Latitudinal Gradients in Climatic Niche Evolution€¦ · Latitudinal Gradients in Climatic Niche Evolution Adam Matthew Lawson Master of Science Department of Ecology and Evolutionary

Latitudinal Gradients in Climatic Niche Evolution

by

Adam Matthew Lawson

A thesis submitted in conformity with the requirements for the degree of Master of Science

Department of Ecology and Evolutionary Biology University of Toronto

© Copyright by Adam Matthew Lawson 2014

ii

Latitudinal Gradients in Climatic Niche Evolution

Adam Matthew Lawson

Master of Science

Department of Ecology and Evolutionary Biology

University of Toronto

2014

Abstract

Either tropical niche divergence or tropical niche conservatism could drive the latitudinal

diversity gradient. Greater niche divergence in the tropics could accelerate reproductive isolation

leading to more rapid species formation. Alternatively, latitudinal asymmetry in niche

conservatism, whereby tropical species are more conserved than high latitude species, could

promote more dispersal in to than out of the tropics, leading to greater tropical richness. Here I

test whether rates of climatic niche evolution vary across the latitudinal gradient for 164 closely

related pairs of species. Using the evolutionary ages at which sister species diverge, and the

niche divergence between them, I applied Brownian motion models to test whether rates of

climatic niche evolution varied with latitude. My results indicate that climatic niche conservatism

is strongest in the tropics. This suggests that the latitudinal diversity gradient is driven by the

inability of tropical to adapt to temperate climates and colonize non-tropical latitudes.

iii

Acknowledgments

I would like to thank Jason Weir for all of his help and guidance during my time as a graduate

student under his supervision.

iv

Table of Contents

Chapter 1

Latitudinal Gradients in Climatic Niche Evolution

pg. 1 - 12

1 Introduction: pg. 1 - 3

2 Methods: pg. 4 - 7

2.1 Data Collection: pg. 4 - 5

2.2 Data Analysis Evolutionary Rate: pg. 5 - 7

2.3 Data Analysis Climatic Specialization: pg. 7

3 Results: pg. 7 - 8

4 Discussion: pg. 9 - 12

Figures and Tables: pg. 13 - 27

References Cited: pg. 28 - 32

v

List of Tables

Table 1 pg. 13

vi

List of Figures

Figure 1 pg. 14

Figure 2 pg. 15

Figure 3 pg. 16

Figure 4 pg. 17 - 27

1

Chapter 1

Latitudinal Gradients in Climatic Niche Evolution

1 Introduction

The existence of the latitudinal diversity gradient is widely accepted, but its cause remains

unknown despite decades of analysis (Mittlebach et al. 2007). Hypotheses for the diversity

gradient generally focus on evolutionary or ecological explanations (Mittlebach et al. 2007).

Evolutionary explanations suggest that diversification rates are higher in the tropics, or that the

tropics have had more time in which diversification could occur (Fischer, 1960). Ecological

explanations often focus on the ecological niche, and factors that might allow for greater number

of niches, and thus species, in the tropics (Wiens and Donoghue 2004, Wiens et al. 2009).

Evolutionary and ecological explanations are not mutually exclusive. For example, the ecological

niche of species may evolve more rapidly at certain latitudes which may influence the latitudinal

diversity gradient. Both high tropical rates of evolution (tropical niche divergence) and low

tropical rates of evolution (tropical niche conservatism) have been proposed to act as drivers of

high tropical species richness (Kozak and Wiens, 2012). Niche divergence results in ecological

differentiation between closely related species through evolutionary time, allowing them to

exploit novel regions of ecological space (Schluter, 2000). Faster divergence in the tropics could

promote greater utilization of available niche space resulting in fast rates of speciation. In

contrast, niche conservatism describes a pattern of stasis in the ecological niche through

evolutionary time (Wiens and Graham 2005). If tropical species are highly conserved in their

niche through time, they might have difficulty adapting to freezing temperatures of temperate

latitudes, and thus exhibit poor reduced dispersal out of the tropics. Whether tropical niche

divergence or tropical niche conservatism drives the diversity gradient is unknown and

disentangling the contributions of each remains a challenge.

Niche divergence between species can promote both local and regional species richness.

At regional scales, strong ecological divergent selection drives diversifying populations into

different niches, where they evolve towards different ecological optima (Slabbekoorn and Smith

2001). This divergent evolution may result in rapid development of reproductive isolation

2

between populations that have adapted to different ecological conditions which, in turn, could

cause a fast rate of speciation (Rundle and Nosil 2005; Price 2008). For example, birds adapting

to different habitat types sing at different pitches (Weir et al. 2012), and these differences in song

can lead to reproductive isolation. Ecological divergence can also result in selection against

maladapted hybrids, resulting in post-zygotic isolation (Schulter 2009). At the local scale,

ecological niche divergence may allow for more rapid attainment of sympatry between newly

formed species (i.e. Weir & Price 2011), allowing for stable coexistence. If rates of ecological

niche divergence were to differ latitudinally, they could provide an explanation for the origin of

the latitudinal richness gradient. For example, more rapid ecological divergence in the tropics

could accelerate the attainment of reproductive isolation and sympatric coexistence, leading to

high tropical species richness (Mittlebach et al. 2007).

