5^. HARVARD UNIVERSITY Library of the Museum of Comparative Zoology H E MC7 '"£5 2 J 1995 GREAT BASIN NATURALIST VOLUME 56 NO 1 — JANUARY 1996 BRIGHAM YOUNG UNIVERSITY GREAT BASIN NATURALIST Editor KiciiAHi) \V. Baumann 29()MLBM PO Box 2()2()() Bri,i;;hani Yoiinjji; University Provo, VT s'Ki()2-()200 801-378-5053 FAX 801-378-3733 Assistant Editor Nathan M. Smiiii 190MLBM PC) Box 26879 Brigham Young University Provo, LIT 84602-6879 801-378-6688 E-mail: NMS(a)HBLLl. BYU.EDU Associate Editor: MiCllAKI. A. liOWKliS HlaiuK l'A|HMiiiK'iilal Farm, Univcrsitx' of Virginia, liox 175, Honxv, \'A 22620 J. H. ('Al.I.AllAN Museum of SoutliwvsttMii Biology, University of New Mexico, Alhuquerciue, NM Mailing address: Box 3140, I lemet, CA 92546 JKI'IKKY J. JOIIANSI'.N Department olliiologv jolm (Carroll llnixiMsity University Heights, Oil'} 1 1 IS Boius C KoxnuAriKi'i- Department ol Isntomology, (Colorado State Uiiiversit\, i'ort Collins, CO 80523 PaulC. Mahsii Center for Kn\ ironmental Studies, Arizona State Universitx, lempe, AZ 85287 SiANLKY D. Smith Department ol Biology University of Nevada-Las Vegas Las Vegas, NV 89154-4004 Paui.T. TUEl.l.KK Department olEnxironmental Hesonrce Sciences University oINevada-lUMio, 1000 \alle\ Hoad Reno, NV 89512 KoiU'Hi (;. WiirrMOHi'. Division ol I'ori'stry, Box 6125, West Virginia Uni\ersit\, Morgantown, \\'\" 26506-6125 Kdilorial Board, jerran T l'"liiiders, C^liairman, Botan\ and Range Science; Duke S. Rt)gers, Zoology; VViUbrd M. Hess, Botany and Range Science; Richard R. Tolman, Zoology. All are at Brigham Young Uni\ersit\. Kx Ollicio lulitorial Board members include Steven L. Taylor College of Biolo,g> and Agriculture; 11. Dnant- Smith, Director, Monte L. Bean Life Science Museum; Richard \V. Baumann, Editor Crcat Basin Naturalisf. The (Ural Hdsiii Ndtnnilist, founded in 1939, is puMished cinarterK !)>■ Brigham Young University. I ni)ul)lished manuscripts that hutluM- our biological undcMstanding ol the Creat Basin and surrounding areas in \\('st("rn Nortli America are accepted lor publie;ition. Subscriptions. Annual snbscriiilions to the Clrcdt lidsiit Ndiiinilist for 1996 are $25 for indixidual sub- scribers (.$.30 outside tlu- Uniti'd States) and $50 for institutions. The price of single issues is $12. .\11 back issues are in i)rint anil ;i\ ailable for sale. .Ml matters pertaining to subseri|itions, b;ick issui>s, or other busi- ness should be ilireeted to the P.ditor C.irat Basin Ndtwdlist, 290 MLBM, PO Box 20200, Brigham Young UnivtMsilN, l'ro\(), UT 84602-0200. Sclu)hirly Kxchanges. l,ibraries or other organizations interested in obtaining the Clrcdt Bdsin NdtnralisI through a continuing exchange of seholarK publications should contact the E.xchange Librarian, 6385 IIBLL, I'O Box 26889, Brigham Young UniMisily Provo, UT 84602-6889. Editorial Production Staff joAunc Abel 'iechnii'al l^ilitor jail Sptiieer Assist;iiit to the lulitor (.iopNiiulit (i'^ 19% In Brigham Young I'niversity OIReial piihlkation ilate; 31 January U)9(i ISSN 0017-3614 1-96 750 16974 The Great Basin Naturalist 1^1 BIJSIIi:i) Al PlU)\(), U TAII, 15V BUICMIAM Y()LIN(; UNIVKliSriY ISSN ()()17-;]()14 VOIAMK 56 31 January 1996 No. 1 Groat Basin Naliiralist 56(1), © 1996, pp. 1-1 1 TEMPORAL AND SPATIAL DLSrUlBUTlON Ol ilK;ilWAY MORTALITY OF MULE DEER ON NEWLY CONSTRUCTED ROADS AT JORDAN ELLE RESERVOIR, UTAH Laura A. Horniii' and jolin A. Bissoiictte^ AliSTKACT: — ill tliis pap rela- tionships. There is limited information that broadK characterizes mule deer use and kill distributions on and near highway systems, or that has investigated the influence of physical 1 an d scape featu res . Study Area The study area is located in the valley between the Wasatch and Uintah mountain ranges of northeastern Utah; the Provo River originates in the Uintah mountains and bisects the valley floor. Segments of three highways — 1995] ITlCIIW AY MOKTALITV DISTRIBUTIONS OF DRER US 40, state routes (SR) 32 and 248, totaliiit^ 47.3 km on the eastern slope of the Wasateh mountains in nortlieastern LI tali — were ehosen for study. Construction of the roadways was completed in 1989 and was necessitated by inundation of existing roads following con- struction of Jordanelle Reservoir. Filling com- menced in spring 1993. Dominant valley habitats consist of mesic meadow, riparian areas, and pasture lands. Surrounding drainage slopes are predomi- nantly within a mountain brush and sage- brush-grass zone (6000-7000 ft elevation), with scattered pinyon pine and juniper. Limited stands of aspen, cottonwood, and willow occur. Mule deer utilize the area as year-long range but usually are forced into the valley bottom during winters with heavy snowfall. Methods Deer roadkill data were collected at least once per week by research personnel from 15 October 1991 to 14 October 1993. UDOT and UDWR personnel assisted with collection efforts during their daily activities. Date, high- way identification, and location of each kill were recorded to the nearest 0.10 mile. Deer initially were classified as adult or fawn; incisors were removed from adult deer for age determination by cementum annuli proce- dures (Low and Cowan 1963). Deer kill zones and nonkill zones were des- ignated based on 1991-1993 deer-highway mortality locations. A minimum of 5 kills per mile had to have occurred for a segment of roadway to be considered a kill zone. A kill zone ended when a section of road did not contain a kill for more than 0.10 mile. We ran- domly selected 4 kill zone and 4 nonkill zone paired locations of 0.10-mile length along each highway and established transects to evaluate respective road alignment and associated habi- tat features. We recorded the distribution of kills over the entire study area, average traffic volume and speed for each highway percent vegeta- tive cover, and topography proximal to area roads. Kills were recorded to the nearest 0.01 mile. Twice monthly spotlight counts were conducted to document deer use and density adjacent to study area roads. Counts were ini- tiated at dark; each count averaged 3.2 h {s = 21 min). We began the spotlight run on a dif- ferent route each night. We drove along both sides of each road at a speed of 45-50 kph and used a hantlheld 400, 000 candlepower spot- light to locate deer. Deer were located to the nearest 0.10 mile. We stopped when deer were spotted to identify sex and age class, dis- tinguishing fawns by size. The activity of deer spotted in the right-of-way was classed as feeding, bedding, walking, or standing. We used statistical correlations to compare deer road- kill locations between years and with locations of live deer. Rangefinder readings were recorded at each 0.1-mile interval to provide an estimate of ob- servable area along each road (Fafarman and DeYoung 1986). Mountainbrush habitats de- creased deer visibility, and some areas along roads were not visible from a vehicle due to roadside rock cuts or steep declines bordered by concrete barriers. From numerous spotlight runs, we calculated the mean maximum visi- ble distance to be 500 m. Deer snow track counts were recorded along the right-of-way once each during the winters 1991-92 and 1992-93 to evaluate deer approaches to the roads. We counted the num- ber of trails within each 0.10-mile intei^val and described them as either parallel or perpen- dicular to the road. A parallel trail continued its direction for at least 30 m. Road alignment, right-of-way width and slope, right-of-way vegetation, and vegetation composition were characterized to a perpen- dicular distance 100 m beyond the right-of- way fence. Each highway was classified as either 4-lane or 2-lane with passing lanes. UDOT recorded traffic speed and volumes for each road during 2 periods: 11 March to 15 March 1992 and 29 June to 5 July 1992. Road alignments at each selected kill and nonkill transect location were described as cun^e, hill, or straight section. A cui-ve or hill was consid- ered part of the road alignment if it was within 100 m of the transect. Deer further than 100 m from the road are unlikely to be involved in an immediate collision (Romin and Dalton 1992); thus, beyond this distance, a hill or curve that would have reduced driver visibility had less significance. We analyzed habitat features during Sep- tember 1993. Stereoscopic aerial photography (1:24,000) was used to describe habitat features. We placed a transparent grid over photographs to determine percent cover (mountain brush Great Basin Naturalist [Volume 55 XJS40 30 M "' I 1^ ^ 10 O 20 M 1^ 10 3 .a. (^ SR248 1^ 1 -# 10 11 12 13 10 11 12 13 SR32 1 3S 30 2S :s 15 eS 10 ^^ s o 1. ^ J Fig. 1. Distribution of deer kill (%) by mile marker on 3 nevvlv built highways (US 40, SR 248, SR 32) at Jordanelle Reservoir, Utah, 1991-1993. and riparian areas) and topographical features at deer-highway mortality locations beginning at the road and extending 1.2 km distant. At each paired kill and nonkill location we estab- lished 3 habitat transect lines aligned perpen- dicular to the road. The transects were spaced 100 m apart and extended through the right- of-way zone 100 m past the right-of-way fence. We measured the length of each habitat along each transect line. Habitats included right-of- way revegetation, mountain brush, sagebrush- grass, grass-forb, aspen, cottonwood, willow, agricultural pastureland, riparian, and river. We calculated the proportion of each habitat present along the combined transect lines for each kill and nonkill location. We identified roadkill and live deer loca- tions, as well as descriptixe roadside features to 0.1 mile, consistent with highway mile marker delineation. We con\erted to metric units for analysis where appropriate. Results Deer locations We documented 397 deer roadkills during the stud\ from 15 October 1991 to 14 October 1993; 278 (5.9 kills/km) kills occurred during 1995] Highway Mortality Distributions of Deer Kill Zone \ ::$$>» Drainage Location Heber 1 Mile Fig. 2. Location of kill zones and associated drainages at Jordanelle Reservoir, Utah, 1991-1993. the 1st year of study (15 October 1991 to 14 October 1992), and 119 (2.5 kills/km) were documented during the 2nd year (15 October 1992 to 14 October 1993). Highway US 40 sustained the highest kill levels: 68% during the 1st year and 55% during the 2nd yean State routes 248 and 32 sustained similar kill levels; during the 1st year we recorded 18% and 14% of the total deer-highway mortality on SR 248 and SR 32, respectively. During the 2nd year we recorded 25% of the total annual kill on SR 248 and 19% on SR 32. Deer kills averaged <20 before the roads were relocated. Nineteen deer kill zones were identified based on the spatial distribution of deer road- kills during both years (Figs. 1, 2). The mean length of kill zones was 1.0 km {s = 0.62). Deer- vehicle collisions along US 40 occurred most frequently between mile markers 6.0 and 9.0 during both the 1st (56%) and 2nd (48%) years of the study. Twenty-eight percent of deer roadkills along US 40 occurred from mile marker 7.0 to 7.9 during the 1st year. Roadkill locations were correlated between years along US 40 at both the 1.0-mile (r = 0.69, P = 0.03) and 0.10-mile (r = 0.56, P < 0.001) interval. Deer kill locations were not signifi- cantly correlated between 1st and 2nd years along SR 32 at either the 1.0-mile (r = -0.14, P = 0.70) or 0.10-mile (r = 0.004, P = 0.968) 6 Great Basin Naturalist [Volume 55 scale. Deer kill locations along SR 248 were significantly correlated at the 1.0-niile intenal (r = 0.72, P = 0.02) but not at the 0.10-mile inten^al (r = 0.18, P = 0.07). Deer spotlight counts were not signifi- cantly correlated to kill locations at the 1.0- mile interval for any road during either year: SR 248 year 1 (r = 6.43, F = 0.19), year 2\r = 0.17, P = 0.61); SR 32 year 1 (r = 0.42, P = 0.23), year 2 (r = 0.12, P = 0.73); US 40 year 1 (r = 0.51, P = 0.14), year 2 (r = 0.15, P = 0.68). However, positive correlations were stronger during the first year. Fort>' percent of spotlighted deer were seen on the right-of-way. We identified the behav- ior of 968 (55%) of the deer along the right-of- way. Thirty-three percent were standing when first observed, 32% were feeding, 12% were bedded, and 23% were walking along the right- of-way or crossing the road. Perpendicular snow tracks were not corre- lated with deer-highway mortality locations (r = 0.29 , P = 0.42). Parallel tracks constituted 48% and 32% of all deer trails counted during the 1st and 2nd years, respectively. Traffic Characteristics Traffic characteristics contributed to deer- highway mortality levels (Table 1). Highway US 40 had the highest (3.7-9.9 times) mean 24-hr traffic totals of the 3 study area roads. Mean traffic speed was highest along US 40 (69.3 mph) from 11 March to 15 March 1992; however, over the 4 July weekend (29 June-5 July 1992), average speed along SR 248 (59.1 mph) was slightly higher than along US 40 (58.9 mph). Volume and speed were somewhat higher along SR 248 than along SR 32 for both test dates. Highway US 40 is a 4-lant> road and SR 248 and SR 32 are 2-lane roads with occasional passing zones. Road alignment (Table 2) was similar for transect kill and nonldll zone loca- tions (x^ = 1.2, df= 2, P = 0.70). Habitat From aerial photographs (1:24,000) we deter- mined that percent cover was greater along US 40 (63%) than along SR 248 (28%) or SR 32 (31%). Designated kill zones had higher mean percent cover (40%) than nonkill zones (29%). Highway deer kill along US 40 was highest in an area (mile markers 6.0-9.0) of 88%) vegeta- tive cover during both the 1st (56%) and 2nd Table 1. Traffic speed and volume of new routes at Jor- danelle Reser\'oir, Utiih, 1992. Speed (mph) Date Location Mean Maximum Vehicles/hr 11 Mareh- US 40 69.3 76.0 172.2 15 March SR248 56.9 72.0 37.9 SR.32 54.0 68.0 17.3 29 June- US 40 58.9 68.0 264.6 5 July SR248 .59.1 63.8 71.4 sr;32 .55.0 68.0 .37.8 (48%) years of study. Low mortality occurred in predominantly sagebrush-grass/wet meadow (mile markers 4.0-5.0) or agiicultm-al zones (mile markers 12.3-12.9) with <20%) cover. Along SR 248, agricultural zones sustained 1 deer (1%) mortality during the 2-yr period. State route 32 sustained 28% of its total deer road- kill in agricultural areas. However, 50% of this kill occurred at mile marker 9.0, located in a riparian area at an agricultural pasture and cliff interface. During spotlight censuses we observed a larger proportion of deer along right-of-ways associated with mountain brush habitat than along agricultural areas (Table 3). Paired / tests of microhabitat features showed no significant difference in proportion of cover 100 m beyond the fence between kill and nonkill locations {t = 0.13, df = 13, P = 0.90). Proportion of cover on the right-of-way ne\'er was higher than 29% for any transect. We examined 19 kill zones and 19 nonkill zones in the study area for associations with drainages (Fig. 2). Since deer-vehicle collisions occuned along nearly all of US 40, we evaluated the 8 highest kill locations along this road. Major drainages intersected the roads in 16 (79%) kill zones. Along US 40, large drainages intersected the highway at 6 (75%) of the kill locations. Two kill zone locations along US 40 weie at highway overpasses (mile markers 4.0 and 8.0); drainages were located within 0.2 miles. Two other kill zones extended past high- way underpasses (mile markers 8.2 and 11.4) Seven (37%) nonkill zones had drainages inter- secting the roads. Howe\er, in 4 of the nonkill zones, drainages were within 0.2 miles of a kill zone. Kill and nonkill locations did not differ in right-of-way widtlis (^ = 1.1, df = 13, P = 0.30). Deer kill per km was greatest when right-of- way areas were inclined rather than declined or level (Table 4). 1995] HiciiwAY Mortality Distiubutions of Deer Tablk 2. Road alitiniiu'iit at paiivtl (n — 42) kill and nonkill locations along stiid\ areas routes at Jordanelle Rcsenoir, Utali. Cunt- Straight Hill Kill Nonkill 15 19 23 21 1.2, df'=2. f = 0.70. T.\BLE 3. Deer observed (% of total deer) along right-of- \\a>'S associated with agricultural or nioinitain bnish habitat t>pes. Habitat US 40 SR 248 SR32 Agricultural Mountain brush 49 19 40 23 44 juvenile 2.5 3.5 AGE CLASS Fig. 3. Deer-highway mortality ages classes {n = 198), Jordanelle Reservoir, Utah, 1991-1993. Table 4. Deer kill per km relative to right-of-way slope relief along both sides of study area roads at Jordanelle Resen'oir Utah, 1991-1993. Road Right-of-way US 40 SR 248 SR32 No incline Incline 1 side Incline 2 sides 6.7 22.3 17.1 0.9 6.8 9.3 2.6 7.1 10.6 Temporal Deer Roadkill Distributions During winter 1991-92, mean monthly snowfall totaled 7.7 cm; mean monthly winter snowfall for 1992-93 was 46.9 cm. Of 397 deer mortalities documented during the study from 15 October 1991 to 14 October 1993, we clas- sified 205 (51.6%) does, 75 (18.9%) bucks, 86 (21.7%) fawns, and 31 (7.8%) unknown. Sixty- four fawns (16.1%) were female and 22 (5.5%) were male (Fig. 3). There was a 57% decrease from 278 (5.9 deer/km) deer roadkills during the 1st year to 119 (2.5 deer/km) roadkills dur- ing the 2nd year. We determined the age of 198 (70.7%) adult deer by cementum annuli techniques. Sixty-seven percent (n = 133) adult kills were < 2.5 yr old. The oldest recorded deer roadkills (2.5%) were 6.5 yr old. The 1992 hunter buck hai-vest from the Kamas district, east of the study area, also indicated a young population (n = 85); 55% yearlings, 15% 2.5 yr old, and 30% > 3.5 yr old (M. Welch, UDWR, personal communication). We located 4378 deer on 39 spotlight trips driving a total of 1845 km. There was a 64.2% decrease from an average 14.6 deer/km^ in the 1st year of the study to 5.23 deer/km^ during the 2nd year UDWR estimated a similar 70% reduction in the deer population on the Kamas District, attributed to the harsh 1992-93 win- ter (M. Welch, UDWR, personal communica- tion). We identified sex and age of 1515 (34.6%) spodighted deer: 987 (65.2%) does, 136 (8.9%) bucks, and 392 (25.9%) fawns. We calculated an obsei'vable area unobstructed by roadside barriers or dense vegetation of 10.98 km^ for the study area. We identified monthly and seasonal peaks in deer mortality (Table 5) by phenological period: fall (September-November), winter (December-Februaiy), spring (March-May), and summer (June-August). The following anal- yses treat the study period as year 1 (15 Octo- ijcr 1991-30 August 1992) and year 2 (1 Sep- tember 1992-14 October 1993), to allow inter- pretation of seasonal deer distributions and roadkill patterns. The highest roadkill peak (25%) occurred during No\'ember 1991. Thirty percent of the mortality in year 1 occurred during the fall even though data collection did not begin until 15 October 1991. Another peak (33%) was evident during the summer of year 1; 15% of the mortality for the year occurred in July. A similar fall peak (52%) occurred dur- ing year 2; 20% of the mortality occurred in October and 19% in November. A relatively large peak (18%) occurred in April. Eleven percent of the mortality occurred during the summer. During year 1, 41.8% of the annual Great Basin Naturalist [Volume 55 Table 5. Seasonal roadldll distributions (%) for each deer class at Jordanelle Resenoir, Utah, October 1991-August 1993. YearI Fall \\'inter Spring Summer Doe 30.0 16.0 10.2 44.0 Buck 14.5 27.3 16.4 41.8 ? fawn 57.4 2S.6 11.4 2.9 6 fawn 47.0 40.0 13.0 0.0 Year 2 Doe 65.4 6.2 13.6 14.8 Buck 52.6 10.5 26.3 10.5 9 fawn 50.0 21.4 28.6 0.0 6 fawn 40.0 40.0 0.0 20.0 buck mortality and 44.8% of doe mortality occurred during summer (Table 5). Fawn mor- tality peaked for both males (47%) and females (57.4%) in the foil. During year 2, the highest mortality among all sex and age classes occuned during fall. Seasonal distributions of deer-highway mor- tality' were compared to observed deer densi- ties during the same periods. Seasonal deer densities and highway mortalities were not significantly correlated (r = 0.54, P = 0.14) over the 2-year period (Fig. 4). For the period of summer 1992 to summer 1993, deer-high- way mortality and deer population density were strongly correlated (r = 0.94, P < 0.01), suggesting a density-dependent relationship. A negative coirelation existed between deer den- sities and kill/density (r = -0.68, F = 0.06). During year 1, observed deer density was low during fall (5.4 deer/km^) and winter (9.9 deer/ kni^) while highway mortality was high (fall = 71 deer, winter = 58 deer). Deer density (2.41 deer/ km^) and highway mortality (18 deer) were low during the 2nd winter Following winter 1992-93 deer density adjacent to study area roads increased slightly during spring (3.3 deer/km^) and summer (3.8 deer/km^). Observed density never reached pre-winter levels. Highway mortality levels of deer also increased (n = 31) in spring 1993 but did not return to pre-winter levels. Kill as a function of density was lower than observed deer den- sity from winter 1992 to winter 1993 but exceeded density following the harsh winter of 1992-93 (Fig. 4)^ The roadkill buckxloe ratio during fall (22.9:100) and early winter (78.9:100) of year 1 was greater than that observed in the living population (fall = 6.7:100, winter = 4.4:100) during the same periods (Table 6). Likewise, 30 25 20 ^ . •s^ 10 I I I I rl □ Year 1 ^ Year 2 I ! I Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul MONTH Fig. 4. Monthly deer4nghway ni()rtalit\ at Jordanelle Reserxoir Utah. 1991-1993. Aug 1995] lllCIIWAV MOKTALITV DiS THIIRITIONS OF DEER Tabi.K (i. Seasonal hiickidoi' latios ol roadkill ami spolliglil cicci- at JorclaiR'lk' He,Sfi"v<)ir, Utah, Ottohcr 1991 -October 1993. Se asons' • Counts' F91 W91 Sp92 Su92 F92 W92 Sp93 Su93 Kill Spotlight 22.9 6.7 7S.9 4.4 75 2.9 44.2 31.3 18.9 5.6 40.0 0.0 45 13.3 16.7 12.5 ■'Kill and spotliiiht count.s are recorded as l)ucks:100 does. ''Winter counts include only December and earK- Jannan'; sprini; counts include only April and May. Bucks arc prolialily underrcpresented. the roadkill huck:cloe ratio during the fall of year 2 (18.9:100) was larger than the ratio of the living population (5.6:100). The summer l:)uck:doe ratio was similar for roadkill and liv- ing populations during both years. For the months June-November 1992, the correlation coefficient between number of fawns involved in vehicular collisions and number observed on spotlight runs was significant: r = 0.84 {P = 0.04). For both summers the fawn:doe ratio of road-killed animals was 8.3:100, higher than the observed fawn: doe ratio (1.4:100) of the living population. Discussion We distinguished aspects of deer mortality based on traffic volume, habitat, topography, and seasonal distribution. Traffic volume signif- icantly influenced overall deer mortality levels. Though total kill in the study area decreased by 57% fi-om the 1st to the 2nd year, roadkills remained higher along US 40 than either SR 248 or SR 32. The 4-lane alignment of US 40 contributed to higher deer kills. Traffic vol- ume was higher and deer-vehicle collisions occurred more frequently along SR 248 than along SR 32 during both years. Vegetative cover along the length of US 40 was greater than along state routes 248 or 32. Likewise, percent cover was higher for desig- nated kill zones compared to nonkill zones. High percent cover appears to attract deer to right-of-ways for foraging. Agricultural areas provide abundant forage away from roadsides and were associated with low deer-vehicle collision levels. Deer usually approached roads along drainages, and higher kill levels occurred near large drainages. The ability to predict kill locations requires that kill locations remain similar over time. Kill location correlations at the 0.10-mile interval were low for SR 248 and SR 32 between the 2 yr The kill locations along US 40 were signifi- cantly correlated; however, most of US 40 was considered a continuous kill zone, which would lead to a correlation simply by coinci- dence. Although drainages provide highway ap- proaches, it is not possible to predict with exact- ness where deer-car collisions will occur based on habitat (% cover) and topography proximal to the roads. Deer often move parallel along the right-of-way after approaching a road. How- ever, inclined right-of-ways flinneled deer along the highway and were associated with higher kills. Low correlations between spotlight and kill locations further suggest that deer did not immediately cross the roads where they entered right-of-way areas. Snow trail counts also indi- cated parallel movement of deer. While seasonal deer-highway mortality dis- tributions tracked large fluctuations in popula- tion levels, behavior associated with life his- toiy activities of deer, e.g. fawning, breeding, and migration, also influenced year-round road- kill levels and composition. During the 2-yr study period, both roadkill and observed deer density levels decreased. When harsh winter conditions (1992-93) reduced population lev- els, deer-highway mortality was proportionally lower. Variability in the association between live deer density and roadkill numbers can be attrib- uted in part to deer-use patterns. Between fall and spring of year 1, highway mortality de- creased and spotlight counts recorded increased deer density. The mild winter that year allowed deer access to large areas and they maintained residence higher on drainage slopes. Weather conditions did not force deer to remain near area roads, although tliey fiequenfly approached and crossed roads. We attributed the initial increase in deer density during spring 1992 to the approach and congregation of deer along right-of-ways for foraging. Fall peaks in deer-highway mortality ap- peared related to activities associated with 10 Great Basin Naturalist [Volume 55 25 DEEDING SEVERE W UNTING SEASON 100 <«'' ^'^'^ # # ^* # SEASON Fig. 5. Seasonal deer-highway mortahty (no.) and den- sity (deer/km2) at Jordanelle Reservoir, Utah, 1991-1993. hunting and breeding during this time (Fig. 5). Deer were moving around the study area more frequently than during other seasons. Proportionally more bucks were involved in vehicular collisions during the fall than were obsei"ved in the population. The lireeding sea- son of mule deer in Utah begins the last few days of October, peaks between 20 November and 2 December, and declines through Janu- ary (Robinette and Gashwiler 1950). During the study, Utah deer and elk hunting seasons occurred from late August through October (T. L. Parkin, UDWR, personal communication). Fawns were involved in deer-vehicle colli- sions most often during the fall and least often during the summer of both years. The fawning period for mule deer in Utah begins appro.xi- mately 5 June, reaches and maintains a peak 11-20 June, and declines through 15 August (Robinc>tte and Gashwiler 1950). Fawns are seen inirec^uently during their first 6-8 wk because their predator defense is based on a "hider" strategy ((ieist 1981). Fawns were absent in the observed population during the sunnner but appeared during the fall. Does were involved in collisions and ob- served more frequently than males during both years. Si.xty-eight percent of adult deer roadkills were does, while 70% of fawns were female during year 1. Similarly, 81% of adult deer killed were does and 87.5% of fawais were female during year 2. Does have heavy energy demands associated with gestation, parturi- tion, and lactation, which may explain their association with high-(jualit>' roadside vegeta- tion and subscciuent high mortalitv rates. Management Recommendations Certain topographic features and vegeta- tion characteristics associated with roads, cou- pled with deer movement dynamics, predis- pose mule deer to highway mortality. Highway alignment and right-of-way topography often function to funnel deer to the right-of-way and encourage movement of deer along the high- way corridor, creating the potential for colli- sions at numerous locations. Roads planned in high deer-use areas that will sustain high traf- fic volumes should be prioritized for mitigative procedures during planning. Mitigative tech- nologies, particularly fencing with crossing stiiictures, should focus on the initial approach of deer to the highway along large drainages and take into account deer spatial dynamics and population trends. Continuing studies designed for species- specific and habitat-specific conditions may fiuther an understanding of why deer-vehicle collisions occur on a spatial and temporal basis, and promote development of appropri- ate pre-construction designs and mitigation strategies. Acknowledgments We thank die United States Bureau of Recla- mation (BOR), Utah Department of Transpor- tation (UDOT), Utah Division of Wildlife Resources (UDWR), and the United States Fish and Wildlife Senace (USFWS) for fund- ing and support provided throughout this study. We extend a special thanks to Lariy B. Dalton (UDWR), whose efforts made imple- mentation of this study possible. We sincerely appreciate the efforts of personnel who assisted with roadkill data collection: UDOT (Kamas maintenance shed crew: Shane W. 13ushell, Doug C. Gines, Ken L. Moon, Tyler K. Page, and Dave H. Sundquist) and Delmar C. Waters, a private contractor. G. David Cook, Justin L. Dalton, Larry B. Dalton, and Herb C Freeman provided valuable assistance dur- ing spotlight counts. Literature Cited Ai.LKN, R. E., AM) D. R. McCl;li,()Uc;ii. 1976. Deer-car accidents in .southern Michigan. Journal of Wildlife Management 40: 317-325. Basiiore, T. L., W. M. Tzilkowski, and E. D. Bellis. 1985. AnaKsis ol deer-vehicle collision sites in 1995] Highway Mortai,i iy Disti{ii5Uti()ns of Deer 11 PcnnsyKaiiia. Journal of Wildlife Manaut'iiK'iit 49; 769-774. Bkii.is, E. D., and H. B. Gkavics. 1971. Deer mortality on a Pennsylvania interstate highway. Jonrnal of Wildlife Management 35: 232-237. Carbaich, B., J. P Vaughan, E. D. Bellis, a.nd H. B. Graxes. 1975. Distribution and activity of white- tailed deer along an interstate highway. Journal oi Wildlife M anagement 39: 570-58 1 . DusEK, G. L., R. J. Mackie, J. D. Herrices, Jr., and B. B. CoMPTON. 1989. Population ecology of white-tailed deer along the lower Yellowstone River. Wildlife Monographs 104: 1-68. Fafarman, K. R., and C. A. DeYoung. 1986. Exaluation of spotlight counts of deer in south Te.xas. Wildlife Society Bulletin 14: 180-185. Geist, V. 1981. Behavior: adaptive strategies in mule deer. Pages 157-223 in O. C. Wallmo, editor, Mule and black-tailed deer of North America. University of Nebraska Press, Lincoln. Goodwin, G. A., and A. L. Ward. 1976. Mule deer mor- tality on Interstate 80 in Wyoming: causes, patterns, and recommendations. USDA Forest Service Research Note RM-332. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. Pages 1-4. Jahn, L. R. 1959. Highway mortality as an index of deer population change. Journal of Wildlife Management 2: 187-196. Kasul, R. L. 1976. Habitat factors associated with mortal- ity of southern Michigan wildlife on an interstate highway. Unpublished master's thesis, Michigan State University, East Lansing. 39 pp. Kramer, A. 1971. Notes on the winter ecology of mule and white-tailed deer in the Cypress Hills, Alberta, Canada. Canadian Field-Naturalist 85; 141-145. . 1973. Interspecific behavior and dispersion of two sympatric deer species. Journal of Wildlife Manage- ment 37; 288-300. Kress, M. J. 1980. The effects of habitat on the distribu- tion of white-tailed deer {Odocoileus virginianus) along a Pennsylvania interstate highway. Unpublished doctoral dissertation, Pennsylvania State University, University Park. 56 pp. Low, W. A., and I. M. Cowan. 1963. Age determination of deer by annular stnacture of dental cementmn. Jour- nal of Wildlife Management 27; 466—471. Mansfield, T. M., and B. D. Miller. 1975. Highway deer-kill district 02 regional study. Caltrans internal report. Sacramento, CA. 49 pp. Myers, G. T. 1969. Deer-auto accidents; serious business. Colorado Outdoors 18; 38-40. Pi'.EK, F W, and E. D. Bellis. 1969. Deer movements and behavior along an interstate highway. Highway Research News 36: 36-42. Puc;lisi, M. J., J. S. Lindzey, and E. D. Bellis. 1974. Fac- tors associated with highway mortality of white- tailed deer. Journal of Wildlife .Management 38: 799-807. Reed, D. F. 1993. Efficacy of methods advocated to reduce cervid-vehicle accidents: research and rationale in North America. Colorado Division Wildlife Resources, Fort Collins, CO. 13 pp. Reeve, A. F 1988. Vehicle-related mortality of mule deer in Nugget Canyon, Wyoming. Wyoming Coopera- tive Fisheries and Wildlife Research Unit, Laramie. 75 pp. Reilly, R. E., and H. E. Green. 1974. Deer mortality on a Michigan interstate highway. Journal of Wildlife Management 38: 16-19. ROBINETIE, W. L., AND J. S. Gashwiler. 1950. Breeding season, productivity, and fawning period of the mule deer in Utah. Journal of Wildlife Management 14: 457-469. RoMiN, L. A., AND J. A. Bissonette. In press. Deer-vehicle collisions: nationwide status of state monitoring activities and mitigation efforts. Wildlife Society Bulletin. RoMiN, L. A., AND L. B. Dalton. 1992. Lack of response by mule deer to wildlife warning whistles. Wildlife Society Bulletin 20; 382-384. SICURANZA, L. P 1979. An ecological study of motor vehi- cle-deer accidents in southern Michigan. Unpub- lished master's thesis, Michigan State University, Lansing. 63 pp. Utah Division of Wildlife Resources. 1992. Utah big game annual report. Reported yearly summar>' of deer-highway mortality 1982-1992. Utah Division Wildlife Resources, Salt Lake City. Vaughan, J. R 1970. Influence of environment on the activity and behavior of white-tailed deer (Odocoileus virginianus). Unpublished doctoral dissertation, Penn- sylvania State University, University Park. 73 pp. Received 9 May 1995 Accepted 9 October 1995 Great Basin Naturalist 56(1), © 1996, pp. 12-21 EXCEPTIONAL FISH YIELD IN A MID-ELEVATION UTAH TROUT RESERVOIR: EFFECTS OF ANGLING REGULATIONS Wayne A. Wurtsbaughl, David Barnard^ and Thomas Pettengill^ Abstract. — We used creel surveys to evaluate how a change from a 6-mon to a year-round fishing season affected the sport fish harvest in East Canyon Reservoir (Utah), a 277-ha mesoeutrophic system. Under the year-round season, fishing effort was 840 angler-h-ha-^-yr-l, and 360 trout ha~l were captured. Catch rates were proportional to estimated trout densities in the resei"voir, ranging fi^om 1.06 during the winter ice fishery, to 0.18 fish angler~lh~l in July. Ninety- nine percent offish hai-vested were rainbow trout {Oncorhijnchus mijkiss). Thirty-two percent of the 300,000 75-mm fin- gerling trout stocked annually were captiu-ed by anglers within 2.5 yr, but return rates varied with the strain and/or size of trout stocked. Annual fish yield was 102 kg/lia, among the highest yet reported for a temperate zone, lacustrine sys- tem. Extending fishing from a 6-mon season to year-round increased the number of fish captured and provided almost twice as many hours of recreational fishing in the reservoir The harvest period was changed from traditional spring-simimer months to primarily a winter-spring fisher\' because relatively few trout sui-vived for more than 6 mon after reaching hanestable size. Although salmonid production in East Canyon Resei^voir is veiy high, the fishery is in a precarious state because high primaiy producti\'it\' dri\'en, in part, b\ cultural eutrophication, makes water quality sub- optimal din-ing midsummer Key words: reservoir, yield, trout, creel, harvest, strains, regulation, productivity, fish, management, growth, Oncorhynchus mykiss. Important goals of lake and reservoir man- agement are to maximize both fish yield and recreational use. Methods of increasing yield include introducing different species or strains, lake fertilization, and modifying fishing regu- lations (Hall and Viin Den Avyle 1986, Stock- ner 1992). Modification of littoral zone escape habitat may also be important (Wurtsbaugh et al. 1975, Trendall 1988, Tibor and Wurtsbaugh 1991). Changes in fishing regulations, how- ever, offer a manager the most flexibility (Carl- ton 1975), and these changes are less likely to damage the ecosystem than are the other meth- ods. In 1985 the State of Utah changed from a 6-mon open season for trout (late May- November) with a daily limit of 8 fish, to a year-round fisheiy with no seasonal closures and a daily limit of 8 fish. To investigate how this management change affected the fisher); we conducted a 1-yr creel survey in 1986 to determine timing and magnitude of harvest from East Canyon Reservoir: we then com- pared these results with harvest characteris- tics measured in the reservoir in 1970 and 1972 under the 6-mon regulation. The 1986 creel survey also allowed us to measure the high fish yield of the reservoir and to relate it to various limnological parameters affecting fish production (Carline 1986). We were also able to investigate how different strains of trout stocked in the resei^voir recruited to the fisheiy (Brauhn and Kincaid 1982, Babey and Berry 1989). This work was part of a comprehensixe study on the ecology and causes of mortality of stocked rainbow trout in mid-elevation reser- voirs in Utah. Study Area East Canyon Resenoir is located at an ele- vation of 1734 m in northern Utah (Morgan County; 4()°54'N, 110°35'W). East Canyon Creek and other minor tributaries of the reser- voir drain a 99,200-ha watershed in the cal- careous Wasatch Moim tains. At full pool the resei^voir is 5.6 km long, 60 m deep, and cov- ers 277 ha (Table 1). The resei-voir is produc- tive, with a mean summer (May-Oct) chloro- ph\ll a concentration of 5.4 mg/ni'^ (1985-86 and 1989-90 mean), and a mean Secchi depth of 4.6 m (W. Wurtsbaugh unpublished data). Blooms of cyanobacteria occur frecjuentK' dur- ing sunmier and lall. Annual total phosphorus (TP) loading of 2.8 g m~-yr~^ is very high 'Dcparliiient of Fisheries and \Vildlife/EcoloR>' Center. Utuli Stale l.'ni\ersit\, Loyan. V'V 84322-5210. ^Utiili Division ol Wildlife Resources, l,59(i West \ortli Temple, Salt Lake C.'it). IT S4I16. 