Content uploaded by David Lahti
Author content
All content in this area was uploaded by David Lahti on Apr 14, 2020
Content may be subject to copyright.
BRIEF COMMUNICATION
doi:10.1111/evo.13934
Rapid evolution by sexual selection in a
wild, invasive mammal
M. Aaron Owen1,2,3 and David C. Lahti1,2
1Department of Biology, Queens College, City University of New York, 65-30 Kissena Blvd., Flushing, NY 11367
2Graduate Subprogram in Ecology, Evolution, and Behavior, City University of New York, 365 5th Ave., New York, NY
10016
3E-mail: m.aaron.owen@gmail.com
Received February 11, 2019
Accepted December 21, 2019
Sexual selection theory provides a framework for investigating the evolution of traits involved in attracting and competing for
mates. Given the sexual function of such traits, studies generally focus on individual interactions (i.e., displays and contests) in ex-
plaining trait origin and persistence. We show that ecological factors can strongly inuence the adaptive value of these traits, and
changes to these factors can lead to rapid evolutionary change. We compared sexually selected traits in the small Indian mongoose
(Urva auropunctata) between their sparsely populated native range and four tropical islands to which they were introduced within
the last 150 years and where, due to a lack of interspecic competition and predation, they have become invasive and densely
populated. Because of a likely increase in encounter rate, we predicted that selection on long-distance chemical advertisement by
males would relax in the introduced range. Accordingly, male, but not female, anal pads (used in scent marking) decreased in size
in relation to both time since introduction and population density, and their relationship to body size and condition weakened.
Concurrently, as predicted by intensied sperm competition, testis size increased following introduction. The small Indian mon-
goose thus experienced an inversion in the relative contributions to tness of two sexual traits, followed by their rapid evolution
in line with ecological changes.
KEY WORDS: Herpestes auropunctatus, invasive species, scent marking, sexual selection, small Indian mongoose, testis size,
Urva auropunctata.
Darwin’s theory of sexual selection has been successful in ex-
plaining sex differences and the function of traits involved
in mate competition and choice in a wide variety of species
(Darwin 1871; Andersson 1994). Most studies have focused on
the origins or maintenance of such traits, whereas less attention
has been paid to how these traits evolve once established (Kirk-
patrick and Ryan 1991; Hill 2015). As such, although evolution-
ary biology is increasingly documenting instances of wild pop-
ulations evolving by natural selection (Kingsolver et al. 2001;
Lahti 2005; Schilthuizen 2013), we have somewhat less under-
standing of how, and under what circumstances, sexually selected
traits are subject to evolutionary change (Wiens 2001; Miller and
Svensson 2014).
The strength, direction, and targets of sexual selection can
be influenced by a variety of ecological and demographic factors
(Evans and Garcia-Gonzalez 2016). For instance, changes in pre-
dation and the availability and distribution of resources may in-
fluence the evolution of sexually selected traits (Heinen-Kay et al.
2015; Evans and Gustafsson 2017). These factors might act indi-
rectly through population changes that in turn affect social inter-
action. For example, changes in resources can affect population
density; and encounter rates with prospective mates will be lower
in sparse populations and higher in dense populations. Long-
distance or temporally persistent forms of advertisement such as
sounds or chemicals might be expected to predominate in sparse
populations, where individuals usually communicate at some dis-
tance in time or space (Umbers et al. 2015). On the other hand,
physical contests and sperm competition would be expected to
predominate in denser populations where individuals routinely
encounter multiple potential mates and rivals (Rittschof 2010;
1
© 2020 The Authors. Evolution © 2020 The Society for the Study of Evolution.
Evolution
BRIEF COMMUNICATION
Figure 1. Rapid evolution of male but not female anal pad size in the small Indian mongoose (Urva auropunctata). Shown is a comparison
of relative anal pad area (measured in mm2, corrected for total body length in mm) of males and females across locations. Circles represent
the average (±SE) for each location. See Table S1 for reported population density estimates. Letter differences correspond to comparisons
where P<0.05. Mongooses were introduced from India to Jamaica in 1872, from Jamaica to Hawaii in 1883, and from Jamaica to St. Croix
in 1884 (Hoagland et al. 1989); and from India to Mauritius in 1902 (Roy et al. 2002).
Buzatto et al. 2015). Following from these expectations is a pre-
diction that sexually selected traits should change if population
density changes dramatically and consistently. Here we test this
prediction by comparing sexually selected traits of the small In-
dian mongoose (Urva auropunctata, formerly Herpestes aurop-
unctatus, hereafter mongoose) from several populations known
to vary in population density.
The mongoose is a promiscuous, generalist carnivore native
to South Asia, whose introduction history has created a natural
experiment. Since 1872, they have been introduced to more than
70, mostly tropical island, locations (Hoagland et al. 1989; Barun
et al. 2011) where, released from interspecific competition and
predation, they have become invasive. Density estimates from
the introduced range average between 51.4 and 65.6 times higher
than in the native range, depending on the method of location
grouping (range: 4.4–294; Table S1).
The mongoose is solitary and non-territorial but is a pro-
lific anal gland scent marker. In males, scent-marking functions
as a long-distance sexual signal (Owen and Lahti 2015). Anal
gland secretion is deposited onto the substrate with the aid of
an unusual fleshy projection surrounding the anus, the anal pad
(Fig. S1). This structure is under sexual selection in males: in
the native range, relative anal pad size (i.e., corrected for total
body length) is 2.38 times larger in males than females (Fig. 1;
Table S3) and, as is expected in most sexually selected traits
(Hamilton and Zuk 1982; Rahman et al. 2013), the anal pad is
both condition dependent and positively predicted by body size
in males but not females (Figs. 2 and 3; Owen and Lahti 2015).
