Limits...
Ecological niche partitioning between Anopheles gambiae molecular forms in Cameroon: the ecological side of speciation.

Simard F, Ayala D, Kamdem GC, Pombi M, Etouna J, Ose K, Fotsing JM, Fontenille D, Besansky NJ, Costantini C - BMC Ecol. (2009)

Bottom Line: Population structure analysis identified three chromosomal clusters, each containing a mixture of M and S specimens.Rather, they are involved in ecological specialization to a similar extent in both genetic backgrounds, and most probably predated lineage splitting between molecular forms.When such mutations occur in portions of the genome where recombination is suppressed, such as the pericentromeric regions known as speciation islands in An. gambiae, they would contribute further to the development of reproductive isolation.

View Article: PubMed Central - HTML - PubMed

Affiliation: Laboratoire de Lutte contre les Insectes Nuisibles, Institut de Recherche pour le Développement, UR016, 911 Av. Agropolis, Cedex 5, Montpellier, France. frederic.simard@ird.fr

ABSTRACT

Background: Speciation among members of the Anopheles gambiae complex is thought to be promoted by disruptive selection and ecological divergence acting on sets of adaptation genes protected from recombination by polymorphic paracentric chromosomal inversions. However, shared chromosomal polymorphisms between the M and S molecular forms of An. gambiae and insufficient information about their relationship with ecological divergence challenge this view. We used Geographic Information Systems, Ecological Niche Factor Analysis, and Bayesian multilocus genetic clustering to explore the nature and extent of ecological and chromosomal differentiation of M and S across all the biogeographic domains of Cameroon in Central Africa, in order to understand the role of chromosomal arrangements in ecological specialisation within and among molecular forms.

Results: Species distribution modelling with presence-only data revealed differences in the ecological niche of both molecular forms and the sibling species, An. arabiensis. The fundamental environmental envelope of the two molecular forms, however, overlapped to a large extent in the rainforest, where they occurred in sympatry. The S form had the greatest niche breadth of all three taxa, whereas An. arabiensis and the M form had the smallest niche overlap. Correspondence analysis of M and S karyotypes confirmed that molecular forms shared similar combinations of chromosomal inversion arrangements in response to the eco-climatic gradient defining the main biogeographic domains occurring across Cameroon. Savanna karyotypes of M and S, however, segregated along the smaller-scale environmental gradient defined by the second ordination axis. Population structure analysis identified three chromosomal clusters, each containing a mixture of M and S specimens. In both M and S, alternative karyotypes were segregating in contrasted environments, in agreement with a strong ecological adaptive value of chromosomal inversions.

Conclusion: Our data suggest that inversions on the second chromosome of An. gambiae are not causal to the evolution of reproductive isolation between the M and S forms. Rather, they are involved in ecological specialization to a similar extent in both genetic backgrounds, and most probably predated lineage splitting between molecular forms. However, because chromosome-2 inversions promote ecological divergence, resulting in spatial and/or temporal isolation between ecotypes, they might favour mutations in other ecologically significant genes to accumulate in unlinked chromosomal regions. When such mutations occur in portions of the genome where recombination is suppressed, such as the pericentromeric regions known as speciation islands in An. gambiae, they would contribute further to the development of reproductive isolation.

Show MeSH

Related in: MedlinePlus

Assignment of An. gambiae karyotypes by multilocus genetic clustering. Genetic cluster analysis using STRUCTURE based on chromosome-2 karyotypes of the M and S form of An. gambiae in Cameroon. Each individual mosquito is represented by a thin horizontal line divided into K = 3 (most likely value of K) coloured segments that represent the individual's estimated membership fraction to each of the K = 3 clusters. A. Output based on chromosomal inversions stratified by decreasing latitude of the specimen's sampling locality (top = North to bottom = South); biogeographic domains are given on the left (cf. Figure 1). B. Same as in A, with specimens sorted according to their molecular form status. A black line separates specimens of the S (above) and M (below) molecular forms and within each form, specimens are sorted by decreasing latitude of the collection locality.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2698860&req=5

Figure 6: Assignment of An. gambiae karyotypes by multilocus genetic clustering. Genetic cluster analysis using STRUCTURE based on chromosome-2 karyotypes of the M and S form of An. gambiae in Cameroon. Each individual mosquito is represented by a thin horizontal line divided into K = 3 (most likely value of K) coloured segments that represent the individual's estimated membership fraction to each of the K = 3 clusters. A. Output based on chromosomal inversions stratified by decreasing latitude of the specimen's sampling locality (top = North to bottom = South); biogeographic domains are given on the left (cf. Figure 1). B. Same as in A, with specimens sorted according to their molecular form status. A black line separates specimens of the S (above) and M (below) molecular forms and within each form, specimens are sorted by decreasing latitude of the collection locality.

