122 
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES 
Series 4, Volume 65, Supplement I 
example, Vonhof and Barclay, 1996, 1997; Waldien et al., 2000; Amett and Hayes, 2009). In the 
western hemlock zone of the western Cascades of Oregon, Waldien et al. (2000) located 73 roosts 
of 21 radio-tagged reproductive females and determined characteristics of roost trees and the for¬ 
est stands in which they occurred. Bats were tracked to several types of structures and were found 
roosting in stumps, snags, trees, and logs in descending order of use. Roost switching was frequent, 
with occupancy averaging 1.2 days (range one to four). Most adult females were solitary, although 
groups of up to 14 were observed. Type of roost used did not vary by reproductive condition, and 
individuals switched among roosts of different structural types (Waldien et al., 2000). Characteris¬ 
tics of roosts in a mature (trees greater than 80 years old), largely unharvested watershed were com¬ 
pared with those in two younger, intensively harvested watersheds. Bats tended to roost in snags in 
older forests and in stumps in younger forests. Roost sites were not significantly closer to available 
water than random sites. Most (18 of 20) roosts in snags were in large-diameter Douglas firs, the 
dominant species of overstory tree in the region. Use of snags in intermediate stages of decay pre¬ 
dominated, and snags were more likely to be used if they were close to other snags in intermediate 
stages of decay (Waldien et al., 2000). (Snags in intermediate stages of decay provide greater 
opportunities to roost under exfoliating bark or in secondary cavities than those in earlier or more 
advanced stages.) Snags located farther from stand edges were less likely to be used as roosts. 
Snags with roosts did not protrude above the surrounding canopy, but their locations in gaps in the 
canopy and near edges probably offered similar benefits of increased exposure to solar radiation 
(Waldien et al., 2000). In the harvested watersheds, this species roosted exclusively in crevices in 
stumps (defined as less than three meters in height), primarily stumps of Douglas fir. Taller stumps 
were more likely to be used than shorter stumps, as were stumps that were more accessible (less 
woody debris or vegetation in the immediate surroundings). 
In a second study of long-eared myotis roost use in Douglas fir forests of western Washington 
and Oregon, Arnett (2007) and Amett and Hayes (2009) augmented the observations by Waldien 
et al. (2000), especially regarding use of snags. Twenty-seven individuals were radio tracked from 
one to 15 days each, with individuals using one to seven unique roosts and switching roosts from 
zero to seven times (Amett and Hayes, 2009). Both male and female M. evotis used snags, downed 
logs, and stumps that tended to be close (less than one kilometer) to water (88% were within 915 
meters of capture sites over ponds). Differences were not observed between sexes in roost use, and 
use of snags was nearly twice as high in stands with abundant snags (Arnett, 2007; Amett and 
Hayes, 2009). Used snags were in stands that did not differ in age from randomly selected stands. 
Douglas fir snags were used more frequently than other species, but no more than their typical fre¬ 
quency among randomly chosen snags. Use of stumps and logs as roosts was higher in stands with 
lower densities of snags (Amett and Hayes, 2009). These results suggest that snags may be pre¬ 
ferred roosts of these bats in coniferous forests of their study area in the Pacific Northwest, but that 
they exploit stumps and logs when snags are less available. 
In forests of British Columbia, long-eared myotis roosted in cavities under loose bark of 
stumps of ponderosa and lodgepole pines in clear-cut areas (Vonhof and Barclay, 1997). Nineteen 
roosts were found in 17 stumps (among 1,542 examined stumps) located in three of 11 searched 
clear-cuts and were occupied nearly exclusively by males and non-reproductive or post-lactating 
females. Clear-cuts with stumps used as roosts generally had less cover (downed logs and vegeta¬ 
tion) over stumps, and stumps tended to face southerly directions; these and additional character¬ 
istics of roosts suggested that both thermal/metabolic advantages and predator avoidance were like¬ 
ly factors of importance in stump-roosting by this species (Vonhof and Barclay, 1997). 
In Yosemite National Park in the California Sierra Nevada Range, long-eared myotis (includ¬ 
ing a maternity colony) were discovered using basal hollows of giant sequoia trees as roosts dur¬ 
ing summer (Pierson et al., 2006). 
