154 
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES 
Series 4, Volume 65, Supplement I 
0.1-hectare plots of forest stands with roosts, characteristics that contributed most to the likelihood 
that a roost would be used included number of other snags greater than 30 centimeters in diameter, 
openness of canopy compared to randomly selected 0.1-hectare sites, and in some models dis¬ 
tances to stream channels (Weller and Zabel, 2001). In Yosemite National Park in the California 
Sierra Nevada Range, two maternity roosts were discovered in basal hollows of giant sequoia trees, 
two in snags of sugar pine, and two other roosts were discovered in snags of ponderosa pine (Pier¬ 
son et al., 2006). Roosts held one to 29 bats, and one female tracked for five days changed roosts 
daily (Pierson et al., 2006). 
Habitat characteristics of thirteen fringed myotis maternity colony sites in Colorado and one 
site in northern New Mexico were investigated by Hayes and Adams (2015), including roosts dis¬ 
covered by radio tracking seven females. No roosts were in trees or snags. Ten sites were on faces 
of rock outcroppings or cliffs, one was in a crevice in a boulder, two were in abandoned mines, and 
one was in an abandoned cabin. Model selection procedures including four landscape-level vari¬ 
ables were used to best describe habitats at nine maternity sites in comparison with randomly 
selected potential roost sites. The top three models were all competitive, and involved combina¬ 
tions of variables that measured grade (% slope), aspect, elevation, and distance to water, with a 
model involving just grade and aspect having the highest weight (Hayes and Adams, 2015). Mater¬ 
nity sites had steeper grades (mean 43% slope) and southerly aspects. Estimated maternity colony 
sizes in rock structures based on visible clusters ranged from four to 30 individuals, whereas 
colonies in each of two abandoned mines numbered about 100 bats. Based on radio tracking and 
other supportive observations, Hayes and Adams (2015) suggested that this species showed high 
daily and long-term (multiple years) fidelity to maternity roost sites in their study regions. In west¬ 
ern Colorado, Neubaum (2017) radio tracked two lactating females to five roosts, all located in 
crevices in high cliffs. 
In the Black Hills of South Dakota, 15 fringed myotis (13 reproductive females) tracked from 
one to 10 days used 36 roosts (27 in rock crevices, nine in tree snags), averaging 1.8 days per roost, 
with exit counts ranging one to 27 bats (Cryan et al., 2001). Only two bats roosted in trees (both 
lactating females), one of which also roosted in rock crevices; other tracked bats did not switch 
between trees and rock crevices (Cryan et al., 2001). Roosts in trees were in cavities or cracks of 
ponderosa pine snags, with none under exfoliating bark (Cryan et al., 2001). Roost trees were 
greater in diameter but did not differ in height compared to randomly selected snags; numbers of 
snags were greater in roost plots than in randomly selected plots and more roosts were on south¬ 
facing slopes than expected (Cryan et al., 2001). The rock crevices were in sandstone and limestone 
boulders and cliff faces with southern exposures. Females roosted in deeper crevices (greater than 
25 centimeters) than males. Most rock roosts were in rock ridges or canyons at the ponderosa pine 
and oak/juniper ecotone, where snags were also plentiful (Cryan et al., 2001). 
Fringed myotis also have been observed roosting in both rock outcroppings and in trees in pon¬ 
derosa pine forests in far northwestern Arizona (Herder and Jackson, 2000). However, Rabe et al. 
(1998a) found that 15 adult females followed by radio tracking during the reproductive season in 
northern Arizona roosted only in ponderosa pine snags, with one exception roosting in a Douglas 
fir snag. Fringed myotis females (n = 16) radio tracked in east-central Arizona ponderosa pine for¬ 
est roosted in 17 snags and two live trees of five species and one rock roost, with mean colony sizes 
of 63 bats observed in exit counts (range 26-86; Saunders, 2015). 
In the Jemez Mountains of New Mexico, seven adults (six lactating females) were tracked for 
three to 18 days and used 11 roosts: all roosts were in rock crevices or solution cavities located nine 
to 23 meters high on cliff walls, most facing southeast (Bogan et al., 1998). Colony sizes varied 
from four to 162, averaging 66 bats (Bogan et al., 1998). In pinon-juniper woodlands of the Galli- 