Niche conservatism could also promote elevated tropical species richness (Mittlebach et

al. 2007). Strong niche conservatism hinders dispersal into regions with novel climates. Recent

studies suggest an asymmetry in dispersal ability with high latitude lineages more readily

colonizing the tropics than the reverse (Smith and Klicka 2010; Smith et al. 2012; Weir et al

2009). Such asymmetrical dispersal could be driven by a number of factors including strong

asymmetries in niche conservatism. As an example, physiological intolerance to temperate

climates may prevent tropical species from expanding into temperate latitudes (Sunday et al

2012; Smith et al. 2012). In contrast, high latitude species are more likely to be adapted to a wide

range of climatic conditions, and as a result may more easily colonize the tropics (Addo-Bediako

et al. 2000; Smith et al. 2012). All else being equal (i.e. diversification rates, etc.) this greater

ease of colonization of the tropics would result in a latitudinal diversity gradient.

Climate is an easily quantifiable aspect of the ecological niche that varies between

geographic localities. The climatic niche is defined as the set of climatic conditions (eg.

temperature, precipitation, seasonality) a species experiences across its natural geographic range

(Hutchinson 1957). During the diversification process there appears to be extensive opportunities

for climatic niche divergence across latitude. For example, both equatorial and boreal latitudes

consist of many habitat types ranging from rainforest to alpine grassland, all of which differ in

climate (Olsen et al. 2001). Despite this, available climatic niche space may not be utilized

equally across latitude during diversification. As an example, generalists have been found to

3

occupy more climatic niche space than specialists (Devictor et al. 2010, Peers et al. 2012). If the

proportion of specialist to generalist clades varies between tropical and extra-tropical regions,

then a latitudinal gradient of climatic niche divergence may result.

Rates of climatic niche evolution have not previously been quantified across latitude.

Previous authors have instead, focused on climatic niche overlap between closely related species.

These studies have produced contradictory results (Hua and Wiens 2010) with reduced niche

overlap occurring in the tropics (Plethodontid salamanders: Kozak & Wiens 2007), at high

latitudes (various high elevation vertebrate groups; Cadena et al 2011), or with no latitudinal

differences detected (frogs; Hua and Wiens 2010). What is now needed are studies that quantify

rates of climatic niche evolution across latitudinal gradients. To do this, the amount of niche

divergence between species must be corrected for the age at which pairs of species diverged from

a common ancestor.

Here I test whether rates of climatic niche evolution vary across the latitudinal gradient

for closely related pairs of New World bird and mammal species. I quantify climatic niche

evolution as a function of the amount of climatic divergence between sister pairs and their time

to most recent common ancestry. Brownian motion and Ornsetin-Ulhenbeck models are used to

test whether rates of climatic niche evolution vary as a function of latitude. In addition, the

climatic ranges of species are calculated, in order to determine whether tropical species are more

climatically specialized than temperate species. If rates of climatic niche divergence are found to

be greatest in the tropics, insipient tropical species would likely become reproductively isolated

rapidly, due to adaptation to different ecological conditions. Under this scenario the tropics

would act as a speciation pump, resulting in high tropical species richness. Alternatively, if rates

of climatic niche evolution are found to be slowest in the tropics, tropical species would likely be

highly conserved in their climatic niche. This strong niche conservatism may cause tropical

species to be intolerant to freezing temperature disallowing colonization into extra tropical

regions. Under this alternative scenario, the inability for tropical species to disperse outside of

the tropics would result in high tropical species richness and drive the latitudinal gradient of

species diversity.

4

2 Methods

2.1 Data Collection

Published molecular phylogenies were used to obtain 111 avian, and 53 mammalian New

World sister pairs. Sister pairs included both sister species and phylogroup splits within species

(sister pairs were never phylogenetically nested within other sister pairs, and are each statistically

independent contrasts). For each sister species, GTR- distances were calculated in PAUP

4.0b10 (Swofford 2002) from cytochrome b sequences obtained from Genbank (see Figure 4 for

accession numbers). Avian and mammalian sister pairs were included if cytochrome b sequences

(of at least 500 base pairs in length) were available for both members of a sister pair, and if

GTR- distances exceed 0.75 percent divergence. Sister species with GTR- distances less than

0.75 percent divergence were excluded because the stochastic nature of mutation along a short

DNA sequence renders time estimates highly inaccurate for young sister species. In order to

determine the relative age of each sister pair, relaxed clock phylogenies were created separately

for birds and mammals in BEAST version1.7.5 (Drummond et al. 2012) using cytochrome B

sequences, and a lognormal relaxed clock (using a Yule speciation prior) with a GTR- model.

Tree topologies were fixed (using Baker (2004) for higher level relationships in birds and

Bininda-Emonds (2007) in mammals, and using a large number of published molecular

phylogenies for relationships between species and genera) and BEAST was used to estimate

branch lengths only. BEAST phylogenies were not time calibrated, and node ages represented

relative rather than absolute time. For both birds and mammals, Bayesian analyses were run for

20 million generations and trees were sampled every 1000 generations, following a burn-in of 10

million generations. Maximum clade credibility trees with median node heights were generated

in TreeAnnotator v1.7.5 (Drummond et al. 2012) (Figure 4). The relative sister pair ages were

obtained from branch lengths along this tree All latitudinal values were obtained from range

maps located on Natureserve (birds: Ridgley et al. 2003; mammals: Patterson et al. 2007, IUCN

2012). Midpoint latitude was used to quantify the latitudinal position of a sister pair. Midpoint

latitude of a sister pair was calculated as the mean absolute midpoint latitude of each sister’s

breeding range. Sister pairs were excluded if the midpoint latitudes of both sisters in a pair

differed by more than 25°, and if the combined latitudinal distribution of both species in a pair

was greater than 45° absolute latitude. This exclusion removed sister pairs with very wide

latitudinal distributions.

