population size and density that inform conservation and management practices for imperiled species. As a result, evaluating the performance of different survey methods across a range of conditions that may be encountered in the field can increase understanding of the time and effort that may be required to ensure that survey results are sufficiently accurate and reliable for conservation goals. We used a spatially explicit agent-based model to simulate four commonly used freshwater mussel field survey methodologies: simple random sampling (SRS), transect random sampling (TRS), adaptive cluster sampling (ACS), and qualitative timed searches (QTS) to investigate the influence of sampling method, spatial distribution, and mussel density on the performance ( i.e., accuracy, precision, and detection rate) of survey techniques. Our analysis suggests that mussel density, spatial distribution, and sampling effort influence sampling accuracy, precision, and species detection for all sampling methods. QTS produces highly variable catch-per-unit-effort (CPUE) metrics when mussels are dense and/or clustered, indicating the technique may be unreliable as a proxy for density. Quantitative methods like SRS and TRS may be well-suited for estimating population characteristics, but a high level of effort may be needed to obtain reasonable accuracy when mussels occur at low densities. ACS may be more efficient for mussels at low densities, but it can be challenging to plan for the level of effort required to complete an ACS protocol. Designing an ecological survey requires careful consideration of research objectives and available resources. Future research may consider the performance of qualitative and quantitative surveys in combination as a means of overcoming some of the practical challenges of applying individual survey methods.

Ecological Indicators↗
USGS2025

Comparing SMART and ELECTRE methods for multi-criteria decision analysis: A case study evaluating conservation strategies

Tradeoffs are part of the inherent challenge of making decisions. Defaulting to a prevalent method can mask methodological variation and potential improvement in decision quality. We applied and compared methods for multi-criteria decision analysis (MCDA) in real-world environmental planning. Decision makers at a global conservation organization formed teams to review and prioritize the allocation of resources across a set of existing strategies across seven teams. Each team evaluated and rated strategies based on a common scale composed of weighted objectives and criteria. Sensitivity analysis included implementing two different MCDA methods, varying strategy ratings, and using two different criteria weighting techniques. The MCDA methods were SMART (Simple Multi-Attribute Rating Technique) and ELECTRE (ELimination and Choice Expressing The REality). We tested user satisfaction between methods, overall rankings between methods, and whether MCDA improved the decision-making process. The methods did not differ in most comparisons, although participants were more likely to use the SMART method in future tradeoff decisions. We found evidence of rank correlation between methods. Lastly, the teams’ final recommendations were consistent with MCDA results. This study highlights the advantages and disadvantages of MCDA in real-world applications.

Environmental and Sustainability Indicators↗
USGS2023

Application of a catch multiple survey analysis for Atlantic horseshoe crab Limulus polyphemus in the Delaware Bay

Objective This paper applies a catch multiple survey analysis (CMSA) to Atlantic horseshoe crabs Limulus polyphemus in the Delaware Bay to generate robust population estimates for harvest management. Currently, horseshoe crabs along the U.S. Atlantic coast are harvested as bait for other fisheries and collected for their blood, which is used in a biomedical industry. The Delaware Bay is home to the largest population of horseshoe crabs and is a significant stopover for shorebirds to rebuild energy by consuming horseshoe crab eggs prior to completing their northward migration. To address this interrelationship, the Adaptive Resource Management (ARM) Framework has been used since 2013 to ensure that horseshoe crab harvest within the region takes into account the forage needs of migratory birds. Since its inception, the ARM Framework has used a single trawl survey's swept area-based population estimates of horseshoe crab relative abundance and a theoretical population model developed primarily from literature-derived values. With more data collected in the region in recent years and other sources of mortality that can now be quantified, a catch survey model can provide horseshoe crab population estimates going forward. Methods A CMSA was used to estimate male and female horseshoe crab population size for 2003–2021 using all quantifiable sources of mortality and three fishery-independent indices of abundance. Result The CMSA results indicated that adult abundance of male and female horseshoe crabs was stable from 2003 to 2013 and then began to increase through 2017, a result that is consistent with stock rebuilding following a period of harvest restrictions as recommended by the ARM Framework. Population estimates were lower in recent years but remained above the levels estimated before implementation of the ARM Framework. In 2021, the CMSA estimated that there were over 6 million mature females and nearly 16 million mature male horseshoe crabs in the region. Conclusion The CMSA provides the best and most comprehensive population estimates of horseshoe crabs in Delaware Bay and will improve modeling efforts within the ARM Framework going forward.

