Search

Conservation of predators – especially large carnivores and those that potentially pose threats to humans – can be controversial among stakeholders who must coexist with them. What is often overlooked, however, are the direct and indirect ecosystem services and disservices predators provide as a result of consumption of herbivores (“predation services”). We used a theoretical predator–prey–economic model to examine when predators are likely to provide a net service to society, by comparing services/disservices to a predator‐free counterfactual scenario. We found that net predator services were strongly dependent on how per‐capita services and disservices of predators and prey changed with abundance (ie assumed marginal value [MV] functions of service/disservice). We suggest that further empirical research is needed into MVs of services/disservices of wildlife, because transferring net services among locations – a common practice – is problematic unless MV functions are known. Rigorously quantifying services/disservices of predators could improve conservation and management outcomes by increasing effective communication to diverse stakeholders.

As species’ distributions shift in response to changing environmental conditions, novel species interactions are emerging, challenging researchers and managers alike. When new species enter existing ecological communities, their effects on fellow guild members are difficult to predict. Competition between species within the same guild can regulate populations and influence community structure, but niche partitioning is a way for sympatric species to coexist through differential use of space, time, and resources. The fisher (Pekania pennanti) is a carnivore species that has recently colonized Southeast Alaska (SEAK), the northern margin of its range. Although currently considered at low density, the fisher has been expanding its distribution in SEAK in recent years and may impact the native SEAK forest carnivore guild through competitive interactions. We examined the potential for competition between fisher, American marten (Martes americana), and ermine (Mustela erminea) in SEAK during the winter of 2018 by investigating patterns of spatial and temporal overlap between the species at both local and landscape scales. Using data collected from remote‐sensed habitat features, climate characteristics, species detections from 25 systematic camera stations (50 unique cameras), and single‐species occupancy models, we found that fisher occupancy was positively correlated with vegetation height, and detection decreased with increased snow density. Using two‐species occupancy models, we found evidence of temporal, rather than spatial, avoidance patterns among the SEAK mustelid species investigated, although our inference was limited by sample size. Fisher, the largest mustelid evaluated, did not exhibit changes in occupancy or detection in the presence of marten or ermine. Likewise, detection and occupancy parameters of the mid‐sized marten and small‐sized ermine were not affected by the presence of the other mustelids. Instead, analysis of diel activity patterns showed that marten and ermine reduced temporal overlap with fisher, exhibiting less activity during times of peak fisher activity. Overall, these results show that mustelids in SEAK displayed a pattern of hierarchical temporal partitioning during winter, with smaller‐bodied mustelids avoiding larger, more dominant mustelids. However, if the fisher population continues to expand its range and abundance in SEAK, species interactions may shift and increase the potential for competition between the fisher, marten, and ermine mustelid guild.

Habitat loss is often the ultimate cause of species endangerment and is also a leading factor inhibiting species recovery. For this reason, species‐at‐risk legislation, policies and plans typically focus on habitat conservation and restoration as mechanisms for recovery. To assess the effectiveness of these instruments in decelerating habitat loss, we evaluated spatiotemporal habitat changes for an iconic endangered species, woodland caribou (Rangifer tarandus caribou). We quantified changes in forest cover, a key proxy of caribou habitat, for all caribou subpopulations in Alberta and British Columbia, Canada. Despite efforts under federal and provincial recovery plans, and requirements listed under Canada's Species at Risk Act, caribou subpopulations lost twice as much habitat as they gained during a 12‐year period (2000–2012). Drivers of habitat loss varied by ecotype, with Boreal and Northern Mountain caribou affected most by forest fire and Southern Mountain caribou affected more by forest harvest. Our case study emphasizes critical gaps between recovery planning and habitat management actions, which are a core expectation under most species‐at‐risk legislation. Loss of caribou habitat from 2000 to 2018 has accelerated. Linear features within caribou ranges have also increased over time, particularly seismic lines within Boreal caribou ranges, and we estimated that only 5% of seismic lines have functionally regenerated. Our findings support the idea that short‐term recovery actions such as predator reductions and translocations will likely just delay caribou extinction in the absence of well‐considered habitat management. Given the magnitude of ongoing habitat change, it is clear that unless the cumulative impacts of land‐uses are effectively addressed through planning and management actions that consider anthropogenic and natural disturbances, we will fail to achieve self‐sustaining woodland caribou populations across much of North America.

