Faculty of Science

EXAMINING THE MORPHOLOGICAL, ECOLOGICAL, AND ENVIRONMENTAL
PREDICTORS OF BRAIN EVOLUTION IN THE ORDER CHIROPTERA
2023 | VALENTINA SANTAMARIA

B.Sc. HONOURS THESIS – BIOLOGY

EXAMINING THE MORPHOLOGICAL, ECOLOGICAL, AND ENVIRONMENTAL
PREDICTORS OF BRAIN EVOLUTION IN THE ORDER CHIROPTERA
by

VALENTINA SANTAMARIA

A THESIS SUBMITTED IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
BACHELOR OF SCIENCE (HONS.)
in the
DEPARTMENT OF BIOLOGICAL SCIENCES
(Ecology and Environmental Biology)

This thesis has been accepted as conforming to the required standards by:
Dr. Matthew Reudink (Ph.D.), Thesis Supervisor, Dept. Biological Sciences
Dr. Nancy Flood (Ph.D.), Co-supervisor, Dept. Biological Sciences
Dr. Mark Rakobowchuk (Ph.D.), External examiner, Dept. Biological Sciences

Dated this 15th day of June, 2023, in Kamloops, British Columbia, Canada
© Valentina Santamaria, 2023

ABSTRACT
Brains evolve in conjunction with an array of morphological, ecological and environmental
factors, ranging from diet to habitat. Although the details vary, most theories relating to brain
evolution suggest that increasing variation, complexity or unpredictability in a species’
environment favours a larger brain or an increase in the size of a specific brain region. In this
phylogenetic comparative study, we investigated the relationship between ecology and brain size
across 204 bat species (Order Chiroptera) using variables related to diet, habitat, climate, and
range. Our findings suggest that different evolutionary patterns exist between the two major groups
of bats, Microchiroptera and Pteropodidae. Diet itself did not have an influence on the whole brain
size of bats from either clade. However, frugivorous species of Microchiroptera have larger
olfactory bulbs and smaller medullas than insectivorous species. In addition to the constituents of
the primary diet, we found that diet breadth contributes to brain evolution: frugivorous
Microchiropterans with narrower diets (i.e., lower diet breadth) had larger olfactory bulbs. Among
Pteropodid species, those with narrower diets had larger hippocampi than those with broader diets.
The size of some brain regions was associated with variables other than diet; for example, within
Pteropodidae, medulla size was impacted by whether species lived on islands. This study
highlights how different brain regions are under different selective pressures, which is consistent
with the mosaic theory of brain evolution. We also found evidence supporting the cognitive buffer
hypothesis, as temperature variability and island-dwelling seemed to have influenced the evolution
of certain brain regions. General patterns of brain evolution are difficult to elucidate even within
a single order, but overall, the dominant result seems to be that diet has had the largest impact on
the individual brain regions, particularly within the Order Chiroptera.
Thesis Supervisor: Professor Dr. Matthew Reudink
Thesis co-supervisor: Professor Dr. Nancy Flood
ii

ACKNOWLEDGEMENTS
I would like to acknowledge that this research was conducted on the unceded territories of
the Tk'emlúps te Secwépemc within Secwépemc'ulucw, the traditional and unceded territory of
the Secwépemc. Thank you to Dr. Matt Reudink and Dr. Nancy Flood for supervising my project
and to Dr. Mark Rakobowchuk for acting as my external supervisor. I would also like to thank Dr.
Mateen Shaikh and Sean Mahoney for their continued support with my statistical analysis, and Dr.
Andrew Iwaniuk for his help developing the methodology of this project.

iii

TABLE OF CONTENTS
ABSTRACT.................................................................................................................................... ii
ACKNOWLEDGEMENTS ........................................................................................................... iii
LIST OF FIGURES ........................................................................................................................ v
LIST OF TABLES ......................................................................................................................... vi
INTRODUCTION .......................................................................................................................... 1
MATERIALS AND METHODS .................................................................................................... 7
RESULTS ..................................................................................................................................... 10
DISCUSSION ............................................................................................................................... 19
LITERATURE CITED ................................................................................................................. 26
APPENDIX A: TABLE A1 .......................................................................................................... 30

iv

LIST OF FIGURES
Figure 1. Path analysis results for Michrochiroptera for a whole brain b hippocampus c medulla
and d olfactory bulbs (n=168). The response variable is outlined in green. Abbreviations:
W weight, WHB whole brain, HIP hippocampus, AREA range area (km 2), ROBV rest of
brain volume, MED medulla, B breadth, OB olfactory bulbs. Arrows indicate direction of
the relationship, while numbers indicate the strength of the relationship and whether the
effect is positive or negative.…………………………………………………………….15
Figure 2. Path analysis results for Pteropodidae for a hippocampus b medulla and c olfactory bulbs
(n=36). The response variable is outlined in green. Abbreviations: ROBV rest of brain
volume, HIP hippocampus, AREA range area (km 2), B breadth, IM Island mainland
classification, MED medulla, H Habitat, OB olfactory bulbs. Arrows indicate direction of
the relationship, while numbers indicate the strength of the relationship and whether the
effect is positive or negative………..………………………………………..…..……….18

v

LIST OF TABLES
Table 1. The component loadings assigned to each component for PC1 and PC2 (n=204)……..11
Table 2. Top Reduced PGLS models of the brain including Microchiroptera (n=168)……...….12
Table 3. Top Reduced PGLS models of the brain including Pteropodidae (n=36)……….……..13
Table 4. Path analysis coefficients for Microchiroptera (n=168)………………………………..16
Table 5. Path analysis coefficients for Pteropodidae (n=36)………………………...…………..19
Table A1. Counts for each discrete ecological variable used (n=204)………………….……….30

vi

INTRODUCTION
Among animal species, brains vary substantially in both size and morphology, and their
evolution is driven by an array of morphological, ecological or environmental factors. Multiple
theories have been proposed to explain how these factors influence the evolution of brain size and
morphology (Eisenberg and Wilson 1978; Iwaniuk and Arnold 2004; Jones and MacLarnon 2004;
Pitnick et al. 2006; Sayol et al. 2016; Sayol et al. 2018). For example, the cognitive buffer
hypothesis proposes that unpredictable environments may favour larger brains due to the
instability or dynamic nature of resource availability within them (Sayol et al. 2018). Similarly,
the habitat complexity hypothesis suggests that closed, dense, or cluttered habitats are more
complex, thus these habitats favour an enlargement of the brain (Safi et al. 2005). The common
thread among many hypotheses is that the more dynamic, unpredictable, or demanding the
environment, the more resourceful the animal must be in order to find food, survive to sexual
maturity and reproduce. Thus, selection may favour changes in brain size and morphology that
improve inventiveness and resourcefulness in these unpredictable environments.
Rather than changing the brain as a whole, selection may alter individual parts of the brain,
as described by the mosaic evolution theory (Barton and Harvey 2000). This theory states that
varying selection pressures over evolutionary time can result in parts of the brain evolving as a
patchwork (Barton and Harvey 2000; Rojas et al. 2013; Sayol et al.2018). In other words, different
parts of the brain may become larger or smaller according to their use or disuse, respectively. For
example, nut-caching birds tend to have a larger hippocampus—the brain region associated with
spatial memory—likely because a larger hippocampus allows the birds to find the nuts they have
hidden (Gould et al. 2013). As such, large-scale comparative approaches that examine the relative

