Advertisement
  • Loading metrics

Ecotype Evolution in Glossina palpalis Subspecies, Major Vectors of Sleeping Sickness

  • Thierry De Meeûs ,

    thierry.demeeus@ird.fr

    Affiliations IRD (INTERTRYP), UMR 177 IRD-CIRAD, Centre International de Recherche-Développement sur l'Elevage en zone Subhumide (CIRDES), Bobo-Dioulasso, Burkina-Faso, IRD (INTERTRYP), UMR177 IRD-CIRAD, TA A-17/G, Campus International de Baillarguet, Montpellier, France

  • Jérémy Bouyer,

    Affiliations IRD (INTERTRYP), UMR177 IRD-CIRAD, TA A-17/G, Campus International de Baillarguet, Montpellier, France, Unité Mixte de Recherche Contrôle des Maladies Animales Exotiques et Emergentes, Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD), Montpellier, France, Unité Mixte de Recherche 1309 Contrôle des Maladies Animales Exotiques et Emergentes, Institut national de la recherche agronomique (INRA), Montpellier, France, Institut Sénégalais de Recherches Agricoles, Laboratoire National d'Elevage et de Recherches Vétérinaires, Dakar—Hann, Sénégal

  • Sophie Ravel,

    Affiliation IRD (INTERTRYP), UMR177 IRD-CIRAD, TA A-17/G, Campus International de Baillarguet, Montpellier, France

  • Philippe Solano

    Affiliations IRD (INTERTRYP), UMR 177 IRD-CIRAD, Centre International de Recherche-Développement sur l'Elevage en zone Subhumide (CIRDES), Bobo-Dioulasso, Burkina-Faso, IRD (INTERTRYP), UMR177 IRD-CIRAD, TA A-17/G, Campus International de Baillarguet, Montpellier, France

Ecotype Evolution in Glossina palpalis Subspecies, Major Vectors of Sleeping Sickness

  • Thierry De Meeûs, 
  • Jérémy Bouyer, 
  • Sophie Ravel, 
  • Philippe Solano
PLOS
x

Abstract

Background

The role of environmental factors in driving adaptive trajectories of living organisms is still being debated. This is even more important to understand when dealing with important neglected diseases and their vectors.

Methodology/Principal Findings

In this paper, we analysed genetic divergence, computed from seven microsatellite loci, of 614 tsetse flies (Glossina palpalis gambiensis and Glossina palpalis palpalis, major vectors of animal and human trypanosomes) from 28 sites of West and Central Africa. We found that the two subspecies are so divergent that they deserve the species status. Controlling for geographic and time distances that separate these samples, which have a significant effect, we found that G. p. gambiensis from different landscapes (Niayes of Senegal, savannah and coastal environments) were significantly genetically different and thus represent different ecotypes or subspecies. We also confirm that G. p. palpalis from Ivory Coast, Cameroon and DRC are strongly divergent.

Conclusions/Significance

These results provide an opportunity to examine whether new tsetse fly ecotypes might display different behaviour, dispersal patterns, host preferences and vectorial capacities. This work also urges a revision of taxonomic status of Glossina palpalis subspecies and highlights again how fast ecological divergence can be, especially in host-parasite-vector systems.

Author Summary

The role of environmental factors in driving adaptive trajectories of living organisms is still being debated. This is even more important to understand when dealing with important and /or neglected diseases and their vectors. In this paper, we analysed genetic divergence, computed from several genetic markers, of 614 tsetse flies (Glossina palpalis gambiensis and Glossina palpalis palpalis, major vectors of animal and human trypanosomes) from 28 sites of West and Central Africa. We found that the two subspecies are so divergent that they deserve the species status. We found that G. p. gambiensis from different landscapes (Niayes of Senegal, savannah and coastal environments) were significantly genetically different, and thus represent different adaptive entities or even subspecies. We also confirm that G. p. palpalis from Ivory Coast, Cameroon and DRC are strongly divergent. These results provide an opportunity to examine whether these different types of tsetse fly might display different behaviour, dispersal patterns, host preferences and vectorial capacities. This work also urges a revision of taxonomic status of Glossina palpalis subspecies and highlights again how fast ecological divergence can be, especially in host-parasite-vector systems.

Introduction

A capital step for species diversification is the initiation of some kind of disruptive selection, driving the newly diverged group of entities to some level of genetic adaptive divergence [1,2]. There has been a continuous debate on the respective role of geography and ecology in speciation, especially the speed at which these factors drive organisms to divergence [3]. These debates are important as they focus on key processes involved in evolution. For parasites and their vectors, the role of ecology and geography in driving divergence has important implications for management, as rapid evolution can occur in response to control practices or introductions to new environments [4]. This can have consequences on dispersal capacity [4], behaviour [5] and vectorial capacities [68].

Tsetse flies (Diptera: Glossinidae) are the sole cyclical vectors of human (HAT or sleeping sickness) and animal (AAT or nagana) African trypanosomoses, two major plagues that are seriously impeding African development [9]. Among these, Glossina palpalis palpalis and Glossina palpalis gambiensis, which are major vectors of both HAT and AAT, have recently been the subject of several population genetics studies (see [10] for a review). These studies, mainly based on spatio-temporal variation at microsatellite loci, have recurrently revealed some degree of genetic divergence, in some cases above the reasonable amount expected from geographically based population structure [1114]. Because control programs against trypanosomoses often rely on tsetse eradication or suppression, it is important to specify the amount of such divergences and, if possible, if it could be linked to some ecological factors. Indeed, adaptive divergence may be correlated to variation in behaviour, host preference (or attractiveness to trapping devices) and vectoring ability.

