Limits...
Consistent phenological shifts in the making of a biodiversity hotspot: the Cape flora.

Warren BH, Bakker FT, Bellstedt DU, Bytebier B, Classen-Bockhoff R, Dreyer LL, Edwards D, Forest F, Galley C, Hardy CR, Linder HP, Muasya AM, Mummenhoff K, Oberlander KC, Quint M, Richardson JE, Savolainen V, Schrire BD, van der Niet T, Verboom GA, Yesson C, Hawkins JA - BMC Evol. Biol. (2011)

Bottom Line: Despite ample evidence on such timescales for local adaptations of populations at specific sites, the long-term impacts of such changes on evolutionary significant units in response to past climatic change have been little documented.Shifts in climate tolerance appear to have been more important in this flora than is currently appreciated, and lineages that underwent such shifts went on to contribute a high proportion of the flora's extant species diversity.That shifts in phenology, on an evolutionary timescale and on such a scale, have not yet been detected for other floras is likely a result of the method used; shifts in flowering phenology cannot be detected in the fossil record.

View Article: PubMed Central - HTML - PubMed

Affiliation: School of Biological Sciences, Lyle Tower, University of Reading, Whiteknights, Reading RG6 6BX, UK. ben.warren@cirad.fr

ABSTRACT

Background: The best documented survival responses of organisms to past climate change on short (glacial-interglacial) timescales are distributional shifts. Despite ample evidence on such timescales for local adaptations of populations at specific sites, the long-term impacts of such changes on evolutionary significant units in response to past climatic change have been little documented. Here we use phylogenies to reconstruct changes in distribution and flowering ecology of the Cape flora--South Africa's biodiversity hotspot--through a period of past (Neogene and Quaternary) changes in the seasonality of rainfall over a timescale of several million years.

Results: Forty-three distributional and phenological shifts consistent with past climatic change occur across the flora, and a comparable number of clades underwent adaptive changes in their flowering phenology (9 clades; half of the clades investigated) as underwent distributional shifts (12 clades; two thirds of the clades investigated). Of extant Cape angiosperm species, 14-41% have been contributed by lineages that show distributional shifts consistent with past climate change, yet a similar proportion (14-55%) arose from lineages that shifted flowering phenology.

Conclusions: Adaptive changes in ecology at the scale we uncover in the Cape and consistent with past climatic change have not been documented for other floras. Shifts in climate tolerance appear to have been more important in this flora than is currently appreciated, and lineages that underwent such shifts went on to contribute a high proportion of the flora's extant species diversity. That shifts in phenology, on an evolutionary timescale and on such a scale, have not yet been detected for other floras is likely a result of the method used; shifts in flowering phenology cannot be detected in the fossil record.

Show MeSH

Related in: MedlinePlus

Phylogeny of Moraea showing reconstructed shifts in flowering duration and distribution under an ML model. All species are in the genus Moraea except where otherwise indicated. Coloured balls at terminals indicate the flowering duration of species, while equivalent coloured pie diagrams at each internal node depict the proportional likelihood for different flowering durations. One month, white; 2 months, blue; 3 months, red; 4 months, green; 8 months, black. Nodes reaching the threshold of two log-likelihood units separating the flowering duration of highest likelihood from alternative flowering durations are marked with an asterisk. Bars attached to the left side of nodes indicate significant support (two log-likelihood units separation) for a reconstructed presence in the west; those to the right side of nodes indicate significant support for a reconstructed presence in the east. Internal nodes with no horizontal bars are those for which the reconstructed distribution is not significantly supported. A vertical bar below a node indicates 80-100% bootstrap support; a bar above the node indicates 50-79% bootstrap support; no vertical bar indicates 0-49% bootstrap support.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3045326&req=5