Quantifying both the lethal and non-lethal (or “risk”) effects of predation has emerged as a major research focus in carnivore-ungulate systems. While numerous studies have examined predation risk and risk effects in recent decades, a lack of standardization in approaches has impeded progress in the field. We provide an overview of five major study design considerations involved in assessing predation risk and responses of prey in carnivore-ungulate systems, highlighting how different design choices can impact the strength and scope of inference. First, we stress the importance of distinguishing measures of predation risk (probability of being killed) from measures of risk effects (costs of antipredator behaviors in response to risk). Second, we recommend explicit consideration of spatial and temporal scales using a standardized framework to facilitate cross-study comparisons. Third, ungulates use visual, auditory, and olfactory sensory pathways to evaluate predation risk. Experiments that manipulate signals of risk (e.g., auditory playbacks or application of predator scent) can be powerful approaches, but the dosages and types of cues need to be carefully considered. Fourth, ungulates usually face threats from multiple predators simultaneously, and we highlight the potential for remote cameras and structural equation modeling to help address this challenge. Fifth, emerging technologies may substantially improve our ability to assess risk. We discuss several promising technologies, such as animal-borne video, unmanned aerial vehicles, and physiological sensors. We conclude with general recommendations for study design, which may improve the utility of predation risk research for the conservation and management of carnivore-ungulate systems.

As global climate change progresses, wildlife management will benefit from knowledge of demographic responses to climatic variation, particularly for species already endangered by other stressors. In Canada, climate change is expected to increasingly impact populations of threatened woodland caribou (Rangifer tarandus caribou) and much focus has been placed on how a warming climate has potentially facilitated the northward expansion of apparent competitors and novel predators. Climate change, however, may also exert more direct effects on caribou populations that are not mediated by predation. These effects include meteorological changes that influence resource availability and energy expenditure. Research on other ungulates suggests that climatic variation may have minimal impact on low-density populations such as woodland caribou because per-capita resources may remain sufficient even in “bad” years. We evaluated this prediction using demographic data from 21 populations in western Canada that were monitored for various intervals between 1994 and 2015. We specifically assessed whether juvenile recruitment and adult female survival were correlated with annual variation in meteorological metrics and plant phenology. Against expectations, we found that both vital rates appeared to be influenced by annual climatic variation. Juvenile recruitment was primarily correlated with variation in phenological conditions in the year prior to birth. Adult female survival was more strongly correlated with meteorological conditions and declined during colder, more variable winters. These responses may be influenced by the life history of woodland caribou, which reside in low-productivity refugia where small climatic changes may result in changes to resources that are sufficient to elicit strong demographic effects. Across all models, explained variation in vital rates was low, suggesting that other factors had greater influence on caribou demography. Nonetheless, given the declining trajectories of many woodland caribou populations, our results highlight the increased relevance of recovery actions when adverse climatic conditions are likely to negatively affect caribou demography.

The pathways through which environmental variability affects population dynamics remain poorly understood, limiting ecological inference and management actions. Here, we use matrix-based population models to examine the vital rate responses to environmental variability and individual traits, and subsequent transient dynamics of the population in response to the environment. Using Sitka black-tailed deer (Odocoileus hemionus sitkensis) in Southeast Alaska as a study system, we modeled effects of inter-annual process variance of covariates on female survival, pregnancy rate, and fetal rate, and summer and winter fawn survival. To examine the influence of environmental variance on population dynamics, we compared asymptotic and transient perturbation analysis (elasticity analysis, a life-table response experiment, and transience simulation). We found that summer fawn survival was primarily determined by black bear (Ursus americanus) predation and was positively influenced by mass at birth and female sex. Winter fawn survival was determined by malnutrition in deep-snow winters and was influenced by an interaction between date of birth and snow depth, with late-born fawns at greater risk in deep-snow winters. Adult female survival was the most influential vital rate based on classic elasticity analysis, however, elasticity analysis based on process variation indicated that winter and summer fawn survival were most variable and thus most influential to variability in population growth. Transient dynamics produced by non-stable stage distributions produced realized annual growth rates different from predicted asymptotic growth rates in all years, emphasizing the importance of winter perturbations to population dynamics of this species.

