![Alternative models (linear mixed-effects models; log transformed Gaussian distribution) and predictor variables explaining activity patterns of Florida bonneted bats (Eumops floridanus) detected at bat houses on Fred C. Babcock-Cecil M. Webb Wildlife Management Area, FL (1 Sept 2015e1 Sept 2018). For each response variable, the top two models are listed with associated DAIC (difference between Akaiki Information Criteria [AIC] score of each alternative model and model with the lowest AIC score) and Akaike weights (u i ).](https://www.researchgate.net/profile/Elizabeth-Braun-De-Torrez-2/publication/344935321/figure/tbl1/AS:951672822038530@1603907890105/Alternative-models-linear-mixed-effects-models-log-transformed-Gaussian-distribution_Q320.jpg)
Content uploaded by Elizabeth Braun de Torrez
Author content
All content in this area was uploaded by Elizabeth Braun de Torrez on Oct 28, 2020
Content may be subject to copyright.
Original Research Article
Evidence of resource-defense polygyny in an endangered
subtropical bat, Eumops floridanus
Elizabeth C. Braun de Torrez
a
,
*
, Jeffery A. Gore
b
, Holly K. Ober
c
a
Florida Fish and Wildlife Conservation Commission, 1105 SW Williston Rd., Gainesville, FL, 32601, USA
b
Florida Fish and Wildlife Conservation Commission, Panama City, FL, 32409, USA
c
University of Florida, Quincy, FL, 32351, USA
article info
Article history:
Received 23 July 2020
Received in revised form 23 September
2020
Accepted 24 September 2020
Keywords:
Activity budget
Animal behavior
Emergence
Endangered species
Florida bonneted bats
Harems
Roost fidelity
abstract
Understanding sociality and animal behavior is critical for developing effective conserva-
tion strategies. Many tropical bat species form harems, where dominant males play key
social roles by defending groups of females directly (female-defense polygyny) and/or the
resources that females need (resource-defense polygyny). The Florida bonneted bat
(Eumops floridanus) is an endangered subtropical species suspected to form harems, but
our understanding of its social structure, reproduction, and behavior is rudimentary. In this
study, we evaluated demographic variation in morphological and behavioral characteris-
tics of Florida bonneted bats to test the hypothesis that this species forms harem groups
and exhibits resource-defense polygyny at roost sites. We used a 4-year dataset of 341
individuals uniquely marked with passive integrated transponders (PIT tags), coupled with
tri-annual capture records, to track activity patterns of bats at five roosts fitted with PIT tag
readers. We identified the likely dominant males and other demographic groups in each
roost using morphometric characteristics and reproductive status. We assessed differences
between sexes and among status categories in three primary metrics: amount of activity at
the roost, time of emergence, and initial foray duration per night. Dominant males
consistently were the most active individuals at roosts and spent the least amount of time
away from roosts during forays, relative to females and other males. Females spent more
time away from roosts than males and shared similar foraging activity patterns regardless
of status. Our findings suggest that Florida bonneted bats form small harem groups that
are active year round. Male bats exhibit characteristics of resource-defense polygyny at
roost sites and a size-biased hierarchy, with the largest reproductively active males
appearing to defend the roost at the expense of time spent foraging. We suggest that the
roost site represents a critical, limited and defendable resource for male Florida bonneted
bats to gain access to females, which has important implications for the conservation and
enhancement of roost sites. Our study highlights the importance of accounting for dif-
ferences in behavior across demographic groups and social roles when considering
resource needs for imperiled species.
©2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
*Corresponding author.
E-mail address: Elizabeth.Braun@myfwc.com (E.C. Braun de Torrez).
Contents lists available at ScienceDirect
Global Ecology and Conservation
journal homepage: http://www.elsevier.com/locate/gecco
https://doi.org/10.1016/j.gecco.2020.e01289
2351-9894/©2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
Global Ecology and Conservation 24 (2020) e01289
1. Introduction
Animals live in complex societies structured largely by the distribution of mating opportunities and availability of key
resources (Emlen and Oring, 1977;Clutton-Brock, 1989). Individuals vary their behavioral patterns to balance energy ex-
penditures with demands associated with foraging, reproduction, predator avoidance, and resource defense; in turn, these
trade-offs can vary based on social context, sex, reproductive status, and season (Lima and Dill, 1990;Thies et al., 2006).
Understanding social structure, social roles, and how behaviors vary across demographic groups is increasingly recognized as
having far-reaching implications for species conservation (Parreira and Chikhi, 2015;Snijders et al., 2017;Brakes et al., 2019;
Goldenberg et al., 2019). Disruption of social connectivityand stability can impact not onlysurvival and success of individuals
but ultimately population viability (Snijders et al., 2017).
Sociality is widespread in bats (Order Chiroptera), a hyper-diverse, gregarious, and globally threatened taxon (O’Shea et al.,
2016). However, few species have been studied in enough detail to fully characterize their social systems or understand how
behavior may differ demographically and seasonally (McCracken and Wilkinson, 2000;Kerth, 2008). Harems, a relatively
stable association of a male with several females, occur in many tropical bat species where, unlike in temperate species,
breeding seasons are not constrained by climate extremes, and male competition for access to females occurs year round
(Campbell et al., 2006;Kerth, 2008;Kerth et al., 2011). Harems can vary both in the compositional stability of individual
members and the duration of their associations, with some groups remaining together for many years (McCracken and
Wilkinson, 2000). A single dominant male typically guards its harem from competing males (e.g., Dwyer, 1970;Williams,
1986;Park, 1991) and thereby obtains greater reproductive success than other males that may be present within the
harem (‘subordinate’) and males outside the harem (‘satellite’), although these other males can also win some copulations
with females (Ortega et al., 2003;Garg et al., 2015;Wilde et al., 2018). Although they may vary in body size and the presence
of sexually dimorphic features, dominant males tend to be larger than other individuals in their harem and exhibit higher
roost fidelity within and/or across breeding seasons (Williams, 1986;McWilliam, 1988;Ortega and Arita, 1999;Storz et al.,
2000).
Dominant males in harem-forming species may defend females directly (female-defense polygyny) and/or defend re-
sources critical to females (resource-defense polygyny) (Emlen and Oring, 1977;McCracken and Wilkinson, 2000;Kerth,
2008). However, these social systems are not mutually exclusive and some species display aspects of both defense strate-
gies depending on the distribution of mates or resources (Kunz et al., 1983;McCracken and Wilkinson, 2000;Günther et al.,
2016). In species exhibiting female-defense polygyny or resource-defense at foraging sites, dominant males forage near fe-
males in their harem to defend the females or a foraging territory (e.g., greater sac-winged bats, Saccopteryx bilineata;
(Bradbury and Vehrencamp, 1976). Nocturnal surveillance of females may benefit males by preventing extra-harem copu-
lations at night. In species exhibiting resource-defense polygyny at roost sites, males often remain close to the roost and
return frequently throughout the night to defend it through vigilance at roost entrances and physical exclusion of conspecific
males while females forage elsewhere (e.g., Indian short-nosed fruit bat, Cynopterus sphinx (Balasingh et al., 1995;Mahandran
et al., 2018); white-throated, round-eared bat, Lophostoma silvicolum (Dechmann et al., 2005). Availability of suitable roost
sites likely drives resource-defense polygyny in many bat species given that roosts are a particularly coveted resource by
females to raise young, are often limited across the landscape, and are easily defended by males (Dechmann et al., 2005;
Kerth, 2008;Wilde et al., 2018).
Several species of Eumops (Molossidae), a widespread genus throughout much of the new world tropics, form mixed sex
colonies, but it is unclear if these are harem groups and what role the males play socially (e.g., Wagner’s bonneted bat, E.
glaucinus (Best et al., 1997), western mastiff bat, E. perotis (Best et al., 1996), dwarf bonneted bat, E. bonariensis,(Bowles et al.,
1990). Males of these species have been noted to have gular-thoracic glands (hereafter “gular glands”), a scent-producing
gland on the ventral surface of the throat that has an unknown function but may contribute to harem maintenance
(Bowles et al., 1990;Best et al., 1996, 1997). In other harem-forming species, males use these glands to scent mark females,
other males, and roost substrates (McCracken and Bradbury, 1981;Heideman et al., 1990;French and Lollar, 1998;Voigt and
von Helversen, 1999;Racey, 2009), which may signal the mating status and individual identity of the male (Adams et al.,
2018).