Delaware, Maryland, New Jersey↗
USGS2023

Assessment of recovery potential for the American horseshoe crab (Limulus polyphemus): An application of the IUCN green status process

According to an International Union for Conservation of Nature (IUCN) Red List assessment (RLA), the American horseshoe crab ( Limulus polyphemus ), an iconic coastal species, is at risk of extirpation in some regions within its range where small and vulnerable populations occur. However, the RLA does not consider future status beyond viability and does not attempt to identify the conservation necessary to effectively mitigate threats and recover the species to full ecological functionality. To aid in conservation planning for vulnerable species, the IUCN developed the Green Status of Species assessment (GSA) process to complement the RLA. This paper describes the application of the GSA process to assess the recovery potential of the American horseshoe crab. First, specific Limulus populations within spatial units for conservation were delineated, and their statuses were defined based on viability and ecological functionality. Then conservation actions were identified that would promote recovery and affect their near- and long-term population status under different conservation scenarios. Horseshoe crab conservation has relied on, and will continue to depend on, effective harvest regulation. However, as currently conceived, conservation is not expected to mitigate habitat loss at the scale required to restore range-wide ecological functionality, primarily because habitat loss is widespread and affected by climate change. Thus, the GSA results, while indicating that there is potential for near-term recovery gains, reveal that long-term recovery is in doubt owing to expected loss of habitat. To conserve critical habitats for spawning and early life stages and achieve ecological functionality, it is imperative to identify and develop conservation plans at appropriate spatial scales. Unfortunately, such plans do not currently exist and need to be established. The GSA Green Score can then serve as a metric for monitoring recovery and gauging the effectiveness of conservation implementation.

Aquatic Conservation: Marine and Freshwater Ecosys↗
USGS2023

Creating conservation strategies with value-focused thinking

Biodiversity and human well-being strategies are only as good as the set of ideas that people think about. This article evaluates value-focused thinking (VFT), a framework that focuses on creating objectives and strategy ideas that are responsive to the objectives. We performed a proof-of-concept study of VFT on six planning teams at a global conservation organization. We developed a package of support materials, including session agendas, a virtual facilitation template, facilitator's guide, and evaluation questionnaires. The study tested whether VFT resulted in a set of quality strategies, resulted in participant satisfaction, and was scalable, meaning that it could be facilitated by someone newly trained in VFT and result in quality strategies and participant satisfaction, as compared to an experienced facilitator. Net response indicated positive quality ratings for the set of strategies per team. Respondents indicated positive satisfaction overall, though it was higher for objectives than for strategies. Among the participants with previous experience, all were at least as satisfied with their VFT strategies compared to previously developed strategies, and none were less satisfied (P = 0.001). Changes in participant satisfaction were not related to facilitator type (P > 0.10). In addition, we found that some participants had a premature sense of shared understanding of important values and interests before entering the study, which VFT strengthened. This study highlights the advantages of structuring the development and evaluation of conservation planning frameworks.

Conservation Biology↗
USGS2023

Estimating recruitment rate and population dynamics at a migratory stopover site using an integrated population model