A recurring challenge for resource managers and decision makers is quantifying the trade-offs associated with alternative recovery actions for threatened species. Structured decision-making approaches can help evaluate such complex problems by formalizing objectives and constraints into functions that quantify the benefits and costs associated with each action. Yet, many of the scientific tools necessary to implement structured decision making require extensive literature review and often involve complex algorithms that make them inaccessible to managers. To address these issues, we integrated available information and developed a decision-support tool that managers can readily use to compare costs and benefits associated with alternative recovery actions for threatened species. Our software can be used to quantitatively estimate and compare the costs and demographic benefits of recovery actions for an iconic threatened species, woodland caribou (Rangifer tarandus caribou). While we use caribou as a case study, our approach to developing this management tool is transferable to other threatened taxa. The tool consists of a generalized matrix population model that is parametrized based on information from the published literature or ongoing experiments. Users can input population parameters (e.g., population size and survival rates) or choose from pre-set caribou subpopulations to estimate changes to populations from implementing recovery actions. The tool estimates the trade-offs associated with seven alternative recovery actions: linear feature restoration (LFR), linear feature deactivation (LFD), maternal penning (MP), conservation breeding (CB), predator exclosure (PE), wolf reduction (WR), and moose reduction (MR). We demonstrate our software by comparing recovery actions for the East Side Athabasca River caribou subpopulation and discuss how this tool can be used under a structured decision-making framework. This case study suggests that our open-source tool can be useful to guide wildlife conservation decisions by explicitly estimating costs and benefits associated with recovery actions, which ultimately helps to bridge the gap between management and science via increased accessible application of current knowledge.

Background Temperatures in arctic-boreal regions are increasing rapidly and pose significant challenges to moose (Alces alces), a heat-sensitive large-bodied mammal. Moose act as ecosystem engineers, by regulating forest carbon and structure, below ground nitrogen cycling processes, and predator-prey dynamics. Previous studies showed that during hotter periods, moose displayed stronger selection for wetland habitats, taller and denser forest canopies, and minimized exposure to solar radiation. However, previous studies regarding moose behavioral thermoregulation occurred in Europe or southern moose range in North America. Understanding whether ambient temperature elicits a behavioral response in high-northern latitude moose populations in North America may be increasingly important as these arctic-boreal systems have been warming at a rate two to three times the global mean. Methods We assessed how Alaska moose habitat selection changed as a function of ambient temperature using a step-selection function approach to identify habitat features important for behavioral thermoregulation in summer (June–August). We used Global Positioning System telemetry locations from four populations of Alaska moose (n = 169) from 2008 to 2016. We assessed model fit using the quasi-likelihood under independence criterion and conduction a leave-one-out cross validation. Results Both male and female moose in all populations increasingly, and nonlinearly, selected for denser canopy cover as ambient temperature increased during summer, where initial increases in the conditional probability of selection were initially sharper then leveled out as canopy density increased above ~ 50%. However, the magnitude of selection response varied by population and sex. In two of the three populations containing both sexes, females demonstrated a stronger selection response for denser canopy at higher temperatures than males. We also observed a stronger selection response in the most southerly and northerly populations compared to populations in the west and central Alaska. Conclusions The impacts of climate change in arctic-boreal regions increase landscape heterogeneity through processes such as increased wildfire intensity and annual area burned, which may significantly alter the thermal environment available to an animal. Understanding habitat selection related to behavioral thermoregulation is a first step toward identifying areas capable of providing thermal relief for moose and other species impacted by climate change in arctic-boreal regions.

Conservation strategies are hindered by a lack of accurate maps of important habitat for many wildlife species, but especially for species inhabiting managed forest landscapes. Prioritizing restoration efforts on Alaska’s Tongass National Forest from past extensive clearcut logging is extremely challenging given the difficulty in accurately mapping its remote, rugged temperate rainforest landscapes. We tested the application of airborne light detection and ranging (LiDAR) technology to build a winter habitat model for Sitka black-tailed deer (Odocoileus hemionus sitkensis), the primary herbivore in the coastal temperate rainforest. We analyzed the importance of geomorphometric and forest structure characteristics as predictors of deer winter habitat selection using Random Forest applied to a 3-year GPS relocation dataset collected from 40 adult female deer. The LiDAR-based habitat model had a predictive performance of 94% (Out-of-bag error = 6%), a 10% lower model error compared to air-photo interpreted polygons and modeled plot data. Random Forest also outperformed analogous resource selection function models based on a comprehensive k-fold cross-validation. Deer habitat selection patterns in the LiDAR-based model were nonlinear across geomorphometric and forest structure predictive variables, and generally supported existing studies of deer habitat selection. Besides improving deer conservation and management on the Tongass National Forest, our approach could greatly enhance the accuracy and resolution of habitat maps used for conservation and restoration planning across large managed forest landscapes.