The markedly higher population densities of mongooses in
their introduced range have likely affected the ways in which in-
dividuals communicate and how males maximize reproductive
success. Higher densities translate into increased encounter rates
relative to the native range, likely by orders of magnitude. This
situation would diminish the utility of long-distance advertising,
thus reducing the ecological relevance of scent marking, and pre-
sumably relaxing selection for large anal pads. Therefore, we pre-
dict that, relative to the native range, in the introduced range anal
pad size will be smaller in males but not different in females, and
that the anal pad’s relationships to body condition and size will
be significantly weaker in males with no change in females. Ad-
ditionally, an increase in mates and rivals likely intensifies post-
copulatory sexual selection via sperm competition. Thus, we also
predict that testis size will have increased following introduction.
We test these predictions by comparing wild mongooses in their
native range of India to those found on four islands of introduc-
tion: Hawaii, Jamaica, Mauritius, and St. Croix, all of which de-
scend from Indian populations.
Materials and Methods
TRAPPING SITES AND METHODS
Trapping of mongooses took place between 2012 and 2016. Mon-
gooses were trapped from an average of five populations per lo-
cation (range 2–7). Trapping protocols differed slightly by loca-
tion. In general, traps were placed opportunistically, rather than
in grids, to maximize capture rate. Traps were baited with sar-
dines, chicken wings, or raw fish, and were covered to minimize
heat stress. In India, traps were set at dawn, checked two to three
times per day, and closed at night to avoid capture of non-target
species; in all other locations, traps were set the day before and
2EVOLUTION 2020
BRIEF COMMUNICATION
Figure 2. Weakening and loss of condition dependence of the anal pad in male but not female small Indian mongooses (Urva aurop-
unctata) following species introduction (from India to four islands). Shown are scatter plots and associated trendlines of body condition
(residuals of an OLS means regression of total body length against mass) against anal pad area of males and females.
Figure 3. Weakening and loss of relationship between body size (PC1) and anal pad size in male but not female small Indian mongooses
(Urva auropunctata) following species introduction (from India to four islands). Shown are scatter plots and associated trendlines of body
size (PC1) against anal pad area of males and females.
checked the following morning, as trapping non-target species
was not a concern.
In India, mongooses were trapped in three areas in two
northern states: Chandrabani, a village located on the edge of the
city of Dehradun, the capital city of the state Uttarakhand; and
the rural outskirts of Jwalapur, Uttarakhand and Saharanpur, Ut-
tar Pradesh. These sites were chosen because they contain similar
habitat types to what have been anecdotally described as favored
by the mongoose (Prater 1971): areas of residential buildings
and small shops, partially completed and current construc-
tion projects, short-to-high vegetation, agricultural fields, and
forests.
In Hawaii, mongooses were collected for removal by the
United States Department of Agriculture Animal and Plant
Health Inspection Service (USDA-APHIS) within the rainforest
and surrounding grasslands of the Keaukaha Military Reservation
in Hilo, Hawaii (Owen and Lahti 2015).
In Jamaica, mongooses were trapped in three areas: on the
campus of the University of West Indies – Mona in Kingston
and surrounding neighborhoods, in a gated community on the
EVOLUTION 2020 3
BRIEF COMMUNICATION
north-central part of the island called Richmond Estates, and in
a rural village outside of Discovery Bay. Traps were set in a va-
riety of locations including forests, residential areas, shrubland,
and grassland.
Mongooses collected on St. Croix were trapped on the west
end of the island in the forest and grassland north of Frederiksted
and on the beach and thick forest and shrubs of Sandy Point Na-
tional Wildlife Refuge; and on the east side of the island on the
beaches and shrubs of Jack’s, Isaac’s, and East End Bays.
Mongooses collected on Mauritius were trapped in six lo-
cations that span much of the island and that can be catego-
rized as either semi-urban or forest. The semi-urban areas include
the neighborhoods of Beau Bassin, Quatre Bornes, and Phoenix,
where traps were placed around homes and in farmland. The
forested areas include Black River Gorges National Park, Bras
D’Eau National Park, and Casela Safari Adventures nature park.
SAMPLE SIZE
A total of 283 individuals were trapped across all locations. Ju-
veniles and subadults were excluded, leaving 232 individuals for
analyses: India: 23 males, 19 females; Hawaii: 34 males, 23 fe-
males; Jamaica: 30 males, 22 females; Mauritius: 16 males, 22
females; and St. Croix: 47 males, 28 females. Age was estimated
from eruption and wear of canines (adult canines are thicker and
longer than those of subadults) and development of genitalia.
MEASUREMENTS
In India, mongooses were anesthetized with a 2.5 mg/kg Ke-
tamine and 0.06 mg/kg Meditomedine injection administered
intra-muscularly in the thigh with a 1 mL syringe (Kreeger 1996).
After measurements were recorded (see below), all individuals
were placed back into their traps until fully recovered and re-
leased at the point of capture. In Hawaii, mongooses were trapped
for removal by the United States Department of Agriculture An-
imal and Plant Health Inspection Service (USDA-APHIS) and
were sacrificed using a custom-built CO2chamber (Owen and
Lahti 2015). In all other locations, mongooses were sacrificed in
the field through inhalation of 5 mL of isoflurane via a nose cone
(Parker et al. 2008).