Consideration of the full annual cycle population dynamics can provide useful insight for conservation efforts, but collecting data needed to estimate demographic parameters is often logistically difficult. For species that breed in remote areas, monitoring is often conducted during migratory stopover or at nonbreeding sites, and the recruitment rate of new breeding adults can be difficult to estimate directly. Here, we present an integrated population model that uses mark-resight and count data to estimate survival probability, population growth rate, and recruitment rate for an Arctic-breeding shorebird of conservation concern, the red knot ( Calidris canutus rufa ), from data collected during spring stopover in Delaware Bay, USA, from 2005 to 2018. At this site, red knots feed primarily on the eggs of spawning horseshoe crabs ( Limulus polyphemus ), a legally harvested species. We used this model to estimate the relationship between horseshoe crab abundance and red knot demographics, which informed a recent revision to the framework used to establish horseshoe crab harvest regulations. Our analysis indicates that the red knot population was most likely stable from 2005 to 2018 (average λ = 1.03, 95% credible interval [CRI]: 0.961, 1.15) despite low recruitment rates (average ρ = 0.088, 95% CRI: 0.012, 0.18). Adult survival probability was positively associated with horseshoe crab abundance in the same year (β = 0.35, 95% CRI: 0.09, 0.63), but we found no effect of horseshoe crab abundance two years previously on recruitment of new adults (β = −0.08, 95% CRI: −0.41, 0.38). Our approach demonstrates the utility of integrated population models for understanding population dynamics, even when data are only available from migratory stopover monitoring.

Delaware, New Jersey↗
USGS2023

When are environmental DNA early detections of invasive species actionable?

Environmental DNA (eDNA) sampling provides sensitive early detection capabilities for recently introduced taxa. However, natural resource managers struggle with how to integrate eDNA results into an early detection rapid response program because positive eDNA detections are not always indicative of an eventual infestation. We used a structured decision making (SDM) framework to evaluate appropriate response actions to hypothetical eDNA early detections of an introduced aquatic plant in Sebago Lake (Maine, USA). The results were juxtaposed to a recent study that used a similar SDM approach to evaluate response actions to hypothetical eDNA early detections of introduced mussels in Jordanelle Reservoir (Utah, USA). We found that eDNA early detections were not actionable in Sebago Lake because the plant's invasion potential was spatially constrained and the current management activities provided acceptable levels of mitigation. In Jordanelle Reservoir, eDNA detections were actionable due to high invasion potential and analyses supported management actions to contain the invasion. The divergent outcomes of the two case studies are related to the unique attributes of the habitats and species, highlighting the utility of the SDM approach when considering an eDNA monitoring program. We use these two case studies to present a general SDM framework and a set of heuristics that can be efficiently applied to eDNA early detection rapid response scenarios and other instances associated with indeterminant eDNA detections, especially when there is an imperative to make decisions as quickly as possible.

Maine↗
USGS2023

A science and management partnership to restore coregonine diversity to the Laurentian Great Lakes

Similar to many freshwater ecosystems, the Laurentian Great Lakes of North America have undergone numerous anthropogenic stressors resulting in considerable loss of biodiversity and habitat. Among Great Lakes fishes, the coregonine sub-family has endured the most extensive declines, including extinction of several species ( Coregonus johannae, C. alpenae , and C. kiyi orientalis ) and at least 10 instances of local extirpations of other species ( C. nigripinnis, C. reighardi, C. zenithicus, C. hoyi , and C. artedi ) across all 5 lakes, much of which occurred prior to the 1960s owing to overfishing, interactions with non-indigenous species, and habitat loss. Despite these declines, no federal-, provincial-, or state-mandated actions were ever implemented to conserve coregonine diversity, potentially because so much of the coregonine declines occurred prior to the enactment of federal conservation legislation. Possible explanations for inaction since enactment of that legislation include insufficient data on biological vulnerability or threats, unresolved taxonomy, and limited support from the fishery management agencies and their stakeholders prior to the 2000s. In recent decades, however, several fishery management agencies have undertaken efforts to re-introduce coregonine diversity. These efforts helped lead to development of a science-based framework to restore coregonines that was universally endorsed by fishery managers representing eight U.S. states, four U.S. tribal organizations, and the province of ON, Canada, in May 2018. The basin-wide framework is based on principles of conservation biology and adaptive management. We describe details of its key steps, including planning, restoring, and evaluating, while also describing recent implementation efforts to develop methods, improve available resources, and enhance coordination across the basin. Although our paper describes a regional effort to restore native coregonines, our adaptive-management approach could be used by other multi-agency stakeholders seeking to conserve or restore native fishes.

