Knowledge about the spatial distribution of seabirds at sea is important for conservation. During marine conservation planning, logistical constraints preclude seabird surveys covering the complete area of interest and spatial distribution of seabirds is frequently inferred from predictive statistical models. Increasingly complex models are available to relate the distribution and abundance of pelagic seabirds to environmental variables, but a comparison of their usefulness for delineating protected areas for seabirds is lacking. Here we compare the performance of five modelling techniques (generalised linear models, generalised additive models, Random Forest, boosted regression trees, and maximum entropy) to predict the distribution of Balearic Shearwaters (Puffinus mauretanicus) along the coast of the western Iberian Peninsula. We used ship transect data from 2004 to 2009 and 13 environmental variables to predict occurrence and density, and evaluated predictive performance of all models using spatially segregated test data. Predicted distribution varied among the different models, although predictive performance varied little. An ensemble prediction that combined results from all five techniques was robust and confirmed the existence of marine important bird areas for Balearic Shearwaters in Portugal and Spain. Our predictions suggested additional areas that would be of high priority for conservation and could be proposed as protected areas. Abundance data were extremely difficult to predict, and none of five modelling techniques provided a reliable prediction of spatial patterns. We advocate the use of ensemble modelling that combines the output of several methods to predict the spatial distribution of seabirds, and use these predictions to target separate surveys assessing the abundance of seabirds in areas of regular use.
Distributions of Species
The two stocks of Steller sea lions (Eumetopias jubatus) in Alaska include an endangered western stock, recently recovering in parts of its range following decades of decline, and an eastern stock which was removed from the U.S. Endangered Species List in 2013 following increasing numbers since the 1970s. Information on overlapping distributions of eastern and western sea lions is needed for management considerations. We analyzed >30,000 sightings collected from 2000–2014 of 2,385 sea lions that were branded as pups at 10 Alaskan rookeries to examine mesoscale (mostly <500km) spatial distribution, geographic range, and geographic population structure based on natal rookery, sex, and age during breeding and non-breeding seasons. Analyses of summary movement measures (e.g., natal rookery, sex, and age-class differences in spatial distribution and geographic range) indicate wide variation in rookery-specific movement patterns. Correlations between movement measures and population dynamics suggested movement patterns could be a function of density dependence. Animals from larger rookeries, and rookeries with slower population growth and lower survival, had wider dispersion than animals from smaller rookeries, or rookeries with high growth and survival. Sea lions from the largest rookery, Forrester Island, where survival and population trends are lowest, were the most widely distributed. Analysis of geographic population structure indicated that animals born in the eastern Aleutian Islands had the most distinct movements and had little overlap with other western sea lions. Northern Southeast Alaska, within the eastern stock, is the area of greatest overlap between stocks, and is important to western animals, especially those born in Prince William Sound. Detailed knowledge of distribution and movements of western sea lions is useful for defining recovery and population trend analysis regions that better reflect dispersion and population structure and provides valuable information to managers as critical habitat is re-evaluated and the location of the stock boundary reconsidered.
Mid-latitude (∼30-60°) seasonally stratifying shelf-seas support a high abundance and diversity of marine predators such as marine mammals and seabirds. However, anthropogenic activities and climate change impacts are driving changes in the distributions and population dynamics of these animals, with negative consequences for ecosystem functioning. Across mid-latitude shelf-seas marine mammals and seabirds are known to forage across a number of oceanographic habitats that structure the spatio-temporal distributions of prey fields. Knowledge of these and the bio-physical mechanisms driving such associations are needed to improve marine management and policy. Here, we provide a concise and easily accessible guide for both researchers and managers of marine systems on the predominant oceanographic habitats that are favoured for foraging by marine mammals and seabirds across mid-latitude shelf-seas. We (1) identify and describe key discrete physical features present across the continental shelf, working inshore from the shelf-edge to the shore line, (2) provide an overview of findings relating to associations between these habitats and marine mammals and seabirds, (3) identify areas for future research and (4) discuss the relevance of such information to conservation management. We show that oceanographic features preferentially foraged at by marine mammals and seabirds include shelf-edge fronts, upwelling and tidal-mixing fronts, offshore banks and internal waves, regions of stratification, and topographically complex coastal areas subject to strong tidal flow. Whilst associations were variable across taxa and through space and time, in the majority of cases interactions between bathymetry and tidal currents appear to play a dominant role, alongside patterns in seasonal stratification and shelf-edge upwelling. We suggest that the ecological significance of these bio-physical structures stems from a capacity to alter the densities, distributions (both horizontally and vertically) and/or behaviours of prey in a persistent and/or predictable manner that increases accessibility for predators, and likely enhances foraging efficiency. Future conservation management should aim to preserve and protect these habitats. This will require adaptive and holistic strategies that are specifically tailored to the characteristics of an oceanographic feature, and where necessary, evolve through space and time in response to spatio-temporal variability. Improved monitoring of animal movements and bio-physical conditions across shelf-seas would aid in this. Areas for future research include multi-disciplinary/trophic studies of the mechanisms linking bio-physical processes, prey and marine mammals and seabirds (which may elucidate the importance of lesser studied features such as bottom fronts and Langmuir circulation cells), alongside a better understanding of how predators perceive their environment and develop foraging strategies during immature/juvenile stages. Estimates of the importance of oceanographic habitat features at a population level should also be obtained. Such information is vital to ensuring the future health of these complex ecosystems, and can be used to assess how anthropogenic activities and future environmental changes will impact the functioning and spatio-temporal dynamics of these bio-physical features and their use by marine predators.