12 1996] Trout Yield in Utah Reskhvoir 13 Table 1. Limnological characteristics of East Canyon Reser\oir, Utah. Data sources: ''Utah Department of Health (1982); ''Merritt et ah 1980. Other data are unpuh- lislii'd data of W. Wurtshaiigh. Elevation-' 1734 m Area (hill pool)-' 277 ha Volume (hdl pool)^' 63,200 nv^ Mean and m;L\inunn depths-'' 23 & 60 m Shoreline length-' 16 km Ch!oroph>ll a (May-Oct) 5.4 /i,g/L Seechi depth (May-Oct) 4.6 m AlkalinitN-' 3.4 mE(iui\' Total hardness" 233 mg/L Total dissolved solids" 328 mg/L Annual phosphorus loading'' 2.8 g ni~^yr~^ Mean water colimin total phosphonis-'' 80 /U.g/L tMit color of fliiorcsccMit pigment (Phinn(>y et al. 1967, Vondracek et al. 1980). Trout grow quickly in East Canyon Reser- voir and enter the fishery within 5 nion. The reservoir is intensively fished due to its prox- imity to 2 major population centers, Salt Lake City and Ogden. Creel sui'veys in the 1970s indicated fishing effort at over 300 angler- hha~^yr~^ Because anglers fish primarily with bait, there is little catch-and-release fish- ing. Most trout captured are less than 350 mm. Schrader (1988), Babey and Beny (1989), and Tabor and Wurtsbaugh (1991) provide additional information on the fish and fishery. (Merritt et al. 1980), and mean water column TP is 80 )ag/L (Utah Department of Health 1982). Algal growth in the reservoir, however, is limited primarily by nitrogen (Wurtsbaugh 1988). The reservoir's water level fluctuates widely because of water withdrawals for irri- gation, and consequently there is little macro- phyte development in the littoral zone. During much of the summer, oxygen concentrations in the hypolimnion drop below 1 mg/L. Epilim- netic temperatures reach 22°C in July, and the reservoir is typically ice covered from late December through March. During much of the year high densities (>10/L) oi Daphnia pulex, D. galeota, and other crustacean zoo- plankton are evident (Tabor and Wurtsbaugh 1991, W. Wurtsbaugh unpublished data). Additional limnological information is given in Table 1. Dominant fishes in the lake, in approximate order of biomass, are Utah suckers {Catosto- mus ardens), redside shiners {Richardsonius balteatus), and rainbow trout {Oncorhynchus mykiss). Less-abundant species are cutthroat trout (O. clarki), brown trout {Salmo trutta), speckled dace {Rhinichthys osculiis), fathead mii.nows {Pimephales promelas), and kokanee (O. .lerka). Rainbow tiout are heavily parasitized by anchor worms {Lernaea cyprinacea- Berry et al. 1991). In late May the Utah Division of Wildlife Resources stocks 300,000 (1080/ha) rainbow trout, approximately 75 mm in length, in East Canyon Reservoir. Fish captured by anglers during our 1986 creel survey were derived from several strains of rainbow trout stocked in 1984-1986 (Table 2). Each strain stocked in the reservoir was spray marked with a differ- Methods Creel data were collected during 1970, 1972, and 1986 by interviewing anglers and by count- ing the total number of anglers on the reser- voir. Sampling effort was stratified by weekday and weekend, month, time of day (morning, midday, and evening), and method of fishing (ice, shore, and boat), with random samples taken within each stratum (Malvestuto 1983). The creel clerk determined the number of fish released and the number, length, and weight (1986 only) of each species or strain kept. In 1986, 25% of the weights were not measured. These were subsequently estimated with an empirically derived length-weight regression for rainbow trout: W = 1.619 10-5 • TL2 949 ; fi2 = 0.95, where W = wet weight in grams and TL = total length in mm. Sample estimates were expanded to provide monthly and seasonal totals for fish harvests and angler use. Details of the methods varied somewhat be- tween surveys in the 1970s and those in 1986. In 1986 we sampled 5 weekdays and 4 week- end/holidays each month of the year In 1970 and 1972 the sampling inteival lasted only fi-om opening day (Memorial Day weekend) through August. Creel surveys in 1970 and 1972 were done on both days of the opening weekend: during the remainder of the sampling period the reservoir was randomly censused on 20 (1970) or 48 d (1972). Because catch infomiation was unavailable for the September-November periods in 1970 and 1972, we restricted com- parison with the 1986 catch statistics to the Januaiy-August intei^val. Nevertheless, in 1986, 14 Great Basin Naturalist [Volume 56 Table 2. Sizes (± standard deviation) and percentages of rainbow trout strains planted in East Canyon Resei-voir from 1984 to 1986, and percent of those fish captured by anglers during 1986. Each year 300,000 fish were stocked in the resei-voir Relative return of each strain was calculated: [100 (% returned / % stocked)] -100. A large (L) and small (S) group of Ten Sleep strain were planted in 1986. Shepherd = Shepherd of the Hills strain. Strain-Size Mean stocked weight (g) ±s % stocked Number captiued % captured Relative return 1984 Kamloop Ten Sleep McConaughy 4.7 ±1.4 5.0 ±1.6 5.8 ±3.1 32 36 32 2,300 2,400 4,500 1985 25 26 49 -22 -28 +53 Kamloop Ten Sleep Shepherd 7.5 ± 2.3 5.4 ±1.4 3.7 ±1.4 33 33 33 33,000 27,700 16,100 1986 43 36 21 +30 +9 -37 Ten Sleep-L Ten Sleep-S Shepherd 4.8 ±1.4 3.1 ±1.1 4.2 ±1.4 25 25 50 3.300 2,800 6,700 26 22 52 +4 -12 +5 85% of the effort and 81% of the annual rain- bow trout harvest occurred by the end of August (see below), indicating that earlier sur- veys provided a reasonable assessment of the fisher)'. During 1986 we identified fluorescent- marked rainbow strains using a portable, bat- teiy-powered black light affixed within a light- exclusion box. Fish captured during the year they were planted were designated age 0, and those captured during the 2nd and 3rd year after planting as age 1 and age 2, respectively. We analyzed creel data with the FORTRAN program WCREEL, supplied by the Utah Divi- sion of Wildlife Resources (B. Schmidt personal communication). Temporal changes in trout abundance in a put-grow-and-take fisheiy such as that in East Canyon Reservoir can be evaluated by the number of fish removed from the system by anglers because relativeK' little mortality occins from other factors after trout attain a harvest- able size. For example, estimated losses of all sizes of trout to birds, which has been shown to bc> important in some Utah resenoirs (Waso- wicz 1991) and elsewhere (Matkowski 1989), accounts for approximately 6% of planted trout in East Canyon Reservoir (R. A. Tabor unpub- lished data). Piscivorous fish eat over 25% of stocked trout, but this loss is negligible once prey reach 150 mm (Wiu-tsbaugh 1987 and unpublished data). Furthermore, because the reser\()ir has a deep release, located in the hyi^olimnion, we believe that few fish emigrate, although we lack quantitative data to support this. Had we used this approach to estimate abun- dance of trout planted in 1985 that reached harvestable size, we would have required creel data from at least 3 consecutive years (1985-1987), or until anglers had removed all of the cohort. Because we measured harvest only during 1986, and thus lacked a long-term data set, we assumed that harvests of age 0 fish in 1985 and age 2 fish in 1987 were simi- lar to the measured harvest of age 0 (stocked 1986) and age 2 (stocked 1984) fish during 1986. Because 80-90% of each strain was har- \'ested as age-1 fish (see below), violations of this assumption should not have seriously affected our analysis. To determine the effect of trout densit\' in the reservoir on monthly success rates for anglers, we graphed the estimated density of fish remaining to be captured from the 1985 cohort against catch per hour for fish in that cohort. At the beginning of the survey in Janu- ar>' 1986, we estimated that 67,400 fish from the 1985 cohort were available in the reser- voir This density was based on total catch of the cohort in 1986 plus an additional 9000 fish estimated to have survived into 1987. Nine thousand (3%) of the 1984 cohort sunived over 1 yr and were captured by anglers in 1986. Fish densities for subsequent months were calculated b>' subtracting the previous month's 1996] Trout Yield in Utah Reservoir 15 liarvest. The resulting regression from tliis analysis may include some bias, since mea- surements of fish densities each month were not independent of each other For this reason we did not calculate statistical significance levels for the regression. Nevertheless, the approach yields a useful estimate of the rela- tionship between abundance and catch rates. To estimate the mortality of trout that were captured and then released by anglers, we assumed a survival rate of 69% (Taylor and White 1992). Because sizes and ages of these released fish were unknown, we assigned pro- portions to the different year classes. Our inter- views with anglers indicated the main reason fish were released was because of small size, but a limited number were also returned be- cause of the presence of ectoparasites {Lernaea) or their scars. We therefore assumed that 90% of returned fish were age 0 (i.e., returned be- cause of small size), 10% were age 1 (returned for cosmetic reasons), and no age 2 were returned. Results The Fisheiy Under a Six-month Season Under the 6-mon open season documented in 1970 and 1972, fishing was concentrated from the opening weekend in late May through August. Fishing on the opening week- end accounted for 16-21% of the estimated total effort, and 28-38% of the rainljow trout harvest (Table 3). Fishing pressure dropped steadily through the summer, and catch rates varied from 0.18 to 0.49 trout/li. Total fishing effort was similar in 1970 and 1972, with the lake providing over 350 angling-h/ha. Anglers harvested an estimated 60,100 rainbow trout during the survey period in 1970, but only 35,600 in 1972 (Table 4). The catch rate for rainbow trout in July 1972 was much lower than in other months. This was due, in part, to anglers fishing for kokanee and a strain of albino rainbow trout that made up 44% of the July hai-vest. The total catch rate of 0.32 fish/h was comparable to other months of the year when kokanee and albino trout were har- vested less (4% of the catch in June and 9% in August). The Year-round Fishery Trout grew rapidly in East Canyon Reser- voir, particularly during their 1st year (Fig. 1). Fish were planted in May at a mean size of 75 mm and 3.8 g. When they first entered the fisheiy in July, they were 178 mm and 77 g. By July, the previous year's cohort of fish had reached 305 mm and 420 g. By the end of the 3rd year, fish had reached 400 mm and 728 g. In 1986 anglers spent over 230,000 h (±9300, .sy) fishing in East Canyon Resei-voir, or 840 angler-hha"lyr"l. Most of these hours were by shore anglers (58%), followed by boat Table 3. Pressure, harvest, and catch rates for rainbow trout for creel surveys conducted in 1970, 1972, and 1986 for the January-August period. Eadier surveys lasted only fiom the opening weekend (Memorial Day — the last weekend in May) through August. In 1986 the state changed to a year-round season, so there was no opening day. Only the Janu- ary-August data of 1986 are shown here to facilitate comparisons between the 2 periods. Total catch for the year is shown in Table 4. Jan-May Opening weekend June Julv August Total Jan-Aug 1970 Effort (h) Harvest (nuinber) Catch rate (fish/h) — 19,100 18,300 0.96 58,600 21,900 0.37 1972 28,600 13,900 0.49 13,600 6,000 0.44 119,900 60,100 0.50 Effort (h) Harvest (number) Catch rate (fish/li) — 22,100 13,400 0.61 40,700 12,600 0.31 1986 32,500 5,800 0.18 11,600 3,800 0.33 106,900 35,600 0.33 Effort (h) Harvest (number) Catch rate (fish/h) 118,000 66,800 0.57 — 42,400 8,300 0.20 22,700 3,200 0.14 15,900 2,500 0.16 199,000 80,800 0.41 16 Great Basin Naturalist [Volume 56 Table 4. Total catch of salinonids from East CaiiNon Reservoir in 1970, 1972, and 1986. In 1970 and 1972 yields were estimated from the start of the fishing season, in June, through August. Data for 1986 show captures during the entire year. Table 3 shows the comparable catch in 1986 from the opening day through August. 450 I 350 1970 1972 1986 z Ul _l _l < 250 Tax.\ Rainbow trout 60,100 35,600 98,960 O 150 Albino rainbow trout^ — 1.200 — 1- Brown trout 200 20 60 Cutthroat trout 0 500 700 50 Kokanee-' — 3,900 100 Tot.^l 60,300 42,220 99,820 "First stocked in 1970 (24%) and ice anglers (18%). The relative dis- tribution of angling type varied seasonally; In January and February, nearly all fishing was done through the ice, but subsequent fishing pressure was dominated by boat and particu- larly shore anglers (Fig. 2A). Total fishing pres- sure reached a peak during May, the period of the traditional opening day. Monthly catch rates for rainbow trout var- ied from a high of 1.06 fish/angler-h in Febru- ary to 0.18 in July (Fig. 2B). Annual catch rates were 0.92 for ice anglers, 0.34 for boat anglers, and 0.30 fish/h for shore anglers. The average for all types of fishing was 0.42 fish/h. Catch rates for ice anglers in Januaiy and Februaiy were the highest for any month or method for the year (Fig. 2B). There was a strong relationsliip between the estimated density of trout from the 1985 cohort remaining to be captured and monthly catch rates for those fish (Fig. 3). In Januaiy and Februaiy when there were more than 200 fish/ha (0.2 fish/m^) in the reservoir, catch rates were over 0.6 fish/angler-h. As densities dropped, however, catch rates declined pro- gressively, reaching a low of 0.1 fish/angler-h in December. We estimate that 99,300 ± 7500 (sy) game fish were removed from East Canyon Reser- voir by anglers in 1986. Of these, 99.1% were rainbow trout, 0.7% were cutthroat trout, 0.1% were kokanee salmon, and <0.1% were brown trout. Sixty-eight percent of the annual har- vest of rainbow trout occurred from Januar\' through May, and 38%; of these were captured in January and Februaiy during the ice-fishing season (Fig. 2C). Rainbow trout planted the previous year (1985) dominated the catch from Januaiy to August of 1986 (Fig. 2C). Rainbow TOTAL LENGTH J 1984 y ' 1985 / w^^ " - 1 - J^986 ■ 700 600 § 500 £ 400 O I 300 200 100 0 - WET WEIGHT ^' - 1 /i984 : - r^J - ^ ^1985 COHORT - •V - -m*^ . ^^1986 , _ ■ YEAR 1 YEAR 2 YEAR 3 Fig. 1. Changes in total lengths (above) and wet weights (below) of the 1984, 1985, and 1986 cohorts of rainbow- trout captured by anglers in East Canyon Resei^voir All fish were captured during 1986 but are plotted over a 2.5- \T period to show long-term growtli rates. Also plotted are initial lengths and weights of the fish stocked in 1986. Total lengths (TL) can be converted to standard lengths (SL) by dividing by 1.15. trout planted in May 1986 first entered the fisheiy at a mean total length of only 178 mm in Jul\', and by October this cohort dominated the harvest. Although age 0 and age 2 fish were important in the fishen' eaiK' and late in the year, 78% of the total catch was of age 1 fish from the 1985 planting. Anglers released 37,000 hooked fish during 1986, giving an estimated mortalitv' of 10,400 fish during the 1st >'ear the>' were in the reser- voir and an additional 1100 in the 2nd xear. Consequently, approximately 4% of stocked fish are lost because of hooking mortalit). About 75% of this mortcilit)' occuned fiom JuK' througli December when small trout first entered the fisheiy. Total fish yield in East Can\on Reservoir during 1986 was 102 kg/ha. Most of the hanest occuned before July (Figs. 2C, 2D). Fish planted the previous year represented 82% of the bio- mass of rainbow trout captured in 1986. 1996] Trout Yield in Uiaii Hkskhvoik 17 50000 ^■K X t- 40000 z o 30000 ^ « 20000 oc 3 O 10000 X ^•^ 0 1.6 1.4 oc 1.2 3 O 1.0 X 0.8 T W 0.6 u. 0.4 StMlO 0.2 0.0 ^„^ 30000 X H- Z o 20000 ^ cc UJ 03 10000 s 3 Z *-' 0 8000 ^ X 6000 1- z o 4000 S o> )^ 2000 : B. CATCH RATES - jn ICE : ^ \ ^ BOAT -■••-. ■-♦-. --■■' ^<.j:r-' SHORE M B ^ ^ .' .■«k ■ . D. YIELD ■ k A. ^TOTAL ■ Biib^ ^ Ota ^ ■iffe*^ lilllllilH^' ^ >^ |i984lHHIIHHI .: <---i:^iiSi JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NCV DEC Fig. 2. Seasonal changes in fishing effort and rainbow trout captured during 1986 in East Canyon Reservoir A, Fre- quency polygon of seasonal changes in effort expended in ice fishery (ice), boat angling, and shore angling (top line shows total fishing effort); B, monthly changes in catch rates for the 3 fishing methods; C, numbers of rainbow trout cap- tured each month during 1986 from the 1984, 1985, and 1986 cohorts of rainbow trout planted in the resenoir; D, total and component yield of trout from each cohort captured during 1986. Hai-vest of Different Strains Four strains were in the reservoir during 1986 as a result of stocking in 1984, 1985, and 1986 (Table 2). Relative proportions of each strain harvested fluctuated seasonally. McConaughy strain from the 1984 stock and Kamloop trout from 1985 were captured more than expected in the winter and spring catches of 1986. In the summer, however, catch rates of Kamloop and Ten Sleep from the 1985 stocking were similar for the rest of the year. Shepherd of the Hills strain stocked in 1985 was han'ested less than the other two strains planted that year. During 1986 there were sig- nificant differences in harvest rates of differ- ent strains planted in 1984 (X^ = 13.34, p < 0.05) and in 1985 (X^ = 7.76, P < 0.05), but not in 1986. A large percentage of each strain stocked in the reservoir was eventually captured by anglers. There were, however, considerable dif- ferences in relative return of different strains. We estimate that 40% of Kamloop, 32% of Ten Sleep, and only 23% of Shepherd of the Hills strain were captured during their first 2 1/2 yr in the reservoir (Fig. 4). For all strains com- bined, 32% of the fish stocked were eventually captured by anglers. 18 Great Basin Naturalist [Volume 56 ^ 1.U ' T ' 1 ■ O F X ^ 0.8 . _ z y ^ ;;: 0.2 - "osXi - 0 J/* J 9 lU " Y - 0.0053 ♦ 0.0031 X r - 0.79 3 O. n n ^_^ . 1 > 1 ■ 100 200 TROUT DENSITY (No. / Ha) 300 Fig. 3. Relationship between monthly estimates of the density of rainbow trout remaining in the 1985 cohort and catch per unit effort (CPUE) for those fish in the reservoir. The CPUE shown here is less than in Figure 2B because it does not include fish from 1984 and 1986 cohorts that were captured, nor the captiue of other species. Letters on graph indicate months. Discussion The fishing regulation change in East Can- yon Resei-voir resulted in an excellent winter ice fishery but poorer summer angling than when a 6-mon season was in effect. In 1970 and 1972 anglers harvested 30-37% of the annual total during the intensive 3-d opening (Table 3), but large numbers of fish still remained in the lake to support a summer fish- ery with catch rates of 0.3-0.5 fish/h. In 1986, however, about 66% of the fish had been har- vested in the winter and spring fisheiy by the time of the traditional opening day. Monthly estimates of pressure during the summer fish- ery (June-August) for 1986 were similar to those in the earlier studies (Table 3), but the sununer harvest was only 33-64% of that in previous years. While failing to maintain the tiaditional catch rate for summer months, the regulation change may have provided a fishery that not only pro- duced increased numbers offish over a longer period of time, but also provided almost twice as many hours of recreational fishing as under the 6-mon open season (Table 3). If the popu- larity of winter angling were to increase sub- stantially, an even larger proportion of trout would be captured then, leaving fewer for the traditional spring and summer fisheries. To spread the catch over a longer period, the State of Utah reduced the winter bag limit to 4 fish subsequent to our study. The differences noted under the different angling regulations must be treated cautiously, however, as only 1 yr of data was available for the year-round season, and substantial between-year differences were noted for the 1970 and 1972 period. Factors such as changing predation pressure from pisci- vores and changes in nutrient loading to the reservoir undoubtedly also contributed to changes in the fishery. Catch rates for the 1985 cohort of fish were clearly related to monthly changes in the den- sity of these fish (Fig. 3), but there may have been additional factors influencing fishing success. Catch rates in February were higher than the prediction based on density'. The rea- son for this is not clear, but it is possible that catch rates were especially high during mid- winter when available food was low. Catch rates in June-August were somewhat below the regression, perhaps because during warm months of the year fish are concentrated in deeper water near the thermocline where they are more difficult for anglers to reach. Catch rates increased, relative to the regression, in the fall (September-November) when the reser- voir began to cool. Despite relative minor sea- sonal shifts, it appears that densities of rain- bow trout available in the reservoir can explain most of the variation in catch rates. Significant differences in the relative har- vest of different strains of rainbow trout were not unexpected, as others have found that strains stocked can have large effects on the fisheiy (e.g., Brauhn and Kincaid 1982, Babe\' and Berry 1989). The poor return for Shep- herd of the Hills strain (Table 2) is consistent with the poor return of this group in East Canyon Reservoir reported by Babey and Beny (1989). Nevertheless, 2 factors confound the interpretation of these results. First, despite efforts to control sizes offish planted, there were sometimes substantial differences in weights of different strains stocked. For each annual cohort, the relative return of a strain was correlated with its size at stocking (Table 2); groups stocked at a large size usualh sunived better than smaller ones. Second, because our creel survey lasted only 1 yr, we could not determine if some strains entered the fisheiy as (luickK as others. For example, the veiy high relatixe return rate of the McConaughy strain in 1986 may be a consequence of a very low catch rate of these fish measured in 1984 and 1996] Tkout Yield in Utah Rkseuvoiii 19 30- ^ 20 H -I 3 10 - El KAMLOOP A— TEN SLEEP ---«-- SHEPHERD T T ? MJJASONDJFMAMJJASOND YEAR 1 YEAR 2 J F M A M J J YEAR 3 TIME IN RESERVOIR Fig. 4. Cumulative monthly increase in the percent of fish captured from 3 strains of trout stocked in 1985. The creel sun'ey was conducted for only 12 mon, but data were expanded to cover a longer period by using information on other cohorts (see te.\t). 1985 (Schrader 1988). Consequently, in our study and in many others (see Babey and Berry 1989) that have investigated the importance of fish strains, results are confounded because strain size and condition were not carefully controlled, and because the harvest of fishes was not measured over their entire life span. The fish yield of 102 kg/ha in East Canyon Reservoir is among the highest yet reported for a temperate zone lake (Morgan et al. 1980, Jones and Hoyer 1982, Schlesinger and Regier 1982) and is as high as yields in many tropical systems (Morgan et al. 1980). It is also high in relation to clilorophyll levels in the lake. Regres- sions with summer chlorophyll levels would predict yields ranging from 4 to 13 kg/ha, de- pending on the model chosen (Ogelsby 1977, Jones and Hoyer 1982; see Carline 1986). A model based on total phosphoiiis would predict salmonid production of only 22 kg/h (Plante and Downing 1993), so the realized yield of 102 kg/ha is far above expectations (Downing and Plante 1993). Even when the weight of fish stocked (5 kg/ha) is subtracted from total yield, hai-vest from this cold-water reservoir is still remarkably high. Several characteristics of the reservoir and fishery may contribute to the high yield. First, high nutrient loading (Merritt et al. 1980) pro- duces high algal productivity that in turn sup- ports a large zooplankton population domi- nated by Daphnia (this, however, does not explain why fish production is higher than that predicted by chlorophyll or phosphorus levels). Second, rainbow trout in East Canyon Reser- voir are primarily first-order carnivores, feed- ing throughout most of their lives on large Daphnia spp. (Tabor et al. in press). They begin feeding on other fish only when they exceed about 370 mm total length (Wurtsbaugh 1987). Third, the management agency takes full ad- vantage of high productivity by stocking large numbers of fish. Fourth, with intense fishing pressure, most of the trout are hai-vested thor- oughly and quickly while they are growing rapidly (Fig. 1). The combined effects of high reservoir productivity, high stocking density, trout feeding close to the base of the food web, and intensive fishing pressure contribute to the very high fish yield. Although East Canyon Reservoir has pro- vided exceptional trout yields, there are indi- cations that high nutrient loading from resi- dential and recreational development in the headwaters of the drainage may be pushing the fisheiy toward collapse. Because the reser- voir is already mesoeutrophic, increased pro- ductivity resulting from development may fur- ther deplete oxygen in the hypolimnion and metalimnion. Oxygen and temperature pro- files we took in July and August 1985 and 1986 demonstrated that water with O2 con- centrations >5 /xg/L was found only at depths above 10 m where temperatures were above 20 Great Basin Naturalist [Volume 56 18° C. Summer metalimnetic and hypolimnetic oxygen concentrations in 1985 and 1986 were much lower than reported for the reservoir during 1978-1980 (Merritt et al. 1980, Utah Department of Health 1982). When oxygen is lost from these layers, fish are forced into the warm epilimnetic water. Because optimal tem- peratures for rainbow trout are near 15-18° C (Hokanson et al. 1977, Wurtsbaugh and Davis 1977), and because O2 concentrations for salmonids should be at or above 5 /Ltg/L (Brett 1979, EPA 1986), the situation in East Canyon Reservoir may become too stressful for rain- bow trout, and they may be squeezed into a narrow metalimnion where conditions are sub- optimal. Indications that trout are stressed include poor growth in midsummer (Fig. 1; Babey and Beny 1989), increases in Lernaea infestation from 20/fish in the 1970s to 40/fish in the late 1980s (T. Pettengill unpublished data), and complete failure of the 1989 and 1991 year-classes subsequent to our field study. Loss of salmonid fisheries with increasing eutrophication is common (Colby et al. 1972). Consequently, urban planners and fisheiy man- agers should limit reservoir nutrient loading to maintain adequate summer oxygen levels and thus ensure that the outstanding family fisheiy for salmonids in the resei'voir is maintained. Acknowledgments We thank D. Neverman and K. Marine for assistance in the field, and G. Blommer for helping in the field and in revising the WCREEL FORTRAN program. Danen Brandt assisted in data analysis and preparation of fig- ures. C. Beny encouraged the stud\' and pro- vided \'aluable criti(|ues of the manuscript. D. Hepworth, R. Tabor, D. Archer, R. Whaley, and two anonymous reviewers provided valu- able comments on drafts of the manuscript. D. Pitman and D. Andriano carried out the origi- nal creel studies in 1970 and 1972. The study was supported by the Utah Division of Wildlife Resources with Federal Sport Fish Restoration funds (F47-R) and was administered bv the USFWS C:()operative iMsh and Wildlife Re- search Unit at Utah State University. Literature Cited B.\l!l';v, {;. J., AND C. R. Bkhuv. 19.S9. Posl-stocking pltIoi- niance of three strains of rainbow trout in a reser- voir. North American Journal of Fisheries Manage- ment 9: 309-315. Berry, C. R. Jr., G. J. Babey, and T. Schrader. 199L Effect of Lernaea eijprinacea (Crustacea: Copepoda) on stoclced rainbow trout (Oncorhijnchus intjkiss). Journal of Wildlife Diseases 27: 206-213. Br.\lh\, J. L., AND H. L. Kincaid. 1982. Sunival, growth, and catchability of rainbow trout of four strains. North American Journal of Fisheries Management 2: 1-10. Brett, J. R. 1979. Environmental factors and growth. Pages 599-675 in W. S. Hoar, D. J. Randall, and J. R. Brett, editors. Fish physiology. Volume III, Bioener- getics and growth. Academic Press, NY. Carline, R. F 1986. Indices as predictors offish commu- nity' traits. Pages 46-56 ('» G. E. Hall and M. J. Van Den Avyle, editors, Resenoir fisheries management: strategies for the 80s. American Fisheries Society; Bethesda, MD. Carlton, F E. 1975. Optimum sustainable yield as a man- agement concept in recreational fisheries. American Fisheries Society Special Publication 9: 45—49. Colby, R J., G. R. Spangler, D. A. Hurley, and A. M. McCoMBlE. 1972. Effects of eutrophication on salmonid communities in oligotrophic lakes. Journal of the Fisheries Research Board of Canada 29: 97.5-983. Dow NiNG, J. A., and C. Pl.\nte. 1993. Production of fish populations in lakes. Canadian Journal of Fisheries and Aquatic Sciences 50: 110-120. EPA (U.S. Environmental Protection Agency). 1986. Quality' criteria for water. EPA report 440/5-86-001. Washington, DC. Hall, G. E., and M. J. Van Den A\tle. 1986. Resei-voir fisheries management: strategies for the 80s. Ameri- can Fisheries Society, Bethesda, MD. 327 pp. Hok.\nson, K. E. E, C. E Kleiner, and T. W. Thors- LU.ND. 1977. Effects of constant temperatures and diel temperature fluctuations on specific growth and mortality rates and yield of juvenile rainbow trout, Salmo gairdneri. Journal of the Fisheries Research Board of Canada 34: 639-648. Jones, J. R., and M. V. Hoyer. 1982. Sportfish harvest predicted by summer chlorophyll-« concentration in midwestern lakes and resenoirs. Transactions of the American Fisheries Society 111: 176-179. Mal\'ESTUTO, S. P 1983. Sampling the recreational fishen. Pages 397-419 in L. A. Nielsen and D. L. Johnson, editors, Fisheiy techniques. American Fisheries Society, Bethesda, MD. M VI'KOWSKI, S. M. D. 1989. Differential susceptibilitx of three species of stocked trout to bird predation. North American Journal of Fisheries Management 9: 184-187. MiTuuiT, L. B., A. W Miller, R. N. Winget, S. R. Rush- forth, AND W. H. Brimhall. 1980. East Canyon Resen'oir water quality assessment. Mountainland Association of Governments, Provo, UT. 193 pp. Morgan, N. C, et al. 1980. Secondary production. Pages 247-340 in E. D. LeCren and R. H. Lowe-McConnell, editors. The functioning of freshwater ecosystems. Cambridge University Pres.s, London. Ogelsry, R. T. 1977. Relationships of fish \ield to lake pin toplankton standing crop, production and moipho- edaphic factors. Journal of the Fisheries Research Board of Canada .34: 2271-2279. 1996] Trout Yield in Utah Rkservoir 21 I'liiNNEY, E. E., D. M. MiiJj'.K, AM) M. L. Daiilberg. 1967. Mass-marking young salnionids with fluores- cent pigment. Transactions oi tlie Ameriian ImsIi- eries Society 96: 157-162. Plante, C, and J. A. Downing. 1993. Relationship of salmonine production to lake trophic status and tem- perature. Canadian Journal of Fisheries and Aquatic Science 50: 1324-1328. SCHLESINGER, D. A., AND H. A. Regier. 1982. climatic and morphoedaphic indices of fish yields from nat- ural lakes. Transactions ol tlic American Fisheries Society 111: 141-150. Schrader, T. M. 1988. Performance of three strains of rainbow trout in East Canyon Reservoir. Unpub- lished master s thesis, Utah State University, Logan. Stockner, J. G. 1992. Lake fertihzation: the enrichment cycle and lake sockeye salmon {Oncorhijnchus iwrka) production. Pages 199-214 in H. D. Smith, L. Mar- golis, and C. C. Wood, editors, Sockeye salmon [Oiico- rJiynchus nerha) population biology and fiiture man- agement. Canadian Special Pubhcation in Fisheries and Aquatic Science 96: 198-215. Tabor, R. A., and W. A. Wurtsbaugh. 1991. Predation risk and the importance of cover for juvenile rainbow trout in lentic systems. Transactions of the American Fisheries Society 120: 728-738. Tabor, R. A., C. Luecke, and W. A. Wurtsb.wgh. In press. Effects of Daphnia availability on the growth and food consumption rates of rainbow trout in two Utah Resei^voirs. Transactions of the American Fish- eries Society. T.WLOR, M. J., and K. R. White. 1992. A meta-anahsis of hooking mortality of nonanadromous trout. North American Journal of Fisheries Management 12: 760-767. Trend.all, J. 1988. Recruitment of juvenile mbuna (Pisces: Cichlidae) to experimental rock shelters in Lake Malawi, Africa. Em ironmental Biology of iMshes 22: 117-132. Utah Department of Health. 1982. State of Utah clean lakes inventory and classification, volume 1. Depart- ment of Health, Salt Lake City, UT. 519 pp. Vondracek, B., W. Wurisbaugh, and J. Cecu. 1980. Mass marking of Gamlmsia. California Mostjuito and Vector Control Association 48: 42—44. Wasowicz, a. F 1991. Influence offish and avian preda- tors upon the trout population of Minersville Reser- voir. Unpublished masters thesis, Utah State Univer- sity, Logan. Wurtsbaugh, W. A. 1987. Importance of predation by adult trout on mortality rates of fingerling rainbow trout stocked in East Canyon Resei^voir, Utah. Pages 14-17 in Proceedings of the Bonneville Chapter of the American Fisheries Society, Salt Lake City, UT, 1987. . 1988. Iron, molybdenum and phosphonis limita- tion of N, fi.xation maintains nitrogen deficiency of plankton in the Great Salt Lake drainage (Utah, USA). Verhandlungen Internationale Vereinigung fur Theoretische unci Angewandte Limnologie 23: 121-130. Wurtsbaugh, W A., and C. E. Davis. 1977. Effects of temperature and ration level on the growth and food con\'ersion efficiency of Sahno gairdneri, Richard- son. Journal of Fish Biologv' 11: 87-98. Wurtsbaugh, W. A., R. W Brocksen, and C. R. Gold- man. 1975. Food and distribution of underyearling brook and rainbow trout in Castle Lake, California. Transactions of the American Fisheries Society 104: 88-95. Received 26 May 1995 Accepted 18 September 1995 Great Basin Naturalist 56(1), © 1996, pp. 22-27 CONSUMPTION OF DIFFUSE KNAPWEED BY TWO SPECIES OF POLYPHAGOUS CRASSHOPPERS (ORTHOPTERA: ACRIDIDAE] IN SOUTHERN IDAHO Dennis J. Fielding^-, M. A. Brusvenl, and L. P Kish^ Abstract. — Consumption of diffuse knapweed {Centaurea diffusa Lam.) by 2 polyphagous grasshopper species, Melanoplus sanguinipes (E) and Oedaleonotus enigma (Scudder), was studied using microhistological analysis of grasshopper crop contents. Grasshoppers were confined to cages containing C. diffusa and Sisymbrium altissimum L., a member of the mustard family known to be readily eaten by these 2 grasshopper species. Preference indices for knap- weed were lower than for S. altissimum in 4 of 5 trials. An uncaged population of A/, sanguinipes on a knapweed-infested site consumed only small amounts of knapweed until late summer when most other plants were senescent. Results sug- gest that diffuse knapweed's low palatability to generalist herbivores may confer to it a competitive advantage over other rangeland plants. Key words: Centaurea diffusa Lam., diffuse knapweed, herbivory, insects, competition. Diffuse and spotted knapweed, Centaurea diffusa Lam. and C. maculosa Lam., respec- tively, were introduced to the Pacific North- west around 1900 (Watson and Renney 1974). Since then they have rapidly spread through- out the area (Fig. 1; Forcella and Hai-vey 1981). Heavy infestations of knapweed reduce produc- tion of more desirable species of forage plants, thus reducing the value of rangeland for graz- ing and wildlife habitat. Several specialist insect herbivores have been introduced in attempts to control knapweed (Story and Ander- son 1978, Maddox 1979). To date, no studies have reported on the consumption of knapweed by polyphagous insect herbivores. Cnicin, a sesquiteipene lactone, is produced by spotted and diffuse knapweed (Drodz 1966, Locken and Kelsey 1987). Pieman (1986) sug- gested that sesquiterpene lactones have toxic effects on many herbivores and may function as deterrents to herbivoiy Locken and Kelsey (1987) suggested that nonpalatability of knap- weeds may afford them a competitive advantage over many other plant species by protecting them from herbivoiy. (grasshoppers (Orthoptera: Acrididae) are a conspicuous and important class of herbivores on rangeland in the \\'est- em U.S. Rangeland grasshopper populations in south- ern Idaho occasionally reach outbreak propor- tions. Two species in particular, Melanoplus sanguinipes (F). and Oedaleonotus enigma (Scudder), are capable of attaining very high densities (>30/m-). Both species feed upon a broad range of forbs (Brusven and Lamley 1971, Banfill and Brusven 1973, Sheldon and Rogers 1978). Pfadt (1992) suggested that an increase in introduced weeds is a factor lead- ing to outbreaks of O. enigma. Fielding and Brusven (1993) found that both species prefer disturbed rangeland habitats dominated by e.xotic annual plants. This study assessed the utilization of diffuse knapweed as food by these 2 grasshopper species to determine if knap- weed represents a significant and expanding resource for grasshoppers and if grasshopper herbivor)' may be a constraint to knapweed populations. Previous studies (Brusven and Lamle\ 1971) have shown Sisymbrium altissimmn L., an intro- duced annual forb, to be preferred by many forb-feeding grasshoppers. Both species of weeds initiate growth as a basal rosette of leaves and later develop erect, sparsely leaved stems that bear flowers. Because C. diffusa is usually a biennial, it does not develop beyond the basal rosette until the 2nd year Sisymbrium alfissi}num constituted a large proportion of the forbs present in this stud)'; therefore utili- zation of C. diffusa and S. altissinuim was compared. ' l)c|>,irliMciit of Plant, Sciil, and luiloniolDjiual Scicntos, Univcrsit) oC Ulalm, Moscow. W 83844-2339. -Present addtcss; PO Box 75010. Uni\eisit\<)l Alaska, Fairbanks, AK 9MT7.">0102. 