Once immobilized or sacrificed, the following were
recorded: mass; lengths of head (tip of nose to occipital bone),
body (occipital bone to base of tail), tail, and the left testis (its
longest axis); and left canine diameters. Additionally, the follow-
ing were recorded in all locations except Hawaii: the left hind
foot length, head width, and neck and chest circumferences. Mass
was measured in the field with a 1 kg Pesola scale with 10 g units
(India, St. Croix, and Mauritius) and in the laboratory using a
digital scale to 0.001 g (Hawaii and Jamaica). Head, body, tail,
and feet lengths were measured using a tape measure with 1 mm
units. Head widths, testis lengths, and canine diameters were con-
sidered the mean of three measurements by digital calipers to
0.01 mm. Testis lengths were considered an accurate proxy for
testis size, as St. Croix males’ testis lengths and masses were
highly correlated (F1,92 =195.5, R2=0.68, P<0.001). Anal pad
area was estimated by photographing the anal pad with a ruler in-
set in the photograph. After isolating the hairless area of skin in
each photograph, anal pad area was estimated to 0.01 mm2using
the analysis feature of Photoshop CS5 (Adobe Systems Inc., San
Jose, CA, USA) (Owen and Lahti 2015).
ANALYSES
Statistical analyses were performed using SYSTAT version 10
(SPSS Inc., Chicago, IL, USA) and R 3.3.1. Two-tailed t-tests
were used to determine morphological differences between the
sexes within each location, and paired t-tests were used to de-
termine within-sex differences in canine diameters and foot and
testis lengths. ANOVAs were used to determine morpholog-
ical differences across locations. Condition was estimated as
the residuals of an OLS regression of total body length and
mass (Schulte-Hostedde et al. 2005; Edelman 2011; White et al.
2012). Condition dependence of the anal pad was estimated
by regressing these residuals against anal pad area (Owen and
Lahti 2015).
A principal components analysis was performed to reduce
the number of correlated variables. Because several morpholog-
ical traits were not collected in Hawaii, only mass, head, body,
and tail lengths, and the top and bottom left canine diameters
were included in the PCA. Additionally, because a large num-
ber of mongooses were missing one or both of their left canines,
the sample available for analyses with the PCs was reduced. PC1
loaded highly across all traits and explained 61.1% of the vari-
ance, and thus we treated it as an overall estimate of body size
(Table S4). PC2 explained 17.2% of the variance and was not
used in any analysis.
ANCOVAs were used to determine the influence of location
on the relationships of anal pad size and condition and body size
using location as a covariate. Degrees of sexual size dimorphism
were calculated as the mean male trait value divided by the mean
female trait value.
ETHICAL NOTE
The mongoose was named by the IUCN as one of the 100 worst
invasive species on the planet for its role in the declines, extirpa-
tions, and extinctions of dozens of species (Lowe et al. 2000). As
such, individuals trapped in areas of introduction were sacrificed
as a means of population control. For data collected in Hawaii,
mongooses were captured and euthanized by the USDA-APHIS
for non-research purposes and their disposal was approved by
USDA-APHIS WS SOP #AC 005.00. Methods carried out in
4EVOLUTION 2020
BRIEF COMMUNICATION
Figure 4. Rapid increase in testis size following introduction in the small Indian mongoose (Urva auropunctata). Shown is a comparison
of relative testis size (measured in mm, corrected for total body length in mm) of males across locations. Circles represent the average
(±S.E.) for each location. See Table S1 for reported population density estimates. Letter differences correspond to comparisons where P
<0.05. These data were not collected in Hawaii.
all other locations were approved by Queens College IACUC
protocol #159.
Results
Male relative anal pad size was significantly smaller in areas of
introduction than in the native range (overall model: F4,144 =
26.31, P<0.001; Tukey post hoc comparisons between Indian
males and males from all introduced locations: all P<0.001;
Fig. 1); no such difference was observed in females (overall
model: F4,106 =2.44, P=0.051; Tukey post hoc comparisons
between Indian females and females from introduced populations
were all nonsignificant: Mauritius: P=0.900; St. Croix: P=
0.755; Hawaii: P=0.900; and Jamaica: P=0.105; Fig 1). Rel-
ative anal pad area differed between female mongooses found on
two introduced islands, Jamaica and Mauritius (Tukey post hoc
comparison: P=0.045).
As predicted from a relaxation of selection, in areas of intro-
duction the condition dependence of male anal pads was signifi-
cantly weaker than in the native range (F4,135 =2.63, P=0.037;
Fig. 2). The relationship between male anal pad size and body
size (PC1) was also significantly weaker in introduced locations
(F4,121 =3.558, P=0.009; Fig. 3). Consistent with the anal pad
being a sexually selected trait only in males, female anal pads
were neither condition dependent nor showed any relationship
with body size in any location and did not differ across locations
(condition: F4,101 =0.274, P=0.894; body size: F4,91 =0.818,
P=0.517; Figs. 2, 3).
As predicted from an intensification of sperm competi-
tion, relative testis size increased following introduction (overall
model: F4,144 =26.31, P<0.001; Tukey post hoc comparisons
between Indian males and males from introduced locations: all
P<0.001; Fig. 4).
The changes observed in males correspond to the time since
mongooses have been introduced to the islands. Mongooses were
introduced from India to Jamaica in 1872, and from this Jamaican
population were introduced to both St. Croix and Hawaii in 1882
and 1883, respectively (Hoagland et al. 1989; Barun et al. 2011).
Additionally, a separate, independent introduction of mongooses
from India to Mauritius occurred in 1902 (Roy et al. 2002). For
Jamaican mongooses, which have experienced the longest time
since introduction, male relative anal pad area was the smallest
of all locations, measuring 66% of the size of their Indian coun-
terparts. For mongooses from Mauritius, which have experienced
the shortest time since introduction, relative male anal pad area
was the largest of all of introduced locations and intermediate in
size between the native range of India and the first area of intro-
duction of Jamaica. Male relative anal pad area for mongooses in
Hawaii and St. Croix were of intermediate sizes to those found on
Mauritius and Jamaica. The strength of condition-dependence of
the anal pad in males follows a similar pattern with respect to time
since introduction: the strongest relationship is found in the native
range, but among introduced locations, the strongest relationship
is observed in Mauritian males and the weakest relationship is
found in mongooses from Jamaica where the anal pad showed
no relationship with condition. Males from Hawaii and St. Croix
again exhibited an intermediate relationship between condition
and anal pad area between those of Mauritius and Jamaica. The
relationship of body size with anal pad area shows a similar trend
in most respects. The strength of this relationship has weakened
significantly with time since introduction relative to the native
range, with this relationship being lost in Jamaican males.