In seasonally varying environments, animals should alter habitat selection through time to avoid the harshest conditions. Winter severity is limiting for many ungulates in high-latitude ecosystems, and quality of habitat is an important determinant of winter survival. Previous studies in Southeast Alaska indicated that Sitka black-tailed deer (Odocoileus hemionus sitkensis) selected old-growth forest that provides both snow interception and forage, but with great variability among studies, years, and geographic areas. Clearcut timber harvest has greatly reduced the extent and quality of old-growth forest. The value of 2nd-growth and old-growth forest types to deer likely depends on snow depth, which is highly variable in space and time. We measured selection for vegetation classes, landscape features, and forage biomass by monitoring 56 GPS-radiocollared adult female deer from 1 January to 1 April between 2011 and 2013. Simultaneously, we measured snow depth across deer home ranges daily. We determined that snow depth had a strong effect on selection for vegetation classes. During periods of low snow, deer selected young 2nd growth but avoided old 2nd growth and high-volume old growth. As snow depths increased, young 2nd growth was avoided and deer selected old 2nd-growth and productive old-growth forests. The composition of vegetation classes within the landscape influenced selection, with deer selecting locally abundant habitats. These behaviors suggest that the widespread distribution of forest patches that provide snow interception and forage biomass may be critical to fulfilling the energetic requirements of deer during winters with snow. Such context-dependent habitat selection is likely widespread among wildlife species in variable environments and should be incorporated into study design and analysis.

In this paper, we provide an overview of positioning systems for moving resources in forest and fire management and review the related literature. Emphasis is placed on the accuracy and range of different localization and location-sharing methods, particularly in forested environments and in the absence of conventional cellular or internet connectivity. We then conduct a second review of literature and concepts related to several emerging, broad themes in data science, including the terms location-based services (LBS), geofences, wearable technology, activity recognition, mesh networking, the Internet of Things (IoT), and big data. Our objective in this second review is to inform how these broader concepts, with implications for networking and analytics, may help to advance natural resource management and science in the future. Based on methods, themes, and concepts that arose in our systematic reviews, we then augmented the paper with additional literature from wildlife and fisheries management, as well as concepts from video object detection, relative positioning, and inventory-tracking that are also used as forms of localization. Based on our reviews of positioning technologies and emerging data science themes, we present a hierarchical model for collecting and sharing data in forest and fire management, and more broadly in the field of natural resources. The model reflects tradeoffs in range and bandwidth when recording, processing, and communicating large quantities of data in time and space to support resource management, science, and public safety in remote areas. In the hierarchical approach, wearable devices and other sensors typically transmit data at short distances using Bluetooth, Bluetooth Low Energy (BLE), or ANT wireless, and smartphones and tablets serve as intermediate data collection and processing hubs for information that can be subsequently transmitted using radio networking systems or satellite communication. Data with greater spatial and temporal complexity is typically processed incrementally at lower tiers, then fused and summarized at higher levels of incident command or resource management. Lastly, we outline several priority areas for future research to advance big data analytics in natural resources.

The decline of top carnivores has released large herbivore populations around the world, incurring socioeconomic costs such as increased animal–vehicle collisions. Attempts to control overabundant deer in the Eastern United States have largely failed, and deer–vehicle collisions (DVCs) continue to rise at alarming rates. We present the first valuation of an ecosystem service provided by large carnivore recolonization, using DVC reduction by cougars as a case study. Our coupled deer population models and socioeconomic valuations revealed that cougars could reduce deer densities and DVCs by 22% in the Eastern United States, preventing 21,400 human injuries, 155 fatalities, and $2.13 billion in avoided costs within 30 years of establishment. Recently established cougars in South Dakota prevent $1.1 million in collision costs annually. Large carnivore restoration could provide valuable ecosystem services through such socio‐ecological cascades, and these benefits could offset the societal costs of coexistence.