22 1996] Knapweed Consumption by Crassiioim'krs 23 Fig. 1. Idaho counties reporting infestations of dirfuse knapweed, Centuurea diffusa. Materials and Methods The study site is about 3 km south of Sho- shone, Idaho (Lincoln County), in a knapweed- infested area that had been seeded with crested wheatgrass {Agropyron ciistatum [L.] Gaertn.) in 1975. Grasshopper food preferences were identified by microhistological analysis of grass- hopper crop contents (Brusven and Mulkern 1960, Sparks and Malechek 1968, Fielding and Bnisven 1992). Grasshoppers were confined to cages so that relative amounts of different plant species could be precisely determined. Five trials were conducted during the summer of 1989: O. enigma 4th- and 5th-instar nymphs in early June; O. enigma adults and M. san- guinipes 4th and 5th instars in late June; M. sanguinipes adults in July and again in August. For each trial, 4 wire-mesh (5-mm pore size), conical cages covering 0.5 m^ each were placed in the field such that at least 1 plant each of C. dijfiisa and Sisymbrium altissimiim L., along with assorted common grasses, occurred within each cage. Twelve to 15 grasshoppers of a sin- gle species were placed in each cage. Grass- hoppers used in the tests were collected from rangeland and placed in the cages within 20 h of collection. A 4-d interval was estimated to be sufficient to completely void previous meals and to accurately assess preferences in choice tests. After 4 d, 10 grasshoppers were removed from each cage and immediately preserved in 95% ethanol for crop analysis. Species composition of plants in each cage was determined on an air-dry basis by clipping and sorting by species aboveground portions of plants in each cage after each trial. Clipped plants were stored in air-tight plastic bags, and fresh weight was obtained within 4 h of clip- ping. Clipped plants were then allowed to air- dry until they quit losing weight (10-15 d), after which dry weights were obtained (to the nearest 0.1 g). Percent moisture of above- ground portions of each plant species was then determined. Plants were rated after each trial according to phenology as follows: 1, vegetative growth only; 2, flowering; 3, seed set; 4, seed maturity; 5, senescent or dormant (USDA-Soil Conser- vation Service 1976). Grass species present included Poa sandbergii Vasey, Bromus tecto- riim L., and Agropyron cristatum. Centaurea dijfusa and S. altissimum composed about 97% of aboveground biomass of forbs. Both 1st- and 2nd-year C. diffusa were present in each of the cages. Other forbs present were Helian- thus annuus L., Lactuca serriola L., and Epilo- bium L. sp. Grasshopper crops were removed and the contents mounted on glass slides in glycerin and safranin stain. Plant fragments in the crops were identified by comparing them with refer- ence slides made from fragments of known plants collected at the study site, similar to the methods described by Fielding and Brusven (1992). Frequency counts were made for each plant species by determining their presence or absence in 20 microscope viewing fields per grasshopper crop. Trichomes, hairs, and pollen were not counted. Frequencies from the 10 grasshoppers per cage were summed. Relative frequency was calculated by dividing the fre- quency of a plant species by the total fre- quency of all plant species (Sparks and Malechek 1968, Pfadt and Lavigne 1982). Holecheck and Gross (1982) demonstrated the 24 Great Basin Naturalist [Volume 56 near equivalence of relative frequency to actual diy weight percentage of plants consumed. Relative availability of different plant species within an area has been shown to influence diet composition in many grasshopper species (Ueckert et al. 1972, Mitchell 1975). To account for the effect of availability on consumption, preference values for plant species constitut- ing more than 10% of either cage or crop con- tents were calculated by dividing relative fre- quency of a plant species in the crops by that species' percentage of the dry-weight of all plants within the cages (Ueckert and Hansen 1971). A preference value >1 indicates feed- ing in greater proportion to the plant's avail- ability, whereas a preference value <1 indi- cates low preference in relation to a plant's availability. Possibly, total diy weight of a plant may not accurately portray the amount of plant mater- ial available to grasshoppers, thus introducing bias into the preference values. In this study our obsei"vations indicated that both species of weeds had similar ratios of leaves to stems. Also, we have obsei^ved grasshoppers feeding on stems of both weed species. Because we had no way to determine more precisely exactly what proportion of the plant was avail- able as food to the grasshoppers, we used total aboveground biomass as a reasonably objec- tive measure of availability. The presence of Ist-year rosettes of C dijfusa in the cages ensured that each replication included a rep- resentative choice of plant material. Differences bet\\'een plant species in rela- tive frequency and preference values were tested using the Wilcoxon 2-sample test (PROC NPARIWAY, SAS 1985), with each cage rep- resenting 1 replication. Comparisons between plant species were made for each trial of a sin- gle grasshopper species and with data from different trials pooled by grasshopper species. The same statistical methods were also used to test for differences in relative frequency and preference values between grasshopper species for C. diffusa and S. (dtissimurn. Food selection was monitored in an uncaged population of M. sanguiwipes near the cage study. Thirty to 50 individuals were collected on each of 5 dates from June through October from an area of ca 1 ha infested with knap- weed. Food preference in this population was determined by microhistological methods de- scribed above. Plant species composition at the site was determined by visual estimates, in 5% incre- ments, of the ground cover of each plant species in forty 0.1-m- quadrats, arranged in 4 transects of 10 quadrats each. Ground cover estimates were made in July and again in October after precipitation caused abundant germination of cheatgrass. Because accurate estimates of food axailability (biomass) in the field were not available, preference values were not calculated and the results are presented for comparative puiposes only. Results Cages were placed such that C. diffusa was equally as abundant as or more abundant than S. altissimum in each trial (Table 1). Percentage moisture of both species of weeds declined throughout the season (Table 1). Sisijmhnum altissimum tended to be slightly more advanced phenologically than C. diffusa throughout the season, partly due to the presence of Ist-year rosettes of C. diffusa in the cages, but also because of earlier flowering bv S. altissimum (Tlible 1). Although C. diffusa constituted a substan- tial percentage (10-46%) of the caged grass- hoppers diet, preference values for C. diffusa were <1 in eveiy trial, indicating that it was not consumed in proportion to its diy weight composition within the cages (Table 1). Prefer- ences values for S. altissimum were > 1 in each trial, indicating that it was consumed in pro- portions greater than its relative availability. After flowering in Jul\', a large portion of the C. diffusa plant material in the crops of i\/. sanguinipes consisted of floral parts (44% and 30% of the C diffusa material consumed, in the July and August trials, respectively). Other forbs represented in in situ caged trials were not present in sufficient quantity to ade- (|uately assess their preference values. More S. altissimum than C. diffusa was con- sumed b>' grasshoppers in 3 of the 5 trials (Table 1). Preference values for S. altissimum were greater than those for C. diffusa in 4 of the trials (Table 1). Combining data from the 3 trials with M. sanguinipes, crop contents and preference values for S. altissimum, 42% and 2.0, respectively, were greater than for C. dif- fusa, 16% and 0.5, respecti\'el\' (Wilcoxon test, P < 0.01 for both tests). For O. enigma, the overall preference value for S. altissimum, 3.5, 1996] Knapweed Consumption by Grasshoppers 25 Table I. Relatixe availaliilit\' and consumption h\' grasshoppers of plant species. Plant PerccTit Mean Relative Relative Mean pheuologic; il moisture clr\ weigh t ; ixailability frequency preference Plant species stage' ol plants in cages in cages^ in crops index 4th- and 5th-instar Oedaleonotus enigma nymphs on 6 June 1989 Ccntaurca dijfusu 1 77 13.4 23 lOa'5 0.38a Sisyinhriuin altissiiiiuiii 1 81 8.8 15 48b 5.06b Other ibrbs 1 85 0.6 1 <1 — A^roptjroit cristatiiin 1-2 57 18.1 31 <1 <0.05 Poo sandbi'i-fiii 4 24 4.7 8 2 — Broiuus tcctoriiin 4 21 12.8 22 32 1.42 Detritus 6 adult Oedaleonotufi 1 enigma on 26 June 1989 Ccntaurca diffusa 1 64 59.0 61 46a 0.76a Sisymbrium altissimum 1-2 67 27.1 28 48a 1.9.3a Other forbs 1-2 79 1.0 1 1 — Agropyron cristatum 3-4 45 6.8 7 0 — Poa sandbergii 5 15 0 0 0 — Bromus tcctoruin 5 12 3.6 3 1 — Detritus 4 — 4th- and 5th -instar Melanoplus sanguinipes nymph s on 26 June 1989 Ccntaurca diffusa 1 64 60.8 59 16a 0.25a Sisymbrium altissimum 1-2 67 25.8 25 74b 3.00b Other forbs 1-2 79 2.1 2 0 — Agropyron cristatum 3-4 45 6.2 6 0 — Poa sandbergii 5 15 4.1 4 5 — Bromus tectorum 5 12 3.1 3 4 — Detritus 2 — adult Melanoplus sanguinipes on 21 July 1989 Ccntaurca diffusa 1-2 63 25.4 29 16a 0.56a Sisymbrium altissimum 2-3 55 24.5 28 44b l.,55b Other forbs 1-2 75 8.8 1 3 — Agropyron cristatum 4 45 7.0 8 4 — Poa sandbergii 5 9 17.5 20 3 0.17 Bromus tectorum 5 14 11.4 13 27 2.52 Detritus 3 — ad ult Melanoplus sanguinipes on 25 August 1989 Centaurea diffusa 1,3-4 22 38.8 38 23a 0.70a Sisymbrium altissimum 4-5 11 18.4 18 24a 1.48b Other forbs 2-3 65 8.2 8 5 — Agropyron cristatum 4 18 15.3 15 3 0.37 Poa sandbergii 5 8 4.1 4 5 — Bromus tectorum 5 7 13.3 13 25 2.83 Detritus 16 — 'l. vegetative giowtli uiiK; 2, fluwfriiig, .3, seed set; 4, seed ni;iturit\'; 5, seneseent ur donnant -Mean (JV = 4) percentage ot aboveground plant biomass (air-dn' basis) within cages •'Means for C. diffusa and S. altissimum within columns of each trial followed by different letters are significantly different, P < 0.0.5, Wilco.xon 2-saniple test. was greater than for C. diffusa, 0.6 (Wilcoxon test, P < 0.05). There was no difference in con- sumption by O. enigma between S. altissimum and C. diffusa, 48% and 27%, respectively (Wilcoxon test, P > 0.05). There were no dif- ferences between the 2 species of grasshop- pers in relative frequency or preference values for either S. altissimum or C. diffusa (Wilcoxon test, F > 0.10 for both comparisons). Of the grass species, only Bromus tectorum was eaten in greater proportion than its per- centage of air-dry biomass. Even though O. enigina is generally considered to be a forb- feeder (Sheldon and Rogers 1978, Pfadt 1992), B. tectorum constituted 32% of the diet of O. e7iigma in early June (Table 1). Adult O. enigma in late June ate very little B. tectorum. Melanop- lus sanguinipes consumed B. tectorum through- out the summer, with 4-27% of its diet com- posed of B. tectorum, even though the grass was completely senescent by 26 June (Table 1). 26 Great Basin Naturalist [Volume 56 Table 2. Relative frequency of food items in crops of Af. sanguinipes on 5 dates and percentage ground cover in JuK and October 1989. Sisyinhriuin (iltissiinit)n Ccntaiirca diffusa Other forbs'* Bromiis tectorutn Other grasses Litter, detritus Relati\e fre(juenc\ of crop components 30 20 14 6 13 June Jul\ Aug Sep Oct 46 23 22 / 6 18 30 32 55 1 19 25 7 24 6 7 9 12 6 76 1 4 15 1 10 9 9 13 7 1 Percentage ground cover July 2 6 <1 October 1 4 <1 16 6 ••Iiicludes rahl)itlirusli iChnisotluimiis rwti'irosiis [Rill] Biitt.l, lupine iLiipiiais L spl, am! suiifloufr Hh'lianthus (iiuuiiis L.l. Knapweed was the most common forb grow- ing on the site where the uncaged population of M. sanguinipes was studied (Table 2). In June, S. oltissimiDn was the largest single food item, but consumption declined as the season pro- gressed. Knapweed was a substantial food item, especially in August and September when it remained succulent after other forbs had dried. After rainfall stimulated germination of B. tec- toriim in late September and October (Table 2), it became the primaiy food item for M. san- guinipes, and forbs constituted onb' a minor portion of the diet. Discussion The evolutionar)' histoiy of an herbivorous species, by shaping its food habits and other life history traits, determines its present rela- tionships with exotic plant species. The 2 grass- hopper species in this study consume a wide variety of plants, especialK' forbs (Banfill and Brusven 1973, Sheldon and Rogers 1978, Pfadt 1992), and will readily accept exotic plant species. Melunophis sangninipcs is a veiy oppor- tiniistic feeder Egg hatch in this species is often spread out over a long period, resulting in a large proportion of a population maturing dur- ing the diy periods typical of late summer in the intermoimtain region. At such times man\ late-maturing plants that still retain some suc- culence, such as rabbitbrush, sagebrush, and some lupine species, are primary food items for M. sanguinipes. The results of this stud> indicate that this was the case with C. diffusa: even though it was not highly preferred 1)\ M. sanguinipes, it was a major food item in late summer when most other plants were dry. SisyniJjriuni alfissiinuin tended to become sene- scent earlier than C. dijfusa, which would re- duce the qualit)' of S. altissimum relative to C. dijfusa, especially when Ist-year rosettes, con- sisting mostly of leaves, are considered. Locken and Kelsey (1987) reported that cnicin concentrations in C. maculosa xaiy con- siderably within and among indi\ idual knap- weed plants. Cnicin is stored within glandular trichomes on the surface of knapweed tissues. Highest concentrations of cnicin were found in leaves surrounding the inflorescence. Only trace quantities were found by Locken and Kelsc)' (1987) in the stem epidermis and flow- ers. Leaf concentrations were lowest in spring and increased with flo\\ ering. We assume that cnicin concentrations in C. dijfusa follow much the same pattern. Variability in cnicin concen- tration may result in selective consumption by grasshoppers of knajDweed tissues w ith low cnicin concentrations. Our results suggest that this is the case: In late-summer trials much of the knapweed tissue consimied by grasshoppers consisted of flo\\'ers. This implies that during years of high grasshopper densities, feeding by grasshoppers, especialK' on the flowers, could result in a modest reduction in seed production in this plant. Residts of this stud)' pro\'ide support for the h> pothesis that knapweed is protected from herbivoiy by its chemical constituents (Pieman 1986, Locken and Kelsey 1987). When com- pared to S'. altissimum, diffuse knapweed was a 2nd-choice food item for these generalist grasshopper species. Its low palatability may confer a competitive acKantage to knapweed when herl)i\on' is a strong selection factor. .\lthough it is conceivable fliat at high densities grasshoppers may consume significant amounts of knapweed and reduce seed production, man> other plants would be affected to a greater degree, thus reducing competition to knapweed. 1996] Knapweed Consumption by Giusshoppers 27 Grasshopper species used in this trial are the dominant species contributing to outbreaks in soutliern Idaho. It appears that increasing knapweed infestations do not represent a sig- nificant increase in food resources for these grasshoppers. However, because knapweed stays green longer during the summer than many other rangeknid phmts, it may provide sustenance for polyphagous grasshoppers dur- ing kvte-summer droughts in southern Idaho. Acknowledgments The authors thank the staff of the Bureau of Land Management s Shoshone District office for technical and logistic support. Russell Biggam assisted with field studies. Bahman Shafii advised on statistical matters. Robert H. Callihan, Dave Koehler, and Don Hostetter provided helpful comments on earlier versions of the manuscript. This study was supported in part by the Bureau of Land Management as Cooperative Agreement ID 910-CA7-05. It is published with the approval of the director of the Idaho Agricultural Experiment Station as paper 94724. Literature Cited Banfill, J. C, AND M. A. Brusven. 1973. Food habits and ecology of grasshoppers in the Seven Devils Moun- tains and Sahnon River breaks of Idaho. Melanderia 12; 1-21, Brusven, M. A., and J. D. Kwiley. 1971. The food habits and ecology of grasshoppers from southern Idaho rangeland. University of Idaho in cooperation with USDA, Project Completion Report 1967-1971. Brusven, M. A., and G. B. Mulkern. 1960. The use of epidermal characteristics for the identification of plans recovered in fragmentary condition from the crops of grasshoppers. North Dakota Agricultural Experiment Station Research Report .3. 11 pp. Drodz, B. 1966. Isolation of cnicin from the herbs of Ccn- faiirea dijfusa Lam. Dissertationes Pharmaceuticae et Pharmacologicae 18: 281-283. Fielding, D. J., and M. A. Brusven. 1992. Food and habi- tat preferences of Melanophi^ sanguinipes and Aiilo- cara elliotti (Orthoptera: Acrididae) on disturbed rangeland in southern Idaho. Journal of Economic Entomologx' 85: 78.3-788. . 1993. Grasshopper (Orthoptera: Acrididae) com- mimity composition and ecological disturbance on southern Idaho rangeland. Environmental Entomol- ogy 22: 71-81. FoRCELLA, E, and S. J. Hahvey. 1981. New and exotic weeds of Montana. Volume II: Migration and distri- bution ol 100 alien weeds in northwestern U.S.A., 1880-1980. Montana Department of Agriculture, Helena. Holecheck, J. L., and B. D. Gross. 1982. Evaluation of diet calculation procedures for microhistological analysis. Joimial of Range Management 3.5: 721-723. LoCKEN, L. J., and R. G. Kelsey. 1987. Cnicin concentra- tions in Centaurea maculosa, spotted knapweed. Biochemical Systematics and Ecology 15: 313-.320. Maddo.x, D. M. 1979. The knapweeds: their economics and biological control in the western states, U.S.A. Rangelands 1: 139-141. Mitchell, J. E. 1975. Variation in food preferences of three grasshopper species (Acrididae; Orthoptera) as a function of food availability. American Midland Naturalist 94; 267-283. Pfadt, R. E. 1992. Valley grasshopper Oedaleonotm enigma species fact sheet. USDA-Animal and Plant Health Inspection Service, Wyoming Agricultural Experi- ment Station Bulletin 912. Pfadt, R. E., and R. J. Lavigne. 1982. Food habits of grasshoppers inhabiting the Pawnee site. University of Wyoming Agricultural Experiment Station Science Monograph 42. 72 pp. PiCMAN, A. K. 1986. Biological activities of sesquiterpene lactones. Biochemical Systematics and Ecology 14: 2.55-281. SAS Institute. 1985. SAS/STAT user's guide. SAS Insti- tute, Carey, NC. Sheldon, J. K., and L. E. Rogers. 1978. Grasshopper food habits within a shrub-steppe community. Oecologia 32: 8.5-92. Sparks, D. R., and J. C. Malechek. 1968. Estimating per- centage dry weights in diets using a microscopic tech- nique. Journal of Range Management 21: 264—265. Story, J. M., and N. L. Anderson. 1978. Release and establishment of Urophora ajfinis (Diptera; Trypeti- dae) on spotted knapweed in western Montana. Envi- ronmental Entomology 7: 44.5-448. Ueckert, D. N., and R. M. Hansen. 1971. Dietary over- lap of grasshoppers on sandhill rangeland in north- easteiTi Colorado. Oecologia 8; 276-295. Ueckert, D. N., R. M. Hansen, and C. Terwilliger, Jr. 1972. Influence of plant fi-equency and certain mor- phological variations on diets of rangeland grasshop- pers. Journal of Range Management 25; 61-6.5. USDA-SoiL CONXERSATION SERVICE. 1976. National range handbook. USDA-SCS, Washington, DC. Watson, A. K., and J. Renney 1974. The biolog>' of Cana- dian weeds. 6. Centaurea dijfusa and C. maculosa. Canadian Journal of Plant Science .54: 687-701. Received 11 July 1994 Accepted 2 February 1995 Great Basin Naturalist 56(1), © 1996, pp. 28-37 FIRE FREQUENCY AND THE VEGETATIVE MOSAIC OF A SPRUCE-FIR FOREST IN NORTHERN UTAH Linda Wadleighl and Michael J. Jenkins^ Abstract. — Fire scar and vegetative analysis were used to constnict a fire histoiy for the Engelmann spruce/sub- alpine fir {Picea engelmannii/Abies lasiocarpa) vegetation type of the Utah State University (USU) T. W. Daniel E.xperi- mental Forest. Three distinct periods of fire frequency were established — presettlenient (1700-1855), settlement (1856-1909), and suppression (1910-1990). Mean fire intei-val (MFI) decreased during the setdement period and greatly increased during the suppression era. The difference was attributed to the influ.x of ignition sources during the settle- ment of nearby Cache Valley, located 40 km to the west. Logging and livestock grazing appear to have led to the reduced MFI, which in turn worked as a factor to create the vegetative mosaic now obsei"ved on the study area. The increase in MFI during the suppression era permitted the advancement of shade-tolerant species in the understoiy of the shade-intolerant lodgepole pine (Pimis contoiia \ar latifoUa) and quaking aspen {Popiihis treiiuilokles). Continued suppression of disturbance fi-om wildfire will allow the lodgepole pine cover type, which experienced the lowest MFI during the settlement period, to be further invaded by shade-tolerant species, decreasing spatial stand diversity and increasing the risk of more intense fires. Key words: fire jrequeney. subalpine spruce-fir joresl. fire sear. Absence of natural fire in wilcUand ecosys- tems, due to removal of fine fuels by livestock, reduction in Native American ignitions, and a suppression policy instituted in the early 1900s has led to extensive alterations in natural vege- tative succession patterns. Human disruption of natural fire regimes in fire-dependent com- munities limited natural diversity and altered the long-term stability of fire-adapted plant species (Heinselman 1973, Gruell 1986, Agee 1993). Previously, natural ecosystems had evolved under episodic fires (Parsons 1981, Gruell 1983). Gruell's (1983) interpretation of paired photos from the Northern Rockies showed early stages of forest succession were more common from 1870 to 1940 than they are today; however, Gruell (1983) also found the absence of fire has contributed to a marked alteration of natural vegetation mosaics by favoring woody species such as shrubs and trees over grasses. Lightning-ignited fires in Engelmann spiiice/ subalpine fir {Picca cn'^chntnuui/ Abies lasio- carpa) forests are less frequent than fires in drier vegetation types. Arno (1980) estimated a fire return interval of 50 to 130 yr for spruce/ fir habitat types. Veblen et al. (1994) found a mean fire-rctnrn intenal of ca 200 \'r in a Kock> Mountain subalpine forest in northwestein Colorado. In these subalpine fir forests, historic fire allowed the dominance of serai species and created a mosaic of species and age com- positions. Where serai species such as lodge- pole pine {Piniis contorta) or aspen {Populus tremidoides) occurred, a higher fire frequency favored their dominance (Bradley et al. 1992). In the lodgepole pine-dominated communi- ties that occur in the lower portion of the sub- alpine fir forest, fire was more frequent with intensit)' depending on amount of precipitation received in the summer months. Abundant evi- dence was found in the lodgepole pine forests of northern Utah of nondestmctive groimd fires, more intense "thinning fires, "stand-replacing fires, and severe double bums" (Arno 1980). Fire histoiy studies provide land managers with estimates of past fire fi-equencies, mean fire-return intenals, and effects of natural fire on stand composition and structure (Arno and Sneck 1977). Such studies help to determine the return inten al of fires on a site, intensit)' and size of fire, effects of past fire on stand dviiamics, and effects of an era of modem sup- pression. Managers may also use the natund fire cycle or regime of an area to determine if the present disturbance regime is within the histor- ical range ol \ariatiou. A variet\ ol techni({ues are used to exaluate fire historx; including 'USDA Fmcsl Service, Onden, IJT ^DepartTiieiit of l-oresl Resotirees, Utah State Univcisit\', Ix)Kaii, UT 84322-5215. Addiess correspoiicleiiee to this autlior. 28 1996] Spkuce-Fih Fire Frequency 29 mapping stand types, correlating tire dates from fire-scarred trees to establisli a fire chronolog)', and determining age-class distri- butions, using increment cores to establish the extent of fires (Arno 1980, T^mde 1979). The objective of this study was to deter- mine il the existing vegetative mosaic of the T. W. Daniel Experimental Forest is correlated with the fire histon' of the study area, primarily, whether fire frequency has changed between 3 distinct periods; presettlement, settlement, and suppression. Additionally, if fire frequency has changed, is that change reflected in the vegetation structure visible today. Study Area The USU T W. Daniel Experimental Forest, located about 40 km east of Logan, Utah, is 1036 ha in area and ranges in elevation from 2377 m to 2651 m (Fig. 1). Topography ranges from higher plateaus dissected by deep drainages to gentle slopes and small meadows. No permanent lakes or streams are wdthin the study area (Schimpf et al. 1980); however, intermittent streams do cany runoff from the site. Winters are cold and wet, and summers are warm and dry. Mean annual precipitation is 104 cm per yr, mostly falling as snow (Hart and Lomas 1979). The major vegetation component is the Engelmann spruce/subalpine fir type in late successional stages, with serai lodgepole pine {Piniis contorta van latifolia) and ciuaking aspen stands, and small meadows distributed througli- out. A young conifer understory consisting pri- marih' of subalpine fir is often present in the aspen stands (Schimpf et al. 1980). Methods Fourteen sampling transects were estab- lished along contours spaced 61 m apart based on slope distance. A continuous log of forest cover type, the predominant vegetative type, was kept along each transect to create a stand map. As the contour intervals were traversed, trees with fire scars were identified and re- corded. The number of fire scars was recorded for each "catface" — an open scar resulting from lire damage. Fire scars are fonned when flames near the trunk raise the temperature of the cambium to a lethal level, or actually consume bark, phloem, and xylem (McBride 1983). Trees with the largest number of sound scars were marked for further studv. Great Salt Lake Fig. 1. Map showing approximate location of the T. W. Daniel Experimental Forest between Cache and Rich counties northeast of Logan, Utah. 30 Great Basin Naturalist [Volume 56 Sixty-two trees with the greatest number of visible, individual fire scars were sampled by taking a partial cross section from the pith to 1 side of the catface (Arno and Sneck 1977). The wedges were sanded and annual growth rings counted, recording the number of years back to each fire and the number between fires. Trees may be scan'ed in a number of ways in- cluding mechanical damage by nearby falling trees, root rot infection, lightning, or strip attacks by mountain pine beetle {Dendroctoniis ponderosae Hopkins, Coleoptera: Scolytidae) (Johnson and Gutsell 1994); however, there were no blue stains, lai^val galleries, or beetle emergence holes in the scars sampled (Stuart et al. 1983), which would suggest they had resulted from causes other than fire. Because pockets of obscured rings or rot may also cause inaccurate counts, tree records were combined into a master fire chronology (Arno and Sneck 1977). Individual tree ring counts were arranged horizontally on paper, geographically ordered so that neighboring trees were adjacent. Ten- year increments were placed on the left verti- cal axis, beginning with the sample year at the top, and the oldest ring year recorded at the bottom. The number of trees scarred in a year was compared to the number of trees suscep- tible to scarring. If a tree was consistently out of order, a number of years was added or sub- tracted to bring it into alignment (Arno and Sneck 1977). The maximum number of years added or subtracted equaled 3; and 16 trees were adjusted. Variable-radius plots were laid out along the sampling transects at a spacing of 200 m. Tree species present were recorded to determine cover type, and a site tree — a dominant or codominant tree on the plot — ^was aged for each species. Increment cores were taken at breast height for each site tree and were adjusted for total age for each species. A 74()th-ha regener- ation plot was recorded, tallying seedlings and saplings by species and diameter, at the center of the variable plot to aid in detennining suc- cessional patterns. Cover type, dated scars, and stand age data collected from these plots were incorporated into a stand map to show the extent of stands that might have resulted from a fire distiu- bance (McBride 1983). The stand map was supplemented by remotely sensed satellite imagery obtained in 1986. Fire frequency, "the number of fires per unit of time ' (Romme 1980), on an area was calculated for 3 fire frequency periods to por- tray the effects of settlement, logging, grazing, and modem fire suppression on the fire regime. Mean fire intei^vals, "an arithmetic average of all fire intervals determined in a designated area' (Romme 1980), were calculated for each period. Determining mean fire intei^vals for distinct land-use periods is useful in under- standing human impact on forest ecology and fire histoiy (McBride 1983). The periods were "suppression" (1910-1990), when U.S. Forest Sei'vice fire suppression was initiated, "settle- ment era" (1856-1910), and "presettlement" (prior to 1856). Mormon pioneers established the first settlement by Europeans in the Cache Valley in 1856 (Bird 1964). The presettlement period began the year just prior to the age of the oldest tree sampled — 1700 (Romme and Despain 1989). A fire history is limited by longevity of trees on the site and durabilit\' of wood exposed when scarred (Heinselmann 1973). Total number of years in each period was then divided by the number of fires in that period to obtain mean fire interval. Docu- mented evidence of historical fires was used to verify dates in the settlement-era and fire-sup- pression periods (Bird 1964). A master fire chronolog)' was developed for each stand experiencing fire in the study area as indicated by scars and the presence of even-aged stands of lodgepole pine (Romme and Despain 1989) or aspen stands (Brown and Simmerman 1986, Debyle et al. 1987). Stands were considered even-aged if deviation in the increment core age of site trees was < 20% (Daniel et al. 1979). Results Three forest cover types consisting of 15 stand types were identified. Species repre- sented in pure stands were lodgepole pine, Engelmann spruce, subalpine fir, and quaking aspen, but the area in pure stands was rela- tively small compared to that of mixed stands: 280 ha in pure stands \ersus 580 ha in mixed stands out of a total 1036 ha. Ol the 15 delineated stand t>'pes, subalpine fir, the climax species in the habitat type pre- sent (Schimpf et al 1980), was a major sec- ondary stand component in 9 types and the 1996] Spkuce-Fik Fikk Fiu:yuENCY 31 Tahle 1. Percent of regeneration by species witliin stand type. Suhalpine fii- is flie primary comi^onent in regenera- tion in all stands except aspen. Stand type Snbalpine fir Engelniann sprnce Aspen Lodgepol pine Donglas- fir DF/PF^' DF/ES/AF LP LP/AF/AS DF/AF DF/ES LP/AF LP/AF/ES ES ES/AF AF AF/AS AS/ES/AF AS 100 0 0 0 0 100 0 0 0 0 75 19 6 0 0 67 11 22 0 0 67 33 0 0 0 67 33 0 0 0 65 35 0 0 0 61 31 0 8 0 60 25 15 0 0 57 32 11 0 0 56 33 11 0 0 52 0 48 0 0 46 27 27 0 0 41 3 53 0 3 "Stand hpe abl)reviatit>ns: AF = siihalpinc lir. AS = aspen, ES = Engelmanii spnice, LP = lodgepole pine, DF = Douglas-lir, PF = linilier pine. principal component in 2. Regeneration sur- veys conducted at each plot showed subalpine fir to be the primaiy regeneration component in 13 of the 15 types (Table 1). Aspen regener- ation was the primary component in the aspen stand t\'pe. Overstoiy ages ranged fi"om 63 to 284 yr in lodgepole pine, 106 yr in aspen, 188 yr in subalpine fir, and 193 yr in Engelmann spruce. Sixty-two fire-scar wedges were collected fiom fire-scarred trees, 22 fiom Engelmann spiiice, 1 fi-om subalpine fir, and 39 fi^om lodge- pole pines. All scar and pith dates were used in the master fire chronology, but only 6 of the spruce scars were used to indicate fire years, while 37 lodgepole pine scars were utilized. The remaining scars were not used due to rings obscured by decay. Sixteen fire years were represented in scar and/or regeneration data. Where scars were not present, but vegetation was even-aged, e.g., stands L20, F24, L18, and L17 (Tables 2, 3), a fire year was determined from the age of dominant lodgepole pine or aspen trees pre- sent. Two of the 16 fire yeai's, 1700 and 1860, were represented solely by age-classes on the site. Two fire years during the settlement per- iod, 1890 and 1895, were documented by Bird (1964). Bird's account stated that numerous small fires were reported in Logan Canyon in 1890, while the 1895 fire year was substanti- ated by a large fire reported in Stump Hollow in Logan Canyon, an area north of the study area (Bird 1964). Those stands where the major component was lodgepole pine exhibited 13 fire years, 4 in the presettlement fire period from 1700 to 1855, 9 in the settlement period from 1856 to 1909, and no fires in the suppression period from 1910 to the present. Ten of the 13 fires were represented by fire scars in the present stands (Table 2). There were 7 fire years in stands in the spruce/fir cover type, which predominantly comprised spruce/fir and secondary compo- nents of lodgepole pine, aspen, Douglas-fir, and limber pine. There were no fires in the pre- settlement period, 6 in the settlement period, and 1 in the suppression period. Five of the 7 fires were recorded by scars and validated by age of the present stand (Table 3). There were 4 fire years in the aspen cover type. Three of those fires were validated by both fire scars and age-class analysis. One fire occurred in the presettlement period and 4 in the settle- ment period (Table 3). Only 1 fire year, 1903, was common to all 3 forest types. Four fire years (1860, 1890, 1902, and 1903) were shared between the spruce/fir and lodgepole pine cover types (Figs. 2, 3). Mean fire inter\'als estimated for the entire study area, for each cover type, and for each fire frequency period are shown in Table 4. Mean fire interval for the entire study area was 18 yr, i.e., a fire occurred about every 18 yr somewhere within the study area. Mean fire interval was shortest in lodgepole pine and longest in aspen. During the presettlement 32 Great Basin Naturalist [Volume 56 Table 2. Fire frequency in the lodgepole cover type by stand and fire year Stands consist of a predominant lodgepole component or mixed species with the priman,' overstoiy component of lodgepole pine. (Adapted fiom Arno and Sneck 1977.) Stands year L2^' L3 L4 L.5 L6 L7 L22 LI I LIO L12 LL5 L20 LLS L17 L9 L13 Suppression period 1942 — — Settlement period 1909 ______ — — _ — — — — — — — 1903 3rl^ 2r Ir 4r 2 1 1 _,■_ — — — — — — — 1902 _______ Ir 1 Ir 1 Ir 1 5r — — 1899 __________ _ _____ 1895 -r____l ___ 1890 ________1 1887 ___-,_____ 1883 _r _ _ Ir — — — — — 1877 -r _ _ -r — 1 _ _ _ I860 _________ 1858 ________2 Presettlement period 1847 -r _r _ 2r — — — — — 1834 ________ ir 1822 ________ Ir 1700 ___,■ _____ aStand description: Lodgepole = L3, L5, L6, L22, L12, L17, L9, L1.3; LP/AF/AS = L2, L7, L1.5; LP/AF = L4. Lll, L20; LP/AF/ES = LIO. LLS, LP = iudgt-pole pine, AF = subalpine fir, AS = aspen, ES = Engelmann spruce. "Digit (L2, etc ...) = number of trees in stand with fire-scar date; r = regeneration in stand, determined from increment cores. Table 3. Rre freciiiencies in the Engelmann spriice/subalpine fir and aspen cover t\'pes. (Adapted from Arno and Sneck 1977.) Aspen Engelmann spnice/suhtilpine fir stands stands Fire year E2'' E4 E5 F21 F24 F7 F23 A3 A8 Supression period 1942 ll' ________ Settlement period 1909 ______1 __ 1903 Ir Ir 2r — — — — 1 Ir 1902 1 ___rl ____ 1899 1 ________ 1890 ______r ___ 1883 _______1 _ 1877 _______! _ 1860 _____r ____ 1858 _________ Presettlement period 1834 _________ 1822 _________ 1700 _________ "Stand description: ES/AF = E2. E.5; AF/AS = F,3, F23; DF/ES = D2; AF/LP = V5. F7; ES/AF/AS = E4, F24: AF/ES = n). Alv'LP/ES = F2L AS = A3. .\S. AS = aspen, AF = subalpine fire, DF = Douglas-fir, ES = Engelmann spruce, and LP = lodgepole pine. •^Digit (1, 2) = number of trees in stand with fire-scar date: r = regeneration in sl:md determined from increment cores that correspond to fire data. 1996] Spruce-Fih Fire pREyuENCY 33 ^ 1 : 2 4.000 0.5 1 1.5 KI L OMET E RS Legend 1847 \ZA 1860 ^ 1877 Regeneration occurring after 1847 ♦ I Regeneration occurring after 1860 \W]\ Regeneration occurring after 1 877 \M.\ Fire scars that recorded the particular fire date Fig. 2. GlS-pi-oduced diagram of fires in the study area from 1700 through 1877 based on stand mapping, regenera- tion, and fire-scar data. 34 Great Basin Naturalist [Volume 56 1 : 24.000 KILOMETERS Legend E^ 1883 ZZ 1890 E] 1902 V77A 1887 en] 1895 E] 1903 3E Regeneration that occurred after the fire date r^\ I Fire scars that recorded the particular fire date Fig. 3. CIS-produced diauiaiii of fiic-s in the stucK area iioiii 1(SS3 through 1942 l)ased on .stand mapping, regenera- tion, and fire-sear data. 1996] Spruce-Fir FikI'; Frequency 35 Table 4. Mean fire intenal In conlt t\pe and firi' irciinenty [leiiod. Mean lire interval i.s an aritlnnelic average in years of tlie nnniher of years in a period dixided by the nnmher ol iires oecurring in that ])eriod. A d()nl)le hyphen denotes that no exidenee of fire occurring in tliat period was found. Ranges of intervals are in parentheses. Presettlement Settlement Suppression Total (1700-1855) (1856-1909) (1910-1988) (289 years) Stud\ area 39 (1-122) 4.9 (1-30) 79 18.1 Cover types ES/AF — 9 (1-30) 79 41.3 LP 39 (12-122) 6 (1-17) — 22.2 AS 156 13.5 (4-16) ~ 57.8 and settlement periods, mean fire intei"val was shortest in lodgepole pine. Mean fire internals were longest in the suppression period (e.g., spruce/fir) or no fires occurred (e.g., lodgepole pine and aspen; Table 4). Discussion Stand Age and Regeneration The widespread occurrence of subalpine fir in the cover types, both in the overstoiy com- ponent and in the regenerating understory, is associated with later stages of succession (Schimpf et al. 1980). Stands sustaining the most recent extensive fires, 1902 and 1903, have less of a subalpine fir component than those not withstanding recent fires (Figs. 2, 3). However, subalpine fir is apparent as a com- ponent of regeneration following these fires and now as a tolerant understoiy. Fire frequencies declined during the last century, a trend that would favor the establish- ment of stands of Engelmann spruce and sub- alpine fir that are less resistant to fire. When a subalpine fir climax is reached, overtopping intolerant serai species, it is not easily replaced due to its tolerant reproduction, unless a dis- turbance interferes, such as fire, insects, dis- ease, or logging (Eyre 1980). Aspen stands also have a component of subalpine fir present and will require a disturbance if they are not to be replaced by the tolerant subalpine fir climax (Mauk and Henderson 1984). Fire Frequencies Compared to the mean fire interval in the presettlement period, there was a large increase in fire frequency in the settlement period in all 3 cover types (Tables 2, 3). Both Bird (1964) and Roberts (1968) stated that ignition sources increased while settlement was occurring in Cache Valley Size and number of fires in the mountains surrounding Cache Valley coincided with the heaviest use period (Bird 1964). The 1880 cen- sus stated 1%-10% of the timbered area of Cache County buiTied, or 5000 to 50,000 acres. Heavy grazing of the period undoubtedly reduced fine fuel loads, but use by loggers and sheepherders increased ignition hazards. Fires were largely untended until 1906, when the U.S. Forest Service arrived. An employee of the U.S. Forest Sei^vice in 1906 stated that 3/4 of the Bear River Forest Reserve (later to become part of the Wasatch-Cache National Forest) had been burned over in tlie last 20 yr, probably due to careless sheepherders (Bird 1964). Fires were recorded in Blacksmith Fork Canyon in 1878, as well as a "large fire in Stump Hollow in Logan Canyon in 1881 (Bird 1964). Compared to the settlement period, fire fi"e- quency decreased during the suppression per- iod and there was no evidence of fire in the lodgepole pine and aspen types. Forest Ser- vice suppression techniques decreased the size and occurrence of fires, which also coincided with a large reduction in allowable grazing, lessening an ignition hazard (Bird 1964). The lack of evidence of fire since 1910 can- not be attributed to deterioration of fire-scar evidence. A fire severe enough to scar stand- ing trees should be recorded in the present stands. The actual fire frequency may be higher than recorded; fires may not have been severe enough to scar trees (Lorimer 1984) or were suppressed before they became extensive. 36 Great Basin Naturalist [Volume 56 Mean fire intervals in all cover types de- creased in the settlement period and increased or there were no fires during the suppression era (Table 4). There were few if any fires found in this study in the presettlement period. The fire scars in aspen may have been lost to natural mortal- ity and decay, and fires may not have been severe enough to produce fire-scarred trees. Evidence of additional fires in the lodgepole pine and spruce/fir cover types may also have been destroyed, and actual mean fire intei"vals for this period may be substantially shorter. Fire hazard in a lodgepole pine stand is highest shortly following a fire due to standing snags and remaining ground fuels from the previous fire, and later when crowns of the tolerant underston' reach into crowns of mature lodgepole pine creating ladder fuels (Brown 1975, Romme 1982). In tlie study area, less fire- resistant Engelmann spruce and subalpine fir have begun to reach into the crowns of the lodgepole pine and aspen stands, increasing fire hazard. Both spruce and fir are highly sus- ceptible to fire, due to their low-branching habits and thin bark (Schimpf et al. 1980). Evi- dently, fuel was also available to allow several nonlethal fires to bum in lodgepole pine stands, as occurred between 1877 and 1903 in the study area. One stand apparently burned 4 times during this 26-yr period, and several areas burned more than once (Table 2). Conclusions The lack of disturbance by fire on the USU T. W. Daniel Experimental Forest in the last 80 yr has allowed succession to proceed towards a climax of subalpine fir. The increase in fire frequency following settlement was probably due to efforts to exploit natvnal resources and the concomitant increase in ignition sources. Freciuent disturbance by fires dining the settlement period resulted in die present mature vegetative mosaic. These earlier frequent fires favored lodgepole pine, and the less-fre(iuent fires of the suppression period favored more tolerant species, as demonstrated 1)\ the abun- dance of subalpine fir regeneration in all coxcr types. The continued lack of disturbance will allow the more tolerant species of subalpine fir and Engelmann spruce to overtop the intoler- ant lodgepole pine and aspen. Eventually the area will lose its diverse appc>arance and will be similar to that in the areas where fire dis- turbance is less frequent. Acknowledgment This research was supported by the Utah Agricultural Experiment Station, Utah State University, Logan, Utah, as Journal Paper 4689. Literature Cited At;EE, J. K. 1993. Fire ecology of Pacific Northwest forests. Island Press, \V;ishington, DC. 493 pp. Arno, S. E 1980. Forest fire histoiy in the Northern Rock- ies. Journal of Forestry 78(8): 460-465. Arno. S. F, and T. D. Petersen. 1983. Variation in esti- mates of fire intei^vals; a closer look at fire histoiy on tlie Bitterroot National Forest. USDA Forest Service Research Paper INT-301. Arno, S. F, and K. M. Sneck. 1977. A method for deter- mining fire history in coniferous forests of the Mountain West. USDA Forest Service, General Technical Report INT-42. Barreit, J. W. 1980. Regional sil\ iculture of the United States. 2nd edition. John Wile\ ic Sons, Inc., New York. Bird, D. W. 1964. A histoiy of timber resource use in the development of Cache Valley, LI tali. Unpublished mas- ter's thesis, Utah State University, Logan. BiuDLEY, A. F, N. V NosTE, AND W. C. FisCHER. 1992. Fire ecology of forests and woodlands in Utah. USDA Forest Sei^vice, General Technical Report INT-287. Brown, J. K. 1974. Handbook for inventorying downed woody material. USDA Forest Senice, General Tech- nicalReport INT-16. Brown, J. K., and D. G. Simmerman. 1986. Appraising fuels and flammabilit)' in western aspen: a prescribed fire guide. LISDA Forest Sendee, General Technic;il Report INT-20.5. Daniel, T W, J. A. Hel.ms, and F S. Baker. 