The Florida bonneted bat, E. floridanus, (Molossidae) has been subject to increased attention because of its recent listing as
federally endangered in the United States (United States Fish and Wildlife Service [USFWS] 2013). Endemic to a small sub-
tropical region (southern Florida, U.S.A), they are suspected to differ ecologically and socially from most other bats in the
temperate United States. However, our understanding of their social structure, reproduction, and nightly and seasonal activity
patterns is rudimentary, precluding development of appropriate recovery efforts. These bats do not appear to migrate or
hibernate, are suspected to form harems with an extended reproductive season (Ober et al., 2017a), and have been found to
roost in small groups (<25 individuals) in tree cavities (Belwood, 1981;Angell and Thompson, 2015;Braun de Torrez et al.,
2016). Like other Eumops, males have gular glands that vary among individuals in size, color, and whether they are open
or closed (Ober et al. 2017a, 2017b). Within Florida bonneted bat roosts, Ober et al. (2017) documented mixed sex groups with
a clear female bias among adults. The majority of roosts contained a single adult male with an open gular gland that
consistently had the greatest mass of any male within the roost and larger testes than males with closed gular glands. These
findings provide some evidence that each roost may have one dominant male, but we do not know anything about the
stability of these suspected harem groups or the role these males play in defense.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
2
In this study, we evaluated demographic variation in morphological and behavioral characteristics of Florida bonneted
bats to test the hypothesis that this species forms year-round harems and exhibits resource-defense polygyny at roost sites. To
do this, we used capture records coupled with passive integrated transponders (PIT tags) to track activity patterns of indi-
vidual bats at roosts outfitted with PIT tag readers that documented the time at which each tagged individual entered or
exited a roost. We assessed seasonal variation and differences among bats of different sex and reproductive status in three
primary metrics: amount of activity at the roost, time of emergence, and foray duration (time spent awayfrom the roost) each
night. If Florida bonneted bats form harems that are defended by a dominant male, we would expect to find one repro-
ductively active male roosting with multiple females that is larger than any other males present and exhibits high roost fi-
delity. If dominant males defend harem females or their resources at both roosting and foraging sites (female-defense or
resource-defense of foraging territories) we would expect males and females to have very similar nightly activity patterns.
In contrast, if dominant males exhibit resource-defense at roost sites while females forage, we would expect males and fe-
males to have different nightly patterns. In this case we would predict that dominant males would be more active at the
roosts, emerge later, and spend less time away from the roost than females and non-dominant males. We assumed that high
activity levels at the roost indicated frequent short flights near the roost or time spent near the entrance to the roost, both of
which support possible roost vigilance and defense.
2. Methods
2.1. Study area
We conducted this study from April 2014 to August 2018 in Fred C. Babcock-Cecil M. Webb Wildlife Management Area
(BWWMA), a 26,611 ha conservation area within the core range of Florida bonneted bats in Florida, U.S.A. where the species
roosts in artificial bat houses and tree cavities (for study area details, see Bailey et al., 2017a;Ober et al., 2017a). Ten matched
pairs of single chamber bat houses (each ca. 8 10
3
cm
3
interior volume) sharing a pole are located in BWWMA and spaced at
distances of 0.34e19.30 km apart (8.68 ±0.46); each matched pair of bat houses is hereafter referred to as a “roost”and
considered to function as one social unit. The property predominantly consists of mesic and hydric slash pine (Pinus elliottii)
flatwoods with a woody shrub and grass understory, and freshwater prairies. During the study period, the average daily
temperature was 23.99 ±0.14
C (7.22e31.67
C), and the area received an average of 147.42 cm of rain each year, with the
majority of precipitation occurring in the wet season (mid-May through October).
2.2. Data collection
To collect data on morphology, demography and colony composition over time, we conducted multi-day capture events
using triple high mist nets at occupied Florida bonneted bat roosts every 4 months (April, August, and December) from 2014
to 2018 (for capture procedure details, see Bailey et al., 2017a;Ober et al., 2017a). For each captured bat we recorded body
mass, forearm length, age (adult/subadult), sex, and reproductive status (females: nonreproductive, pregnant, lactating, or
postlactating; males: nonreproductive [testes abdominal], reproductive [testes descended]). As in Ober et al. (2017),we
classified the gular glands on males as “open,”“closed,”or “not apparent,”and measured gular glands (bare patch width),
gular gland openings (width), and descended testes (width, length).
To identify individual bats, we uniquely marked all unmarked bats during each capture event with PIT tags (12 mm,
115 mg, 134.2 kHz FDXB PIT-tags [Biomark Inc., Boise, ID, USA]; see Bailey et al., 2017a;Ober et al., 2017a). Permanent PIT tag
readers (Biomark IS1001 RFID transceivers) were installed at seven roosts between 2014 and 2018 and powered by solar
panels (150 W with 12 V 10 A solar charge controller). Antennas fitted at the base of each bat house detected PIT-tagged bats
each time they passed the plane of the antenna and transmitted data to the PIT tag reader at that roost. The real time of each
reader was set to reflect the U.S. Naval Observatory Master Clock Time every 1e6 months and used to calculate and adjust the
data for the average time drift per day of each reader (x¼1.79 ±0.03 s/day).
All research followed American Society of Mammalogists (ASM) guidelines for research on live animals (Sikes and Animal
Care Use Committee of the American Society of Mammalogists, 2016) and was in accordance with the following permits:
University of Florida IACUC (# 201308070), USFWS (#TE 23583B-1); it was conducted under the Cooperative Agreement
between the United States Fish and Wildlife Service and Florida Fish and Wildlife Conservation Commission for the con-
servation of threatened and endangered species under Section 6(c) of the Endangered Species Act.
2.3. Data analysis
2.3.1. PIT tag data adjustments
We restricted our analyses to five roosts (paired bat houses) that recorded PIT tag data simultaneously for 3 full years (1
Sept 2015e1 Sept 2018). However, we included in our analyses all individuals that were PIT tagged from the start of the
project (April 2014) or captured at alternate roost sites within the study area because many of these bats were subsequently
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
3
detected at the focal bat houses within our analysis period. We conducted all data analyses in the statistical software language
R (v. 3.5.1) in conjunction with R studio (v. 1.1.456).
We first assigned each detection a “bat night”that reflected the date of each detection between 1200 on one day and 1159
on the following day. We generated sunset and sunrise times for each bat night (function sunrise. set, R package Stream-
Metabolism) and calculated the number of minutes past sunset for each bat detection. Because the time between sunset and
sunrise varies seasonally, we calculated night length as the number of hours between sunset and sunrise and accounted for
this when calculating proportion of time spent foraging each night. To minimize misleading detections occurring from bats
moving near the PIT tag antennas at the roost entranceduring the day and immediately prior to emergence, we eliminated all
detections between sunrise and 10 min after sunset, the earliest time bats have been visually observed to first emerge at
BWWMA (unpublished data). For bats that were detected at multiple roosts on the same night, we assigned bats to the roost
where they were detected closest to sunset, so that we included only bats that had presumably roosted there during the day
and excluded bats that may have visited after emerging from other roosts during the night (0.5% of detections excluded).
To merge our capture dataset with our PIT tag dataset, we assigned the capture data (e.g., reproductive status) associated
with each individual from the date of the most recent capture event until the date of the subsequent capture event. If a bat
was not captured during one capture event but detected by PIT tag readers during subsequent weeks or months, that bat was
excluded from analysis during that time period because of unknown status. To account for potential short-term changes in
behavior due to the capture efforts, we removed all data from dates of capture and 1 day after capture from our analyses. We
accounted for PIT tag loss by identifying individuals that had been PIT tagged multiple times. To do this, we "took wing tissue
biopsies (4 mm) and" screened all bats captured without a PIT tag for matching genotypes using a microsatellite panel of 22
loci (J. Austin unpublished data). The probability of identity (PID) of the marker panel (probability that two individuals have
identical genotypes by chance) and the PIDsibs (probability that two individual siblings have identical genotypes by chance)
were 1.8x10^-11 and 2.2x10^-4 respectively, giving us high confidence that the genotype profiles correctly identified bats that
lost PIT tags. We documented 32 cases of retagged bats (8.6% of total tagged bats) and matched newly inserted PIT tag
numbers with the old tag numbers starting on the dates the new PIT tags were reinserted.