Unmanned aircraft systems (UAS) – also called unmanned aerial vehicles (UAVs) or drones – are an emerging tool that may provide a safer, more cost-effective, and quieter alternative to traditional research methods. We review examples where UAS have been used to document wildlife abundance, behavior, and habitat, and illustrate the strengths and weaknesses of this technology with two case studies. We summarize research on behavioral responses of wildlife to UAS, and discuss the need to understand how recreational and commercial applications of this technology could disturb certain species. Currently, the widespread implementation of UAS by scientists is limited by flight range, regulatory frameworks, and a lack of validation. UAS are most effective when used to examine smaller areas close to their launch sites, whereas manned aircraft are recommended for surveying greater distances. The growing demand for UAS in research and industry is driving rapid regulatory and technological progress, which in turn will make them more accessible and effective as analytical tools.

Science communication is seen as critical for the disciplines of ecology and conservation, where research products are often used to shape policy and decision making. Scientists are increasing their online media communication, via social media and news. Such media engagement has been thought to influence or predict traditional metrics of scholarship, such as citation rates. Here, we measure the association between citation rates and the Altmetric Attention Score—an indicator of the amount and reach of the attention an article has received—along with other forms of bibliometric performance (year published, journal impact factor, and article type). We found that Attention Score was positively correlated with citation rates. However, in recent years, we detected increasing media exposure did not relate to the equivalent citations as in earlier years; signalling a diminishing return on investment. Citations correlated with journal impact factors up to ∼13, but then plateaued, demonstrating that maximizing citations does not require publishing in the highest-impact journals. We conclude that ecology and conservation researchers can increase exposure of their research through social media engagement and, simultaneously, enhance their performance under traditional measures of scholarly activity.

Summary 1. Neonate survival is a key life history trait, yet remains challenging to measure in wild populations because neonates can be difficult to capture at birth. Estimates of survival from neonates that are opportunistically captured might be inaccurate because some individuals die before sampling, resulting in data that are left truncated. The resulting overestimation of survival rates can further affect ecological inference through biased estimates of covariate effects in survival models, yet is not addressed in most studies of animal survival. Here, we quantify the effects of left truncation on survival estimates and subsequent ecological inference. 2. Vaginal implant transmitters (VITs) enable capture of ungulates at birth, yielding data without left truncation. The effects of left truncation on survival estimation were quantified using age-dependent survival models for VIT and opportunistically captured neonatal deer. Differences in daily survival rates (DSRs) and cumulative survival probability were calculated for the first 70 days of life. In addition, left truncation was simulated by removing fawns that died during the first 1 or 2 days of life from the VIT-caught sample, isolating the effect of left truncation. 3. Cumulative probability of survival during the first 70 days of life was overestimated by 7–23% for fawns caught opportunistically compared with those caught by VIT, depending on model design. Differences in DSRs were large at age 1 day, but had converged by age 30 days. Simulated left truncation resulted in overestimates of survival of up to 31%. Model selection and covariate coefficients were strongly affected by left truncation, producing spurious ecological inference, including changes to sign and/or magnitude of inferred effects of all covariates. 4. We recommend (i) every effort be made to capture neonates; (ii) consistent capture methods, using at least in part non-truncating techniques, be implemented across years and study areas; and (iii) exclusion of left-truncated data from survival estimates until DSRs converge with those calculated from non-truncated data. This work emphasizes the importance of accounting for left truncation in survival estimation for any species with strong age-dependent survival in order to prevent biased conclusions produced by sampling method rather than true ecological effects.

