Multivariate classification of animal communication signals: A simulation-based comparison of alternative signal processing procedures using electric fishes

Evolutionary studies of communication can benefit from classification procedures that allow individual animals to be assigned to groups (e.g. species) on the basis of high-dimension data representing their signals. Prior to classification, signals are usually transformed by a signal processing procedure into structural features. Applications of these signal processing procedures to animal communication have been largely restricted to the manual or semi-automated identification of landmark features from graphical representations of signals. Nonetheless, theory predicts that automated time-frequency-based digital signal processing (DSP) procedures can represent signals more efficiently (using fewer features) than can landmark procedures or frequency-based DSP – allowing more accurate classification. Moreover, DSP procedures are objective in that they require little previous knowledge of signal diversity, and are relatively free from potentially ungrounded assumptions of cross-taxon homology. Using a model data set of electric organ discharge waveforms from five sympatric species of the electric fish Gymnotus, we adopted an exhaustive simulation approach to investigate the classificatory performance of different signal processing procedures. We considered a landmark procedure, a frequency-based DSP procedure (the fast Fourier transform), and two kinds of time-frequency-based DSP procedures (a short-time Fourier transform, and several implementations of the discrete wavelet transform -DWT). The features derived from each of these signal processing procedures were then subjected to dimension reduction procedures to separate those features which permit the most effective discrimination among groups of signalers. We considered four alternative dimension reduction methods. Finally, each combination of reduced data was submitted to classification by linear discriminant analysis. Our results support theoretical predictions that time-frequency DSP procedures (especially DWT) permit more efficient discrimination of groups. The performance of signal processing was found to depend largely upon the dimension reduction procedure employed, and upon the number of resulting features. Because the best combinations of procedures are dataset-dependent and difficult to predict, we conclude that simulations of the kind described here, or at least simplified versions of them, should be routinely executed before classification of animal signals - especially unfamiliar ones.     

From Sea to Sea: Canada's Three Oceans of Biodiversity

Evaluating and understanding biodiversity in marine ecosystems are both necessary and challenging for conservation. This paper compiles and summarizes current knowledge of the diversity of marine taxa in Canada''s three oceans while recognizing that this compilation is incomplete and will change in the future. That Canada has the longest coastline in the world and incorporates distinctly different biogeographic provinces and ecoregions (e.g., temperate through ice-covered areas) constrains this analysis. The taxonomic groups presented here include microbes, phytoplankton, macroalgae, zooplankton, benthic infauna, fishes, and marine mammals. The minimum number of species or taxa compiled here is 15,988 for the three Canadian oceans. However, this number clearly underestimates in several ways the total number of taxa present. First, there are significant gaps in the published literature. Second, the diversity of many habitats has not been compiled for all taxonomic groups (e.g., intertidal rocky shores, deep sea), and data compilations are based on short-term, directed research programs or longer-term monitoring activities with limited spatial resolution. Third, the biodiversity of large organisms is well known, but this is not true of smaller organisms. Finally, the greatest constraint on this summary is the willingness and capacity of those who collected the data to make it available to those interested in biodiversity meta-analyses. Confirmation of identities and intercomparison of studies are also constrained by the disturbing rate of decline in the number of taxonomists and systematists specializing on marine taxa in Canada. This decline is mostly the result of retirements of current specialists and to a lack of training and employment opportunities for new ones. Considering the difficulties encountered in compiling an overview of biogeographic data and the diversity of species or taxa in Canada''s three oceans, this synthesis is intended to serve as a biodiversity baseline for a new program on marine biodiversity, the Canadian Healthy Ocean Network. A major effort needs to be undertaken to establish a complete baseline of Canadian marine biodiversity of all taxonomic groups, especially if we are to understand and conserve this part of Canada''s natural heritage.     