strength of ecological, biological, and environmental factors on the evolution of different brain
regions may provide crucial insights into brain evolution.
Bats (Order Chiroptera) comprise a speciose mammalian, whose approximately 1,400
species live in a broad range of environments (Safi et al. 2005; Dechmann and Safi 2009). As a
result, the order encompasses wide variation in body morphology, behaviour, life history, and brain
size and structure. Due to bats’ success in colonizing a range of habitats, they are a useful model
for studying adaptive radiation and how different selective pressures can change morphology,
including that of the brain (Hutcheon et al. 2002; Dechmann and Safi 2009).
As seems to be true for birds, bat brain evolution might be constrained by the energetic
costs associated with flight (Dechmann and Safi 2009; McGuire and Guglielmo 2009). For
example, migratory birds have smaller brains than non-migratory birds after controlling for body
size; this is likely due to the trade-offs between the energetic demands of long-distance migration
and growing expensive neural tissues (Vincze 2016). So much energy is required for flight that
any changes in the size of the brain or its parts must have strong selective advantages. Thus,
comparative study of bats may provide insight into the selective pressures that most influence the
mammalian brain.
Absolute brain size is strongly correlated with body size (Dechmann and Safi 2009).
Allometry accounts for most of the variation in the brain size; thus, the remaining percentage of
brain volume and how brain mass is distributed across different brain regions must be explained
by evolutionary pressures other than those affecting body size overall (Dechmann and Safi 2009).
This correlation with body size can be controlled for by using residual brain mass (Jones and
MacLarnon 2004; Pitnick et al. 2006; McGuire and Ratcliffe 2011; Smaers et al. 2012; Sayol et
al. 2018) or less commonly, relative brain mass (Yao et al. 2012; Todorov et al. 2021 ). In this

2

study we used whole brain volume as a response variable and control for body weight including it
as a main effect within our analysis. Similarly, to control for brain size when analyzing different
brain parts, we used volume of the rest of brain (ROB) as a predictor variable in the model
(methods follow: Chambers et al. 2021; DeCasien and Higham 2019).
Diet is likely associated with brain evolution. For example, phytophagous bats (those that
feed on nectar, leaves, and fruit) have been shown to have larger brains relative to their body size
than do carnivorous bats (Eisenberg and Wilson 1978; Hutcheon et al. 2002; Jones and MacLarnon
2004). Eisenberg and Wilson (1978) hypothesized that demanding foraging strategies (e.g.,
searching for fruit or nectar) would select for larger brains. They suggest that large brains in
frugivores may have evolved since the location of fruit is more unpredictable throughout the year
than are insects; locating fruit requires more sensory input and spatiotemporal memory than does
echolocation. Consistent with this, Eisenberg and Wilson (1978) found that aerial insectivores had
the smallest brains. A more recent study by Rojas et al. (2013), using phylogenetic generalized
least squares (PGLS) models to compare the dietary specialization and relative brain size of two
bat families did not find any significant correlations between diet type, the degree of diet
specialization and brain size; they also did not find that phytophagous bats had larger brains than
insectivorous bats. The authors proposed that future studies should focus on the relative size of
specific brain structures in bats, rather than on the overall size of the brain (Rojas et al. 2013).
When examining specific brain regions, Hutcheon et al. (2002) found that fruit bats had
larger olfactory bulbs and hippocampi than non-fruit eaters. Furthermore, echolocating bats had
larger auditory nuclei than fruit bats, which cannot echolocate (Hutcheon et al. 2002). This
suggests bats’ brains are morphologically adapted for the most efficient form of food seeking for
their diet (Hutcheon et al. 2002). For example, phytophagous bats detect fruit by scent and have

3

large hippocampi, which increases their spatial memory and ability to recall the location of ripe
fruit.
Though selection may favour increases in overall brain size or the size of specific brain
regions, due to the high metabolic demands of brain tissue, brain size may be constrained by the
energetic needs of other organs and tissues—a trade-off referred to as the direct metabolic
constraint hypothesis or the expensive tissue hypothesis (Jones and MacLarnon 2004). However,
there is little evidence that energetic constraints have impacted brain size significantly in bats
(Jones and MacLarnon 2004; Dechmann and Safi 2009). When tested against covariates such as
basal metabolic rate (BMR), intestine length, and litter size, and brain size does not seem to be
related to any of these variables (Jones and MacLarnon 2004). In addition, Jones and MacLarnon
(2004) found no support for the maternal energy hypothesis which proposes that a higher BMR
and longer gestation period will result in a larger brain. Another study found that among
polyandrous species, males have large testes and small brains due to an energetic tradeoff between
tissue types (Pitnick et al. 2006). However, in a reanalysis using different covariates than those
selected by Pitnick et al. (2006), Dechmann and Safi (2009) found no correlation between bats’
mating system and brain size. It has been suggested that the original correlation was due to a lack
of covariates that considered the foraging ecology of the bat species examined (Healy and Rowe
2007; Dechmann and Safi 2009). These findings suggest that future comparative studies should
include a multitude of covariates that may directly or indirectly influence brain size (Healy and
Rowe 2007; Dechmann and Safi 2009 ).
Just as is true of expensive tissues, expensive behaviours such as migration can constrain
brain size. For example, migratory bat species have smaller brains and neocortices than nonmigratory species, perhaps due to a trade-off between the high energetic demands of sustaining

4

brain tissue and traveling long distances (McGuire and Ratcliffe 2011); alternatively, nonmigratory bats that face a harsh, cold environment could benefit from having a larger brain. Which
might increase their chances of survival by allowing them to be more resourceful (McGuire and
Ratcliffe 2011). Such a relationship between migratory status and brain size has also been found
in many bird species (McGuire and Guglielmo 2009). However, for bats, migratory status is
difficult to characterize due to high variability at both a population and individual level (McGuire
and Ratcliffe 2011; Fleming 2019).
Regardless of the energetic costs associated with larger brains, the cognitive buffer
hypothesis proposes that heightened intelligence allows an individual to adapt quickly to a
changing habitat (Sayol et al. 2018). Though named differently in different papers (flexibility
hypothesis, migratory precursor hypothesis, island hypothesis, foraging hypothesis), this idea has
been tested using migration, habitat complexity, and island occupation as covariates influencing
brain size (Safi et al. 2005; McGuire and Guglielmo 2009; Sayol et al. 2018). Safi and Dechmann
(2005) found that a larger hippocampus was associated with larger wing area and a larger wing
area was correlated with greater habitat complexity (Safi and Dechmann 2005). They suggest that
those traits would enhance the ability to maneuver and forage (respectively) in closed habitats.
However, they did not incorporate habitat complexity, wing area and brain size within the same
analysis; therefore, the correlation between habitat complexity and brain size was never directly
tested (Safi and Dechmann 2005).
Living on islands may also influence brain evolution, as island habitats restrict dispersal
more than similar mainland habitats, limiting gene flow and enhancing local adaptation. In
addition, large brained, and thus more innovative individuals, may be more resilient to the
environmental changes on islands as they are unable to disperse when unfavorable conditions arise

