Determining the optimal age for recording the retinal vascular pattern image of lambs

Abstract

Newborn Ripollesa lambs (n = 143) were used to assess the optimal age at which the vascular pattern of the retina can be used as a reference for identification and traceability. Retinal images from both eyes were recorded from birth to yearling (d 1, 8, 30, 82, 180, and 388 of age) in duplicate (2,534 images) using a digital camera specially designed for livestock (Optibrand, Fort Collins, CO). Intra- and inter-age image comparisons (9,316 pairs of images) were carried out, and matching score (MS) was used as the exclusion criterion of lamb identity (MS <70). Retinal images were used for verifying the identity of live lambs of 4 Spanish commercial categories: milk-fed lambs (MF; "lechal," 1 mo of age and <12 kg of BW, n = 136); light fattened lambs (LF; "recental," 3 mo of age and ~25 kg of BW, n = 134); recruited replacement lambs (RR; "ovino mayor," 6 mo of age and ~35 kg of BW, n = 59); and yearling replacement lambs (YR; >12 mo of age and ~50 kg of BW, n = 25). Values of MS were treated with a model based on the 1-inflated bivariate beta distribution, and treated data were compared by using a likelihood ratio test. Intra-age image comparisons showed that average MS and percentage of images with MS ≥70 increased (P < 0.05) with lamb age from d 1 to 82 (81.0 ± 1.0 to 95.9 ± 0.4, and 75.8 to 100%, respectively) and reached a plateau thereafter. Inter-age retinal image comparisons for MF lambs at 30 d showed that the percentage of images with MS ≥70 was greater when reference images at 8 d were used instead of 1-d images (94.8 vs. 87.4%, respectively; P < 0.05). In LF lambs, 30-d retinal images matched better than those at 8 d of age (99.6 vs. 93.6%, respectively; P < 0.05). Regarding RR and YR lambs, the 82-d images allowed the confirmation of 100% lamb identity, whereas 8-d images showed a smaller (P < 0.05) percentage of matching (87.9 and 89.8%, respectively, for RR and YR lambs; P > 0.05); no differences were detected for 30-d images (97.4 and 98.0%, respectively, for RR and YR lambs; P > 0.05). Total percentage of matching was achieved when images were obtained from older lambs (180 and 388 d). In conclusion, retinal imaging was a useful tool for verifying the identity and auditing the traceability of live lambs from suckling to yearling. Matching scores were satisfactory when the reference retinal images were obtained from 1-mo-old or older lambs.