jbi12135-sup-0001-appendixS1

advertisement
1
Journal of Biogeography
SUPPORTING INFORMATION
Replicated radiations of the alpine genus Androsace (Primulaceae) driven by range
expansion and convergent key innovations
Cristina Roquet, Florian C. Boucher, Wilfried Thuiller and Sébastien Lavergne
Appendix S1 Supplementary methods
Study group
Androsace is a plant genus of c. 110 species, most of them found in alpine habitats. The map
below (Fig. S1) shows that a high proportion of Androsace species are found in mountainous
areas. It should be noted that this map shows occurrence data, and thus it may not reflect the
true species richness.
Figure S1 World map of Androsace species richness with a resolution of 1°. The grey layer shows areas
with an altitude higher than 1000 m. Occurrence data was extracted from the Global Biodiversity Information Facility database (GBIF ; http://www.gbif.org/).
2
Biogeographical inference
Biogeographical analyses were conducted with LAGRANGE (Ree & Smith, 2008) at two scales: continental and regional. At the continental scale, we aimed to search for broader patterns in the
biogeographic reconstruction and to test which continental delimitation (Fig. S1) suits best
Androsace species based on their likelihood.
Figure S2 Four biogeographical models with different continental delimitation that where tested with the
dispersal–extinction–cladogenesis method. The differences (highlighted by circles) concern whether the
Caucasus and Asia Minor regions are grouped with Europe or with the rest of Asia; and whether the
Beringian Asian region is grouped with North America or with the rest of Asia. The highest-scoring model
was A.
At the regional scale, two types of biogeographical models were compared: a baseline
model without dispersal constraints; and a stepping-stone model with different dispersal probabilities for neighbouring and non-neighbouring areas (Fig. S3), testing for a range of probabilities: we first tested 10 values equally ranged from 0.1 to 1, and because the best value was obtained with 0.1, we then tested 10 values equally ranged from 0.01 to 0.1. Neighbour areas were
defined as areas that are adjacent or at least that do not have another area or sea between them
that could act as a barrier, except for Arctic Asia and Arctic North America, which were considered as neighbouring areas due to the lability of the Bering Strait, where a land bridge has emerged during various geological periods (Hopkins, 1967; Wen, 1999). The species’ geographical
ranges were defined by integrating data from the Global Biodiversity Information Facility
database (GBIF, http://www.gbif.org/) and information of several floras: Flora of China (Hu &
Kelso, 1996); Flora Europaea (Ferguson, 1972); Flora Ibérica (Kress, 1997); Flora of North
America (Cholewa & Kelso, 2009); Flora of Pakistan (Nasir, 1984); and Flora of the U.R.S.S.
(Shishkin & Bobrov, 1952).
3
Figure S3 Diagram showing the stepping-stone biogeographical model used with LAGRANGE. Arrows show
dispersals for which probability was set as equal to 1, while dispersals between areas without arrows
were set to a lower value. All dispersals are bidirectional.
Climatic vicariance analysis
To detect whether diversification in Androsace has been influenced by climatic vicariance, we
used the statistical analyses implemented in SEEVA (Struwe et al., 2011). The rationale behind
SEEVA is that phylogenetic nodes (i.e. cladistic splits) can reflect ecological or geographical splits
(i.e. ecological or spatial vicariance). Statistically, it can be evaluated whether nodes are associated with specific ecological splits. SEEVA tests this expectation with a null hypothesis that ecological and phylogenetic splits are independent. For instance, if species’ divergences were systematically associated with events of climatic vicariance, one would conclude that climate caused parapatric speciation in the study group. To address this issue, we tested whether there was a significant divergence for the climatic niche between all sister lineages within Androsace with SEEVA.
This software works using data from individuals or populations and a phylogeny. Quantitative data is grouped into quantiles; a contingency table for each node and variable is created;
and a Fisher exact test is performed to test for skewed patterns between sister clades. Because
an independent test is computed for each node (here, 41 in total), a Bonferroni correction (Rice,
1989) has to be applied to define the significance level of P. In addition, SEEVA computes an index
of divergence (D, with values from 0 to 1) for each node and variable, which allows a comparison
of the strength of phylogenetic–ecological associations of different variables for a given node.
Occurrence data for the SEEVA analyses was extracted from Boucher et al. (2012), for
which 19 Bioclim variables were obtained for 51 species from the WordClim database (Hijmans
et al., 2005; http://worldclim.org/). We selected 5 variables based on a principal components
analysis (PCA): annual mean temperature; temperature seasonality; isothermality; annual precipitation; and precipitation seasonality. The phylogeny used corresponded to the 50% majorityrule consensus phylogeny from Boucher et al. (2012)
4
REFERENCES
Boucher, F.C., Thuiller, W., Roquet, C., Douzet, R., Aubert, S., Alvarez, N. & Lavergne, S. (2012)
Reconstructing the origins of high-alpine niches and cushion life form in the genus
Androsace s.l. (Primulaceae). Evolution, 66, 1255–1268.
Cholewa, A.F. & Kelso, S. (2009) Primulaceae. Flora of North America (ed. by Flora of North
America Editorial Committee), pp. 257–264. New York.
Ferguson, K.I. (1972) Androsace L. Flora Europaea (ed. by T.G. Tutin, V.H. Heywood, N.A. Burges,
D.M. Moore, D.H. Valentine, S.M. Walters and D.A. Webb), pp. 20–23. Cambridge University Press, Cambridge.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G. & Jarvis, A. (2005) Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology, 25,
1965–1978.
Hopkins, D.M. (1967) The Bering land bridge. Stanford University Press, Palo Alto, CA.
Hu, C.M. & Kelso, S. (1996) Primulaceae. Flora of China (ed. by Z-Y. Wu and P.H. Raven), pp. 118–
119. Science Press, Beijing.
Kress, A. (1997) Androsace L. Ebenaceae–Saxifrgaceae. Flora ibérica, Vol. 5. (ed. by S. Castroviejo,
C. Aedo, M. Laínz, R. Morales, F. Muñoz Garmendia, G. Nieto Feliner and J. Paiva), pp. 22–
40. Real Jardín Botánico, CSIC, Madrid.
Nasir, Y.J. (1984) Androsace. Flora of Pakistan (ed. by E. Nasir and S.I. Ali), p. 74. University of
Karachi, Karachi.
Ree, R.H. & Smith, S.A. (2008) Maximum likelihood inference of geographic range evolution by
dispersal, local extinction, and cladogenesis. Systematic Biology, 57, 4–14.
Rice, W.R. (1989) Analyzing tables of statistical tests. Evolution, 43, 223–225.
Shishkin, B.K. & Bobrov, E.G. (1952) Androsace. Flora of the U.S.S.R. (Flora SSSR) (ed. by B.K.
Shishkin and E.G. Bobrov), pp. 217–242. Akademii Nauk SSSR, Leningrad.
Struwe, L., Smouse, P.E., Heiberg, E., Haag, S. & Lathrop, R.G. (2011) Spatial evolutionary and ecological vicariance analysis (SEEVA), a novel approach to biogeography and speciation research, with an example from Brazilian Gentianaceae. Journal of Biogeography, 38, 1841–
1854.
Wen, J. (1999) Evolution of eastern Asian and eastern North American disjunct distributions in
flowering plants. Annual Review of Ecology and Systematics, 30, 421–455.
Download