In this paper, we combined and synthesized published and unpublished microsatellite data sets of these two taxa from populations sampled in West Africa and central Africa. We analysed the whole data set in order to evaluate the genetic divergence between the two taxa as assessed with microsatellite markers and then we analysed separately G. p. gambiensis and G. p. palpalis in order to assess the respective role of geographic distance, date of capture (time distance), landscape type and river basin in determining the level of genetic divergence of tsetse flies. The observed levels of divergence provide support for changes in the taxonomic status of these subspecies. Furthermore, based on both genetic and ecological criteria, we propose that several additional taxonomic groups should be recognized. The importance of these findings in developing novel control strategies and facilitating future research endeavours is discussed. Subspecies G. p. gambiensis and G. p. palpalis may have split no more than 13000 years ago [15,16]. The ecotypes evidenced in the present study necessarily are much younger and illustrate on how swift ecological divergence can be.

Material and Methods

Study sites

Study sites are located as represented in Fig. 1. The species, country, landscape type, river basin, date of capture, GPS coordinates and sample sizes are presented in Table 1. Raw data are available in S1 Table.

thumbnail
Fig 1. Geographic locations of sampled tsetse flies.

Glossina palpalis gambiensis samples come from Dakar, Pout, Missira, S1 (Senegal 1) and S3 (Senegal 3) (Senegal); Banjul North and Banjul South (Banjul) (Gambia); Magnokhoun (Mag*), Falessadé, Bani and Guéckédou (Guinea); M8 (Mali 8) and M12 (Mali 12) (Mali); Bleni, Tou* (Toussiana), Bama, Ban* (Banzon), Dar* (Darsalamy), Nia* (Niafongo) and Folonzo (Burkina-Faso). Glossina palpalis palpalis samples come from Bonon, Aniassue, Azaguié and Abidjan (Ivory Coast); Fontem, Bafia Bipindi and Campo (Cameroon); and Malanga (Democratic Republic of Congo). * indicates abbreviated names. For G. p. gambiensis sites, Niayes sites are in purple, coastal sites in blue and savannah sites in green; for G. p. palpalis sites, Ivory-Coast sites are in brown, Cameroon in light brown and DRC in black (see also Fig. 2).

https://doi.org/10.1371/journal.pntd.0003497.g001

thumbnail
Table 1. Characteristics of subsamples of Glossina palpalis gambiensis (Gpg) and Glossina palpalis palpalis (Gpp) in different countries and different sites.

https://doi.org/10.1371/journal.pntd.0003497.t001

Sampling

Most of the samples studied in this paper were already used and genotyped for publications relating to other, though related purposes. These papers are cited in Table 1 and sites samples can be seen in the Fig. 1. Folonzo sample was never published and was sampled during April 2007 following the same method as in [17]. Guekedou sample was never published and was sampled during May 2007 following the same procedure as in [18]; Senegal 1 and Senegal 3 samples were never published and were kindly provided by the Insect Pest Control Laboratory, Joint FAO/IAEA Program of Nuclear Techniques in Food and Agriculture and sampling followed the same procedures as described in [9]. Azaguié sample was never published and was sampled and genotyped for another project of our team by S. Ravel in collaboration with Dr D. Kaba (Pierre Richet / Institut National de Santé Publique, Abidjan, Ivory Coast) and Dr G. Acapovi-Yao (Laboratoire de Zoologie, Université d’Abidjan-Cocody, Abidjan, Ivory Coast) (Acapovi-Yao et al., manuscript in preparation). Published papers are available at http://gemi.mpl.ird.fr/SiteSGASS/SiteTDM/ArtiPDF.html. These 28 samples summed to 614 genotyped individuals, with 9 unpublished samples.

Genotyping

Genotyping of unpublished data followed the same protocol as described in [17] and [19]. For Mali 12, Mali 8, Senegal 1 and Senegal 3, the genotypes of the flies were kindly provided by the Insect Pest Control Laboratory, Joint FAO/IAEA Program of Nuclear Techniques in Food and Agriculture and protocols were the same. Some of analyses undertaken do not tolerate missing data. For the sake of consistency between all analyses, only complete genotypes at seven loci were kept. These loci were: Gpg55.3 (X linked) [20]; B104 (X linked), B110 (X linked) and C102 that were kindly supplied by A. Robinson, Insect Pest Control Laboratory (formerly Entomology Unit), Food and Agricultural Organization of the United Nations/International Atomic Energy Agency [FAO/IAEA], Agriculture and Biotechnology Laboratories, Seibersdorf, Austria; pGp13 (X linked) and pGp24 [21]; and GPCAG [22]. Protocols followed what was described in references cited above (e.g. [18]). All genotyping were handled or supervised by the same person (SR) who ensured perfect calibration of allele sizes across sub-samples. A total of 614 tsetse flies from 28 sites displayed a full genotype at the seven microsatellite loci. All genotypic data were coded as they appeared, hence males were coded as homozygous at X-linked loci. Sex information was missing in samples from Gambia and assessed through genotypes found on X-linked loci.

Data analyses

All data sets were built in appropriate text files and converted with Create V 1.1 [23] into the appropriate format as needed except for bootstrap analysis with Phylip for which we used Convert V 1.31 [24].

