INVESTIGADORES
RIVERA Paula Cecilia
congresos y reuniones científicas
Título:
Niche divergence in Boa constrictor: a conservation approach
Autor/es:
RIVERA PAULA C.; DI COLA VALERIA; CHIARAVIGLIO MARGARITA
Lugar:
Arecibo
Reunión:
Simposio; Caribbean and Latin American Boa Group Meeting; 2013
Institución organizadora:
Caribbean and Latin American Boid Specialist Group
Resumen:
Boa constrictor is the most widely distributed boid species in America, it is distributed from northern Mexico to central Argentina. It occurs in numerous biogeographic domains, to the point that Boa constrictor is considered to be a plastic species, which can adapt to different environments. On the other hand, the knowledge of its intraspecific taxonomy is vastly discussed. Around fourteen subspecies have been described until now, but only 8 or 9 are usually recognized. Among them three are distributed in islands (and some authors considered them as true distinct species), two are local endemism and the remaining four are widely distributed (Boa constrictor amarali, Boa constrictor constrictor, Boa constrictor imperator and Boa constrictor occidentalis).
Assessing conservation status of widely distributed lineages is now a point of controversy. Generally these lineages are not considered as conservation issues as are endemic ones. This could be due to the fact that the classification of threatened species in Red Lists is constructed to be applied primarily on a global scale, not on a national scale or an environmental scale. This is the case of the four widely distributed subspecies of Boa constrictor and that is why we decided to perform this study with the objectives of estimating their potential distribution and to identify the environmental requirements that distinguish their ecological niches. Finally, based on these data we discussed about their conservation status taking into account landscape modification of the different biogeographic regions where these lineages occur.
The analysis of species-environment relationship has been a key question in ecology due to its influence in the distribution patterns. The quantification of this relationship represents the essence of the environmental niche modeling. This methodology relates field observations with environmental predictor variables in order to characterize favorable environmental conditions for the occurrence of the species.
We gathered, from museum records, a total of 113 points of presence for Boa constrictor occidentalis, 136 for Boa constrictor amarali, 110 for Boa constrictor constrictor and 132 for Boa constrictor imperator. Points of presence in areas where more than two lineages were recorded were not incorporated into the study, as well as points in areas outside the known distribution for the lineage in order to avoid misidentifications.
Two types of variables were considered as predictors: topographic (altitude), and bioclimatic, both obtained from WorldClim. All variables were post-processed at a pixel size of 2.5 arc-minutes resolution. Resulting data layers cover South and Central America and the south of North America. To avoid over-parameterizing the analyses we performed Pearson correlation tests. For variables that were highly correlated, we chose the variable more biologically easy to interpret. After that, six independent variables were selected: altitude, Annual mean temperature (BIO1), Temperature seasonality (BIO4), Annual precipitation (BIO12), Precipitation seasonality (BIO15), Precipitation in the warmest quarter (BIO18). Models were constructed with Maxent , using an equal training sensitivity and specificity cumulative threshold. This approach is more restrictive than the minimum training presence threshold. I make this clarification because the models we obtained were very restrictive indicating areas where the lineage is found as unsuitable.
The model for B.c. occidentalis shows that this lineage is restricted to the Chaco biogeographic region in Argentina. However, it did not indicate its presence in the Paraguayan Chaco. According to the model, B. c. amarali distributes mainly in the Cerrado region in Brasil, Bolivia and Paraguay, but also it occurs in the north of the Paranaense rainforest. On the other hand, the northern region of the Cerrado is not indicated for this species, maybe because no presence data of this area were included in the model. Regarding B. c. constrictor, it is the lineage most widely distributed, occupying the Amazonas and the Neotropical Savanna in the north of Venezuela, but also the model indicates as suitable areas located in the Caatinga and in the Atlantic Rainforest of Northeast of Brazil. For this lineage the model also overpredicts its distribution in mainland and insular Central America. This is because geographical barriers to dispersal are not taken into account when modeling the potential distribution. The model for B. c. imperator shows that it is distributed in very diverse biogeographic regions including the arid regions of Sonora and Ecuador, the arid and semiarid forest and savannas of the Pacific Mexican cost, Mexican Gulf, East of Central America and Magdalena and the humid forests of the Yucatan Peninsula, Chiapas, West of the Isthmus of Panamá and Chocó. Probably due to this niche amplitude, the model also indicates several other regions as suitable. We performed a principal component analysis using the values of the six independent environmental variables of each pixel with a presence record to examine the overall level of divergence in environmental space among these four subspecies and to assess which environmental variables influence their separation. This analysis reveals that the two first components explain 69% of the variation. The principal component one, responsible for the 44% of the variation, depicts a gradient from subtropical sites with a marked cold and dry winter, to tropical sites with higher temperatures and precipitations. The principal component two, responsible for the 25% of the variation, represents a gradient from sites with lower to higher altitude and with seasonal precipitation to localities with year-round rainfall. We can observe that B. occidentalis and B c. amarali show cohesiveness in their environmental space. On the other hand, B c. constrictor shows great niche amplitude and its overlaps with B. c. amarali. This could be due to the fact that some specimens of B. c. constrictor may be incorrectly classified or that actually these two lineages present similar environmental requirements. Regarding B. c. imperator, the niche amplitude we observed in the distribution map it is also presented here. It overlaps almost all the other lineages and also presents the site with the most extreme environmental variability.
Finally, we analyzed these results from a conservation approach considering the anthropogenic processes that occurs in the different biogeographic region where the boas occurs.
Regarding B. c. occidentalis: it is a regional endemism of the Chaco. This region is a largely threatened region, strongly affected by extensive livestock raising, extractive forestry and poorly planned agricultural expansion, being the woodland in the Chaco region lost at an annual rate of 2.2%. B. c. amarali is regional endemism of the Cerrado.This region is severely affected also by extractive forestry and agricultural expansion. Although the entire distribution of the lineage is not clear, it appears to occur allopatrically only in the south of the Cerrado region, an area heavily modified by human activities.
In the case of B. c. constrictor, this lineage is the most widely distributed, it occurs in numerous countries. National conservation status instead a global one is necessary to preserve its different populations. Furthermore, since this lineage occurs in very different biogeographic regions, we may ask if there are different lineages under the name of B. c. constrictor. And finally, regarding B. c. imperator, its wide niche amplitude could indicate that maybe we are considering different lineages under the same name. Being able to efficiently and accurately delimit species is one of the most basic and important aspects in conservation. Especially when we are dealing with lineages distributed in several countries and biogeographic regions and hence exposed to different threats to their conservation.
This is a first approximation of niche modeling in the subspecies of Boa constrictor. New analyses using modeling technique ensembles and other methodologies to estimate thresholds could improve the models. But also integrating this methodology with morphological and genetic analysis could improve the result in dubious lineages like B. c. contrictor or B. c. imperator.