Mentions: As a first approach, each chromosomal inversion was considered as a bi-allelic locus; subsequently, due to overlap and linkage between some inversions, we repeated the analysis identifying loci by chromosomal inversion systems as defined by [1,23,25]. A few specimens carrying the inversions 2Rj and 2Rbk (N = 7, Table 3) were omitted in the latter analysis, without detectable changes in the output. Results by the two approaches were concordant: both detected three genetic clusters in the dataset (Additional file 5). All clusters were composed of specimens belonging to both the M and S forms, suggesting that alternative chromosomal arrangements did not segregate at contrasting frequencies between molecular forms (Additional file 6). The first cluster (Cluster 1, identified by the green colour in Figure 6) corresponded essentially to the FOREST chromosomal form [1]. It was composed mainly of specimens carrying the monomorphic standard arrangement at all inversion systems on chromosome-2 (karyotypes 0000000), with additional low-level polymorphism for inversions 2Rb and 2La only in the S form (Additional file 6). As shown in Figure 6A, where individuals are arranged according to biogeographic domain and latitude of collection, specimens belonging to this cluster were mainly collected in the rainforest area (Central Plateau and Atlantic Coast). Mosquitoes collected in savanna biotopes were partitioned into two distinct genetic clusters based on their karyotype (identified by the yellow and red colours in Figure 6), without further clustering by geography with respect to latitude. A clinal pattern of relative abundance of the two clusters, however, was somewhat apparent (Figure 6). Cluster 2 (in yellow) was mainly composed of specimens polymorphic for the 2Rb and 2La inversions only (Additional file 6), with a high frequency of inverted homozygotes. It closely resembles the cluster grouping together most S form specimens sampled in the dry savannas of Burkina Faso [45], and typically corresponds to the SAVANNA chromosomal form of Coluzzi et al. [1], which is widespread throughout Africa. This cluster intergraded with Cluster 1 at the ecotone between the humid southern savannas and the rainforest, as can be seen in Figure 6A, where some specimens with mixed ancestry (i.e. individuals whose probabilities of membership to Cluster 1 and 2 are substantial; these individuals are identified by their bars being almost equally subdivided in green and yellow) were apparent in this area, especially at the margins of the highlands. Intergradation between Clusters 1 and 2 was more obvious in S than in M, probably because of the sparseness of the savanna populations of the latter taxon (Figure 6B). Finally, Cluster 3 (identified by red bars in Figure 6) was characterized by a high level of polymorphism for each inversion system (Additional file 6). This cluster comprised individuals carrying the 2Rbc and 2Rd arrangements, in combination or not with other inversions. In our samples, the frequency of mosquitoes belonging to this cluster increased when moving northwards (Figure 6). All M and S individuals from Cluster 3 carried karyotypes that would put them under the (polymorphic) SAVANNA chromosomal form described by Coluzzi et al. [1] from collections carried out in Mali [25] and Nigeria [23]. However, contrary to expectations [1,74] and despite considerable geographical overlap and sympatry throughout the savanna areas of Cameroon (Figure 6A), there was lower occurrence of mixed ancestry between the two SAVANNA clusters (2 and 3) than between each of them and cluster 1 (FOREST). To quantify the degree of intergradation between clusters, we classified any mosquito with >10% probability to belong to more than one cluster as an admixed individual (due to hybridization or shared ancestry). We detected 166 admixed individuals between Clusters 1 and 2 (8.4% of the total number of specimens included in the analysis), 81 between Clusters 1 and 3 (4.1%) and 51 (2.6%) between Clusters 2 and 3. Among these, 7 specimens (0.4%) could be assigned to any of the three clusters with probability >10%. Overall, these results indicate that the boundaries between chromosomally-defined clusters were permeable and that they assorted independently of molecular form status. Because a wealth of evidence substantiates the ecological and reproductive unity of the molecular forms, it is likely that the distribution of chromosomal arrangements that we observed merely reflected directional selection acting on different karyotypes in alternative environments, within both M and S.


