Predictive models are central to many scientific disciplines and vital for informing management in a rapidly changing world. However, limited understanding of the accuracy and precision of models transferred to novel conditions (their 'transferability') undermines confidence in their predictions. Here, 50 experts identified priority knowledge gaps which, if filled, will most improve model transfers. These are summarized into six technical and six fundamental challenges, which underlie the combined need to intensify research on the determinants of ecological predictability, including species traits and data quality, and develop best practices for transferring models. Of high importance is the identification of a widely applicable set of transferability metrics, with appropriate tools to quantify the sources and impacts of prediction uncertainty under novel conditions. Predicting the UnknownPredictions facilitate the formulation of quantitative, testable hypotheses that can be refined and validated empirically [1]. Predictive models have thus become ubiquitous in numerous scientific disciplines, including ecology [2], where they provide means for mapping species distributions, explaining population trends, or quantifying the risks of biological invasions and disease outbreaks (e.g., [3,4]). The practical value of predictive models in supporting policy and decision making has therefore grown rapidly (Box 1) [5]. With that has come an increasing desire to predict (see Glossary) the state of ecological features (e.g., species, habitats) and our likely impacts upon them [5], prompting a shift from explanatory models to anticipatory predictions [2]. However, in many situations, severe data deficiencies preclude the development of specific models, and the collection of new data can be prohibitively costly or simply impossible [6]. It is in this context that interest in transferable models (i.e., those that can be legitimately projected beyond the spatial and temporal bounds of their underlying data [7]) has grown.Transferred models must balance the tradeoff between estimation and prediction bias and variance (homogenization versus nontransferability, sensu [8]). Ultimately, models that can Highlights Models transferred to novel conditions could provide predictions in data-poor scenarios, contributing to more informed management decisions.The determinants of ecological predictability are, however, still insufficiently understood.Predictions from transferred ecological models are affected by species' traits, sampling biases, biotic interactions, nonstationarity, and the degree of environmental dissimilarity between reference and target systems.We synthesize six technical and six fundamental challenges that, if resolved, will catalyze practical and conceptual advances in model transfers.We propose that the most immediate obstacle to improving understanding lies in the absence of a widely applicable set of metrics for assessing transferability, and that encouraging the development of models grounded in well-established mech...

Cetacean-habitat modeling, although still in the early stages of development, represents a potentially powerful tool for predicting cetacean distributions and understanding the ecological processes determining these distributions. Marine ecosystems vary temporally on diel to decadal scales and spatially on scales from several meters to 1000s of kilometers. Many cetacean species are wideranging and respond to this variability by changes in distribution patterns. Cetacean-habitat models have already been used to incorporate this variability into management applications, including improvement of abundance estimates, development of marine protected areas, and understanding cetacean-fisheries interactions. We present a review of the development of cetacean-habitat models, organized according to the primary steps involved in the modeling process. Topics covered include purposes for which cetacean-habitat models are developed, scale issues in marine ecosystems, cetacean and habitat data collection, descriptive and statistical modeling techniques, model selection, and model evaluation. To date, descriptive statistical techniques have been used to explore cetacean-habitat relationships for selected species in specific areas; the numbers of species and geographic areas examined using computationally intensive statistic modeling techniques are considerably less, and the development of models to test specific hypotheses about the ecological processes determining cetacean distributions has just begun. Future directions in cetacean-habitat modeling span a wide range of possibilities, from development of basic modeling techniques to addressing important ecological questions.

Aim -Enhanced management of areas important for marine biodiversity are now obligations under a range of international treaties. Tracking data provide unparalleled information on the distribution of marine taxa, but there are no agreed guidelines that ensure these data are used consistently to identify biodiversity hotspots and inform marine management decisions. Here we develop methods to standardise the analysis of tracking data to identify sites of conservation importance at global and regional scales.Location -We applied these methods to the largest available compilation of seabird tracking data, covering 60 species, collected from 55 deployment locations ranging from the poles to the tropics.Methods -Key developments include a test for pseudo-replication to assess the independence of two groups of tracking data, an objective approach to define species-specific smoothing parameters (h values) for kernel density estimation based on area-restricted search behaviour, and an analysis to determine whether sites identified from tracked individuals are also representative for the wider population.

These datasets and accompanying syntheses provide a greater understanding of fundamental ecosystem processes in the Southern Ocean, support modelling of predator distributions under future climate scenarios and create inputs that can be incorporated into decision making processes by management authorities. In this data paper, we present the compiled tracking data from research groups that have worked in the Antarctic since the 1990s. The data are publicly available through biodiversity.aq and the Ocean Biogeographic Information System. The archive includes tracking data from over 70 contributors across 12 national Antarctic programs, and includes data from 17 predator species, 4060 individual animals, and over 2.9 million observed locations.Scientific Data | (2020) 7:94 | https://doi.org/10.1038/s41597-020-0406-x www.nature.com/scientificdata www.nature.com/scientificdata/ circum-Antarctic synthesis yet exists that crosses species boundaries. This deficiency prompted the Expert Group on Birds and Marine Mammals (EG-BAMM) and the Expert Group on Antarctic Biodiversity Informatics (EGABI) of the Scientific Committee on Antarctic Research (SCAR; www.scar.org) to initiate in 2010 the Retrospective Analysis of Antarctic Tracking Data (RAATD). RAATD aims to advance our understanding of fundamental and applied questions in a data-driven way, matching research priorities already identified by the SCAR Horizon Scan 9,21 and key questions in animal movement ecology 22 . For these reasons, we worked on the collation, validation and preparation of tracking data collected south of 45 °S. Data from over seventy contributors (Data Contacts and Citations 23 ) were collated. This database includes information from seventeen predator species, 4,060 individuals and over 2.9 million at-sea locations. To exploit this unique dataset, RAATD is undertaking a multi-species assessment of habitat use for higher predators in the Southern Ocean 24 .RAATD will provide a greater understanding of predator distributions under varying climate regimes, and provide outputs that can inform spatial management and planning decisions by management authorities such as the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR; www.ccamlr.org). Our synopsis and analysis of multi-predator tracking data will also highlight regional or seasonal data-gaps.Scientific Data | (2020) 7:94 | https://doi.