Genetic distances were computed with MSA 4.05 [25]. We used a Cavalli-Sforza and Edwards chord distance matrix [26] for dendrogram construction with a Neighbour-joining tree (NJTree) [27] and for regression analyses, as recommended [28,29]. The NJTree dendrogram showing relationships between all tsetse subsamples was built with Mega V 5 [30]. Robustness of nodes was assessed through 1000 bootstraps over loci with Phylip v 3.68 [31]. For that purpose, nodes in G. p. gambiensis were studied after rooting the tree with Malanga subsample (DRC), while for G. p. palpalis nodes, tree was rooted with Banjul North subsample (Gambia). Sample sizes are represented in Table 1.

Relationships between genetic distances and the other parameters were tested with partial Mantel tests (for robustness) and also explored with linear regressions (for illustrations and strength of signals measures). Explanatory variables and factors were the subspecies distance (whether the two compared samples contain the same subspecies or not), geographic distance (in km), time distance (in days), landscape and river basin (same or not). The different landscapes and river basins are presented in Table 1. For the Mantel tests, these factors were coded as 0 when the two sites compared shared the same value (e.g. both G. p. gambiensis) or 1 when different. For the linear regression, factors were coded as "Same" when similar in both sites and a combination of two modalities when different (e.g. Savannah-Coast). Mantel test for global data was undertaken to test for the effect of sub-speciation between G. p. palpalis and G. p. gambiensis. Because there are probably interactions with this effect, other factors were then analyzed more precisely within each subspecies separately. Partial Mantel tests were undertaken under Fstat V 2.9.3 [32] (updated from [33]) with 10000 Monte-Carlo randomizations of genetic distance matrix items. We also undertook Principal Component Analyses (PCA) on each sub-species data set. For this we used PCAGen 1.2.1 [34] that works on allele frequencies in subsamples and reorganize the data into a multidimensional space the metric of which is equivalent to Wright's FST [35], i.e. the part of inbreeding that is explained by population subdivision. The significance of the first axes was tested with the broken stick criterion [36] and also with 10000 permutations of individuals across subsamples. We then submitted subsample coordinates of each significant axis to multiple regressions. The general model to start with was always of the form Axisi ∼ Lat + Long + Day + Landscape + RiverBasin + Lat:Long where i identifies the axis number being investigated, Lat and Long mean the latitudinal and longitudinal GPS coordinates in degrees, Day means the number of days after the oldest sub-sample, Lanscape is as described above, RiverBasin is the name of the river basin as described above and ":" stands for interaction between two explanatory variables. Here the variables were weighted for subsample sizes.

All multiple regressions were undertaken under R [37] using sample sizes as weights. For all linear regressions, the best (minimum) model was selected after a stepwise procedure, using the Akaike Information Criterion [38], significance tested with a F test and multiple comparisons (when useful) were done with the Student-Newman-Keuls (SNK) test. Order of entry of explanatory variables matters both in Fstat (Mantel) and R analyses. We thus chose to enter these variables following an order of decreasing importance we thought they would have: geographic distance, time, landscape, basin and interactions (if any). Ecological factors were entered last to make sure the response was controlled for the other parameters.

Null alleles and X-linked loci produce an artificial excess of inbreeding in subsamples that should not affect the tests in any predictable direction but a decrease in power. These issues are thus relevant only in those cases where tests do not appear significant. A coming work involving one of the authors (TDM) will be devoted to the robustness of different genetic distances to such issues (manuscript in preparation).

NJTrees were also built without X-linked loci, on females only and on males only. This did not change the general aspect of the tree even if a few populations happened to branch in slightly different places. These NJTrees can be seen in S1 File.

Data accessibility

The complete data set is available in S1 Table.

Results

Dendrogram

Geographic locations, landscape types and genetic relationships between all subsamples are presented in Table 1, Fig. 1 and Fig. 2. It can be seen that the two subspecies are clearly separated. In G. p. gambiensis, distinction between Savannah, Niayes and Coastal populations, in some instances, overcome geographic differentiation. This is particularly clear for samples from Gambia and Senegal (Fig. 1). For instance, as can be seen from Figs. 1 and 2, subsamples Senegal 1 and 3 are genetically closer to Burkina-Faso and Mali sites (Savannah) than from the geographically closer Dakar, Pout (Niayes), Missira, Banjul North and South (Coast). In G. p. palpalis, Central and Western African flies are clearly separated and geography seems to be the predominant factor within each of the two zones. Again, genetic distances are quite pronounced and bootstrap values relatively high.

thumbnail
Fig 2. NJTree based on Cavalli-Sforza and Edwards chord distance between all subsamples of G. palpalis s.l. from West and Central Africa.

G. palpalis gambiensis subsamples are in green (Savannah), purple (Niayes) and blue (Coast) respectively. G. palpalis palpalis subsamples are in brown (Ivory Coast), light brown (Cameroon) and black (DRC) respectively. Bootstrap values above 500 (out of 1000) are indicated. For geographic position see Fig. 1.

https://doi.org/10.1371/journal.pntd.0003497.g002

Partial Mantel, PCA and multiple regression approaches

Results of partial Mantel test for the whole data set provided a highly significant contribution of subspecies (partial R2 = 0.65, P-value<0.0001).

