Successful carnivore identification with faecal DNA across a fragmented Amazonian landscape

The use of scat surveys to obtain DNA has been well documented in temperate areas, where DNA preservation may be more effective than in tropical forests. Samples obtained in the tropics are often exposed to high humidity, warm temperatures, frequent rain and intense sunlight, all of which can rapidly degrade DNA. Despite these potential problems, we demonstrate successful mtDNA amplification and sequencing for faeces of carnivores collected in tropical conditions and quantify how sample condition and environmental variables influence the success of PCR amplification and species identification. Additionally, the feasibility of genotyping nuclear microsatellites from jaguar (Panthera onca) faeces was investigated. From October 2007 to December 2008, 93 faecal samples were collected in the southern Brazilian Amazon. A total of eight carnivore species was successfully identified from 71% of all samples obtained. Information theoretic analysis revealed that the number of PCR attempts before a successful sequence was an important negative predictor across all three responses (success of species identification, success of species identification from the first sequence and PCR amplification success), whereas the relative importance of the other three predictors (sample condition, season and distance from forest edge) varied between the three responses. Nuclear microsatellite amplification from jaguar faeces had lower success rates (15-44%) compared with those of the mtDNA marker. Our results show that DNA obtained from faecal samples works efficiently for carnivore species identification in the Amazon forest and also shows potential for nuclear DNA analysis, thus providing a valuable tool for genetic, ecological and conservation studies.

Successful carnivore identification with faecal DNA across a fragmented Amazonian landscape

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Predator species identification from saliva at kill sites with limited remains

Thesis (Master's)--University of Washington, 2016-06

Preliminary guide to the identification of the early life history stages of Gerreid fishes of the Western Central Atlantic

The family Gerreidae contains four genera and 13 species that occur in the western central North Atlantic. Adult gerreids are small to medium size fishes that are abundant in coastal waters, bays, and estuaries in tropical and warm temperate regions and sometimes occur in freshwaters. They are generally associate~ with grassy or open bottoms, but not with reefs. Gerreids are silvery fishes, with deeply forked tails, and extremely protrusible mouth that points downward when protracted. They apparently feed on bottom-dwelling organisms and at least one species (Eucinostomus gula) shows a distinct transition, during the juvenile period, from a planktivore (exclusively copepods) to a carnivore that includes a diet of almost solely polychaetes (Carr & Adams, 1973; Robins and Ray, 1987; Murdy et al., 1997). (PDF contains 10 pages)

Texture for Script identification

The problem of determining the script and language of a document image has a number of important applications in the field of document analysis, such as indexing and sorting of large collections of such images, or as a precursor to optical character recognition (OCR). In this paper, we investigate the use of texture as a tool for determining the script of a document image, based on the observation that text has a distinct visual texture. An experimental evaluation of a number of commonly used texture features is conducted on a newly created script database, providing a qualitative measure of which features are most appropriate for this task. Strategies for improving classification results in situations with limited training data and multiple font types are also proposed.

Speaker identification using higher order spectral phase features and their effectiveness vis-a-vis Mel-Cepstral features

The effectiveness of higher-order spectral (HOS) phase features in speaker recognition is investigated by comparison with Mel Cepstral features on the same speech data. HOS phase features retain phase information from the Fourier spectrum unlikeMel–frequency Cepstral coefficients (MFCC). Gaussian mixture models are constructed from Mel– Cepstral features and HOS features, respectively, for the same data from various speakers in the Switchboard telephone Speech Corpus. Feature clusters, model parameters and classification performance are analyzed. HOS phase features on their own provide a correct identification rate of about 97% on the chosen subset of the corpus. This is the same level of accuracy as provided by MFCCs. Cluster plots and model parameters are compared to show that HOS phase features can provide complementary information to better discriminate between speakers.