For relative testis size, whereas mongooses from all areas of
introduction had larger values compared to the native range,
the changes are not associated with time since introduction;
EVOLUTION 2020 5
BRIEF COMMUNICATION
in fact, trait values do not significantly differ across the study
locations.
The relationship between the observed changes in males and
population density of each location shows a similar pattern to
that of time since introduction, albeit with less precision. In gen-
eral, reported population density estimates for areas of intro-
duction are several times higher compared to the native range
(Table S1). The median population density estimate for Mauri-
tius is 4.4 times higher than for India and for Jamaica is 31.0
times higher. However, the median population density estimates
for St. Croix and Hawaii are the highest at 58.3 and 46.7 times
higher than India, respectively. Further, the estimates reported in
Table S1 span more than 50 years of research and utilize differ-
ent techniques for density estimation (with varying levels of so-
phistication), which is reflected in the wide variation in reported
estimates within, and presumably between, locations. Addition-
ally, habitat type, proximity to resources, and proximity to human
habituation are known to influence density estimates (Quinn and
Whisson 2005; Mahmood et al. 2011). Locations also differ in the
number of studies estimating density, with Jamaica and Mauritius
each having only a single estimate.
Discussion
Collectively our results indicate the repeated evolution of two
sexually selected traits in opposite directions, specifically as pre-
dicted from changes in social interaction. To our knowledge,
this is the first documented case of rapid evolution of sexu-
ally selected traits in a wild mammal, artificial selection aside
(Tiilikainen et al. 2010; Pigeon et al. 2016). Our results suggest
the following sequence of events: (I) mongooses introduced to is-
lands were released from interspecific competition and predation,
resulting in a dramatic increase in population density relative to
thenativerange.(II) These high densities increased encounter
rates, which diminished the ecological relevance of long-distance
chemical advertising, as mating decisions could be based instead
on direct assessments in real time. (III) This relaxed selection on
scent marking favored lower investment in the anal pad, leading
to a reduction in its size, condition dependence, and relationship
with body size in males. Concurrently, (IV) high population den-
sities increased sperm competition, as the number of mating part-
ners likely increased in this promiscuous species, favoring males
with larger quantities of sperm. This, in turn, led to (V)anevolu-
tionary increase in testis size.
The utility of long-distance chemical advertising was pre-
dicted to decrease in males following introduction, and while
scent marking frequency was not quantified directly, the tool used
for the application of scent, the anal pad, shrunk in size. A mor-
phological feature generally decreases in size and functional in-
tegrity when it loses utility, although the rate at which this evolu-
tionary decay occurs is variable and depends on the costs of con-
tinuing to construct and maintain the trait (Lahti et al. 2009). The
costs of the anal pad are unknown, but theory predicts that sexual
selection should drive male display traits away from an optimum
value with respect to all other sources of selection, whereas cor-
responding female trait values should be at or near the optimum
(Bonduriansky 2007). Sexual size dimorphism (SSD) of the anal
pad decreased between the native and introduced ranges between
63 to 88 percentage points for raw values (Table S2) and between
54 and 70 percentage points for relative values (Table S3), de-
pending on population. This marked reduction in SSD, specifi-
cally owing to a decrease in male anal pad size rather than an
increase in females, is consistent with an evolutionary change in
males towards the selective optimum following relaxed sexual se-
lection on scent marking. Further evidence of a reduction of anal
pad function is lent by its partial or complete decoupling from
male condition and body size. Sexual selection in general does
not appear to be relaxed, however, but changed: the consistent in-
crease in testis size in all introduced populations indicates a shift
towards a more influential role for sperm competition in male
reproductive success.
Several aspects of the results indicate evolution by sex-
specific selection on males as the primary mechanism for the
observed trait changes. First, changes to the anal pad in size,
condition-dependence, and its relationship to body size were ob-
served only in males and not in females (Figs. 1–3). Second, we
collected several additional morphological measurements in or-
der to determine whether any other traits, besides the two we pre-
dicted, changed consistently following introduction. Of all traits
measured, the only other strong and uniformly directional change
observed following introduction was in hind foot length, but this
was observed in both sexes (Fig. S2). Third, comparisons of vari-
ance across locations of relative anal pad and testis size follow
predictions of relaxed and increased selection, respectively. Fol-
lowing a relaxation of selection, trait variance might increase,
decrease, or be unaffected (Lahti et al. 2009), but after an in-
crease in selection variance is expected to increase (Fisher 1930).
Coefficients of variation (CV) of introduced males’ relative anal
pad size are unchanged compared to males of the native range
(Fig. S3), while CVs of introduced males’ relative testis size are
generally greater relative to native range males, with higher CVs
in locations with earlier introductions (Fig. S4).