Figure 2: Phylogeny of Moraea showing reconstructed shifts in flowering duration and distribution under an ML model. All species are in the genus Moraea except where otherwise indicated. Coloured balls at terminals indicate the flowering duration of species, while equivalent coloured pie diagrams at each internal node depict the proportional likelihood for different flowering durations. One month, white; 2 months, blue; 3 months, red; 4 months, green; 8 months, black. Nodes reaching the threshold of two log-likelihood units separating the flowering duration of highest likelihood from alternative flowering durations are marked with an asterisk. Bars attached to the left side of nodes indicate significant support (two log-likelihood units separation) for a reconstructed presence in the west; those to the right side of nodes indicate significant support for a reconstructed presence in the east. Internal nodes with no horizontal bars are those for which the reconstructed distribution is not significantly supported. A vertical bar below a node indicates 80-100% bootstrap support; a bar above the node indicates 50-79% bootstrap support; no vertical bar indicates 0-49% bootstrap support.

Mentions: Our phylogenetic reconstructions demonstrate that 9 clades have undergone shifts in flowering pattern consistent with predictions based on past climate change (Figure 2; Table 1). How the number of shifts per Cape clade are counted depends on the treatment of nodes for which character states are unresolved; we use the basal-most possible position of shifts (as in additional file 1: AF1.pdf) in order to produce figures that are comparable across Cape clades. On this basis, shifts in flowering pattern consistent with predictions consist of one (Figure 2, and see AF1.pdf, Trees 5, 8, 10, 12 &13), two (AF1.pdf, Tree 6) or four (AF1.pdf, Tree 5) shifts in flowering duration per Cape clade, and one (AF1.pdf, Trees 3, 5, 6 & 9) or two (AF1.pdf, Tree 5) shifts in the timing of flowering per Cape clade, making a total of eleven shifts in flowering duration and six shifts in the timing of flowering detected across the sampled flora. By contrast, four clades have undergone shifts in flowering pattern contrary to predictions (Table 1, additional file 2: AF2.pdf); these consist of one (AF1.pdf, Trees 5 & 10) and two (AF1.pdf, Tree 14) shifts in flowering duration per Cape clade, and one shift in the timing of flowering per Cape clade (AF1.pdf, Trees 5, 10 & 12), making a total of four shifts in flowering duration and three shifts in the timing of flowering contrary to predictions detected across the flora.


Consistent phenological shifts in the making of a biodiversity hotspot: the Cape flora.

Warren BH, Bakker FT, Bellstedt DU, Bytebier B, Classen-Bockhoff R, Dreyer LL, Edwards D, Forest F, Galley C, Hardy CR, Linder HP, Muasya AM, Mummenhoff K, Oberlander KC, Quint M, Richardson JE, Savolainen V, Schrire BD, van der Niet T, Verboom GA, Yesson C, Hawkins JA - BMC Evol. Biol. (2011)

Phylogeny of Moraea showing reconstructed shifts in flowering duration and distribution under an ML model. All species are in the genus Moraea except where otherwise indicated. Coloured balls at terminals indicate the flowering duration of species, while equivalent coloured pie diagrams at each internal node depict the proportional likelihood for different flowering durations. One month, white; 2 months, blue; 3 months, red; 4 months, green; 8 months, black. Nodes reaching the threshold of two log-likelihood units separating the flowering duration of highest likelihood from alternative flowering durations are marked with an asterisk. Bars attached to the left side of nodes indicate significant support (two log-likelihood units separation) for a reconstructed presence in the west; those to the right side of nodes indicate significant support for a reconstructed presence in the east. Internal nodes with no horizontal bars are those for which the reconstructed distribution is not significantly supported. A vertical bar below a node indicates 80-100% bootstrap support; a bar above the node indicates 50-79% bootstrap support; no vertical bar indicates 0-49% bootstrap support.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3045326&req=5