Seabirds select suitable habitats at sea, but these habitats may be strongly impacted by marine spatial planning, including the construction of offshore wind farms (OWFs) and the associated ship traffic. Loons (Gavia spp.) are particularly vulnerable to anthropogenic activities and are also of high conservation status, making them particularly relevant to marine planning processes. We investigated the effects of OWF construction and ship traffic on Loon distributions in the German North Sea on a large spatial scale, using a ‘before–after’ control impact analysis approach and a long-term data set. Many OWFs were built in or close to core areas of Loon distributions. Loons showed significant shifts in their distribution in the ‘after’ period and subsequently aggregated between two OWF clusters, indicating the remaining suitable habitat. The decrease in Loon abundance became significant as far as about 16 km from the closest OWF. Ship traffic also had a significant negative impact on Loons, indicating that OWFs deterred Loons through the combined effect of ship traffic and the wind turbines themselves. This study provides the first analysis of the extensive effects of OWFs and ships on Loons on a large spatial scale. The results provide an essential baseline for future marine spatial planning processes in the German North Sea and elsewhere.
Seagrasses form one of the most ecologically important and productive three-dimensional habitats in coastal seas. Knowing the global distribution of seagrass meadows is essential for conservation and blue carbon estimates. Here, we modelled the global distribution of seagrass using 43,037 occurrence records and 13 environmental variables within the modelling software MaxEnt at 30 arc sec resolution (c. 1 km at the equator). We found that sea surface temperature and distance from land contributed most in predicting seagrass distribution globally. Comparison of summing models for individual species, genera, and families found that a model combining all species occurrence records best fitted the known geographic distribution. In addition, this model fills geographic gaps in previous maps. We predicted the seagrass biome may occupy 1,646,788 km2, more than double previous global estimates. Applications for this dataset include blue carbon estimates, spatial planning such as for designing Marine Protected Areas, environmental sensitivity mapping, and monitoring of change in biome cover.
To understand and predict current and future distributions of animals under a changing climate it is essential to establish historical ranges as baselines against which distribution shifts can be assessed. Management approaches also require comprehension of temporal variability in spatial distributions that can occur over shorter time scales, such as inter-annually or seasonally. Focussing on the Southern Ocean, one of the most rapidly changing environments on Earth, we used Species Distribution Models (SDMs) and satellite ocean data to reconstruct the likely historical foraging habitats of Antarctic fur seals (Arctocephalus gazella) from three populations during the non-breeding winter (Marion Island, Bird Island and Cape Shirreff), to assess whether habitat quality has changed in recent decades. We then quantified temporal variability in distributions to assess overlap with management areas (CCAMLR – Commission for the Conservation of Antarctic Marine Living Resources) and the potential for competition with fisheries. Despite notable physical ocean changes, the quality of foraging habitat during the non-breeding season has remained relatively consistent over 20 years at Marion and Bird Islands, but less so at Cape Shirreff, where reduced sea ice cover has improved habitat accessibility. Spatio-temporally explicit SDMs identified variability in habitats across the winter. Some areas overlapped significantly with fisheries activities, suggesting a potential for competition for prey resources at several key periods. A significant component of core habitat at all populations was not within the CCAMLR Convention Area. Although organisations such as CCAMLR adopt a precautionary, ecosystem-based approach to fisheries management, changes to the physical environment and developments in the fishing industry can affect how dependant species are impacted. The hindcasting of historical spatial distributions shown here are baselines against which future changes can be assessed. Given recent proposals for a system of marine protected areas (MPAs) in the Southern Ocean, our results can be used in the design and evaluation of MPAs, be they static or dynamic. Our study also demonstrates that the core habitat of species may fall outside of areas of active management, providing an important context for the interpretation of monitoring programs and management efforts.