2.3.2. Response variables
We calculated three primary metrics of activity for each bat on each night: 1) Roost activity (total number of detections),
which represents how active each individual bat was near the entrance of the roost, and may indicate a bat leaving/entering
the roost to forage, leaving/entering the roost to fly nearby, or positioning itself on the landing pad just outside the roost
entrance; 2) Emergence times (minutes past sunset that each bat exited each house for the first time). We required that the 1st
detection must be 1.5 h past sunset. This threshold was used because the majority of detections (90%) were within this
defined time period, suggesting this constituted the primary emergence of the colony and excluded extraneous detections
later in the night (Supplementary Materials S1); and 3) Initial foray duration (time between first and second consecutive
detections), which we presumed to be the primary evening foraging foray. We then calculated the proportion of the night
spent foraging by standardizing all initial foray durations by their respective night length (hours between sunset and sunrise)
to account for seasonal variation in time available for foraging. Finally, to further describe activity patterns we calculated (but
did not statistically analyze) the number of forays and the total time spent away from the roost per night. For the number of
forays, we counted the paired detections at each roost (odd numbered detections presumed to be a bat exiting the roost [i.e.,
detection 1 is initial emergence for first foray] and even numbered detections presumed to be the same bat entering the roost
[i.e., detection 2 is bat returning from first foray]). For the total time spent away from the roost, we calculated the time between
each of these paired detections (i.e., time between presumed roost exit and subsequent return) and added them together for
each night. We excluded observations with an odd number of total detections on any given night because we were unable to
discern where the error may have occurred (i.e., false detection or missed detection).
2.3.3. Predictor variables
We considered several reproductive status categories in our analyses (hereafter referred to as “status”); both adults and
subadults were included in these categories and were separated by reproductive status and morphometrics rather than by
age. For females, we considered pregnant and lactating individuals as “reproductive females”and post-lactating and non-
reproductive individuals as “non-reproductive females.”For males, we defined status using several objective criteria
related to their potential status in a harem. First, based on previous evidence that there may be one large, dominant adult
male per harem (Ober et al., 2017a), we identified the most likely dominant males captured from each roost at each capture
event where at least one female was also present. We selected the adult male with the greatest body mass and indication of
reproductive activity (open gular gland and/or descendedtestes) that was present in the roost where captured forat least 10%
of the total days between capture events. If two adult males from the same roost differed by <3 g in body mass and had
similar reproductive characteristics, we considered them to be equal candidates and classified both as potentially dominant.
For the remaining males, we defined the following status categories: contender 2 males with open gular glands AND
descended testes, contender 3 males with open gular glands OR descended testes, and non-reproductive males with closed or
not-apparent gular glands AND abdominal testes. We labeled males as contender males rather than subordinates because we
were uncertain of their role within the roost. To visually verify that our defined male status categories explained observed
variation in male characteristics, we conducted a principle components analysis using the following scaled continuous
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
4
variables: body mass, forearm length, body condition index (forearm length/body mass; BCI), testes size (length and width),
and gular gland size (opening width and gland width) and overlaid our defined male status categories (Fig. 1).
2.3.4. Statistical analyses
We first tested for differences in size (body mass and forearm length) and roost fidelity among bats of different status
categories using an analysis of variance (ANOVA), followed by Tukey HSD tests for multiple comparisons. For individuals
captured more than once, we used a mean value to represent each size characteristic for each individual that was reported in
the same status category multiple times. We calculated roost fidelity (%) as the number of nights each bat was present in the
roost where it was captured from one capture event to the next out of the total number of nights that any bat was present
during that period.
We tested for differences in activity patterns (roost activity, emergence times, and initial foray duration) among individuals
of different sex and status by constructing two sets of linear mixed-effects models (LMMs; Gaussian distribution, function
lme, R package nlme (Pinheiro et al., 2012); for each of the fixed effects sex and status. To meet assumptions of normality, we
applied a natural log transformation to each response variable. All models included a random effect of status nested within
individual PIT tag identification number, nested within roost, which accounted for the fact that (i) bats from the same roost
may have activity more similar to one another than to bats from other roosts, (ii) activity from the same bat on multiple nights
was not independent, and (iii) individual bats changed status categories over the study period. We also included an auto-
regressive correlation structure (coAR1) term, which accounts for inherent temporal autocorrelation in repeated measures.
For each response variable, we fit a null model and a set of single or double variable alternative models that included one of
the fixed effects (sex, status) and/or one of three temperature variables: minimum, maximum or average daily dry bulb
temperatures collected from the nearest weather station (Punta Gorda, National Oceanic and Atmospheric Association
[NOAA]). For the set of status models, we also included the best model that included sex to compare the relative strength of
the two fixed effects. We included different measures of daily temperature as possible covariates because bats arewell known
to vary their activity patterns with temperature (Hayes, 1997) and Florida bonneted bats appear to be sensitive to cold (United
States Fish and Wildlife Service [USFWS] 2013;Bailey et al., 2017), but it is unknown what metric of temperature best explains
variation in activity. We used Akaike’s Information Criterion (AIC) to rank each set of models (Burnham, 2002). Models within
2
D
AIC of each other were considered to have equivalent support. We used maximum likelihood (ML) and Laplace approx-
imations to allow for model comparisons and calculated parameter estimates with REML (Bolker et al., 2009;Pinheiro et al.,
2018). We tested the effect of status category by using a likelihood ratio test to compare two nested models (LRT, function
anova, R package stats; significance threshold
a
¼0.05), and Tukey pairwise comparisons between levels of significant factors
Fig. 1. Lef t panel ePrinciple components analysis of male Florida bonneted bats (Eumops floridanus) captured at five bat houses at Fred C. Babcock-Cecil M. Webb
Wildlife Management Area, FL (1 Sept 2015e1 Sept 2018) using scaled continuous variables: body mass (g), forearm length (mm), body condition index (forearm
length/body mass; BCI), testes size (length and width), and gular gland size (gular gland width and opening width). Each dot represents an individual male bat
during one capture event and each color and shaded area represents a status category for that bat classified at the time of that capture event (status categories
defined in methods). Legend abbreviations: “Dom”¼Dominant males, “C2”¼Contender 2 males, “C3”¼Contender 3 males, “NR”¼Non-reproductive males.
Right panel eExample of a bat classified as a dominant male with an open gular gland and descended testes.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
5
(function glht (R package multcomp)), which adjusts significance values for multiple comparisons. All reported errors are
standard error of the mean (±SE) unless otherwise noted.
3. Results
3.1. Roost composition
We uniquely tagged 341 Florida bonneted bats (206 females and 135 males) across all roosts (10 bat houses and 2 tree
roosts). We recaptured 207 bats at least once, with each bat captured an average of 3.05 ±0.14 times (range 1e11). During our
defined analysis period (1 Sept 2015e1 Sept 2018), individual bats were detected by capture or recorded by PIT tag readers at
an average of 2.18 ±0.08 different roosts (range 1e7). The number of individuals captured per occupied roost ranged from one
to 34 individuals (x¼14.83 ±1.30) and was female biased in most cases (sex ratio: x¼2.78 ±0.25, range 1e6), except in three
cases where only males were present (out of 41 total roost capture events, Supplementary Materials S2). Within each roost,
the number of females ranged from 0 to 24 individuals (x¼9.83 ±1.03), and the number of males from one to 10 males (x¼
3.83 ±0.36). Reproductive females were captured in nearly every capture event but occurred most frequently in April and
only rarely in December (Supplementary Materials S2). We classified the one or two reproductively active adult males that
had the greatest mass in each roost during each capture event as the most likely dominant males (Fig.1). On five occasions, we
assigned dominant status to males that were not captured (i.e., present and did not emerge from roosts) but that we
determined from PIT tag data were consistently present during that period and had been classified as the dominant males in
that roost before and after the capture event in question. Other males (contender 2, contender 3, and/or non-reproductive
males) were present in addition to dominant males during 84.2% (32 of 38) of roost capture events where any females
were captured (Supplementary Materials S2).