For G. p. gambiensis, partial Mantel test highlighted two major factors that best explained genetic distance between subsamples (Table 2). The first is geographic distance, which explains 41% of the variance, followed by landscape distances that explain 10% of the variance of genetic distances and are highly significant. Other parameters (time and river basin) contribute little to the coefficient of determination R2, though significantly so. Regarding the linear model, the stepwise procedure could not simplify the model. Nevertheless, river basin distances did not display consistent results since the response mainly was due to higher genetic distances between sites from the same basin as compared to other comparisons. This incoherence, which probably comes from interaction with geographic distance, led us to remove this factor from the analysis. Results are presented in Fig. 3. The total R2 = 0.66. Here the main factor is geographic distance, followed by landscape distance. Both explained not less than 62% of the total genetic variance (which is quite big given the variation expected for genetic distances). Time explained very little of the variance though significant: the more time between subsamples, the more genetic divergence between them. For landscape distances, paired comparisons led to the conclusion that genetic distances between similar landscapes were smaller than any other comparison, and that Niayes subsamples were always genetically more distant from the other sites than any other comparison.

thumbnail
Fig 3. Results of multiple regressions between partials for genetic distances (DCS&E) and different parameters with 95% confidence intervals, percentage of total variance explained (R2) and corresponding P-value (F test): G. palpalis gambiensis for geographic distance (3a), time distance (DT) (3b) and landscape distance (3c); Glossina palpalis palpalis for geographic distance (DG) (3d).

For (3c), landscape comparisons with different letters are significantly different (SNK test). Statistics are given here only to illustrate the strength of the signals though the nature of the data does not allow using these formally.

https://doi.org/10.1371/journal.pntd.0003497.g003

thumbnail
Table 2. Results of partial Mantel test for Glossina palpalis gambiensis.

https://doi.org/10.1371/journal.pntd.0003497.t002

In G. p. palpalis subsamples, the Mantel test, partial R2 and corresponding P-value are presented in Table 3. Only geographic distance displayed a significant effect here, with 34% of the variance explained. For the multiple regression approach, only geography seemed to play a significant role (partial R2 = 0.34, P-value = 0.0002) (Fig. 3D). In particular, very high bootstrap values are observed between Central and West Africa and between Cameroon and RDC subsamples.

thumbnail
Table 3. Results of partial Mantel test for Glossina palpalis palpalis.

https://doi.org/10.1371/journal.pntd.0003497.t003

For PCA analysis of G. p. gambiensis sub-samples, the first two axes appeared significant both with the broken stick criterion and with permutation testing (P-value≤0.0001 for Axis 1 and P-value = 0.0345 for Axis 2, permutation test). Axes 1 and 2 represent 41% and 16% of total inertia respectively. After stepwise procedures, Axis 1 is explained by all initial variables but Day (Table 4). By far the two most important variables are the latitude and the landscape that explain respectively 66% and 29% of the total variance in axis 1 (both P-values<0.0001). For the second axis, the minimum model was Axis2 ∼ Lat + Long + Landscape + RiverBasin (Table 5). Here, the most important variables are Lanscape and Latitude that respectively explain 39% and 30% of the total variance in axis 2 (P-values<0.00001).

thumbnail
Table 4. Minimum model obtained for the regression of PCA axis 1 coordinates of Glossina palpalis gambiensis.

https://doi.org/10.1371/journal.pntd.0003497.t004

thumbnail
Table 5. Same as for Table 4 but for the second PCA axis of Glossina palpalis gambiensis subsamples.

https://doi.org/10.1371/journal.pntd.0003497.t005

For PCA analyses of G. p. palpalis sub-samples, the first three axes appeared significant both with the broken stick and permutation tests, with permutation P-value≤0.0001 for the two first axes and P-value = 0.011 for the third. They respectively represent 34, 27 and 17% of total inertia respectively. Here, variable Landscape was not introduced as it does not vary in the sampled zones for G. p. palpalis. For axis 1, no simplification of the model was possible (Table 6). The only significant effect comes from the latitude which explains 95% of the total variance on axis 1 (P-value = 0.0038). On axis 2, only two variables stayed in the model after the stepwise procedure (Table 7). However only longitude really mattered and explained no less than 94% of axis 2 (P-value≤0.0001). Finally, for axis 3, the minimum model was Axis3 ∼ Lat + Long + RiverBasin + Lat:Long and the most important explanatory variables were the river basin, explaining 81% of axis 3 variance (P-value = 0.0053), and the interaction between latitudinal and longitudinal coordinates that explained 14% of axis 3 variance (P-value = 0.0291) (Table 8).

thumbnail
Table 6. Results for the model obtained for the regression of PCA axis 1 coordinates of Glossina palpalis palpalis.

https://doi.org/10.1371/journal.pntd.0003497.t006

thumbnail
Table 7. As for Table 6 but with axis 2 of Glossina palpalis palpalis sub-samples PCA coordinates.

https://doi.org/10.1371/journal.pntd.0003497.t007

thumbnail
Table 8. As for Table 7 but with axis 3 of Glossina palpalis palpalis sub-samples PCA coordinates.

https://doi.org/10.1371/journal.pntd.0003497.t008

Discussion

The importance of geographic distance for determining genetic relationships between tsetse populations has been recurrently reported [9,19,39]. Its predominant effect above the effect of river basin was an expected result, at least for G. p. gambiensis [9] and is newly demonstrated here for G. p. palpalis.

The genetic distance that separates the two subspecies and the high bootstrap level obtained with microsatellite markers (known for their homoplasic nature) are advocating for a revision of the nomenclature of those taxa as different species. This is also in line with the biological definition of species, although the usefulness of such a concept is debatable [40], since the heterozygous males of the F1 crossing between these taxa are completely sterile [41,42] which leads to a very sharp allopatry between them in Ivory Coast [43,44]. Moreover, these taxa can be discriminated on a morphological basis using the size of the palette of the inferior claspers (larger in G. p. gambiensis) and the length of hairs on the inferior claspers (longer in G. p. gambiensis) [45]. This is even more justified as we also find evidence in the present paper of the existence of subunits within these two taxa, some of which are of an ecological nature.

