Biogeography of Caribbean weevils highlights the importance of founder-event speciation - Guanyang Zhang
Download as PPTX, PDF0 likes1,272 views
The document discusses model-based biogeographic inference methods, emphasizing the addition of founder-event speciation and jump dispersal in analyzing Caribbean biodiversity, specifically within the Curculionidae family. It highlights the significance of different biogeographic models (DIVA, DEC, BayArea, and Biogeobears) in understanding ancestral ranges and the role of inter-island dispersal and in situ diversification as primary drivers of speciation. Key findings illustrate that incorporating jump dispersal leads to more accurate estimations of ancestral range evolution and supports the notion that islands contribute to continental diversity.
Biogeography of Caribbean weevils highlights the importance of founder-event speciation - Guanyang Zhang
- 1. Guanyang (GY) Zhang @GYZhang2 Usmaan Barashat, Nick Matzke & Nico Franz Arizona State University taxonbytes.org somanyinsects.org Society for Systematic Biologists 2015 Standalone Meeting, Ann Arbor, Michigan 05/21/2015 © Carlos De Soto Molinari
- 2. Model-based biogeographic inference methods (e.g., DEC, DIVA, BayArea) assume certain range-evolution processes. Dispersal Process Before After Ranges Modified from Matzke (2013) Frontiers of Biogeography A A B B A, B—area
- 3. Model-based biogeographic inference methods (e.g., DEC, DIVA, BayArea) assume certain range-evolution processes. Dispersal Process Before After Ranges Modified from Matzke (2013) Frontiers of Biogeography Extinction A A A A B B B B A, B—area
- 4. Model-based biogeographic inference methods (e.g., DEC, DIVA, BayArea) assume certain range-evolution processes. Process Before After Ranges Modified from Matzke (2013) Frontiers of Biogeography Range copying/ Sympatry A A A Tree A A A A, B—area
- 5. Model-based biogeographic inference methods (e.g., DEC, DIVA, BayArea) assume certain range-evolution processes. Process Before After Ranges Modified from Matzke (2013) Frontiers of Biogeography Vicariance A B AB B B BA A A Tree A, B—area
- 6. Matzke (2014 Syst Biol) added founder-event speciation process into existing models of geographic range evolution. Founder-event speciation/ Jump dispersal (+J) DEC—Dispersal-Extinction-Cladogenesis (Ree et al., 2005; Ree & Smith, 2008) DIVA—DIspersal-Vicariance Analysis (Ronquist, 1997) BayArea—Bayesian biogeographic analysis (Landis et al., 2013) A A B B Process Before After Ranges Tree A B A
- 7. BioGeoBEARS implements several models, adds founder-event jump dispersal (J) and enables model selection. Different biogeographic models (DIVA, DEC, BayArea, BioGeoBEARS supermodel) allow a certain set of biogeographic processes. BioGeoBEARS specifies each process with a parameter that can be turned on or off, including “jump dispersal (j)”. Matzke (2013) Frontiers of Biogeography Processes Models BioGeo BEARS
- 8. High levels of diversity and single-island endemicity in the Caribbean. Ideal for testing for founder-event speciation & studying Caribbean biogeography. • Curculionidae: Entiminae (broad-noses) • Generalist plant-feeders • Larvae (aka caterpillar) live in soil • Caribbean: ~85 species; >95% endemic • Central America: ~47 spp. • South America: ~12 spp. E. nicaraguensis Bovie E. quadrivittatus (Olivier) E. quinquedecimpunctatus (Olivier) E. roseipes (Chevrolat) Images and specimen data @
- 9. 65 ingroup species, 26 outgroup. Sampled major Greater Antillean islands, Lesser Antilles, and Central America. 6 genes, ~4800 bp. Phylogenetic reconstruction and fossil-calibrated molecular dating with BEAST v1.8 . Three fossil calibration points
- 10. What are the effects of modeling founder-event jump dispersal (J) on ancestral range evolution estimation? Most probable ancestral ranges were inferred to be a single area. DIVA+J DIVA (w/o J) Ancestral ranges as a single area Ancestral ranges widespread at some nodes
- 11. What are the effects of modeling founder-event jump dispersal (J) on ancestral range evolution estimation? Most probable ancestral ranges were inferred to be a single area. Jump dispersal accounted for virtually all range evolution events. DIVA+JDIVA+J Dispersal Sympatry (subset) Vicariance Vicariance Vicariance Extinction A variety of range-evolution processes Jump dispersal DIVA (w/o J) Only jump dispersal was inferred
- 12. Do biogeographic models with founder-event jump dispersal (J) provide a better fit? Model Parameter estimates Likelihood- ratio test Ln L d e j P DIVALIKE -163.1 0.0052 0.0010 0.000 DIVALIKE+J -115.4 0.0000 0.0000 0.063 1.5E-22 Parameter estimates Likelihood- ratio test AIC analysis Ln L d e j P weight ratio -163.1 0.0052 0.0010 0.000 -115.4 0.0000 0.0000 0.063 1.5E-22 2.02E+20 Strong support from statistical tests Adding jump dispersal led to much higher log likelihood
- 13. Inter-island dispersal and within-area in situ diversification were the main drivers of speciation. Jump dispersal occurred at 25% of all ancestral nodes (n=64), and range-copying (viz. in situ diversification) ~75%. Cuba Cuba (in situ diversification) 25 MYA