Fourth, the results allow us to distinguish evolutionary
change from the founder effect. If randomly biased character-
istics of the introduced individuals were mainly responsible for
the observed changes, we would not expect the same directional
changes across introduced populations, especially considering
that the Mauritian population was introduced from India indepen-
dently from the Jamaican population. The founder effect would
also predict a greater difference in the population most recently
6EVOLUTION 2020
BRIEF COMMUNICATION
introduced, with the smallest propagule size, and with subsequent
introductions from island to island. None of these predictions
was validated; on the contrary, changes were positively correlated
with time since introduction, at least for the anal pad, consistent
with evolution by natural selection, and were unrelated to propag-
ule size and introduction sequence. Furthermore, while the intro-
duction histories and propagule sizes of mongooses to different
areas around the world are generally well-documented (Hoagland
et al. 1989; Barun et al. 2011), Barun et al. (2013) conclude that
for many areas (including Jamaica and Mauritius) the historical
record does not always align with the expected allelic variation
from reported propagule sizes. Their findings suggest that, at
least for some islands, the original propagule size may have been
underestimated or that additional introductions may have taken
place. If either of these scenarios is true for populations of mon-
gooses investigated in this study, the additional genetic variation
introduced to the respective island populations lends further evi-
dence that our results are not consistent with the founder effect.
A final point is that an increase in population density might
have resulted in an increase in direct agonistic competition for
mates, but our morphological measurements do not indicate that
this has occurred: SSD in total body length and relative top canine
diameter, two features that would enhance male physical com-
petitiveness, either decreased, showed no change, or increased
by only one to two percentage points (Tables S2 and S3). Evi-
dence for rapid evolution in response to sperm competition but
not male agonistic competition is to be expected, as male mon-
gooses have not been observed to defend territories or females
(Roy et al. 2002; Owen and Lahti 2015).
While our results are consistent with a shift in selection pres-
sures on two traits as predicted from changes in encounter rates,
our study has several limitations. First, we did not measure en-
counter rates of mongooses in any location. However, with den-
sity estimates between 4 and 58 times higher in the introduced
locations relative to the native range, individuals in introduced lo-
cations are almost certainly encountering conspecifics at a much
higher rate than in the native range. Second, we do not estimate
reproductive success for males of different relative anal pad ar-
eas. However, the characteristics of condition-dependence, a pos-
itive relationship to body size only in males, and strong sexual
size dimorphism, together suggest this trait is indeed important
for males in maximizing reproductive success (Owen and Lahti
2015). Finally, we did not compare scent-marking frequency be-
tween the native and introduced ranges to test for its decline.
However, a change in male trait values towards those of females
is an outcome predicted by sexual selection theory when selec-
tion on the male trait has relaxed (Bonduriansky 2007); the re-
duction in male trait values we observed, leading to a reduc-
tion in sexual size dimorphism of the anal pad, accords with this
prediction.
Density estimates in areas of introduction are not reliable
enough to make robust inferences comparing evolution in the dif-
ferent islands. Most critical to our hypotheses is rather their gen-
eral comparison to the density of the native population in India.
Nevertheless, our results do suggest that relative anal pad area
tends to be smaller in relatively denser locations, where its rela-
tionship to both condition and body size is weaker or nonexis-
tent compared to the least dense locations; and relative testis size
tends to be larger in denser locations.
Research has not commonly focused on evolutionary change
in established sexually selected traits. Our results demonstrate
that such traits can change dramatically over even modest his-
torical time scales (<150 years). Moreover, the nature of these
changes highlights the importance of ecological influences on
sexually selected traits. Because sexual selection operates be-
tween members of a single species, we might be tempted to con-
sider the evolution of such traits as being subject only to social
influences, namely mate choice and competition. Like any trait
under natural selection, however, the utility or adaptive value of a
sexually selected trait is influenced by the environment in which
it is expressed (Endler 1988; Candolin and Vlieger 2013). Thus,
a change in ecological factors that have little, if anything, to do
with mating, such as resource availability and population density,
can lead to evolutionary changes in sexually selected traits (Evans
and Garcia-Gonzalez 2016). This being the case, we expect that
with the intensification of climate change and its corresponding
ecological changes, more instances of rapid evolution of sexually
selected traits will be observed in the future.
AUTHOR CONTRIBUTIONS
M.A.O. and D.C.L. designed the study. M.A.O. collected and analyzed
the data. M.A.O. and D.C.L. prepared the manuscript.
DATA ARCHIVING
Data can be found on Dryad: https://datadryad.org/stash/dataset/https://
doi.org/10.5061/dryad.rxwdbrv4j
CONFLICT OF INTEREST
The authors declare no conflict of interest.
ACKNOWLEDGEMENTS
This work was supported by grants to M.A.O. from the Fulbright Foun-
dation; Sigma Xi; the American Society of Mammalogists; The Graduate
Center, City University of New York; and Queens College, City Univer-
sity of New York. This work was also supported by grants to D.C.L. from
Queens College, City University of New York (startup funds and Re-
search Enhancement Funds), and PSC-CUNY grant #67646-0045. We
thank W. Pitt, Y.V. Jhala, H. Guleria, B. Wilson, R. Robinson, Camilo
Trench, and Denise Henry, B. Hoagland, and V. Tatayah for logistical
and technical support in the field. We also thank A. Richards, D. Ruben-
stein, J. Munshi-South, B. Dumont, and M. Baker for their discussion of
ideas and comments.
EVOLUTION 2020 7
BRIEF COMMUNICATION
LITERATURE CITED
Andersson, M. 1994. Sexual selection. Princeton Univ. Press, Princeton, New
Jersey.
Barun, A., C. C. Hanson, K. J. Campbell, and D. Simberloff. 2011. A review
of small Indian mongoose management and eradications on islands. Pp.
17–25 in C. R. Veitch, M. N. Clout, and D. R. Towns, eds. Island inva-
sives: eradication and management. IUCN, Gland, Switzerland.
Barun, A., M. L. Niemiller, B. M. Fitzpatrick, J. A. Fordyce, and D.
Simberloff. 2013. Can genetic data confirm or refute historical records?
The island invasion of the small Indian mongoose (Herpestes auropunc-
tatus). Biol. Invasions 15:2243–2251.
Bonduriansky, R. 2007. The evolution of condition-dependent sexual dimor-
phism. Am. Nat. 169:9–19.