5

(Sayol et al. 2018; Gavriilidi et al. 2022). When comparing island bird species with sister taxa
found on the mainland, Sayol et al. (2018) found a strong relationship between bird brain size and
island occupation: island birds had larger brains (Sayol et al. 2018). Their results suggest that this
increase in brain size evolved after colonization, which supports the theory that islands are
characterized by different selective pressures than the mainland (Sayol et al. 2018).
Previous studies on bat brain evolution have mostly used Phylogenetic Independent
Contrasts, a method of analysis that assumes complete independence among taxa (ie., it constructs
sets of taxa that are independent with respect to the evolution of the traits of interest). However,
Rojas et al. (2012) suggested a realistic approach should account for the varying degrees of
independence among taxa. This approach is best accomplished by a Phylogenetic Generalized
Least Squares (PGLS) analysis; thus, it has gained popularity in comparative studies over the past
decade (Rojas et al. 2013; Sayol et al. 2018). To our knowledge, there have been no large-scale
phylogenetic studies on bat brain evolution that incorporate multiple aspects of bat ecology while
using PGLS. Using PGLS we can examine patterns of evolution across whole families, orders, or
clades. PGLS does not, however, allow us to discern whether specific predictor variables have a
direct or indirect effect on brain size. To parse the nature of the relationships among covariates
and changes in brain morphology, we can adopt a Phylogenetic Path Analysis (PPA). This allows
us to determine whether the factors evolutionarily associated with brain size, as identified in the
PGLS, have a direct or indirect effect, as well as determine the strength and directionality of the
relationships (van der Bijl 2018). This type of analysis could aid in determining which hypothesis
or hypotheses for brain evolution best explains the patterns seen in bats.
In this study we ask how ecological and biological traits influence brain size and
morphology across the entire Order Chiroptera. Specifically, using PGLS we test whether

6

variables, such as diet, climate, or island-dwelling are correlated with a change in overall brain
size, as well as the size of specific brain regions. Based on previous studies, we predict that diet
and habitat will be important predictors of brain size (Safi et al. 2005, Safi and Dechman 2005).
In addition, we examine temperature and annual precipitation, factors that have been linked to
body size, and potentially to energetic constraints (Jones and MacLarnon 2004). Because birds
and bats are both volant species that may share similar energetic constraints and occupy similar
niches, we predict that they may also experience similar changes in brain morphology associated
with their ecology, such as an increase in brain size in island species (McGuire and Guglielmo
2009; Sayol 2018). After identifying factors associated with differences in brain morphology, we
then employ a path analysis to determine the strength of these relationships and whether these
relationships are direct or indirect.

MATERIALS AND METHODS
Data
Data on the brain volume of 204 species of bats were obtained from Comparative
Neurobiology in Chiroptera (Baron et al. 1996). This source includes data on the volume of the
following brain parts for each species: the hippocampus, olfactory bulbs (one data point for both
bulbs), and medulla. The overall body size for each species is also provided.
Information on the ecology of bats was gathered from the Handbook of the Mammals of the
World (2019) and from the International Union for Conservation of Nature (IUCN 2022). The data
includes primary diet (fruit, seeds, insects, nectar/pollen, vertebrates, fish, or blood), habitat (from
among forest, artificial/terrestrial, savannah, open or cave), and range size estimates (km 2) from
IUCN range maps. See Table A1 in Appendix A for counts of the number of species in each
category for diet and habitat variables. Bats with diets consisting of fish (n=1), blood (n=2) or
7

vertebrates (n=5) were combined into a new diet category denoted as “Other” due to small sample
sizes for each. Seed diets (n=3) were recategorized as fruit diets (now n=52) for the same reason.
For a full list of diet and habitat categories with sample sizes please refer to Table A1 in Appendix
A.
The data from the IUCN also included information on island occupancy. Islands were defined
as land masses with areas less than or equal to 785 753 km 2, which is the size of Papua New Guinea
(Benitez-Lopez et al. 2021). Island (n=49) and mainland (n=155) species were designated based
on whether at least 70 % of their range fell into one of these 2 categories (Table A1). If the species’
range was evenly divided between island and mainland environments, it was denoted as a mainland
species.
Climate data was obtained from the World Clim2 website (www.worldclim.org). Climate
variables included temperature seasonality (changes in temperature over the year), annual
temperature range, isothermality (which relates day-to-night temperature oscillations to summerto-winter oscillations), mean temperature of the coldest quarter, mean temperature of the coldest
month, mean temperature of the driest quarter, precipitation seasonality, precipitation in the driest
month, precipitation in the driest quarter, precipitation in the coldest quarter, and annual
precipitation. A Principal Component Analysis (PCA) performed on the total set of temperature
and precipitation variables. These variables were then used as covariates in our Phylogenetic
Generalized Least Squares (PGLS) analysis and Phylogenetic Path Analysis (PPA).
Because diet could have a significant influence on brain morphology, the diet breadth of the
bats was considered along with their primary food source. We obtained data on bat diet breadth
data from Wilman et al. (2014) and Sayol et al. (2018). The dietary usage data estimated the
frequency with which each species consumed different foods (invertebrates, vertebrates, fish, fruit,

8

nectar, and seeds) and scaled them from 0 (not used) to 100 % (primarily used [Wilman et al.
2014]). Next, we created a matrix for each type of food the bats consumed and used the matrix to
calculate overall diet breadth. Rao’s quadratic entropy, a formula from the R package indicspecies
was used to characterize the diversity (i.e., breadth) of a species’ diet was used (De Caceres and
Legendre 2009).
Only one phytophagous species (Brachyphylla cavernum) included non-fruit plant matter in
their diet, thus plant matter was reclassified to fruit to simplify the analysis. One species,
Phylloderma stenops, had a diet breadth 4 times larger than the next highest species (0.1 vs 0.025);
it was thus considered an outlier and removed from subsequent analyses.
Linear Regression
A simple linear regression was performed on brain volume and body weight in the
statistical analysis software JASP (JASP Team 2023).
Phylogenetic Generalized Least Squares (PGLS)
To examine the effects of phylogeny we began with a full mammalian phylogeny from
Upham et al. (2019) and trimmed it to only include bat species. We then performed a series of
analyses using R studio. A phylogenetic generalized least squares (PGLS) model from the nlme
package in R was used to control for relationships among taxa (Revell 2012). When testing whole
brain size as a response variable, allometric correction was performed by placing body weight
directly in the model as a predictor variable. To control for brain volume when using a portion of
the brain (e.g., amygdala), the rest of brain (ROB) method was used (DeCasien and Higham 2019).
Specifically, the weight of the brain part was subtracted from the whole brain volume and used as
a predictor variable in the model. Next, we performed a stepwise elimination of non-significant
terms to arrive at a final best fit model.