5

For each species, climatic data was obtained for each georeferenced coordinate of known

occurrence. An alternative approach would systematically sample across a species range map

(e.g. Botero et al 2013), but published range maps, especially from the tropics, are highly

inaccurate (Jetz et al. 2008). In total ~ 275,000 locality records (latitude and longitude) were

obtained from ORNIS 2 (http://ornis2.ornisnet.org/), MaNIS (http://manisnet.org/) and the

Global Biodiversity Information Facility (GBIF) (http://data.gbif.org/welcome.htm). On a

species level, individual coordinates were excluded if they did not occur within a species range,

as specified by Natureserve range maps. With the exception of species known from less than 5

localities, species (and the resulting sister pair) were excluded if less than 5 unique locality

records were obtained. Climatic niche was calculated for only allopatric and parapatric sister

pairs, because sympatric sister pairs occupy the same geographic range and experience similar

climates.

Climatic information was obtained from WorldClim (http://www.worldclim.org/), a

database that integrates climatic data from a global distribution of weather stations (Hijmans et

al. 2005). Climatic data was obtained separately for each coordinate using a global resolution of

30 arc seconds (~ 1km2). I chose 48 variables to quantify the overall climatic conditions a species

encounters. These included 36 temperature variables (maximum, minimum, and mean for each

month of the year) and 12 precipitation variables (mean for each month of the year). Because

seasonality is not synchronous in different geographic regions, the monthly maximum,

minimum, and mean temperatures and precipitations were each sorted from highest to lowest

values for each locality. This reduces overestimation of climatic divergence between sister pairs

with asynchronous seasonality.

2.2 Data Analysis: Evolutionary Rate

To quantify climatic niche, principal component analyses (PCA hereafter) were

performed on all log transformed climatic variables for all sample localities in a species range.

Separate PCA analyses were performed on birds and mammals. For each of the first three

principal components (PC’s), the midpoint value along that PC was calculated for each species. I

used midpoints rather than means to represent the climatic niche of each species, because means

6

are more heavily skewed by densely surveyed regions within a species range. Euclidean distance

between the midpoints of species within a sister pair were used as a measure of climatic

divergence between sister pairs.

Both Brownian motion (BM) and Ornstein-Uhlenbeck (OU) models were used to

compare evolutionary rates of climatic niche across the latitudinal gradient. The BM model

estimates a single parameter, the evolutionary rate, β, and assumes no limit on trait divergence

(i.e. climatic niche divergence) through time. The OU model adds an additional parameter, α,

which acts as a constraint on trait divergence so that trait divergence cannot increase indefinitely

through time. Under the OU model, α represents a “pull” towards an intermediate trait value

between each sister pair (assumed to be the ancestral trait value). This prevents strong trait

divergence away from the ancestral trait value and acts as an evolutionary constraint . An OU

model is appropriate when trait space is finite as is expected to be true for climatic variables. At

high values of α (i.e. a high constraint), evolutionary divergence becomes difficult. As α

approaches 0, the OU model collapses to the simpler BM model in which evolution is not

constrained. To determine if latitude drives rates of climatic niche evolution, I use maximum

likelihood (MLE) to fit of a BM and OU models in which a single rate (and constraint for OU) of

evolution was estimated for all sister pairs, to models in which evolutionary rate (and constraint

for OU) was allowed to vary linearly with latitude or to change after a latitudinal breakpoint. If

the climatic dataset best supports a model where evolutionary rate strongly increases or decreases

across latitude there would be verification of for tropical niche conservatism and tropical niche

divergence, respectively. All MLE analyses were performed in R using the package EvoRAG

(Weir 2014). The likelihood functions for the single rate, linear and breakpoint models are

formulated here:

( ) ∏

√ (

) ,

(BM null model) = ,

(BM model with latitude) = ( ),

(BM two rate model) = ( ( ) ( ))

(OU model) = ( α)(1-exp(α )),

7

(OU mode with latitude) = (( ) ( ))(1-exp(α )),

(OU two rate model) = ( α)(1-exp(α )), where ( )

( ) ( ) ( )

where D is the Euclidean distance between species pairs, T is the relative age of each sister pair,

bβ and bα are the slope, and cβ, and cα are the intercept parameters describing the linear change

of β and α respectively as a function of Latitude (Lat). Model fit was determined using Akaike

Information Criterion (AICc). The best fit model is considered to be the one with the lowest

AICc.

Maximum likelihood analyses were performed on an additional dataset to determine

whether avian results were an artifact of high latitude migration. The dataset used only

temperate, northern hemisphere sister pairs from the original avian dataset and used climatic data

from the months of April to September (a rough estimation of the high latitude breeding season).

2.3 Data Analysis: Climatic Specialization

To determine whether tropical species are more climatically specialized than temperate

species, the climatic ranges of each species were calculated for PC1 to PC3. Specialized species

are expected to have lower climatic ranges than generalist species. Pagel’s lambda (Pagel 1999)

was used to determine if the climatic ranges of species increased with latitude, while correcting

for phylogenetic relatedness between species. The phylogenies in Figure 4 were used for

phylogenetic regression, and Pagel’s lambda was performed using the APE package (Paradis et

al. 2004) in R.

3 Results

The first three PCs explained 94 and 96 percent of the variance in birds (PC1 61%, PC2

20% PC3 13%) and mammals (PC1 71%, PC2 20% PC3 5%), respectively. Rates of climatic

niche evolution were estimated separately for each of the first three PC’s. All three PCs represent

nearly identical aspects of climate in both birds and mammals (fig. 1a, 1b). PC1 represents

8

temperature, PC2 represents precipitation and PC3 represents seasonality of both temperature

and precipitation (fig. 1).