The stronger impact of river basins on G. p. gambiensis than on G. p. palpalis is not surprising, taking into account that the savannah environment of the former makes it much more difficult to cross the interfluve than the dense forest environment of the latter. Time did not play a very pronounced effect on G. p. gambiensis and apparently had no effect on G. p. palpalis. For the latter, smaller sample sizes are probably the cause of this absence of detectable effect. For G. p. gambiensis, the significance of the effect is in line with genetic drift due to small effective population sizes that could be estimated in several studies in these taxa [9,10,12,1719,39,46] but also in other tsetse taxa (see [47] for review). It highlights the need to take into account this factor in population genetics studies and the necessity to avoid pooling individuals that do not belong to the same cohort, in particular to estimate population differentiation, isolation by distance and migration.

In G. p. gambiensis, an important and significant effect of landscape where tsetse flies are found was evidenced. Interestingly, in several instances, genetic distances between subsamples from different landscapes are far above those between subsamples from the same landscape, even between the most remote ones. This strong impact of landscape was confirmed by the regression analyses where this variable explained as much, and sometimes more, the genetic composition of G. p. gambiensis sub-samples. Our study also confirms the genetic isolation of G. p. gambiensis from the Niayes [12,48] which has led to an eradication program in Senegal (http://www.fao.org/news/story/en/item/211898/icode/). It is clear from the different analyses that tsetse from the Niayes (Senegal) represent an objective subspecies, adapted to a specific environment [48,49]. This subspecies is able to reproduce in the complete absence of perennial hydrographic network. Moreover, tsetse from savannah and those from coastal landscapes also represent original diverged entities that can deserve the denomination of ecotypes, if not subspecies. There is however no pre- or post-mating barriers between these taxa, as evidenced by successful mating observed between tsetse flies from the Niayes and savannah tsetse flies from Mali and Burkina-Faso [50]. They can thus be considered as subspecies. It is the first time that such subspecies are evidenced. It has to be underlined that the discovery of these ecotypes may have important consequences. In particular, data from several studies made in the coastal part of Guinea have shown that the G. p. gambiensis ecotype caught in the sleeping sickness foci of this country do not display any infection with the pathogenic trypanosomes usually identified (including human and animal trypanosomes) in this species. Nevertheless, G. p. gambiensis is the only vector of sleeping sickness there [51,52]. To what extent the fact that they constitute a distinct ecotype can be linked to a different vector capacity remains to be documented, but may be of paramount importance for control programmes against both human and animal trypanosomoses. It was also demonstrated that Trypanosoma brucei gambiense from Guinea were genetically very different than those from Ivory Coast, and that this was probably due to the fact that they were not transmitted by the same tsetse taxa, i.e. G. p. gambiensis of the coastal landscape for T. b. gambiense from Guinea, and G. p. palpalis for the T. b. gambiense from Ivory Coast [53].

For G. p. palpalis, the very high bootstrap values observed between Central and West Africa and between Cameroon and RDC subsamples suggest subspeciation, if not more, in the ecological sense of it (adaptively divergent but not necessarily sexually isolated entities, see [40,54]). The existence of three subspecies (or even species) separating flies from West Africa (Ivory Coast), South of Cameroon, Equatorial Guinea and DRC has already been suggested, based on mtDNA (COI) [13] and there are probably more than that [55]. Here, our seven microsatellite loci provide a strong confirmation that G. p. palpalis is a strongly heterogeneous taxon. Moreover, [56] found significant differences in the morphology of the head between G. p. palpalis from West Africa and DRC. Regression analyses on PCA axes also highlighted the relevance of river basins. Nevertheless, many sites in the range of this species are missing (Gabon, Nigeria, Benin, Togo and Ghana) and other environmental measures are missing as well. Future studies, implying GIS approaches should bring more information and more precision on the mechanisms of ecotype and population delimitations in tsetse flies.

These observations are not only of academic interest as they have important repercussion as regard to vector control. Such ecological entities might represent different cases as regard to control success and reinvasion probabilities. It is thus key that such newly defined entities be ecologically characterized in order to compare their respective ecology (hygrometric and temperature preferences, host preferences, symbiotic flora and vector competences). Mating preferences or differential competitiveness may also alter the success of sterile insect technique (SIT) if inappropriate ecotypes are released in the wrong environment. This thus opens the gate to many and very productive new research topics on trypanosomes and their vectors. It also highlights how useful genetic markers can be in exploring the ecology of difficult organisms. Finally, our results call for an urgent taxonomic review of the status of G. palpalis subspecies. The split between G. p. gambiensis and G. p. palpalis was dated as old as 3.2 million years according to COI mtDNA assuming molecular clock [13]. Nevertheless, this result is based on a single mtDNA marker known to behave very oddly at the beginning of a split (for less than 1 million year the divergence can vary from 4 to 20%) [57]. Moreover, because of their lack of neutrality [5860], mtDNA markers might not be ideal to estimate divergence time. We thus prefer relying on experts of the life history of tsetse flies who dated the split between the two sub-species around 13000 years ago (around 91000 tsetse fly generations) when the initial forest was separated into two isolated masses by drought [15,16]. It is probably the most parsimonious interpretation of tsetse flies history. The mean genetic distance between the two taxa is 0.65 (which is very high for a distance bonded to 1). It is 0.48 between savannah and coast subsamples, 0.53 between savannah and the Niayes and 0.5 between coast and the Niayes. Assuming constant microsatellite divergence with time, we can extrapolate that the ecological split in G. palpalis gambiensis occurred around 10000 years ago (around 70000 generations), hence at the end of last glaciation. These estimates probably correspond to considerable overestimates as divergence speed probably strongly decreased as the two sub-species increased in population size (which tends to freeze genetic drift) when meteorological constraints were progressively relaxed at the end of the Würm ice age. These results provide another powerful illustration on how swift ecological divergences can occur, in particular in host-parasite-vector systems [4,54].