9

Phylogenetic Path analysis
Using the R package phylopath, we performed a phylogenetic path analysis (1) to identify
the strength and directionality of the factors identified as important in the PGLS and (2) to
determine whether the effects were direct or indirect (van der Bijl 2018). The C statistic ((ΔCIC2)
was used to rank the path models. The best model was selected for analysis if ΔCIC2 was > 2;
however, we performed model averaging of the top models when ΔCIC2 was < 2.
Since categorical data cannot be included in path analysis, diet, habitat, and
island/mainland classification data were changed to binary numerical variables. Specifically, diet
was recategorized as either frugivorous or other and habitat was recategorized as either open or
closed. Open habitats included those described as artificial/terrestrial, savannah, or open, and
closed habitats included caves and forests. The Pteropodidae species included in the analysis all
lived in open environments; thus, the habitat data was recategorized to be other and forest. open
and artificial/terrestrial habitats were included in the "other" category.

RESULTS
Linear regression analysis showed that 90.5 % of the variation in brain size could be explained
by variation in body mass. This finding suggests that 9.5 % of the variation in brain mass is due to
other factors or potentially random noise. The mean brain volume was 837.08 ± 1444.56 cm 3 SD.
Principal Component Analysis
We performed a Principal Component Analysis (PCA) to collapse our climate variables. PC1
explained 49.7 % of the variation in our climate variable and related primarily to variation in
temperature. The variables which loaded heavily on PC1 wer temperature seasonality, temperature
annual range, isothermality, mean temperature of the coldest quarter-year, minimum temperature
of the coldest month, and mean temperature of the driest quarter-year (Table 1). PC2 explained
10

21.6% of the variation in our climate variables and related mainly to variation in precipitation. As
the loadings on PC2 show, it describes mainly precipitation seasonality (the amount of variation
in precipitation within a year), total annual precipitation, and the amount of precipitation during
driest month of the year, and driest quarter-year (Table 1. Together, PC1 and PC2 account for 71.3
% of the variation in our climate data
Table 1. The largest loadings for each variable on PC1 and PC2 (n=204).
Component Loadings/Variables
Temperature Seasonality
Annual Range in Temperature
Isothermality
Mean Temperature of the Coldest Quarter
Minimum Temperature of the Coldest Month
Mean Temperature of the Driest Quarter
Precipitation Seasonality
Precipitation in the Driest Month
Precipitation in the Driest Quarter
Precipitation in the Coldest Quarter
Annual Precipitation

PC1
-1.11
-0.97
0.97
0.82
0.80
0.72

PC2

-1.05
0.93
0.93
0.77
0.46

Predictors of brain size and morphology: Microchiroptera
Within the Microchiroptera, diet type (X2 = 9.58, p < 0.05, n = 168) and PC1 (X2 = 7.4, p
< 0.01, n = 168) showed an association with whole brain size (Table 2). Diet type (X 2 = 26.45, p
< 0.0001, n = 168) was the only variable other than the rest of brain (ROB) volume that showed a
significant correlation with hippocampal volume (Table 2). As seen in Table 2, the size of the
medulla was also significantly correlated with diet type (X 2 = 11.58, p < 0.01, n = 168) and also
with range area (X2 = 4.22, p < 0.05, n = 168). Finally, diet breadth (X2 = 13.02, p < 0.001, n =
168) and diet type (X2 = 18.77, p < 0.001, n = 168) were significantly correlated with olfactory
bulb volume (Table 2). We did not find a difference in the size of any brain parts among species
inhabiting different habitats or those living on islands rather than on the mainland.

11

Table 2. Top Reduced PGLS models of the brain including Microchiroptera (n=168).
Predictor
Variables

X2

p

Brain

Weight
Diet Type
PC1

358.20
9.58
7.40

<0.0001
0.022
0.0065

Hippocampus

ROB Volume
Diet Type

708.55
26.45

<0.0001
<0.0001

Medulla

ROB Volume
Diet Type
Range Area

667.19
11.58
4.22

<0.0001
0.0090
0.04

Olfactory bulbs

ROB Volume
Diet Breadth
Diet Type

503.27
13.02
18.77

<0.0001
0.00031
0.00031

Response Variable

Degrees of
Freedom
1
3
1
1
1
3
1
1
1
1
1
1
3

Predictors of brain size and morphology: Pteropodidae
In the fruit bat family Pteropodidae, none of the ecological predictors tested were
associated with whole brain size (Table 3). Pteropodid species with broader diets had a larger
hippocampal volume (X2 = 5.92, p < 0.05, n = 36). Medulla volume was marginally associated
with island-dwelling (X2 = 3.95, p < 0.05, n = 36). Olfactory bulb volume had the most complicated
associations with the covariables, as diet breadth (X 2 = 21.03, p < 0.0001, n = 36), diet type (X2 =
7.64, p < 0.01, n = 36), habitat (X2 = 5.84, p < 0.05, n = 36)., and island-dwelling (X2 = 16.76, p <
0.0001, n = 36) all emerged as significant correlates in the reduced model (Table 3).

12

Table 3. Top Reduced PGLS models of the brain including Pteropodidae (n=36).
Predictor
Degrees of
Response Variable
X2
p
Variables
Freedom
Hippocampus

ROB Volume
Diet Breadth

565.29
5.92

<0.0001
0.015

1
1

Medulla

ROB Volume
Island-Dwelling

317.37
3.95

<0.0001
0.047

1
1

Olfactory bulbs

ROB Volume
Diet Breadth
Habitat Type
Diet Type
Island-Dwelling

487.39
21.03
5.8435
7.638
16.7555

<0.0001
<0.0001
0.016
0.0057
<0.0001

1
1
2
1
1

Path analysis: Microchiroptera
From the path analysis, we were able to determine the strength and directionality of the
relationships identified in the PGLS. Within the Microchiropteran clade, the best model showed a
negative and direct effect of PC1 (the climate variable which represents temperature variation) on
whole brain volume and weight; however, the effect for both was small (Figure 1a, Table 4). Our
diet variable, which was changed to a binary variable representing a frugivorous or nonfrugivorous diet, did not have an effect on any variable, but PC1 had a negative affect on diet
(Figure 1a, Table 4). These findings suggests that diet does not have an impact on brain size, and
that temperature and weight are much stronger determinants of whole brain size than diet.
Diet had a direct effect on ROB volume, but an indirect effect on hippocampus volume
(Figure 1b, Table 4). Diet does not seem to directly influence hippocampus volume in
Microchiroptera. Hippocampus size is largely determined by the size of the brain, and only
indirectly by diet.
13

Diet had a direct and negative effect on medulla size (Figure 1b, Table 4): the path analysis
showed that frugivorous Microchiropterans have smaller medulla than insectivorous bats. Range
area also had a direct negative effect on medulla size, but it also had a small positive effect on diet
(Figure 1c, Table 4).
Olfactory bulb size was strongly correlated with diet type and breadth in Microchiroptera.
First, frugivory had a negative effect on diet breadth; thus, fruit eating species are more specialized
than species with other diets (Figure 1d, Table 4). Second, frugivory also has a direct and positive
effect on brain size. Third, diet breadth had a negative and direct influence on olfactory bulb size
(Figure 1d, Table 4). Thus, species with large olfactory bulbs had smaller diet breadth. Our analysis
suggests that frugivorous species with specialized diets, or small diet breadths had the largest
olfactory bulbs among the Microchiroptera.