Diversity and distribution of marine microbial eukaryotes in the Arctic Ocean and adjacent seas

We analyzed microbial eukaryote diversity in perennially cold arctic marine waters by using 18S rRNA gene clone libraries. Samples were collected during concurrent oceanographic missions to opposite sides of the Arctic Ocean Basin and encompassed five distinct water masses. Two deep water Arctic Ocean sites and the convergence of the Greenland, Norwegian, and Barents Seas were sampled from 28 August to 2 September 2002. An additional sample was obtained from the Beaufort Sea (Canada) in early October 2002. The ribotypes were diverse, with different communities among sites and between the upper mixed layer and just below the halocline. Eukaryotes from the remote Canada Basin contained new phylotypes belonging to the radiolarian orders Acantharea, Polycystinea, and Taxopodida. A novel group within the photosynthetic stramenopiles was also identified. One sample closest to the interior of the Canada Basin yielded only four major taxa, and all but two of the sequences recovered belonged to the polar diatom Fragilariopsis and a radiolarian. Overall, 42% of the sequences were <98% similar to any sequences in GenBank. Moreover, 15% of these were <95% similar to previously recovered sequences, which is indicative of endemic or undersampled taxa in the North Polar environment. The cold, stable Arctic Ocean is a threatened environment, and climate change could result in significant loss of global microbial biodiversity.     

The gait of Australopithecus

A biomechanical analysis of the pelvic and femoral samples available for Australopithecus is presented. No feature of these samples was found to distinguish their gait pattern from that of modern man or to differ in the two presently recognized allomorphs of Australopithecus. Morphological differences between Australopithecus and modern man appear to be the result of different degrees of encephalization rather than any difference in locomotor adaptation.     

Understory bird communities in Amazonian rainforest fragments: species turnover through 25 years post-isolation in recovering landscapes

Inferences about species loss following habitat conversion are typically drawn from short-term surveys, which cannot reconstruct long-term temporal dynamics of extinction and colonization. A long-term view can be critical, however, to determine the stability of communities within fragments. Likewise, landscape dynamics must be considered, as second growth structure and overall forest cover contribute to processes in fragments. Here we examine bird communities in 11 Amazonian rainforest fragments of 1-100 ha, beginning before the fragments were isolated in the 1980s, and continuing through 2007. Using a method that accounts for imperfect detection, we estimated extinction and colonization based on standardized mist-net surveys within discreet time intervals (1-2 preisolation samples and 4-5 post-isolation samples). Between preisolation and 2007, all fragments lost species in an area-dependent fashion, with loss of as few as <10% of preisolation species from 100-ha fragments, but up to 70% in 1-ha fragments. Analysis of individual time intervals revealed that the 2007 result was not due to gradual species loss beginning at isolation; both extinction and colonization occurred in every time interval. In the last two samples, 2000 and 2007, extinction and colonization were approximately balanced. Further, 97 of 101 species netted before isolation were detected in at least one fragment in 2007. Although a small subset of species is extremely vulnerable to fragmentation, and predictably goes extinct in fragments, developing second growth in the matrix around fragments encourages recolonization in our landscapes. Species richness in these fragments now reflects local turnover, not long-term attrition of species. We expect that similar processes could be operating in other fragmented systems that show unexpectedly low extinction.     

Is Niagara Falls a barrier to gene flow in riverine fishes? A test using genome‐wide SNP data from seven native species

Since the early Holocene, fish population genetics in the Laurentian Great Lakes have been shaped by the dual influences of habitat structure and post‐glacial dispersal. Riverscape genetics theory predicts that longitudinal habitat corridors and unidirectional downstream water‐flow drive the downstream accumulation of genetic diversity, whereas post‐glacial dispersal theory predicts that fish genetic diversity should decrease with increasing distance from glacial refugia. This study examines populations of seven native fish species codistributed above and below the 58 m high Niagara Falls – a hypothesized barrier to gene flow in aquatic species. A better understanding of Niagara Falls’ role as a barrier to gene flow and dispersal is needed to identify drivers of Great Lakes genetic diversity and guide strategies to limit exotic species invasions. We used genome‐wide SNPs and coalescent models to test whether populations are: (a) genetically distinct, consistent with the Niagara Falls barrier hypothesis; (b) more genetically diverse upstream, consistent with post‐glacial expansion theory, or downstream, consistent with the riverscape habitat theory; and (c) have migrated either upstream or downstream past Niagara Falls. We found that genetic diversity is consistently greater below Niagara Falls and the falls are an effective barrier to migration, but two species have probably dispersed upstream past the falls after glacial retreat yet before opening of the Welland Canal. Models restricting migration to after opening of the Welland Canal were generally rejected. These results help explain how river habitat features affect aquatic species’ genetic diversity and highlight the need to better understand post‐glacial dispersal pathways.