1979. Princi- ples of silviculture. McGraw-Hill, Inc., New York. Debvle, N. V, C. D. Bevins, and W. C. Fischer. 1987. Wildfire occurrence in aspen in the interior western United States. Western Journal of Applied Forestn 2(3): 73-76. Eyre, F H. 1980. Forest cover types of the United States and Canada. Societ\' of American Foresters, Bethesda, MD. Gruell, G. E. 1983. Fire and vegetative trends in the Northern Rockies: interpretations from 1871-1982 photographs. USDA Forest Senice, General Techni- cal Report INT-158. . 1986. The importance of lire in the Greater Yellow- stone Ecosvstem. Western Wildlands, I'all 1986; 14-18. Mart, G. E., and D. A. Lomas. 1979. Effects of clearcut- ting on soil water depletion in an Engelmann spruce stand. Water Resources Research 1.5: 1598-1602. llllNSELMAN, M. L. 1973. Fire in the virgin forests of the Boundan Waters Canoe Area, Minnesota. Quateman Research ,3: 329-382. Johnson, E. A, and Gutsell, S. L. 1994. Fire frequent) models, methods and interpretations. Advances in I-Aologieal Research 25: 2.39-287. 1996] Sphuce-Fih Fire Frequency 37 LoiUMER, C. G. 1984. Mrthodolo^ical considerations in the analysis of forest clisturl)anee history. Canadian Journal of Forestiy Research 15: 200-213. Mauk, R. L., .\nd J. A. Henderson. 1984. Coniferous for- est habitat types of northern Utah. USl^A Forest Sei^vice, General Technical Report INT-170. McBride, J. R. 1983. Analysis of tree rings and fire scars to establish fire history. Tree-Ring Bulletin 43: .51-67. Pakson.s, D. J. 1981. Role of fire management in maintain- ing natural ecosystems. Pages 469-488 /';; Proceed- ings of the conference — fire regimes and ecosystem properties. US DA Forest Senice, General Technical Report \VO-26. Roberts, R. B. 1968. Histon' of the Cache National For- est. 3 volumes. USDA Forest Sei-vice, Washington, DC. ROMME, W. H. 1980. Committee chairman of fire histoiy terminolog\ : report of the ad hoc committee. Pages 1135-1137 in Proceedings of fire histoiy workshop, 20-24 October 1980, Tucson, AZ. USDA Forest Ser- vice, General Technical Report RM-81. . 1982. Fire and landscape diversity in subalpine forests of Yellowstone National Park. Ecological Monographs 52: 199-221. RoMMK, W. II., and D. G. Despain. 1989. Historical per- spective on the Yellowstone fires of 1988. BioScience 39: 695-699. SciiiMPK, D. J., J. A. Henderson, and J. A. MacMahon. 1980. Some aspects of succession in the spruce-fir forest zone of northern Utah. Great Basin Naturalist 40: 1-26. Stuart, f. D., D. R. Geiszler, R. I. Gara, and J. K. Acee. 1983. Mountain pine beetle scarring of lodgepole pine in south central Oregon. Forest Ecology Man- agement 5: 207-214. Tande, C. F 1979. F"ire histoiy and vegetative patterns of coniferous forests in Jasper National Park, Alberta. Canadian Journal of Botany 57: 1912-1931. Veblen, T. T, K. S. Hadley, E. M. Nel, T. Kitzberger, M. Reid, and R. VlLl^LBA. 1994. Disturbance regime and disturbance interactions in a Rocky Mountain subalpine forest. Journal of Ecolog)' 82: 125-135. Received 18 November 1994 Accepted 2 October 1995 Great Basin Naturalist 56(1), © 1996, pp. 38-47 ARIZONA DISTRIBUTION OF THREE SONORAN DESERT ANURANS: BUFO RETIFORMIS, GASTROPHRYNE OLIVACEA, AND PTERNOHYLA FODIENS Brian K. Sullivan 1, Robert W. Bowker^, Keith B. Malmos^, and Erik W. A. Gergus'^ Abstract. — We surveyed historic collecting localities in south central Arizona dining JuK, August, and September 1993-94 to determine the presence of 3 little-known Sonoran Desert anurans, Biifo retifoniiis. Gastrophnjne olivacea, and Pternoliyla fodiois. All 3 species were present at most historic localities visited under appropriate conditions (fol- lowing rainfall in JuK' and August). Pternohijla fodiens was restricted to San Simon Wash and associated tributaries in south central Pima County. Gastrophnjne olivacea ranged from Vekol Valley in extreme southern Maricopa County south to the Mexican border, and southeast near Tucson and Nogales in Pima and Santa Cruz counties. Bitfo retifonnis occuiTcd over the widest area, from southern Rainbow Valley in Maricopa Coimty southwest to the vicinit>' of Organ Pipe Cactus National Monument, and southeast to the vicinity of Tucson and Sasabe in Pima County. Key tcords: Bufo retifonnis, Gastrophiyne olivacea, Pternohyla fodiens, historic distribution, present distribution, ainpliibian decline, Arizona, Sonoran Desert. Three relatively little-known anurans, Bufo retiforDiis, Gastrophnjne olivacea, and Pterno- hyla fodiens, occur in the Sonoran Desert in south central Arizona. Although placed in sep- arate families (Bufonidae, Microhylidae, and Hylidae, respectively), they are superficially similar in behavioral ecology. Each is inactive for more than 10 mon each year, emerging only to reproduce and forage following intense rainfall during the summer "monsoon' season. All exhibit "explosive" breeding behavior (Wells 1977) in which males form high-density aggre- gations for a few nights (sometimes only one) following a major rainstorm and call to attract females. Within Arizona all 3 species are largely restricted to a small portion of the Sonoran Desert in the extreme south central part of the state, so it is perhaps not suiprising that they are relatively unknowii. Indeed, Bufo retifonnis was described in 1951 based on specimens collected southeast of Ajo in 1948 (Sanders and Smith 1951), and Pternohyla fodiens was first documented in Aiizona in 1957 (Chrapliw>' and Williams 1957, Williams and ChraplivvT 1958). Given limited information on these Arizona aniuans, this investigation was undertaken in 1993 and 1994 to ascertain th(>ir present dis- tribution in Maricopa, Pima, Pinal, and Santa Cruz counties, Arizona. First, we describe methods used in conducting the suney. Then, for each target species sun^eyed, we describe distinguishing acoustic characteristics and out- line historic and present distributions. Last, we present observations on breeding behavior. Materials and Methods Suney Methods All surveys were conducted along paved roads throughout the known ranges of the 3 target species following rainstorms during July, August, and September 1993-94. Given the highly unpredictable and variable nature of summer rainfall and the need for monitor- ing the entire south central portion of Arizona, we could only crudely estimate (e.g., weather reports) the appropriateness of field condi- tions (i.e., le\el of rainfall) for anuran activit\ prior to each field excursion. Whenever suffi- cient rainfall appeared to have fallen in the study area, we traveled to that particular area on the night of the rainfall exent, or the fol- lowing night, to surve\' for amphibians along roadways. Frequently, 2-3 nights of surveying occurred for each rainfall exent. Occasionally, siu\'e\- plans were adjusted to take advantage of local conditions (e.g., localized flooding). I|)i-pai(iiit'iil of Life Sciences, Arizona State Universit\- West, PC) Box 37101). Phoeni.v, .\/. 85069. -Department of Biolosy, Clendale Comnuniit>' College, Glendale. AZ 85302. ^Department of Zoology; Arizona State University', Tenipe, AZ 85287. 38 1996] SONOIUN Dksert Anuiuns 39 To conduct surveys we drove slowly (40-65 kniph) along paved roadways scanning for anurans on the road surface and listening lor chorus acti\'it\' adjacent to the roadway. Most roads in the study area are located in valley floodplains crossed by numerous washes so that collection of large lain pools immediately ad- jacent to roadways occurs commonly. If insuf- ficient rainfall had occurred so that anuran surface activity was initiated but no chorusing activit\' was apparent (i.e., no calling or breed- ing), we continued driving, scanning for and recording all anurans foimd on the road. When activity was relatively high (e.g., >20 anurans/ km) and/or associated with an area of interest (e.g., historic or suspected locality for one of the target species), we recorded eveiy individ- ual anuran seen on the roadway (for a minimum of 1 km) until lack of moisture resulted in reduced anuran activity (e.g., <5 anurans/km). Whenever we detected choiaising activity or pools of water along the roadway, we stopped and scanned the area adjacent to the roadway. If none of the target species were detected either visually or acoustically, we resumed the road survey. If target species were present, we attempted to record a series of voucher calls (see below) and collect a small series of voucher specimens {N < 10). Unfortunately, summer rainfall in south central Arizona was below average during the sui-vey period, resulting in few actual breeding aggregations. All speci- mens are deposited in the ASU Vertebrate Collection. Field Observations Each target species possesses distinctive vocalizations. Advertisement calls were recorded in the field with a Marantz PMD 430 stereo recorder and Sennheiser ME 80 microphone with K3-U power module, or a Sony WM-D6C cassette recorder and Sony ECM-909 stereo microphone. Males generally ceased calling when they were approached {Gastrophryne and Pternohyla were easily disturbed); only if the observer remained relatively motionless would apparently normal calling behavior be resumed. Release calls were recorded either in the field or in the laboratory by gently com- pressing the sides of a male held between thumb and forefinger directly above a micro- phone (following Sullivan 1992). Only slight pressure was necessary to elicit a series of re- lease calls. Cloacal temperatures were measured with a Weber quick-recording thermometer within 5 sec of recording the final advertise- ment call or release call. Water and air tem- peratures were generally within 3°C] of cloacal temperatures during field recordings. Acoustic Analysis Advertisement calls were digitized with a DATA Precision model 610 plug-in digitizer at a sampling rate of 10 kHz (Nyquist frequency = 5 kHz) and analyzed with a DATA Precision 6000 waveform analyzer. Release calls were digitized at a capture rate of 22 kHz on a Macin- tosh LC computer using a Farallon Corpora- tion MacRecorder and analyzed with Sound- Edit software (version 2.03). Call durations were measured to the nearest 0.01 sec with the Waveform analyzer (<2 sec) or with a stop- watch. Pulse rates of advertisement calls were measured over a 0.5-sec interval spanning the call midpoint; all pulses were counted to deter- mine the pulse rate of release calls using the oscilloscope mode of SoundEdit. Dominant fre- quencies were estimated to the nearest 10 Hz over a 0.25-sec intei^val spanning call mid- points using the waveform analyzer. Neither advertisement nor release calls are frequency modulated to any large extent in any of the 3 anurans under study. For each male used in analysis of advertisement and release calls, mean values were generated for each of the 3 call variables from 3 or more calls. Historic Distributions We obtained specimen listings from the fol- lowing institutions: American Museum of Nat- ural Histoiy (AMNH), Arizona State Univer- sity (ASU), Brigham Young University (BYU), California Academy of Sciences (CAS), Carne- gie Museum of Natural Histoiy (CMNH), Los Angeles County Museum (LACM), Museum of Vertebrate Zoology (MVZ), University of Arizona (UA), University of Michigan Museum of Zoology (UMMZ), University of New Mex- ico (UNM), and United States National Muse- um (USNM). It is important to note that we examined only specimens deposited in the ASU collection and a portion of those housed at the USNM. We assume that anurans listed by the other institutions are conectly identified. Given that these 3 anurans are quite distinct from other Sonoran Desert forms and therefore un- likely to be misidentified, it seems reasonable to accept these listings in lieu of a physical 40 Great Basin Naturalist [Volume 56 examination of all specimens. We did, however, obtain detailed information from collectors for any specimen collected outside or on die periph- ery of the range (e.g., San Xavier region). Results and Discussion Bufo retifonnis Relative to other toads (genus Bufo) found in south central Arizona, B. retifonnis pos- sesses an unusually high-pitched, short-dura- tion advertisement call, often described as an "insect-like buzz" (see Stebbins 1985, Hulse 1978). However, given similarities in adver- tisement calls of B. retiformis and G. olivacea, identification based on calls can only be confi- dently determined with analysis of signals in the laboratory (Sullivan unpublished data). On average, B. retiformis calls are longer (/x = 3.0 sec, range = 2.0—4.3 sec at approximately 26° C body temperature) and lower in frequency ilJL = 3112 Hz) than calls of Gastrophryne (typically 1-2 sec duration at =4000 Hz). Historic distribution. — Bifo retiformis is known from west central Sonora and south central Arizona (Hulse 1978; Fig. 1). Since it was described in 1951, this anuran has been obsei-ved in Arizona at sites ranging from near San Cristobal Wash, just west of Organ Pipe Cactus National Monument, north to tribu- taries of Waterman Wash near Mobile, south- east to the vicinity of Tucson (San Xavier Mis- sion), and southwest to the international bor- der near Sasabe. Across this region it occurs in creosote flats, upland saguaro-palo verde asso- ciations, and relatively high-elevation (>900 m) desert grassland. One historic locality deserves special dis- cussion: southern Vekol Valley, Pinal Count)'. At this site Jones et al. (1983) reported bodi B. retiformis and B. debilis. We have examined the single voucher specimens for B. retiformis (USNM 252797) and B. debilis (USNM 252776; SVL = 43 mm, reproductive female) and deter- mined by comparison with juvt-niles in the ASU collection (ASU 23099-23102) that the putative B. debilis is not simply a juvenile B. retijormis. Using the morphometric methods proposed 1)\ Ferguson and Lowe (1969), we scored diis indixidual close to B. debilis in all respects; hence, the B. debilis individual can- not be disnussed as a simiije nnsidentification or hybrid. The presence of /i debilis well with- in the range of R retiformis is especially prob- lematic. No B. debilis have been recorded from appropriate habitat spanning the 240 km be- tween Vekol Valley and the otherwise western- most previous locality for this eastern relative of B. retiformis (near Benson, Arizona). Unfor- tunately, we were unable to sui'vey Vekol Val- ley when conditions were suitable for anuran activity. Present distribution. — In 1993-94, we obsei^ved B. retiformis at or near most historic localities, except San Xavier and Vekol Valley, and at additional sites (Fig. 1). They were especially abundant along Indian Route (IR) 15, 0-40 km north of Quijotoa, associated with the Santa Rosa Wash floodplain. Surveys in which every anuran was identified along a roadway segment (1-65 km) revealed that B. retiformis constituted up to 63% of all anurans sighted on this route (Table 1), whereas they were absent or composed a small proportion (<1%) of total anurans sighted on roadways on the peripheiy of their distribution near Mobile and Sasabe (Table 1). Similarly, this toad was not abundant along State Route (SR) 85 near Organ Pipe Cactus National Monument. Din- ing 1993 and 1994 we never obserxed this species on SR 85 or SR 86 in this westernmost portion of the range. Philip Rosen (personal communication) has observed only a few B. retifonnis near the international border, and a number of individuals near Why, Arizona, dur- ing the course of extensive fieldwork near Organ Pipe Cactus National Monument over the past 6 yr Contraiy to the suggestion ol Hulse (1978; see also Nickerson and Mays 1968), Bufo reti- formis does not appear to be expanding its range northward into areas of agricultural activ- ity (e.g., soutiiern Pinal County). We conducted many sin"\ eys in southern Pinal County: south of Stantield and south of Arizona City, 2 areas directly north of known localities for B. reti- formis (Fig 1). We also extensively surveyed the Avra Valley region, Pima County, immedi- ately west of Tucson, and the \icinity of Mobile, Maricopa County. These habitats are similar to areas inhabited by B. retiformis directb' to the south or west, except that agricultural activity is relati\t'K higher in these areas. It appears that B. retiformis is less conmion on die periph- eiy of its range: near Organ Pipe Cactus Nation- al Monument in the west, near Mobile in the noitli, and in Altar Vallev in the east. 1996J SoNoiUN Desert Anuiuns a) Historic collecting localities for Bufo retiformis in south central Arizona. 41 b) Recent collecting localities for Bufo retiformis in south central Arizona. Fig. 1. Map of a) historic distribution (•) and b) present distribution (•) of Bufo retifonim in south central Arizona. Breeding ACTiviri'. — Like many explosive breeding desert anurans, B. retiformis will take advantage of a variety of water sources for repro- duction. We observed chorusing activity in cattle tanks and roadside pools associated with washes. We obsei-ved B. retiformis breeding in the same pool with all other explosive breed- ing anurans that occur in south central Arizona: B. ulvarius, B. cogmitus, B. punctatus, Gastro- phnjne olivacea, Pternohylafodiens, Scaphiopus concha, and Spea multipUcata. We never ob- served B. retiformis breeding in the absence of other anurans — minimally, B. cognatus and S. concha bred sympatrically with B. retiformis. Male B. retiformis typically call positioned beneath vegetation (e.g., small shrubs or grass), 42 Great Basin NATUii\LiST [Volume 56 Table 1. Numbers of anurans individualK' identified on road surface over a specified distance. Bal = B. alvarius, Bco = B. cognatus, Bpu = B. punctatiis. Ere = B. rctifonnis, Sco = Scaphiupus cotichii, IR = Indian Route, SR = State Route. MM = mile marker Location (appro.ximate) Siuvey distance (km) Species Date Bal m Bco {9c) Bpu m Bre {%) Sco m Total 7/18/94 SR 286 40 31 (33) 13 (14) — — 49 (53) 93 7/28/94 .\rizona Cit>' 24 4 (40) 2 (20) — — 4 (40) 10 7/29/94 SR 286 72 13 (18) 13 (18) 5 (7) — 39 (56) 70 8/7/94 Mobile 25 5 (18) — 3 (11) 1 (3) 19 (68) 28 8/8/94 SR 286 24 3 (23) 4 (31) — — 6 (46) 13 8/8/94 Mobile 30 10 (14) 4 (6) 5 (7) — 51 (73) 70 8/13/94 IR 15. MM 11 4.8 1 (3) 1 (3) 1 (3) 9 (28) 20 (63) 32 8/15/94 IR 15, M.\l 11 3.4 2 (25) — — 5 (63) 1 (13) 8 9/10/94 Stanfield 5.3 64 (75) 9 (10) 1 (1) — 11 (13) 85 1-5 111 from the water's edge. Amplexus is ini- tiated on land with the t>'picall\' larger female earning the male to water for o\iposition. In high-density aggregations, satellite males can be common — we saw as man\' as 3 non-calling males near 1 calling male. Chorusing males and ample.xing pairs were obsen^ed on onK' 4 occasions. Three breeding aggregations along IR 15 were relatively large and located at sites used regularly in the past (e.g., 1984, 1986, 1988; Sullivan and Bowker unpublished). At mile marker (MM) 18.7 on IR 15 north of Quijotoa, a large aggregation formed in a shallow roadside pool (8/9/93). Unfortunately, direct counts of all indi\iduals present were not possible due to restricted property access, but complete counts of all males and females along an open section of the pool shoreline (23 calling and satellite males, 5 females in 75 m) allow a rough mini- mum estimate of >2()0 males and females for the entire pool (=600 m circumference). Obsenations at a 2nd site that same night, a cattle tank (=25 X 50 m) near MM 8.5, nortli of Quijotoa, indicate a thriving population in spite of hybridization with B. piinctatiis (see below). On the 1st night (8/9/93) following hea\->- rainfall in this area, we counted 20 male B. retifonnis at 0300, calling with numerous B. alvariiis, B. cognatus, and B. punctatus. On the following night (8/10/93), appro.\imat(>l\- 40 male B. retifonnis were obsened, in additit)ii to a niiiiimimi of 5 pairs in ample.xus. A 3rd l^reeding aggregation (8/25/94) at a roadside pool (=50 X 25 m) at MM 11 on IR 15 west of Santa Rosa comprised 19 calling males and 5 amplexing pairs (direct count of all individu- als). In contrast to these relativeb' vigorous aggregations, onK' 6 males and a single female were obsened at a "first-night" choiois (8/20/93) in a large cattle tank (=25 X 75 m) near Gun- sight Wash along SR 85. Hybridization with Bvfo puxctatvs. — Bowker and Sullivan (1991) documented a naturally occurring hybrid between B. reti- fonnis and B. punctatus. and we obsened 3 additional h\ brids during oiu" in\estigation (all in August 1993). These Inbrids were obsened along IR 15, 10-20 km north of Quijotoa. Hybrids are intermediate to the 2 parental forms and unlikeK to be confused with an> other anurans in the \ icinit). Gi\ en the appar- ent rareness of hybrids, it is unlikeK that they present a significant concern for the popula- tion status of either parental form. Hxbridization between B. punctatus and B. retifor))iis is somewhat surprising gi\en dra- matic differences in their advertisement calls and habitat preferences (Ferguson and Lowe 1969). Three factors may facilitate li\bridiza- tion between B. punctatus and B. retifonnis along IR 15 north of Quijotoa. First, along IR 15 we observed relatively high numbers of B. retifonnis compared to B. punctatus. and we also noted satellite males near calling males in these aggregations. Male mating tactics such 1996] SoNORAN Desert Anurans 43 as active searching and satellite behavior can in- crease the probability of heterospecific crosses since these tactics subvert active choice by females. Second, although B. retifonnis is typi- cally found in desert flats and B. punctatus generally occurs in rockier, upland regions, the "hybrid zone" along IR 15 (MM 6-12) rep- resents a transition between lowland (Lower Colorado River Subdivision) and upland (Ari- zona Upland Subdivision) desert habitats that would allow coexistence of both species. Third, habitat modification at the site, namely, road construction and development of cattle tanks, may overcome ecological separation between the species and provide opportunities for hybridization. Gastrophryne olivacea As noted above, the advertisement call of G. olivacea can be confused with B. retifonnis. In the hand, this small, narrow-mouthed toad cannot be confused with any other species found in Arizona (Nelson 1972a, 1972b, 1973, Stebbins 1985). Identification based on calls (insect-like buzz) alone must be corroborated by laboratoiy acoustic analysis. Although Lowe (1964) listed G. carolinensis from the mountains near Nogales, Arizona, Nelson (1972a, 1972b) showed that these indi- viduals do not differ significantly from nearby populations of G. olivacea from lower-eleva- tion sites. Having examined specimens from throughout the range in Arizona, we concur with Nelson that only a single taxon occurs north of the international boundary. Historic distribution. — The range of G. olivacea largely overlaps that of B. retifonnis (Fig. 2), except in Santa Cruz County (e.g., near Pena Blanca) where Gastrophryne occurs farther east. Of the 3 anurans surveyed, this species occurs in the widest variety of habitats in Arizona, ranging from low-elevation cre- osote flats through grasslands to oak-woodland communities near Ruby, Arizona (>1200 m). Wake (1961) reported calling G. olivacea 4.8 km southeast of Ajo. Because no individu- als were visually confirmed and because of the difficulty of identifying this species by call, we are inclined to discount the record. Present distribution. — In 1993-94 we obsei-ved G. olivacea at most historic localities except those on the eastern margin of the study area (San Xavier and vicinity of Pena Blanca), and at some new sites (Fig. 2). We observed a small chorus near Lukeville, just north of the international boundaiy, a site that extends the range of Gastrophryne approxi- mately 58 km southwest of the previous west- ernmost locality (San Simon Wash, SR 86) in the United States. Philip Rosen (personal com- munication) suggests that Gastr()})hryne is more abundant in Mexico to the south and southeast of Lukeville. The absence of previous distribu- tional records from Organ Pipe Cactus National Monument substantiates the notion that G. olivacea reaches its northwestern range limit in this area. We were unable to document G. olivacea anywhere along SR 286 (Altar Valley, Buenos Aires Refuge) in spite of apparently adequate habitat and the presence of G. olivacea to the east. Philip Rosen (1994 personal communica- tion) obsei-ved a number of G. olivacea breed- ing choruses in southwestern Santa Cruz County, just east of the Buenos Aires Refuge boundary, during summer 1994. Hence, this species likely occurs in the area but, like B. retifonnis, may be less abundant along SR 286. We did not find G. olivacea in the vicinity of San Xavier Mission or along SR 289, although we visited these sites after rainfall on several occasions. Our failure to document Gastrophryne in areas with appropriate habi- tat may be an artifact of its secretive habits (i.e., individuals may not come on road sur- faces) and small size (i.e., they are difficult to detect when on a road). Breeding activit\'. — Gastrophnjne olivacea aie usually well concealed in vegetation when calling and possess a call that is extremely dif- ficult to localize. They call next to water sources or from floating vegetation. Male satellite activity was not observed. Although G. olivacea has been observed in choruses with all other sympatrically breeding anurans (see above list- ing under B. retifonnis), on many occasions we observed it in large, relatively monotypic aggregations (e.g., MM 26.7 and 35, IR 15). In these areas Gastrophryne often breeds in dense stands of mesquite shrubs growing in the flood- plain of Santa Rosa Wash. Choruses of Gastrophryne are easily de- tected, and we were led to a number of new Gastrophryne localities by their distinctive vocalizations. Because of their secretive nature, we never observed pairs in amplexus, and thus no definitive estimates of population size were obtained for breeding choruses of G. olivacea. 44 Great Basin Naturalist [Volume 56 a) Historic collecting localities for Gastrophryne olivacea in south central Arizona. b) Recent collecting localities for Gastrophryne olivacea in south central Arizona. Fig. 2. Map of a) historic' distrihutioii (•) aiul h) present distriliiition (•) of Gastrophryne olivacea in sonth central Ari- zona. By walking the perimeter oi rain-formed i)o()ls, we obtained rough estimates of >2()() ealling males at 2 sites along IR 15, 43 and 56 km north of Quijotoa, respectively, on recent (8/9/93) and previous sinveys (19(S4: Sulh\an and Bowker impublished). UnfortunateK, since these pools contained considerable vegetation (mesquite shrubs, grass), chorus sizes can onl> be considered approximate (individual toads were not visually verified). B\ contrast, at Luke- \ ille (8/9/94) ouK- 5 calling males were present in a small pool (5 X 10 X 0.25 m). Rain had fallen the previous 2 nights (8/7-8/8), and sev- eral small egg masses were obseiA'ed. Ptcniohyla fodicns Hie advertisement call oi Ptcrnohijlafodiens is a distinctixe "wonk" repeated at a relatively high rate (2/sec: "wonk- wonk- wonk . . . ," etc; see Trueb 1969). Males also produce a call, 19961 SoNOiuN Desert Anukans 45 a) Historic collecting localities for Pternohyla fodiens in south central Arizona. b) Recent collecting localities for Pternohyla fodiens in south central Arizona. 1 ' -MARICOPA COUNTY -,^4^' ^ J r—(¥>. 4 \ ... \^ ' Hickiwan ■- ^ '^* \ t^yj ' \ Organ Pipe / m PINAL COUNTY Queens Well ^Quijotoa Sells PIMA COUNTY I SANTA CRUZ COUNTY .Nogales Fig. 3. Map of a) historic distribution (. ) aiid h) present distribution (•) oi Pternohyla fodiens in south central Arizona. which, based on simihirities with other hyhds, can be tentatively classified as a territorial call. This putative territorial call sounds much like the advertisement call of Pseiidacris triseriata or the sound of a finger sliding across a comb. Historic distribution. — This anuran has been obsei-ved at a few sites (Fig. 3). All locali- ties but Santa Rosa Wash are associated with washes that flow south toward Mexico: San Simon Wash, and its 2 largest tributaries, Hicld- wan and Sells washes. Randy Babb (personal communication) has heard the distinctive vocalization of Pternohyla fodiens many times and visually identified at least 1 individual approximately 16 km north of Quijotoa, west of IR 15, in the floodplain of Santa Rosa Wash. Present distribution. — In 1993-94 we obsei-ved P. fodiens at most historic localities except Santa Rosa Wash and the vicinity of Sells, and at some additional sites (Fig. 3). 46 Great Basin Naturalist [Volume 56 More than the other target species, P. fodiens is found in association with washes. The 2 new locahties we documented are both associated with small tributaries of Sells Wash, a tribu- taiy of San Simon Wash. During the preparation of this report, Thomas R. Jones and Ross J. Timmons (per- sonal communication) found a single male P. fodiens near Santa Rosa Wash, 1 km north of the Pinal County line and west of IR 15 (12 July 1995). This record confirms the presence of P. fodiens in Santa Rosa Wash, well north of the San Simon Wash system. Pteniohyla fodiens is only rarely found on road surfaces, although specimens can be taken near washes when roads are wet (e.g., SR 86 at San Simon Wash). Similar to Gastrophrync, Pternohyla can be easily missed unless chorus activity is underway when a survey is con- ducted. Because of their extremely explosive breeding habits and the lack of sufficient rain- fall near Sells during the survey period, it is not surprising that we obsei-ved no Pternohyla at the historic localities along Sells Wash near SR 86. Breeding activit\\ — We observed breed- ing aggregations of Pternohyla fodiens only in rain-formed pools associated with washes. Calling males are always in or near water, and of the 3 survey anurans Pternohyla seems more dependent on heavy rainfall to initiate breeding activity. This species appears to exhibit the most explosive mating system of the 3 species. We never obsei"ved Pternohyla chorusing more than 36 h after rainfall; by contrast, both Gastrophryne and Bufo were observed in chorus activity 1-4 nights follow- ing rainfall. The only significant Pternohyla chorusing that we observed occiured near Hickiwan (7/13/93) and San Simon Wash (7/13/93). Although direct coimts were not possible, esti- mates from chorusing intensities suggest that dozens, if not hundreds, of calling males may have been present at San Simon Wash along SR 86; however, only a single pair in amplexus was obser\'ed. Large aggregations of Pterno- hyla have been observed at these sites regu- larly over the past 30 yr (Sullivan and Bowker unpublished). Summary Oui- siu-\ey indicates that all 3 target species are present at most historic localities in south central Arizona. We documented range exten- sions to the northwest and southeast for B. retiformis (Mobile/SR 286) and to the south- west for Gastrophryne olivacea (Lukeville). These forms probably occur at all historic localities, since our inability to verify their presence at some sites undoubtedly resulted from the absence of sufficient rainfall. It is critical to note that our survey methods, although allowing rapid coverage of a rela- tively large area, were limited by unpredict- able rainfall and the secretive nature of the target species (especially Pternohyla and Gas- trophryne). Unless chorusing activity was undei^way when we visited an area, the pres- ence of any of the 3 forms may have been overlooked. In the absence of chorusing activ- ity, Bufo retiformis was the only target species regularly found on road surfaces. Minimally, the presence of these anurans at most historic localities suggests no widespread decline as experienced by other anuran amphib- ians in the United States (e.g., ranid frogs of the Southwest; Michael Sredl personal com- munication). Future work should address esti- mation of population levels through mark- recapture methods in conjunction with inten- sive monitoring of single sites throughout as many consecutive activity' periods (June— Sep- tember) as possible. An understanding of fac- tors contributing to variations in species abun- dance will require long-term study. Acknowledgments This research was supported by an IIPAM award (192004) from the Arizona Game and Fish Department Heritage Fund. We grate- fulK' acknowledge the assistance of the Tohono O Odham Nation, especially the Department of Public Safety personnel. In addition, Henn' Ramon of the Hickiwan District, Norbert Manuel of the Sells District, and Madeline Sakiestewa and Jefford Francisco of the Babo- quivari District were especially helpful in co- ordinating activities. Mike Demlong, Robert Dudley, Matthew Goode, Matthew Flowers, and Michael Sredl pro\'ided assistance with field observations. Randy Babb, Darrel Frost, Jeff Howland, K. Bnice Jones, Thomas R. Jones, Cla\t()n Ma\, Phil Rosen, Cecil Schwiillie, Nonn Scott, and Michael Sredl graciousK' shared their field records and experiences. 1996] SoNORAN Desert Anurans 47 LiTER.\TURE Cited Appendix 1 BOWKER, R. W., AND B. K. SULLlVAN. 1991. Anura: Bitfo punctatus X B. retiformis natural liybridization. Her- petological Review 22: 54. CnRAPLiwT, R S., AND K. L. Williams. 1957. A species of frog new to the fauna of the United States: Ptemo- hyla fodiens Boulenger. Chicago Academy of Sci- ence, Natural Histoiy Miscellaneous Publication 160: 1-2. Ferguson, J. H., and C. H. Lowe. 1969. Evolutionaiy relationships of the Bufo punctatus group. American Midland Naturalist 81; 435-446. Hl'LSE, A. C. 1978. Bufo rctifonnis: Sonoran green toad. Catalogue of American Amphibians and Reptiles 207: 1-2. Jones, K. B., L. Porzer Kepner, and W. G. Kepner. 1983. Anurans of Vekol Valley, central Arizona. Southwestern Naturalist 28: 469—170. Lowe, C. H. 1964. The vertebrates of Arizona. University of Arizona Press, Tucson. Nelson, C. E. 1972a. Gastrophnjne olivacea: western narrow-mouthed toad. Catalogue of American Amphibians and Reptiles 122: 1-4. . 1972b. Systematic studies of the North American microhylid genus Gastrophnjne. Journal of Herpe- tology6(2): 111-137. . 1973. Gastrophnjne: narrow-mouthed toads. Cat- alogue of American Amphibians and Reptiles 134: 1-2. NiCKERSON, M. A., AND C. E. Mays. 1968. Bufo retiformis Sanders and Smith from the Santa Rosa Valley, Pima County, Aiizona. Journal of Heipetology 1(1-4): 103. Sanders, O., and H. M. Smith. 1951. Geographic varia- tion in toads of the dehilis group of Bufo. Field and Laboratory 19(4): 141-160. Stebbins, R. C. 1985. A fieldguide to western reptiles and amphibians. Houghton Mifflin Press. 589 pp. Sullivan, B. K. 1992. Calling behavior of the southwest- ern toad [Bufo inicroscaphus). Herpetologica 48: 383-389. Trueb, L. 1969. Pternohijla fodiens: bunowing treefi^ogs. Catalogue of American Amphibians and Reptiles 77: 1-4. Wake, D. B. 1961. The distribution of the Sinaloa nanow- mouthed toad Gastrophnjne mazatlanensis (Taylor). Southern California Academy of Science Bulletin 60(2): 88-92. Wells, K. D. 1977. The social behaviour of anuran amphibians. Animal Behaviour 25: 666-693. Williams, K. L., and E S. Chrapli\w. 1958. Selected records of amphibians and reptiles from Arizona. Transactions of the Kansas Academy of Science 61: 299-301. Specimen numbers for historic collecting localities for Bufo retiformis, GastropJmjne olivacea, and Pternohijla fodiens. Institutional abbreviations: AM Nil = American Museum of Natural Histoiy, ASU = Arizona State Uni- versity vertebrate collection, BYU = Brigham Young Uni- versity collection, CAS = California Academy of Sciences, CMNH = Carnegie Museum of Natural History, LACM = Los Angeles County Museum, MVZ = Museum of Ver- tebrate Zoology, UAZ = University of Arizona, UMMZ = University of Michigan Museum of Zoology, UNM = University of New Mexico, USNM = United States National Museum. Bufo retiformis: AMNH 59189, 60671, 85357-65, 91953- 54, 1022.34-36; ASU 3298-3300, 3894-3902, 3942-48, 8002, 8004, 8005, 22775-76, 23099-102, 23252, 24038-39, 24273-74, 25552-53; BYU 42119; CAS 91.501-04, 94390- 95, 98055-56, 188354-55; CMNH 51562, .53841-42, 538.55, 63,520, 89782-95; LACM 26086-88, 64180-84, 88380-400, 91833, 105719, 11.5266-314, 12.3234-41, 137788-89; MVZ 71906-07, 73751-52, 74206-32, 76620-28, 81269, 139130, 180219-22, 180358-59; UAZ 12369-75, 14848-49, 25847- 48, 31381, 4,3011; UMMZ 133460, 1,36,395, 134077; UNM 30993-995, 31268, 40207, 41686-87; USNM 226443-45, 24,5988, 252797, 322966. Gastrophnjne olivacea: AMNH 88986, 91971-80, 119746; ASU 14014, 22059-60, 22224-25, 22969-70, 22771-74, 2,3095, 23411, 24259-60, 25664-66; CMNH 63138-,39; LACM 26576-81, 91896, 115511, 112480, 12,3293; MVZ 49479-,504, 58922, 72304-05; UAZ 26993-96, 29101-04, 29107, 42187-91, 38181, 35163-64, 38179, 38200-01, 38180, 38197-99, 29027; USNM 252817; UMMZ 136400, 75737-38, 757,53, 92300. Pternohyla fodiens: AMNH 91964-70, 95147; ASU 3301, 1,39,52-68, 22777-80, 24276, 25,556-61; CAS 91505; CMNH 63188-89; LACM 90170-82, 11,5447-75; MVZ 71905, 73747-48, 80104-21, 81271, 178447, 76629-,33; UNM 40201, 40204. Received 4 May 1995 Accepted 1 September 1995 Great Basin Naturalist 56(1), © 1996, pp. 48-53 HABITAT AFFINITIES OF BATS FROM NORTHEASTERN NEVADA Mark A. Ports l and Peter V Bradley^ Abstract. — Bat surveys were completed in 6 habitat types in eastern Nevada between 1980 and 1994. Twelve species of bats and 578 individuals were identified fioni 33 trap localities in 144 trap nights. There were weak correlations between bat species richness and Januan maximum temperatures (0.728, P < 0.05) and mean annual days widi 0° C or lower (-0.704, P < 0.05). Bat species richness exhibited no correlation with annual normal precipitation, Januaiy mini- mum temperatiu-es, July minimum temperatures, and July maximum temperatures. It appears that bat species richness is highest in portions of northeastern Nevada typified by sedimentary' deposits (limestone, dolomite). Igneous mountain ranges (basalt, volcanic ash) generally had moderate bat species richness, and metamoiphic mountain ranges (quartzite) t^'pically had low bat species richness. Notable range extensions include Antrozoiis paUidus (from central Nye Countv' north to the Nevada-Idalio border, approximately 450 km), Tadarida brasiliensis (approximately 350 km north), and Pip- istrellus hespenis (approximately 350 km north). Also, the presence of Lasiomjcteris noctivagans. Ldaiunis cinereus, and Corijiiorluiuis fownsendii was confirmed. Key uords: bats, Chiropfera. Nevada, habitat. Although the distribution of mammals of the Great Basin has been studied in some detail (Hall 1946, Dun-ant 1952, Brouai 1971, Thomp- son and Mead 1982, Wells 1983, Grayson 1987), bats remain poorly known. There are verv" few recent records of bats from the northern Great Basin of Oregon, Idalio, and Nevada (Hall 1946, Durrant 1952, Larrison and Johnson 1981). Here we present new information on habitat affinities and distribution of 12 species of bats from eastern and northeastern Nevada. Such information may prove valuable to land man- agers and wildlife biologists who make deci- sions on how to deal with the impact of human activities on bats. Methods Study Area Northeastern Nevada is part of the Great Basin Division of the Intermountain Floristic Region (Holmgren 1972), an area of continen- tal climate with fairly hot summers and cold, snowy winters. Some 30 north/south-trending fault-block mountain ranges (3000— 1000 m) are separated by high-ele\'ation (15()()-2()0() m) xeric basins. Mountain ranges in northern Elko, Eureka, Humboldt, and Lander counties are mosth' igneous and metamorphic fault blocks, coNcred with \ arious mountain brush communities and fragmented coniferous and deciduous forests. Perennial streams produce riparian habitats in most canyons. Vertical cliffs and stands of de- ciduous and coniferous trees provide sites for da\' roosting and shelter for maternity acti\i- ties. Valle\' floors are mosth' xeric, co\ered with salt-tolerant shrubs {Atriplex spp., Sarco- batus spp.) and sagebrush {Artemisia spp.). Occasional perennial streams extend onto val- le\' floors and are lined with narrow coiridors of deciduous woodlands and mesic shrubs. Mountain ranges in eastern Nevada (White Pine and southern Eiueka and Lander coun- ties) are predominantK' limestone and dolomite fault blocks and tend to have more xeric plant communities. A large number of natural caves and vertical cliff sites provide excellent habi- tats for bat maternit) and hibernation roosts. Natural perennial springs found near the val- ley/mountain fault lines often provide the only dependable water for miles around. Contigu- ous coniferous forests on some of the higher mountain slopes provide suitable tree roosts. Abandoned mine shafts and adits are abundant in northeastern Nevada and are criticalK' im- portant to some bat species, botli siunmer and winter. Sur\ e\' Methods Sun eys began in the smnmer of 1980 and extended through the fall of 1994. Capture 'Biolo)^- Departim-nt, Great Basin Colliue, 1.500 College Parkway, Elko. N\' 89801. ^Nevada Division orWikllilc, 137.5 Mountain Citv Hut.. Elko, NV 89801. 