Why More Leaflets? The Role of Natural Selection in Shaping the Spatial Pattern of Leaf-Shape Variation in Oxytropis diversifolia (Fabaceae) and Two Close Relatives

Abstract

Leaf shape exhibits tremendous diversity in angiosperms. It has long been argued that leaf shape can affect major physiological and ecological properties of plants and thus is likely to be adaptive, but the evolutionary evidence is still scarce. Oxytropis diversifolia (Fabaceae) is polymorphic for leaf shape (1 leaflet, 1–3 leaflets, and 3 leaflets) and exhibits clinal variation in steppes of Nei Mongol, China. With two close relatives predominantly fixed for one phenotype as comparison (Oxytropis neimonggolica with 1 leaflet and Oxytropis leptophylla with 5–13 leaflets), we used a comprehensive cline-fitting approach to assess the role of natural selection in shaping the spatial pattern of leaf-shape variation in this system. For 551 individuals sampled from 22 populations, we quantified leaf-morphological differentiation, evaluated patterns of neutral genetic variation using five chloroplast DNA intergenic regions and 11 nuclear microsatellite loci, and performed microhabitat and macroclimatic-association analyses. We found that 1-leaflet proportions in O. diversifolia populations significantly increased from west to east, and three phenotypes also differed in leaflet-blade size. However, compared with the other two species, populations of O. diversifolia showed little neutral genetic differentiation, and no population structure was detected at either marker. We further revealed that the leaf-shape cline could largely be explained by three macroclimatic variables, with leaflet number decreasing and leaflet-blade size increasing with annual precipitation and showing the reverse trends with temperature seasonality and isothermality. Our results suggest that spatially varying abiotic environmental factors contribute to shape the leaf-shape cline in O. diversifolia, while the interspecific pattern may be due to both local adaptation and historical events.

Tension zone trapped by exogenous cline: Analysis of a narrow hybrid zone between two parapatricOxytropisspecies (Fabaceae)

“…Significant heterozygote deficits were observed at 1 to 10 loci in each population, and in total 138 out of 209 locus‐specific tests was significant (Table S6). The systematic pattern of heterozygote deficiency found in O. diversifolia and O. leptophylla is likely due to the self‐compatibility and/or inbreeding (Wang et al, 2021), while in the hybrid zone, the pattern is not significantly different from that of parental species ( p = .38; Table 2A).…”

“…In this study, the two Oxytropis species are separated by a narrow hybrid zone that can be maintained by a balance between dispersal and divergent selection. This inference is based on the shape of the phenotypic and genetic transition (i.e., sigmoidal clines), the narrow width relative to the dispersal ability (Wang et al, 2021), and the elevated variance at the center (Barton & Hewitt, 1985; Stankowski et al, 2021). However, solely from the sigmoidal cline, it is difficult to distinguish between selection against hybrids (i.e., “endogenous” selection) and adaptation to different environments (i.e., “exogenous” selection), because endogenous and exogenous selection can generate clines of similar shape (Barton & Gale, 1993; Barton & Hewitt, 1989; Kruuk et al, 1999).…”

“…Introgression is emerging as an important source of novel genetic variation, alongside with standing variation and mutation, and it can be adaptive when such introgressed alleles are maintained by natural selection (termed “adaptive introgression”, Abbott et al, 2013; Suarez‐Gonzalez et al, 2018). A previous study revealed that the clinal variation in leaf shape in O. diversifolia can be attribute to natural selection imposed by spatially varying macroclimatic factors (Wang et al, 2021). It would be interesting to test whether leaf‐shape polymorphism in O. diversifolia was obtained through hybridization and introgression from O. leptophylla to the 1‐leaflet phenotype.…”

“…For the two populations in the hybrid zone, 11 nuclear microsatellite loci (N745892, N145635, N2724893, N2717495, N178451, N161850, N49251, N350553, N935993, N2528349, and N2697375; Table S2) were scored across all 84 individuals according to the methods described by Wang et al (2018). Five cpDNA intergenic regions ( trn T‐ psb D, pet N‐ psb M, trn S‐ trn G, psb E‐ pet L, and rpl 16 intron; Table S3) were screened for 43 and 21 individuals in G and HUA, respectively, according to the methods described by Wang et al (2021). All of the sequences were deposited in GenBank under accession numbers MW511796 – MW512115.…”

“…This survey was used to estimate population size (i.e., the total number of individuals) and population density (population size divided by total area). Silica‐gel dried leaf samples were collected, using methods described in Wang et al (2021). In total, 507 individuals were collected, of which 321 were from O. diversifolia populations, 102 were from O. leptophylla populations, and 84 were from the hybrid zone.…”