Dominant males tended to be larger (greater body mass, testes size and gular gland size) and be more roost faithful than
other males (Figs. 1 and 2). Body mass (ANOVA, F
5, 309
¼44.67, p<0.001), forearm length (F
5, 309
¼11.25, p<0.001), and roost
fidelity (F
5, 531
¼13.04, p<0.001) differed among male and female status categories. Dominant males and reproductive
females were significantly heavier than bats in all other categories except contender 2 males, which did not differ from
dominant males (Table 1;Fig. 2A). Forearm lengths were significantly greater in bats of all male status categories than all
female status categories, except between reproductive females and non-reproductive males, but did not differ among status
categories within each sex (Fig. 2B). Dominant males were present at the roosts where they were captured from 6 to 650 days
x¼195.05 ±43.90) during the study period and had significantly higher average fidelity to that roost between capture events
than males in the other status categories but did not differ from females (Fig. 2C).
3.2. Activity patterns
After filtering the PIT tag dataset based on our described criteria, we recorded a total of 183,349 nighttime detections over
the 3-year study period. PIT tag readers detected 256 of the 341 uniquely tagged bats (males: N¼100, females: N¼156) at
least once in the five focal roosts during the study period. For all three metrics of activity (roost activity, emergence times, and
initial foray duration), sex and status categories were included in the best LMMs, with a temperature variable improving all
models (Table 2).
Fig. 2. Differences among Florida bonneted bats (Eumops floridanus) in each status category captured at five bat houses at Fred C. Babcock-Cecil M. Webb Wildlife
Management Area, FL (1 Sept 2015e1 Sept 2018) for: A) body mass (g), B) forearm length (mm), and C) roost fidelity (%) to the roost where bats were captured
from one capture event to the next, standardized by the maximum number of days that any bat was present within that roost. Different letters indicate significant
differences. Status category abbreviations: RF ¼Reproductive females, NR F ¼Non-reproductive females, Dom M ¼Dominant males, C2 M ¼Contender 2 males,
C3 M ¼Contender 3 males, NR M ¼Non-reproductive males.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
6
3.2.1. Roost activity
Males were detected significantly more frequently at roosts each night (x¼3.57 ±0.22 detections/night, range: 1e305)
than females (x¼2.90 ±0.32 detections/night, range: 1e250), and both sexes were more active with increasing daily average
temperature (Table 2). Status was also a significant predictor of variation in roost activity (LRT:
c
2
¼77.43, p<0.001), and was
a better predictor than sex (Table 2). Dominant males were significantly more active at roosts than individuals in all other
status categories (Appendix A;Table 3). Seasonally, males exhibited peaks in roost activity in February and July/August, driven
primarily by individuals classified as the dominant males (Fig. 3A&B). Reproductive females, non-reproductive females, and
non-reproductive males had consistently low roost activity throughout the year, with a slight increase from June to
September.
3.2.2. Emergence times
On average, males emerged slightly earlier (x¼34.68 ±0.77 min past sunset) than females (x¼36.93 ±0.43 min past
sunset), with both sexes emerging closer to sunset as minimum daily temperature increased (Table 2). Again there was a
significant effect of status (LRT:
c
2
¼46.67, p<0.001) and minimum daily temperature on emergence times, with status
outperforming sex, but the patterns were less clear (Table 2). Dominant males, contender 3, and non-reproductive males
emerged earlier than reproductive females and contender 2 males (Appendix A). Non-reproductive males also emerged
earlier than non-reproductive females. No other pair-wise comparisons were statistically significant. Emergence times varied
seasonally, occurring closer to sunset as days lengthened from winter to summer (January to July/August) and occurring later
as days shortened again into winter (August to December; Fig. 3C&D).
Table 1
Mean (±SE) mass, forearm length, and body condition index (BCI; mass/forearm length) of Florida bonneted bats (Eumops floridanus) captured at Fred C.
Babcock-Cecil M. Webb Wildlife Management Area, FL (1 Sept 2015e1 Sept 2018). Total sample size exceeds the number of unique individuals captured
because some individuals captured multiple times exhibited different status during different capture events.
Sex Status nMass (g) Forearm (mm) BCI
Female Reproductive 74 48.14 ±0.41 62.89 ±0.14 0.766 ±0.006
Non-reproductive 125 40.87 ±0.43 62.78 ±0.10 0.651 ±0.007
Male Dominant 19 48.00 ±0.88 63.83 ±0.21 0.752 ±0.014
Contender2 29 45.07 ±0.64 63.77 ±0.24 0.707 ±0.010
Contender3 35 41.87 ±0.72 64.01 ±0.17 0.654 ±0.011
Non-reproductive 33 37.94 ±0.75 63.50 ±0.19 0.597 ±0.012
Table 2
Alternative models (linear mixed-effects models; log transformed Gaussian distribution) and predictor variables explaining activity patterns of Florida
bonneted bats (Eumops floridanus) detected at bat houses on Fred C. Babcock-Cecil M. Webb Wildlife Management Area, FL (1 Sept 2015e1 Sept 2018). For
each response variable, the top two models are listed with associated
D
AIC (difference between Akaiki Information Criteria [AIC] score of each alternative
model and model with the lowest AIC score) and Akaike weights (
u
i
).
Model - Parameters K
D
AIC
u
i
ROOST ACTIVITY
Sex þDaily Average Temperature 8 0.00 1.00
Daily Average Temperature 7 37.74 0.00
Null (intercept only) 6 1445.93 0.00
Status þDaily Average Temperature 12 0.00 1.00
Sex þDaily Average Temperature 8 43.97 0.00
Null (intercept only) 6 952.00 0.00
EMERGENCE TIMES
Sex þDaily Minimum Temperature 8 0.00 0.84
Daily Minimum Temperature 6 3.26 0.16
Null (intercept only) 6 2586.89 0.00
Status þDaily Minimum Temperature 13 0.00 1.00
Sex þDaily Average Temperature 8 28.97 0.00
Null (intercept only) 6 1837.60 0.00
INITIAL FORAY DURATION
Sex þ(Daily Maximum Temperature)^2 9 0.00 1.00
Daily Maximum Temperature 6 3.26 0.16
Null (intercept only) 6 2586.89 0.00
Status þ(Daily Maximum Temperature)^2 13 0.00 1.00
(Daily Maximum Temperature)^2 8 67.42 0.00
Null (intercept only) 6 18709.27 0.00
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
7
Table 3
Summary of activity patterns for Florida bonneted bats (Eumops floridanus) detected with PIT Tag readers at bat houses on Fred C. Babcock-Cecil M. Webb
Wildlife Management Area, FL. Mean (±SE) values are listed for all roosts over the study period (1 Sept 2015e1 Sept 2018). Samples sizes (n) differ slightly
because of the filtering criteria applied to the dataset before each analysis. Total sample size exceeds the number of unique individuals captured because
some individuals captured multiple times exhibited different status during different capture events. Status category abbreviations: Repro F ¼Reproductive
females, NR F ¼Non-reproductive females, Dom M ¼Dominant males, C2 M ¼Contender 2 males, C3 M ¼Contender 3 males, NR M ¼Non-reproductive
males.
