O’SHEA, CRYAN & BOGAN: UNITED STATES BAT SPECIES OF CONCERN 
211 
of recruitment through reproduction can be more variable, are not well studied, and can be posi¬ 
tively biased by studying reproduction in female bats only at maternity colonies. First among these 
is natality expressed as the proportion of females that reproduce each year, and second is age at first 
reproduction. In many species of bats, non-reproductive females do not require the same roosting 
conditions that are necessary for females raising young, and as seen in some of the species of 
concern, non-reproductive females can function similar to males in using different foraging and 
roosting habitats than reproductive females. Thus not all females may be available for sampling at 
roosts or nearby habitats throughout their adulthood, and the usual assumption that intervals 
between births are one year may not always be valid. The biases in interpreting reproduction data 
from samples taken at maternity colonies was recognized nearly 50 years ago by Davis and 
Barbour (1970:261), who stated “The percentage of reproductive failure, which needs to be known 
to understand population dynamics, cannot be obtained from nursery colonies.” Similarly, although 
they had limited data to contrast between reproductive rates of females captured at maternity roosts 
and those captured away from roosts, Barclay et al. (2004:691) noted that “caution should be taken 
when assessing reproductive rates from colonies and extrapolating to the entire population”. Data 
acquired from long-term studies of tagged individuals have been successfully combined with 
multistate, robust-design models to provide more reliable estimates of breeding probabilities in 
other taxa, such as sea turtles (Kendall and Bjorkland, 2001) and manatees (Kendall et al., 2003). 
Employing these techniques offers a potential way to approach the problem, although we are aware 
of only two cases where these approaches have been applied to estimating breeding probabilities 
of U.S. bats, both common species that were studied where they roosted in colonies in buildings 
(Frick et al., 2010b; O’Shea et al., 2010). 
Estimates of the proportion of adult females that reproduce indicate that this is the most 
variable aspect of recruitment in the species of concern. In our literature review, we noted that stud¬ 
ies of several of these species have shown variability in these proportions due to direct and indi¬ 
rect climatic effects, including intense storms (the red fruit bat), drought (western small-footed 
myotis, long-eared myotis, long-legged myotis), and long periods of cool rainy weather in summer 
(Yuma myotis). Calculations of the proportion reproductive individuals based on samples taken in 
mist nets at places (primarily over water) other than at maternity colonies strongly indicate that not 
all females breed in any given year. Cumulative totals of such data indicated crude proportions of 
females reproductive for the following seven U.S. species of concern: 77% (68 of 89) for spotted 
bats, 95% (96 of 101) for Allen’s big-eared bats, 56% (96 of 172) for western small-footed myotis 
(also 56% of 351 females in Alberta, Canada), 68% (412 of 603) for long-eared myotis, 65% (100 
of 155) for Arizona myotis, 75% (282 of 375) for fringed myotis, and 42% (383 of 910 bats) for 
long-legged myotis. Data based on captures over water may also have biases because heavier, preg¬ 
nant bats may be less able to avoid nets, lactating bats require more water than non-reproductive 
bats, and in mountainous regions the elevation of a site may influence the proportion of reproduc¬ 
tive females. 
Female bats are often assumed to give birth at age one year. However, there are no data for this 
parameter for 13 of the 20 species; simple generalizations of birth at one year without supporting 
age data appear in the literature for two others (California leaf-nosed bat and Yuma myotis, but with 
one study noting only 42% of the latter breed as one-year-olds, based on unpublished data of Frick 
et al. [2007] from California). A generalization that not all first-year females breed appears for one 
species (long-legged myotis; Druecker, 1972) without supporting age data. Evidence from one-year 
old bats marked as juveniles is available only for four species of concern: a small proportion of 
one-year old females give birth in the Arizona myotis (one of 35) and Townsend’s big-eared bat 
(nine of 34), whereas all of 46 marked one-year-old southeastern myotis and all of 39 female cave 
myotis were reproductive. 
