Reconciling multiple data sources to improve accuracy of large-scale prediction of forest disease incidence

Melding wildlife surveys to improve conservation inference

Integrated models are a popular tool for analyzing species of conservation concern. Species of conservation concern are often monitored by multiple entities that generate several datasets. Individually, these datasets may be insufficient for guiding management due to low spatio-temporal resolution, biased sampling, or large observational uncertainty. Integrated models provide an approach for assimilating multiple datasets in a coherent framework that can compensate for these deficiencies. While conventional integrated models have been used to assimilate count data with surveys of survival, fecundity, and harvest, they can also assimilate ecological surveys that have differing spatio-temporal regions and observational uncertainties. Motivated by independent aerial and ground surveys of lesser prairie-chicken, we developed an integrated modeling approach that assimilates density estimates derived from surveys with distinct sources of observational error into a joint framework that provides shared inference on spatio-temporal trends. We model these data using a Bayesian Markov melding approach and apply several data augmentation strategies for efficient sampling. In a simulation study, we show that our integrated model improved predictive performance relative to models for analyzing the surveys independently. We use the integrated model to facilitate prediction of lesser prairie-chicken density at unsampled regions and perform a sensitivity analysis to quantify the inferential cost associated with reduced survey effort.

Integrating presence-only and occupancy data to model habitat use for the northernmost population of jaguars

Species distribution models (SDMs) have become an essential tool for the management and conservation of imperiled species. However, many at-risk species are rare and characterized by limited data on their spatial distribution and habitat relationships. This has led to the development of SDMs that integrate multiple types and sources of data to leverage more information and provide improved predictions of habitat associations. We developed a novel integrated species distribution model to predict habitat suitability for jaguars (Panthera onca) in the border region between northern Mexico and the southwestern USA. Our model combined presence-only and occupancy data to identify key environmental correlates, and we used model results to develop a probability of use map. We adopted a logistic regression modeling framework, which we found to be more straightforward and less computationally intensive to fit than Poisson point process-based models. Model results suggested that high terrain ruggedness and the presence of riparian vegetation were most strongly related to habitat use by jaguars in our study region. Our best model, on average, predicted that there is currently 25,463 km² of usable habitat in our study region. The United States portion of the study region, which makes up 38.6% of the total area, contained 40.6% of the total usable habitat. Even though there have been few detections of jaguars in the southwestern USA in recent decades, our results suggest that protection of currently suitable habitats, along with increased conservation efforts, could significantly contribute to the recovery of jaguars in the USA.

Gaining Insights about Forest Health Prescriptions from Loggers and Foresters: Understudied Voices in the Human Dimensions of Forest Health

Maintaining healthy forests requires multiple individuals, including foresters who develop timber sale silvicultural prescriptions and loggers who implement those prescriptions, resulting in the transplantation of forest health science into workable management plans. However, data on the experiences, attitudes, and opinions of these two groups are often missing when developing or refining forest health treatment strategies. To explore the role that these groups play in sustaining forest health, we examined timber sale administrators' and loggers' perspectives on treatment approaches for eastern spruce dwarf mistletoe (Arceuthobium pusillum) (ESDM), a parasitic plant native to Minnesota that increases mortality and reduces growth rate and regeneration success of black spruce (Picea mariana). While ESDM has been managed for decades in black spruce stands in Minnesota, little is known about the effectiveness of the management approaches. Data were gathered through interviews and focus groups with loggers, as well as an online survey and focus groups with foresters who administer timber sales. Study participants identified a range of field-based barriers, knowledge gaps, and uncertainties that hamper the ability to effectively implement ESDM treatment strategies as designed, including financial, administrative, informational, policy-related, and environmental factors. These factors have a significant bearing on the ability to effectively implement ESDM treatment approaches; yet may be factors that were not known or considered when developing treatment strategies. This case study underscores the value of nurturing a science-management partnership to ensure that a broad set of voices are considered when developing or revising forest health treatment strategies.

