Diversification and Evolutionary Dynamics in Tropical Montane Regions

Anolis heterodermus. This Andean species can live up to 3700 meter above sea level! Photo by Jhan C. Salazar.

Much like when I was an undergrad, I found myself trying to decide what to do for my PhD thesis, as I had to change or rethink most of it after COVID happened. Months before my proposal, I came across a really cool paper by someone from my hometown (Cali), Julian A. Velasco (Velasco et al. 2020). In this paper, Velasco and colleagues addressed two questions: first, whether common ideas about heat and food availability explain why animal sizes vary geographically; and second, how fast those sizes changed as the planet cooled over the last 66 million years. They used a massive dataset containing georeferenced information for 379 species of anoles.

When it came the time to write and present my proposal in 2021, I wrote a chapter based on Julian’s data. However, I didn’t know Julian at the time, nor did I have the original dataset of nearly 25,000 georeferenced individuals. Luckily, the following year, the Colombian Congress of Herpetology was held at my alma mater in Cali, Universidad Icesi, and Julian was a keynote speaker. I emailed him a few months before the congress to set up a meeting and discuss the project I had in mind.

Although the paper I’m discussing today is related to that original study, it actually grew from a comment made by another collaborator, Adam Algar. During a revision, he asked: “Is it worth thinking about which processes have contributed to highland anole diversification or speciation?”

Now, let’s talk about highland anoles and diversification. For this study, I used the georeferenced records to extract two types of data from publicly available rasters: topographic complexity (measured as roughness) and past climate-change velocity (capturing the speed of local shifts in temperature and precipitation). Our study was recently published in Global Ecology and Biogeography; Salazar et al. 2026. We used topographic complexity and past climatic-change velocity to test two hypotheses:

High-elevation environments promote higher speciation rates.
Greater topographic complexity and climatic stability (low past climactic-change velocity) positively influence speciation.

We ran three main analyses using those data

CoMET – to test whether the observed signal could be explained by shifts in diversification rates across the phylogeny of extant species
GeoHiSSE – to test whether high-elevation habitats act as “species pumps”.
PGLS – to evaluate the relationship between topographic complexity, past climatic-change velocity, and diversification rate.

In total, we analyzed 303 species. One major challenge was defining the “highland” vs. “lowland” boundary, as there is no exact elevational threshold. We settled on 700m as our primary criterion, based on significant turnover in Caribbean anoles (Frishkoff et al. 2022), though we also tested 300m and 1500m thresholds for comparison.

But what did we find? We found that the pace of new species appearing actually slowed down as we got closer to the present (CoMET analysis). This slowdown occurred right when Central America started getting tectonically active and kept dropping until the Andes finished rising, at which point things finally stabilized especially after a big shift in the Miocene about 7 million years ago.

For the GeoHiSSE, we found that the ancestral Anolis was a “generalist” species (Figure 1) with high net diversification rate. The GeoHiSSE analysis also inferred that most transitions to highland habitats occurred during the late Eocene and the Oligocene (Figure 1b). Additionally, for speciation rate and net diversification, we found that lowland species had higher speciation rates and net diversification compared to their highland counterparts (Figure 1c).

Figure 1. (a; top) Tree-wide diversification rates over time, as inferred from the CoMET analysis. The purple line represents the estimated diversification rate (with the mean and 95% highest posterior density) based on the anole phylogeny show in (b). (a; bottom) Bayes Factor (BF) evidence for a shift in speciation rates across the tree. The dashed line marks the threshold for “positive” support (2 ln BF > 2; Kass and Raftery, 1995). (b) Ancestral range reconstruction of the GeoHiSSE results. The outlines of the branches are colored by net diversification rates, with white indicating low rates and red indicating high rates. The inner parts of the branches are colored by reconstructed ancestral states: lowland species (green), highland species (yellow), and widespread species (brown). Different clades of anole lizards are also shown. (c) Estimates of net diversification (top) and speciation (bottom) in highland than lowland areas.

Lastly, we found no significant correlation between topographic complexity or past climatic-change velocity and diversification rate (Figure 3a, 3b). However, we separated our 303 species into five clades, we found that for the species in the Draconura clade we found a significant correlation between topographic complexity and past climatic-change velocity with speciation rate (Figure 3c, 3d).

Figure 2. Relationships between mean speciation rate (MeanDR) with topographic complexity and climatic-change velocity, for all anole species (a, b), and the species from the Draconura clade (c, d). The black regression line represents the results of the phylogenetic generalized least squares (PGLS). Points represent individual species. Yellow points represent highland species, while the green points represent lowland species.

To answer our two questions:

Do high-elevation environments act as engines for faster speciation? No, we found that lowland species have higher speciation rates compared to highland species.
Does a combination of rugged topography and a stable, predictable climate lead to higher rates of new species forming? No, speciation rates are not related to greater topographic complexity, past climatic-change velocity, or mean annual temperature.

Our findings challenge assumptions about the direct role of topographic complexity in speciation, highlighting the need to consider multifaceted ecological and biogeographical factors driving evolutionary processes. By disentangling the relative contributions of climatic stability (i.e., lower past climatic-change velocity) and topographic heterogeneity, this study highlights the complex dynamics shaping biodiversity in tropical mountains and calls for further integrative approaches to understand species diversification in the face of climatic and geological change.

In our study, we treated all Neotropical mountains as uniform. Moving forward, research needs elevation thresholds that reflect regional realities rather than arbitrary numbers, accounting for local variations in climate and topography. A holistic approach, integrating climatic stability with topographic complexity, is essential to understanding how montane species evolve and adapt. By refining these boundaries, we can better predict how global climate change will shift evolutionary trajectories and prioritize conservation in these complex landscapes.

References

Frishkoff, L. O., Lertzman-Lepofsky, G., & Mahler, D. L. (2022). Evolutionary opportunity and the limits of community similarity in replicate radiations of island lizards. Ecology Letters, 25(11), 2384–2396.
Kass, R. E., & Raftery, A. E. (1995). Bayes factors. Journal of the American Statistical Association, 90(430), 773–795.
Salazar, J. C., Algar, A. C., Poe, S., Losos, J. B., & Velasco, J. A. (2026). Diversification and evolutionary dynamic in tropical montane regions. Global Ecology and Biogeography. 35(3), e70218.
Velasco, J. A., Villalobos, F., Diniz-Filho, J. A. F., Poe, S., & Flores-Villela, O. (2020). Macroecology and macroevolution of body size in Anolis lizards. Ecography, 43(6), 812-822.