Ecological niche partitioning between Anopheles gambiae molecular forms in Cameroon: the ecological side of speciation.

Simard F, Ayala D, Kamdem GC, Pombi M, Etouna J, Ose K, Fotsing JM, Fontenille D, Besansky NJ, Costantini C - BMC Ecol. (2009)

Assignment of An. gambiae karyotypes by multilocus genetic clustering. Genetic cluster analysis using STRUCTURE based on chromosome-2 karyotypes of the M and S form of An. gambiae in Cameroon. Each individual mosquito is represented by a thin horizontal line divided into K = 3 (most likely value of K) coloured segments that represent the individual's estimated membership fraction to each of the K = 3 clusters. A. Output based on chromosomal inversions stratified by decreasing latitude of the specimen's sampling locality (top = North to bottom = South); biogeographic domains are given on the left (cf. Figure 1). B. Same as in A, with specimens sorted according to their molecular form status. A black line separates specimens of the S (above) and M (below) molecular forms and within each form, specimens are sorted by decreasing latitude of the collection locality.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2698860&req=5

Figure 6: Assignment of An. gambiae karyotypes by multilocus genetic clustering. Genetic cluster analysis using STRUCTURE based on chromosome-2 karyotypes of the M and S form of An. gambiae in Cameroon. Each individual mosquito is represented by a thin horizontal line divided into K = 3 (most likely value of K) coloured segments that represent the individual's estimated membership fraction to each of the K = 3 clusters. A. Output based on chromosomal inversions stratified by decreasing latitude of the specimen's sampling locality (top = North to bottom = South); biogeographic domains are given on the left (cf. Figure 1). B. Same as in A, with specimens sorted according to their molecular form status. A black line separates specimens of the S (above) and M (below) molecular forms and within each form, specimens are sorted by decreasing latitude of the collection locality.
Mentions: As a first approach, each chromosomal inversion was considered as a bi-allelic locus; subsequently, due to overlap and linkage between some inversions, we repeated the analysis identifying loci by chromosomal inversion systems as defined by [1,23,25]. A few specimens carrying the inversions 2Rj and 2Rbk (N = 7, Table 3) were omitted in the latter analysis, without detectable changes in the output. Results by the two approaches were concordant: both detected three genetic clusters in the dataset (Additional file 5). All clusters were composed of specimens belonging to both the M and S forms, suggesting that alternative chromosomal arrangements did not segregate at contrasting frequencies between molecular forms (Additional file 6). The first cluster (Cluster 1, identified by the green colour in Figure 6) corresponded essentially to the FOREST chromosomal form [1]. It was composed mainly of specimens carrying the monomorphic standard arrangement at all inversion systems on chromosome-2 (karyotypes 0000000), with additional low-level polymorphism for inversions 2Rb and 2La only in the S form (Additional file 6). As shown in Figure 6A, where individuals are arranged according to biogeographic domain and latitude of collection, specimens belonging to this cluster were mainly collected in the rainforest area (Central Plateau and Atlantic Coast). Mosquitoes collected in savanna biotopes were partitioned into two distinct genetic clusters based on their karyotype (identified by the yellow and red colours in Figure 6), without further clustering by geography with respect to latitude. A clinal pattern of relative abundance of the two clusters, however, was somewhat apparent (Figure 6). Cluster 2 (in yellow) was mainly composed of specimens polymorphic for the 2Rb and 2La inversions only (Additional file 6), with a high frequency of inverted homozygotes. It closely resembles the cluster grouping together most S form specimens sampled in the dry savannas of Burkina Faso [45], and typically corresponds to the SAVANNA chromosomal form of Coluzzi et al. [1], which is widespread throughout Africa. This cluster intergraded with Cluster 1 at the ecotone between the humid southern savannas and the rainforest, as can be seen in Figure 6A, where some specimens with mixed ancestry (i.e. individuals whose probabilities of membership to Cluster 1 and 2 are substantial; these individuals are identified by their bars being almost equally subdivided in green and yellow) were apparent in this area, especially at the margins of the highlands. Intergradation between Clusters 1 and 2 was more obvious in S than in M, probably because of the sparseness of the savanna populations of the latter taxon (Figure 6B). Finally, Cluster 3 (identified by red bars in Figure 6) was characterized by a high level of polymorphism for each inversion system (Additional file 6). This cluster comprised individuals carrying the 2Rbc and 2Rd arrangements, in combination or not with other inversions. In our samples, the frequency of mosquitoes belonging to this cluster increased when moving northwards (Figure 6). All M and S individuals from Cluster 3 carried karyotypes that would put them under the (polymorphic) SAVANNA chromosomal form described by Coluzzi et al. [1] from collections carried out in Mali [25] and Nigeria [23]. However, contrary to expectations [1,74] and despite considerable geographical overlap and sympatry throughout the savanna areas of Cameroon (Figure 6A), there was lower occurrence of mixed ancestry between the two SAVANNA clusters (2 and 3) than between each of them and cluster 1 (FOREST). To quantify the degree of intergradation between clusters, we classified any mosquito with >10% probability to belong to more than one cluster as an admixed individual (due to hybridization or shared ancestry). We detected 166 admixed individuals between Clusters 1 and 2 (8.4% of the total number of specimens included in the analysis), 81 between Clusters 1 and 3 (4.1%) and 51 (2.6%) between Clusters 2 and 3. Among these, 7 specimens (0.4%) could be assigned to any of the three clusters with probability >10%. Overall, these results indicate that the boundaries between chromosomally-defined clusters were permeable and that they assorted independently of molecular form status. Because a wealth of evidence substantiates the ecological and reproductive unity of the molecular forms, it is likely that the distribution of chromosomal arrangements that we observed merely reflected directional selection acting on different karyotypes in alternative environments, within both M and S.