Figure 2: Phylogeny of Moraea showing reconstructed shifts in flowering duration and distribution under an ML model. All species are in the genus Moraea except where otherwise indicated. Coloured balls at terminals indicate the flowering duration of species, while equivalent coloured pie diagrams at each internal node depict the proportional likelihood for different flowering durations. One month, white; 2 months, blue; 3 months, red; 4 months, green; 8 months, black. Nodes reaching the threshold of two log-likelihood units separating the flowering duration of highest likelihood from alternative flowering durations are marked with an asterisk. Bars attached to the left side of nodes indicate significant support (two log-likelihood units separation) for a reconstructed presence in the west; those to the right side of nodes indicate significant support for a reconstructed presence in the east. Internal nodes with no horizontal bars are those for which the reconstructed distribution is not significantly supported. A vertical bar below a node indicates 80-100% bootstrap support; a bar above the node indicates 50-79% bootstrap support; no vertical bar indicates 0-49% bootstrap support.
Mentions: Our phylogenetic reconstructions demonstrate that 9 clades have undergone shifts in flowering pattern consistent with predictions based on past climate change (Figure 2; Table 1). How the number of shifts per Cape clade are counted depends on the treatment of nodes for which character states are unresolved; we use the basal-most possible position of shifts (as in additional file 1: AF1.pdf) in order to produce figures that are comparable across Cape clades. On this basis, shifts in flowering pattern consistent with predictions consist of one (Figure 2, and see AF1.pdf, Trees 5, 8, 10, 12 &13), two (AF1.pdf, Tree 6) or four (AF1.pdf, Tree 5) shifts in flowering duration per Cape clade, and one (AF1.pdf, Trees 3, 5, 6 & 9) or two (AF1.pdf, Tree 5) shifts in the timing of flowering per Cape clade, making a total of eleven shifts in flowering duration and six shifts in the timing of flowering detected across the sampled flora. By contrast, four clades have undergone shifts in flowering pattern contrary to predictions (Table 1, additional file 2: AF2.pdf); these consist of one (AF1.pdf, Trees 5 & 10) and two (AF1.pdf, Tree 14) shifts in flowering duration per Cape clade, and one shift in the timing of flowering per Cape clade (AF1.pdf, Trees 5, 10 & 12), making a total of four shifts in flowering duration and three shifts in the timing of flowering contrary to predictions detected across the flora.

Bottom Line: Despite ample evidence on such timescales for local adaptations of populations at specific sites, the long-term impacts of such changes on evolutionary significant units in response to past climatic change have been little documented.Shifts in climate tolerance appear to have been more important in this flora than is currently appreciated, and lineages that underwent such shifts went on to contribute a high proportion of the flora's extant species diversity.That shifts in phenology, on an evolutionary timescale and on such a scale, have not yet been detected for other floras is likely a result of the method used; shifts in flowering phenology cannot be detected in the fossil record.

View Article: PubMed Central - HTML - PubMed

Affiliation: School of Biological Sciences, Lyle Tower, University of Reading, Whiteknights, Reading RG6 6BX, UK. ben.warren@cirad.fr

ABSTRACT

Background: The best documented survival responses of organisms to past climate change on short (glacial-interglacial) timescales are distributional shifts. Despite ample evidence on such timescales for local adaptations of populations at specific sites, the long-term impacts of such changes on evolutionary significant units in response to past climatic change have been little documented. Here we use phylogenies to reconstruct changes in distribution and flowering ecology of the Cape flora--South Africa's biodiversity hotspot--through a period of past (Neogene and Quaternary) changes in the seasonality of rainfall over a timescale of several million years.

Results: Forty-three distributional and phenological shifts consistent with past climatic change occur across the flora, and a comparable number of clades underwent adaptive changes in their flowering phenology (9 clades; half of the clades investigated) as underwent distributional shifts (12 clades; two thirds of the clades investigated). Of extant Cape angiosperm species, 14-41% have been contributed by lineages that show distributional shifts consistent with past climate change, yet a similar proportion (14-55%) arose from lineages that shifted flowering phenology.

Conclusions: Adaptive changes in ecology at the scale we uncover in the Cape and consistent with past climatic change have not been documented for other floras. Shifts in climate tolerance appear to have been more important in this flora than is currently appreciated, and lineages that underwent such shifts went on to contribute a high proportion of the flora's extant species diversity. That shifts in phenology, on an evolutionary timescale and on such a scale, have not yet been detected for other floras is likely a result of the method used; shifts in flowering phenology cannot be detected in the fossil record.

Show MeSH
Related in: MedlinePlus