Loliginid squids constitute marine resources of increasing importance in shelf ecosystems off the coast of South Brazil. However, the existing information and knowledge about the occurrence of early-life stages and causes of distributional patterns are insufficient. Here, we have revisited Brazilian historical plankton samples obtained from 11 oceanographic surveys to identify paralarvae and their abundances over time. The study area and time period cover the region between Cabo de São Tomé (22°S) and Cananéia (25°S) at depths down to 200 m from 1991 to 2005. Of the 246 paralarvae quantified, ~50% were identified to the genus or species level, including Doryteuthis spp. (D. sanpaulensis and D. plei), Lolliguncula brevis and a single specimen of Pickfordiateuthis pulchella. Paralarval occurrence and abundance peaked in different areas and were associated with distinct oceanographic conditions: D. sanpaulensis occurred in the northern region associated with cold waters and upwelling events, D. plei occurred primarily in the southern region of the study area and in warmer waters, and L. brevis was found in shallow and low salinity waters in the estuarine region off the coast of Santos. Overall, the highest abundance of paralarvae occurred in the nearshore, northernmost areas during summer, and this can be associated with the observed retention mechanisms caused by local circulation, seasonal upwelling, the intrusion of nutrient-rich waters, and spawning peaks. The present study provides new information and evidence for loliginid patterns in the area that may potentially be useful for better understanding the recruitment patterns and fishery assessments of squid populations.
Understanding population dynamics in broadly distributed marine species with cryptic life history stages is challenging. Information on the population dynamics of sea turtles tends to be biased toward females, due to their accessibility for study on nesting beaches. Males are encountered only at sea; there is little information about their migratory routes, residence areas, foraging zones, and population boundaries. In particular, male leatherbacks (Dermochelys coriacea) are quite elusive; little is known about adult and juvenile male distribution or behavior. The at-sea distribution of male turtles from different breeding populations is not known. Here, 122 captured or stranded male leatherback turtles from the USA, Turkey, France, and Canada (collected 1997–2012) were assigned to one of nine Atlantic basin populations using genetic analysis with microsatellite DNA markers. We found that all turtles originated from western Atlantic nesting beaches (Trinidad 55%, French Guiana 31%, and Costa Rica 14%). Although genetic data for other Atlantic nesting populations were represented in the assignment analysis (St. Croix, Brazil, Florida, and Africa (west and south), none of the male leatherbacks included in this study were shown to originate from these populations. This was an unexpected result based on estimated source population sizes. One stranded turtle from Turkey was assigned to French Guiana, while others that were stranded in France were from Trinidad or French Guiana breeding populations. For 12 male leatherbacks in our dataset, natal origins determined from the genetic assignment tests were compared to published satellite and flipper tag information to provide evidence of natal homing for male leatherbacks, which corroborated our genetic findings. Our focused study on male leatherback natal origins provides information not previously known for this cryptic, but essential component of the breeding population. This method should provide a guideline for future studies, with the ultimate goal of improving management and conservation strategies for threatened and endangered species by taking the male component of the breeding population into account.
Ocean currents profoundly impact all life in the oceans and over a broad size spectra species may show both horizontal and vertical movements to stay on preferred locations. As a corollary it might be expected that individuals in preferred oceanic habitats may simply drift with flows. We explored these scenarios by both satellite tracking young pelagic loggerhead turtles and examining the genetic structuring of individuals on coastal foraging areas across the Mediterranean in relation to ocean flows measured both with Lagrangian drifters and a numerical ocean circulation model for the area. Both patterns of movement (n = 18 turtles ranging in size from 41.2 to 68.5 cm CCL tracked for up to 460 days) and genetic structuring (n = 165 individuals from six sites across the ocean basin) suggested that ocean flows profoundly impact the movements of immature turtles and suggest a pattern of largely passive drift within an ocean basin that, throughout, is broadly favourable for developing loggerhead turtles. The situation contrasts with more heterogeneous habitats in the Atlantic and Pacific, where larger amounts of directional swimming may be required to avoid sub-optimum areas.
Atlantic sturgeon (Acipenser oxyrinchus oxyrinchus) is an endangered species that migrate through, and occupy the coastal waters of the mid-Atlantic Bight where they interact with anthropogenic activities. Measures to understand and avoid Atlantic sturgeon that take into consideration the dynamic nature of their habitat may reduce harmful interactions. In this study, we matched fisheries independent biotelemetry observations of Atlantic sturgeon with daily satellite observations to construct a time resolved spatial distribution model of Atlantic sturgeon. We determined that depth, day-of-year, sea surface temperature, and light absorption by seawater were the most important predictors of Atlantic sturgeon occurrence. Demographic factors, such as sex and river-of-origin were of secondary importance. We found strong spatial differences in spring and fall migration patterns, when anthropogenic interactions peak. Our cross-validated models correctly identified > 88% of biotelemetry observations in our study region. Our models also correctly identified ∼64% of bycatch observations throughout the year. However, during their migrations, when harmful interactions were highest, our models correctly identified ∼90% of fisheries dependent observations. We suggest that this model can be used for guidance to managers and stakeholders to reduce interactions with this highly imperiled species, thereby enhancing conservation and recovery efforts.