Multiple regression analyses of traits contributing to the skin value of South African ostriches

Abstract

The contribution of various skin traits to determining skin value was investigated in South African slaughter ostriches. Processed skins (n = 747) from the South African Black ostrich breed slaughtered between 2003 and 2019 and aged 282–588 d were used. Tannery data (skin size, skin grade, and quill value) were used to calculate individual skin price based on the prices paid to producers in October 2020. The PROC REG model of the SAS statistical package was used for stepwise multiple regression analyses to investigate the influence of various skin traits on skin price. Traits included skin size, skin grade, quill value, nodule size score, nodule shape score, hair follicle score, and pitting score. Pearson correlations coefficient (r) among variables were calculated using Minitab software. Skin grade accounted for most of the variation in the monetary value of individual skins at 76.66%, followed by skin size and quill value with respective contributions of 5.12% and 1.42% to the cumulative R2 of the final model. When quill value was omitted from the model, nodule shape contributed 0.26% to the variance in skin value. When skin grade was not modelled, skin size accounted for most of the variation in skin price; pitting score contributed only 0.62% to the cumulative R². Pearson correlations confirmed that skin grade was the dominant regression variable for skin value (r = -0.88); quill value and skin size were significant, but less important. No strong correlations were observed between skin grade and quill value. Therefore, it seems that skin grade is the dominant force determining the price currently paid to ostrich farmers in South Africa.