Optimising sampling techniques and estimating sampling variance of fleece quality attributes in alpacas

Huacaya and Suri alpacas (n = 120) of varying age, live weight (LWT) and sex (female, male) were selected randomly from four farms in southern Australia. At shearing, fleeces were divided into four components: saddle (S), neck (N), pieces (P; front and back legs, belly, apron) and the midside sample (MS). Components were weighed, sampled using the grid sampling technique and fleece attributes measured: clean washing yield (CWY), mean fibre diameter (MFD), coefficient of variation of the MFD (CV(D)), incidence of medullated fibres (Med), mean medullated fibre diameter (MedMFD) and coefficient of variation of the MedMFD (MedCV(D)). The MS and saddle grid sample (SGS) were used to create models to predict the fleece attribute of the total fleece (TF), saddle and neck fibre. For each fleece attribute MS had lower values than SGS and TF (P < 0.005) and SGS, except for CWY, had lower values than the P and TF (P < 0.005). The means were: MFD MS 27.5 mum, S 28.8 mum, N 28.7 mum, P 37.6 mum, TF 31.2 mum; CV(D) MS 24.3%, S 27.0%, N 28.6%, P 30.6%, TF 28.1%; CWY MS 90.2%, S 91.4%, N 88.9%, P 92.8%; Med 24.4%, S 33.1%, P 44.5%, TF 35.2%; MedNLFD MS 32.7 mum, S 34.4 mum, P 41.1 mum, TF 36.0 mum; MedCV(D) MS 19.4%, S 22.3%, P 25.9%, TF 23.4%. The MS was found to be an appropriate sample from which to predict the MFD and CWY CV(D) was only satisfactorily predicted by the SGS (r = 0.88), with the exception of the neck fleece, for which neither the MS nor SGS could provide an accurate predictive model. The MS did not sufficiently account for the variation in Med (r = 0.73-0.79). The SGS gave accurate prediction of Med (r = 0.98). Sex effects were detected in models for TFMFD, NMFD and TFCV(D). LWT effects were detected in models for NMFD, NCV(D) and TFMedMFD. SGS often gave a more accurate prediction of a fleece attribute but it requires the removal of the entire fleece, whereas MS can be removed by shearing a small area or can be removed during shearing with a minimum of effort. Sampling variance for SGS was generally two to four times greater than the sampling variance for MS with the 95% confidence limits (CLs) for SGS being about double those of MS for most parameters except for clean washing yield (CWY) which were similar. Sampling variance for the incidence of medullated fibres in SGS was very high. The large 95% CL for all the tested fibre attributes indicate that alpaca breeders and advisors need to consider taking suitable duplicate measurements and other precautions during breeding and animal selling programs. Crown Copyright (C) 2002 Published by Elsevier Science B.V. All rights reserved.