Sex Status nRoost activity
(detections per night)
Emergence time (mins
after sunset)
Initial Foray Duration (hours,
percent of night)
No. forays
per night
Total foray duration (hours,
percent of night)
Female Repro
F
90
e91
2.71 ±0.07 37.58 ±0.53 4.02 ±0.12 (37.3 ±1.1%) 1.36 ±0.03 4.84 ±0.13 (44.9 ±1.1%)
NR F 134
e136
2.96 ±0.37 36.61 ±0.52 3.93 ±0.11 (33.4 ±0.9%) 1.64 ±0.37 4.52 ±0.12 (38.4 ±1.0%)
Male Dom
M
21
e23
5.49 ±0.61 33.11 ±1.30 2.40 ±0.25 (21.0 ±2.3%) 2.54 ±0.25 3.67 ±0.27 (32.1 ±2.5%)
C2 M 26
e28
3.08 ±0.28 43.60 ±2.63 2.79 ±0.31 (25.0 ±2.9%) 1.55 ±0.12 3.43 ±0.32 (30.6 ±3.0%)
C3 M 45
e50
3.00 ±033 34.56 ±0.85 3.29 ±0.22 (28.4 ±2.1%) 1.45 ±0.13 3.59 ±0.22 (30.9 ±2.0%)
NR M 33
e34
2.91 ±0.21 31.92 ±1.07 3.24 ±0.27 (26.9 ±2.4%) 1.39 ±0.09 3.59 ±0.30 (29.8 ±2.6%)
Fig. 3. Seasonal activity patterns of Florida bonneted bats (Eumops floridanus) detected by PIT tag readers at five bat houses in Fred C. Babcock-Cecil M. Webb
Wildlife Management Area, FL (1 Sept 2015e1 Sept 2018). Figures represent mean monthly values for A) roost activity (number of detections per night) by sex B)
roost activity by status categories, C) emergence times (first detection per night per bat) by sex, and D) emergence times by status categories.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
8
3.2.3. Forays
The number of forays per bat per night varied widely from one to 125 (x¼1.65 ±0.13). Across all status categories bats
spent the most time away from the roost each night during the initial foray (x¼3.56 ±0.08 h), which amounted to 31.1 ±0.8%
of the night, with decreasing time spent on each subsequent foray (Fig. 4). Males had significantly shorter initial foray du-
rations (x¼2.95 ±0.14 h; 25.5 ±0.1%) than females (x¼3.95 ±0.09 h; 34.6 ±0.8%; Table 2,Fig. 5). The temperature term in
the best model was quadratic: both male and female foray duration initially increased with increasing daily maximum
temperature until around 30
C at which point they decreased with higher temperatures (Table 2). Status (LRT:
c
2
¼77.43,
p<0.001) and maximum daily temperature were also significant predictors for initial foray duration, again outperforming the
model including sex (Table 2). Dominant males had significantly shorter initial forays than reproductive females, non-
reproductive females, and contender 3 males (Appendix A;Table 3;Fig. 5). Contender 2, contender 3, and non-
reproductive males also had shorter initial forays than both female categories. No other pairwise comparisons were statis-
tically significant. Seasonally, both males and females showed slightly reduced foray durations during colder months
(DecembereJanuary; Fig. 5). Dominant males had consistently low initial foray durations throughout the year, while the other
male status categories were much more variable. Non-reproductive and reproductive females followed similar patterns
seasonally, showing increased initial foray durations in the early spring (FebruaryeMarch) and summer (JuneeAugust; Fig. 5).
In addition to having the shortest initial forays overall, dominant males exhibited a greater number of forays per night than
females and other males but spent a similar amount of total time away from the roost each night as other males (Table 3).
4. Discussion
We provide multiple lines of evidence that Florida bonneted bats form small harem groups that are active year round and
exhibit resource-defense polygyny at roost sites, a social system much more typical of tropical species than temperate species
(Kerth, 2008). Similar to Ober et al. (2017a), we found strongly female-biased sex ratios and support for each harem group
having a dominant male that is a large, reproductively active adult with an open gular gland. As in other species exhibiting
resource-defense polygyny, the roost site may represent a critical and defendable resource for male Florida bonneted bats to
gain access to females.
4.1. Roost defense
If these bats exhibited female-defense polygyny or resource-defense at foraging sites, we would have expected nightly
activity of males and females to be very similar. In contrast, we found that activity patterns between sexes differed across all
metrics, suggesting that male and female Florida bonneted bats behaved differently and foraged separately at night. Our
results are supported by a recent study that used Global Positioning System (GPS tags to track movements of 20 Florida
bonneted bats and found that males had smaller home ranges, shorter foray loop lengths and traveled shorter maximum
distances from their roosts than females (Webb, 2018). Additionally, similar to Ober et al. (2017b), we found that males in all
status categories had larger forearm lengths than females (and hence larger wings). It is possible that the reduced wing
loading from larger wings may benefit males by conferring greater maneuverability and lower energy expenditure if they
invest considerable time flying near roosts during defense. In greater sac-winged bats (S. bilineata), males are thought to have
evolved to be smaller (with lower wing loading) than females to reduce energetic costs of territorial maneuvers and courtship
displays involving prolonged periods of hovering at the roost (Voigt et al., 2005).
Fig. 4. Foray durations of Florida bonneted bats (Eumops floridanus) detected by PIT tag readers at five bat houses on Fred C. Babcock-Cecil M. Webb Wildlife
Management Area, FL (1 Sept 2015e1 Sept 2018). Boxplots represent the time (hours) that bats spent away from the roost during the first four forays each night
for individuals of different status categories.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
9
The largest male Florida bonneted bats in each roost, all of which were reproductively active (open gular glands and/or
descended testes), were consistently detected more frequently at roost entrances each night than other individuals. These
males also had higher fidelity to roost sites than smaller males. We suspect these are the dominant males attempting to
defend their roosts and suggest there may be a size-biased male hierarchy, similar to that observed for Jamaican fruit bats
(Ortega and Arita, 1999) and many other vertebrates in which larger males have a competitive advantage during mate
guarding and aggressive displays (Archer, 1988). Although they spent similar total time away from the roost each night as the
other males, these large males returned to the roost more quickly after the initial foray and subsequently engaged in addi-
tional shorter forays, returning more frequently to the roost throughout the night. By remaining vigilant at the roost entrance,
dominant males may be able to deter or physically exclude intruder males from entering the roost. In other harem-forming
species, it is suspected that frequent activity at the roost by the dominant male provides unpredictability for competing males
attempting to enter even if there is no physical interaction between males (Kunz et al., 1998;McCracken and Wilkinson,
2000). Contrary to our prediction that dominant males would emerge later than females, we found that males overall
emerged slightly earlier than females, with dominant males emerging earlier than reproductive but not non-reproductive
females. This differs from Indian short-nosed fruit bats (C. sphinx), where dominant harem males emerged after all the fe-
males had left the roost (Mahandran et al., 2018). Rather than remaining in the roostto defend it until the colony has emerged,
dominant male Florida bonneted bats tend to emerge earlier, which we suspect allows them to discourage other males as the
harem emerges.
Our findings that dominant male Florida bonneted bats spent more time at the roost and less time foraging than females
and other males align closely with those in other harem-forming bat species that defend roost sites (e.g., spotted-winged fruit
bat, Balionycteris maculate; Seba’s short-tailed fruit bat, Carollia perspicillata (McCracken and Wilkinson, 2000). In greater
spear-nosed bats (Phyllostomus hastatus; characterized as a female-defense mating system but exhibiting characteristics of
resource-defense at roosts), dominant males spent 25% less time away from the roost and had 3.5 times more forays each
night than females (Kunz et al., 1998). These activity patterns are comparable to our study where, on average, dominant males
spent 40% less time away from the roost during the initial foray and made nearly twice as many forays as females. This
increased time spent at roosts may lead to an energetic trade-off for males balancing mate acquisition and harem mainte-
nance with their own need for energy intake. Mahandran et al. (2018) found that harem male Indian short-nosed fruit bats
invest around 50% of their nightly activities in roost guarding and upkeep; this time investment was positively associated
with female group size, suggesting that males with the largest harems spent the least amount of time foraging. In our study,
we estimated that dominant male Florida bonneted bats spent an average of 67.9% of each night at the roost. In addition to
reduced time spent foraging, the cost of roost defense, vigilance, and harem maintenance may be energetically expensive for
males. In greater sac-winged bats, the metabolic rates of males increased with the number of females in their harems because
of the greater number of flight maneuvers they performed to maintain and defend these larger harems (Voigt and von
Helversen, 1999;Voigt et al., 2001). In our study, we were unable to quantify the proportion of time male Florida bonn-
eted bats spent at the roost in defense versus resting or how energy expenditures may change with harem size; we
recommend further research to better understand energetic tradeoffs and harem dynamics.
4.2. Presence of additional non-dominant males in harem groups
Within many harem-forming bat species, a single adult male typically guards the harem (e.g., little free-tailed bat,
Chaerephon pumila (McWilliam, 1988); common vampire bat, Desmodus rotundus (Park,1991); Seba’s short-tailed bat, Carollia
Fig. 5. Seasonal changes in the initial foray duration (proportion of time spent away from roosts relative to the total night length) for Florida bonneted bats
(Eumops floridanus) detected by PIT tag readers at bat houses in Fred C. Babcock-Cecil M. Webb Wildlife Management Area, FL (1 Sept 2015e1 Sept 2018). Figures
represent mean monthly values for individuals of different A) sex and B) status categories.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
10
perspicillata (Williams, 1986). However, multiple adult males may also be present. In Jamaican fruit bats, subordinate males
can be part of large harem groups (Ortega and Arita, 1999); these males tend to be reproductively active adults (descended
testes) and slightly smaller than dominant males but larger than non-reproductive (abdominal testes) satellite males that
roost peripherally to the harems. In our study, nearly all roost groups had what we could consider smaller subordinate
(contender 2 and 3) and satellite (non-reproductive) males in addition to the dominant male.In several cases there were two
males in a roost that were both classified as dominant because of indistinguishable morphological characteristics. It is
possible that fission occurs in larger groups and these could be examples of two harems within the same roost or where a
second large male competes for the dominant position. Because we do not have visual recordings, we don’t know if there is
any subgroup separation within the roosts or if satellite males roost apart from the main group. Further, in the instances
where bats roosted in separate bat houses on the same pole, we were unable to distinguish which bats occupied each house;
thus, some of our groups may represent two distinct harem groups or contain satellite males.
