Morphological and Chemical Diversity and Antioxidant Capacity of the Service Tree (Sorbus domestica L.) Fruits from Two Eco-Geographical Regions

Abstract

Service tree, Sorbus domestica L., is a rare and neglected wild fruit tree species of southern and central Europe. Being distributed in different eco-geographical regions, with fragmented and low-density populations, S. domestica represents an interesting model case for investigating patterns of within- and between-population diversity at geographical and environmental scales. This study aimed to analyze the proximate composition, antioxidant activity, and morphometric fruit characteristics. We examined the diversity and population divergences of 49 S. domestica individuals originating from seven populations across continental and Mediterranean eco-geographical regions. In addition, tests of isolation by distance and environment were performed to detect the magnitude of divergence explained by geographic and environmental variables. Significant differences between the studied populations were found in almost all of the studied morphometric and chemical fruit characteristics. The studied service tree populations were characterized by high phenotypic variation despite the low number of trees per population. Model-based population structure analysis using morphometric and chemical fruit characteristics revealed three groups of service tree populations. We concluded that non-effective pollen and seed dispersal along with genetic drift and specific environmental factors resulted in a distinct phenotype with a specific chemical composition in the isolated island population. In addition, a pattern of isolation by the environment was revealed. We infer that morphological and chemical differences between the studied populations in the true service tree from different eco-geographical regions were mediated by adaptation to the specific environmental conditions.

Keywords: adaptation; chemical analysis; fruit tree species; island populations; isolation by distance; isolation by environment; morphometric analysis; population variability.

Figure 1

Results of the multivariate statistical methods for fruit and seed morphometric analysis and locations of the seven sampled Sorbus domestica populations. (A) Geographical distribution of three groups of populations detected from K-means clustering (the proportions of the ancestry of each population in each of the defined clusters are color-coded: cluster A– green, cluster B–yellow, cluster C–blue); (B) Hierarchical tree dendrogram; (C) Barplot with posterior probabilities of classification of each individual into each group from the results of the classification discriminant analysis. Acronyms of populations: P01—Psunj; P02—Tounj; P03—Istria; P04—Novi Vinodolski; P05—Split; P06—Brač; P07—Konavle.

Figure 2

Results of the multivariate statistical methods for fruit chemical analysis and locations of the seven sampled Sorbus domestica populations. (A) Geographical distribution of three groups of populations detected from K-means clustering (the proportions of the ancestry of each population in each of the defined clusters are color-coded: cluster A– green, cluster B–yellow, cluster C–blue); (B) Hierarchical tree dendrogram; (C) Barplot with posterior probabilities of classification of each individual into each group from the results of the classification discriminant analysis. Acronyms of populations: P01—Psunj; P02—Tounj; P03—Istria; P04—Novi Vinodolski; P05—Split; P06—Brač; P07—Konavle.

Figure 3

The boundaries obtained with Monmoniers’s maximum difference algorithm of Euclidean distances among populations revealed by program BARRIER. (A) The first three boundaries (a–c) for morphometric variation; and (B) the first two boundaries (a and b) for chemical variation. The thickness of barriers is related to their importance. Acronyms of populations: P01—Psunj; P02—Tounj; P03—Istria; P04—Novi Vinodolski; P05—Split; P06—Brač; P07—Konavle.

Figure 4

Biplot of the principal component analysis based on ten fruit and seed morphometric characteristics in studied Sorbus domestica populations. Acronyms of populations: P01 (Psunj), P02 (Tounj), P03 (Istria), P04 (Novi Vinodolski), P05 (Split), P06 (Brač), and P07 (Konavle). Each point represents a single tree, with a color corresponding to cluster from K-means clustering method and origin of population: continental populations—cluster A (green); island population Brač—cluster B (yellow); and coastal Mediterranean populations—cluster C (blue).

Figure 5

Biplot of the principal component analysis based on ten chemical traits in studied Sorbus domestica populations. Acronyms of populations: P01 (Psunj), P02 (Tounj), P03 (Istria), P04 (Novi Vinodolski), P05 (Split), P06 (Brač), and P07 (Konavle). Each point represents a single tree, with a color corresponding to cluster from K-means clustering method and origin of population: continental populations—cluster A (green); island population Brač—cluster B (yellow); and coastal Mediterranean populations—cluster C (blue).

Figure 6

Isolation-by-distance and Isolation-by-environmental distance. Scatter plots of simple Mantel tests showing the relationships between (A) geographic and morphological distances (r = –0.0729, p = 0.58), (B) environmental and morphological distances (r = 0.3814, p = 0.05), (C) geographic and chemical distances (r = –0.0848, p = 0.47), (D) environmental and chemical distances (r = 0.0706, p = 0.37), and (E) morphological and chemical distances (r = 0.6129, p = 0.02).