Remote Sensing of Antarctic Penguin Populations
Penguins, high trophic-level predators almost exclusively confined to the Southern Ocean, are believed to be particularly susceptible to the unprecedented climatic changes that are currently being experienced in the region. Indeed, the two species of interest to this research, the chinstrap and gentoo penguins, are designated as ‘indicator species’ or sentinels of change within the natural environment by the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR), the responsible international agency for conserving Antarctic marine life. However, despite the intrinsic role that the species play, there is a dearth of knowledge about even basic demographic and biological aspects (census, distribution, habitat requirements, lifecycles) due, in the main, to the significant environmental and logistical barriers that are presented when considering field surveys within the region. As such, the potential of remote sensing applications and aligned software are beginning to be realised and are proving particularly apt at augmenting the data collected from the more traditional methods of ground-surveys and the laborious counting of species manually from imagery.
To test this belief, freely-available ‘open-source’ software was used to design and develop research-specific methodological approaches to provide both population census information and to calculate nesting densities from aerial photography taken of the Cape Shirreff rookery, Livingston Island, the South Shetland Islands; with open-source software explicitly chosen in preference to commercial packages to test the potential of and for such software and the approaches described herein to be used by all, regardless of background and experience. The methodological approaches developed produced very favourable results: for population census, the counts were within 5% of the actual in-situ ground-counts recorded by the US Antarctic Marine Living Resources (US AMLR) programme; whilst nest-to-nest distances and colony density calculations correlated very well with the (admittedly, limited) published data, indicating that the adopted approaches described herein may be reliably utilised for future surveys, albeit with some modifications. Two further, unheralded, revelations emerged: firstly, that nest-to-nest distances of and between the two species increased markedly within congeneric colonies when compared to those colonies where only one species is nesting; whilst, secondly, the colonies are situated within two quite narrow bands within the rookery, leaving a broad swath of ostensibly suitable territory uncolonized. Whilst the reasons are somewhat uncertain, these observations further illustrate the imperative need for a concerted research campaign of appropriate spatio-temporal extent.