Supporting Information

S1 Table. Raw data of Glossina palpalis gambiensis and Glossina palpalis palpalis with Species, Country, Site, Landscape, Date, Day, Bassin, FAO_hydroshed, Sex, SubSample and genotypes at loci X55_3, XpGp13, pGp24, XB104, XB110, C102 and GPCAG.

https://doi.org/10.1371/journal.pntd.0003497.s001

(TXT)

S1 File. NJTrees obtained with chord distance matrices on autosomal loci only, females only and males only.

https://doi.org/10.1371/journal.pntd.0003497.s002

(PPTX)

Acknowledgments

We are very grateful to the Insect Pest Control Laboratory, Joint FAO/IAEA Program of Nuclear Techniques in Food and Agriculture for providing some of the tsetse samples and genotypes of four additional samples in this study. We thank our scientific collaborators and the direction of our partners institute from Institut Pierre Richet Côte d’Ivoire, in particular D. Kaba and G. Acapovi-Yao for the Azaguié sample, CIRDES Bobo-Dioulasso Burkina Faso, and the national control program against sleeping sickness of Guinea for kindly providing us some of the tsetse samples. We wish to thank anonymous referees that reviewed early and late versions of the present paper that helped improving it.

Author Contributions

Conceived and designed the experiments: TDM JB SR PS. Performed the experiments: SR. Analyzed the data: TDM. Contributed reagents/materials/analysis tools: JB SR PS. Wrote the paper: TDM JB SR PS.