It is unclear if the additional males within the harem are a cost to the dominant male through increased male-male in-
teractions, competition for copulations with females, and/or a decrease in paternity. In Jamaican fruit bats, subordinates may
actually benefit the dominant male by participating in roost defense and reducing the number of visits from foreign males,
allowing the harem sizes to be larger (Ortega et al., 2003). In turn, although the subordinates do not receive much benefitin
terms of immediate increased paternity, their long-term reproductive success may increase because they are more likely to
take over the dominant male position once it is vacated. Florida bonneted bats may similarly exhibit a male status hierarchy in
which multiple reproductively active males occur in one roost, with the largest male occupying the dominant role until he is
out-competed or otherwise disappears. Anecdotally, in one roost we observed that a bat defined as a contender 2 male
increased his roost activity immediately following the apparent disappearance (no more detections on PIT tag reader) of the
male that had been defined as dominant for nearly 3 years; this contender 2 male was subsequently identified as the new
dominant male at that roost in the next capture event based on his size relative to other males. We recommend further
research into male turnover, reproductive success, and colony relatedness to determine the benefits conferred by male hi-
erarchical status.
4.3. Seasonality
We confirmed that Florida bonneted bats are active in and out of roosts year round, which differs from another species of
Eumops in the United States (western mastiff bat, E. perotis) that changes roosts seasonally and may undergo local migrations
(Best et al., 1996). As we predicted, temperature influenced the bats’activity patterns: as temperature increased, bats went
into and out of the roost more often, emerged earlier, and spentmore time foraging (until a maximum daily temperature of ca.
30
C). This supports previous research suggesting that the likelihood of detection of this species increases as temperature
increases (Bailey et al., 2017) and other studies showing that many bat species increase foraging activity as temperature
increases, likely in response to insect prey (Lewis and Dibley, 1970;Hayes, 1997;Meyer et al., 2016). Further investigation into
the influence of abiotic factors is warranted but beyond the scope of this paper.
Despite temperature being an important predictor, it did not entirely explain the observed seasonal activity patterns,
particularly for roost activity and initial foray duration. The peaks in male roost activity in early spring and late summer and
the peak in female initial foray duration in early spring may correspond toreproductivecycles. However, our understanding of
reproduction in Florida bonneted bats remains limited and our interpretations of seasonal activity patterns are only spec-
ulative. This species appears to be aseasonally polyestrous with the primary period of parturition beginning in early May at
which point non-volant pups are present in roosts (Ober et al., 2017a). Although pregnant females have been reported in
many months of the year (AprileSeptember; Barbour and Davis, 1969;Belwood, 1981;Robson, 1989;Ober et al., 2017a), the
prevalence of pregnant females captured in April in our study suggests that peak pregnancy occurs in the spring. Thus, longer
initial foray durations may be required for females in the early spring during early pregnancy and in the summer during
lactation when energetic demands are highest. Similarly, the peaks in male activity in February and July/August may
correspond to mating and greater male-male competition at roosts. We do not know when mating occurs for this species;
however, based on the appearance of pups in early May and the estimated gestation period of a closely related species, E.
perotis, (ca. 80e90 days; Ammerman et al., 2012), we suspect that copulation mayoccur around the time that we observed the
February peak in dominant male activity. While the spike in activity in late summer may also correspond to copulation
(pregnant females captured in late August), it could be associated with increased movements by individuals among roost
groups when pups born in May become volant, as has been observed in other harem forming species (Ortega and Arita, 1999).
Contender 2 and 3 males that appeared to be reproductively active but were not the largest males in the roost also showed
slight increases in activity during these times; it is unclear whether these increases in activity indicate possible competition
for the dominant position and/or attempts to copulate with females or may aid in deterring outside males, as has been
observed in other species (Ortega et al., 2003). More research on the timing of reproduction and corresponding competitive
interactions is needed to fully understand the seasonal activity patterns for this species.
5. Conclusions and conservation implications
We provide evidence that Florida bonneted bats form year-round harems and exhibit resource-defense polygyny, with the
largest reproductively active males in each roost defending roost sites. The activity patterns of these males suggest they play
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
11
distinct social roles. Dominant males may act as key contributors to the overall stability and viability of the social group or
population, and loss of such individuals may have a disproportionate effect on the resiliency of the population (Snijders et al.,
2017). Although we do not know the level of competition for mates and roost sites, we suspect that roosts are limited and
declining (United States Fish and Wildlife Service [USFWS] 2013;Braun de Torrez et al., 2016) and that competition may
increase as suitable roost sites continue to decline across the species range. In regions with few roost structures, males may
face increasing pressure to defend roost sites at the expense of foraging, and high male turnover may lead to colony instability
and reduced reproductive success (Kerth, 2008;Kerth et al., 2011). This has been seen in other species exhibiting resource-
defense polygyny at roost sites (Jamaican fruit bats), where harems in areas with abundant roost sites are quite stable (Kunz
et al., 1983), while harems using highly contested and limited roost sites have greater male turnover (Morrison, 1980).
Increased aggressive interactions can lead to higher metabolic output, stress, injury, and spread of disease in many animals
(Snijders et al., 2017), including bats (Kunz et al., 1998;McCracken and Bradbury, 1981;Voigt et al., 2001). In Florida bonneted
bats, we do not know how males defend roosts (e.g., physical interaction, territorial displays) or the energetic costs associated
with defense. We also do not understand how male-male competition may vary with proximity to other roosts. Males may
face the greatest competition from other males in neighboring harem groups (Ortega and Arita, 1999), or individuals may be
less aggressive toward their immediate neighbors because they are familiar individuals that pose less of a threat (Temeles,
1994). Understanding how roost distribution affects colony stability in the context of resource-defense polygyny has crit-
ical implications for installing configurations of artificial roosts as a conservation strategy for Florida bonneted bats (e.g.,
identification of roost proximity guidelines). Finally, if dominant males spend time closer to the roost at the expense of
foraging, we need to consider the quality of foraging resources near roost sites in addition to sites where females may be
foraging farther away. We encourage further investigation into male competition, colony stability, and reproductive success
for Florida bonneted bats across their range. Broadly, our study highlights the importance of accounting for differences in
behavior across demographic groups and social roles when considering resource needs and conservation plans for imperiled
species.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Acknowledgements
We thank the many people who have assisted with data collection and management, including but not limited to, J. Myers,
K. Smith, C. Pope, L. Smith, K. Mobley, W. Gurley, S. Sofferin,M. Berger, T. Doonan, R. Arwood, M. Wallrichs, M. Bailey, and many
other biologists and staff at Fred C. Babcock-Cecil M. Webb Wildlife Management Area. This research was funded by the
Florida Fish and Wildlife Conservation Commission, United States Fish and Wildlife Service, and Bat Conservation
International.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.gecco.2020.e01289.
Appendix A
Coefficient estimates and pairwise comparisons (Tukey post-hoc tests adjusted for multiple comparisons) for significant
predictor variables are reported for the best model only. Status category abbreviations: Repro F ¼Reproductive females, NR F
¼Non-reproductive females, Dom M ¼Dominant males, C2 M ¼Contender 2 males, C3 M ¼Contender 3 males, NR M ¼Non-
reproductive males.