14

Figure 1. Path analysis results for Michrochiroptera for a Whole brain b Hippocampus c Medulla
and d Olfactory bulbs (n=168). The response variable is outlined in green. Abbreviations: W
weight, WHB whole brain, HIP hippocampus, AREA range area (km2), ROBV rest of brain
volume, MED medulla, B breadth, OB olfactory bulbs. Arrows indicate direction of the
relationship, while numbers indicate the strength of the relationship and whether the effect is
positive or negative.

15

Table 4. Path analysis coefficients for Microchiroptera (n=168).
Response
Variable

Factor 1

Factor 2

Path
Coefficient

Standard
Error

Brain

Weight
PC1
PC1
PC1

→
→
→
→

Brain Volume
Diet
Weight
Brain Volume

-0.75
-0.02
-0.10
-0.08

0.03
0.03
0.06
0.03

Hippocampus

ROB Volume
Diet Type

→
→

Hippocampus
ROB Volume

0.90
0.25

0.03
0.16

Medulla

ROB Volume
Diet Type
Range Area
Range Area

→
→
→
→

Medulla
Medulla
Medulla
Diet Type

1.00
-0.11
-0.05
0.01

0.03
0.07
0.02
0.03

Olfactory bulbs

ROB Volume
Diet Breadth
Diet Type
Diet Type

→
→
→
→

Olfactory Bulbs
Olfactory Bulbs
Olfactory Bulbs
Diet Breadth

0.76

0.03
0.05
0.07
0.10

-0.14
0.08
-0.41

Path Analysis: Pteropodidae
Within the Family Pteropodidae, we found patterns opposite to those found for the
Microchiroptera. Unlike the other bat families, the hippocampus size of Pteropodidae is directly
influenced by diet breadth (Figure 2a, Table 5). Unexpectedly, fruit bats with a large breadth seem
to have a smaller hippocampus than those with a small breadth (Figure 2a, Table 5). We also found
a negative relationship between island-dwelling and medulla size (Figure 2b, Table 5). Fruit bats
that live on islands have smaller medulla than those that live on islands.
Thus, overall, the path analysis, showed that the olfactory bulbs of Pteropodids are
indirectly and directly influenced by diet type, diet breadth, habitat type, and island-dwelling
(Figure 2c, Table 5). Olfactory bulb evolution is more complicated for Pteropodidae than for
16

Microchiroptera. Our path analysis showed that frugivorous species with a larger diet breadth had
small olfactory bulbs (Figure 2c, Table 5). Thus, among the Pteropodidae, fruit eating generalists
with small olfactory bulbs seem to be the norm. From this analysis it is not possible to know if
specialist fruit bats have larger olfactory bulbs as in Microchiroptera. However, our results suggest
that pollen/nectar eaters that are specialists have larger olfactory bulbs than those that are
generalists. Our findings suggest that phytophagous (plant-eating; i.e., nectivores and frugivores)
species from both bat clades could face similar selective pressures affecting olfactory bulb size,
but these patterns are complex and difficult to understand.

17

Figure 2. Path analysis results for Pteropodidae for a Hippocampus b Medulla and c Olfactory
bulbs (n=36). The response variable is outlined in green. Abbreviations: ROBV rest of brain
volume, HIP hippocampus, AREA range area (km2), B breadth, IM Island mainland classification,
MED medulla, H Habitat, OB olfactory bulbs. Arrows indicate direction of the relationship, while
numbers indicate the strength of the relationship and whether the effect is positive or negative.

18

Table 5. Path analysis coefficients for Pteropodidae (n=36).
Response
Variable

Factor 1

Factor 2

Path
Coefficient

Standard
Error

Hippocampus

ROB Volume
Diet Breadth

→
→

Hippocampus
Hippocampus

0.88
-0.09

0.04
0.38

Medulla

ROB Volume
Island Dwelling

→
→

Medulla
Medulla

0.78
-0.06

0.04
0.03

Olfactory bulbs ROB Volume
Island Dwelling
Island-Dwelling
Habitat Type
Habitat Type
Diet Type
Diet Type
Diet Breadth

→
→
→
→
→
→
→
→

Olfactory Bulbs
Olfactory Bulbs
Habitat Type
Olfactory Bulbs
Diet Type
Olfactory Bulbs
Diet Breadth
Olfactory Bulbs

1.04
0.11
-0.38
-0.23
0.05
0.12
0.07
-0.10

0.05
0.0
0.34
0.05
0.18
0.04
0.18
0.04

DISCUSSION
Whether the size of the brain, or the size of specific areas of the brain, in bats is influenced
by diet, habitat, or energetic constraints has been widely debated (Healy and Rowe 2007). While
diet, habitat, and environment were all associated with brain evolution, our findings demonstrate
that the nature of the selective pressures that have influenced whole brain, hippocampus, medulla
and olfactory bulbs size are complex and differ between the clades Microchiroptera and
Pteropodidae. Taken together, our results support the mosaic evolution theory, suggesting that the
evolution of individual brain parts has been influenced by different selective pressures in a variety
of ways.
Whole brain

19

We found that whole brain size in Microchiroptera was negatively associated with
temperature (PC1). The climate variable PC1, which is comprised primarily of temperature
variables, had a direct effect on brain size in the Microchiroptera. Many of the PC1 component
loadings pertained to temperature variability, such as isothermality and temperature seasonality.
Thus, bat species with a higher PC1 score likely come from regions where temperature is variable
over the year. A larger brain size could allow bats living in regions with larger fluctuations in
temperatures to have more flexible behaviours or a wider range of behaviours, which would help
to buffer them against environmental changes, consistent with the predictions of the cognitive
buffer hypothesis (Sayol et al. 2018).
Few studies have investigated the effects of climate change on the mammalian brain, and
the few that exist have mainly focused on human evolution. One recent macroevolutionary study
on the Genus Homo found that, even after controlling for the effects of body size, sex, phylogeny,
and geography (Stibel 2022), larger brains evolved during periods of global cooling. It is possible
that both bat and human evolution are affected by environmental stochasticity in that daily and
seasonal temperature fluctuations may influence the size of the brain. The cognitive buffer
hypothesis suggests that stochasticity and fluctuations in climate could favor a capacity for
individuals to produce novel behaviours that would improve the chances of survival in a changing
environment. However, Stibel’s (2002) results did not reveal if temperature variation influenced
the evolution of brain size directly, or only indirectly through an effect on body size.
Rubalcaba et al. (2022) found that there was a trade off between flight ability in bats and
thermoregulation in colder environments; larger wing surface area increases heat loss but reduces
energy expenditure since it makes flight easier. Bats that live in colder environments tended to
have a lower wing surface area due to the energetic constraints associated with living in cold