Euclidean distances are shown in figure 2 for PCs that best support a model where

climatic niche evolution changes across latitude. For the avian dataset, maximum likelihood

results for PC1, PC2, and PC3 supported models in which rates of climatic niche divergence

increased with latitude (Table 1a). PC1 best supported an OU linear model where alpha

decreased across latitude (α at 0° latitude = 1.73, α at 60° latitude = 1.13) and beta increased

across latitude (β at 0° latitude = 29.57, β at 60° latitude = 144.54) (Fig 2a). PC1 did not provide

significantly less support for the OU two rate model (AICc = 1.45). Here, both alpha and beta

increased at 39° latitude. The best supported model for PC2 was an OU linear two rate model,

where the breakpoint was calculated at 5 degrees latitude. Because this breakpoint does not

represent a division between tropical and temperate regions, PC2 warrants no further analysis.

For PC3, the best supported model was the OU linear model. Here, evolution rate increased

across latitude (β at 0° latitude = 0.66, β at 60° latitude = 9.92) while evolutionary constraint

decreased across latitude (α at 0° latitude = 0.77, α at 60° latitude = 0.23) (Fig 2b).

For mammals, PC3 best supported a model where latitude increased linearly, while PC1

and PC2 best supported OU constant (null) models. The best fit model for PC3 was a BM

constant two rate model, with a breakpoint calculated at 24 degrees latitude (β at 0-23° latitude =

0.27, β at 24-60° latitude = 0.80) (Fig 2c).

For the migration control test (North American temperate bird data for summer months

only) PC1 supported a model in which climatic niche divergence increases across latitude while

PC2 and PC3 supported null models where climatic niche divergence did not change latitudinally

(Table 1c). The best fit model for PC1 was a BM constant two rate model, with a breakpoint

calculated at 23 degrees latitude (β at 0-23° latitude = 3.25, β at 24-60° latitude =26.16).

Climatic range size variation across latitude is displayed in Figure 3. The climatic ranges

of mammals increased significantly across latitude within PC1 (slope = 0.021, p = 0.004), but the

result was not significant for PC2 (slope = 0.010, p = 0.130) or PC3 (slope = 0.015, p = 0.477).

For birds, the climatic ranges of PC1 (slope = 0.004, p < 0.001) and PC3 (slope = 0.015, p <

9

0.001) increased significantly across latitude while the climatic ranges within PC2 (slope =

0.006, p = 0.081) did not increase significantly across latitude (Figure 3b).

4 Discussion

My results indicate faster evolutionary divergence in climatic niche at high latitudes for

temperature (PC1) in birds and seasonality (PC3) in birds and mammals. In contrast, no climatic

axes of variation supported faster evolutionary divergence in the tropics. These results reject the

role of climatic niche divergence as a driver of high tropical species richness, and instead support

the role of niche conservatism in the tropics as a potential driver of the latitudinal diversity

gradient.

Strong climatic niche divergence in temperate species could be driven by extensive

climatic fluctuations at high latitudes that occurred during the Plio-Pleistocene glacial cycles

(Weir & Schluter 2004) and earlier periods. In contrast, the tropics have experienced less severe

climate fluctuations (Bush et al. 1990, Colinvaux et al. 1996). The comparatively greater stability

of paleoclimates in the tropics may have allowed for a high degree of specialization in climate,

with tropical species occupying narrow temperature ranges, and exhibiting reduced seasonality

(Janzen 1967). The reduced climatic ranges exhibited by tropical birds (for temperature and

seasonality) and mammals (temperature) support these predictions. Ecological specialization

reduces evolutionary divergence of climatic niche by limiting a species’ ability to readily adapt

to novel climatic conditions (Futuyma and Moreno 1988). For example, climatically specialized

tropical species may experience difficulty adapting physiologically to freezing conditions, and

are thereby limited in their ability to colonize temperate latitudes (Smith and Klicka 2010;

Hawkins et al. 2006). In contrast, increased seasonality at high latitudes promotes more climatic

niche generalization (Deutsch et al. 2008). This generalization allows high latitude species to

more easily adapt to tropical climates, increasing the probability of their colonization of tropical

regions. All else being equal, the resulting asymmetry in colonization ability between tropical

and temperate regions promotes the buildup of high tropical species richness. The tropical

conservatism hypothesis (Wiens & Donoghue 2004; Wiens et al. 2006; Mittlebach et al 2007)

posits that clades preferentially originate in the tropics and only rarely disperse out of the tropics

10

due to physiological limitations to freezing temperatures. The results of this study are consistent

with this hypothesis, but do not require that clades preferentially originate in the tropics. Rather,

the greater ease with which high latitude species colonize the tropics is sufficient to generate and

maintain a latitudinal diversity gradient.

The results of this study are highly congruent with New World avian colonization

patterns that occurred before the completion of the Central American Landbridge between North

and South America (Weir et al. 2009, Smith and Klicka 2010). During this time, many Nearctic

species were able to colonize South America, but relatively few South American species were

able to colonize North America (Weir et al. 2009; Smith and Klicka 2010) demonstrating an

asymmetry in dispersal ability. My finding of a latitudinal asymmetry in climatic niche

conservatism may be the cause of this differential dispersal ability between Nearctic and South

American derived Neotropical taxa. Completion of the landbridge precipitated a wave of avian

and mammalian dispersalists which easily colonized tropical regions of North America from

South America (Weir et al. 2009). However, this wave of invading species were generally unable

to colonize beyond the northern extent of tropical forest in central Mexico, suggesting that

tropical niche conservatism limited their dispersal out of the tropics. South American derived

mammal groups that colonized tropical regions of North America likewise seldom extended their

range north of the tropics.