Buzatto, B. A., J. D. Roberts, and L. W. Simmons. 2015. Sperm competition
and the evolution of precopulatory weapons: Increasing male density
promotes sperm competition and reduces selection on arm strength in a
chorusing frog. Evolution 69:2613–2624.
Candolin, U., and L. Vlieger. 2013. Estimating the dynamics of sexual selec-
tion in changing environments. Evol. Biol. 40:589–600.
Darwin, C. 1871. The descent of man, and selection in relation to sex. John
Murray, London.
Edelman, A. J. 2011. Sex-specific effects of size and condition on timing of
natal dispersal in kangaroo rats. Behav. Ecol. 22:776–783.
Endler, J. A. 1988. Sexual selection and predation risk in guppies. Nature
332:593–594.
Evans, J. P., and F. Garcia-Gonzalez. 2016. The total opportunity for sexual
selection and the integration of pre- and post-mating episodes of sexual
selection in a complex world. J. Evol. Biol. 29:2338–2361.
Evans, S. R., and L. Gustafsson. 2017. Climate change upends selection on
ornamentation in a wild bird. Nat. Ecol. Evol. 1:0039
Fisher, R. A. 1930. The genetical theory of natural selection. Carendon Press,
Oxford.
Hamilton, W. D., and M. Zuk. 1982. Heritable true fitness and bright birds: A
role for parasites. Science 218:384–387.
Heinen-Kay, J. L., K. E. Morris, N. A. Ryan, S. L. Byerley, R. E. Venezia,
M. N. Petersen, and R. B. Langerhans. 2015. A trade-off between nat-
ural and sexual selection underlies diversification of a sexual signal.
Behavioral Ecology 26:533–542.
Hill, G. E. 2015. Sexiness, individual condition, and species identity: The
information signaled by ornaments and assessed by choosing females.
Evolutionary Biology 42:251–259.
Hoagland, D. B., G. R. Horst, and C. W. Kilpatrick. 1989. Biogeography and
population biology of the mongoose in the West Indies. Pp. 628–645
in C. A. Woods, ed. Biogeography of the West Indies. Sandhill Crane
Press, Gainseville, FL.
Kingsolver, J. G., H. E. Hoekstra, J. M. Hoekstra, D. Berrigan, S. N. Vignieri,
C. E. Hill, A. Hoang, P. Gibert, and P. Beerli. 2001. The strength of phe-
notypic selection in natural populations. American Naturalist 157:245–
261.
Kirkpatrick, M., and M. J. Ryan. 1991. The evolution of mating preferences
and the paradox of the lek. Nature 350:33–38.
Kreeger, T. K. 1996. Handbook of wildlife chemical immobilization. Interna-
tional Wildlife Veterinary Services Inc., Larammie, WY.
Lahti, D. C. 2005. Evolution of bird eggs in the absence of cuckoo parasitism.
Proc. Natl. Acad. Sci. USA 102:18057–18062.
Lahti, D. C., N. A. Johnson, B. C. Ajie, S. P. Otto, A. P. Hendry, D. T.
Blumstein, R. G. Coss, K. Donohue, and S. A. Foster. 2009. Relaxed
selection in the wild. Trends in Ecology & Evolution 24:487–496.
Lowe, S., M. Browen, S. Boudjelas, and M. De Poorter. 2000. 100 of the
World’s Worst Invasive Alien Species - A selection from the Global In-
vasive Species Database. Published by The Invasive Species Specialist
Group (ISSG) a specialist group of the Species Survival Commission
(SSC) of the World Conservation Union (IUCN), 12pp. First published
as special lift-out in Aliens 12, December 2000.
Mahmood, T., I. Hussain, and M. S. Nadeem. 2011. Population estimates,
habitat preference and the diet of small Indian mongoose (Herpestes
javanicus) in Potohar Plateau, Pakistan. Pakistan Journal of Zoology
43:103–111.
Miller, C. W., and E. I. Svensson. 2014. Sexual Selection in Complex Envi-
ronments. Annu. Rev. Entomol 59:427–445.
Owen, M. A., and D. C. Lahti. 2015. Sexual dimorphism and condition de-
pendence in the anal pad of the small Indian mongoose (Herpestes au-
ropunctatus). Can. J. Zool. 93:397–402.
Parker, W. T., L. I. Muller, R. R. Gerhardt, D. P. O’Rourke, and E. C. Ramsay.
2008. Field use of isoflurane for safe squirrel and woodrat anesthesia. J.
Wildlife Manag. 72:1262–1266.
Pigeon, G., M. Festa-Bianchet, D. W. Coltman, and F. Pelletier. 2016. Intense
selective hunting leads to artificial evolution in horn size. Evol. Appl.
9:521–530.
Prater, S. H. 1971. The book of Indian animals. Oxford Univ. Press, New
Delhi, India.
Quinn, J. H., and D. A. Whisson. 2005. The effects of anthropogenic food on
the spatial behaviour of small Indian mongooses (Herpestes javanicus)
in a subtropical rainforest. J. Zool. 267:339–350.
Rahman, M. M., J. L. Kelley, and J. P. Evans. 2013. Condition-dependent ex-
pression of pre- and postcopulatory sexual traits in guppies. Ecol. Evol.
3:2197–2213.
Rittschof, C. C. 2010. Male density affects large-male advantage in the golden
silk spider, Nephila clavipes. Behav. Ecol. 21:979–985.
Roy, S. S., C. G. Jones, and S. Harris. 2002. An ecological basis for control of
the mongoose Herpestes javanicus in Mauritius: is eradication possible?
Pp. 266–273 in C. R. Veitch, and M. N. Clout, eds. Turning the tide: the
eradication of invasive species. IUCN SSC Invasive Species Specialist
Group, Gland, Switzerland and Cambridge, UK.