Laurentian Great Lakes↗
USGS2023

A proposed methodology for conducting threats assessments within the Great Lakes Coregonines restoration framework

This document serves to fulfill the Coregonine Threats Assessment Science Team’s charge of providing a written recommendation for a methodology to conduct threats assessments for Great Lakes coregonines within the Coregonine Restoration Framework (CRF). Through a series of team meetings that included presentations by experts on five candidate threats assessment frameworks followed by structured deliberations, we came to consensus to recommend the threats assessment framework used by Fisheries and Oceans Canada under Canada’s Species at Risk Act, with three modifications: (1) a conceptual modeling step, (2) the use of a “point spreading” approach to incorporate uncertainty when scoring threats, and (3) the use of a modified Delphi or “estimate-talk-estimate” approach when scoring key elements in the assessment. We recommend that this approach be applied to the spatial units delineated by the CRF Resolve Taxonomy and Gap Analysis science teams. In brief, the assessment process includes providing background information on the spatial unit and threats under assessment, constructing a conceptual model linking threats to key processes and vital rates, and scoring or ranking threats across six elements: likelihood of occurrence, level of impact, strength of evidence, unit-level threat occurrence, unit-level threat frequency, and unit-level threat extent. We provide detailed instructions for completing each step of the assessment and generating associated results, with particular attention paid to our suggested modifications. The Coregonine Threats Assessment Science Team also conducted two test runs to assess the applicability and effectiveness of our recommended framework for Great Lakes coregonine populations and their threats. We conducted these test runs on two examples of Great Lakes coregonines that represented two extremes of data availability, as well as two different management contexts. We chose Kiyi (Coregonus kiyi) in Lake Ontario as an example of a data-poor, extirpated population, and we chose Cisco (Coregonus artedi) in Lake Superior as an example of a data-rich, extant population. We provide the results of these test runs in Appendices 1-2. We also describe the lessons we learned from these test runs throughout this document and highlighted them in the “Recommendations for avoiding challenges during application” section.

Report↗
USGS2022

Multispecies approaches to status assessments in support of endangered species classifications

Multispecies risk assessments have developed within many international conservation programs, reflecting a widespread need for efficiency. Under the United States Endangered Species Act (ESA), multispecies assessments ultimately lead to species-level listing decisions. Although this approach provides opportunities for improved efficiency, it also risks overwhelming or biasing the assessment process and would benefit from clear guidance for practitioners. We reviewed multispecies assessments conducted between 1993 and 2019 for ESA listing decisions to identify the ecological basis for combining species, the assessment approach used, and the policy factors influencing their efficacy. We identified 42 cases covering 359 species. Most assessments (81%) included two to five species, although the maximum was 82. A common theme involved grouping narrow endemics or habitat specialists based on taxonomic relatedness, similar distributions, and common threats to persistence. All assessments included a combined threats analysis, but few employed a common species' response model or expert elicitation process. Although ESA risk assessments are distinct from policy decisions, most assessments (50%) supported decisions that all species warranted endangered status. Available guidance has generally emphasized ecological similarity as the key attribute leading to successful multispecies assessments. The challenge with consistently selecting species based on qualitative proxies such as common distributions or threats to persistence is that ecological patterns and processes are scale dependent. Focusing instead on the assessment methods and their potential for bias and increased efficiency may provide a stronger basis for developing consistent and transparent guidance.

Conservation Science and Practice↗
USGS2022

Using structured decision making to evaluate potential management responses to detection of dreissenid mussel (Dreissena spp.) environmental DNA

