3.3. Genetic relationship with baldness
Crown baldness was present in 8/15 King Island scrubtits, including in both sexes and all three subpopulations, but none of 55 Tasmanian scrubtits. Logistic regression across all genotyped individuals showed a weak negative relationship between multi-locus heterozygosity (MLH) and the probability of crown baldness occurrence (β = -22.152. se = 8.788, z = -2.52, p = 0.017, Figure 6a). However, this relationship disappeared when the analysis was restricted to only King Island scrubtits (β = 11.56, se = 21.8, z = 0.53, p = 0.60, Figure 6b), suggesting baldness is more likely to occur in the King Island scrubtit population than the Tasmanian scrubtit population, but baldness in King Island scrubtits is not linked to (relatively) low genome-wide MLH.
After accounting for population structure (k = 3) and adjusting the p -values, the LFMM identified six loci that were significantly associated with baldness (Figure S11). Of the six candidate SNPs, three were genic and three were non-genic (Table S7). Three SNPs were located on the same assembled contig, including one in the DOCK11 gene involved in regulation of filopodium assembly. Filopodia have been implicated in feather follicle formation and feather branching in chickens (Cheng et al., 2018), suggesting a possible role of this gene in baldness.