O’SHEA, CRYAN & BOGAN: UNITED STATES BAT SPECIES OF CONCERN 
179 
A subsequent study investigated snag use by long-legged myotis at roost (microscale), stand 
and landscape (mesoscale), and landscape (megascale) levels in forests consisting of ponderosa 
pine and other species in Washington, Oregon, and Idaho (Lacki et al., 2010). A total of 153 adult 
females were radio tracked to 395 roosts on six watersheds and roost use was compared with ran¬ 
domly selected roosts that were verified as lacking use by bats. The importance of different scales 
in influencing roost use varied by region. In Washington and Oregon, the amount of surface area 
of a snag that was covered with exfoliating bark provided the best model for explaining roost site 
selection by females (with greater amounts favored), whereas in Idaho measurements of the degree 
of fragmentation and foraging habitat quality provided the highest-ranking model (Lacki et al., 
2010). The most important characteristics in Washington and Idaho were amount of exfoliating 
bark on the snag, the height of the snag, and whether the top was broken (intact tops favored); no 
stand-landscape or landscape level characteristics had high importance values. In Idaho, roost scale 
characteristics had low importance values, but four characteristics were important at the higher lev¬ 
els: live tree density, stand size, amount of edge within 750 meters (less favored), and number of 
forested stands within 750 meters (fewer favored; Lacki et al., 2010). Proximity of roosts to water 
or variables indicating greater ease of roost-switching were not important in this study, but char¬ 
acteristics that indicated greater forest fragmentation had more negative associations with roost 
use. This study emphasized the regional differences that can occur among factors of importance to 
roosting bats (Lacki et al., 2010). 
Over the longer term, Lacki et al. (2012) followed fates of 339 snag roosts used by long-legged 
myotis in Washington, Oregon, and Idaho. Half-lives of snags were less than three years after dis¬ 
covery. Snag persistence varied with region and species of tree (fir snags were least persistent), and 
snags that were shorter in height, larger in diameter, and had fewer remaining branches were like¬ 
ly to persist the longest (Lacki et al., 2012). 
Vonhof and Barclay (1996) followed two radio-tagged individuals (male and female) in mixed 
forests of southern British Columbia, and they found that roosts were switched an average of 11 
days for the female and five days for the male, with distances between roost trees averaging 28 
meters in two moves of the female, and 206 meters in one move of the male. These five roosts were 
all under loose bark of unspecified species of conifers (Vonhof and Barclay, 1996). This species 
has been found roosting in cavities in both conifers and deciduous trees in central British Colum¬ 
bia (Psyllakis and Brigham, 2006) and under bark of a ponderosa pine in southeastern Montana 
(Jones et al., 1973). 
In Yosemite National Park in the California Sierra Nevada mountain range, two male radio- 
tagged long-legged myotis were discovered roosting during summer in basal hollows of giant 
sequoia trees, under bark in snags of sugar pine, and in a ponderosa pine snag (Pierson et al., 2006). 
They roosted in basal hollows of legacy trees (large old trees that have been spared during harvest 
or other disturbances) in commercial redwood forests of northern California, where they were the 
species most frequently identified by DNA analysis of fecal pellets collected from these hollows 
(Mazurek and Zielinski, 2004; Zielinski et al., 2007). 
Roosting habits of this species also have been studied in forests of the interior western states. 
In northern Arizona, warm season roosts of 13 radio-tracked adult females were all located in pon¬ 
derosa pine snags (Rabe et al., 1998a). In a ponderosa pine forest undergoing extensive manage¬ 
ment for restoration of historic characteristics on Mount Trumbull in northwestern Arizona, this 
species roosted primarily in ponderosa pine snags and switched roosts every one to five days 
(Herder and Jackson, 2000). Utilized snags were taller and larger in diameter, were on lower slopes, 
and had more exfoliating bark than randomly selected snags. Roost snags were also located closer 
to drainages and forest openings, had less dense canopy cover, were in stands with larger trees, and 
