COMPARISON OF RUMEN AND FECAL ANALYSIS TO ESTIMATE MOOSE DIETS

Large ungulate management and research projects often require estimate s of diet composition, and indirect methods are often the only feasibl e alternatives. We compared diet estimates based on moose (Alces alces gigas) rumen and fecal samples taken from the same animal on the Copp er River Delta, Alaska. Twenty-five animals were sampled, the majority of which were killed by hunters during September. Three procedures we re used, a standard microhistological technique on both rumen and feca l samples and a macroscopic technique on rumen samples. The macroscopi c analysis of rumen samples identified contents to species and specifi c plant parts. However, the results of macroscopic rumen analysis can be affected by seasonal changes in rumen particle size. Microhistologi cal analysis of both rumen and Fecal samples could not differentiate a mong willow species (Salis spp.), the dominant food item, nor could pl ant parts be distinguished. These are major shortcomings of this proce dure. However, when willows were identified only to genus, diets estim ated by the 3 procedures differed (P < 0.001) only for minor items (<5 % of the diet), or items that were unidentifiable microhistologically (i.e., wood fragments, catkins). Quantifying rumen and Fecal winter sa mples was difficult. Each procedure had shortcomings, and a combinatio n of procedures may have to be used, particularly if study objectives require large sample sizes from a variety of individual animals. Knowl edge of the strengths and weaknesses of diet estimation procedures all ows investigators to assess methods in relation to project objectives and when evaluating the results of other studies.