Schilthuizen, M. 2013. Rapid, habitat-related evolution of land snail colour
morphs on reclaimed land. Heredity 110:247–252.
Schulte-Hostedde, A. I., J. S. Millar, and G. J. Hickling. 2005. Condition
dependence of testis size in small mammals. Evol. Ecol. Res. 7:143–
149.
Tiilikainen, R., T. Nygren, J. Pusenius, and V. Ruusila. 2010. Variation in
growth pattern of male moose Alces alces after two contrasted periods
of hunting. Ann. Zool. Fenn. 47:159–172.
Umbers, K. D. L., M. R. E. Symonds, and H. Kokko. 2015. The mothematics
of female pheromone signaling: strategies for aging virgins. Am. Nat.
185:417–432.
White, A. J., R. G. Poulin, B. Wissel, J. L. Doucette, and C. M. Somers.
2012. Agricultural land use alters trophic status and population density
of deer mice (Peromyscus maniculatus) on the North American Great
Plains. Can. J. Zool. 90:868–874.
Wiens, J. J. 2001. Widespread loss of sexually selected traits: how the peacock
lost its spots. Trends Ecol. Evol. 16:517–523.
Associate Editor: M. Maan
Handling Editor: D. Hall
8EVOLUTION 2020
BRIEF COMMUNICATION
Supporting Information
Additional supporting information may be found online in the Supporting Information section at the end of the article.
Fig. S1. Typical anal pads of female (a) and male (b) small Indian mongooses (Urva 4auropunctata), demonstrating sexual dimorphism.
Fig. S2. Lack of consistent change in sexual dimorphism in nonfocal morphological traits after introduction in the small Indian 21 mongoose (Urva
auropunctata).
Fig. S3. Lack of change of variance in relative anal pad area following introduction in male small 40 Indian mongooses (Urva auropunctata).
Fig. S4. Increase in variance in relative testis size following introduction in the small Indian 73 mongoose (Urva auropunctata).
Tab l e S 1 . Reported population densities of the small Indian mongoose (Urva auropunctata).
Tab l e S 2 . Degrees of sexual size dimorphism (SSD) of morphological traits, using raw trait values, in five populations of the small 97 Indian mongoose
(Urva auropunctata); and the difference in SSD between the native and introduced ranges.
Tab l e S 3 . Degrees of sexual size dimorphism (SSD) of morphological traits, using relative (i.e., body-length corrected) trait values, in 109 five populations
of the small Indian mongoose (Urva auropunctata); and the difference in SSD between the native and introduced 110 ranges.
Tab l e S 4 . Eigenvalues and eigenvectors of the two principal components (PCs) that cumulatively accounted for 78% of the variation 124 in the measured
morphological traits of the small Indian mongoose (Urva auropunctata).
EVOLUTION 2020 9
1
Supplementary Information
1
2
3
Fig. S1. Typical anal pads of female (a) and male (b) small Indian mongooses (Urva
4
auropunctata), demonstrating sexual dimorphism. A: anus, AP: anal pad (photo MAO, 2012,
5
Hawaii).
6
7
8
9
10
11
12
13
14
15
16
17
18
2
19
20
3
Fig. S2. Lack of consistent change in sexual dimorphism in nonfocal morphological traits after introduction in the small Indian
21
mongoose (Urva auropunctata). Shown are comparisons of multiple morphological traits of both sexes across locations. Circles
22
represent male (blue) and female (red) means (± S.E.) for each location. Letter differences correspond to male (blue) and female (red)
23
comparisons where P < 0.05. Relative values are those corrected for total body length. Data for traits in panels B, C, H, and I were not
24
collected in Hawaii.
25
26
27
4
28
29
30
31
32
33
34
35
36
37
38
39
Fig. S3. Lack of change of variance in relative anal pad area following introduction in male small
40
Indian mongooses (Urva auropunctata). Shown is a comparison of coefficients of variation
41
(circles ± S.E.) across locations.
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
61
62
63
64
65
66
67
68
69
70
71
72
Fig. S4. Increase in variance in relative testis size following introduction in the small Indian
73
mongoose (Urva auropunctata). Shown is a comparison of coefficients of variation (circles ±
74
S.E.) across locations. These data were not collected in Hawaii.
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
6
Table S1. Reported population densities of the small Indian mongoose (Urva auropunctata).
93
94
Per Population
Per Land Mass
Mean
density
(per/ha)
Variation about the
mean (per/ha)
Location
Mean
density
(per/ha)
Location
Native
Range
0.083
Pakistan1
0.084
Pakistan
0.085
Introduced
Range
10.19
3.076 (S.E.)
Antigua2
9.45
Antigua
5.88
1.86 (S.E.)
9.71
2.94 (S.E.)
12.47
3.72 (S.E.)
8.99
2.74 (S.E.)
0.7
Fiji3
0.7
Fiji
10.4
8.2-12.6 (S.E.)
Grenada4
6.35
Grenada
3.2
2.5-4 (S.E.)
5.9
4.9-7.2 (S.E.)
4.0
2.7-5.2 (S.E.)
4.7
4-5.4 (S.E.)
9.9
7.7 -12.1 (S.E.)
24.7
Hawaii5
9.78
Hawaii
0.72
0.65-1.94 (C.I.)
Hawaii6
3.92
1.88-31.4 (C.I.)
2.6
1-7 (R)
Jamaica7
2.6
Jamaica
0.37
0.26-0.52 (R)
Mauritius8
0.37
Mauritius
2.5
Puerto Rico9
2.30
Puerto Rico
1.83
Puerto Rico10
4.6
2-8 (R)
Puerto Rico11
0.57
0.21 (F); 0.36 (M)
Puerto Rico12
2.02
1.00-6.14 (C.I.)