48 1996] Bats from Northeastern Nevada 49 methods included mist nets, hand capture, and harp trap (Kunz and Kurta 1990). Mist nets and the harp trap were used over perennial streams, small springs, beaver ponds, livestock tanks, in forest canopies, and adjacent to mine shafts, adits, and natural caves. Captured bats were identified, sexed, reproductive status recorded, aged, weighed, and then released. Some indi\'idiials were taken as voucher speci- mens and are temporarily held in tlie vertebrate collection of Great Basin College. S. Altenbach (personal commimication) and M. OFairell (per- sonal communication) assisted in identifica- tions. Localities were identified on 1:100,000 scale metric topographic maps. To describe habitat affinities, we delineated 6 general habitat types for the region: C-river canyons in igneous or metamoiphic rock, above low-gradient, perennial streams lined with Cottonwood {Popiilus spp.), willow {Salix spp.), and mesic shrubs {Roso spp. and Ribes spp.), elevation approximately 2200 m; S-foothill and valley springs, with or without deciduous trees and a surrounding area of salt-tolerant shrubs {Athplex spp., Sarcobatiis spp.) or mountain brush {Artemisia spp., Amelanchier spp., Sambiicus spp., Syrnphoricarpos occiden- talis, Purshia tridentata) communities, eleva- tion approximately 2000 m; F-mid- to high- elevation coniferous forests of juniper {Jiini- penis osteosperma), fir {Abies concolor and A. lasiocarpa), spruce {Picea engehnannii), and pine {Pinits monophylla, P. flexilis, and P. lon- gaeva) often with cliff sites and natural caves in the proximity, elevation approximately 230O-.3000 m; D-mid- to high-elevation decidu- ous forests of aspen {Populus tremuloides), Cot- tonwood {Populus spp.), and mesic shrubs {Amelanchier spp., Prunus spp., Betula occi- dentalis, Ahius tentdfolia) often along high- gradient, perennial streams, elevation approxi- mately 2300-2800 m; U-natural caves and underground mine shafts/adits with surround- ing plant communities described in habitats C, F, S, and D; and B-buildings in towns and on ranches. There may also be additional important bat habitats not yet identified in this region. Results and Discussion A total of 578 individuals of 12 species of bats were identified from 33 trap localities in 144 trap nights from eastern and northeastern Nevada (Tables 1, 2 and Appendix 1). Three species of Myotis, (M. evotis, M. volans, and M. ciliolahrum) were the most widespread (Appendix 1) and had the highest occurrence (Tables 1, 2) of bats from eastern Nevada. M. evotis was one of the most abundant species of Myotis in eastern Nevada and occurred in all habitats except towns and around buildings. This species is most often associated with mid- elevation pinyon pine and Utah juniper wood- lands (Manning and Jones 1989). We, too, found this species to be most abundant in this habitat type (localities 8, 9, and 18, TdhXe 1). M. evotis depended heavily on the presence of natural springs within these woodlands as their sole source of water. M. volans was also found to utilize a variety of habitats in eastern Nevada, including pinyon-juniper woodlands such as those found near Old Man's Cave. Eight lac- tating females were examined at this site, sug- gesting a nearby nurseiy colony. Upon release, 4 individuals flew into the cave while the oth- ers flew to nearby rock outcrops. The litera- ture suggests that this species uses cracks in cliff sites and areas beneath bark as roost sites and caves only as hibernacula (Warner and Czaplewski 1984). It is possible that M. volans is using caves in easteiTi Nevada as maternity roosts, although more data are needed to con- firm this. M. ciliolabrum also occurred in a variety of habitats in eastern Nevada (Table 1), including river canyons with sunounding sage- brush deserts (locality 14, Appendix 1). Lamson and Johnson (1981) found this species in simi- lar canyon and desert habitat in central Idaho. Only 6 individuals of M. htcifiigus were caught. This species was uncommon and more restricted in its habitat affinities. Unidentified specimens of Myotis were sent to Dr. Scott Altenbach and Dr. Mike O'Farrell to deter- mine whether or not M. californiciis is present in this region (Table 2, Myotis spp.). Tentative identifications suggest that M. californicus may be found in southern White Pine County, while M. ciliolabrum is more common in the remainder of the region. The 3 high-elevation, tree-roosting species (L. noctivagans, E. fusciis, and L. cinereus) were found in order of decreasing occuirence (Table 1). These species were found repeatedly in several mountain ranges of eastern Nevada that have a combination of coniferous and/or deciduous trees (aspen, cottonwood, white fir, subalpine fir, and Engelmann spruce) for 50 Great Basin Naturalist [Volume 56 Table 1. Occurrence of bat species by locality (see Appendix 1). Habitat affinities (C-river canyons, S-springs, F-high-elevation coniferous forests. D-mid-elevation deciduous forests, U-underground caves and mines, B-buildings) for each species and relative frequencies for each species examined. Bat species Localities (Appendix 1) Habitat affinities Mtjotis ciliolabrinn Mijotis evotis Mijotis hicifugiis Mijotis volans Lasiiinis cinereus Lasionycteris nocfivagans Eptesicus fnsciis Pipisf reikis hespcnis Conjnorhiniis townsendii Antrozous palUdus Tadarida hrasiliensis 2,6,8,9, 10, 11, 12,14, 17, 20, 25, 26, 29, 32, 33 1,3,4,6.8,9,11, 12, 15-19, 21, 22, 25, 32, 33 5, 12, 15-17 1, 2, 6, 7, 9-12, 15, 17-19, 24, 25, 27, 32 10, 17, 20 10-12, 17, 23, 28, 29, 32 10, 12, 17, 23, 26, 29, 32 10,29 5, 9, 10, 13-15. 24-27. 30. 32 10, 14, 15, 25 10, 29. 31. 32 C, S, E D, U, B C, S, E D, U C, E D, U C, S, E D, U S.ED C, S, E D, B C, S, E D, U, B S, B C, S, U C, S, U S. U. B roosting and open water in the form of beaver ponds, stock tanks, and perennial streams for foraging and drinking sites. In the mountains of the West, these 3 species are known to com- monly forage together in similar habitats along with 2-4 species of Mijotis (Kunz 1982). In eastern Nevada high-elevation deciduous and coniferous forests are limited to watered drainages and north-facing slopes in the larger mountain ranges. This suggests that these species are uncommon when compared to populations in the northern Rocky Mountains and may be negatively impacted by deteriora- tion, fiagmentation, and/or total removal of for- est habitats by hard-rock mining, livestock graz- ing, and logging. Foothills covered with pinyon pine and Utah juniper, caves, and river canyons with high cliffs provided habitats for 2 lower-elevation breeding species, Corynorhinus townsendii and A. pallidus. C. townsendii had 4 times the fre- quency of occurrence as A. pallidus and appeared to be more evenly distributed across the region (Table I). C. townsendii and A. pal- lidus depend heavily on cliff sites, natural caves, and mine shafts/adits for maternity, hiberna- tion, and day roosts in eastern Nevada. They are found to utilize similar situations in other arid regions of the West, such as California, Montana, Washington, and Utah (Kunz and Martin 1982). Hermanson and O'Shea (1983) rarely found A. pallidus using caves, but rather found them depending heavily on crevices and cliff sites for maternity roosts, day roosts, and hibernacula. We found this species using caves (localities 15, 25), cliff sites (14), and val- ley springs (10) in eastern Nevada. A large, historic colony of T. hrasiliensis was found occupied in July 1994. Vandalism may have caused this population to roost else- where in 1992 and 1993. Outside of Las Vegas and Reno, this colony is the largest known concentration of mammals in Nevada. Based on visual techniques suggested by Kunz and Kurta (1990), we estimate the population at between 54,000 and 82,000 animals. P. hesperus was found in low numbers in this region. Two individuals were caught 320 km apart, and no meaningful habitat patterns were identified for this species. Species found in and around abandoned mine shafts and adits included C. townsendii, M. ciliolahnnn, and M. volans. C. townsendii was found using mines dining both winter and summer Mijotis species were found only in summer Pat Brown (personal communication) recently docimiented a maternity colony of Antrozous pallidus in an abandoned mine shaft in northern Lander Count)' as well. Climatological data from Elko in the north- eastern part of the state, Ely in the east central, and Las Vegas in the south were compared to 1996] Bats from Northeastern Nevada 51 Table 2. Number of bats examined, percent freciuency by species, and nnmber of specimens collected and preserved from eastern Nevada (1980-1994). Bat species Number of bats % frequency Specimens collected Mijotis ciliolabnnn 73 Mijotis evotis 112 Mijotis Iticifiif^u,s 6 Mijotis volans 186 Mijotis spp. 16 Lasiiiriis cinereus 3 Lasioni/cteris noctivagans 39 Eptesicus fiiscits 52 Pipistrellus Hesperus 2 Corynorhinm townsendii 69 Antrozoiis paUidiis 15 Tudarida brasiliensis'^ 5 TO'IAL 578 13.0 19.0 0.4 32.0 3.0 0.1 7.0 10.0 0.1 12.0 3.0 0.4 100.0 2 3 1 3 2 1 4 2 0 1 1 2 22 ''Roost cavern not included in calculations. bat species richness from each of these regions (Hall 1946, Durrant 1952). Pearson's 3i and Spearman's Rho tests were used to test for correlations. Bat species richness exhibited no correlation with the following climatologi- cal data: annual normal precipitation, January minimum temperatures, July minimum tem- peratures, and July maximum temperatures. There were weak correlations between bat species richness and January maximum tem- peratures (Pearson's % 0.728, P < 0.05) and mean annual days with 0° C or lower (Pear- son's % -0.704, P < 0.05). Bat records were pooled by mountain ranges with similar rock types — sedimentaiy, igneous, or metamoiphic. Bat species richness was high- est in portions of northeastern Nevada typified by sedimentary rock (limestone, dolomite). Igneous mountain ranges (basalt, volcanic ash) generally had moderate bat species richness, and metamorphic mountain ranges (quartzite) typically had low bat species richness. Several bat localities from eastern Nevada represent notable range extensions. Four locali- ties (10, 14, 15 and 25, Appendix 1) for A. pal- lidus extend its range from central Nye County (Hall 1946) north to the Nevada and Idaho border, approximately 450 km. Two specimens of T. hrasiliensis at Swallow Canyon (locality 10, Appendix 1), the recent confirmation of a large roost colony, and the two specimens from Elko (locality 29, Appendix 1) represent the first records of this species for Elko and White Pine counties (Hall 1946) and extend its range approximately 350 km north. The capture of single specimens of P. hesperus at Swallow Canyon (locality 10, Appendix 1) and in Elko (locality 29, Appendix 1) also suggest a north- ern range extension and, based on spring and late-summer capture dates, may represent mi- grating individuals. Although certain bat species have long been suspected of occuning in this region (Hall 1946, Durrant 1952, Kunz 1982, Kunz and Martin 1982), the localities listed in Appendix 1 rep- resent the first range confirmations for L. noc- tivagans, L. cinereus, and C. townsendii in east- em and northeastern Nevada. On examination of contributing abiotic fac- tors such as geological features, precipitation, and average temperatm^es, one can see patterns in eastern Nevada's bat fauna beginning to emerge. The greatest diversity of bat species from eastern Nevada was recorded in east central Nevada. The lower maximum January temperatures and more annual days below 0° C in east central Nevada contradicted the cor- relations in our data and suggested that factors other than climate were contributing to zoo- geographical patterns. East central Nevada's mountain ranges are primarily sedimentary in nature and provide abundant caves, cliff sites, and high-elevation forests for roosting and hibernation. In northeastern Nevada most of the mountain ranges are igneous or metamor- phic in structure, thus reducing the number of potential roost sites for bats. Climatic factors undoubtedly play a large role in defining bat 52 Great Basin Naturalist [Volume 56 distribution. However, the density of suitable roost sites may prove to be an even greater influence on bat distribution where roost site availability becomes a limiting factor Inasmuch as most bat species probably do not migrate more than 1500 km from maternity roosts to hibernacula (Hill and Smith 1992), an abun- dance of suitable hibernation roosts would probably provide any given bat fauna the best chance of survival in an area where severe winters are commonplace. Manning, R. W., and J. K. Jones, Jr. 1989. Myotis evotis. Mammalian Species 329; 1-5. Thompson, R. S., and J. I. Mead. 1982. Late Quaternary environments and biogeography in the Great Basin. Quaternary Research 17: 39-55. Warner, R. M., and N. J. Czaplewski. 1984. Myotis volans. Mammahan Species 224: 1^. Wells, R V. 1983. Paleobiogeography of montane islands in the Great Basin since the last glaciopluvial. Eco- logical Monographs 53: 341-382. Received 24 March 1995 Accepted 15 August 1995 Acknowledgments We wish to thank the numerous people who accompanied us in the field, especially our families, Lois and Susan and the boys, Roger, Mark, Boden, and Jedediah. Thanks also to Dr Scott Altenbach (University of New Mex- ico); Dr. Mike O'Fanell (O'Farrell Wildlife Consulting); Lariy Hyslop and Len Seymour (Great Basin College); Linda White-Trifaro and Mitchell White (USFS); Cristi Baldino and Vidal Davila (Great Basin National Park); Curt Baughman, Lany Gilbertson, Sara Gran- tham, Gary Herron, Rory Lamp, and Tyler Turnipseed (Nevada Division of Wildlife); and the Northeastern Nevada Naturalists. Literature Cited Brown, J. H. 1971. Mammals on mountaintops: nonequi- lihrium insular biogeography. American Naturalist 105: 467-478. DURRANT, S. 1952. Mammals of Utah: taxonomy and dis- tribution. University of Kansas, Museum of Natural Histoiy Publication 6: 1-159. Grayson, D. K. 1987. The biogeographic histoiy of small mammals in the Great Basin: obsen'ations on the last 20,000 years. Journal of Manmialogy 68: 359-375. Hall, E. R. 1946. The mammals of Nevada. University of Galifornia Press, Berkeley. 710 pp. Hermanson, J. W, and T. J. O'Shea. 1983. Antrozoiis pal- lidus. Mammalian Species 213: 1-8. Hill, J. E., and J. D. Smith. 1992. Bats: a natural histoiy University of Te.xas Press, Austin. 243 pp. HoLMCREN, N. H. 1972. Plant geography of the Inter- mountain Region. Pages 77-161 in A. Cronquist, N. H. Holmgren, and J. L. Reveal editors, Intermountain flora. Volume 1. Hafner Publishing Go., New York. KuNZ, T. H. 1982. Lcmonycteris noctiixigens. Mannnalian Species 172: 1-50. Kunz, T. H., and Allen Kurta. 1990. Capture methods and holding devices. Pages 1-29 in T. H. Kunz, edi- tor. Ecological and behavioral methods for the stud\' of bats. Smithsonian institution Press, Washington, DG. 533 pp. Kunz, T. H., and R. A. Martin. 1982. Plecotus tnivnsendii. Mammalian Species 175: 1-6. Larrlson, E. J., and D. R. Johnson. 1981. Manmuils of Idaho. University of Idaho Press, Moscow. 166 pp. Appendix 1 Bat Survey Localities and Animals E.xamined 1. Stump Greek, 8.2 mi S and 7.6 mi W of Northfork, Independence Mountains, Elko Go., Nevada. T40N, R53E, SWl/4 sec 12. 2325 m. 17 Julv 1980, Mijotis evotis (1), M. volans (2). 2. Sheep Greek, 8.5 mi S and 7.8 mi W of Northfork, Independence Mountains, Elko Go., Nevada. T40N, R53E, NWl/4 sec 13. 2320 m. 6-7 August 1980, Myotis volans (1 lactating female), M. ciliolabnnn (1 male). 3. Jim Greek, 10.4 mi S and 7.2 mi W of Northfork, Independence Mountains, Elko Co., Nevada. T40N, R53E, NEl/4 sec 25. 2155 m. 15 July 1981, Myotis evotis (2 nonscrotal males). 4. Jarbidge River, 5.5 mi S and 1.2 mi E of Jarbidge, Jarbidge Mountains, Elko Co., Nevada. T45N, R58E, SEl/4 sec 10. 2460 m. 26 July 1981, Myotis evotis (1). 5. Northfork of tlie Humboldt River, 12.4 mi S and 2.5 mi E of Northfork, Elko Co., Nevada. T.39N, R55E, center sec 3. 1850 m. 7 Sept. 1981, Corynorhinus townsendii (1); 30 August 1989, Myotis hicifugus (1). 6. Mouth of Cave Creek, Ruby Lake National Wildlife Refuge, east slope of the Ruby Mountains, Elko Co., Nevada. T27N, R57E, SWl/4 sec 24. 1850 m. 25 July 1986, Myotis volans (2), M. evotis (1): 15 June 1987, Myotis evotis (1), M. ciliolabnim (1). 7. Ferguson Springs, 1/4 mi W of Ferguson Station on St. Hwy 93, Elko Co^, Nevada. T30N, R69E, NEl/4 sec 33. 187.5 111. 17 Sept. 1989, Myotis volans (1). 8. Arizona Springs, southeast end of the East Hum- boldt Range, Elko Co., Nevada. T33N, R61E, SWl/4 sec 20. 2050 m. 21 June 1991, Myotis evotis (9 males, 18 lac- tating females), M. ciliolahrwn (3 males). 9. Sidehill Spring, 6.4 mi S and 11.8 mi W of Wend- over, Goshute Mountains, Elko Co., Nevada. T32N, R68E, SWl/4 sec 14. 22.55 m. 7 June 1991, Myotis evotis (6 males, 2 lactating females), M. volans (4 males, 2 lactating females), .\/. ciliolahruni (1), Corynorhinus townsendii, 1 male. 10. Swallow Canyon, spring site at the mouth of the canyon. Snake Range, White Pine Co., Nevada. TllN, i^68E, sec 5. 2100 iii. 21 August 1991, A/(/()^/,s' ciliolahrwn (1), Lasionycteris noctivagans (1 male, 2 females), Lasiurus cinerius (\ male), Tadarida hrasiliensis (2 males), Antro- zous pallidas (1 lactating female); .30 August 1991, Myotis volans (1 male), Lasionycteris noctivagans (19 males), Pip- istrelhis hesperiis (1 male), Eptesiciis fiisciis (2 males), Cory- norhinus tounscn(hi (1 male); 22 August 1994, Myotis 1996] Bats from Northeastern Nevada 53 Lolaus (8), M. evotis (1), M. ciliolahrwu (11), Coryiwrliiiiiis townscndii (1), Lasiomictens noctiv(i;i:,(ut.s (2), Ei)tc'sicits fiiscus (1). 11. Headwaters of" McCall Cheek, Bull Run Moun- tains, Elko Co., Nevada. T45N, R52E, middle sec 23. 2420 m. 6 July 1991, Mijotis volans (2), M. evotis (1), M. ciliolahrum (1), Lasiomjcteris noctivagans (2). 12. Man's River, 6.5 mi S and 2 mi W of Maiy's River Peak, Jarbidge Mountains, Elko Co., Nevada. T44N, R58E, SWl/4 sec 35. 2220 m. 30 July 1990, Mtjotis evotis (2 males), M. ciliolahnnn (1 lactating female), Eptesicus fiiscus (1 lactating female); 31 July 1990, Mijotis lucifugus (2 males), M. evotis (1 lactating female), M. volans (1 male, 2 lactating females, 5 nonlactating females), M. ciliolahrum (2 females), Eptesicus fuscus (1 male, 1 female), Lasiomjc- teris noctivagans (2 males); 1 August 1990, Mijotis volans (2 males, 3 lactating females), M. evotis (1 lactating female), Eptesicus fuscus (2 males), Lasiomjcteris noctivagans (2 males). 13. Complex of mine shafts in Snowstorm Mountains, canyon 1.5 mi N of Midas, Elko Co., Nevada. T39N, R46E, NVVl/4 sec 16. 1950 m. 31 May 1992, Conjnorhinus town- scndii (3). 14. Salmon Falls Creek, 1.6 mi W of Jackpot, Elko Co., Nevada. T47N, R64E, center sec 10. 1500 m. 23 May 1992, Mtjotis ciliolahrum (1); 24 June 1992, Mijotis ciliolahrum (1), Antrozous pallidus (3 males, 1 lactating female), Coryno- rhimts townscndii (1 lactating female). 15. Goshute Cave^ Cheny Creek Range, White Pine Co., Nevada. T25N, R63E. 20 June 1992, Mijotis evotis (2), M. lucifugits (1), Conjnorhinus townscndii (3), Antro- zous pallidus (3); 16 August 1992, Mijotis evotis (2 males, 4 lactating females), M. volans (2 scrotal males), Mijotis spp. (either ciliolahrum or californiciis) (1 scrotal male), Antro- zous pallidus (3 scrotal males, 1 nonscrotal male), Conjno- rhinus townsendii (5 scrotal males). 16. Bruneau River, junction of Cottonwood Creek and the Bruneau, Elko Co., Nevada. 1725 m. T45N, R57E, NWl/4 sec 20. 7 June 1992, Mijotis evotis (1), M. lucifugus (1 pregnant female); 22 July 1992, Mijotis evotis (1). 17. Mill Creek, 1.6 miN and 2.4 mi W of Jack Creek Campground, Independence Range, Elko Co., Nevada. T42N, R53E, SWl/4 sec 16. 2620 m. 15 July 1992, Mijotis evotis (1), M. ciliolahrum (5 males, 6 lactating females), M. volans (1), M. lucifugus (1), Eptesicus fuscus (3 males, 4 lactating females), Lasiomjcteris noctivagans (3), Lasiiirus cinereus (1). 18. Water Canyon and Buck Springs, southwest slope of the Ruby Mountains, White Pine Co., Nevada. T25N, R56E, NW'l/4 sec 1 and T26N, R56E, center of sec 35, respectively. 2300 m. 6 July 1992, Mijotis evotis (3 scrotal males, 2 females), M. volans (2 scrotal males, 3 females), M. californiciis (1 scrotal male); 22 July 1993, Mijotis evo- tis (17), M. volans (2), M. californicus (2). 19. Middlefork of Doby George Creek, 1.2 mi S of Maggie Creek Summit, Bull Run Mountains, Elko Co., Nevada. 2050 m. 27 July 1992, Mijotis evotis (4), M. volans (1). •^Because of the sensitivity of natural caves, location descriptions are li; ited to township and range information. 20. Horse Creek, 5.2 mi W and 0.4 mi N of Secret Pass, East Humboldt Range, Elko Co., Nevada. T34N, R61E, NEl/4 sec 16. 2520 m. 4 August 1993, Myotis cilio- lahrum (7), Lasiurus cinereus (1). 21. USES campground on Northfork of Berry Creek, Schell Creek Range, White Pine Co., Nevada. T17N, R65E, SEl/4 sec 10. 2550 m. 9 July 1993, Myotis evotis (3 lactating females). 22. Worthington Canyon, Schell Creek Range, White Pine Co., Nevada. T17N,'R65E, center sec 16. 2550 m. 10 July 1993, Myotis evotis (3). 23. Currant Creek, USES campgrounds, 1.8 mi E and 0.8 mi S of Currant Mountain, White Pine Co., Nevada. 2650 m. 11 July 1993, Eptesicus fuscus (1), Lasiomjcterus noctivagans (1). 24. Old Man's Cave, North Snake Range, White Pine Co., Nevada. T15N, R70E. 16 August 1993, Corijnorhinus townscndii (4 scrotal males, 1 nonscrotal male, 4 lactating females, 4 nonlactating females), Myotis volans (1 scrotal male, 5 nonscrotal males, 9 females), Myotis spp. (2 males, 1 lactating female, 1 nonlactating female); 7 September 1994, Corijnorhinus townsendii (7 males, 17 femtiles), Myotis volans (2 females). 25. Snake Creek Cave, Snake Creek, South Snake Range, White Pine Co., Nevada. T12N, R70E. 17 August 1993, Myotis ciliolahrum (1 scrotal male, 3 females), M. californiciis (1 lactating female), M. evotis (1 male, 1 female), M. volans (1 female), Corijnorhinus townsendii (1 scrotal male), Antrozous pallidus (3 scrotal males). 26. Pescio Cave, Schell Creek Range, White Pine Co., Nevada. T19N, R64E. 18 August 1993, Myotis ciliolahrum (2 scrotal males), M. californicus (1 scrotal male, 2 females), Eptesicus fuscus (1 scrotal male), Conjnorhinus townsendii (1 scrotal male, 1 lactating female). 27. Mine shafts near Emigrant Canyon, Edna Moun- tain, Humboldt Co., Nevada. T36N, R40E, sec 36. 1400 m. 28 Sept. 1993, Myotis volans (1), M. ciliolahrum (2), Corijnorhinus townsendii (3). 28. North Fork Little Humboldt River, 3.5 mi S and 9 mi E of Table Mountain, Santa Rosa Range, Humboldt Co., Nevada. T44N, R41E, sec 1. 2270 m. 10 August 1991, Lasiomjcterus noctivagans (1). 29. Elko, town center, Elko Co., Nevada. T34N, R55E, center sec 15. 22 Sept. 1992 and 23 Sept. 1991, Lasiomjc- terus noctivagans (1); 15 May 1992, Pipistrellus hesperus (1); 19 Aug. 1991, Myotis ciliolahrum (1); 15 July 1992, Eptesicus fuscus maternity roost; 15 Nov. 1994 and 6 Jan. 1995, Tadarida hrasiliensis (2). 30. Mine shaft near Contact, Elko Co., Nevada. T45N, R64E, sec 19. 1800 m. 21 Dec. 1993, Conjnorhinus town- sendii (3 hibernating). 31. Cave in Spring Valley, White Pine Co., Nevada. T15N, R68E. 2300 m. 27 July 1994, Tadarida hrasiliensis roost (54,000-82,000). 32. Muiphy Wash, South Snake Range, White Pine Co., Nevada. TION, R68E, sec 2. 2250 m. 29 July 1994, Cory- norhinus townsendii (3), Lasiomjcterus noctivagans (1), Myotis evotis (11), M. volans (42), Myotis spp. (2), Eptesi- cus fuscus (1), Tadarida hrasiliensis (1); 21 Sept. 1994, Myotis volans (35), M. evotis (4), M. ciliolahrum (1). 33. Rock Creek, Sheep Creek Range, Eureka Co., Nevada. T34N, R48E, sec 8. 1450 m. 21 May 1994, Myotis ciliolahrum (10), M. evotis (1). Great Basin Naturalist 56(1), © 1996, pp. 54-58 NUPTIAL, PRE-, AND POSTNUPTIAL ACTIVITY OF THE THATCHING ANT FORMICA OBSCURIPES FOREL, IN COLORADO John R. Conway^ Abstract. — Obsei-vations and excavations of thatching ant nests from 1990 to 1994 at 2560 m in Colorado provided infomiation on the numbers and behavior of males and winged and wingless queens. Nuptial activit)' was compared to that reported by other investigators at lower altitudes. Reprodiictives were obsewed from 24 June to 15 August. Activity was greatest in 1993 when reprodiictives were on 10 of 98 mounds in the area. Mating and swarming occuned on rab- bitbmsh 4 m from 1 nest 2-6 July. The number of wingless queens in 4 excavated nests varied fioni 0 to 198. Key words: nuptial flight, Formica obscinipes, Colorado, thatching ant. Information on the reproductive activity of the thatching ant, Formica ohscuripes Forel, in Colorado is sparse (Gregg 1963). The puipose of this study is to help remedy the deficiency and to compare nuptial and pre- and postnup- tial activity of the thatching ant at high alti- tude in Colorado with similar studies on this species at lower elevations in North Dakota (McCook 1884, Weber 1935, Kannowski 1963, Wheeler and Wheeler 1963), Michigan (Talbot 1959, 1972), Illinois (Herbers 1978, 1979), Idaho (Cole 1932), and Nevada (Clark and Comanor 1972). The Nevada site north of Reno at 1550 m most closely approximates the Colo- rado study area in elevation and vegetation. Mating flight plays a major role in the reproduction and dispersal of most social in- sects (Holldobler and Wilson 1990). Males and queens of F. ohscuripes fly to "swarming grounds" as reported by Talbot (1972). There males lly back and forth in search of queens, which alight on low vegetation and release pheromones to atti'act males (Cheilx et al. 1993). Materials and Methods The main Colorado study area (64.6 X 114 m) has 85 mounds and is dominated by big sagebrush {Artemisia tridentata Nuttall). It is adjacent to a quaking aspen grove {Popuhis trermdoides Michau.x) at an elevation of about 2560 m. The site is located in Gunnison C^ount)' north of Blue Mesa Resenoir and west of Soap Creek road. Other plants in the study area are Chrysothamnus nauseosus (Pallas) Britton (rub- ber rabbitbrush), Purshia tridentata (Pursh) de Candolle (antelope bitterbiTish), Lupimis argen- teus Pursh (silvery lupine), SympJioricarpos rotiimlifolius A. Gray (mountain snowberry), Rosa woodsii Lindley (Woods rose), Urtica gracilis Alton (stinging nettle), Penstemon strictus Bentham (Mancos penstemon), Ipo- mopsis aggregata (Pursh) Grant ssp. aggregata (trumpet gilia), 1 Saskatoon serviceberry tree {Amelanchier alnifolia var pumila), and 1 Doug- las-fir {Pseiidotsiiga sp.). Observations in this area took place on 5-6 August 1990; 20-28 June, 22-27 July 13-15 August, 12-13 Sep- tember, and 11 October 1992; 28 June-16 August 1993; and 29 June-31 July and 14-16 August 1994. Observations before 20 June were not possible due to academic commit- ments. A nest was excavated on each of the following dates: 6 August 1990, 27-28 June 1992, 12-14 July 1993, and 11-25 July 1994. The 1993 mound was poisoned with 1 1/2 cups Hi-Yield ant killer granules (Diazinon) wetted down with about 2 gal of water prior to excavation. Results and Discussion Reproductives Reproductives (males, winged and wingless queens) were observed in Colorado from 24 June to 15 August over 3 summers. Activity was greatest in 1993 when reproductives were found on 10 mounds scattered among 98 nests in the area: males, winged queens, and wing- less queens on 5 mounds; males and winged 'Department of Biology, Univcrsit\ of Scranton, Scranton, P.V 18.510. 54 1996] Thatchinc Ant in Colorado 55 queens on 3 mounds; a winged queen on 1 mound; and a wingless queen on 1 mound. Observations of both male and female alates on Colorado mounds support Herbers s (1978) observations that some nests produce a mix- ture of sexes. We were unable to confirm reports that some nests produce all males or all females (Kannowski 1963, Herbers 1978), or that a changeover from early all-male flights to later all-female ones occurs (Talbot 1959, 1972, Clark and Comanor 1972). Males. — Males were observed on 8 mounds from 28 June to 13 July 1993 and at 1 mound on 5-6 July 1994. Males seemed to prefer the shady side of 1 mound built around a fencepost. Workers sometimes chased males and once one carried a male on a mound. Oth- ers have reported males earlier in the year. Talbot (1959, 1972) saw males flying 16-24 June, and Clark and Comanor (1972) saw males from 15 April to 4 May. Although males were observed from 0740 to 1635 hours in Colorado, they were most numerous and flew from 0938 to 1101. Talbot (1959) saw them fly even earlier, between 0608 and 1000. Clark and Comanor (1972) also saw morning flights, but noted males through- out the day (0840 to 1445). The largest number of males on 1 Colorado mound was 10 on 3 July 1993, about the same maximum per mound (12) reported by Clark and Comanor (1972). Herbers (1979) noted up to 1264 males. Talbot (1959, 1972) reported even more males (up to 4500) but noted that the ratio of males to females varies from colony to colony and from flight to flight. One male was found in a Colorado nest exca- vated in July 1993; none were in 3 other exca- vated nests. Wheeler and Wheeler (1963) re- ported males in nests from 23 May to 12 July. Winged queens. — Winged queens were observed on 9 Colorado mounds from 28 June to 16 July 1993, and one was on a mound on 5 July and 10 July 1994. Workers pulled queens by their wings and antennae on mounds and were in turn sometimes dragged by queens. Queens were noted with tattered, spread, and partial wings from 30 June to 6 July. Others reported winged queens at nests earlier and later in the season than in Colorado. Clark and Comanor (1972) saw them as early as 1 May, and Wheeler and Wheeler (1963) reported winged females in nests as late as 8 August. Winged queens were observed from 0654 to 1640 hours in Colorado, but most often in the morning. Clark and Comanor (1972) also saw them throughout the day, from 0830 to 1720. Those found later in the day were pre- sumably remnants of the morning activity. The maximum number of winged queens on 1 Colorado mound was about 50 on 3 July 1993. Odiers reported greater numbers per nest: 78 (Clark and Comanor 1972) and 230 (Talbot 1959). Winged queens were more abundant than males on Colorado mounds as reported by Clark and Comanor (1972), except on 1 occasion when males were more numerous. No winged queens were found in 4 excavated Colorado nests. Wingless queens. — Dealation was not ob- sei'ved in Colorado, but wingless queens were seen on 6 mounds and on trails from 24 June to 15 August between 0757 and 1742 hours. The greatest number on 1 mound was 7. Wing- less queens were usually sunounded by a group of workers on the mounds who often pulled them by their antennae and legs and some- times lunged at queens as if attacking them. Some were carried on the trails by workers. Dead wingless queens were observed being carried on a mound and a nearby dirt road. The number of wingless queens in 4 nests excavated in Colorado varied greatly: 0, 1, 32, and 198. Five of the 198 queens from 1 nest were found with numerous workers amid a clump of rabbitbrush roots 1.5 m away from the excavated moimd. Workers probably moved the queens along a trail from the main nest to a secondaiy nest at the rabbitbrush for safety during the prolonged excavation. Kannowski (1963) stated that many species of Formica have more than 1 dealate queen per colony, and Cole (1932) reported 2 or more per F. obsciiripes nest. The significance of the highly varible number of dealated queens per Colorado nest is unclear, and more excavations are necessary to determine the normal state of affairs. Observations of wingless queens on trails suggest that they may be transferred be- tween mounds or adopted by existing colonies after the nuptial flight (Weber 1935). Flight Season and Period The time of year during which alates of a species in a given area fly is termed the flight season. Kannowski (1963) noted that species such as F. obscuripes, with a large geographical 56 Great Basin Naturalist [Volume 56 distribution, may have a very long flight sea- son over their range. In Colorado, queens flew 1-8 July and males 1-9 July. Although others noted flights as early as 1 May (Clark and Comanor 1972) and as late as September (McCook 1884), flights were more common in June and July (Cole 1932, Weber 1935, Talbot 1972). Talbot (1972) noted that the flight sea- son varies greatly fi-om colony to colony in any year and that colonies may have 5-16 flights. Interestingly, she found that colonies in shel- tered nests or those on west-facing slopes flew later than those on open east slopes. Each ant species has a flight period — the time of day that flights take place. Kannowski (1959) reported that most species of Formica have early morning flights. Queens flew be- tween 0950 and 1141, and males between 0938 and 1101 in Colorado. Colorado flights did not begin as early (0500) or end as early (0750) as some reported by Talbot (1959) in Michigan, perhaps due to colder temperatures at high altitude in the morning. Reproductive activity subsided at Colorado nests between 1040 and 1107, or approximately at the same times (1030-1145) reported by T^ilbot (1972). Emergence and Positioning Reproductive emergence and positioning be- havior in Colorado is similar to that reported by Kannowski (1963) and Weber (1935). Mates emerged, walked around, and went back into the entrances before leaving the mound and climbing nearby structures. Workers some- times chased emerging alates or held onto their wings; at other times they seemed to ignore the sexuals. Males ignore winged queens at this time. Winged queens left Colorado mounds 1-8 July 1993 between 0818 and 1145 hours. Winged queens and males were found on the ground as far away as 7.85 m and 5.28 m from the mounds, respectively. Reproductives often climb prior to flight. In Colorado they climbed nearby sagebrush, rabbitbrush, lupine, and grass, as well as dead sagebrush and a fencepost protruding from mounds. At the most active mound thev climbed 3 sagebrushes, 0.48-0.89 m high, and 0.91-2.57 m away. Others have reported alates on nearby sagebmsh and rabbitbmsh (Clark and Comanoi- 1972), grass and herbs (Weber 1935), and tim- othy and bluegrass (TlUbot 1959). Although a number of Colorado reproduc- tives flew from their perches, many did not. Some queens descended 1-6 min after arrival, and one was pulled down by workers. Kan- nowski (1963) saw some alates wait longer (10-30 min) before flying from their perches. Tapping and blowing on perched queens did not induce them to fly. A correlation between temperature and emergence and positioning was noted by Tal- bot (1972). She reported that alates began leaving mounds when the air temperature reached 17.2° C and began climbing plants at temperatures above 18.3° C. Flights In Colorado alates flew from grass, sage- brush, rabbitbrush, and lupine; a few took off from the ground. Prior to flying, some queens released their front legs and fanned their wings, as reported by Kannowski (1963). On the other hand, Talbot (1959) reported that queens flew quickly with little preliminary wing fluttering. One Colorado queen flew east at least 13.1 m at an estimated altitude of 4 m. Another flight lasted about 20 sec at an estimated alti- tude of 9 m. Other winged queens moved away from mounds by alternately walking on the ground and making short, low flights between plants. One queen using this method moved 7.85 m away from a mound over a period of 37 min. Most queen flights were low and down- hill to the east. Males generally had short (2.5 cm-1.5 m), flitting or hovering flights about a meter above the ground, sometimes reland- ing on the same vegetation from which they departed. Reproductive activity was greatest in Colo- rado on clear, warm, windless days. All investi- gators agree that these are the most favorable conditions for flight. Wind supressed repro- ductive activity at 0918 hours on 3 Jul\- 1993. Weber (1935) noted alates leaving the nest when the air temperature was above 15.5° C, humidity exceeded 50%, and the sk-)- was clear. Others reported first flights at an air tempera- ture at least 5 ° C higher. A Colorado male flew at 22.7° C. Talbot (1972) reported that alates flew at temperatures between 20.5° C and 27.2° C, and Clark and Comanor (1972) saw flights between 20.5° C and 26.5° C, but at a relative humidity of only about 18%. Talbot (1959, 1972) noted that wind gusts, rain, low temperatures, and dark skies stopped flights, and wet grass and gray skies delayed flying. 1996] Thatching Ant in Coloiuuo 57 Colorado flights involved relatively few re- prodiictives, but reports in the literature vary considerably. Weber (1935) believed there is no marriage flight because only 1 sexual or a few sexuals fly at a time. Kannowski (1963) saw 1 mass flight, but noted most flights were sparse or moderate. Talbot (1959), on the other hand, reported that 695 females and an esti- mated 4500 males flew over time. Rates of fly- ing of 4-14 queens/min and 1-10 males/min have been reported (Talbot 1959, Clark and Comanor 1972). There appears to be no agreement on the flight pattern. Talbot (1959) noted that most queens flew downhill and westward, but some had short, sporadic flights from plant to plant or to the ground as sometimes observed in Colorado. Colorado flights were generally at low altitude (estimate 4-9 m), downhill, and eastward toward the sun. Kannowski (1963) also noted that alates fly in the general direc- tion of greatest light intensity. Others report that flights are often upward and out of view (12 m or more; Weber 1935, Kannowski 1963, Clark and Comanor 1972). Swarming and Mating Swarming is the process whereby alates aggregate to mate in the air or on the ground and vegetation (Kannowski 1963). Most swann- ing and mating in Colorado occurred 2-6 July 1993 bet^veen 1008 and 1125 hours on rabbit- brush 4.01 m from 1 mound. Mating was also obsei^ved on rabbitbiaish beside another mound on 2 July and 6 July 1993. Talbot (1972) noted swarming earlier in the year and over a longer time period, namely, 4-17 June between 0700 and 1200. Swarming in Colorado was similar to that described by Kannowski and Johnson (1969) and Talbot (1972). Queens anived first on rab- bitbrush, followed by males. Queens perched on the upper parts of plants often with their heads down and their abdomens pointing upward or toward the nest. Presumably they emit a pheromone to attract males (Kannowski and Johnson 1969, Walter et al. 1993). Once the female's pheromone is detected, males fly upwind to the general location of the female, fly quickly from stem to stem until they find her, alight, and then attempt to mate (Kan- nowski 1963). After mating, males usually fly off while the queen remains and sometimes inspects her abdomen. Up to 7 in copulo alates were noted at 1 time at the Colorado swarming site 4.01 m away, 6 pairs on rabbitbrush and 1 pair on an adjacent lupine. Some pairs fell off the plants. One queen appeared to mate 2 or 3 times. Kannowski (1963) reported a queen mating 4 times. Two Colorado males tried to simultane- ously mate with a queen for 1 min 40 sec and remained attached to each other for 20 sec after the queen left. Talbot (1972) noted 3 or 4 males tiying to mate a queen, and Kannowski (1963) reported a single male may mate sev- eral times before flying away. The durations of 6 Colorado matings ranged from 1 min 40 sec to 3 min 40 sec (mean = 2 min 43 sec), or within the 1- to 5-min dura- tions reported by Talbot (1972). Talbot (1959, 1972) noted larger, more diverse, and more heavily populated swarming areas than the small rabbitbrush area in Colo- rado. Some of her swarming areas were over short grass; others were on shrubs. One swami- ing area involved thousands of males hovering over hundreds of females from 3 colonies and covered an oval-shaped area 27.5 X 11 m. Males usually flew near grass level, but some- times as high as 1.2-1.5 m. Another swarming area shifted somewhat from day to day and increased to approximately 41.3 X 32.1 m. She found that these areas were maintained throughout the flying season, and some were used year after year. Conclusions Preliminary studies of the reproductive behavior of the thatching ant, F. obscuripes, in Colorado are in general agreement with the literature. Time constraints on our seasonal obsei"vations probably explain why we did not observe reproductive behavior as early in the year as that reported in the literature. The most notable finding was the paucity of repro- ductive activity: swarming and mating were obsei-ved only 2-6 July 1993; 9 of 98 mounds (9%) in the area had winged reproductives; mating occurred near 2 mounds (2%); and a swaniiing area was found 4.01 m from 1 mound (1%). The numbers of males and winged queens were relatively low and the swarming area was small. Other notable findings were the highly variable number (0-198) of dealated queens per nest and the almost complete absence of winged alates in excavated nests. 