References

  1. 1. Kirkpatrick M, Ravigné V (2002) Speciation by natural and sexual selection: models and experiments. American Naturalist 159 Suppl 3: S22–S35. pmid:18707367
  2. 2. De Meeûs T (2000) Adaptive diversity, specialisation, habitat preference and parasites. In: Poulin R, Morand S, Skorping A, editors. Evolutionary Biology of Host Parasite Relationships: Theory Meets Reality. Amsterdam.: Elsevier. pp. 27–42.
  3. 3. Kempf F, Boulinier T, De Meeûs T, Arnathau C, McCoy KD (2009) Recent evolution of host-associated divergence in the seabird tick Ixodes uriae. Molecular Ecology 18: 4450–4462. pmid:19793353
  4. 4. De Meeûs T, Koffi BB, Barré N, de Garine-Wichatitsky M, Chevillon C (2010) Swift sympatric adaptation of a species of cattle tick to a new deer host in New-Caledonia. Infection Genetics and Evolution 10: 976–983. pmid:20601171
  5. 5. Gimonneau G, Bouyer J, Morand S, Diabate A, Simard F (2010) A behavioral mechanism underlying ecological divergence in the malaria mosquito Anopheles gambiae. Behavioral Ecology 21: 1087–1092. pmid:22476108
  6. 6. Duneau D, Boulinier T, Gómez-Díaz E, Petersen A, Tveraa T, et al. (2008) Prevalence and diversity of Lyme borreliosis bacteria in marine birds. Infection Genetics and Evolution 8: 352–359. pmid:18394972
  7. 7. McCoy KD (2008) The population genetic structure of vectors and our understanding of disease epidemiology. Parasite 15: 444–448. pmid:18814720
  8. 8. McCoy KD, Chapuis E, Tirard C, Boulinier T, Michalakis Y, et al. (2005) Recurrent evolution of host-specialized races in a globally distributed parasite. Proceedings Of The Royal Society B-Biological Sciences 272: 2389–2395. pmid:16243689
  9. 9. Kone N, Bouyer J, Ravel S, Vreysen MJB, Domagni KT, et al. (2011) Contrasting population structures of two vectors of African trypanosomoses in Burkina Faso: consequences for control. PLoS Neglected Tropical Diseases 5: e1217. pmid:21738812
  10. 10. Solano P, Ravel S, De Meeûs T (2010) How can tsetse population genetics contribute to African trypanosomiasis control? Trends in Parasitology 26: 255–263. pmid:20202905
  11. 11. Ravel S, De Meeûs T, Dujardin JP, Zeze DG, Gooding RH, et al. (2007) The tsetse fly Glossina palpalis palpalis is composed of several genetically differentiated small populations in the sleeping sickness focus of Bonon, Côte d'Ivoire. Infection Genetics and Evolution 7: 116–125. pmid:16890499
  12. 12. Solano P, Kaba D, Ravel S, Dyer NA, Sall B, et al. (2010) Population genetics as a tool to select tsetse control strategies: suppression or eradication of Glossina palpalis gambiensis in the Niayes of Senegal. PLoS Neglected Tropical Diseases 4: e69.
  13. 13. Dyer NA, Furtado A, Cano J, Ferreira F, Afonso MO, et al. (2009) Evidence for a discrete evolutionary lineage within Equatorial Guinea suggests that the tsetse fly Glossina palpalis palpalis exists as a species complex. Molecular Ecology 18: 3268–3282. pmid:19619197
  14. 14. Solano P, De la Rocque S, Cuisance D, Geoffroy B, De Meeûs T, et al. (1999) Intraspecific variability in natural populations of Glossina palpalis gambiensis from West Africa, revealed by genetic and morphometric analyses. Medical and Veterinary Entomology 13: 401–407. pmid:10608229
  15. 15. De Meeûs T, Ravel S, Rayaisse JB, Kaba D, Courtin F, et al. (2014) Genetic correlations within and between isolated tsetse populations: what can we learn? Acta Tropica 138S: S6–S11.
  16. 16. Challier A, Gouteux JP, Coosemans M (1983) La limite géographique entre les sous-espèces Glossina palpalis palpalis (Rob.-Desv.) et G. palpalis gambiensis Vanderplank (Diptera: Glossinidae) en Afrique occidentale. Cahiers ORSTOM Série Entomologie Médicale et Parasitologie 21: 207–220.
  17. 17. De Meeûs T, Ravel S, Rayaisse J-B, Courtin F, Solano P (2012) Understanding local population genetics of tsetse: the case of an isolated population of Glossina palpalis gambiensis in Burkina Faso. Infection Genetics and Evolution 12: 1229–1234. pmid:22546274
  18. 18. Solano P, Ravel S, Bouyer J, Camara M, Kagbadouno MS, et al. (2009) The population structure of Glossina palpalis gambiensis from island and continental locations in coastal Guinea. PLoS Neglected Tropical Diseases 3: e392. pmid:19290038
  19. 19. Mélachio T, , , Simo G, Ravel S, De Meeûs T, Causse S, et al. (2011) Population genetics of Glossina palpalis palpalis from central African sleeping sickness foci. Parasites and Vectors 4: 140. pmid:21767402
  20. 20. Solano P, Duvallet G, Dumas V, Cuisance D, Cuny G (1997) Microsatellite markers for genetic population studies in Glossina palpalis (Diptera: Glossinidae). Acta Tropica 65: 175–180. pmid:9177579
  21. 21. Luna C, Bonizzoni MB, Cheng Q, Aksoy S, Zheng L (2001) Microsatellite polymorphism in the tsetse fies (Diptera: Glossinidae). Journal of Medical Entomology 38: 376–381. pmid:11372961
  22. 22. Baker MD, Krafsur ES (2001) Identification and properties of microsatellite markers in tsetse flies Glossina morsitans sensu lato (Diptera: Glossinidae). Molecular Ecology Notes 1: 234–236. pmid:16479272
  23. 23. Coombs JA, Letcher BH, Nislow KH (2008) CREATE: a software to create input files from diploid genotypic data for 52 genetic software programs. Molecular Ecology Resources 8: 578–580. pmid:21585837
  24. 24. Glaubitz JC (2004) CONVERT: A user-friendly program to reformat diploid genotypic data for commonly used population genetic software packages. Molecular Ecology Notes 4: 309–310.
  25. 25. Dieringer D, Schlötterer C (2002) Microsatellite analyser (MSA): a platform independent analysis tool for large microsatellite data sets. Molecular Ecology Notes 3: 167–169.
  26. 26. Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: model and estimation procedures. American Journal of Human Genetics 19: 233–257. pmid:6026583
  27. 27. Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406–425. pmid:3447015
  28. 28. Takezaki N, Nei M (1996) Genetic distances and reconstruction of phylogenetic trees from microsatellite DNA. Genetics 144: 389–399. pmid:8878702
  29. 29. De Meeûs T, McCoy KD, Prugnolle F, Chevillon C, Durand P, et al. (2007) Population genetics and molecular epidemiology or how to "débusquer la bête". Infection Genetics and Evolution 7: 308–332. pmid:16949350
  30. 30. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, et al. (2011) MEGA5: Molecular Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary Distance, and Maximum Parsimony Methods. Molecular Biology and Evolution 28: 2731–2739. pmid:21546353
