Shearwaters and petrels (hereafter petrels) are highly adapted seabirds that occur across all the world's oceans. Petrels are a threatened seabird group comprising 124 species. They have bet-hedging life histories typified by extended chick rearing periods, low fecundity, high adult survival, strong philopatry, monogamy and long-term mate fidelity and are thus vulnerable to change. Anthropogenic alterations on land and at sea have led to a poor conservation status of many petrels with 52 (42%) threatened

[1] A dynamic model is constructed for interactive silicon, nitrogen, sulfur processing in and below Arctic sea ice, by ecosystems residing in the lower few centimeters of the distributed pack. A biogeochemically active bottom layer supporting sources/sinks for the pennate diatoms is appended to thickness categories of a global sea ice code. Nutrients transfer from the ocean mixed layer to drive algal growth, while sulfur metabolites are reinjected from the ice interface. Freeze, flux, flush and melt processes are linked to multielement geocycling for the entire high-latitude regime. Major element kinetics are optimized initially to reproduce chlorophyll observations, which extend across the seasons. Principal influences on biomass are solute exchange velocity at the solid interface, optical averaging in active ice and cell retention against ablation. The sulfur mechanism encompasses open water features such as accumulation of particulate dimethyl sulfoniopropionate, grazing and other disruptive releases, plus bacterial/enzymatic conversion to volatile dimethyl sulfide. For baseline settings, the mixed layer trace gas distribution matches sparging measurements where they are available. However, concentrations rise to well over 10 nM in remote, unsampled locations. Peak contributions are supported by ice grazing, mortality and fractional melting. The model bottom layer adds substantially to a ring maximum of reduced sulfur chemistry that may be dominant across the marginal Arctic environment. Sensitivity tests on this scenario include variation of cell sulfur composition and remineralization, routings/chemical time scales, and the physical dimension of water layers. An alternate possibility that peripheral additions are small cannot be excluded from the outcomes. It is concluded that seagoing dimethyl sulfide data are far too sparse at the present time to distinguish sulfur-ice production levels.Citation: Elliott, S., C. Deal, G. Humphries, E. Hunke, N. Jeffery, M. Jin, M. Levasseur, and J. Stefels (2012), Pan-Arctic simulation of coupled nutrient-sulfur cycling due to sea ice biology: Preliminary results,

The Mapping Application for Penguin Populations and Projected Dynamics (MAPPPD) is a web-based, open access, decision-support tool designed to assist scientists, non-governmental organisations and policy-makers working to meet the management objectives as set forth by the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) and other components of the Antarctic Treaty System (ATS) (that is, Consultative Meetings and the ATS Committee on Environmental Protection). MAPPPD was designed specifically to complement existing efforts such as the CCAMLR Ecosystem Monitoring Program (CEMP) and the ATS site guidelines for visitors. The database underlying MAPPPD includes all publicly available (published and unpublished) count data on emperor, gentoo, Adélie and chinstrap penguins in Antarctica. Penguin population models are used to assimilate available data into estimates of abundance for each site and year. Results are easily aggregated across multiple sites to obtain abundance estimates over any user-defined area of interest. A front end web interface located at www.penguinmap.com provides free and ready access to the most recent count and modelled data, and can act as a facilitator for data transfer between scientists and Antarctic stakeholders to help inform management decisions for the continent.

Most cetacean species are wide-ranging and highly mobile, creating significant challenges for researchers by limiting the scope of data that can be collected and leaving large areas un-surveyed. Aerial surveys have proven an effective way to locate and study cetacean movements but are costly and limited in spatial extent. Here we present a semi-automated pipeline for whale detection from very high-resolution (sub-meter) satellite imagery that makes use of a convolutional neural network (CNN). We trained ResNet, and DenseNet CNNs using down-scaled aerial imagery and tested each model on 31 cm-resolution imagery obtained from the WorldView-3 sensor. Satellite imagery was tiled and the trained algorithms were used to classify whether or not a tile was likely to contain a whale. Our best model correctly classified 100% of tiles with whales, and 94% of tiles containing only water. All model architectures performed well, with learning rate controlling performance more than architecture. While the resolution of commercially-available satellite imagery continues to make whale identification a challenging problem, our approach provides the means to efficiently eliminate areas without whales and, in doing so, greatly accelerates ocean surveys for large cetaceans.

Colonially-breeding seabirds have long served as indicator species for the health of the oceans on which they depend. Abundance and breeding data are repeatedly collected at fixed study sites in the hopes that changes in abundance and productivity may be useful for adaptive management of marine resources, but their suitability for this purpose is often unknown. To address this, we fit a Bayesian population dynamics model that includes process and observation error to all known Adélie penguin abundance data (1982–2015) in the Antarctic, covering >95% of their population globally. We find that process error exceeds observation error in this system, and that continent-wide “year effects” strongly influence population growth rates. Our findings have important implications for the use of Adélie penguins in Southern Ocean feedback management, and suggest that aggregating abundance across space provides the fastest reliable signal of true population change for species whose dynamics are driven by stochastic processes.

We present a first compilation, quantification and summary of 27 seabird species presence data for north of the Arctic circle (>66 degrees latitude North) and the ice-free period (summer). For species names, we use several taxonomically valid online databases [Integrated Taxonomic Information System (ITIS), AviBase, 4 letter species codes of the American Ornithological Union (AOU), The British List 2000, taxonomic serial numbers TSNs, World Register of Marine Species (WORMS) and APHIA ID] allowing for a compatible taxonomic species cross-walk, and subsequent applications, e.g., phylogenies. Based on the data mining and machine learning RandomForest algorithm, and 26 environmental publicly available Geographic Information Systems (GIS) layers, we built 27 predictive seabird models based on public open access data archives such as the Global Biodiversity Information Facility (GBIF), North Pacific Pelagic Seabird Database (NPPSD) and PIROP database (in OBIS-Seamap). Model-prediction scenarios using pseudo-absence and expert-derived absence were run; aspatial and spatial model assessment metrics were applied. Further, we used an additional species model performance metric based on the best publicly available Arctic seabird colony location datasets compiled by the authors using digital and literature sources. The obtained models perform reasonably: from poor (only a few coastal species with low samples) to very high (many pelagic species). In compliance with data policies of the International Polar Year (IPY) and similar initiatives, data and models are documented with FGDC NBII metadata and publicly available online for further improvement, sustainability applications, synergy, and intellectual explorations in times of a global biodiversity, ocean and Arctic crisis.

No abstract

No abstract