58 Great Basin Naturalist [Volume 56 Further studies are needed to determine whether our findings are anomahes or whether they represent the normal state of affairs for this species at high altitude. Acknowledgments I thank 4 University of Scranton students, John Bridge, Tom Sabalaske, Antliony Musingo, and Jeanne Rohan, who conducted fieldwork in Colorado in 1993-94. Support for this re- search was provided by a grant from the Howard Hughes Medical Institute through the Undergraduate Biological Sciences Edu- cation Program. Barry C. Johnston, ecologist at the U.S. Forest Service in Gunnison, Colo- rado, identified plant specimens. Literature Cited Cherix, D., et at. 1993. Attraction of the sexes in Formica htgiibris Zett. Insectes Sociaux 40: 319-324. Clark, W. H., and R L. Comanor. 1972. Flights of the western thatching ant, Formica obsciiripes Forel, in Nevada. Great Basin Naturalist .32: 202-207. Cole, A. C, Jr. 1932. The thatching ant, Formica obscuripes Forel. Psyche 39: 30-33. Gregg, R. E. 1963. The ants of Colorado. University' of Colorado Press, Boulder 792 pp. Herbers, J. M. 1978. Trends in sex ratios of the reproduc- tive broods of Formica obscuripes. Annals of the Entomological Society of America 71: 791-793. . 1979. The evolution of sex-ratio strategies in Hymenopteran societies. American Naturalist 114; 818-8.34. HOLLDOBLER, B., AND E. O. WlLSON. 1990. The ants. The Belknap Press of Harxard University Press, Cam- bridge, MA. 732 pp. Kannowski, R B. 1963. The flight activities of formicine ants. Symposia Genetica et Biologica Italica 12: 74-102. Kannovvskl R B., and R. L. Johnson. 1969. Male patrol- ling behaviour and sex attraction in ants of the genus Fonnica. Animal Behaviour 17: 42.5—129. McCooK, H. C. 1884. The nifous or thatching ant of Dakota and Colorado. Proceedings of the Academy of Nat- ural Sciences, Philadelphia, part 1: .57-6.5. Talbot, M. 1959. Flight activities of two species of ants of the genus Formica. American Midland Naturalist 61: 124-132. . 1972. Flights and swarms of the ant Formica obscuripes Forel. Journal of the Kansas Entomologi- cal Society 45: 254-258. Walter, F, et al. 1993. Identification of the sex phero- mone of an ant, Formica lugubiis. Naturwissenschaften 80: 30-34. Weber, N. A. 1935. The biology of the thatching ant Formica obscuripes Forel in North Dakota. Ecological Monographs 5: 16.5-206. Wheeler, G. C., and J. Wheeler. 1963. The ants of Nortli Dakota. University of North Dakota Press, Grand Forks. 326 pp. Received 17 Jaituanj 1995 Accepted 21 June 1995 Great Basin Naturalist 56(1), © 1996, pp. 59-72 TRACHYTES KALISZEWSKU, N. SE (ACARI: UROPODINA), FROM THE GREAT BASIN (UTAH, USA), WITH REMARKS ON THE HABITATS AND DISTRIBUTION OF THE MEMBERS OF THE GENUS TRACHYTES Jerzy Bloszyk^ and Pawe4 Szymkowiak^ Abstract. — Trachytes kaliszewshii, n. sp., is described fiom the Great Basin, Utah, USA. SEM photography illustrates moiphological detail. An annotated list is included of cuirently recognized species of the genus Trachytes, with comments on their distribution and habitat characteristics. Key words: mites. Trachytes kaliszewskii, Uropodina, Great Basin, Utah. Mites of the genus Trachytes Michael, 1894, are a morphologically distinct entity of the Uropodina. The genus consists of 31 species known mainly from the Palearctic region of Europe and Japan. Wisniewsld and Hirschmann (1993) mention two species from the USA: T. aegrota (C. L. Koch, 1841) and T. traegardhi (Hirschmann and Zirngiebl-Nicol, 1969). Tra- chytes traegardhi is regarded as nominum nudum. The USA listing for T. aegrota is con- sidered either a mistake in determination or an accidental introduction. Taxonomic studies on mites of the genus Trachytes are found in Hirshmann and Zirn- giebl-Nicol (1969), Hutu (1983), and Pecina (1970). Information on their biology, ecology, and zoogeography is found in Athias-Binche (1978, 1979, 1980, 1981, 1985), Pecina (1980), Bloszyk (1980, 1982, 1984, 1985, 1990, 1991, 1992, 1993), Bloszyk and Athias-Binche (1985), Bloszyk and Miko (1990), Bloszyk and Ols- zanowski (1985a, 1985b, 1985c, 1986), and Bloszyk et al. (1984). We found a new species of the genus Tra- chytes in soil collected from Rock Canyon near Provo, Utah, USA. It is most similar to those described by Hiramatsu (1979, 1980) from Japan: T. aoki and T. onishii. Moiphological dif- ferences between our species, those mentioned from Japan, and Trachytes aegrota are shown in Table 1. Our new species is dedicated to the Polish acarologist. Dr. Marek Kaliszewski, who was a faculty member at Brigham Young Uni- versity/, Provo, Utah, USA, until 1993, when he died tragically in an automobile accident. Systematic Status of the Genus Trachytes Michael SUPERFAMILY. — Polyaspidoidea sensu Athias- Binche & Evans, 1981 Family. — Trachytidae Tragardh, 1938 Genus. — Trachytes Michael, 1894 Type species. — Celano aegrota C. L. Koch, 1841 { = Trachynotus pyrifonnis Kramer, 1876) Mites of middle size, strongly sclerotized, dorsoventrally flattened. Idiosoma triangular, "vertex" distinct with smoodi or slightly seiTated edges. Corniculus simple, laciniae longer than corniculi. Hypostomatic setae: hi very long, simple; h2 shorter than hi, simple; h3 very long, massive; h4 very short, serrated. Fixed digit of the chelicera longer than moveable digit, shaiply pointed distally. Base of tritoster- num wide, not covered by coxae I. Trachytes kaliszewskii, n. sp. Diagnosis. — The fonn of the body is typical for the genus Trachytes Michael. Vertex with lamella. Dorsal shield with polygonal patteiTi and irregular cavities in central part (similar to T aegrota). Marginal shield is not divided as in European species, without polygonal pattern. Dorsal setae long and massive. Small pygidial shield present in female. Epigynial shield tiape- zoidal with net pattern, front margin slightly convex and produced laterally into little corns. Sternal setae short. Operculum of male rounded, with a pair of long genital setae. Ventroanal shield separated from sternal and metapodal shields by a wide zone of interscutal membrane. 'Department of Animal Taxonomy and Ecolog\', Adam Mickiewicz University, Szamarzewskiego 91A, 60-569 Poznari, Poland. 59 60 Great Basin Naturalist [Volume 55 Ventral setae long. One pair of paranal setae. Postanal seta present. Adult female. — Length of idiosoma 900- 907 ^tni, width 535-574 /xni. Dorsmn: Lamellae with characteristic pat- tern. Marginal shield not divided posteriorK; \\ith irregular caxities in posterior part. Dorsal shield with poKgonal pattern lateralh' and irregular cavities in central and posterior parts (Figs. 1, 10, 11). Dorsal setae long and mas- sive. Two pairs of setae on vertex; no unpaired medial dorsal setae. Marginal setae on small scutellae; 4 pairs of setae situated medialK on marginal shields. Pygidial shield \\ith pattern as on marginal shield. Veiitruin: Sternal shield (Fig. 2) fused to parapodals. Ventroanal shield separated from steiTial and metapodal shields by a zone of in- terscutal membrane bearing 4 pairs of platelets (Fig. 13). Sternal shield smooth, bearing 5 pairs oi short stenial setae. Setae: stl situated between coxae II at the \eye\ of their front margins; st2 and st3 placed abo\ e anterior edge of epig> - nium; st4 and st5 situated laterally of epig)- nium. Opisthogastric setae generally long, simple or delicateh' serrated, most anterior pair short, similar to sternal setae. First pair of opisthogastric setae situated below posterior margin of epig\iiium, 2nd pair on metapodal shields, with 4 pairs on interscutal membrane and 2 pairs on ventroanal shield. One pair of adanal setae; short and serrated. Postanal seta long. E.xopodal and metapodal shields with o\'al or irregular cavities. Ventroanal shield smooth anteriorly, with polygonal patterns in the posterior regions. Epigynial shield trapezoidal, with front margin slightly convex and produced laterally into little corns; measurements: 175-199 /xm length and 137-156 fim width (N = 3). Sur- face of epig>aiium with delicate polygonal net in anterior and central areas. Peritrema simple, without poststigmatic section, extending from the level of the poste- rior border of the foramen pedale III (with stigma) to beyond coxae II. GnatJwsoma: Laciniae (internal mala) longer than corniculi, serrated. Hypostomatic setae (Fig. 4) smooth except for setae /i4 which are delicately serrated; hi very long, /j2 shorter than hi, /i3 long as hi but more massi\'e, h4 shorter than h2. Three transversal rows of h\pognathal denticles between setae ^3 and /i4. Appendages: Shape of chelicerae typical for Trachytes; fixed digit of the chelicera longer than moxeable digit, shaped distalK. Pedipalp xentral, setae of trochanter (vl, v2) massive and serrated (Fig. 5). Shape of legs tvpical for family. Tarsi of legs II-IV \\'ith 4 long setae (3 times longer than Table 1. SiininiaiA of major differences between closely related Trachytes species. Character T. acp'ota T. (loki T. onishii T. kaliszewskii Sex parthenogenic bisexual ■? bise.xual Female Lamella transverse trans\erse trans\'erse oblong Setae on interscutal nienibr me absent present absent present Unpaired mediodorsal seta present absent absent absent Bod\- incasmements (in fim ) 600 X 68.5 400 X 450 400 X 600 535-574 X 900-907 H\postomal setae h3 simple massive massive massive Setae on ventroanal shield different ecjual equal equal Epig\-ninm smootii smooth with poKgon al net V'ential seta on metapodal s lields long short short long Seta Pa short short short long 1995] Tlh\CHYTES KALISZEWSKII, N. SH, I-^HOM UTAH 61 Fig, 1. Trachytes kaliszeicskii. n. sp., dorsal view of female idiosoma. 62 Great Basin Naturalist [Volume 55 Fig. 2. Trachytes kaliszcwskii, n. sp., \ential view oi icinalf idiosonia. 1995] TrACHYTI'S hiMJSZEWSKII, N. SH, FKOM UlAll 63 Fig. 3. Trachytes kaliszewskii, n. sp., ventral view of male idiosoma. 64 Great Basin Naturalist [Volume 55 Figs. 4—5. Trachytes kaliszew.skiL n. sp., female: 4, gnatliosoma, ventral view; 5, ventral setae of palpal trochanter. others), small claws, and a veiy long distal seta. Shape of dorsal setae on tarsus, tibia, genu, and femur of legs I as in the genera Polijaspis and Polyaspimis. Chaetotaxy of legs I and IV is shown in detail in Figures 6 and 7. Sexual dimorphism observed on femora II (Figs. 8, 9). Adult male. — Bodv measurements 830- 862 Aim X 538-540 )Ltm.' Dorsum: Male dorsum slightly changed in posterior part; pygidial shield absent (Fig. 12). Sculpture and dorsal chaetotaxy as in the female. Ventnim: SteiTial shield with numerous oval cavities and bearing 5 pairs of short sternal setae (Fig. 3). Genital operculum rounded (74-79 X 72 fxm), located a little below coxae IV, with 1 pair of long genital setae. Opistho- soma separated b)' transverse suture with in- terscutal membrane. Seven pairs of long ven- tral setae on rounded platelets; 1st pair short, located below operculum. With 1 pair of deli- cately serrated adanal setae and long unpaired postanal seta (Pa). Opisthosoma with poly- gonal sculpture on metapodal and anal shields and small oval cavities on central portion. Deutonymph. — Body measurements 624 X 396 lam. Dorsum: Dorsum with polygonal pattern (Fig. 15). Podonotal shield trapezoidal, fused with lamellae. Mesonotal shields large, trian- gular, with 4 setae. Pygidial shield arched, with 2 pairs of setae. Dorsal setae strong, mas- sive. Setae on interscutal membrane and mar- ginal setae inserted on small platelets. Ventrum: Ventrum with polygonal pattern (Fig. 16). Sternal shield elongated, with 5 pairs of short sternal setae; most posterior pair deli- cately senated. Opisthogastric setae situated on interscutal membrane, delicately serrated, sit- ting on small platelets. Large ventroanal shield with 2 pairs of short adanal setae (Ad), postanal seta (Pa) longer than Ad; both setae serrated. Protonymph. — Bodv measurement 528 X 295 Aim. Dorsum: Dorsum with poKgonal pattern (Fig. 17). Podonotal shield trapezoidal. Meso- notal shields large, oval-triangular, without setae. P\gidial shield arched, with 2 strong, massive setae. Dorsal setae strong, massive. No setae on intersutal membrane. Marginal setae numerous, inserted on small platelets. Ventrum: Sternal shield smooth, elongate, with 4 pairs of simple sternal setae (Fig. 18). Four massive, serrated opisthogastric setae 1995] Trachytes kaliszewsku, n. sf., from Utah 65 Figs. 6-9. Trachytes kaliszewskii, n. sp., legs chaetotaxy: 6, leg I of female; 7, leg IV of female; 8, chaetotaxy of male femora II; 9, chaetotaxy of female femora II. situated on intersutal membrane. Large ven- troanal shield with 1 pair simple adanal setae and a long postanal seta. Material examined. — All specimens were collected from soil under a maple tree in Rock Canyon near Provo, Utah, 10 September 1992; leg. J. Bloszyk (holotype and 5 paratype females, 7 paratype males, 7 deutonymphs, 5 proto- nymphs). The holotype is deposited in the Canadian National Collection, Biosystematics Research Cenbe, Ottawa, Canada. Paratypes are deposited 66 Great Basin Natur.'VLIst [Volume 55 Figs. 10-14. Trachytes kaliszewskii, n. sp.; 10, dorsal polygonal pattern of feniiile (550X); 11, female, general dorsal view (llOX); 12, posterior part of male idiosoma (220X); 13, opisthosoma of female (200X); 14, marginal setae of female (750X). in the Monte L. Bean Life Science Museum, Brigham Young Univer.sity, Provo, Utah, USA; in CSIRO, Canberra, Austraha; and in J. Bloszyk's collection (Acarological Association, ul. Lisowsldego, 16/1, 61-606 Poznari, Poland). List of the Trachytes Species with Remarks on Distribution and Habit.at Preferences Hirshmann (1993) listed 31 species refer- able to the genus Trachytes. In view of the 1995] Trachytes kaliszewsku, n. sp., from Utah 67 Fig. 15. Trachytes kaliszewskiu n. sp., dorsal view of deutonymph idiosonia. above, we recognize 31 species in the genus Trachytes as follows^: Trachytes aegrota (C. L. Koch, 1841) is one of the most numerous Uropodine species in -Some data from Poland originate from an unpublished investigation carried out by J. Bloszyk in the thematic program Bank of Invertebrate Fauna; data on the distribution may be found in Hirschmann (1979, 1993). Hufu (1973, 1983), Hiramatsu (1979, 1980), and Athias-Binche (1981). central Europe. This species is parthenogenetic and nonphoretic; males are rarely found (sex ratio is 1:10,000). This eurytopic species lives in all kinds of biotypes, but it prefers forest lit- ter. It most often occurs below 500 m elevation but is considered a tychoalpine species (i.e., lives in the mountains as well as the lowlands). In Poland the spring-summer season is the best time to observe the larva. 68 Great Basin Natuiulist [Volume 55 Fiu;. 16. Trachiilcs kalisznvskiL n. sp., vcMitral view ol dcutonymiili idiosoina. Trachytes aoki I liramatsu, 1979. Japan. In litter. Trachytes arcuatus Hirschmann and Zirn- gicbl-Nicol, 1969. Austria, Koniania, H unwary. Habitat unknown. Trachytes hah)'clioalpine species. The best time to observe the lai'va is during the spring-summer season. Trachytes pecinaia Iluju, 1983. Romania. In htter. Trachytes pi Berlese, 1910. West and Cen- tral Europe. In htter. Trachytes romanica Huju, 1983. Romania. In litter. Trachytes splendkla Huju, 1983. East Car- pathian species — Romania, Poland, Slovakia. In litter and moss. Trachytes stammeri Hirschmann and Zirn- giebl-Nicol, 1969. Locality and biotype un- known. Trachytes tesquorwn Pecina, 1980. Czech Republic. In grass. Trachytes traeghardi Hirschmann and Zirn- giebl-Nicol, 1969. Locality and biotype un- known. Trachytes tubifer Berlese, 1914. Italy, Austria. In litter. Trachytes welhournia Moraza, 1989. Spain. In litter. Trachytes wisniewski Huju, 1983. Romania. In litter. Acknowledgments Dr. J. Bloszyk wishes to thank the adminis- trators and workers of the Department of Zoology and Monte L. Bean Life Science Museum at Brigham Young University (BYU), Provo, Utah, USA, for providing facilities and an atmosphere that encouraged scholarship. The authors are greatly indebted to Dr. Richard Baumann, Department of Zoology, BYU, for his kind help in reviewing the manu- script and for his judicious remarks and advice; and to Dr. John S. Gardner, electron micros- copist from BYU, for his valuable scanning photography. This study was completed with financial assistance from the Department of Zoology, Brigham Young University, and Acarological Association (Poznaii, Poland). Literature Cited Athias-Binche, E 1978. Etude quantitative des Uropodes edaphiques de la hetraie de la Tillaie en foret de Fontainebleau (Acariens, Anactinotriches). Revue d'Ecologie et de Biologie du Sol 1.5: 67-88. . 1979. Effects of some soil features on a uropodide mite community in the Massane forest (Pyrenees- Orientales, France). Pages 567-.573 in Recent advances in acarology — proceedings of the 5th International Congress on Acarology. . 1980. Contribution a la connaissance des Uro- podidcs librcs (Arachnides: Anactinotriches) de quelques ecosysternes forestiers Europeens. These d'Etat, Universite de Paris VI, Paris. 1981. Differents types de structures des peuple- ments d'Uropodides cdaphicjues de trois ecosysternes forestiers (Arachnides: Anactinotriches). Acta Oeco- logica-Oecologia Ceneralis 2: 153-169. BiDSZYK, J. 1980. Mites of the genus Trachytes Michael, 1894 (Acari: Mesostigmata) in Poland. Prace Komisji Biologicznej, PTPN 54: 5-52. . 1982. Uropodina Polski (Acari, Mesostigmata). Thesis, Biblioteka Glowna UAM, Poznan. 543 pp. . 1984. Altitudinal distribution of the Uropodina fauna (Acari) in Poland. Przeglad Zoologiczny 28: 69-71. . 1985. Contribution to knowledge of the mites in the mole nests {Talpa eiiropea L.). I. Uropodina (Acari, Mesostigmata). Przeglad Zoologiczny 29: 175-181. . 1990. Fauna of Uropodina mites (Acari: Meso- stigmata) of decayed tree stumps and hollows in Poland. Zeszyty Problemowe Post^pow Nauk Rol- niczych 373: 217-2.35. . 1991. State of investigation of Uropodina (Acari: Anctinotrichida) in Polish National Parks. Parki Nar- odowe i Rezei-waty Przyrody 10 (1,2): 115-122. . 1992. Materials to the knowledge of the acaro- fauna of Roztocze Upland. III. Uropodina (Acari: Mesostigmata). Fragmenta Faunistica 35 (11): 323-344. . 1993. Uropodina (Acari: Mesostigmata) of pine forests in Poland. Fragmenta Flumistica 36 (11): 175-183. BiDSZYK, J., AND E Athias-Binche. 1985. Urban ecosys- tems and ecological studies: example of soil uropo- did community in Poznan Park. Pages 278-282 in Soil fauna and soil fertility — proceedings of the 9th International Colloquium on Soil Zoology. BU5SZYK, J., AND L. MiKO. 1990. Podna fauna Pienin. I. Uropodina (Acarina: Anactinotiichida). Entomologicke Problemy 20: 21-47. BtDSZYK, J., AND Z. Olszanowski. 1985a. Contribution to the knowledge of mites of birds nests. I. Uropodina and Nothroidea (Acari: Mesostigmata et Oribatida). Przeglad Zoologiczny 29: 69-74. . 1985b. Mites of the genus Trachytes Michael, 1894 (Acari: Mesostigmata) in Poland. III. Sporadic appearance of males in some populations of partheno- genetic species. Przeglad Zoologiczny 29: 313-316. . 1985c. Contribution to the knowledge of biology of some Uropodina (Acari: Anactinotrichida) juvenile stages. Przeglad Zoologiczny 29: 487-490. _. 1986. Contribution to the knowledge of mites of ant hills in Poland (Acari: Uropodina). Przeglad Zoo- logiczny .30: 191-196. BiDSZYK, J., I. ChOJNACKI, AND M. Kaliszewski. 1984. Study on the mites of the genus Trachytes Michael, 1894. I. Seasonal population changes of Trachytes aegrota (Koch, 1841) in deciduous resei^ves "Jakubowa" and "Las Gr^dowy" near Pniewy, Poland. Pages 893- 900 in D. A. Griffiths and C. E. Bowman, editors, Acarology VI. Volume II. Ellis Hai"wood, Chichester Hirschmann, W. 1979. Bestinmibare Uropodiden-Arten der Erde (ca. 1200 Arten), geordnet nach dem Gang- system Hirschmann, 1979 und nach Adulten Gnippen (Stadien, Heimatliinder, Synonym, Literatur). Acarolo- gie (Niiemberg) 26: 15-57. 72 Great Basin Naturalist [Volume 55 HmscHMANN, W, AND I. ZiRNGlEBL-NicOL. 1969. Gangsys- tematik der Parasitifonnes Teil 57. Typus der Gattung Trachytes Michael, 1894. Acarologie (Niirnberg) 12; 76-81. HiRAMATSU, N. 1979. Gangsystematik der Parasitiformes Teil 3322. Stadien einer neuen Traclujtes-Art aus Japan (Uropodini, Uropodinae). Acarologie (Niirn- berg) 25: 76-77. . 1980. Gangsystematik der Parasitiformes Teil 360. Teilgang und Stadien von 2 neuen Trachytes -Arten aus Japan (Uropodini, Uropodinae). Acarologie (Niirn- berg) 27; 26-27. HUXU, M. 1973. Gangsystematik der Parasitifonnes Teil 145. Zur Kenntnis der Uropodiden-Fauna Rumaniens Neue Uropodiden-Arten der Gattungen Trachytes Michael, 1894, Dinychus (Kramer, 1886) und Tra- chyuropoda (Berlese, 1888). Hirschmann u. Zinigiebl- Nicol 1961 nov. comb. Acarologie (Niirnberg) 19; 45-51. . 1983. Gangsystematik der Parasitiformes Teil 428. Teilgange, Stadien von 6 neuen Trachytes- hxien aus Rumanien und Schweden (Uropodini, Uropodinae). Acarologie (Niiniberg) 30; 51-66. PeCi.na, P 1970. Czechoslovak uropodid mites of the genus Trachytes Michael, 1894 (Acari, Mesostigmata). Acta Universitatis Garolinae, Biologica 1969; 39-59. . 1980. Additional knowledge of members of the genus Trachytes Michael, 1894 (Acari, Mesostig- mata) from Czechoslovakia. Acta Universitatis Gar- olinae, Biologica 1978; 389-407. WiSmewski, J., a.nd W. Hirschmann. 1993. Gangsystem- atik der Parasitiformes Teil 548. Katalog der Gang- gattengen, Untergattungen, Giiippen und Alien der Uropodiden der Erde. Acarologie (Niirnberg) 40; 1-220. Received 22 September 1994 Accepted 25 September 1 995 Great Basin Naturalist 56(1), © 1996, pp. 73-84 PRODUCTIVITY, FOOD HABITS, AND ASSOCIATED VARIABLES OF BARN OWLS UTILIZING NEST BOXES IN NORTH CENTRAL UTAH Sandra J. Loonian^ Dennis L. Shirley^, and Clayton M. White^ Abstract. — Productivit)' and food habits of the Bam Owl {Tyto alha) utilizing nest bo.xes in Juab, Utah, and Salt Lake counties, Ut;ili, during 1979-1984 were examined. Average clutch size was 5.8 eggs for the 6-yr period; mean number fledged was 3.9 yoimg per successfiil nest. While severe weather during the 1981-82 winter did not result in a significant decrease in productivit)' during the 1982 breeding season, it may have resulted in a significant oveiproduction of female \oung. BaiTi Owls in north central Utah fed almost exclusively on mammalian species, particularly Microtus spp. Differ- ences in clutch size between areas and years may be a response to availability as well as abundance of prey. Key words: Barn Owl food. Barn Owl reproduction, nest boxes, Utah, Tyto alba. The Barn Owl {Tyto alba) is a nearly cosmo- politan species that uses diverse nest sites, in- cluding man-made ones (Voous 1988). Although Barn Owls were reported in Utah as early as 1899 (Smith and Marti 1976), they were con- sidered uncommon and rare breeders prior to 1976 (Smith and Marti 1976). The first Barn Owl nesting record was reported by Behle (1941) near Kanab in Kane County. Woodbury et al. (1949) proposed that Barn Owls were probably residents and widely distributed in valleys and lower elevations throughout the state. Smith et al. (1972, 1974) and Smith and Marti (1976) presented information on Barn Owl food habits, nesting ecology, and distribu- tion throughout the state. While these studies indicated prey was abundant in irrigated agri- cultural areas, nesting sites were not adequate in those areas to allow growth of the popula- tion (Marti et al. 1979). Marti et al. (1979) installed 8 nest boxes in abandoned concrete silos in north central Utah during 1977 and an additional 22 in 1978 in an effort to increase numbers of nesting Bam Owls. Of those boxes, 50% were used by breeding owls in 1977 and 80% in 1978. A total of 154 young fledged fiom nest boxes during the 2 yr In 1979 a similar program of installing nest boxes in silos was adopted in central Utah by the Utah Division of Wildlife Besources (UDWR). Between 1979 and 1984, 41 nest boxes were installed in Juab, Utah, and Salt Lake counties. An ongoing investigation of Barn Owl population and feeding habits was undertaken in 1979. Herein we document reproductive activities, dispersal, sui^vival, and food habits of Barn Owls utilizing these nest boxes from 1979 to 1984. Study Area This study was conducted on the 15- to 25- km-wide strip of farmland and suburban area between the Wasatch Mountains on the east and Utah Lake on the west. The climate is arid, characterized by hot, diy summers, cold win- ters, and cool, wet springs. Precipitation aver- ages 40 cm annually, falling mainly as winter snow. Extensive agricultural irrigation and the presence of a large freshwater lake have cre- ated broad areas of habitat, especially for voles {Microtus spp.), a major Barn Owl prey. Trees occur sporadically along rivers and irrigation canals and on farmsteads. Preliminary surveys by UDWR in 1979 revealed that 50 silos were used for roosts by Barn Owls, as indicated by presence of regur- gitated pellets, fecal stain, and/or presence of owls. Silos were in rural or semirural areas throughout the counties and generally close to corn or alfalfa fields; a few were located in suburban areas within 2 km of an agricultural area (dairy or cattle ranch). Silos not used by fanners provided roosting owls protection from Department of Zoolog)-, Brigham Young University, Prove, UT 84602. Present address: Department of Biological Sciences, and Institute of Arctic Biology, Universit\' of Alaska, Fairbanks, AK 99775. ^Utah Division of Wildlife Resources, Regional Office, Springville, UT S4663. Department of Zoology, Brigham Young University, Provo, UT 84602. 73 74 Great Basin Naturalist [Volume 56 predation and disturbance; however, none pro- vided adequate nest sites. Most bams and other structures in the area also lacked adequate nesting sites. Forty-one wooden nest boxes were built, after Marti et al. (1979), and installed between 1979 and 1984 (18 installed in 1979, 6 in 1980, 5 in 1981, 9 in 1982, and 1 each in 1983 and 1984). Three nest cavities (2 in silos and 1 in a school building) were discovered and moni- tored during these years; data from these sites are included herein. Methods All nest boxes were examined at least once monthly throughout the year to determine presence of adult owls or fresh regurgitated pellets. Behavior of adults was recorded on all visits, and adults were caught and banded if possible. Pellets were collected during each visit. Presence of cached food and prey remains inside boxes and on silo floors was noted. Sites where nesting occurred were visited appro.ximately eveiy 2 wk throughout the breed- ing season, Januaiy-August, in 1979-1981 and 1984. During 1982 and 1983, a study to develop a sexing technique (Looman 1985) was started, and therefore we increased our efforts and vis- ited active nest boxes more frequently (usually once a week) throughout most of the breeding season (May-August) during these years. Nests were considered active if an adult owl was obseived in the nestbox or signs of recent occu- pation were evident (i.e., eggs, eggshells, fresh pellets in nestbox, nestlings). Onset of egg lay- ing was determined by direct observation or by backdating from known-age nestlings or date of fledging. For backdating, we used 30 d as an incubation period (Smith et al. 1974, Marti 1992), with 2 d between individual eggs (Bunn et al. 1982). Clutch size and productivity' (fledgling number) data were determined by direct obsei^vation. Behavior of adults and nesdings was recorded at each visit. All young were banded when approximately 5-6 wk old, and during 1982 and 1983 each young was weighed at fledging (approximately 8 wk) and sexed according to the sexing method described by Looman (1985). While pellets collected during a 5-yr period (1979-1983) were available for food habit assessment, only pellets collected in 1982 and 1983 were separated into 4 time group- ings, each representing a seasonal period of Barn Owl activity and roughly corresponding with 1 of the 4 seasons. The spring period (March-May) corresponded with early repro- ductive activities, summer (June-August) with adult attentiveness to fledgling but still depen- dent young. The autumn period (September- November) included abandonment and subse- quent dispersal of most young, and winter (December-Februaiy) corresponded with the period that remaining owls moved into well- protected residential structures. Pellet analysis followed Marti (1974). Verte- brate prey remains were identified by compar- ison with mammal (see Durrant 1952) and bird specimens at M. L. Bean Museum, Brigham Young University. Prey weights for estimation of biomass were means obtained from these specimens and from reported weight esti- mates (Marti 1974, Steenhof 1983). Estimated age of prey for use in biomass calculations was based on cranial features (ossification of sutures and auditory bullae and tooth eruption and wear). Diversity of Barn Owl diet was determined using the multivariate statistical package MVSP (Kovach 1987). To allow comparisons with other published diversity indices of Bam Owl diet, diversity indices were calculated using the modified Shannon-Weiner diversity^ index formula n = -i(pi){\ogpi\ i=l where s is the number of species and p,- is the proportion of the number of indixiduals in the ith species. Species evenness (E = H/log2; Magurran 1988) was also calculated. Results Breeding Chronoloy Dates of onset of egg laying range from earh- Januaiy (date obtained by backdating) through early August, with 36% commencing egg lay- ing during the first half of March and 25% beginning in late FebiTiaiy (Fig. 1). The earliest date on which eggs were obsened in a box was February 12, the latest September 14 (eggs and nestlings observed). Length of the nesting season for this popu- lation, defined as the period from deposition of first egg to fledging of last young, averaged 6.6 mon for the 5->'r period (range 4.0 mon in 1996] Barn Owls in North Central Utah 75 15 -T 10 - 5 - ONSET OF EGGLAYING 1979-1983 23 123 9023 9023 1 1 1 1 12 012 012 012 012 012 123 90123 123 90123 3 3 3 1 3 1 3 123 123 9 9012 0 0 0 903 23 1-15 16-31 Jan 1-14 16-28 Feb 1-15 16-31 Mar "1 1 1-15 16-30 Apr NESTS 9=1979 n= 7 0=1980 n=16 1=1981 n=20 2=1982 n=19 3=1983 n=23 1213' 1-15 16-31 May — I 1-15 Jun — T" 1-15 July 1-15 Aug 'Onset of second clutch Fig. L Dates of first egg laying by Bam Owls in north central Utah, 1979-1983. 1979 to 9.8 mon in 1983). This is long com- pared to 5.3 mon in south Texas (Otteni et al. 1972) and in Utah (Smith and Marti 1976) dur- ing 1974 and 1975; no late autumn nests were found, however, as have been previously found in Utah (Smith et al. 1970). Individual nesting cycles, from deposition of first egg to fledging of last young in the nest, were approximately 3.3 (3.25 ± 0.2, n = 10) mon in length. Where egg deposition intervals were known, the intei-val was 2 d between eggs (2.1 ± 3, n = 10); this is similar to deposition data (2.3 d) found for Barn Owls in Springville during 1973 (Smith et al. 1974). Known incubation times averaged 32.3 d (±3 d, n = 10). Fledging occurred at 62 d (±4 d), and young remained in the area until approximately 13 wk of age. Similar incubation and fledging times are reported for Barn Owls elsewhere (Pickwell 1948, Reese 1972, Smith et al. 1974). Nests Owls made no attempt at nest construction. However, prenesting behavior of adults, in which they spent a great deal of time at the nest site, resulted in a layer of broken down pellets, incidental feathers, and fecal material which produced a soft bed for eggs. Eggs were laid in a shallow area in the middle. Productivity Four hundred twenty-eight young were fledged from 104 (106 including 2nd broods) nest boxes over a 6-yr period (Table 1), averaging 3.9 young/box with a nest failure rate of 16.6%. Productivity ranged from 0.8 young fledged/ box (2.0 young/active box) and a failure rate of 25% in 1979, to 4.37 young fledged/box (5.4 young/active box) and a failure rate of 9.1% in 1981. Mean clutch size for the 5-yr period was 5.8 eggs/clutch (±1.72) and ranged from 5.3 (1979, 1983) to 6.5 (1981) (Table 2). Modal clutch size was 7 (22%); modal brood size was 7 (21%) (Table 3). Clutch size in 19 nests in 1982 ranged from 2 to 10 eggs and averaged 5.8 (±2.0); broods in these nests ranged from 2 to 7 and averaged 4.0 (±1.9) young hatched/ nest. Thirty-one percent of eggs failed to hatch, and nestling mortality was approximately 8%. Productivity in 16 nests where young success- fully fledged averaged 4.4 (±1.4); however, productivity fell to 3.7 (±2.1) young fledged/ total nesting attempt. Clutch size in 23 nests in 1983 ranged from 3 to 9 and averaged 5.3 (±1.8) (Table 2). Brood number ranged fi-om 2 to 8 and averaged 3.95 (±2.1) young hatched/nest. Twenty-five per- cent of the eggs failed to hatch, and nestling 76 Great Basin Naturalist [Volume 56 Table 1. Productivity of Bam Owls using artificial nest boxes in Juab, Utah, and Salt Lake counties, Utah, 1979-1984. 1979 1980 1981 1982 1983 1984 Total .V •s # nest boxes suneved 20 25 27 29 28 29 158 26.3 3.44 # boxes used as nests 8 16 22 19 23^ 17 106 17.5 5.39 # fledged 16 63 118 71 80 80 428 71.3 33.04 # Hedged/box (.? ) 0.8 2.5 4.4 2.3 2.9 2.8 — 2.6 1.15 # fledged/used box (x ) 2.0 3.9 5.4 3.7 3.5 4.7 — 4.0 1.16 # unsuccessful boxes 2 4 2 2 3 2 15 2.5 .83 % unsuccessful boxes 25 25 9.1 15.8 13.0 11.8 — 15.8 7.25 ^Single nests at which 2iid liroods occurred are counted twice. mortality was 12.5%. Nests that successfully fledged young averaged 4.0 (±1.8) fledglings, but net productivity for total attempt was 3.5 (±2.2). Lower clutch sizes (2, 3, 4 eggs/clutch) had a relatively higher percent success than larger clutches (>4 eggs/clutch); however, clutch sizes of 8 produced the highest number of fledglings (x = 5.3 ± 3.8, n = 3). Clutches with 5 (n = 8) and 10 (n = 1) eggs were least productive, with approximately 50% hatching and fledging success. Seven-egg clutches were among the more productive clutch sizes, fledg- ing an average of 5 young (±2.3), with 82% hatching success and 71% fledging success. Three instances of 2nd broods occurred (Table 1). One female (1982) produced 7 fledg- lings from 1 silo and then from another silo located approximately 200 m away produced 4 fledglings from a 2nd clutch. The "alternate" nest site was consistently used for roosting throughout the previous winter and spring by a male and during the latter part of the first nesting period by the nesting pair Since only the female of the nesting pair was banded, it is not known whether the male using the "alter- nate site" during winter and spring was a mem- ber of the nesting pair, or whether the same male fathered both clutches. The 2nd and 3rd instances of 2nd brood occurred in 1983. Each female produced both clutches in the same box. Of 19 Barn Owl nesting attempts in 1982 with known outcome, 3 failed to fledge young (15.8% failure); in 1983, 3 of 23 nests failed to fledge young (13.0% failure). Nest failures were believed to have occurred during incubation or shortly after eggs hatched, judging from the lack of accumulation of fecal matter and fresh pellets. Reasons for most nest failures are un- known, but 1 case of failure was due to human disturl)ance (use of silo for silage storage). Other probable causes were loss of 1 or more parents or desertion, particularly in 1983, when clutches were abandoned after a long, cool, wet period followdng egg laying. Although reasons for all brood reductions are unknown, some ma>' be attributable to human disturbance, particularly where there was evidence of human activit\' at silos. Fratri- cide may have accounted for at least 2 brood reductions, where remains of young were in the nestbox or in pellets. Two reductions were investigator related and occurred when nest- lings fell fiom the nestbox after the adult female flushed. Sex Ratios Of 65 fledglings sexed in 1982, 26 were males and 39 females; this is a significant overproduction of females (x^= 2.6, 0.5 < P < 0.10; df = 1). However, the number of males and females produced during 1983 (of 49 fledglings sexed: 26 females, 23 males) was not significantly different from the expected 1:1 ratio. Dispersal Thirty-five juveniles banded in the study area between 1979 and 1983 were recovered. Of these, 61% were within 25 km of their natal site, 12% within 50 km, and the remainder within 350 km. Most recovered juxeniles (54%) dispersing more than 25 km tended to fly northwest, with most live returns found occu- pying nestboxes in northern Utah. Twenty- three percent dispersed to the southwest. Eleven (31%) recovered owls were less than 6 mon old; these were mostly within 1 km of the natal site and probably died while dispers- ing. Nineteen (54%) were approximately 1 yr old when recovered, 3 (9%) were recovered approximately 2 yr after banding, and 2 birds were 3 xr old when recovered alive. One was captured as a breeding bird at her natal site 3 vr in a row. 1996] Barn Ow ls in Nohiii Ckntkal Utah 77 Tabi.K 2. Clutch sizes (% of yearly total) of Barn Owls in Jnah, LItah, and Salt Lake counties, Utah, 1979-1983. 