358
since grouped all samples of C. angustifolia, while the other cluster brought
together samples of C. balansae and C. saltensis (Cbal-Csal hereafter). PCoA had
indicated analogous results. Next, we carried out analyses on Structure for
C. angustifolia and Cbal-Csal, independently. Both analyses showed two clusters
(best K = 2).
To display the latitudinal distribution of the genetic clusters estimated by both
independent Bayesian analyses, we constructed a first map representing each of
the 20 populations (Table 13.1) with the proportion of assignment to each genetic
cluster (Fig. 13.4a). In this map, pie charts at sampled locations show the proportion of membership of populations to C. angustifolia, C. balansae, and C. saltensis. The two clusters of C. angustifolia (cluster I depicted in violet and cluster II
depicted in lilac; Fig. 13.4a) were broadly represented among the populations,
which suggests that populations behaved as metapopulations associated with latitudinal and altitudinal distributions across the Yungas (Table 13.1). There was a
tendency to find populations located at higher latitudes and lower altitudes (Los
Pizarros, La Florida Provincial Park, El Siambón, Sunchal, Metán, and El Arenal)
with a higher percentage of membership in cluster I (74–100%). Populations
located towards lower latitudes and higher altitudes (La Ramada, Calilegua
National Park, San Andrés, and Baritú National Park) showed substantial increase
in the percentage of membership in cluster II (43–100%) while exhibiting varying levels of admixture between clusters I and II (Fig. 13.4a); these results are in
agreement with those presented above concerning patterns of genetic diversity
found in C. angustifolia and reported by Inza et al. (2012) when this species was
addressed independently.
To a certain extent, the two clusters of Cbal-Csal were congruent with
taxonomical- delimited species: cluster I (depicted in red; Fig. 13.4a) brought
together samples of C. balansae, while cluster II (depicted in green; Fig. 13.4a)
brought together samples of C. saltensis. Accordingly, Río Seco FF, Río Seco
FSB, Yuto, and Apolinario Saravia populations constitute genetically pure populations of C. balansae. In contrast, four populations that were defined a priori as
C. balansae (Calilegua National Park, Acambuco, San Andrés, and Pintascayo)
contained varying levels of admixture between clusters I and II (Fig. 13.4a). These
results suggest a hybrid or introgressed origin for these four populations. The two
populations of C. saltensis showed percentage of assignment in cluster II >97%
for most of the samples, with the exception of four samples belonging to Calilegua
National Park and San Andrés populations, which exhibited levels of admixture of
clusters I and II >45%. In addition, Pintascayo – a population sampled as C. balansae according to morphological characteristics – showed percentage of membership >83% in cluster II, the C. saltensis cluster. Actually, 4 out of 11 samples of
Pintascayo showed percentage of membership ≥99% in the C. saltensis cluster,
which suggests that these four samples detained a large contribution of the C. saltensis genome.
N. Zelener et al.
since grouped all samples of C. angustifolia, while the other cluster brought
together samples of C. balansae and C. saltensis (Cbal-Csal hereafter). PCoA had
indicated analogous results. Next, we carried out analyses on Structure for
C. angustifolia and Cbal-Csal, independently. Both analyses showed two clusters
(best K = 2).
To display the latitudinal distribution of the genetic clusters estimated by both
independent Bayesian analyses, we constructed a first map representing each of
the 20 populations (Table 13.1) with the proportion of assignment to each genetic
cluster (Fig. 13.4a). In this map, pie charts at sampled locations show the proportion of membership of populations to C. angustifolia, C. balansae, and C. saltensis. The two clusters of C. angustifolia (cluster I depicted in violet and cluster II
depicted in lilac; Fig. 13.4a) were broadly represented among the populations,
which suggests that populations behaved as metapopulations associated with latitudinal and altitudinal distributions across the Yungas (Table 13.1). There was a
tendency to find populations located at higher latitudes and lower altitudes (Los
Pizarros, La Florida Provincial Park, El Siambón, Sunchal, Metán, and El Arenal)
with a higher percentage of membership in cluster I (74–100%). Populations
located towards lower latitudes and higher altitudes (La Ramada, Calilegua
National Park, San Andrés, and Baritú National Park) showed substantial increase
in the percentage of membership in cluster II (43–100%) while exhibiting varying levels of admixture between clusters I and II (Fig. 13.4a); these results are in
agreement with those presented above concerning patterns of genetic diversity
found in C. angustifolia and reported by Inza et al. (2012) when this species was
addressed independently.
To a certain extent, the two clusters of Cbal-Csal were congruent with
taxonomical- delimited species: cluster I (depicted in red; Fig. 13.4a) brought
together samples of C. balansae, while cluster II (depicted in green; Fig. 13.4a)
brought together samples of C. saltensis. Accordingly, Río Seco FF, Río Seco
FSB, Yuto, and Apolinario Saravia populations constitute genetically pure populations of C. balansae. In contrast, four populations that were defined a priori as
C. balansae (Calilegua National Park, Acambuco, San Andrés, and Pintascayo)
contained varying levels of admixture between clusters I and II (Fig. 13.4a). These
results suggest a hybrid or introgressed origin for these four populations. The two
populations of C. saltensis showed percentage of assignment in cluster II >97%
for most of the samples, with the exception of four samples belonging to Calilegua
National Park and San Andrés populations, which exhibited levels of admixture of
clusters I and II >45%. In addition, Pintascayo – a population sampled as C. balansae according to morphological characteristics – showed percentage of membership >83% in cluster II, the C. saltensis cluster. Actually, 4 out of 11 samples of
Pintascayo showed percentage of membership ≥99% in the C. saltensis cluster,
which suggests that these four samples detained a large contribution of the C. saltensis genome.
N. Zelener et al.