Response Variable Model Parameters ePairwise comparisons
b
Estimates SE p-value
Roost activity Sex þDaily Average Temperature Intercept 0.261 0.059 <0.001
Sex: Male - Female 0.156 0.030 <0.001
Daily Average Temperature 0.014 0.000 <0.001
Status þDaily Average Temperature Intercept 0.210 0.071 0.003
Daily Average Temperature 0.013 0.000 <0.001
Status: Repro F - Dom M 0.451 0.049 <0.001
Status: NR F - Dom M 0.464 0.047 <0.001
Status: C2 M - Dom M 0.366 0.053 <0.001
Status: C3 M - Dom M 0.367 0.050 <0.001
Status: NR M - Dom M 0.352 0.057 <0.001
Status: NR F - Repro F 0.013 0.020 0.984
Status: C2 M - Repro F 0.086 0.042 0.299
Status: C3 M - Repro F 0.085 0.036 0.170
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
12
(continued )
Response Variable Model Parameters ePairwise comparisons
b
Estimates SE p-value
Status: NR M - Repro F 0.099 0.043 0.189
Status: C2 M - NR F 0.099 0.041 0.136
Status: C3 M - NR F 0.098 0.035 0.052
Status: NR M - NR F 0.112 0.042 0.078
Status: C3 M - C2 M 0.001 0.042 1.000
Status: NR M - C2 M 0.013 0.049 1.000
Status: NR M - C3 M 0.014 0.046 1.000
Emergence times Sex þDaily Minimum Temperature Intercept 4.266 0.026 <0.001
Sex: Male - Female 0.049 0.021 0.020
Daily Minimum Temperature 0.010 0.000 <0.001
Status þDaily Minimum Temperature Intercept 4.164 0.042 <0.001
Status: Repro F - Dom M 0.106 0.036 0.033
Status: NR F - Dom M 0.075 0.035 0.243
Status: C2 M - Dom M 0.154 0.040 0.002
Status: C3 M - Dom M 0.014 0.037 0.999
Status: NR M - Dom M 0.050 0.042 0.831
Status: NR F - Repro F 0.031 0.015 0.293
Status: C2 M - Repro F 0.048 0.032 0.629
Status: C3 M - Repro F 0.091 0.026 0.006
Status: NR M - Repro F 0.156 0.031 <0.001
Status: C2 M - NR F 0.079 0.031 0.094
Status: C3 M - NR F 0.060 0.025 0.141
Status: NR M - NR F 0.125 0.030 <0.001
Status: C3 M - C2 M 0.139 0.032 <0.001
Status: NR M - C2 M 0.204 0.037 <0.001
Status: NR M - C3 M 0.064 0.034 0.379
Daily Minimum Temperature 0.010 0.000 <0.001
Initial foray duration Sex þ(Daily Maximum Temperature)^2 Intercept 1.360 0.113 <0.001
Sex: Male - Female 0.699 0.067 <0.000
Daily Maximum Temperature, 1 18.463 1.498 <0.001
Daily Maximum Temperature, 2 37.632 1.300 <0.001
Status þ(Daily Maximum Temperature)^2 Intercept 1.380 0.105 <0.001
Status: Repro F - Dom M 0.715 0.102 <0.001
Status: NR F - Dom M 0.738 0.100 <0.001
Status: C2 M - Dom M 0.152 0.109 0.715
Status: C3 M - Dom M 0.306 0.104 0.035
Status: NR M - Dom M 0.229 0.121 0.385
Status: NR F - Repro F 0.023 0.038 0.990
Status: C2 M - Repro F 0.564 0.090 <0.001
Status: C3 M - Repro F 0.410 0.080 <0.001
Status: NR M - Repro F 0.486 0.094 <0.001
Status: C2 M - NR F 0.586 0.088 <0.001
Status: C3 M - NR F 0.433 0.077 <0.001
Status: NR M - NR F 0.509 0.092 <0.001
Status: C3 M - C2 M 0.154 0.088 0.471
Status: NR M - C2 M 0.078 0.103 0.972
Status: NR M - C3 M 0.076 0.099 0.970
Daily Maximum Temperature, 1 18.346 1.548 <0.001
Daily Maximum Temperature, 2 30.554 1.291 <0.001
References
Adams, D.M., Li, Y., Wilkinson, G.S., 2018. Male scent gland signals mating status in greater spear-nosed bats, Phyllostomus hastatus. J. Chem. Ecol. 44,
975e986.
Ammerman, L.K., Hice, C.L., Schmidly, D.J., 2012. Bats of Texas. Texas A&M University Press, College Station.
Angell, E., Thompson, G., 2015. Second record of a natural Florida Bonneted Bat (Eumops floridanus) roost. Fla. Field Nat. 43, 185e188.
Archer, J., 1988. The Behavioural Biology of Aggression. Cambridge University Press, NYC, New York.
Bailey, A.M., McCleery, R.A., Ober, H.K., Pine, W.E., 2017a. First demographic estimates for endangered Florida bonneted bats suggestyear-round recruitment
and low apparent survival. J. Mammal. 98, 551e559.
Bailey, A.M., Ober, H.K., Sovie, A., McCleery, R., 2017b. Impact of land use and climate on the distribution of the endangered Florida bonneted bat. J. Mammal.
98, 1586e1593.
Balasingh, J., Koilraj, J., Kunz, T.H., 1995. Tent construction by the short-nosed fruit bat Cynopterus sphinx (Chiroptera: Pteropodidae) in southern India.
Ethology 100, 210e229.
Barbour, R.W., Davis, W.H., 1969. Bats of America. University Press of Kentucky, Lexington.
Belwood, J.J., 1981. Wagner’s mastiff bat, Eumops glaucinus floridanus, (Molossidae) in southwestern Florida. J. Mammal. 62, 411e413.
Best, T.L., Kiser, M.W., Freeman, P.W., 1996. Eumops perotis. Mamm. Species 534, 1e8.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
13
Best, T.L., Kiser, M.W., Rainey, J.C., 1997. Eumops glaucinus. Mamm. Species 551, 1e6.
Bolker, B.M., Brooks, M.E., Clark, C.J., Geange, S.W., Poulsen, J.R., Stevens, M.H.H., White, J.-S.S., 2009. Generalized linear mixed models: A practical guide for
ecology and evolution. Trends in Ecology &Evolution 24, 127e135. https://doi.org/10.1016/j.tree.2008.10.008.
Bowles, J.B., Heideman, P.D., Erickson, K.R., 1990. Observations on six species of free-tailed bats (Molossidae) from Yucatan, Mexico. SW. Nat. 35, 151e157.
Bradbury, J.W., Vehrencamp, S.L., 1976. Social organization and foraging in Emballonurid bats. Behav. Ecol. Sociobiol. 1, 337e381.
Brakes, P., Dall, S.R.X., Aplin, L.M., Bearhop, S., Carroll, E.L.,Ciucci, P., Fishlock, V., Ford, J.K.B., Garland, E.C., Keith, S.A., McGregor, P.K., Mesnick, S.L., Noad, M.J.,
di Sciara, G.N., Robbins, M.M., Simmonds, M.P., Spina, F., Thornton, A., Wade, P.R., Whiting, M.J., Williams, J., Rendell, L., Whitehead, H., Whiten,A.,
Rutz, C., 2019. Animal cultures matter for conservation. Science 363, 1032e1034.
Braun de Torrez, E.C., Ober, H.K., McCleery, R.A., 2016. Use of a multi-tactic approach to locate an endangered Florida Bonneted Bat roost. SE. Nat. 15,
235e242.
Burnham, K.P., Anderson, D.R., 2002. Model selection and multimodel inference: an information-theoretic approach. Springer Science, New York, New York,
USA.
Campbell, P., Akbar, Z., Adnan, A.M., Kunz, T.H., 2006. Resource distribution and social structure in harem-forming Old World fruit bats: variationsona
polygynous theme. Anim. Behav. 72, 687e698.
Clutton-Brock, T., 1989. Mammalian mating systems. Proc. R. Soc. Lond. B Biol. Sci. 236, 339e372.
Dechmann, D.K.N., Kalko, E.K.V., K€
onig, B., Kerth, G., 2005. Mating system of a Neotropical roost-making bat: the white-throated, round-eared bat,
Lophostoma silvicolum (Chiroptera: Phyllostomidae). Behav. Ecol. Sociobiol. 58, 316e325.
Dwyer, P.D., 1970. Social organization of the bat Myotis adversus. Science 168, 1006e1008 .
Emlen, S.T., Oring, L.W., 1977. Ecology, sexual selection, and the evolution of mating systems. Science 197, 215e223.
French, B., Lollar, A., 1998. Observations on the reproductive behavior of captive Tadarida brasiliensis mexicana. SW. Nat. 43, 484e490.
