Table 1: Genetic diversity and eco-genetic structure
within diploid and allopolyploid Aegilops species. For each
species, Ng individuals were genotyped at plastid and
nuclear loci to estimate allelic richness (AR), observed heterozygosity
(Ho), genetic diversity corrected for sample size (h ) and with
ordered alleles (v ). The eco-genetic structure was assessed
through (partial) Mantel tests associating genetic (G), climatic (E) and
geographical (Geo) distances among individuals within each species and
decomposed into significant isolation-by-distance (G ~
Geo) and isolation-by-environment (G ~ E |
Geo). Range filling represents the fraction of suitable areas based on
the climatic niche that is occupied by the species within the study
area.
† Genome composition of each species according
to van Slageren (1994), with maternal progenitor underlined
‡ Associations tested through 20000
permutations with p-values as *** < 0.001, ** <
0.01, * < 0.05 and ns for non-significant and confirmed
through Redundancy Analysis whose significance is indicated in between
brackets.
Fig. 1: Genetic and ecological additivity of diploid progenitors
as a null hypothesis to analyze micro-evolution of allopolyploid
species. Multilocus genotypes are shown for diploid (stars) and
allopolyploid (diamonds) individuals over an adaptive landscape
presenting fitness as grey isolines. On the left, ecological niches of
diploid species are overlaid as shaded areas delimiting populations
existing across the environmental space. Allopolyploids combine
divergent suites of genes (shown as – or + between brackets) from
diploid species (AoAo and
BoBo) and are therefore expected to
present genetic additivity
(AeAeBeBe).
Fixed heterozygosity for genes underlying the ecological niche of each
diploid progenitor is expected to provide positive population growth
over combined ranges of environmental conditions tolerated by progenitor
species. As a null hypothesis, allopolyploid speciation therefore
produces new species presenting the ecological additivity of their
progenitors, with an observed niche
(AoAoBoBo)
similar to the expected niche (i.e.
AeAeBeBe).
On the right, main scenarios of possible ecological niche evolution
during the establishment of allopolyploids are presented. Ecological
niches of allopolyploids can retain the additivity of niches from their
diploid progenitors. Significant deviation from such expected additivity
is however characteristic of niche shifts in allopolyploids. Niche
contraction occurs when allopolyploids do not outcompete diploid
progenitors and occupy a subset of their niches. In contrast, niche
expansion beyond the combined range of conditions tolerated by diploid
progenitors is coherent with evolutionary novelties commonly postulated
to promote allopolyploid establishment in face of their competition.
Analyses relying on such expected additivity of diploids adequately test
for multivariate niche conservatism vs shift in allopolyploids
and, circumventing ambiguous pairwise comparisons with divergent
progenitors, transcend diagnostics based on niche centroids (crosses)
and/or niche breadths (braces along univariate ecological gradients).
Verifiable conclusions are accordingly reached regarding niche
conservatism vs novelty in allopolyploids responding to variable
competition from diploids in space and time.
Figure 2: Evolution of Aegilops allopolyploid species
following the merging of diploid progenitors. A. Phylogenetic
relationships within and among four diploids species Ae. tauschii(DD), Ae. caudata (CC), Ae. umbellulata (UU) and Ae.
comosa (MM), and four derived allopolyploid species Ae.
cylindrica (DDCC), Ae. triuncialis (UUCC), Ae. geniculata(UUMM) and Ae. crassa (DDMM with tetraploids, 4x, and hexaploids,
6x). Origins of allopolyploid species were inferred through independent
multi-labelled trees based on 30 nuclear loci anchored in the phylogeny
of diploid wheats (Fig. S3). Divergence from maternal (plain line) and
paternal (dashed line) progenitors are reported with corresponding
dates. Tips present accessions from genetic clusters identified with the
same loci, labelled according to Fig. 3. Maternal progenitor species of
allopolyploids, identified using plastid loci, are underlined (Fig. S2).B. Nuclear genomic composition of the four allopolyploid
species estimated as the proportion of conserved (blue), derived
(orange) and lost (grey) haplotypes across loci as compared to those
segregating within their diploid progenitors (Table S2).
Figure 3: Spatial genetic structure within four diploid and four
derived allopolyploid Aegilops species. Genotyped accessions
(circles) are assigned to genetic clusters inferred by STRUCTURE within
each species, colored according to the stacked barplots underneath and
approximately delimited across distribution ranges. Contact zones
between genetic clusters showing admixed accessions are indicated by red
stripes. Labels are otherwise as in Fig. 2.
Figure 4: Climatic niche dynamics of allopolyploidAegilops species. Climatic niche of each Aegilops species
in the common climatic space (100% and 50% indicated as solid and
dashed contour lines) along the first two principal components based on
20 bioclimatic factors shown in the central correlation circle
(abbreviations in Fig. S6). The observed niche of each polyploid species
is compared to the expected additivity of climatic niches modeled from
both its diploid progenitors (blue + green). The fraction of the
expected niche effectively occupied by the allopolyploid species
characterizes niche stability (blue), whereas the unoccupied fraction is
referred to as niche contraction (green). Climatic space occupied by the
allopolyploids while not predicted by the additivity of diploid
progenitors indicates niche expansion (red). For each allopolyploid,
niche centroids of diploid progenitors are indicated as orange stars
labelled with their genome composition. Labels are otherwise as in Fig.
2.