Niche divergence can accelerate reproductive isolation leading to ecological speciation,

whereby extrinsic postzygotic isolation causes hybrids with intermediate ecological preferences

to be selected against (Rundle and Nosil 2005; Schluter 2009). My results indicate that climatic

niche is more divergent at high latitudes, suggesting that ecological speciation occurs faster

there. These results are consistent with previous estimates of accelerated speciation (Weir &

Schluter 2007) and with increased opportunity for incipient speciation (i.e. elevated subspecies

richness) in birds and mammals at high latitudes (Botero et al 2014). This poleward increase in

speciation cannot explain the latitudinal diversity gradient and suggests that other evolutionary

factors such as extinction and dispersal are the key drivers (e.g. Roy & Goldberg 2007;

Mittlebach et al. 2007). Previous studies have estimated extinction in birds, mammals (Weir &

Schluter 2007) and marine bivalves (Jablonsky et al. 2006) to be elevated at high latitudes. In

combination with my study, these results suggest that rates of speciation are elevated at high

11

latitudes, but that low temperate species richness is maintained due to accelerated extinction at

high latitudes and elevated rates of immigration into the tropics.

This study is not the first to discover a positive relationship between latitude and factors

that promote reproductive isolation. Recent research suggests the presence of a latitudinal

gradient in the evolution of sexually selected traits (e.g. length and syllable diversity of avian

song, plumage colouration), where sister pair divergence between such traits increases with

latitude (Martin et al. 2010; Weir and Wheatcroft 2011; Weir et al. 2012). Greater divergence in

climatic niche at high latitudes may have driven the latitudinal increase in divergence of sexually

selected characters. Under climatic niche divergence, populations may evolve towards different

optima that select for dissimilar sexual signals (Baldassarre et al. 2013). Both ecological and

sexually selected traits support a latitudinal pattern where divergent selection is strongest at high

latitudes. Additionally, these studies provide examples of both prezygotic (colour and song

divergence) and postzygotic (niche divergence) isolation occurring most rapidly towards the

poles.

In this study, I analyzed evolution of the realized climatic niche (the climatic conditions

that are present within a species’ geographic range) and not the fundamental climate niche (the

climatic tolerances of a species) (Hutchinson 1957). It is the realized climatic niche that exerts

environmental selection on a species. Given that tropical climate has been relatively stable over

long periods of time, the selection exerted by the realized niche should result in physiological

adaptation and specialization. If tropical species are more specialized, as many studies suggest

(Deutsch et al. 2008; Addo-Bediako et al. 2000), then I predict that the discrepancy between their

fundamental and realized niches will be less extensive in the tropics than at high latitudes. If true,

then the latitudinal asymmetry in niche conservatism should apply to the fundamental niche as

well.

This study is not without limitation. First, most sister pairs in my dataset originated

during the Pliocene and Pleistocene, suggesting that my results may be heavily influenced by

glacial cycles that occurred at this time. Whether my results apply to earlier time periods is

unknown. Second, I follow previous authors (e.g. Botero et al. 2014, Cadena et al. 2012) in

estimating climatic niche within the breeding range of each species, despite the fact that many

12

high latitude avian species are migratory. To determine the sensitivity of the results to migration,

I restricted the avian dataset to Nearctic species using climatic data only from the breeding

period when all Nearctic species are present. The results for this analysis for temperature (PC1)

showed the same pattern of increased divergence with increasing latitude (PC3, seasonality was

not analyzed because I restricted this analysis to spring and summer months), suggesting that my

results are not likely to be invalidated by migration.

I find that tropical species are more specialized in their climatic niche than temperate

species and that rates of climatic niche evolution are greater at high latitudes than in the tropics.

My results suggest faster divergent selection at high latitudes, which could drive faster ecological

speciation where species richness is lowest. This result is inconsistent with the model that views

the tropics as a species pump. Instead, my data show latitudinal asymmetries in niche

conservatism and degree of climatic specialization, which should promote differences in

dispersal ability between tropical and temperate latitudes. What are now needed are direct

estimates of dispersal differences between tropical and temperate regions.

13

Table 1. Likelihoods and support for Brownian Motion and Ornstein-Ulhenbeck models of

climatic niche evolution for Avian (a) and Mammalian (b) and non-migratory Avian (c) sister

pairs. ΔAICc scores (AICc for each model – smallest AICc score) and Akaike Weights (wAIC)

are used as metrics of model support. The best-fit model has the smallest ΔAICc value of 0

(bold). Akaike weights indicate the probability of fit for each model. N indicates the number of

parameters in each model

1a)

1b)

1c)

Migration Test

PC1 PC2 PC3

MODEL N LogLikelihood ΔAICc wAIC LogLikelihood ΔAICc wAIC LogLikelihood ΔAICc wAIC

BM Constant 1 -58.212 6.615 0.020 -38.838 2.592 0.107 -20.305 1.095 0.266

BM Two Rate 3 -52.905 0.000 0.535 -36.371 1.658 0.171 -19.982 4.449 0.050

BM Linear 2 -56.658 5.506 0.034 -37.958 2.833 0.095 -20.273 3.031 0.101

OU Constant 2 -55.231 2.653 0.142 -36.542 0.000 0.392 -18.757 0.000 0.461

OU Linear 4 -53.370 2.930 0.124 -36.367 3.649 0.063 -18.610 3.704 0.072

OU Two Rate 5 -52.210 2.610 0.145 -34.356 1.658 0.171 -18.061 4.449 0.050

Aves

PC1

PC2

PC3

MODEL N LogLikelihood ΔAICc wAIC LogLikelihood ΔAICc wAIC LogLikelihood ΔAICc wAIC