1979 1980 1981 1982 1983 'ihtal # nests (7) (16) (20) (19) (23) (85) # eggs 1 0 0 0 0 0 0 2 0 0 1(5) 1(5) 0 2(2) 3 0 1(6) 0 1(5) 5 (22) 7(8) 4 2 (28.6) 1(6) 2(10) 3(16) 4(17) 12 (14) 5 2 (28.6) 3(19) 1(5) 4(21) 4(17) 14 (16) 6 2 (28.6) 5(31) 4(20) 2(10) 4(17) 17 (20) 7 1(14) 5(31) 5 (25) 5 (26) 3 (13) 19 (22) 8 0 1(6) 7(35) 1(5) 2(9) 11(13) 9 0 0 0 1(5) 1(4) 2(2) 10 0 0 0 1(5) 0 1(1) Total eggs 37 95 130 111 121 494 Mean {s) 5.3(1.13) 5.9(1.29) 6.5(1.67) 5.8 (2.03) 5.3(1.81) 5.8 (1.72) Mortality Collision with automobiles, shooting, acci- dents, and severe winter weather coupled with food shortage have been cited as causes of mortality of adult Bani Owls (Henny 1969, Fleay 1972, Smith and Marti 1976). At least 12 road- kills were seen during summer and autumn 1982 in the study area, and accidental deaths occur frequently, particularly with dispersing juveniles (Smith and Marti 1976). Of 9 known accidental deaths of fledglings in 1982 and 7 in 1983, most were due to collisions with cars. During the winter of 1981-82, at least 55 dead Barn Owls were found in north central Utah. During this same period, Marti and Wagner (1985) reported 77 dead Barn Owls in northern Utah. These birds were emaciated and death was attributed to starvation result- ing fi'om cold weather and deep snow. During the period most deaths occurred, mean tem- peratures were -9.7°C, 2.4° below normal. Snow cover was estimated at 20-25 cm, and this likely interfered with capture of Microtus spp., the Barn Owl's main prey. Additional Observations Adults and fledglings were not color marked; however, on 1 occasion, a banded fledgling from 1 silo was found among a same-age brood in a nearby (ca 0.75 km) silo. The fledgling was 9 wk old and was present at the nearby silo on 2 different occasions. Activity at the silo was monitored the night of the discoveiy, and the "foster" fledgling was observed accepting food brought by the adults. No territorial behavior was noted by adults or fledglings on this occa- sion. The only occurrence of territorial behav- ior noted during the 1982-83 period was aggressive behavior by a female Bam Owl nest- ing in a silo in Lehi toward an American Kestrel {Falco sparverius) nesting in a nearby building. Pellet and Prey Analysis A total of 2179 individual prey items were identified from 888 pellets and pellet frag- ments gathered from silo floors. An additional 44 prey items were identified from remains on silo floors (Table 4). At least 16 mammal species (94% of total prey), 11 bird species (4.8%), and 4 insect groups (0.5%) were identified. By individuals, Microtus spp. (ca 77%) and Per- oiuyscus spp. (ca 7%) accounted for over 84% of total prey. Other important mammalian species included the western harvest mouse {Reithrodontomys megalotis), house mouse {Miis mmculiis), and pocket gopher {Thomomys spp.), although none constituted over 3% on an annual basis. The European Starling {Sturnus vulgaris) and Yellow-headed Blackbird {Xan- thocephalus xanthocephalus) were the most frequently taken birds, each comprising 1% of the total prey. Percent frequency of each class of food identified was strongly correlated with per- centage biomass of the same class of food. Mammals (over 94% by number) made over 92% by biomass, while birds (over 4% by num- ber) made over 7% by biomass. Microtus spp. made up a large proportion (73%) of the bio- mass, with M. montanus alone accounting for 38% of the biomass consumed (Table 5). Seasonal comparisons of prey (Appendix 1) indicate that changes in relative abundance of prey items occurred during the study. Some 78 Great Basin Naturalist [Volume 56 Table 3. Number of nestlings (% yearh' total) fledged from artificial nest boxes in Juab, Utah, and Salt Lake counties, 1979-1984. 1979 1980 1981 1982 1983 1984 Total # nests (6) (12) (20) (16) {■20f (15) (89) # fledged (%) 1 2(33) 0 0 0 1(5) 1(7) 4(4) 2 0 2(17) 1(5) 1 (12.6) 2(10) 0 6(8) 3 2 (33) 0 0 3 (19) 7(35) 3 (20) 15 (17) 4 2 (33) 1(8) 4(20) 5 (31) 4(20) 1(7) 17 (19) 5 0 3 (25) 0 4(25) 1(5) 0 8(9) 6 0 2(17) 6 (30) 1(6) 3 (15) 5(33) 17 (17) 7 0 4(33) 8(40) 2 (13) 1(5) 4(27) 19 (21) 8 0 0 1 (5) 0 1(5) 1(7) 3 (3) Mean (s) 2.7(1.4) 5.3 (1.8) 5.9(1.6) 4.4(1.5) 4.0 (1.8) 5.3 (2.0) 4.8 (1.9) "Counts 2nd clutches in single nests twice changes appeared to be seasonal, while others may be of a long-term nature. While Microtiis was the most heavily used group throughout the collecting period, it was used much more frequently during winter and spring. Peroimjs- ciis spp. and Thomomys spp. were more fre- quent in pellets collected during summer and autumn months. Sorex spp. were present in pellets during autumn, winter, and spring but not summer. Birds were used throughout the year but were least represented during summer. No sin- gle bird species was represented in pellets fi"om all 4 seasons; however, the European Star- ling, House Sparrow {Passer domesticus), and Red-winged Blackbird {Agelaiiis phoeniceiis) were represented in 3 seasons. Analysis of prey diversity (Table 5) gives further characterization of the Barn Owl prey base. Prey species diversity of Barn Owls in north central Utah was 2.96; ma.\imum diver- sity possible was 3.434. While this shows some variation and an ability to take locally abun- dant prey species, it indicates a degree of sin- gular specialization on Microtiis spp. Diversity of north central Utah Barn Owl's food habits is roughly similar to recorded values observed in other areas in North America and Europe (Selleck and Glading 1943, Hawbecker 1945, Evans and Emlen 1947, Uttendorfer 1952, Glue 1974, Marti 1974), but it is higher than values reported from the same area in 1976 (Smith and Marti 1976; T^ible 5). Evenness, the actual diversity of prey base as a percentage of maximum diversity possible, was 59%; this indi- cates Barn Owls were not sampling possible prey evenly, but rather were taking a higher percentage of more common species. Food Brought to Nest Food stockpiles were found at most nests during the incubation period. Stockpiling began slightly before deposition of the first egg and continued throughout the hatching period. Initial stockpiles were small, 2-5 prey items, but stockpile sizes increased as the season pro- gressed. The largest stockpile consisted of 23 microtines, 3 starlings, and 16 Yellow-headed Blackbirds. Wallace (1948) reported a stock- pile of 190 mammals, primarily rodents. At least 9 prey species were recorded: 53% microtines, 28% Yellow-headed Blackbirds, 6% starlings, and 3% each of Red-winged Blackbirds and deer mice. Other species were the Brown-headed Cowbird {Molothrus ater, 1.5%), Black-billed Magpie {Pico pica, 1.5%), vagrant shrew {Sorex vagrans, 1%), and Noi-way rat {Rattns norvegiciis, 1%). Discussion Breeding and Productivity It appears that variability of clutch size in Barn Owls is more closely related to factors other than latitude. The 5-yr mean clutch size (5.8 eggs/clutch) for north central Utah (Lat. 39°— 40°N) reported herein was much higher than average clutch size of 4.2 eggs reported for areas of higher latitude, as well as for a breeding colony studied in the same area in 1973 (Smith et al. 1974); however, this was much lower than the 4->'r mean clutch size of 7.0 eggs reported by Marti and Wagner (1985) for northern Utah Barn Owls (Lat. 41 °N). Additionally, there was a wide discrepancy between the niodal clutch and brood sizes 1996] Barn Owls in North Central Utah 79 Table 4. Total prey identifietl for Barn Owls utilizing artificial nest boxes in Juab, Utah, and Salt Lake counties Utah 1982-83, Number Percent Total Percent total i^re\ species frequency biomass biomass Mammals Microtus pcunsijlvanicus 2L5 9.8 8600.0 9.5 Microtiis immtanus 887 40.4 35480.0 38.7 Microtus longicaudus 377 15.4 15080.0 14.7 Microtus spp. 239 10.9 9560.0 10.4 Mus musculus 51 2.3 969.0 1.1 Neotonm cinerea 9 0.4 2493.0 2.7 Pcromijscus maniculatus 102 4.7 2142.0 2.3 Pcromyscus tniei 2 0.1 42.0 <.l Pcromijscus spp. 63 2.9 1323.0 1.4 Sorex cinereus 6 0.3 30.0 <.l Sorex obscurus 14 0.6 84.0 .1 Sorex vagrans 18 0.8 108.0 .1 Sorex spp. 28 1.3 154.0 .2 Spennophihts variegatus 1 tr. 177.0 .2 Rattus norvegicus 5 0.2 1100.0 1.2 Reithrodontoimjs megalotis 32 1.5 350.0 .4 Thomomijs bottae 50 2.3 4250.0 4.6 Thomomys talpoides 1 tr 85.0 .1 Mephitis mephitis 5 0.2 4110.0 4.5 Total mammalian individuals 2105 94.6 86236.0 92.4 Birds Agelaius phoeniceus 9 0.4 432.0 .5 Cohimba livia 1 tr. .332.0 .4 Icterus galbula 7 0.3 231.0 .2 Molothrus ater 1 tr 41.0 <.l Passer domesticus 8 0.4 216.0 .2 Passerculus sandwichensis 2 0.1 42.0 <.l Pica pica 2 0.1 360.0 .4 Sturnus vulgaris 23 1.0 1817.0 2.0 Turdus migratorius 1 tr. 79.0 <.l Tyto alba 1 tr. 525.0 .6 Xanthocephalus xanthocephalus 23 1.0 1702.0 1.8 Unidentified birds 30 1.4 1260.0 1.4 Total avian individuals 108 5.9 7037.0 7.5 Total vertebrate individuals 2213 99.5 93273.0 99.9 Invertebr.'\tes Carabidae 2 0.1 .4 <.l Tenebrionidae 3 0.1 1.8 <.l Orthoptera 2 0.1 1.2 <.l Unidentified Coleopterans 3 0.1 .6 <.l Total invertebrate individuals 10 0.5 4.0 <.l Total prey individuals 2223 100.0 93281.0 100.0 ''tr = trace reported herein (7, clutch; 7, brood) and those reported elsewhere (Bunn et al. 1982 [5, 2], Ottenietal. 1972 [5, 3]). Lack (1949) found mean clutch size of owls to increase with latitude and abundance of rodents. Otteni et al. (1972) found that clutch size for 112 clutches in southern Texas (Lat. 28° N) averaged 4.9 and was identical to aver- age clutch size for 68 Maryland clutches (Lat. 38°- 43 °N; Henny 1969). A mean clutch size of 5.3 eggs for Barn Owls nesting in Switzer- land (Lat. 46°-47°N) was also reported by Henny (1969); Glue (1974) reported an aver- age clutch size of 4.7 in Great Britain (Lat. 50°-55°N). Lack (1954) suggested the number of eggs laid by each species has been established to correspond with the number of young that can 80 Great Basin Naturalist [Volume 56 Table 5. DiversiW indices oi Barn Owl piedation for Utah and odier areas. # prey items # prey species Diversity-' Location Mammals Birds Source Utah North central 2173 16 11 2.96 this stud>^ Box Elder Co. 178 8 1 2.31 Smith and Marti 1976 Utah Co. 3004 12 12 1.45 Smith and Marti 1976 California Southern 933 10 13 2.19 Selleck and Glading 1943 Central 948 20 11 3.10 Hawbecker 1945 Sierras 513 8 0 1.95 Fitch 1947 Northern 739 8 6 + 2.41 Evans and Emlen 1947 Colorado 4366 6 16 2.76 Marti 1974 Idaho 202 9+ 1 + 1.79 Roth and Powers 1979 Michigan 6815 5 13 0.98 Wallace 1948 Ohio 1060 9 5 0.98 Phillips 1951 Pennsylvania 6165 7 17 1.46 Latham 1950 Texas 2056 6+ 10 3.35 Otteni et al. 1972 Chile 3417 13+ 0 2.82 Herrera and Jaksic 1980 England 47865 8+ 17 2.29 Glue 1974 3546 8 0 1.60 Webster 1973 Germany 76664 51 32 2.69 Uttendorfer 1952 Spain 12351 11 + 0 2.11 Herrera and Jaksic 1980 ^Diversih. calculated usiny .Sliannon-Weiner's diversih index (H ): -'^(POilogPi) be successfully raised, and successful rearing is based on the amount of food available and provided to young by adults. Otteni et al. (1972) found that southern Texas Barn Owls seemed to adjust reproductive efforts to rodent popula- tion fluctuations. They produced slightly low- ered mean clutch size and number of complete clutches during periods of lower rodent prey population sizes and increased the number of young raised/pair during periods of abundant rodent prey populations. Similar findings were reported in Europe by Glue (data from Bunn et al. 1982) and Baudvin (1975), whose studies indicated that variations in fledging success were entirely linked to vole numbers. Marti and Wagner (1985) reported that a winter die- off of northern Utah Barn Owls in 1981-82 resulted in a later egg-laying season, a 40% decline in breeding attempts, and a decline in average clutch size from 7.0 to 5.8 eggs; how- ever, decline in productivity was not paral- leled in our study area during this period. These findings indicate that Barn Owl produc- tivity may be closely tied to availability of prey, and that differences between clutch and brood sizes reported herein, and those reported in the same and in different areas of the Barn Owl range are likely correlated with fluctua- tions in prey populations and weather as they affect prey availability. Production of 2nd broods by Barn Owls is thought to be triggered by an abundance of prey (Honer 1963). All 3 pairs producing 2nd clutches during this study, 1 in 1982 and 2 in 1983, successfully fledged young from their 1st brood. In these cases, deposition of the 1st egg of the 2nd brood occurred several weeks after the last young of the 1st brood fledged. Second broods are often less successful than 1st broods, since pre>' numbers decline later in the season when hatchlings still require feeding (Bunn et al. 1982). This was not the case with our obsei'vations. All 3 second nests were successful, with 2 nests 100% successful in hatching and fledging, and 1 sustaining 60% mortality of eggs but 100% success in fledging young. Furthermore, the 3 pairs successfully fledged 27 young for the breeding season, an average of 9 young per pair. Henny (1969) suggested that in northern environments high biotic potential of Barn Owls may serve as a "built-in compensating fac- tor" that affords protection against low years in rodent cycles and allows rapid restoration of Barn Owl populations to previous "good rodent year" size. Second-clutching during 1996] Bahn Owls in North Central Utah 81 1982 and 1983 nia\' be 1 response to lowered population numbers resulting from the winter die-off of 1982 and abandoned clutches result- ing from cool, wet weather following egg deposition in 1983. Food Habits Barn Owls in Juab, Utah, and Salt Lake counties sustained themselves almost exclu- sively by consuming mammals and birds, de- spite seasonal abundance of large invertebrates, reptiles, and amphibians. Year-round presence of microtine species in the diet is in agree- ment with other data throughout the Barn Owl's range (Wallace 1948, Phillips 1951, Otteni et al. 1972, Smith et al. 1972, Webster 1973, Lovari 1974, Lovari et al. 1976, Smith and Marti 1976, Roth and Powers 1979, Her- rera and Jaksic 1980). Webster (1973) and Wallace (1948) noted that numbers of secondary prey species captured by Barn Owls are inversely proportional to numbers of microtines captured, particularly when Sorcidae spp. form the main alternative to Microtinae. Although Sorex spp. were uti- lized frequently by Barn Owls in north central Utah, no inverse relationship could be seen between proportions of Sorex spp. and Micro- tiis spp. An inverse relationship was noted for proportions of Microtus spp. and Feromyscus spp. Peromyscus spp. were clearly the main alternative to Microtus spp. In studies where numbers of secondary prey species are inversely proportional to numbers of microtines, the correlation has been linked with relative pro- portions of woodland and open areas in the owls' territories (Bunn et al. 1982). Woodlands exist in isolated areas throughout the study area, adjacent to lakes, streams, and foothills, but open field areas are more common. Thus, during summer and autumn, one or both adults may have been foraging more frequently in woodland areas (represented by Peromyscus spp.) than in open field areas (represented by Microtus spp.). During winter and spring, for- aging may have shifted more to open field habitats. Alternatively, increased occurrence or availability of Peromyscus spp. resulting from increased reproductive activity during summer and autumn months may account for the shift in diet. Only a few unusual prey items are notewor- thy: predation on a group of striped skunks {Mephitis mephitis: 2 adults, 3 juveniles) at a silo in Nephi [C. Marti (personal communica- tion) doubts that the owl would have killed so large an animal, but the evidence foimd clearly indicated that owls nonetheless fed on skimks]; presence of a stockpiled rock squirrel [Sper- mophilous variagatus) and a sora {Porzana car- oHmi); cannibalism indicated by presence of a juvenile Barn Owl skull among loose pellets collected in autumn, as well as the discovery of what looked like a partly consumed juvenile Barn Owl in another nestbox. Cannibalism has been reported in Califor- nia (Henny 1969) during years when food sup- plies were low, and Baudvin (1975) reported cannibalism as the major source of Barn Owl nestling mortality in France. Often during this study, owlets (as well as eggs) seem to have "disappeared " without a trace. These may have been cannibalized, they may have died and been moved to another site, they may have been eaten by an adult or a sibling, or they may have been predated by another species. While asynchro- nous hatching characteristic of Barn Owls is thought to facilitate cannibalism (O Connor 1978), care should be taken in ascribing Barn Owl remains in pellets to cannibalism. Sex Ratios Mendenhall (1983) reported an equal pro- duction of sexes in captive Barn Owls at Patuxent Wildlife Research Center, Maiyland, but data from the wild are few. The higher proportion of female fledglings observed in north central Utah during 1982 was significant (X-= 2.6, P < 0.10; df = 1), particularly in view of the high adult winter-kill observed during the severe winter of 1981-82, and the hypothesis of sex-biased brood reduction favoring female offspring during periods of food (or other environmental) stress (Howe 1977, Newton 1979, Bildstein 1981) is supported. While a single season's deviation fiom expected unity could well be stochastic, differential pro- duction of sexes during environmentally stress- ful periods has been obsei-ved in a number of vertebrate groups (Howe 1977, Bull 1980, Charnov 1982). Polygynous behavior by Bani Owls (Baudvin 1975, Bunn et al. 1982, Marti 1990) should be considered when addressing the differential sex ratio. Differential sex ratios among polygy- nous birds are fairly well established (Newton 82 Great Basin Naturalist [Volume 56 1979, Fiala 1981, Charnov 1982). Polygynous species tend to show differential production of sexes more fiequendy dian monogamous species (Lack 1954, Verner 1964, Zimmerman 1966), although hypotheses regarding proximate and ultimate causes vaiy. Olsen and Cockbuni (1991) have shown that raptors frequently have a nat- urally biased sex allocation toward females. The reasons for such an allocation were not clear although their data did not implicate polygyny. No verified polygynous behavior was noted during this study; however, the close association between the "foster" fledgling and parents of a separate brood reported herein indicates a possibility of shared parentage, particularly since the foster fledgling's natal site was so close. Unfortunately, adult males from either silo were never captured for band- ing, so pairing was unknown. An alternative explanation of the "foster" fledgling behavior is that the dispersing fledgling obsei-ved adults leaving and entering the adjacent silo, and in stereotypic behavioral fashion it followed the adults. Once near the nest, normal brood beg- ging would have elicited feeding response from the adults. Further information on Barn Owl mating behavior and dispersal is needed to elucidate the differential production of females observed during this study. More importantly, documen- tation of sex ratios, both at birth and fledging, over many years is required to place the ob- sei-ved skewed sex ratio into perspective. Addendum: Since the final editing of this paper a major review of Barn Owls by Marti (1992) appeared. One should consult that paper for recent details relevent to our findings. Acknowledgments Financial support was provided by the Associated Students Research Council and Zoology Department at Biigham Young Univer- sity. Logistical support was provided by the Utah Division of Wildlife Resources. Field assistance was provided by D. Boyce, K. Fris- tensky, J. Hebdon, K. Keller, R. Meese, K. Rauhaufer, K. Rhodes, and S. Stewart. Norma Konrad assisted in identifying and aging mam- malian prey remains. We thank B. Sample, J. Flinders, C. Marti, H. D. Smith, R. C. Whit- more, and an anonymous reviewer for com- ments on the manuscript. Literature Cited Baudvin, H. 1975. Biologie de reproduction de la chou- ette effraine {Ttjto alba) en Cote d'Ore; premiers resultates. La Jean le Blanc 14; 1-15. Behle, W. H. 1941. Barn Owls nesting at Kanab, Utah. Condor 43:160. BiLDSTEiN, K. L. 1981. Reversed sexual size dimorphism in raptors; selective forces acting on non-breeding birds. Abstract, Proceedings of the 1981 annual meet- ing of the Raptor Research Foundation, October 1981, Montreal, Quebec. Bull, J. 1980. Sex determination in reptiles. Quarterly Review of Biology 55: 3-21. Bunn, D. S., a. B. Warburton, and R. D. S. Wilson. 1982. The Barn Owl. Buteo Books, Vermillion, SD. 264 pp. Charnov, E. 1982. The theoiy of sex allocation. Princeton University Press, Princeton, NJ. 355 pp. DURIUNT, S. D. 1952. Mammals of Utah. Museum of Nat- ural History, University of Kansas Publications 6: 1-549. Evans, E C, and J. T. Emlen, Jr. 1947. Ecological notes on the prey selected by a Barn Owl. Condor 49; .3-9. Fiala, K. L. 1981. Sex ratio constancy in the Red-winged Blackbird. Evolution 35: 898-910. Fitch, H. 1947. Predation by owls in the Sierran foothills of California. Condor 49:137-154. Fleay, D. 1972. Nightwatchmen of bush and plain. Tap- linger Publishing, Co., NY. Glue, D. E. 1974. Food of the Bam Owl in Britain and Ireland. Bird Study 21: 200-210. Hawbecker, a. C. 1945. Food habits of the Barn Owl. Condor 47: 161-166. Henny, C. J. 1969. Geographical variation in mortality rates and production requirements of the Bam Owl Tyto alba. Bird-banding 40: 277-290. Herrera, C, and F M. Jaksic. 1980. Feeding ecology of the Barn Owl in central Chile and southern Spain: a comparative study. Auk 97: 760-767. HONER, M. R. 1963. Obsei-vations on the Barn Owl {Tijfo alba guttata) in die Netlierlands in relation to its ecol- ogy' and population fluctuations. Ardea 51; 158-195. Howe, H. F 1977. Se.x-ratio adjustment in the Common Crackle. Science 198; 744-745. Kovach, W. L. 1987. MVSP— Multivariate Statistical Pack- age, version 1.3, user's manual. Lack, D. 1949. The significance of clutch size. Part 1: Intra- specific variations. Ibis 89: 302-352. . 1954. The natural regulation of animal numbers. Claredon Press, Oxford. Lathanl R. M. 1950. The food of predaceous animals in northeastern United States. Final report, Pittman- Robertson Project 36-R, PennsyKania Game Com- mission. Logman, S. J. 1985. Pr()clucti\ it\, food habits and sexing of Barn Owls in Utah. Unpublished masters thesis, Brigham Young University, Provo, UT. Lo\'ARl, S. 1974. The feeding habits of four raptors in cen- tral Italy Raptor Research 8; 45-57. LovARi, S., A. J. Renzoni, and R. Fondi. 1976. The predator}' liabits of the Barn Owl {Tyto alba) in rela- tion to the vegetation cover. Bulletin of Zoolog\' 43: 173-191. 1996] Barn Owls in North Central Utah 83 MAGURa\N, A. E. 1988. Ecological diversity and its mea- surement. Princeton University Press, Princeton, NJ. 179 pp. Marti, C. D. 1974. Feeding ecology of four sympatric owls. Condor 76: 45-61. . 1990. Same-nest polygxny in the Bam Owl. Condor 92:261-263. . 1992. Barn Owl. //*; A. Poole, R Stettenheim, and E Gill, editors. The birds of North America, No. 1. Academy of Natural Sciences, Philadelphia, and American Ornithologists" Union, Washington, DC. Marti, C. D., and P W. Wagner. 1985. Winter mortality in common Barn Owls and its effect on population density and reproduction. Condor 87: 111-115. Marti, C. D., R W Wagner, .\nd K. W. Denne. 1979. Nest boxes for the management of Barn Owls. Wildlife Society Bulletin 7: 145-148. Mendenhall, v. 1983. Captive Barn Owl breeding at Patu.xent Wildlife Research Center, Laurel, MD. Unpublished manuscript. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD. 399 pp. O'Connor, R. J. 1978. Brood reduction in birds: selection for fratricide, infanticide and suicide? Animal Be- haviour 26: 79-96. Olsen, R D., and a. Cockburn. 1991. Female-biased sex allocation in Peregrine Falcons and other raptors. Behavioral Ecology and Sociobiology 28: 417—423. Otteni, L. C, E. G. Bolen, and C. Cottam. 1972. Preda- tor-prey relationships and reproduction of the Barn Owl in southern Texas. Wilson Bulletin 84; 434-448. Phillips, R. S. 1951. Food of the Barn Owl, Tyto alba pratincohi, in Hancock County, Ohio. Auk 68; 239-241. Picknvell, G. 1948. Barn Owl growth and behaviourisms. Auk 65: 339-373. Reese, J. G. 1972. A Chesapeake Barn Owl population. Auk 89; 106-114. Roth, D., and L. R. Powers. 1979. Comparative feeding and roosting habits of three sympatric owls in south- western Idaho. Murrelett 60; 12-15. Selleck, D. M., and B. Gladinc;. 1943. Food habits of nestling Bam Owls and Maish Hawks at l^une Lakes, California, as determined by the cage nest method. California Fish and Game 29: 122-131. Smith, D. G., C. R. Wilson, and H. H. Frost 1970. Fall nesting of Barn Owls in Utah. Condor 72: 492. . 1972. Seasonal food habits of Barn Owls in Utah. Great Basin Naturalist 32: 229-234. . 1974. Histoiy and ecology of a colony of Barn Owls in Utah. Condor 76: 131-136. Smith, D. G., and C. D. Marti. 1976. Distributional sta- tus and ecology of Barn Owls in Utah. Raptor Research 10: 33-44. Steenhof, K. 1983. Prey weights for calculating percent biomass in raptor diets. Raptor Research 17: 15-27. Uttendorfer, O. 1952. Neue Ergebnisse uber die Erna- ghrung der Greifvogel und Eulen. Eugen Ulmer, Stuttgart. Verner, J. 1964. Evolution of polygamy in the Long- billed Marsh Wren. Evolution 18; 252-261. Voous, K. H. 1988. Owls of the northern hemisphere. MIT Press, Cambridge, MA. WalL/\ce, G. J. 1948. The Barn Owl in Michigan; its dis- tribution, natural histoiy and food habits. Technical Bulletin of the Agricultural Experiment Station, Michigan 208: 1-61. Webster, J. A. 1973. Seasonal variation in mammal con- tents of Barn Owl castings. Bird Study 20: 185-196. Woodbury, A. M., C. Cottam, and J. W Sugden. 1949. Annotated checklist of birds in Utah. University of Utah Biological Series 39: 1-40. Zimmerman, J. L. 1966. Polygyny in the Dickcissel. Auk 83; 534-546. Received 9 August 1995 Accepted 14 September 1995 (Appendix 1 begins on the following page. 84 Great Basin Naturalist [Volume 56 Appendix 1. Vertebrate food items taken seasonally bv Barn Owls in Juab, Utah, and Salt Lake eounties, LI tab, 1982-83. Season Spring ' Summe .1, Autumn ^■ Winter 1 Totab' %f Total % Total % Total % Mammals Mephitis mephitis 0 0 0 0 5 0 0 Mierotus longieaiidus 191 87 54 50 108 73 24 100 Microtus pennsijhaniciis 138 75 7 33 41 64 29 67 Microtus montamts 489 100 67 100 280 100 51 100 Microtus spp. 109 — 19 — 78 — 33 — Miis iniiscuhis 25 56 3 33 21 64 2 67 Neotoma cinerea 1 6 0 0 8 46 0 0 Peromijseus maniculatus 27 44 23 67 49 91 3 67 Peromyscus truei 0 0 0 0 2 18 0 0 Peromijseus spp. 16 — 11 — 36 — 0 — Rattus norvegictis 4 6 0 0 1 9 0 0 Reithrodontomijs megalotis 20 69 1 17 8 46 3 67 Sorex vagrans 12 44 0 0 3 18 3 67 Sorex cinereus 5 19 0 0 0 0 1 33 Sorex obscurus 14 25 0 0 0 0 0 0 Sorex spp. 22 — 0 — 3 — 3 — Spermophilus variegatus 0 0 1 17 0 0 0 0 Thomomijs hottae 6 31 6 50 38 46 0 0 Thomomijs talpoidcs 1 6 0 0 0 0 0 0 Total mammalian individuals 1080 — 192 — 681 — 152 — Birds Agelaius phoenieeus 5 25 1 17 3 18 0 0 Cohimba livia 1 6 0 0 0 0 0 0 leterus galhuhi 3 19 0 0 4 27 0 0 Molothrus (iter 1 6 0 0 0 0 0 0 Passer domesticiis 4 6 0 0 2 18 2 67 Passercuhis sandwichensis 0 0 2 17 0 0 0 0 Pica pica 2 12 0 0 0 0 0 0 Stiirnus vulgaris 14 38 0 0 7 55 2 67 Turdiis migraforiiis 0 0 0 0 1 9 0 0 Tijto alba 0 0 0 0 1 0 0 0 Xanthoeephahis xanthocepluil US 19 31 0 0 4 18 0 0 Unidentified birds 11 2 13 4 Total avian individuals 60 — 5 — 35 — 8 — Total vertebrate individuals 1140 197 716 160 ^total pellets collected; 467; total nests surveyed: 15 "total pellets collected: 61; total nests surveyed: 6 '^total pellets collected: 287; total nests surveyed: 11 "total pellets collected: 76; total nests surveyed: 3 •"total individuals identified 'fre(|iiency of occurrence in nests sui'veyed Great Basin Naturalist 56(1), © 1996, pp. 85-86 ASTRAGALUS LAXMANNII JACQUIN (LEGUMINOSAE) IN NORTH AMERICA R. C. Barnehyi and S. L. Welsh- Key words: Astragalus laxmannii, nomenclature, North America. In a recent article Podlech (1993) proposed lectotypes for two names that have impHca- tions in the flora of North America, i.e., A. lax- DUDUiii lacquin and A. adsiirgens Pallas. Both names have been used in the literature of American Astragalus in application to the one species that has been generally accepted in modern times as A. adsiirgens sens. lat. (Bameby 1964). Podlech's typifications may be summarized: Astragalus laxmannii Jacquin, Hort. Vindob. 3: 22, Tab. .34. 1776. Lectotypus (Podlech, Sendtnera 1: 270. 1993): "Planta culta in Horto Vindobonensi e seminibus a Laxmann e Sibiria (Samen von Pallas erhalten, siehe Pallas, Sp. Astra- gal, p. 39. 1800). Specimen a Jacquin missum (BM!)." Astragalus adsurgens Pallas, Sp. Astragal. 40. 1800. "Crescit hie Astragalus tantum in regionibus Trans- Baicalensibus, cum A. Laxmanni promiscue, frequens ad Selengam, Ononem, circa Tarei-noor, et usque in Mongo- liae desertum." Type (Podlech 1993): "Transbaicalia, ad Selengam, Pal- las (BM!); Onenem circa Tareinoor, Pallas (BM!); Syn- bi'pen." "Lectot>'pus: ad Selengam, Pallas (BM!)." Following examination of the proposals by Podlech, we obtained pertinent specimens on loan from The Natural History Museum (BM) in London, through the courtesy of A. R. Vick- ery. There are 7 pertinent sheets at BM, 6 from the Pallas herbarium and 1 from the Jacquin herbarium, none of them annotated by Podlech. The sheet from the Jacquin col- lection is labeled "Astragalus Laxmannii. Jack. Hort.VB." and has a notation on the back side, "Herbar NJ Jacquin." It bears a single plant with a branched caudex, several stems, and inflorescences with withered flowers and early fruit. The plant fits well within the characteri- zation of A. adsiirgens var adsurgens as de- scribed by Barneby (1964). It is certainly the plant chosen as lectotype by Podlech. Among the 6 specimens from the Pallas herbarium, 2 bear the designation Astragalus adsurgens and the additional notation, 'lax- mannii. " The other 4 are annotated A. laxman- nii. One of the specimens labeled A. adsur- gens has 2 notations, 1 at the top, "ad [Tarei- noor, crossed out] Selengam," and 1 below the specimen, "Specimen drawn in plate 31. Pall." This specimen (49221 BM) is the undoubted lectotype for A. adsurgens. It is mounted with at least 3 other fragments of the same species. Sheet 49227 (BM!), bearing a "Type Specimen" label and with the name A. adsurgens, is likely a paratype. One (49222 BM!) of the 4 sheets, all bearing the name laxmannii, also has a notation, "ad Selengam," and another, "ad Tareinoor." They are possible paratypes of A. adsurgens and are mounted with 2 other frag- ments. Sheets 49223, 49224, 49225 (all BM!) are all A. laxmannii (as annotated), but appar- ently they are nomenclaturally irrelevant. Two infraspecific taxa have been recog- nized within A. adsurgens in the flora of North America. Their names require nomenclatural realignment within A. laxmannii, as follows: Astragalus laxmannii var. robustior (Hooker) Barneby & Welsh, comb, nov., based on A. adsurgens var. robustior Hooker, Fl. Bor.-Amer. 1: 149. 1831. Astragalus nitidus var. robustior (Hooker) M.E. Jones, Contr. W. Bot. 10: 64. 1902. Astragalus adsurgens ssp. robustior (Hooker) Welsh, Iowa State J. Sci. 37: 357. 1963. Astragalus laxmannii var. tananaicus (Hulten) Barneby & Welsh, comb, nov., based on A. tananaicus Hulten, Fl. Alaska & Yukon 1763. 1959, a substitute for A. viciifolius Hulten, Ark. Bot. 33B: 1, fig. 1. 1947 (non A. viciaefolius DC. 1802). A. adsurgens var. tananaicus (Hulten) Barneby, Mem. New York Bot. Card. 13: 616. 1964. 'The New York Botanical Garden, Bron.x, NY 10458-5126. -Herbarium. M. L. Bean Life Science Museum, Brighani Young University, Prove, UT 84602. 85 86 Great Basin Naturalist [Volume 56 References Pall.\s, E S. 1800. Astragalus adswgens Pallas. Species Astragalorum. Godofiedi Martini, Lipsiae. Barneby, R. C. 1964. Atlas of North American species of PoDLECH, D. 1993. Miscellaneous notes on Astragalus. Astragalus. Memoirs of the New York Botanical Gar- Sendtnera 1: 270. den 13: 1-1188. Jacquin, N. J. 1776. Astragalus laxmannii Jacquin. Hortus Received 19 July 1995 botanicus vindobonensis 3: 22. Tab. 37. Accepted 5 September 1995 Great Basin Naturalist 56(1), © 1996, pp. 87-89 INTERMOUNTAIN MOVEMENT BY MEXICAN SPOTTED OWLS {STRIX OCCIDENTALIS LUCIDA) R. J. Gutierrez^'^, Mark E. Seamans^, and M. Zachariah Peeiy^ Key icorch: Strix occidentalis, Spotted Oivl, dispersal. The Mexican Spotted Owl {Strix occiden- talis liicida) is a threatened subspecies in the United States (USDI 1993). Both the Mexican and Cahfornia (S. o. occidentalis) Spotted Owl subspecies are distributed as fragmented pop- ulations across their respective ranges (USDI 1993, LaHaye et al. 1994). However, it is not known whether these distributional patterns represent metapopulations or are the result of isolation events because no cases of interpop- ulation (i.e., inteniiountain) dispersal have been published. A true metapopulation structure would depend on dispersal among populations (Levins 1970, Gutien-ez and Hairison in press). In the course of extensive banding of juve- nile (n = 95), subadult (n = 21), and adult {n = 57) Mexican Spotted Owls in the Tularosa Mountains, New Mexico, we recorded 3 cases of owl movement among mountain ranges. We report herein the circumstances of these movements. Our study area is in west central New Mex- ico in the Tularosa Mountains (Fig. 1). We attempted to capture and color mark every Spotted Owl during 1991-1995 in a 323-km'2 study area (approximately 70% of the Tularosa Mountain range) using the methods of Forsman (1983). In 1994 we established random sample quadrats to estimate owl densities in areas surrounding the Tularosa Mountains. The following movements were recorded: 1. We banded an adult female owl on 24 May 1994. This bird was paired with an adult male. A female was heard vocalizing from this territory as late as 13 luly 1994. This female was found dead near Deming, New Mexico, on 19 January 1995. The bird was autopsied by a veterinarian in Las Cruces, New Mexico, who said probable cause of death was electrocution. which was consistent with circumstances lead- ing to the bird's discoveiy (i.e., found below a power pole where an electrical transformer short had occurred). Although the bird was 68 g lighter in weight when recovered than when banded, it was in good condition (i.e., no indi- cation of stai"vation or poor health). The bird was recovered approximately 187 km south southeast of its banding location (Fig. 1). Of particular interest was the fact that the bird probably crossed several mountain ranges before it entered treeless Chihauhuan desert grassland where it was recovered. The nearest suitable owl habitat (e.g., mixed-conifer or pine-oak forest [Pinus ponderosa/Quercus spp.]) was in the Animas Mountains, a straight- line distance of approximately 80 km. The mountain range nearest (approximately 20 km) the bird's final location was the Florida Moun- tains. The highest peak in these mountains is a prominent landmark (maximum elevation 2224 m) in the desert, but it contains no suit- able owl habitat (Fig. 1). We surveyed this bird's territory in early spring 1995. The male from 1994 was still pres- ent at the historical location, but we could not detect a female. However, by June we obsei-ved an adult female roosting with this male. There- fore, the female recovered at Deming apparent- ly left her mate, a relatively uncommon event among tenitorial Spotted Owls (Gutierrez et al. 1995). 2. In 1993 we banded a juvenile female owl that we recaptured 56 km west northwest of its natal site in 1994 on Escudilla Mountain, Arizona (Fig. 1). This mountain is part of the San Francisco Mountain Range. This female was paired at the time of capture and had no young. ' Department of Wildlife, Humboldt State Universit\', Areata, CA 95521. -Send reprint requests to Department of Wildlife. Humboldt State University, Areata, C\ 95521. 87 Great Basin Naturalist [Volume 56 Shortest distance and direction between owl capture sites and relocation FtT^TT ^ Fig. 1. Shortest distance and direction between banding location and final location of dispersing Mexican Spotted Owls in New Mexico. Shaded area represents all forested/woodland areas whether or not diey are suitable habitat for Spotted Owls. Numbered lines correspond to nunil>ers in text and do not inipK actual dispersal route of tlie liird. 1996] Notes 89 3. In 1992 we bunded a jnvenile female owl whieh we reeaptnred in 1994 in the Mogollon Mountains, New Mexieo, 22 km south of its natal site (Fig. 1). This female was paired at the time of capture and had no young. Considering that no examples of intermoim- tain movements ha\'e been recorded among more extensively studied California Spotted Owl populations (LaHaye et al. 1992, 1994), these observations are notable. For example, between 1987 and 1995, approximately 750 juvenile and adult California Spotted Owls were banded in the San Bernardino, San Jacinto, Palomar, and San Gabriel mountain ranges with no subsequent recoveries in another mountain range (LaHaye et al. 1994). Our observation of female-only emigration out of the Tularosa Mountains is consistent with the general obsei-vation of female-biased dispersal in birds (Greenwood 1980). Further, during our study we relocated a total of 10 dis- persing juveniles in subsequent years. Of these, 8 (5 males, 3 females) dispersed within the Tularosa Mountains. The 5 females dis- persed an average of 21.8 km (range = 7.75- 56.32 km, s — 20.0) while the 5 males dis- persed an average of 5.8 km (range = 2.04- 12.58, s = 4.0). Thus, these females dispersed farther than males (Mann-Whitney U one- tailed test, ^-value = -2.194, P = 0.0158), which also supports the idea of female-biased dispersal in Mexican Spotted Owls. These intermountain movements also are consistent with a metapopulation structure (Levins et al. 1970, Gutierrez and Harrison in press). In addition, while Spotted Owls are known to be obligate dispersers (Gutierrez et al. 1995), the long-distance movement by an adult female does not fit the general model of Spotted Owl dispersal (Gutierrez et al. 1985) in which juveniles are the more likely long- distance dispersers. However, dispersal car- ries risks, such as predation, starvation, and accidents while traveling in unfamiliar habi- tats. Even though the adult we banded acci- dentally died, it is possible that adult birds, which have greater experience, may have a higher probability of success when crossing desert grasslands or otherwise unsuitable habitats in the Southwest than juveniles, who have little hunting and predator-avoidance experience. Thus, while studies of juvenile Spotted Owl dispersal are essential to the study of metapopulation dynamics (Gutierrez and Harrison in press), the role of dispersing adults in maintaining metapopulation struc- ture should be considered carefully. Acknowledgments We thank the following field assistants: V Baxter, J. Bamesberger, D. Juliano, E. Gunder- shaug, B. Kwasny, W Michael, W. Moore, and M. Stauber D. Kristan and G. deSobrino read the manuscript. The Rocky Mountain Forest and Range Experiment Station provided fund- ing for the project (Contract #53-82FT-4-07 to RJG). Literature Cited FORSMAN, E. D. 1983. Methods and materials for locating and capturing Spotted Owls. USDA Forest Service, General Technical Report PNW-I62. Pacific North- west Forest and Range E.xperiment Station, Port- land, OR. Greenwood, P J. 1980. Mating systems, philopatiy and dispersal in birds and mammals. Animal Behavior 28: 1140-1162. Gutierrez, R. J., and S. Harrison. In press. Applications of metapopulation theory to Spotted Owl manage- ment: a histor)' and critique. In D. R. McGullough, editor, Metapopulations: wildlife management and conservation. Island Press, Covelo, CA. Gutierrez, R. J., A. B. Franklin, and W. S. LaHaye. 1995. Spotted Owl. In: A. Poole and F Gill, editors, The birds of North America, No. 179. The Academy of Natural Sciences, Philadelphia, PA, and The American Ornithologists' Union, Washington, DC. Gutierrez, R. }., A. Fr/\nklin, W. LaHaye, V J. Meretsky, AND J. P Ward, Jr. 1985. Juvenile Spotted Owl dis- persal in northwestern Galifomia: preliminary results. Pages 60-63 in R. J. Gutierrez and A. B. Garey, edi- tors. Ecology and management of the Spotted Owl in the Pacific Northwest. USDA Forest Service, General Technical Report PNW-185. Pacific North- west Forest and Range Experiment Station, Port- land, OR. LaHaye, W. S., R. J. Gutierrez, and H. R. AKgAKAYA. 