Quantifying Soluble Carbohydrates in Tropical Leaves Using a Portable Mid-Infrared Sensor: Implications for Primate Feeding Ecology

摘要

Identifying the nutritional composition of food items has significant ramifications for primate feeding ecology, which, in turn, influences investigations of primate sociality, cognition, and conservation. The aim of our study was to analyze water soluble carbohydrate (WSC) concentrations in the leaves of trees common to the Diani Forest of Kenya. Many of these leaves are consumed by black and white colobus monkeys (Colobus angolensis palliatus). We assessed whether the infrared spectral data collected using a portable spectrometer can be used to accurately predict WSC concentrations. WSC content was first quantified using the phenol-sulfuric acid method for young and mature leaves of 24 species and ranged from 1.15% to 9.16% dry weight. Spectral data were recorded with a spectrometer equipped with an attenuated total reflectance accessory (Agilent Cary 630) and analyzed using partial least squares regression. The spectral region from 1600 cm(-1) to 1000 cm(-1) gave unique polysaccharide bands associated with carboxyl, acetyl, and glycosidic linkages of sugar residues. The multivariate analysis gave excellent performance parameters with correlation coefficient (r2) of 0.95 and standard error of cross validation of 0.6% WSC. We found that IR spectroscopy provides a rapid and accurate technique for analyzing WSC concentrations and offers primatologists many advantages over wet chemistry methods for analyzing nutritional composition. (C) 2016 Wiley Periodicals, Inc.