Bottom Line: Population structure analysis identified three chromosomal clusters, each containing a mixture of M and S specimens.Rather, they are involved in ecological specialization to a similar extent in both genetic backgrounds, and most probably predated lineage splitting between molecular forms.When such mutations occur in portions of the genome where recombination is suppressed, such as the pericentromeric regions known as speciation islands in An. gambiae, they would contribute further to the development of reproductive isolation.

View Article: PubMed Central - HTML - PubMed

Affiliation: Laboratoire de Lutte contre les Insectes Nuisibles, Institut de Recherche pour le Développement, UR016, 911 Av. Agropolis, Cedex 5, Montpellier, France. frederic.simard@ird.fr

ABSTRACT

Background: Speciation among members of the Anopheles gambiae complex is thought to be promoted by disruptive selection and ecological divergence acting on sets of adaptation genes protected from recombination by polymorphic paracentric chromosomal inversions. However, shared chromosomal polymorphisms between the M and S molecular forms of An. gambiae and insufficient information about their relationship with ecological divergence challenge this view. We used Geographic Information Systems, Ecological Niche Factor Analysis, and Bayesian multilocus genetic clustering to explore the nature and extent of ecological and chromosomal differentiation of M and S across all the biogeographic domains of Cameroon in Central Africa, in order to understand the role of chromosomal arrangements in ecological specialisation within and among molecular forms.

Results: Species distribution modelling with presence-only data revealed differences in the ecological niche of both molecular forms and the sibling species, An. arabiensis. The fundamental environmental envelope of the two molecular forms, however, overlapped to a large extent in the rainforest, where they occurred in sympatry. The S form had the greatest niche breadth of all three taxa, whereas An. arabiensis and the M form had the smallest niche overlap. Correspondence analysis of M and S karyotypes confirmed that molecular forms shared similar combinations of chromosomal inversion arrangements in response to the eco-climatic gradient defining the main biogeographic domains occurring across Cameroon. Savanna karyotypes of M and S, however, segregated along the smaller-scale environmental gradient defined by the second ordination axis. Population structure analysis identified three chromosomal clusters, each containing a mixture of M and S specimens. In both M and S, alternative karyotypes were segregating in contrasted environments, in agreement with a strong ecological adaptive value of chromosomal inversions.

Conclusion: Our data suggest that inversions on the second chromosome of An. gambiae are not causal to the evolution of reproductive isolation between the M and S forms. Rather, they are involved in ecological specialization to a similar extent in both genetic backgrounds, and most probably predated lineage splitting between molecular forms. However, because chromosome-2 inversions promote ecological divergence, resulting in spatial and/or temporal isolation between ecotypes, they might favour mutations in other ecologically significant genes to accumulate in unlinked chromosomal regions. When such mutations occur in portions of the genome where recombination is suppressed, such as the pericentromeric regions known as speciation islands in An. gambiae, they would contribute further to the development of reproductive isolation.

Show MeSH
Related in: MedlinePlus