20

environments (Rubalcaba et al. 2022). The authors suggested that selection related to
thermoregulation was stronger in colder regions where bats expended more energy on heat
production; however, they did not test if the effect changed in regions with highly variable
temperature as they primarily used yearly temperature averages in their analysis (Rubalcaba et al.
2022). Our study suggests that the temperature itself may not constrain the size of the brain in
Microchiroptera, but rather the fluctuations in temperature may have an important influence on
brain evolution. Our finding could be the result of thermoregulation and energy tradeoffs, or
because a larger brain is required for bats to survive in a wide range of environmental conditions.
Previous studies have found a correlation between diet type and whole brain size; however,
our analysis suggests that diet has no influence on the brain. Among Pteropodids, whole brain size
was not associated with any ecological factors at all. Thus, our results suggest that in general, diet
type has not directly influenced the evolution of brain size in bats.
Hippocampus
The size of the hippocampus within Microchiroptera was associated with diet type;
however, the path analysis showed that the effect of diet on the hippocampus is indirect, acting
through its direct effect on the rest of brain (ROB) volume. This result provides a more
comprehensive understanding of the evolution of the hippocampus in frugivorous
Microchiropterans, as the only other study to examine hippocampus size across multiple bat
species did not include PGLS or path analysis (Hutcheon et al. 2002). Like Hutcheon et al. (2002),
we did find an association between diet and hippocampus size; however, the relationship is
indirect, and driven by the effect of diet on the ROB volume.

21

In contrast to the findings for Microchiroptera, diet breadth had a direct and negative effect
on hippocampus size in the Pteropodids. Specialist fruit bats have larger hippocampi than
nectivores with broader diets breadths. The hippocampus is responsible for spatial awareness and
memory (Liberti III et al. 2022). Thus, we presumed that bats with larger diet breadths would
require a larger hippocampus to forage for food from various locations and to remember these
locations. The relationship between brain size and diet breadth in other animals is unclear;
however, research suggests a tendency for generalist species to have larger brain size residuals or
score higher on tests of cognitive ability (Chambers et al. 2021; Henke-von der Malsburg 2020;
Mac Lean et al). We found the opposite to be true among the Pteropodidae.
There are two possible explanations for our results: (1) fruit contains more calories than
nectar and pollen, which could loosen the energetic constraints on the brain (DeCasien et al. 2017),
and (2) food is more unpredictably and patchily distributed for longer periods of the year for fruit
or flower eating species (Heithaus 1982). First, similar to our result, DeCasien et al. (2017) found
that among primates, frugivorous species had larger brains than folivores. They also found that
diet type was a better predictor of brain size than other behavioural factors such as sociality
(DeCasien et al. 2017). These authors suggested that frugivorous diets are less energy constrained
than other diet types because fruits are rich in calories and do not require large energy input for
digestion. Lower energy constraints could allow for other tissues, such as those of the brain, to
grow. Although DeCasien et al. (2017) investigated whole brain size, energy-dense foods such as
fruit may also be influencing the evolution of the hippocampus in Pteropodidae. Second, although
obligate fruit eaters only eat one type of food source, there is a wide variation in the size, shape,
color, smell, seasonality, distribution, and nutrient content of fruit. The variation among plants,
and the fact that fruit distribution can change overnight could result in greater selection pressures

22

related to finding food and remembering its location, driving the evolution of a larger hippocampus
(Heithaus 1982).
Medulla
After completing the path analysis for Microchiroptera, we found that the effect of diet on
the medulla was negative and direct: non-frugivorous bats have larger medullae (or medullas) than
frugivorous bats. In mammals, the medulla oblongata is part of the brainstem and relays
information from the brain to the spinal cord (Sricharoenvej et al 2008). It controls involuntary
physiological processes such as heart rate, blood pressure and respiration (Sricharoenvej et al.
2008). It is unclear why insectivores have a larger medulla than frugivores, but it could be due to
trade-offs with other brain parts that seem to be associated with frugivory, such as the olfactory
bulbs.
Range size also had a direct and negative effect on the medulla size in Microchiroptera;
bats species with more restricted ranges have larger medullae. In Pteropodids, medulla size was
correlated with island-dwelling. The relationship was direct and negative, meaning that bats living
on islands have smaller medullae than bats living on the mainland. Because islands inherently have
a limited area it is possible that range size is still responsible for this relationship, however it is
unclear why. Islands restrict species ranges and gene flow and can have more erratic climates than
the mainland (Sayol et al. 2018). These characteristics make living on islands a strong selective
force; thus, living on islands may be impacting certain species of bats within the Pteropodidae.
Olfactory bulbs
Like other studies, we found that fruit eating bats in both Microchiroptera and Pteropodidae
have larger olfactory bulbs than bats that eat other types of food (Bhatnagar & Kallen 1974; Barton
et al. 1995; Eiting et al. 2014; Hall et al. 2020). However, we did not find that frugivorous

23

Pteropodids had larger olfactory bulbs than nectivores. Interestingly, we also found that
Microchiropteran fruit eating bats with smaller diet breadths had larger olfactory bulbs than those
with larger diet breadths. Bats specializing on nectar and pollen also had larger olfactory bulbs
than generalist frugivores in the Pteropodidae. We could expect the opposite outcome if we
consider that larger olfactory bulbs may contain a wider diversity of olfactory receptors for finding
a variety of food types. However, it may be that olfactory bulb size provides bats with higher
sensitivity to smell, rather than the ability to detect a broader spectrum of smells. Frugivorous bats
rely heavily on olfaction to gather their food, especially the species that are specialists. As
mentioned, specialist fruit or flower eaters encounter a wide variety of fruit or flower types and
display flexible behaviours when foraging for fruit (Heithaus 1982). In this case, the absolute size
of the olfactory bulbs could represent increased sensitivity for, and focus on, a particular food
source.
The similarities between Microchiropteran and Pteropodid olfactory bulb evolution end
with diet; although habitat complexity also plays a role in the evolution of the Pteropodidae
olfactory bulbs, this is not true for Microchiropterans. Island dwelling has a positive and direct
effect on the size of the olfactory bulbs in the Pteropodidae. In accordance with the cognitive buffer
hypothesis, bats may require stronger olfaction when there is environmental stochasticity in order
to adjust to changes in food sources. Islands may provide increased habitat complexity and
environmental stochasticity, which would drive an increase in the size of the olfactory bulbs over
evolutionary time. Perhaps the fruit sources on islands are less predictable than those on the
mainland, which results in bats requiring an enhanced sense of smell.
We also found that living in a forested environment seems to result in a decrease in the size
of the olfactory bulbs, compared to living in shrublands or artificial/terrestrial habitats such as