Garg, K.M., Chattopadhyay, B., Swami Doss, D.P., Kumar, A.K.V., Kandula, S., Ramakrishnan, U., 2015. Males and females gain differentially from sociality in a
promiscuous fruit bat Cynopterus sphinx. PloS One 10, e0122180.
Goldenberg, S.Z., Owen, M.A., Brown, J.L., Wittemyer, G., Oo, Z.M., Leimgruber, P., 2019. Increasing conservation translocation success by building social
functionality in released populations. Global Ecol. Conserv. 18, e00604.
Günther, L., Lopez, M.D., Kn€
ornschild, M., Reid, K., Nagy, M., Mayer, F., 2016. From resource to female defence: the impact of roosting ecology on a bat’s
mating strategy. Roy. Soc. Open Sci. 3, 160503.
Hayes, J., 1997. Temporal variation in activity of bats and the design of echolocation-monitoring studies. J. Mammal. 78, 514e524.
Heideman, P.D., Erickson, K.R., Bowles, J.B., 1990. Notes on the breeding biology, gular gland and roost habits of Molossus sinaloae (Chiroptera, Molossidae).
Z. Saugetierkunde 55, 303e307.
Kerth, G., 2008. Causes and consequences of sociality in bats. Bioscience 58, 1e10.
Kerth, G., Perony, N., Schweitzer, F., 2011. Bats are able to maintain long-term social relationships despite the high fission-fusion dynamics of their groups.
Proc. Biol. Sci. 278, 2761e2767.
Kunz, T.H., August, P.V., Burnett, C.D., 1983. Harem social organization in cave roosting Artibeus jamaicensis (Chiroptera: Phyllostomidae). Biotropica 15,
133e138.
Kunz, T.H., Robson, S.K., Nagy, K., 1998. Economy of harem maintenance in the greater spear-nosed bat, Phyllostomus hastatus. J. Mammal. 79, 631e642.
Lewis, T., Dibley, G.C., 1970. Air movement near windbreaks and a hypothesis of the mechanism of the accumulation of airborne insects. Ann. Appl. Biol. 66,
477e484.
Lima, S.L., Dill, L.M., 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Can. J. Zool. 68, 619e640.
Mahandran, V., Murugan, C.M., Nathan, P.T., 2018. Effect of female group size on harem male roosting behavior of the Indian short-nosed fruit bat Cyn-
opterus sphinx. Acta Ethol. 21, 43e49.
McCracken, G.F., Bradbury, J.W., 1981. Social organization and kinship in the polygynous bat, Phyllostomus hastatus. Behav. Ecol. Sociobiol. 8, 11e34.
McCracken, G.F., Wilkinson, G.S., 2000. Bat mating systems. In: Crichton, E.G., Krutzsch, P.H. (Eds.), Reproductive Biology of Bats. Academic Press, New York,
NY, pp. 321e362.
McWilliam, A.N., 1988. Social organisation of the bat Tadarida (Chaerephon) pumila (Chiroptera: Molossidae) in Ghana, west Africa. Ethology 77, 115e124.
Meyer, G.A., Senulis, J.A., Reinartz, J.A., 2016. Effects of temperature and availability of insect prey on bat emergence from hibernation in spring. J. Mammal.
97, 1623e1633.
Morrison, D.W., 1980. Foraging and day-roosting dynamics of canopy fruit bats in Panama. J. Mammal. 61, 20e29.
O’Shea, T.J., Cryan, P.M., Hayman, D.T.S., Plowright, R.K., Streicker, D.G., 2016. Multiple mortality events in bats: a global review. Mamm Rev. 46, 175e190.
Ober, H.K., Braun de Torrez, E.C., Gore, J.A., Bailey, A.M., Myers, J.K., Smith, K.N., McCleery, R.A., 2017a. Social organization of an endangered subtropical
species, Eumops floridanus, the Florida bonneted bat. Mammalia 81, 375e383.
Ober, H.K., Braun de Torrez, E.C., McCleery, R.A., Bailey Amanda, M., Gore, J.A., 2017b. Sexual dimorphism in the endangered Florida bonneted bat, Eumops
floridanus (Chiroptera: Molossidae). Fla. Sci. 80, 38e48.
Ortega, J., Arita, H.T., 1999. Structure and social dynamics of harem groups in Artibeus jamaicensis (Chiroptera: Phyllostomidae). J. Mammal. 80, 1173e1185 .
Ortega, J., Maldonado, J.E., Wilkinson, G.S., Arita, H.T., Fleischer, R.C., 2003. Male dominance, paternity, and relatedness in the Jamaican fruit-eating bat
(Artibeus jamaicensis). Mol. Ecol. 12, 2409e2415.
Park, S.R., 1991. Development of social structure in a captive colony of the common vampire, Desmodus rotundus. Ethology 89, 335e341.
Parreira, B.R., Chikhi, L., 2015. On some genetic consequences of social structure, mating systems, dispersal, and sampling. Proc. Natl. Acad. Sci. Unit. States
Am. 112, E3318eE3326.
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., R Development Core Team, 2012. Nlme: linear and nonlinear mixed effects models. R package version 3, 1e105.
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., R. Development Core Team, 2018. Nlme: Linear and nonlinear mixed effects models. R package version 3. R
Foundation for Statistical Computing, Vienna, pp. 1e137. Available at. https://CRAN.Rproject.org/package¼nlme.
Racey, P.A., 2009. Reproductive assessment of bats. In: Kunz, T., Parsons, S. (Eds.), Ecological and Behavioral Methods for the Study of Bats. Johns Hopkins
University Press, Baltimore, MD, pp. 249e264.
Robson, M.S., 1989. Status Survey of the Florida Mastiff Bat. Final Performance Report. Bureau of Nongame Wildlife, Division of Wildlife, FL Game &Fresh
Water Fish Commission.
Sikes, R.S., Animal Care Use Committee of the American Society of Mammalogists, 2016. 2016 Guidelines of the American Society of Mammalogists for the
use of wild mammals in research and education. J. Mammal. 97, 663e688.
Snijders, L., Blumstein, D.T., Stanley, C.R., Franks, D.W., 2017. Animal social network theory can help wildlife conservation. Trends Ecol. Evol. 32, 567e577.
Storz, J.F., Balasingh, J., Thiruchenthil, N., Emmanuel, K., Kunz, T.H., 2000. Dispersion and site fidelity in a tent-roosting population of the short-nosed fruit
bat (Cynopterus sphinx) in Southern India. J. Trop. Ecol. 16, 117e131.
Temeles, E.J., 1994. The role of neighbors in territorial systems: when are they ’dear enemies’? Anim. Behav. 47, 339e350.
Thies, W., Kalko, E.K.V., Schnitzler, H.U., 2006. Influence of environment and resource availability on activity patterns of Carollia castanea (Phyllostomidae)
in Panama. J. Mammal. 87, 331e338.
United States Fish and Wildlife Service [USFWS], 2013. Endangered and threatened wildlife and plants: endangered species status for the Florida bonneted
bat. Fed. Regist. 78 (191), 61004e61043.
Voigt, C.C., Heckel, G., Mayer, F., 2005. Sexual selection favours small and symmetric males in the polygynous greater sac-winged bat Saccopteryx bilineata
(Emballonuridae, Chiroptera). Behav. Ecol. Sociobiol. 57, 457e464.
Voigt, C.C., von Helversen, O., 1999. Storage and display of odor by male Saccopteryx bilineata (Chiroptera; Emballonuridae). Behav. Ecol. Sociobiol. 47,
29e40.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
14
Voigt, C.C., von Helversen, O., Michener, R., Kunz, T.H., 2001. The economics of harem maintenance in the sac-winged bat, Saccopteryx bilineata (Embal-
lonuridae). Behav. Ecol. Sociobiol. 50, 31e36.
Webb, E., 2018. Foraging Ecology of the Florida Bonneted Bat, Eumops Floridanus. Thesis. University of Florida, Gainesville, FL.
Wilde, L.R., Günther, L., Mayer, F., Kn€
ornschild, M., Nagy, M., 2018. Thermoregulatory requirements shape mating opportunities of male proboscis bats.
Front. Ecol. Evol. 6, 1e14.
Williams, C.F., 1986. Social organization of the bat, Carollia perspicillata (Chiroptera: Phyllostomidae). Ethology 71, 265e282.
E.C. Braun de Torrez, J.A. Gore and H.K. Ober Global Ecology and Conservation 24 (2020) e01289
15