Impact of High-Cadence Earth Observation in Maize Crop Phenology Classification

For farmers, policymakers, and government agencies, it is critical to accurately define agricultural crop phenology and its spatial-temporal variability. At the moment, two approaches are utilized to report crop phenology. On one hand, land surface phenology provides information about the overall trend, whereas weekly reports from USDA-NASS provide information about the development of particular crops at the regional level. High-cadence earth observations might help to improve the accuracy of these estimations and bring more precise crop phenology classifications closer to what farmers demand. The second component of the proposed solution requires the use of robust classifiers (e.g., random forest, RF) capable of successfully managing large data sets. To evaluate this solution, this study compared the output of a RF classifier model using weather, two different satellite sources (Planet Fusion; PF and Sentinel-2; S-2), and ground truth data to improve maize (Zea mays L.) crop phenology classification using two regions of Kansas (Southwest and Central) as a testbed during the 2017 growing season. Our findings suggests that high temporal resolution (PF) data can significantly improve crop classification metrics (f1-score = 0.94) relative to S-2 (f1-score = 0.86). Additionally, a decline in the f1-score between 0.74 and 0.60 was obtained when we assessed the ability of S-2 to extend the temporal forecast for crop phenology. This research highlights the critical nature of very high temporal resolution (daily) earth observation data for crop monitoring and decision making in agriculture.

What processes must we understand to forecast regional-scale population dynamics?

An urgent challenge facing biologists is predicting the regional-scale population dynamics of species facing environmental change. Biologists suggest that we must move beyond predictions based on phenomenological models and instead base predictions on underlying processes. For example, population biologists, evolutionary biologists, community ecologists and ecophysiologists all argue that the respective processes they study are essential. Must our models include processes from all of these fields? We argue that answering this critical question is ultimately an empirical exercise requiring a substantial amount of data that have not been integrated for any system to date. To motivate and facilitate the necessary data collection and integration, we first review the potential importance of each mechanism for skilful prediction. We then develop a conceptual framework based on reaction norms, and propose a hierarchical Bayesian statistical framework to integrate processes affecting reaction norms at different scales. The ambitious research programme we advocate is rapidly becoming feasible due to novel collaborations, datasets and analytical tools.

Clustered versus catastrophic global vertebrate declines

Recent analyses have reported catastrophic global declines in vertebrate populations^1,2. However, the distillation of many trends into a global mean index obscures the variation that can inform conservation measures and can be sensitive to analytical decisions. For example, previous analyses have estimated a mean vertebrate decline of more than 50% since 1970 (Living Planet Index²). Here we show, however, that this estimate is driven by less than 3% of vertebrate populations; if these extremely declining populations are excluded, the global trend switches to an increase. The sensitivity of global mean trends to outliers suggests that more informative indices are needed. We propose an alternative approach, which identifies clusters of extreme decline (or increase) that differ statistically from the majority of population trends. We show that, of taxonomic-geographic systems in the Living Planet Index, 16 systems contain clusters of extreme decline (comprising around 1% of populations; these extreme declines occur disproportionately in larger animals) and 7 contain extreme increases (around 0.4% of populations). The remaining 98.6% of populations across all systems showed no mean global trend. However, when analysed separately, three systems were declining strongly with high certainty (all in the Indo-Pacific region) and seven were declining strongly but with less certainty (mostly reptile and amphibian groups). Accounting for extreme clusters fundamentally alters the interpretation of global vertebrate trends and should be used to help to prioritize conservation efforts.

The impact of RHDV-K5 on rabbit populations in Australia: an evaluation of citizen science surveys to monitor rabbit abundance

The increasing popularity of citizen science in ecological research has created opportunities for data collection from large teams of observers that are widely dispersed. We established a citizen science program to complement the release of a new variant of the rabbit biological control agent, rabbit haemorrhagic disease virus (RHDV), known colloquially as K5, across Australia. We evaluated the impact of K5 on the national rabbit population and compared citizen science and professionally-collected spotlight count data. Of the citizen science sites (n = 219), 93% indicated a decrease in rabbit abundance following the release of K5. The overall finite monthly growth rate in rabbit abundance was estimated as 0.66 (95%CI, 0.26, 1.03), averaging a monthly reduction of 34% at the citizen science sites one month after the release. No such declines were observed at the professionally monitored sites (n = 22). The citizen science data submissions may have been unconsciously biased or the number of professional sites may have been insufficient to detect a change. Citizen science participation also declined by 56% over the post-release period. Future programs should ensure the use of blinded trials to check for unconscious bias and consider how incentives and/or the good will of the participants can be maintained throughout the program.