BM Constant 1 -266.339 47.244 0.000 -207.357 26.380 0.000 -143.643 51.338 0.000

BM Two Rate 3 -245.875 10.316 0.004 -204.138 23.942 0.000 -119.550 7.152 0.026

BM Linear 2 -252.120 20.806 0.000 -204.411 22.488 0.000 -121.291 8.634 0.012

OU Constant 2 -249.737 16.041 0.000 -195.687 5.040 0.065 -131.769 29.590 0.000

OU Two Rate 5 -239.443 1.453 0.324 -190.167 0.000 0.809 -117.432 6.916 0.029

OU Linear 4 -239.717 0.000 0.671 -193.030 3.727 0.126 -114.974 0.000 0.932

Mammals

PC1 PC2 PC3

MODEL N LogLikelihood ΔAICc wAIC LogLikelihood ΔAICc wAIC LogLikelihood ΔAICc wAIC

BM Constant 1 -109.913 22.782 0.000 -87.315 15.220 0.000 -62.704 2.408 0.111

BM Two Rate 3 -107.283 21.522 0.000 -85.599 15.789 0.000 -59.500 0.000 0.371

BM Linear 2 -107.153 19.263 0.000 -86.297 15.184 0.000 -62.001 3.001 0.083

OU Constant 2 -97.522 0.000 0.660 -78.705 0.000 0.648 -61.552 2.103 0.130

OU Two Rate 5 -95.707 2.370 0.202 -77.208 3.006 0.144 -57.817 0.633 0.270

OU Linear 4 -97.081 3.118 0.139 -77.844 2.278 0.270 -60.868 4.736 0.035

14

(B) Mammalian Species

(A) Avian Species

-0.4 -0.2 0 0.2 0.4

T_min_1

T_min_5

T_min_9

T_max_1

T_max_5

T_max_9

T_mean_1

T_mean_5

T_mean_9

Precip_1

Precip_5

Precip_9

PC1

-0.4 -0.2 0 0.2 0.4

T_min_1

T_min_5

T_min_9

T_max_1

T_max_5

T_max_9

T_mean_1

T_mean_5

T_mean_9

Precip_1

Precip_5

Precip_9

PC2

-0.4 -0.2 0 0.2 0.4

T_min_1

T_min_5

T_min_9

T_max_1

T_max_5

T_max_9

T_mean_1

T_mean_5

T_mean_9

Precip_1

Precip_5

Precip_9

PC3

Figure 1. Variable loadings patterns for the first three principal components of climatic niche measurements

extracted from the correlation matrix. Tmin represents minimum temperate for all months of the year, Tmax

represents maximum temperature for all months of the year, Tmean represents mean temperature for all months of

the year and Precip represent mean precipitation for all months of the year. Because climatic data was sorted,

numerical vales do not represent months in annual order.

T m

in

1

4

8

1

2

T m

ax

4

8

1

2

T

min

4

8

12

Pre

cip

4

8

1

2

T m

in

1

4

8

1

2

T m

ax

4

8

1

2

T m

in

4

8

1

2

P

reci

p

4

8

1

2

T m

in

1

4

8

1

2

T

max

4

8

12

T

min

4

8

12

Pre

cip

4

8

1

2

T m

in

1

4

8

1

2

T m

ax

4

8

1

2 T

mea

n

4

8

12

Pre

cip

4

8

1

2

T m

in

1

4

8

1

2

T m

ax

4

8

12

T m

ean

4

8

12

Pre

cip

4

8

1

2

T m

in

1

4

8

1

2 T

max

4

8

12

T m

ean

4

8

12

Pre

cip

4

8

1

2

15

ai) aii) aiii)

bi) bii) biii)

ci) cii) ciii)

Figure 2. Climatic divergence through time (branch length) for (a) PC1 Aves, (b) PC3 Aves and (c) PC3

Mammalia. PCs that best supported a model in which climatic divergence changes across latitude are

displayed. Branch lengths represent relative time and are in units of expected mutations per 100 base pairs

per million years. Graphs labeled (i) represent climatic divergence for tropical sister pairs (those with an

absolute midpoint latitude between 0° and 23°) and graphs labeled (ii) represent climatic divergence for

temperate sister pairs (those with an absolute midpoint latitude between 23° and 57° for birds and 23° and

65° for mammals). Graphs labeled (iii) demonstrate how climatic niche evolution is expected to change

after 3 branch length units of time across the latitudinal gradient based on maximum likelihood estimates

of evolution rate (β) and constrain (α) at each latitude under the best supported model.

0

2

4

6

8

10

12

14

16

0 2 4 6 8 10

Cli

ma

tic

Div

erg

ence

Branch Length

0

2

4

6

8

10

12

14

16

0 2 4 6 8 10

Cli

ma

tic

Div

erg

ence

Branch Length

0

1

2

3

4

5

6

0 2 4 6 8 10

Cli

ma

tic

Div

erg

ence

Branch Length

0

1

2

3

4

5

6

0 2 4 6 8 10

Cli

ma

tic

Div

erg

ence

Branch Length

0

1

2

3

4

5

6

7

8

0 5 10 15 20

Cli

ma

tic

Div

erg

ence

Branch Length

0

1

2

3

4

5

6

7

8

0 5 10 15 20

Cli

ma

tic

Div

erg

ence

Branch Length

0

1

2

3

4

5

6

7

0 10 20 30 40 50 60 70

Exp

ecte

d C

lim

ati

c

Div

erg

ence

Absolute Latitude

0

0.5

1

1.5

2

2.5

3

0 10 20 30 40 50 60 70

Exp

ecte

d C

lim

ati

c

Div

erg

ence

Absolute Latitude

0

0.2

0.4

0.6

0.8

1

0 10 20 30 40 50 60 70

Exp

ecte

d C

lim

ati

c

Div

erg

ence

Absolute Latitude

Tropical (0° to 23°) Temperate (23°+)