Predators and prey assort themselves relative to each other, the availability of resources and refuges, and the temporal and spatial scale of their interaction. Predictive models of predator distributions often rely on these relationships by incorporating data on environmental variability and prey availability to determine predator habitat selection patterns. This approach to predictive modeling holds true in marine systems where observations of predators are logistically difficult, emphasizing the need for accurate models. In this paper, we ask whether including prey distribution data in fine-scale predictive models of bottlenose dolphin (Tursiops truncatus) habitat selection in Florida Bay, Florida, U.S.A., improves predictive capacity. Environmental characteristics are often used as predictor variables in habitat models of top marine predators with the assumption that they act as proxies of prey distribution. We examine the validity of this assumption by comparing the response of dolphin distribution and fish catch rates to the same environmental variables. Next, the predictive capacities of four models, with and without prey distribution data, are tested to determine whether dolphin habitat selection can be predicted without recourse to describing the distribution of their prey. The final analysis determines the accuracy of predictive maps of dolphin distribution produced by modeling areas of high fish catch based on significant environmental characteristics. We use spatial analysis and independent data sets to train and test the models. Our results indicate that, due to high habitat heterogeneity and the spatial variability of prey patches, fine-scale models of dolphin habitat selection in coastal habitats will be more successful if environmental variables are used as predictor variables of predator distributions rather than relying on prey data as explanatory variables. However, predictive modeling of prey distribution as the response variable based on environmental variability did produce high predictive performance of dolphin habitat selection, particularly foraging habitat.

Species distribution models (SDMs) are increasingly applied in conservation management to predict suitable habitat for poorly known populations. High predictive performance of SDMs is evident in validations performed within the model calibration area (interpolation), but few studies have assessed SDM transferability to novel areas (extrapolation), particularly across large spatial scales or pelagic ecosystems. We performed rigorous SDM validation tests on distribution data from three populations of a long-ranging marine predator, the grey petrel Procellaria cinerea, to assess model transferability across the Southern Hemisphere (25-65°S). Oceanographic data were combined with tracks of grey petrels from two remote sub-Antarctic islands (Antipodes and Kerguelen) using boosted regression trees to generate three SDMs: one for each island population, and a combined model. The predictive performance of these models was assessed using withheld tracking data from within the model calibration areas (interpolation), and from a third population, Marion Island (extrapolation). Predictive performance was assessed using k-fold cross validation and point biserial correlation. The two population-specific SDMs included the same predictor variables and suggested birds responded to the same broad-scale oceanographic influences. However, all model validation tests, including of the combined model, determined strong interpolation but weak extrapolation capabilities. These results indicate that habitat use reflects both its availability and bird preferences, such that the realized distribution patterns differ for each population. The spatial predictions by the three SDMs were compared with tracking data and fishing effort to demonstrate the conservation pitfalls of extrapolating SDMs outside calibration regions. This exercise revealed that SDM predictions would have led to an underestimate of overlap with fishing effort and potentially misinformed bycatch mitigation efforts. Although SDMs can elucidate potential distribution patterns relative to large-scale climatic and oceanographic conditions, knowledge of local habitat availability and preferences is necessary to understand and successfully predict region-specific realized distribution patterns.

Small unmanned aircraft systems (sUASs) are fostering novel approaches to marine mammal research, including baleen whale photogrammetry, by providing new observational perspectives. We collected vertical images of 89 gray and 6 blue whales using low cost sUASs to examine the accuracy of image based morphometry. Moreover, measurements from 192 images of a 1 m calibration object were used to examine four different scaling correction models. Results indicate that a linear mixed model including an error term for flight and date contained 0.17 m less error and 0.25 m less bias than no correction. We used the propagation uncertainty law to examine error contributions from scaling and image measurement (digitization) to determine that digitization accounted for 97% of total variance. Additionally, we present a new whale body size metric termed Body Area Index (BAI). BAI is scale invariant and is independent of body length (R2 = 0.11), enabling comparisons of body size within and among populations, and over time. With this study we present a three program analysis suite that measures baleen whales and compensates for lens distortion and corrects scaling error to produce 11 morphometric attributes from sUAS imagery. The program is freely available and is expected to improve processing efficiency and analytical continuity.

In the Northwest Atlantic the distribution of coastal bottlenose dolphins (Tursiups truncatus) overlaps with that of the offshore ecotype. We hypothesized that the distribution of the two ecotypes could be delineated by depth and/or distance from shore, facilitating their identification during surveys. We obtained 304 skin biopsy samples and identified each as either coastal or offshore using analysis of mitochondria1 DNA. We then interpreted the spatial distribution of coastal and offshore forms using spatial analysis. Using a Classification and Regression Tree (CART) analysis, we found a statistically significant break in ecotype distribution at 34 km from shore. In waters beyond 34 km from shore and deeper than 34 m, all bottlenose dolphins were of the offshore ecotype. Within 7.5 km of shore, all 65 samples were of the coastal ecotype. Between these two areas only nine samples were collected, so the genetic composition of bottlenose dolphins in this area remains poorly known. To enhance our understanding of the spatial distribution of the two ecotypes, future research should obtain more biopsy samples in this zone. Nevertheless, our results indicate that a conservative abundance estimate for the coastal ecotype could be generated from surveys of bottlenose dolphins within 7.5 km of shore.