- 14. Neotropical mainland was colonized by Caribbean island species in the early Miocene (~18 MYA) Caribbean Central America
- 15. Current finding corroborates an emerging idea that islands can serve as a source for continental diversity. Bellemain & Ricklefs (2008) TREE
- 16. Biogeographic patterns in Caribbean weevils are reminiscent of those found in Caribbean Anolis lizards, frogs and snakes. Extensive in situ diversification, limited inter-island dispersal and/or island-continent reverse colonization. Alföldi et al. (2011) Nature MainlandCaribbean Caribbean 93-species Anolis phylogeny
- 17. Biogeographic patterns in Caribbean weevils are reminiscent of those found in Caribbean Anolis lizards, frogs and snakes. Heinicke, Duellman & Hedges (2007) PNAS - eleutherodactyline frogs Hedges, Couloux & Vidal (2009) Zotaxa - alsophiine snakes Credit: M. Lammertink http://science.psu.edu/ Credit: S. B. HedgeS
- 18. NSF DEB-1155984 USDA (Agreement No. 58-1275-1- 335 Anyi Mazo Vargas, Dr. Robert Anderson (sequences/specimens) Dr. Steve Davis (AMNH), Dr. Conrad Labandeira (USNM) (fossils) Albert Deler Hernandez, Franklyn Cala Riquelme (field assistance in Cuba)
Editor's Notes
- #3: This slide 12; cumulative 12 Model-based historical biogeographic inference methods each assume a certain set of range-evolution processes, such as dispersal…
- #4: This slide 3; cumulative 15 Extinction [1000, 2000]
- #5: This slide 6; cumulative 21 Range-copying or sympatry [1000, 2000]
- #6: This slide 3; cumulative 24 And vicariance [1000, 2000]
- #7: This slide 30; cumulative 54 In a recent study published in “Systematic Biology” Nick Matzke added founder-event speciation as a new biogeographic process into models of range evolution, which is also called jump dispersal. At cladogenesis one daughter lineage colonizes a new range while the other inherits the ancestral range.
- #8: This slide 12, cumulative 66 The R package BioGeoBEARS implements previous processes as well as jump dispersal. This package also allows you to statistically test biogeographic models just like selecting DNA substitution models.
- #9: This slide 33, cumulative 99 s My study organisms are broad-nosed weevils in the Exophthalmus genus complex. More than 80 species have been described from the Caribbean and nearly all of them are endemic to a single island. This group represents an ideal case for testing for founder-event speciation as well as studying aspects of Caribbean biogeography.
- #10: This slide 10 s, cumulative 109 s I reconstructed a dated molecular phylogeny. The specimen samples covered major Caribbean islands as well as the mainland.
- #11: This slide 22, cumulative 131 s The first question I asked was “What are the effects of modeling founder-event jump dispersal on ancestral range evolution estimation?” For simplicity, let’s focus on just part of the tree. On the left, with jump dispersal the most probable ancestral ranges were always estimated to be a single area. This contrasts to the widespread ancestors in the inference without jump dispersal, shown on the right.
- #12: This slide 12, cumulative 143 Besides, on the left all the biogeographic processes were jump dispersal events. Whereas without jump dispersal, I would have to infer several range-evolution processes, some of which require ad hoc explanations.
- #13: This slide 20s, cumulative 163 s Do models with jump-dispersal provide a better fit? First, inferences with jump dispersal had log likelihood values higher by nearly 50 units. Second, two statistical tests, likelihood-ratio test and Akaike information criterion weights, both strongly favored models with jump dispersal.
- #14: This slide 9, cumulative 192 s Based on the biogeographic inference with jump dispersal, it appears that inter-island dispersal and in situ diversification were the main driving forces of speciation in these weevils. However, jump dispersal occurred infrequently. In contrast, range-copying was inferred at most of the nodes, leading to extensive within-area in situ diversification. For instance, at least two radiations occurred within the island of Cuba during the last 10 Million years.
- #15: This slide 10 s, cumulative 202 s I also recovered a rather intriguing result. Central America was colonized by a Caribbean ancestor, leading to a large continental radiation.
- #16: This slide 9, cumulative 219 This result corroborates an emerging idea that islands could serve as a source for continental diversity.
- #17: This slide 22, cumulative 241 To further put my study into context, the various biogeographic patterns I recovered in these weevils, namely extensive in situ diversification, limited inter-island dispersal, and island-to-continent reverse colonization, are quite similar to those found in the Anolis lizards…
- #18: This slide 22, cumulative 241 eleutherodactyline frogs and alsophiine snakes. Hence, I provide further evidence towards generalizing theories of Caribbean biogeography.
- #19: This slide 6 s, cumulative 247 s. With that I would like to thank NSF and USDA for funding and several colleagues. Thank you for your attention and I will entertain questions.