A hierarchical Bayesian approach for handling missing classification data

Ecologists use classifications of individuals in categories to understand composition of populations and communities. These categories might be defined by demographics, functional traits, or species. Assignment of categories is often imperfect, but frequently treated as observations without error. When individuals are observed but not classified, these "partial" observations must be modified to include the missing data mechanism to avoid spurious inference.We developed two hierarchical Bayesian models to overcome the assumption of perfect assignment to mutually exclusive categories in the multinomial distribution of categorical counts, when classifications are missing. These models incorporate auxiliary information to adjust the posterior distributions of the proportions of membership in categories. In one model, we use an empirical Bayes approach, where a subset of data from one year serves as a prior for the missing data the next. In the other approach, we use a small random sample of data within a year to inform the distribution of the missing data.We performed a simulation to show the bias that occurs when partial observations were ignored and demonstrated the altered inference for the estimation of demographic ratios. We applied our models to demographic classifications of elk (Cervus elaphus nelsoni) to demonstrate improved inference for the proportions of sex and stage classes.We developed multiple modeling approaches using a generalizable nested multinomial structure to account for partially observed data that were missing not at random for classification counts. Accounting for classification uncertainty is important to accurately understand the composition of populations and communities in ecological studies.

Comparison of large-scale citizen science data and long-term study data for phenology modeling

Large‐scale observational data from citizen science efforts are becoming increasingly common in ecology, and researchers often choose between these and data from intensive local‐scale studies for their analyses. This choice has potential trade‐offs related to spatial scale, observer variance, and interannual variability. Here we explored this issue with phenology by comparing models built using data from the large‐scale, citizen science USA National Phenology Network (USA‐NPN) effort with models built using data from more intensive studies at Long Term Ecological Research (LTER) sites. We built statistical and process based phenology models for species common to each data set. From these models, we compared parameter estimates, estimates of phenological events, and out‐of‐sample errors between models derived from both USA‐NPN and LTER data. We found that model parameter estimates for the same species were most similar between the two data sets when using simple models, but parameter estimates varied widely as model complexity increased. Despite this, estimates for the date of phenological events and out‐of‐sample errors were similar, regardless of the model chosen. Predictions for USA‐NPN data had the lowest error when using models built from the USA‐NPN data, while LTER predictions were best made using LTER‐derived models, confirming that models perform best when applied at the same scale they were built. This difference in the cross‐scale model comparison is likely due to variation in phenological requirements within species. Models using the USA‐NPN data set can integrate parameters over a large spatial scale while those using an LTER data set can only estimate parameters for a single location. Accordingly, the choice of data set depends on the research question. Inferences about species‐specific phenological requirements are best made with LTER data, and if USA‐NPN or similar data are all that is available, then analyses should be limited to simple models. Large‐scale predictive modeling is best done with the larger‐scale USA‐NPN data, which has high spatial representation and a large regional species pool. LTER data sets, on the other hand, have high site fidelity and thus characterize inter‐annual variability extremely well. Future research aimed at forecasting phenology events for particular species over larger scales should develop models that integrate the strengths of both data sets.

Predicting wildlife road-crossing probability from roadkill data using occupancy-detection models

Wildlife-vehicle collisions (WVC) represent a major threat for wildlife and understanding how WVC spatial patterns relate to surrounding land cover can provide valuable information for deciding where to implement mitigation measures. However, these relations may be heavily biased as many casualties are undetected in roadkill surveys, e.g. due to scavenger activity, which may ultimately jeopardize conservation actions. We suggest using occupancy models to overcome imperfect detection issues, assuming that 'occupancy' represents the preference for crossing the road in a given site, i.e. is a proxy for the roadkill risk; and that the 'detectability' is the joint probability of an animal being hit in the crossing site and its carcass being detected afterwards. Our main objective was to assess the roadkill risk along roads while accounting for imperfect detection issues and relate it to land cover information. We conducted roadkill surveys over 114 km in nine different roads, biweekly, for five years (total of 484 surveys), and developed a Bayesian hierarchical occupancy model to assess the roadkill risk for the six most road-killed taxa for each road section and season (WET and DRY). Overall, we estimated a higher roadkill risk in road sections surrounded by agriculture, open habitats; and a higher detectability within the 4-lane road sections. Our modeling framework has a great potential to overcome the limitations related to imperfect detection when assessing the roadkill risk and may become an important tool to predict which road sections have a higher mortality risk.