16

ai) aii) aiii)

bi) bii) biii)

Figure 3. The climatic range sizes of species and their corresponding midpoint latitudes for ai) mammals

PC1, aii) mammals PC2, aiii) mammals PC3, bi) birds PC1, bii) birds PC2 and biii) birds PC3. Black lines

indicate phylogenetically corrected regression lines (using Pagel’ lambda).

Absolute Latitude

Cli

mati

c R

an

ge

Absolute Latitude

Cli

mati

c R

an

ge

Absolute Latitude

Cli

mati

c R

an

ge

Absolute Latitude

Cli

mati

c R

an

ge

Absolute Latitude

Cli

mati

c R

an

ge

Cli

mati

c R

an

ge

Absolute Latitude

17

A)

18

19

Figure 4A Mammalian maximum clades credibility phylogenies created in BEAST v1.7.5.

Branches tips are labelled with the name of each species and their corresponding Genbank

Ascension number.

20

B)

21

22

23

24

25

26

27

Figure 4B Avian maximum clades credibility phylogenies created in BEAST v1.7.5.

Branches tips are labelled with the name of each species.

28

References Cited

Addo-Bediako, A., Chown, S. L. and Gaston, K. J. 2000. Thermal tolerance, climatic variability

and latitude. Proc. Biol. Sci. 267, 739-745.

Barker FK, Cibois A, Schikler P, Feinstein J, Cracraft J. 2004. Phylogeny and diversification of

the largest avian radiation. Proc Natl Acad Sci USA. 101: 11040-11045.

Baldassarre, D. T., Thomassen, H. A., Karubian, J., and Webster, M. S. 2013. The role of

ecological variation in driving divergence of sexual and non-sexual traits in the Red-backed

Fairy-wren (Malurus melanocephalus). BMC Evolutionary Biology 13, 75.

Bininda-Emonds, O.R.P., Cardillo, M., Jones, K.E., MacPhee, R.D.E., Beck, R.M.D., Grenyer,

R., Price, S.A., Vos, R.A., Gittleman, J.L., Purvis, A., 2008. Corrigendum: The delayed rise of

present-day mammals. Nature. 456, 274.

Botero C,A, Dor R, McCain C,M, Safran R,J. 2014. Environmental harshness is positively

correlated with intraspecific divergence in mammals and birds. Molecular Ecology. In Press.

Bush, M.B. & De Oliveira, P.E. 2006. The rise and fall of the refugial hypothesis of

Amazonian speciation: A paleo-ecological perspective. Biota Neotropica, 6.

Cadena, C.D., Kozak, K.H., Gomez, J.P., Parra, J.L., McCain, C.M., Bowie, R.C.K. 2012.

Latitude, elevational climatic zonation and speciation in New World vertebrates. Proc. R. Soc.

Lond. B 279: 194–201.

Colinvaux, P. A., De Oliveira, P. E., Moreno, J. E., Miller, M. C. & Bush, M. B. 1996. A long

pollen record from lowland Amazonia: Forest and cooling in glacial times. Science. 275, 85– 88.

Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor CK, Haak DC, Martin PR. 2008.

Impacts of climate warming on terrestrial ectotherms across latitude. P Natl Acad Sci USA. 105:

6668 – 6672.

29

Devictor V, Clavel J, Julliard R, Lavergne S, Mouillot D. 2010. Defining and measuring

ecological specialization. J Appl Ecol 47: 15–25.

Drummond AJ, Suchard MA, Xie D & Rambaut. 2012. A Bayesian phylogenetics with BEAUti

and the BEAST 1.7. Molecular Biology And Evolution. 29:1969-1973.

Fischer, A.G. 1960. Latitudinal variations in organic diversity. Evolution, 14, 64–81.

Futuyma, D. J., and G. Moreno. 1988. The evolution of ecological specialization. Ann. Rev. Ecol.

Syst. 19: 207–33.

Hawkins, B.A., Diniz-Filho, J.A.F., Jaramillo, C.A. & Soeller, S.A. 2006. Post-eocene climate

change, niche conservatism, and the latitudinal diversity gradient of New World birds. Journal of

Biogeography, 33 , 770– 780.

Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones and A. Jarvis. 2005. Very high resolution

interpolated climate surfaces for global land areas. International Journal of Climatology. 25:

1965-1978.

Hua, X., and J. J. Wiens. 2010. Latitutudinal variation in speciation mechanisms in frogs.

Evolution 64: 429–443.

Hutchinson, G.E. 1957. A treatise on limnology. Wiley and Sons, New York.

Jablonski, D. 2006. Out of the tropics: evolutionary dynamics of the latitudinal diversity

gradient. Science. 314: 102–106.

Janzen, D. H. 1967 Why mountain passes are higher in the tropics. Am. Nat. 101, 233–249.

Jetz, W., Sekercioglu, C.H. & Watson, J.E.M. 2008. Ecological correlates and conservation

implications of overestimating species geographic ranges. Conservation Biology. 22, 110–119.

Kozak, K. H., and J. J. Wiens. 2007. Climatic zonation drives latitudinal variation in speciation

mechanisms. Proceedings of the Royal Society of London. 274: 2995–3003.