Background. The Alexander Archipelago wolf, inhabiting the coastal temperate rainforest of North America, was recently evaluated for protection under the U.S. Endangered Species Act, but ultimately was not listed. Stressors thought to be impacting the population include about habitat alteration from industrial timber harvest and subsequent declines in prey (deer), increased human-caused mortality, and climate change. Methods. To evaluate how these factors likely will affect future abundance of wolves and deer, we constructed a model linking wolf and deer population dynamics to environmental conditions and management regulations. We restricted our model to Prince of Wales and outlying islands, because this area is partially isolated, is the focus of timber harvest in the region, and has the most empirical data available for model parameterization. We examined 6 combinations of future timber harvest, winter severity, wolf harvest regulations, and roads on population dynamics of deer and wolves, developed by a panel of experts. Results. Outcomes across scenarios after 30 years varied, with changes in wolf abundance ranging from a 156% increase to a 41% decline, whereas deer abundance declined from 10−37% after 30 years. Mean percentage of the 31 pack areas that were vacant after 30 years ranged from 0 to 67%, indicating that environmental conditions strongly affected pack success. Variation in wolf abundance was driven primarily by changes in wolf harvest regulations, with smaller contributions from road density, forest succession, and severe-winter frequency. Given current low estimated wolf numbers and continued legal and illegal harvest, this raises conservation concerns for the future of wolves in our study area. In addition, we found that wolf declines could be greater if wolves rely more heavily on deer in the future, for instance if salmon availability declines under future climate change, but also that reduction of deer hunting could increase wolf abundance. Discussion. The potential importance of illegal harvest in wolf population dynamics needs further acknowledgement and treatment. However, changes to harvest regulations, which would not reduce unreported harvest, could still be a powerful tool for management of this small, declining, and insular population. Nevertheless, maximum abundance of wolves appears to be curtailed by the steady decline in carrying capacity forecast across all timber management plans currently under consideration. Although we evaluated factors affecting wolf abundance individually, we encourage a holistic approach to management of this predator-prey system in an altered ecosystem.

Animals play an important role in the seed dispersal of many plants. It is increasingly recognized, however, that the actions of a single disperser rarely determine a seed's fate and final location; rather, multiple abiotic or animal dispersal vectors are involved. Some carnivores act as secondary dispersers by preying on primary seed dispersers or seed predators, inadvertently consuming seeds contained in their prey's digestive tracts and later depositing viable seeds, a process known as diploendozoochory. Carnivores occupy an array of ecological niches and thus range broadly on the landscape. Consequently, secondary seed dispersal by carnivores could have important consequences for plant dispersal outcomes, with implications for ecosystem functioning under a changing climate and across disturbed landscapes where dispersal may be otherwise limited. For example, trophic downgrading through the loss of carnivores may reduce or eliminate diploendozoochory and thus compromise population connectivity for lower trophic levels. We review the literature on diploendozoochory and conclude that the ecological impact of a secondary vs. primary seed disperser depends on the relative dispersal distances, germination success, and the proportion of seeds exposed to secondary dispersal by carnivores. None of the studies up to present day have been able to rigorously assess the ecological significance of this process. We provide a framework of the components that determine the significance of diploendozoochory across systems and identify the components that must be addressed in future studies attempting to assess the ecological importance of diploendozoochory.

New-hoof growth is regarded as the most reliable metric for predicting age of newborn ungulates, but variation in estimated age among hoof-growth equations that have been developed may affect estimates of survival in staggered-entry models. We used known-age newborns to evaluate variation in age estimates among existing hoof-growth equations and to determine the consequences of that variation on survival estimates. During 2001–2009, we captured and radiocollared 174 newborn (≤24-hrs old) ungulates: 76 white-tailed deer (Odocoileus virginianus) in Minnesota and South Dakota, 61 mule deer (O. hemionus) in California, and 37 pronghorn (Antilocapra americana) in South Dakota. Estimated age of known-age newborns differed among hoof-growth models and varied by >15 days for white-tailed deer, >20 days for mule deer, and >10 days for pronghorn. Accuracy (i.e., the proportion of neonates assigned to the correct age) in aging newborns using published equations ranged from 0.0% to 39.4% in white-tailed deer, 0.0% to 3.3% in mule deer, and was 0.0% for pronghorns. Results of survival modeling indicated that variability in estimates of age-at-capture affected short-term estimates of survival (i.e., 30 days) for white-tailed deer and mule deer, and survival estimates over a longer time frame (i.e., 120 days) for mule deer. Conversely, survival estimates for pronghorn were not affected by estimates of age. Our analyses indicate that modeling survival in daily intervals is too fine a temporal scale when age-at-capture is unknown given the potential inaccuracies among equations used to estimate age of neonates. Instead, weekly survival intervals are more appropriate because most models accurately predicted ages within 1 week of the known age. Variation among results of neonatal-age models on short- and long-term estimates of survival for known-age young emphasizes the importance of selecting an appropriate hoof-growth equation and appropriately defining intervals (i.e., weekly versus daily) for estimating survival.