Puerto Rico13
6.4
St. Croix4
4.85
St. Croix
3.2
3.4
6.4
2-14 (R)
St. Croix7
1.0
0.7-1.5 (S.E.)
Trinidad4
2.5
Trinidad
4.0
2.2-5.4 (S.E.)
5.74
4.55
Total Mean
S.E. = standard error; C.I. = 95% confidence interval; R = range; F = female; M = male
Superscripts correspond to reference number: 1Mahmood et al., Pakistan J. Zool (43:103-
111), 2Corn & Conroy, J. Mammal. (79:1009-1015), 3Gorman, J. Zool. (187:65-73), 4Nellis
& Everard, Stud. Fauna Curacao Caribb. Is. (64:1-162), 5Seaman, Trans. N. Am. Wildl.
7
Conf. (17:188-197), 6Pitt et al., Biol. Invas. (17:1743-1759), 7Roy et al., Turning the Tide:
The Eradication of Invasive Species. IUCN SSC Invasive Species Specialist Group, Gland,
Switzerland and Cambridge, UK. (266-273), 8Hoagland et al. Biogeography of the West
Indies (1989:628-645), 9Pimentel, J. Mammal. (36:62-68), 10Vilella, Biotropica (30:120-
125), 11Horst et al. Biogeography of the West Indies: Patterns and Perspectives (2001:409-
424), 12Quinn & Whisson J. Zool. (267:339-350), 13Johnson et al. J. Wildlife Manage
(80:37-47).
95
96
8
Table S2. Degrees of sexual size dimorphism (SSD) of morphological traits, using raw trait values, in five populations of the small
97
Indian mongoose (Urva auropunctata); and the difference in SSD between the native and introduced ranges. SSD is calculated as the
98
average male trait value divided by the average female trait value. Traits are ranked according to the magnitude of SSD.
99
100
Native
Introduced
Difference
India
Mauritius
St. Croix
Hawaii
Jamaica
Min
Max
Direction1
Significant2
Anal Pad
2.71
2.08
2.02
1.83
2.04
0.63
0.88
-
Y
Mass
1.45
1.48
1.46
1.44
1.56
0.01
0.11
-/+
N
B Canine
1.17
1.14
1.09
1.09
1.14
0.03
0.08
-
N
Chest
1.18
1.18
1.17
1.25
0.00
0.07
-/+/=
N
Tail
1.14
1.11
1.12
1.07
1.11
0.02
0.07
-
N
T Canine
1.16
1.17
1.09
1.11
1.13
0.01
0.07
-/+
N
Body Length
1.14
1.12
1.12
1.08
1.11
0.02
0.06
-
N
Head Width
1.15
1.09
1.13
1.15
0.00
0.06
-/=
N
Head
1.07
1.11
1.09
1.08
1.10
0.01
0.04
+
N
Neck
1.18
1.14
1.16
1.18
0.00
0.04
-/=
N
Hind Foot
1.14
1.12
1.11
1.12
0.02
0.03
-
N
Body
1.13
1.13
1.12
1.11
1.11
0.00
0.02
-/=
N
B = bottom; T = top; Y = yes; N = no
101
1Direction of difference between the native range and different introduced range locations
102
2Determined by the absence of overlap of 95% confidence intervals for the difference of means between the sexes between the native
103
and different introduced range locations.
104
105
106
107
108
9
Table S3. Degrees of sexual size dimorphism (SSD) of morphological traits, using relative (i.e., body-length corrected) trait values, in
109
five populations of the small Indian mongoose (Urva auropunctata); and the difference in SSD between the native and introduced
110
ranges. SSD is calculated as the average male trait value divided by the average female trait value. Traits are ranked according to the
111
magnitude of SSD.
112
113
Native
Introduced
Difference
India
Mauritius
St. Croix
Hawaii
Jamaica
Min
Max
Direction1
Significant2
Anal Pad
2.38
1.84
1.77
1.68
1.84
0.54
0.70
-
Y
Mass
1.28
1.32
1.30
1.32
1.41
0.02
0.13
+
N
Chest
1.04
1.05
1.05
1.12
0.01
0.08
+
N
Head Width
1.02
0.96
1.02
1.04
0.00
0.06
-/=/+
N
B Canine
1.05
1.00
0.99
1.01
1.03
0.02
0.06
-
N
Head
0.94
0.99
0.99
0.99
0.99
0.05
0.05
+
N
Neck
1.04
1.01
1.04
1.06
0.00
0.03
-/=/+
N
Tail
1.00
0.99
0.99
0.98
1.00
0.00
0.02
-/=
N
T Canine
1.02
1.04
0.99
1.03
1.02
0.00
0.03
-/=/+
N
Body
1.00
1.01
1.01
1.01
1.00
0.00
0.01
=/+
N
Hind Foot
1.00
0.99
1.00
1.01
0.00
0.01
-/=/+
N
B = bottom; T = top; Y = yes; N = no
114
1Direction of difference between the native range and different introduced range locations
115
2Determined by the absence of overlap of 95% confidence intervals for the difference of means between the sexes between the native
116
and different introduced range locations.
117
118
119
120
121
122
123
10
Table S4. Eigenvalues and eigenvectors of the two principal components (PCs) that cumulatively accounted for 78% of the variation
124
in the measured morphological traits of the small Indian mongoose (Urva auropunctata).
125
Eigenvector of morphological trait
Principal
Component
Eigenvalue
Variation
Explained (%)
Body
Mass
Head
Tail
Bot.
Canine
Top
Canine
PC1
3.66
61.1
0.87
0.91
0.81
0.73
0.40
0.85
PC2
1.03
17.2
0.18
0.13
0.19
0.27
-0.88
-0.31
126
11
END