30

Kozak, K. H., and J. J. Wiens. 2012. Phylogeny, ecology, and the origins of climate-richness

relationships. Ecology. 93: S167–S181.

Martin, P. R., Montgomerie, R. & Loughheed, S. 2009. Rapid sympatry explains greater color

pattern divergence in high latitude birds. Evolution. 64, 336 – 347.

Mittelbach, G. G. et al. 2007. Evolution and the latitudinal diversity gradient: speciation,

extinction and biogeography. Ecology Letters. 10, 315–331.

Olson DM, Dinerstein E, Wikramanayake ED, Burgess ND, Powell GVN, Underwood

EC, Kassem KR. 2001. Terrestrial ecoregions of the worlds: a new map of life on Earth.

BioScience. 51, 933–938.

Pagel, M. 1999. Inferring the historical patterns of biological evolution.

Nature. 401, 877 – 884.

Paradis E., Claude J. & Strimmer K. 2004. APE: analyses of phylogenetics and evolution in R

language. Bioinformatics. 20: 289-290.

Patterson, B. D., G. Ceballos, W. Sechrest, M. F. Tognelli, T. Brooks, L. Luna, P. Ortega, I.

Salazar, and B. E. Young. 2007. Digital Distribution Maps of the Mammals of the Western

Hemisphere, version 3.0. NatureServe, Arlington, Virginia, USA.

Peers, M.J.L., Thornton, D.H., Murray, D.L. 2007. Reconsidering the Specialist-Generalist

Paradigm in Niche Breadth Dynamics: Resource Gradient Selection by Canada Lynx and

Bobcat. Plos One. 7: (12).

Price, T. D. 2008. Speciation in birds. Roberts, Greenwood Village, CO.

Ridgely, R.S., T.F. Allnutt, T. Brooks, D.K. MicNicol, D.W. Mehlman, B. E. Young, and J. R.

Zook. 2003. Digital distribution maps of the birds of the Western Hemisphere. Version 1.0.

NatureServe, Arlington, VA.

31

Roy, K., and E. E. Goldberg. 2007. Origination, extinction and dispersal: integrative models for

understanding present-day diversity gradients. American Naturalist. 170: S71–S85.

Rundle HD, Nosil P. 2005. Ecological speciation. Ecology Letters. 8: 336–35.

Slabbekoorn, H. & Smith, T. B. 2002. Bird song, ecology and speciation. Phil. Trans. R. Soc.

Lond. B. 357, 493–503.

Schluter, D. 2000. The Ecology of Adaptive Radiation. Oxford University Press, Oxford.

Schluter, D. 2009. Evidence for ecological speciation and its alternative. Science. 323: 737-741.

Smith, B.T. & Klicka, J. 2010. The profound influence of the Late-Pliocene Panamanian uplift

on the exchange, diversification, and distribution of New World Birds. Ecography. 33, 333–342.

Smith, B.T., Bryson, R.W., Houston, D, Klicka, J. 2012. An asymmetry in niche conservatism

contributes to the latitudinal species diversity gradient in New World vertebrates. Ecology

Letters. 15, 1318 – 1325.

Sobel, J.M., Chen, G.F., Watt, L.R. & Schemske, D.W. 2010. The biology of speciation.

Evolution. 64: 295–315.

Sunday, J. M., Bates, A.E., Dulvy, N.K. 2012. Thermal tolerance and the global redistribution

of animals. Nature Climate Change. 2, 686–690.

Swofford, D. L. 2002. PAUP* 4.0b10: phylogenetic analysis using parsimony (*and other

methods). Sunderland, MA: Sinauer Associates.

Weir, J. T. & D. Schluter. 2004. Ice sheets promote speciation in boreal birds. Proceedings of the

Royal Society of London Series B. 271: 1881 - 1887.

Weir, J. T. & D. Schluter. 2007. The latitudinal gradient in recent speciation and extinction rates

of birds and mammals. Science. 315: 1574 - 1576.

32

Weir J. T., Bermingham E, Schluter D. 2009.The Great American Biotic Interchange in birds.

Proceedings of the National Academy of Science. 106: 21737–21742.

Weir, J. T. and T. Price. 2011. Limits to speciation inferred from times to secondary sympatry

and ages of hybridizing species along a latitudinal gradient. American Naturalist. 177: 462-469.

Weir, J. T., and D. J. Wheatcroft. 2011. A latitudinal gradient in evolutionary rates of bird song

complexity and length. Proc. R. Soc. Lond. B. 78: 1713–1720.

Weir, J. T., D. Wheatcroft, & T. Price. 2012. The role of ecological constraint in driving the

evolution of avian song frequency across a latitudinal gradient. Evolution. 66: 2773-2783.

Weir, J. T. 2014. EvoRAG: Evolutionary Rates Across Gradients. R package version 1.0

http://CRAN.R-project.org/.

Wiens, J. J., and M. J. Donoghue. 2004. Historical biogeography, ecology, and species richness.

Trends in Ecology and Evolution. 19: 639–644.

Wiens, J. J. and C. H. Graham. 2005. Niche conservatism: integrating evolution, ecology, and

conservation biology. Annual Review of Ecology, Evolution, and Systematics. 36: 519–539.

Wiens, J. J., C. H. Graham, D. S. Moen, S. A. Smith, and T. W. Reeder. 2006. Evolutionary and

ecological causes of the latitudinal diversity gradient in hylid frogs: treefrog trees unearth the

roots of high tropical diversity. American Naturalist. 168: 579-596.