24

pastureland or urban areas. Thus, our results reject the cognitive buffer hypothesis, as forested
areas are the most complex (Safi et al. 2005). However, many bats classified as living on
"artificial/terrestrial" lands also inhabit other habitat types such as forests. It might be useful to
investigate this relationship further with a higher resolution index of habitat variability and
complexity such as an Enhanced Vegetation Index which was used by Sayol et at. (2018).
Limitations
In contrast to a previous study (Safi et al. 2005), we did not find a strong relationship
between habitat complexity and the size of various brain parts in bats. It is possible we simply did
not examine the brain structures that are influenced by habitat complexity. Alternatively, the lack
of support could be due to limitations in our habitat data. In nature, habitat type or preference is
not neatly divided into categories for each species, as it was in our analysis. A more accurate
estimation of the complexity of the habitats in which each species lives should be made, as we did
for diet, in which “breadth” was considered.
To make sense of the results, we had to separate the Microchiroptera from the Pteropodidae
(Yao et al. 2012). The large differences between Pteropodidae and the rest of the bats makes it
very difficult to find broad patterns for all bats. If we chose to partition the bats in a different
way—using only monophyletic groups, for example, it may have changed our conclusions.
As with any correlational study using PGLS and path analysis, these analytical tools
themselves have limitations. PGLS results depend on the variables included in the model and the
observations chosen, such as habitat or diet type of the species, for the models (Wartel 2019).
Thus, if we had classified habitat type differently, or included more variables such as hibernation,
migration, or basal metabolic rate, our results may have been different.

25

We had much more data available on the brain sizes of frugivorous Pteropodids (n=28),
but far less for the pollen and nectar eaters (n = 8). It is possible that nectar eaters and fruit eaters
share similar patterns of evolution as both depend on patchily distributed food sources from plants
that may not be available year-round. A smaller sample size of nectivores than frugivores may not
be ecologically meaningful, as differences may be found where none really exist as the statistical
power of an analysis decreases.
Many phylogenetic comparative studies on the evolution of the brain assume that brain size
is a direct index of cognitive ability. However, the relationship between brain size and cognition
is unclear (Healy & Rowe 2021). For future studies, other brain measurements, such as neuron
density, could be used to compare the cognitive abilities of animals (Dicke & Roth 2016).
Our results indicate that climate, foraging methods, diet breadth, range size, and habitat
can influence the size and morphology of the brain in bats. Different selective pressures have
impacted the evolution of the hippocampus, medulla and olfactory bulbs, within the groups
Microchiroptera and Pteropodidae. These findings are consistent with the mosaic evolution
hypothesis, which proposes that each part of the brain evolves independently to a certain degree.
The medulla and olfactory bulbs were both directly influenced by living in stochastic environments
such as islands or regions with high temperature variability—thereby supporting the cognitive
buffer hypothesis. We also found that diet type and diet breadth were associated with the size of
the olfactory bulbs in both bat clades. Although body size is the primary determinant for brain size
in bats, many other ecological factors can limit or enhance the evolution of a section of the brain.
This study represents the largest, most comprehensive study of bat brain evolution to date. Our
study demonstrates that brain evolution in Chiroptera is complex, varying between subgroups of
the order, and among different brain regions.

26

LITERATURE CITED
Baron G, Stephan H, Frahm HD. 1996. Macromorphology, Brain Structures, Tables and Atlases.
1st Ed. Birkhauser Basel. (Comparative Neurobiology in Chiroptera; vol 1).
Barton RA, Harvey PH. 2000. Mosaic evolution of brain structure in mammals. Nature.
405(6790):10055-1058. doi:10.1038/35016580.
Barton RA, Purvis A, Harvey PH. 1995. Evolutionary radiation of visual and olfactory brain
system in primates, bats and insectivores. Philos. Trans. R. Soc. Lond., B, Biol. Sci.
348(1326):381-192. doi:10.1098/rstb.1995.0076.
Benítez-López A, Santini L, Gallego-Zamorano J, Milá B, Walkden P, Huijbregts MAJ, Tobias
JA. 2021. The island rule explains consistent patterns of body size evolution in terrestrial
vertebrates. Nat Ecol Evol. 5(6):768-786. doi:10.1038/s41559-021-01426-y.
Bhatnagar KP, Kallen FC. 1974. Cribriform plate of ethmoid, olfactory bulb and olfactory acuity
in forty species of bats. J Morphol. 142(1):71-89. doi:10.1002/jmor.1051420104.
van der Bijl W. 2018. phylopath: Easy phylogenetic path analysis in R. PeerJ. 6:e4718.
doi:10.7717/peerj.4718.
De Cáceres M, Legendre P. 2009. Associations between species and groups of sites: Indices and
statistical inference. Ecology. 90(12):3566-3574. doi:10.1890/08-1823.1.
Chambers HR, Heldstab SA, O’Hara SJ. 2021. Why big brains? A comparison of models for both
primate and carnivore brain size evolution. PLoS One. 16(12):e0261185.
doi:10.1371/journal.pone.0261185.
DeCasien AR, Higham JP. 2019. Primate mosaic brain evolution reflects selection on sensory and
cognitive specialization. Nat Ecol Evol. 3(10):1483-1493. doi:10.1038/s41559-019-09690.
Dechmann DKN, Safi K. 2009. Comparative studies of brain evolution: A critical insight from the
Chiroptera. Biol. Rev. 84(1):161–172. doi:10.1111/j.1469-185X.2008.00067.x.
Dicke U, Roth G. 2016. Neuronal factors determining high intelligence. Philos. Trans. R. Soc.
Lond., B, Biol. Sci. 371(1685):20150180. doi:10.1098/rstb.2015.0180.
Eisenberg JF, Wilson DE. 1978. Relative Brain Size and Feeding Strategies in the Chiroptera.
Evolution. 32(4):740-751. doi:10.2307/2407489.
Eiting TP, Smith TD, Dumont ER. 2014. Olfactory Epithelium in the Olfactory Recess: A Case
Study in New World Leaf-Nosed Bats. Anat. Rec. 297(11):2105-2122.
doi:10.1002/ar.23030.
Gavriilidi I, De Meester G, Damme R Van, Baeckens S. 2022. How to behave when marooned:
the behavioural componenet of the island syndrome remains unexplored. Biol Lett.
18(4):1–23. doi:10.1098/rsbl.2022.0030.
Gould KL, Gilbertson KE, Hrvol AJ, Nelson JC, Seyfer AL, Brantner RM, Kamil AC. 2013.
Differences in relative hippocampus volume and number of hippocampus neurons among
five corvid species. Brain Behav Evol. 81(1):56–70. doi:10.1159/000345560.
Hall RP, Mutumi GL, Hedrick BP, Yohe LR, Sadier A, Davies KTJ, Rossiter SJ, Sears K, Dávalos
LM, Dumont ER. 2021. Find the food first: An omnivorous sensory morphotype predates
biomechanical specialization for plant based diets in phyllostomid bats. Evolution.
75(11):2791-2801. doi:10.1111/evo.14270.
Healy SD, Rowe C. 2007. A critique of comparative studies of brain size. Proceedings of the Royal
Society B: Biological Sciences. 274(1609):453–464. doi:10.1098/rspb.2006.3748.