Understanding the demographic drivers of realized population growth rates

Identifying the demographic parameters (e.g., reproduction, survival, dispersal) that most influence population dynamics can increase conservation effectiveness and enhance ecological understanding. Life table response experiments (LTRE) aim to decompose the effects of change in parameters on past demographic outcomes (e.g., population growth rates). But the vast majority of LTREs and other retrospective population analyses have focused on decomposing asymptotic population growth rates, which do not account for the dynamic interplay between population structure and vital rates that shape realized population growth rates (λt=Nt+1/Nt) in time-varying environments. We provide an empirical means to overcome these shortcomings by merging recently developed "transient life-table response experiments" with integrated population models (IPMs). IPMs allow for the estimation of latent population structure and other demographic parameters that are required for transient LTRE analysis, and Bayesian versions additionally allow for complete error propagation from the estimation of demographic parameters to derivations of realized population growth rates and perturbation analyses of growth rates. By integrating available monitoring data for Lesser Scaup over 60 yr, and conducting transient LTREs on IPM estimates, we found that the contribution of juvenile female survival to long-term variation in realized population growth rates was 1.6 and 3.7 times larger than that of adult female survival and fecundity, respectively. But a persistent long-term decline in fecundity explained 92% of the decline in abundance between 1983 and 2006. In contrast, an improvement in adult female survival drove the modest recovery in Lesser Scaup abundance since 2006, indicating that the most important demographic drivers of Lesser Scaup population dynamics are temporally dynamic. In addition to resolving uncertainty about Lesser Scaup population dynamics, the merger of IPMs with transient LTREs will strengthen our understanding of demography for many species as we aim to conserve biodiversity during an era of non-stationary global change.

Determining Occurrence Dynamics when False Positives Occur: Estimating the Range Dynamics of Wolves from Public Survey Data

Large-scale presence-absence monitoring programs have great promise for many conservation applications. Their value can be limited by potential incorrect inferences owing to observational errors, especially when data are collected by the public. To combat this, previous analytical methods have focused on addressing non-detection from public survey data. Misclassification errors have received less attention but are also likely to be a common component of public surveys, as well as many other data types. We derive estimators for dynamic occupancy parameters (extinction and colonization), focusing on the case where certainty can be assumed for a subset of detections. We demonstrate how to simultaneously account for non-detection (false negatives) and misclassification (false positives) when estimating occurrence parameters for gray wolves in northern Montana from 2007-2010. Our primary data source for the analysis was observations by deer and elk hunters, reported as part of the state's annual hunter survey. This data was supplemented with data from known locations of radio-collared wolves. We found that occupancy was relatively stable during the years of the study and wolves were largely restricted to the highest quality habitats in the study area. Transitions in the occupancy status of sites were rare, as occupied sites almost always remained occupied and unoccupied sites remained unoccupied. Failing to account for false positives led to over estimation of both the area inhabited by wolves and the frequency of turnover. The ability to properly account for both false negatives and false positives is an important step to improve inferences for conservation from large-scale public surveys. The approach we propose will improve our understanding of the status of wolf populations and is relevant to many other data types where false positives are a component of observations.

Landscape epidemiology of emerging infectious diseases in natural and human-altered ecosystems

A central challenge to studying emerging infectious diseases (EIDs) is a landscape dilemma: Our best empirical understanding of disease dynamics occurs at local scales, whereas pathogen invasions and management occur over broad spatial extents. The burgeoning field of landscape epidemiology integrates concepts and approaches from disease ecology with the macroscale lens of landscape ecology, enabling examination of disease across spatiotemporal scales in complex environmental settings. We review the state of the field and describe analytical frontiers that show promise for advancement, focusing on natural and human-altered ecosystems. Concepts fundamental to practicing landscape epidemiology are discussed, including spatial scale, static versus dynamic modeling, spatially implicit versus explicit approaches, selection of ecologically meaningful variables, and inference versus prediction. We highlight studies that have advanced the field by incorporating multiscale analyses, landscape connectivity, and dynamic modeling. Future research directions include understanding disease as a component of interacting ecological disturbances, scaling up the ecological impacts of disease, and examining disease dynamics as a coupled human-natural system.