Environmental (e)DNA tools are sensitive and cost-effective for early detection of invasive species. However, the uncertainty associated with the interpretation of positive eDNA detections makes it challenging to determine appropriate natural resource management responses. Multiple sources of error can give rise to positive detections of eDNA in a sample when individuals of that species are not present at the site or a widespread infestation is not imminent. Acting on an erroneous eDNA inference could result in needless costs or reductions in desirable resources. Alternatively, failure to rapidly act on eDNA results that truly indicate invader presence could compound negative impacts and lead to high, long-term costs to manage infestations. We used a structured decision making (SDM) process, which incorporates tradeoffs and uncertainties, to evaluate appropriate response actions following hypothetical eDNA detections of invasive dreissenid mussel ( Dreissena spp.) eDNA in Jordanelle Reservoir, Utah (USA). We worked with decision-makers and stakeholders to identify objectives and discrete management action alternatives to assess consequences and tradeoffs. Alternatives ranged from no action to intensive and expensive control efforts. The best performing alternative was delayed containment described by immediate attempts to confirm the eDNA detections using non-molecular sampling techniques followed by mandatory watercraft exit inspections to prevent dreissenid mussel spread to regional water bodies. Non-molecular sampling increased public support for management by demonstrating a commitment to monitor the invasion state before action, whereas containment decreased likelihood of regional spread to other waters. Delayed containment had the lowest downside risk, and the highest upside gains relative to other alternative actions. Sensitivity analyses showed our results to be robust to parameter and outcome uncertainty.

Management of Biological Invasions↗
USGS2021

Telemetry reveals migratory drivers and disparate space use across seasons and age-groups in American horseshoe crabs

Identifying mechanisms that underpin animal migration patterns and examining variability in space use within populations is crucial for understanding population dynamics and management implications. In this study, we quantified the migration rates, seasonal changes in migratory connectivity, and residency across population demographics (age and sex) to understand the proximate cues of migration timing in American horseshoe crabs ( Limulus polyphemus ). Juvenile ( n = 25) and adult ( n = 70) horseshoe crabs were tracked with acoustic telemetry techniques for a 3-yr period in Moriches Bay, NY. Connectivity metrics and residency probability were quantified through spatial network analysis and empirically derived Markov Chain models (EDMC), respectively. The migratory probability of adult horseshoe crabs between Moriches Bay and the Atlantic Ocean was estimated to be 41.0% (95% CI: 34.0–59.8); in contrast, only 8% (95% CI: 1.2–31.6) of juveniles migrated into the ocean. Migration timing was influenced by the interaction of photoperiod and temperature, revealing seasonal differences in migration timing and a 50% narrower range of photoperiod and temperature over which fall migrations occurred compared to spring. Sex-specific differences in space use and connectivity within each season were largely absent; however, centralized habitats were important for maintaining connectivity across all seasons. EDMC results revealed that when standardized to the number of horseshoe crab detections on each receiver, the centrally located habitats in Moriches Bay and Inlet accounted for >50% of the total relative residency probability within most seasons, indicating these areas may be preferred by adult horseshoe crabs. Ontogenetic differences in maximum spatial extent, space use, and connectivity were observed in the bay, as juveniles exhibited lower linkages between locations ( n = 4) relative to adults ( n = 13) during the same temporal period. Our work highlights the application of novel quantitative approaches for addressing the movement dynamics of horseshoe crabs that can be readily applied to other taxa in the context of wildlife conservation.

New York↗
USGS2021

Adaptive two-stage inverse sampling design to estimate density, abundance, and occupancy of rare and clustered populations

Sampling rare and clustered populations is challenging because of the effort required to find rare units. Heuristically, a practitioner would prefer to discontinue sampling in areas where rare units of interest are apparently extremely sparse or absent. We take advantage of the characteristics of inverse sampling to adaptively inform practitioners when it is efficient to move on to sample new areas. We introduce Adaptive Two-stage Inverse Sampling (ATIS), which is designed to leave a selected area after observation of an a priori number of only non-rare units and to continue sampling in the area when rare units are observed. ATIS is efficient in many cases and yields more rare units than conventional sampling for a rare and clustered population. We derive unbiased estimators of population total and variance. We also introduce an easy-to-compute estimator, which is nearly as efficient as the unbiased estimator. A simulation study on a rare plant population of buttercups ( Ranunculus ) shows that ATIS even with the easy-to-compute estimator is more efficient than its conventional sampling counterparts and is more efficient than Two-stage Adaptive Cluster Sampling (TACS) for small and moderate final sample sizes. Additional simulations reveal that ATIS is efficient for binary data (e.g., presence or absence) whereas TACS is inefficient for binary data. The overall results indicate that ATIS is consistently efficient compared to conventional sampling and to adaptive cluster sampling in some important cases.

PLoS ONE↗