27

Heithaus ER. 1982. Coevolution between bats and plants. In: Kunz, TH, editor. Ecology of bats.
Boston, MA: Springer. p.327-367.
Henke-von der Malsburg J, Kappeler PM, Fichtel C. 2020. Linking ecology and cognition: does
ecological specialisation predict cognitive test performance? Behav Ecol Sociobiol.
74:154. doi:10.1007/s00265-020-02923-z.
Hutcheon JM, Kirsch JAW, Garland T. 2002. A comparative analysis of brain size in relation to
foraging ecology and phylogeny in the Chiroptera. Brain Behav Evol. 60(3):165–180.
doi:10.1159/000065938.
International Union for Conservation of Nature (IUCN). (2022). https://www.iucn.org/
Iwaniuk AN, Arnold KE. 2004. Is cooperative breeding associated with bigger brains? A
comparative test in the Corvida (Passeriformes). Ethology. 110(3):203-220.
doi:10.1111/j.1439-0310.2003.00957.x.
JASP Team (2023). JASP (Version 0.17.1) [Computer software].
Jones KE, MacLarnon AM. 2004. Affording larger brains: testing hypotheses of mammalian brain
evolution on bats. Am Nat. 164(1):20-31. doi:10.1086/421334.
Liberti WA, Schmid TA, Forli A, Snyder M, Yartsev MM. 2022. A stable hippocampal code in
freely flying bats. Nature. 604(7904):98-103. doi:10.1038/s41586-022-04560-0.
MacLean EL, Hare B, Nun CL, Addess E, Amic F, Anderson RC, Aureli F, Baker JM, Bania AE,
Barnard AM, et al. 2014. The evolution of self-control. Proc Natl Acad Sci U S A.
111(20):2140-2148. doi:10.1073/pnas.1323533111.
McGuire LP, Guglielmo CG. 2009. What can birds tell us about the migration physiology of bats?
J Mammal. 90(6):1290–1297. doi:10.1644/09-MAMM-S-084R.1.
McGuire LP, Ratcliffe JM. 2011. Light enough to travel: migratory bats have smaller brains, but
not larger hippocampi, than sedentary species. Biol Lett. 7(2):233–236.
doi:10.1098/rsbl.2010.0744.
Pitnick S, Jones KE, Wilkinson GS. 2006. Mating system and brain size in bats. Proc. Royal Soc.
B, Biol. Sci. 273(1587):719–724. doi:10.1098/rspb.2005.3367.
Revell LJ. 2012. phytools: An R package for phylogenetic comparative biology (and other things).
Methods Ecol Evol. 3(2):217-223. doi:10.1111/j.2041-210X.2011.00169.x.
Rojas D, Mancina CA, Flores-Martínez JJ, Navarro L. 2013. Phylogenetic signal, feeding
behaviour and brain volume in Neotropical bats. J Evol Biol. 26(9):1925–1933.
doi:10.1111/jeb.12190. http://tree.bio.ed.ac.uk/soft-.
Rubalcaba JG, Gouveia SF, Villalobos F, Cruz-Neto AP, Castro MG, Amado TF, Martinez PA,
Navas CA, Dobrovolski R, Diniz-Filho JAF, et al. 2022. Physical constraints on
thermoregulation and flight drive morphological evolution in bats. Proc Natl Acad Sci U S
A. 119(15):e2103745119. doi:10.1073/pnas.2103745119.
Safi K, Dechmann DKN. 2005. Adaptation of brain regions to habitat complexity: a comparative
analysis in bats (Chiroptera). Proc. Royal Soc. B, Biol. Sci. 272(1559):179–186.
doi:10.1098/rspb.2004.2924.
Safi K, Seid MA, Dechmann DKN. 2005. Bigger is not always better: When brains get smaller.
Biol Lett. 1(3):283–286. doi:10.1098/rsbl.2005.0333.
Sayol F, Downing PA, Iwaniuk AN, Maspons J, Sol D. 2018. Predictable evolution towards larger
brains in birds colonizing oceanic islands. Nat Commun. 9:2820. doi:10.1038/s41467-01805280-8.
Sayol F, Maspons J, Lapiedra O, Iwaniuk AN, Székely T, Sol D. 2016. Environmental variation
and the evolution of large brains in birds. Nat Commun. 7:13971.
28

Smaers JB, Dechmann DKN, Goswami A, Soligo C, Safi K. 2012. Comparative analyses of
evolutionary rates reveal different pathways to encephalization in bats, carnivorans, and
primates. Proc Natl Acad Sci. 109(44):18006-18011. doi:10.1073/pnas.1212181109.
Sricharoenvej S, Niyomchan A, Lanlua P, Piyawinijwong S, Roongruangchai J. 2008.
Microvasculature of the medulla oblongata in the Lyle’s flying fox (Pteropus lylei). Journal
of Veterinary Medicine Series C: Anatomia Histologia Embryologia. 37(6):401-407.
doi:10.1111/j.1439-0264.2008.00856.x.
Stibel JM. 2022. Climate Change Influences Brain Size in Humans. Brain Behav Evol. 98(2):93106. doi:10.1159/000528710.
Todorov OS, Blomberg SP, Goswami A, Sears K, Drhlík P, Peters J, Weisbecker V. 2021. Testing
hypotheses of marsupial brain size variation using phylogenetic multiple imputations and
a Bayesian comparative framework. Proc. Royal Soc. B, Biol Sci. 288(1947):20210394.
doi:10.1098/rspb.2021.0394.
Wartel A, Lindenfors P, Lind J. 2019. Whatever you want: Inconsistent results are the rule, not the
exception, in the study of primate brain evolution. PLoS One. 14(7) :e0218655.
doi:10.1371/journal.pone.0218655.
Wilman H, Belmaker J, Simpson J, de la Rosa C, Rivadeneira MM, Jetz W. 2014. EltonTraits 1.0:
Species-level foraging attributes of the world’s birds and mammals. Ecology. 95(7):20272027. doi:10.1890/13-1917.1.
Wilson DE, Mittermeier RA. 2019. Bats. Barcelona (ES): Lynx Edicions. (Handbook of the
Mammals of the World; vol. 9).
Yao L, Brown JP, Stampanoni M, Marone F, Isler K, Martin RD. 2012. Evolutionary change in
the brain size of bats. Brain Behav Evol. 80(1):15–25. doi:10.1159/000338324

29

APPENDIX A
Table A1. Counts for each discrete ecological variable used (n=204).
Variable
Count
Primary Diet
Blood, Fish, and Vertebrates
8
Fruit
52
Insects
126
Nectar/Pollen
18
Habitats
Open
22
Artificial/Terrestrial
47
Savana
10
Forest
101
Caves
24
Island/Mainland Classification
Island
49
Mainland
155
N
204

30