  31. 31. Felsenstein J (2008) PHYLIP: Phylogeny Inference Package. Seattle, WA 98195-5065, USA Department of Genome Sciences and Department of Biology, University of Washington, Box 355065, available at http://evolution.gs.washington.edu/phylip.html.
  32. 32. Goudet J (2002) FSTAT, a program to estimate and test gene diversities and fixation indices (version 2.9.3.2). Available from http://www.unil.ch/izea/softwares/fstat.html. Updated from Goudet (1995).
  33. 33. Goudet J (1995) FSTAT (Version 1.2): A computer program to calculate F-statistics. Journal of Heredity 86: 485–486.
  34. 34. Goudet J (1999) PCA-GEN for Windows. Available from http://www2.unil.ch/popgen/softwares/pcagen.htm.
  35. 35. Wright S (1965) The interpretation of population structure by F-statistics with special regard to system of mating. Evolution 19: 395–420.
  36. 36. Frontier S (1976) Etude de la décroissance des valeurs propres dans une analyse en composantes principales: comparaison avec le modèle du bâton brisé. Journal of Experimental Marine Biology and Ecology 25: 67–75.
  37. 37. R-Development-core-team (2013) R: A Language and Environment for Statistical Computing. 3.0.2 ed: R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org, ISBN 3-900051-07-0. pmid:23519455
  38. 38. Akaike H (1974) A new look at the statistical model identification. IEEE Transactions on Automatic Control 19: 716–723.
  39. 39. Bouyer J, Balenghien T, Ravel S, Vial L, Sidibé I, et al. (2009) Population sizes and dispersal pattern of tsetse flies: rolling on the river? Molecular Ecology 18: 2787–2797. pmid:19457176
  40. 40. De Meeûs T, Durand P, Renaud F (2003) Species concepts: what for? Trends in Parasitology 19: 425–427. pmid:14519574
  41. 41. Van der Planck FL (1948) Experiments in cross-breeding tsetse flies. Annals of Tropical Medicine & Parasitology 42: 131–152. pmid:25046131
  42. 42. Gooding RH, Solano P, Ravel S (2004) X-chromosome mapping experiments suggest occurrence of cryptic species in the tsetse fly Glossina palpalis palpalis. Canadian Journal of Zoology-Revue Canadienne De Zoologie 82: 1902–1909.
  43. 43. Challier A, Gouteux JP, Coosemans M (1983) La limite géographique entre les sous-espèces de Glossina palpalis palpalis (Rob.-Desv.) et Glossina palpalis gambiensis Vanderplanck en Afrique occidentale. Cahiers ORSTOM, Série Entomologie médicale et Parasitologie 12: 207–220.
  44. 44. Van der Planck FL (1949) The classification of Glossina palpalis, including the description of new subspecies and hybrids. Proceedings of the Royal Society of London 18: 69–77.
  45. 45. de Barros Machado A (1954) Révision systématique des Glossines du groupe palpalis (Diptera). Diamang (Companhia de Diamantes de Angola) Publicações Culturais 22: 1–190. pmid:20099407
  46. 46. Bouyer J, Ravel S, Guerrini L, Dujardin JP, Sidibé I, et al. (2010) Population structure of Glossina palpalis gambiensis (Diptera: Glossinidae) between river basins in Burkina-Faso: consequences for area-wide integrated pest management. Infection Genetics and Evolution 10.
  47. 47. Gooding RH, Krafsur ES (2005) Tsetse genetics: contributions to biology, systematics, and control of tsetse flies. Annual Review of Entomology 50: 101–123. pmid:15355235
  48. 48. Bouyer J, Seck MT, Sall B, Guerrini L, Vreysen MJB (2010) Stratified entomological sampling in preparation of an area-wide integrated pest management programme: the example of Glossina palpalis gambiensis in the Niayes of Senegal. Journal of Medical Entomology 47: 543–552. pmid:20695269
  49. 49. Touré S (1974) Note sur quelques particularités dans l’habitat de Glossina palpalis gambiensis Vanderplank, 1949 (Diptera, Glossinidae) observées au Sénégal. Revue d'Elevage et de Médecine Vétérinaire des Pays Tropicaux 27: 81–94.
  50. 50. Mutika GN, Kabore I, Seck MT, Sall B, Bouyer J, et al. (2012) Mating performance of Glossina palpalis gambiensis strains from Burkina Faso, Mali and Senegal. Entomologia Experimentalis et Applicata 146: 177–185.
  51. 51. Kagbadouno M, Camara M, Bouyer J, Hervouet JP, Courtin F, et al. (2009) Tsetse elimination: its interest and feasibility in the historical sleeping sickness focus of Loos Islands, Guinea. Parasite-Journal De La Societe Francaise De Parasitologie 16: 29–35. pmid:19353949
  52. 52. Kagbadouno MS, Camara M, Rouamba J, Rayaisse JB, Traoré IS, et al. (2012) Epidemiology of sleeping sickness in boffa (Guinea): where are the trypanosomes? PLoS Neglected Tropical Diseases 6: e1949. pmid:23272259
  53. 53. Koffi M, De Meeûs T, Bucheton B, Solano P, Camara M, et al. (2009) Population genetics of Trypanosoma brucei gambiense, the agent of sleeping sickness in Western Africa. Proceedings of the National Academy of Sciences of the United States of America 106: 209–214. pmid:19106297
  54. 54. De Meeûs T, Michalakis Y, Renaud F (1998) Santa Rosalia revisited: or why are there so many kinds of parasites in 'the garden of earthly delights'? Parasitology Today 14: 10–13. pmid:17040683
  55. 55. Cordon-Obras C, Cano J, Knapp J, Nebreda P, Ndong-Mabale N, et al. (2014) Glossina palpalis palpalis populations from Equatorial Guinea belong to distinct allopatric clades. Parasites & Vectors 7: 31.
  56. 56. Gouteux JP, Dagnogo M (1985) Homogénéité morphologique des genitalia mâles de Glossina palpalis palpalis (Diptera: Muscidae) en Côte d'Ivoire. Cahier de l'ORSTOM, Série Entomologie Médicale et Parasitologie 23: 55–59.
  57. 57. Papadopoulou A, Anastasiou I, Vogler AP (2010) Revisiting the insect mitochondrial molecular clock: the mid-Aegean trench calibration. Molecular Biology and Evolution 27: 1659–1672. pmid:20167609
  58. 58. Bazin E, Glemin S, Galtier N (2006) Population size does not influence mitochondrial genetic diversity in animals. Science 312: 570–572. pmid:16645093
  59. 59. Galtier N, Jobson RW, Nabholz B, Glémin S, Blier PU (2009) Mitochondrial whims: metabolic rate, longevity and the rate of molecular evolution. Biology Letters 5: 413–416. pmid:19324654
  60. 60. Gerber AS, Loggins R, Kumar S, Dowling TE (2001) Does nonneutral evolution shape observed patterns of DNA variation in animal mitochondrial genomes? Annual Review of Genetics 35: 539–566. pmid:11700293
  61. 61. Kaba D, Ravel S, Acapovi-Yao G, Solano P, Allou K, et al. (2012) Phenetic and genetic structure of tsetse fly populations (Glossina palpalis palpalis) in southern Ivory Coast. Parasites & Vectors 5: 153.