1994. Spotted Owl metapopulation dynamics in southern California. Journal of Animal Ecology 63: 775-785. LaHaye, W. S., R. J. Gutierrez, and D.' R. Call. 1992. Demography of an insular population of Spotted Owls {Sitrix occidentalis occidcntalis). Pages 803-814 in D. R. McGullough and R. H. Barrett, editors. Wildlife 2001: populations. Elsevier, New York. Levins, R. 1970. Extinction. Lectures on Mathematics in the Life Sciences 2: 75-107. USDI. 1993. Final nile to list the Mexican Spotted Owl as a threatened species. Federal Register Volume 58, Number 49: 14248-14271. Received 25 July 1995 Accepted 2 October 1995 Great Basin Naturalist 56(1), © 1996, pp. 90-92 LIMBER PINE AND BEARS Heniy E. McCutchen^ Key words: Umber pine, black bears, food habits, Rocky Mountains. Limber pine [Pinus flexilis) is not consid- ered a fall food for black bears {Ursus ameri- canus) or grizzly bears {Ursus arctos) in the Rocky Mountain region of the United States. Previous studies have found that other nut- bearing plant species such as whitebark pine {P. alhicaulis) and Gambel oak {Quercus gam- belii) are preferred over limber pine by bears (Kendall 1983, Mace and Jonkel 1986, Beck 1991). However, these studies have been con- ducted only in areas where limber pine is in sympatry with other hard-mast species. During a study of black bears from 1984 to 1992 (McCutchen 1993) in Rocky Mountain National Park, it became apparent that bears utilized limber pine some years. This paper reports on that use of limber pine and dis- cusses the implications. Rocky Mountain National Park, encompas- sing 107,000 ha, contains elevations among the highest in the continental U.S., ranging from 2440 m to 4345 m. Nearly 1/3 of the area is alpine tundra above a 3200-m timberline. Below timberline, on the upper slopes, is a subalpine zone of Englemann spruce {Picea englemannii) and subalpine fir {Abies lasio- carpa). Extensive stands of lodgepole pine {Pinus contorta) and scattered stands of limber pine intermixed with other species are on middle slopes. At lower elevations Douglas-fir {Pseudotsuga menziesii) and ponderosa pine {Pinus ponderosa) are common. Limber pine is not found west of the Continental Divide in the park, and the species makes up only about 1% of the forest cover (Hess 1991). Between 1984 and 1991 I captured 40 indi- vidual black bears in and adjacent to Rocky Mountain National Park with culvert traps, with Aldrich foot snares, or at denning sites. Twenty-six bears were radiocollared (Telonics, Mesa, Arizona). Between 1984 and 1991, 9 sub- adult and adult bears (4 females, 5 males) were captured and radiocollared on the east side of the park. Radiolocations were primarily col- lected by triangulation from automobile or by hiking. I occasionally used snow tracking to determine bear activities. I used a vegetation type map (Hess 1991) to determine the pro- portion of radiolocations in stands containing limber pine. Bear scats were collected and analyzed at the Composition Analysis Labora- tory, Fort Collins, Colorado, on a gross and microhistological scale (Sparks and Malacheck 1968). Two female bears (2 and 3) were monitored intensively in 1985 and 1986 (McCutchen 1989). Bear 3, a 3-year-old, was captured on 6 August 1985. In 1985 bear 3 spent a signifi- cant amount of time in limber pine stands in fall. During the summer, from 6 August to 3 September, she stayed below 3047 m. We located her 11 times, and none of these loca- tions were in limber pine. In fall, after annual plant senescence began to occur, she made a migration to near timberline and sta\'ed above 3047 m for the next month. From 3 September to 15 October we located her 14 times; 12 of these locations were in stands containing lim- ber pine. On 23 September and again on 11 October I tracked her in the snow and disco\'- ered that she had been feeding on nuts of lim- ber pine cones cached in red squirrel {Tamias- ciurius hudsonicus) middens. At each feeding site the area was littered with cone cores and scales, indicating that she spent considerable time removing nuts from cones. At 2 bed sites, 4 scats were found that consisted almost entirely of limber pine nut shells. She was radiolocated in stands containing limber pine until 15 October, when she moved and denned on 17 October The amount of time spent feeding in limber pine stands was high when calculated in rela- tion to die amount of time I estimate she was iNatioiial Biological Sunoy, Colorado Plateau Held Station, Northern Arizona Universit\\ Box 5(iI4. Klafistafl, AZ 860n. 90 1996] Notes 91 Tabi.K 1. Radiotixes ot black hears in and out ol linil)L'r pine stands in Koek) Mountain National Park, 1985-1990. Year 1985 1986 1987 1988 1989 1990 Bear # Out In % Out In 9f Out In % Out In % Out In % Out In % 2 30 I 3 39 0 0 11 0 0 8 1 11 10 2 17 11 I 8 3 13 12 48 47 0 0 5 1 17 — — — 15 3 17 12 1 8 12 — — — — — — 8 1 11 8 0 0 17 3 15 10 2 17 Total 43 13 86 0 24 2 16 1 42 8 33 4 out of the den during the year. Her emergence date from the den in 1985 was unknown be- cause she had not yet been captured. However, she emerged from the den in 1986 on 9 May. Assuming she emerged in 1985 about the same time (9 May) and denned on 17 October, she was out for about 160 d. During 1985 she fed in hmber pine areas from 16 September to 15 October, a period of 30 d, or 19% of her active time during the year. Bear 2 did not utihze hmber pine to the same extent as bear 3 in 1985. Ahhough home ranges of both were adjacent (McCutchen 1989), bear 2 was located in a Hmber pine area only once out of 30 radiolocations. Bears 2 and 3 were again intensively radiomonitored in 1986 but were not observed to use limber pine (Table 1). Another obsei^vation of bear use of limber pine habitat was made in 1991. A radiocollared 3-year-old male moved into bear 3's home range during emigration from his natal range about 20 km to the south. On 6 December he was tracked in the snow and was found to have dug up squirrel caches of limber pine cones and nuts. He was radiolocated in limber pine areas until 17 December. Further analysis of radiolocations from bears on the east side of the park indicated that 4 of 9 (bears 2, 3, 12, 24) had been located in limber pine habitat at least once, 3 of these several times (Table 1). Percentage of time indi- vidual bears were found in limber pine stands varied from 0% to 48%. Of 272 total radioloca- tions, bears were found in forest types con- taining limber pine 28 times, or 10.3%. The importance of limber pine for bears in the park during the 1985 radiotracking opera- tion was reinforced by 14 scat samples col- lected during that year. Four of these (29%) consisted almost entirely of limber pine seeds. A review of the literature on bear research north and south of the park in the Rocky Mountain region suggests that limber pine is not important if other hard-mast species are present. Black and grizzly bears fed on white- bark pine but not limber pine in Yellowstone National Park (Kendall 1983) and in northern Montana (Mace and Jonkel 1986). Aune and Kasworm (1989) found essentially no grizzly use of limber pine in 10 yr of study in the Montana Front Range. In Montana, Idaho, and most of Wyoming, whitebark pine is either the sole hard-mast species or is more common than limber pine. In south central Colorado, Beck (1991) found that black bears made long- distance movements to feed on acorns of Gambel oak but not on limber pine. However, there are areas in Colorado and Wyoming where limber pine is the sole hard-mast-pro- ducing species present and may be important to bears (Fig. 1). Bear preference for whitebark pine and Gambel oak over limber pine is probably related to several factors. Limber pine seeds are smaller than the other two, producing 10.8 X 10^ seeds/kg as compared to whitebark pine at 5.7 X 103 seeds/kg (McCaughey and Schmidt 1990) and Gambel oak at 1.3 X 10^ /kg (Haiper et al. 1985). Limber pine generally produces large seed crops at wide and irregular intervals with small amounts produced nearly every season. Whitebark pine seeds are produced at frequent and regular intervals (Harlow et al. 1979) with good crops produced at intervals of 3-5 yr (McCaughey and Schmidt 1990). In south central Colorado, Beck (1991) found Gambel oak production to be quite regular with only 1 massive acorn crop failure in 10 yr. From the limited number of obsei"vations of limber pine use by bears in Rocky Mountain National Park, I suggest that if limber pine is the only hard mast available during certain years, perhaps years of limited production of other foods, it may be an important food source for the survival of bears. This hypothe- sis needs to be tested by further research. 92 Great Basin Naturalist [Volume 56 I :::J Limber pine only 1 I Otiier species preferred Fig. 1. Distribution map of limber pine in the Rocky Mountain region in relation to other hard-mast bear foods based on Little (1971). Limber pine only (shaded fill) des- ignates areas where limber pine occurs exclusive of any other hard-mast species. In this area limber pine may be important to bears as a food source. Other species pre- ferred (dotted fill) designates areas of limber pine distrib- ution where other hard-mast species, whitebark pine to the north and Gambel oak to the south, dominate and are preferred by bears over limber pine. (Note: State and county boundaiies are shown to locate limber pine range; GNP'== Glacier National I'ark. YNP = Yellowstone National Park, RMP = Rocky Mountain National Park, MP = Monarch Pass.) Acknowledgments Funding for this research was provided by the National Park Service. I thank David Stevens, Robert Schiller, and the staff of Rocky Mountain National Park for their support on this project. Literature Cited Aune, K., and W. Kasvvorm. 1989. Final report East Front grizzly studies. Montana Department of Fish, Wildlife and Parks, Helena. 332 pp. Bec:k, T. 1991. Black bears of west-central Colorado. Technical Publication 39. Colorado Division of Wildlife, Fort Collins, CO. 86 pp. Harlow, W. M., E. S. Harrar, and F M. White. 1979. Textbook of dendrology. 6th edition. McGraw-Hill, New York. 510 pp. Harper, K. T, E J. W.^gstaff, and L. Kunzler. 1985. Biology and management of the Gambel oak vegeta- tive type; a literature review. General Technical Report I NT- 179. U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station, Ogden, UT. 31 pp. Hess, K. 1991. Description and e\aluation of co\'er t>pes in the Rock^ Mountain National Park. Final report to Rocky Mountain National Park, Colorado. Januan- 1991. 195 pp. Kendall, K. C. 1983. Use of pine nuts by grizzly and black bears in the Yellowstone area. International Confer- ence of Bear Research and Management 5: 166-173. Little, E. L., Jr. 1971. Atlas of United States trees. Vol- ume 1. Conifers and important hardwoods. Miscella- neous Publication 1146. U.S. Department of Agricul- ture, Forest Sei-vice, Washington, DC. 200 pp. Mace, R. D., and C. J. Jonkel. 1986. Local food habits of the grizzly bear in Montana. International Conference on Bear Research and Management 6; 105-110. McCaughey, W. W, and \V. C. Schmidt. 1990. Autecolog> of whitebark pine. Pages 85-96 in W. C. Schmidt and K. J. McDonald, editors. Proceedings of a sym- posium on whitebark pine ecos\stems; ecology' and management of a high mountain resource, Bozeman, MX 29-31 March 1989. General Technical Report INT-270. U.S. Department of Agriculture, Forest Ser- vice, Intermountain Forest and Range Reseaich Sta- tion, Ogden, UT. 386 pp. McCuTCHEN, H. E. 1989. Cnptic beha\'ior of black bears (Ursits americanus) in Rocky Mountain National Park, Colorado. International Conference on Bear Research and Management 8; 65-72. . 1993. Ecolog>' of a high mountain black bear pop- ulation in relation to land use at Rocky Mountain NP Park Science 13; 25-27. Sharks, D. R., and J. C. M.-\lachek. 1968. Estimating per- centage diy weight in diets using a microscopic tech- niciue. Journal of Range Management 21: 264-265. Received 4 January 1995 Accepted 14 August 1995 Great Basin Naturalist 56(1), © 1996, p. 93 BOOK REVIEW Utah Wildflowers: A Field Guide to Northern and Central Mountains and Valleys. Richard J. Shaw. Utah State University Press, Logan, UT. 1995. $12.95 softback. ' Wildflovver books belong to a genre of pub- lications specifically designed for people who wish to see and identify pretty flowers. The wildflowers of the region covered by this handsomely designed book are certainly wor- thy of such a publication. It is conveniently sized for carrying into the field and presents species by flower color, as in many other books of this kind. This enables the user to find potential identities of plants encountered in the field. As in practically all other wildflower books, the writer confronts the enigma of presenting an overall view of the plant or emphasizing the flowers alone. It is the impossibility again of having a wide-angle telephoto lens. The images are clear and shaip, and if the user is able to make the comparison of flowers alone, then the book will be very useful as an identifica- tion tool. The author of the book also had to make arbitrary decisions on which examples to treat. There are more than a thousand species of flowering plants in the region covered by this book, which treats some 92 of them. Those presented are, however, beautiful. This book should be enjoyed for more than its usefiilness in identification. It can be viewed in those times of year, and in those places, where wildflowers are not flowering. The pho- tos will add chanii and understanding by them- selves. The author and the press responsible for production of this book should be compli- mented. Stanley L. Welsh Life Science Museum Brigham Young University Provo, UT 84602 93 INFORMATION FOR AUTHORS The Great Basin Naturalist welcomes previously uiipuhlished manuscripts pertaining to the biologi- cal natural history of western North America. Preference will he given to concise manuscripts of up to 12,000 words. Simple species lists are dis- couraged. SUBMIT MANUSCRIPTS to Richard W. Baumann, Editor, Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young University, Provo, UT 84602-0200. A cover letter accompanying the man- uscript must include phone number(s) of the author submitting the manuscript, and FAX number and E-mail address when applicable; the letter must also provide information describing the extent to which data, text, or illustrations have been used in other papers or books that are published, in press, submitted, or soon to be submitted elsewhere. Authors should adhere to the following guidelines; manuscripts not so prepared may be returned for revision. MANUSCRIPT PREPARATION. In general, the Great Bosin Naturalist follows recommendations in Scientific Style and Format: The CBE Manual for Authors, Editors, and Publishers, 6th edition (Council of Biology Editors, Inc., II South LaSalle Street, Suite 1400, Chicago, IL 60603, USA; phone 312-201-0101; F.\x 312-201-0214). We do, however, differ in our treatment of entries in Literature Cited. Authors may consult Vol. 51, No. 2 of this journal for specific instructions on format; these instruc- tions. Guidelines for Manuscripts Submitted TO THE Great Basin Naturalist, are printed at the back of the issue. Also, check the most recent issue of the Great Basin Naturalist for changes. TYPE AND DOUBLE SPACE all materials, including literature cited, table headings, and figure legends. Avoid hyphenated words at the right-hand margins. Underline words to be printed in italics. Use stan- dard bond (22x28 cm), leaving 2.5-cm margins on all sides. SUBMIT 3 COPIES of the manuscript and the origi- nal on a 5.25- or 3.5-inch disk utilizing WordPerfect 5.1 or above. Number all pages and assemble each copy separately: title page, abstract and key words, text, acknowledgments, literature cited, appendices, tables, figure legends, figures. TITLE PAGE includes an infomiative title no longer than 15 words, names and addresses of authors, a running head of fewer than 40 letters and spaces, footnotes to indicate change of address and author to whom correspondence should be addressed if other than the first author. ABSTRACT states the purpose, methods, results, and conclusions of the research. It is followed by 6-12 key words, listed in order of decreasing importance, to be used for indexing. TEXT has centered main headings printed in all capital letters; second-level headings are centered in upper- and lowercase letters; third-level head- ings begin paragraphs. VOUCHER SPECIMENS. Authors are encouraged to designate, properly prepare, label, and deposit high-cjuality voucher specimens and cultures docu- menting their research in an established permanent collection, and to cite the repository in publication. REFERENCES IN THE TEXT are cited by author and date: e.g., Nfartin (1989) or (Martin 1989). Multiple citations should be separated by commas and listed in chronological order. Use "et al. " after name of first author for citations having more than two authors. ACKNOWLEDGMENTS, under a centered main heading, include special publication numbers when appropriate. LITERATURE CITED, also under a centered main heading, lists references alphabetically in the fol- lowing formats: Mack, G. D., and L. D. Flake. 1980. Habitat rela- tionships of waterfowl broods on South Dakota stock ponds. Journal of Wildlife Management 44: 695-700. Sousa, W. P 1985. Disturbance and patch dynamics on rocky intertidal shores. Pages 101-124 in S. T. A. Pickett and P S. White, eds.. The ecolo- gy of natural disturbance and patch dynamics. Academic Press, New York. Coulson, R. N., and J. A. Witter. 1984. Forest ento- mology: ecology and management. John Wiley and Sons, Inc., New York. 669 pp. TABLES are double spaced on separate sheets and designed to fit the width of either a single column or a page. Use lowercase letters to indicate foot- notes. PHOTOCOPIES OF FIGURES are submitted initially with the manuscript; editors may suggest changes. Lettering on figures should be large enough to withstand reduction to one- or two-column width. Originals must be no larger than 22x28 cm. NOTES. If the manuscript would be more appro- priate as a short communication or note, follow the above instructions but do not include an abstract. A CHARGE of $50 per page is made for articles published; the rate for individual subscribers will be $35 per page. However, manuscripts with com- plex tables and/or numerous half-tones will be assessed an additional charge. Reprints may be pur- chased at the time of publication (an order form is sent with the proofs). FINAL CHECK: • Cover letter explaining any duplication of information and providing phone number(s), FAX number, and E-mail address • 3 copies of the manuscript and WordPerfect diskette • Conformity with instructions • Photocopies of illustrations (ISSN 001 7-3614) GREAT BASIN NATURALIST Vol 56 no lJanuaryl996 CONTENTS Articles Temporal and spatial dishibution of highway mortality of mule deer on newly con- structed roads at Jordanelle Reservoir, Utah Laura A. Romin and John A. Bissonette 1 Exceptional fish yield in a mid-elevation Utah trout reservoir: effects of angling regulations Wayne A. Wurtsbaugh, David Barnard, and Thomas Pettengill 1 2 Consumption of diffuse knapweed by two species of polyphagous grasshoppers (Orthoptera: Acrididae) in southern Idaho Dennis J. Fielding, M. A. Brusven, and L. E Kish 22 Fire frequency and the vegetative mosaic of a spruce-fir forest in northern Utah Linda Wadleigh and Michael J. Jenkins 28 Arizona distribution of three Sonoran Desert anurans: Bufo retiforrnis, Gastrophryne olivacea, and Pternohijla fodiens Brian K. Sullivan, Robert W. Bowker, Keith B. Malmos, and Erik W. A. Gergus 38 Habitat affinities of bats from northeastern Nevada Mark A. Ports and Peter V Bradley 48 Nuptial, pre, and postnuptial activity of the thatching ant, Formica ohscuripes Forel, in Colorado John R. Conway 54 Trachytes kaliszewskii n. sp. (Acari: Uropodina) from the Great Basin (Utah, USA), with remarks on the habitats and distribution of the members of the genus Trachytes Jerzy Bloszyk and Pawel Szymkowiak 59 Productivity, food habits, and associated variables of Barn Owls utilizing nest boxes in north central Utah Sandra J. Looman, Dennis L. Shirley, and Clayton M. White 73 Notes Astragalus laxmannii Jacquin (Leguminosae) in North America R. C. Barneby and S. L. Welsh 85 Intermountain movement by Mexican Spotted Owls {Strix occidentalis lucida) R. J. Gutierrez, Mark E. Seamans, and M. Zachariah Peery 87 Limber pine and bears Henry E. McCutchen 90 Book Review Utah wildflowers: a field guide to northern and central mountains and valleys Richard J. Shaw Stanley L. Welsh 93 H E ^A!\> GREAT BASIN NATURALIST VOLUME 56 N2 2 — APRIL 1996 BRIGHAM YOUNG UNIVERSITY GREAT BASIN NATURALIST Editor Richard W. Baumann 290 MLBM PO Box 20200 Brigham Young University Provo, UT 84602-0200 801-378-5053 FAX 801-378-3733 Assistant Editor Nathan M. Smith 190 MLBM PO Box 26879 Brigham Young University Provo, UT 84602-6879 801-378-6688 E-mail: NMS@HBLL1.BYU.EDU Associate Editors Michael A. Bovvers Blandy Experimental Fann, University of Virginia, Box 175, Boyce, VA 22620 J. R. Callahan Museum of Southwestern Biology, University' of New Mexico, Albuquerque, NM Mailing address: Box 3140, Hemet, CA 92546 Jeffrey J. Johansen Department of Biology, John Carroll University University Heights, OH 441 18 Boris C. Kondratieff Department of Entomology', Colorado State Universitv, Fort Collins, CO 80523 Paul C. Marsh Center for Environmental Studies, Arizona State University, Tempe, AZ 85287 Stanley D. Smith Department of Biology University of Nevada-Las Vegas Las Vegas, NV 89154-4004 Paul T. Tueller Department of Environmental Resource Sciences Universitv of Nevada-Reno, 1000 Vallev Road Reno, NV 89512 Robert C. Whitmore Division of Forestry, Box 6125, West Virginia University, Morgantown, WV 26506-6125 Editorial Board. Jerran T. Flinders, Chairman, Botany and Range Science; Duke S. Rogers, Zoolog}'; Wilford M. Hess, Botany and Range Science; Richard R. Tolman, Zoology. All are at Brigham '\bung University. Ex Officio Editorial Board members include Steven L. Taylor, College of Biology and Agriculture; H. Duane Smith, Director, Monte L. Bean Life Science Museum; Richard W. Baumann, Editor, Great Basin Naturalist. The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University. Unpublished manuscripts that Rniher our biological understanding of the Great Basin and suirounding areas in western North America are accepted for publication. Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1996 are $25 for individual sub- scribers ($30 outside the United States) and $50 for institutions. The price of single issues is $12. All back issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other busi- ness should be directed to the Editor, Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young University, Provo, UT 84602-0200. Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin Naturalist through a continuing exchange of scholarlv publications should contact the Exchange Librarian, 6385 HBLL, PO Box 26889, Brigham \bung University, Provo, UT 84602-6889. Editorial Production Staff JoAnne Abel Technical Editor Jan Spencer Assistant to the Editor Copyright © 1996 by Brigham Young University Official publication date: 29 April 1996 ISSN 0017-3614 4-96 750 17922 The Great Basin Naturalist Published at Proxo, Utah, by Brigham Young University ISSN 0017-3614 Volume 56 30 April 1996 No. 2 Great Basin Naturalist 56(2), © 1996, pp. 95-118 SELECTING WILDERNESS AREAS TO CONSERVE UTAH'S BIOLOGICAL DIVERSITY Diane W. Davidson l, William D. Newmark-, Jack W. Sites, Jr.'^ Dennis K. Shiozawa'^, Eric A. Rickart-, Kimball T. Harper"', and Robert R. Keiter^ Abstr.\c:t. — Congress is currently evaluating the wilderness status of Bureau of Land Management (BLM) public lands in Utah. Wilderness areas play many important roles, and one critical role is the consei'vation of biological diver- sity. We propose that objectives for conser\'ing biodiversity on BLM lands in Utah be to (1) ensure the long-term popu- lation viability of native animal and plant species, (2) maintain the critical ecological and evolutionaiy processes upon which these species depend, and (3) preserve the full range of commimities, successional stages, and environmental gra- dients. To achieve these objectives, wilderness areas should be selected so as to protect large, contiguous areas, augment existing protected areas, buffer wilderness areas with multiple-use public lands, interconnect existing protected areas with dispersal and movement corridors, conserve entire watersheds and elevational gradients, protect native communi- ties from invasions of e.xotic species, protect sites of maximum species diversity, protect sites with rare and endemic species, and protect habitats of threatened and endangered species. We use a few comparatively well-studied ta.xa as examples to highlight the importance of particular BLM lands. Key words: wilderness, biodiversity, conservation. Utah, Bureau of Land Management, endemic species, exotic species, cryptobiotic soils, plants, bees, vertebrates. The Wilderness Act and Biodiversity historical value" (16 U.S. Code, § 1131 [c][4]). Ecological concerns have also figured promi- In the Wilderness Act of 1964, Congress nently in several congressional wilderness endorsed the presentation of federal land in its bills for Bureau of Land Management (BLM) natural state (16 U.S. Code, Sections 1131-36). public lands. Both the Alaska National Interest Congress plainly anticipated that ecological Lands Conservation Act, 16 U.S. Code, § 3101 considerations w^ere an important dimension (b), and the California Desert Protection Act, of the wilderness concept, since the act pro- 103 Public Law 433 Section 2 (b) (1) (B) (1994), vides that wilderness may contain "ecological" expressly acknowledge that wilderness designa- features of "scientific, educational, scenic, or tion is intended to protect important ecological 1 15fpartiiu-iit ot BinloKx', Unh'ersih- of Utah, Salt Lake Git)'. UT S4112. ^Utah Museum of Natural History, Universit\- of Ut;di, Salt Lake City, UT 84112. ■'Department of Zoology, Brigham Young University, Provo, UT 84602. "^Department of Botan\', Brigham Young University, Provo. UT 84602. ^College of Law, Uni%ersit\ of Utah, Salt Lake Cit\', UT 84112. 95 96 Great Basin Naturalist [Volume 56 values. Among the significant ecological func- tions of wilderness areas is their role in con- sei^ving biological diversity (biodiversity). In Utah, undeveloped public lands admin- istered by the BLM (Fig. 1) can potentially play a key role in conserving the state's natural heritage. The BLM is now pursuing an ecosys- tem management policy designed to ensure sustainable ecological processes and biological diversity on lands under its jurisdiction (Depart- ment of the Interior 1994). By using these same criteria to designate wilderness areas. Congress could not only advance the BLM's ecosystem management goals but also reduce conflict over the agency's multiple-use lands (e.g., by diminishing the risk of future endan- gered species listings and the accompanying regulatoiy limitations). Over the long tenn, it is both cheaper and easier to protect species in aggregate in their intact, functioning ecosys- tems than to conserve them individually in fragmented and decimated populations under the Endangered Species Act. In short, the use of biological and ecologi- cal criteria to designate BLM wilderness areas in Utah is consistent with the legal concept of wilderness and would help to avoid future conflicts over resource management. BioDivERSiTi' Defined Biological diversity — the variety of life in a given area — includes three hierarchical com- ponents: genetic diversity, species diversity, and ecosystem diversity (e.g.. National Research Council 1978, Wilson 1988, Reid and Miller 1989, Raven 1992). Cenetic diversity refers to the variety of genes within species. Depletion of genetic diversity during population bottle- necks, or because of inbreeding within frag- mented and isolated populations, can threaten a species' sundval by reducing the capacity of organisms to adapt to changing environments (Soule and Wilcox 1980, Frankel and Soule 1981). Species diversity, or the number of species within a region (species richness), can be divided into three major components (Whittaker 1972): alpha diversity {a), the num- ber of species in a homogeneous habitat; beta diversity (/3), the rate of species-turnover across habitats; and gamma diversity (y), the total number of species observed in all habi- tats within a region. Finally, ecosystem diver- Fig. 1. Map of the state of Utah showing (in black) loca- tions of all existing roadless areas proposed for BLM wilderness status. The BLM formally studied a suliset of these areas and recommended a portion ot studied lands for wilderness status. Data are from a Department of Inte- rior map of BLM Wilderness Study Areas, BLM Proposed Wilderness, and the Utah Wilderness Coalition's BLM Wilderness Proposal. County boundaries also are shown. Isolated moiuitain ranges in Utah's western deserts are identified as follows: a = Deep Creek; b = Fish Springs; c — House range, and d = Newfoundland range (not for- mally proposed or studied for wilderness designation). On the Colorado Plateau, e = the Henn' Mountains. sity consists of the xariety of major ecological communities within areas that are heteroge- neous in their physical attributes, for example, in elevation or soil type. Genetic, species, and ecosystem di\'ersity all result from both interactions bet\\'een organ- isms and their environments, and interactions of organisms with one another. The physical environment sets limits on wliich species can inhabit an area, and interactions among those species determine which are most abundant. Strategies for preserving biodixcrsitx' must therefore take note of all li\ ing things in the landscape, and the linkages among them. Finally, since different species specialize on different stages of natural disturbance cycles, it is important to presei-ve a range of commu- nities and ecosystems representing all stages in the disturbance cvcle. 1996] WlLDKRNESS SELECTION FOR BlODIN KHSITY 97 Objectives The success of conserving biological diver- sit) within a s\ stem ot" protected areas can only be assessed in relationship to a series of selected objectives. We propose that the con- sei-vation of Utah's biological diversity depends on (1) ensuring die long-term viability of native plant and animal populations, (2) maintaining the criticd ecological and evolutionaiy processes upon which these species depend, and (3) pro- tecting the full range of communities, succes- sional stages, and environmental gradients (e.g., lUCN 1978, MacKinnon et al. 1986, Noss 1992). Both the size of the network of protected areas and the selection of individual wilderness areas should be guided by these 3 goals. Although it is possible to presence a small sub- set of species and genotypes in zoological and botanical gardens, communities and species interactions must be consei-ved in situ. Large areas with minimal human intrusion, and with natural processes reasonabK' intact, are critical elements of an in situ conservation strategy; tliey provide protection for fiagile habitats, such as easily eroded soils, and preserve habitat for reclusive species. Moreover, wilderness areas offer natural ecosystems some protection from the biological invasions that have devastated many communities, especially plant communi- ties, across Utah. Here we describe a strategy, based upon widely accepted principles of conservation biology (see e.g., Primack 1993, Meffe and Carroll 1994), for both selecting critical sites for wilderness designation and determining the amount of habitat that should be pre- served as wilderness (see also Babbitt 1995). Criteria for Selection Viable Populations Utah contains approximately 3000 indige- nous plant species and varieties and about 584 vertebrate species. Viable populations for most of these plants and animals can be ensured by focusing, within ecological communities, on species for which the risk of extinction is greatest. Risk-prone species typically include those with small populations, large home range requirements, low reproductive poten- tial, restricted geographic ranges, or large temporal xariation in population size (Brown 1971, Willis 1974, Terborgh and Winter 1980, Diamond 1984, Pimm et al. 1988, Belovsky et al. 1994, Newmark 1995). Many top predators have several of these traits. On BLM lands in Utah, examples of such organisms are river otter {Lutra (•anadi'iisi.s) and both Bald and Golden Eagles {Haliaeefus lencoc('})lialu.s and AqiiiJa chrijsaetos). Risk-prone plants include Holmgren locoweed {Astragalus hobngrenio- nim) and Jones cycladenia {Cijclaclenia huinilis var. jonesii), which have highly specific sub- strate recjuirements . Viability of populations depends on both the level of risk one is willing to accept, and the time frame over which one wishes to consene the population (Shaffer 1981, Schonewald-Cox 1983, Soule 1987). In general, both survival time and the likelihood of population persis- tence increase with population size. A level of risk and persistence that is commonly pro- posed as a management goal is a 99% chance of sui-vival for 1000 years (e.g., Belovsky 1987, Armbruster and Lande 1993). For large carnivores, the minimum viable population necessary to ensure a 99% chance of survival for 1000 vears is estimated to be approximately 10,000-100,000 individuals (Be- lovskv 1987). In habitat area, this is equivalent to 100,000-1,000,000 km2, or 2.5-25 million acres. Although this area requirement may seem remarkably large, documented losses of mammalian species from among the largest of North American national parks (e.g., the 10,328-km^ Yellowstone-Grand Teton park assemblage) during the last 90 years make clear the importance of protecting large areas (Newmark 1987, 1995). Maintenance of Ecological and Evolutionary Processes In selecting wilderness areas, one must take care to ensure the maintenance of the ecological and evolutionary processes upon which all plant and animal species depend (Pickett and Thompson 1978, Kushlan 1979). Among the most important of these processes are natural disturbance and recovery cycles. Ideally, criteria for the selection of wilderness areas should include information on fre- quency, size, and longevity of natural distur- bances. Protected areas should be large enough to contain minimum critical areas of the entire range of recovery stages for each community type (Pickett and Thompson 1978). In western North America, natural dis- turbance regimes can encompass tens of thou- sands to millions of acres, as witnessed by the recent and extensive wildfires in Yellowstone National Park (Christensen et al. 1989). 98 Great Basin Natueulist [Volume 56 Two other critical ecological processes are migration and dispersal of terrestrial organ- isms across landscapes, and of aquatic species within watersheds. The selection of wilder- ness areas requires that attention be given to ensuring that migratoiT pathways are open to organisms migrating seasonally along eleva- tional gradients. Of particular importance is the need to maintain winter ranges and migra- tory routes of large mammals such as mule deer {Odocoileus hemioniis), elk {Cervus elaphiis), and moose {Alces dices). Interactions among competitors, and be- tween predators and prey, are integral aspects of natural ecosystems and should be pre- served. For example, in the southwestern deserts of the United States, the direct and indirect effects of seed predation on plant community structure have been documented in long-term experiments manipulating densi- ties of rodent and ant grani\'ores (Daxidson et al. 19S4, Samson et al. 1992). These effects include transformation ol a shrubland into a grassland biome (Brown and Heske 1990). Special care must be taken to consei^ve popu- lations of predators with large area require- ments, because extinctions of these species can alter whole communities (e.g., by leading to outbreak densities of prey, which then over- exploit their plant resources). Some of the strongest e\'idence for such "trophic cascades" comes from the Greater Yellowstone Ecosys- tem, where intensive browsing by elk has greatly altered many riparian zones by the re- moval of willows (genus Salix), and has elimi- nated aspen seedlings {Popiihis fremiiloides) recruiting from seeds and rhizomes shortly after the extensive 1988 fires. Huge contem- porary elk herds, numbering ~ 40,000 individ- uals in the park, and 20,000 in the northern herd alone, are likely the result of reductions in the full complement of large predators (Kav 1990, Wagner et al. 1995). Gonsidcrable evi'- dence also suggests that deer and elk herds in Utah average significant!)' larger at present than during any extended period in the histor- ical past (Durrant 1950, Julander 1962, Haiper 1986). Strategies for Selecting Wilderness Areas Landscape-wide Priorities Given the large area requirements of many extinction-prone Utah species, it is important to protect large, contiguous land blocks. In designating wilderness areas, high priority should be given to lands whose selection would enlarge and connect existing protected areas (e.g., national parks, wildlife refuges, and Forest Service wilderness areas) and thus enhance the viability of animal and plant pop- ulations (Newmark 1985, Salwasser et al. 1987, Noss 1992, Grumbine 1994). By themselves, BLM wilderness areas in Utah clearly cannot satisfy the huge area requirements noted above as requisite for maintaining viable populations of large carnivores. However, when linked to other public lands (e.g., Utahs national parks, and wilderness areas in other states), BLM wilderness in Utah can be a key component of strategies for long-term presei"vation of biolog- ical diversity. Other high-priority areas are those which, alone or together with other protected areas, encompass entire watersheds. In addition to affording direct benefits to humans, watershed protection is the most effective means of con- sei"ving the aquatic and riparian communities that account for a disproportionate fraction of both species diversity and endangered and threatened species in arid western North America (Miller 1961, Minckley and Deacon 1968, 1990, Holden et al. 1974, Johnson et al. 1977, Cross 1985, Knopf 1985, Moyle and Williams 1990). Moreover, since populations of riparian species are usually isolated from similar communities in other drainage systems, species losses from these environments are not easily remedied b>' natural recolonization. A 3rd priority in selecting wilderness sites is land that fomis or helps to complete the pro- tection of entire elevational gradients, for example, in isolated mountain ranges of the Great Basin. Scant attention paid to consemng these gradients in the past is evident in the restriction of most national parks and wilder- ness areas in western North America to higher elevation sites. Designation of wilderness in comparatively low elevation BLM lands would afford protection to regions of greatest species richness for man\' organisms (e.g., mammals, birds, amphibians, insects, and trees) whose diversity generally declines with elevation throughout much of western North America (Harris 1984, Ste^'ens 1992). Optimal Design Goals 1 If BLM wilderness areas are to contribute substantialK' to the preser\'ation of biodiver- sitv in Utah, then site selection must take into 1996] Wilderness Selection for BiODiVERsm- 99 Buffer Zone account tlie 3 general goals outlined above. H ^'-^ wilderness Ideally, BLM wilderness lands should form an <&% ^ . o .., ,. fijif;: Forest Service Wilderness interconnected core zone of roadless lands w hen combined with otlier federal wilderness H National Park service areas, national and state parks, and wildlife refuges (Fig. 2). Special attention should be Public Lands given to linking roadless lands so as to pre- clude further fragmentation of natural habitat. Inagmentation, or the transformation of an unbroken block of natural habitat into a num- ber of smaller patches separated by altered habitats, reduces population sizes, increases their isolation, and threatens their long-term viability. It is one of the greatest threats to bio- logical dixersit)' worldwide (Wilcox and Mur- phy 1985, Wilcove et al. 1986, Saunders et al. < 1991). Across diverse habitats, there are numer- ous examples of species extinctions precipi- 'KWi;;-^-;.— core zone tated by both natural and human-induced habitat fragmentation (e.g., Brown 1971, Ter- borgh and Winter 1980, Diamond 1984, Fig. 2. An example of a preferred arrangement of Heaney 1984, Patterson 1984, Newmark 1987, wilderness and multiple-use federal and state lands to 1991, 1995, Case and Codv 1988, Soule et al. conserve biological diversity. Wilderness areas adminis- irvoo n 1 L ^ inm\ ' teied bv the Bureau of Land Management, Forest Service, 1988, Bolger et al. 1991 . m ^- i d i c j u- i i ^^/i uf c \ 1 rr National rark bervice, and Fish and Wikllile Service Adjacent multiple-use lands can buffer ,]^o^,i,i fo^m a contiguous core zone in which the most human impacts on biological diversity within extinction-prone species in Utah can be protected. Multi- wilderness areas. Such lands can be expected pie-use lands can effectively buffer this core zone and to pro\'ide marginal habitat for tlie manv species P'^^'de additional marginal habitat to species that are pri- .1 . 1. • i 1 • -1 1. ■ I.- marilv restricted to roadless areas. tliat are restricted prnnanly to more pristme wilderness regions. Thus, proposed wilder- ness areas surrounded bv public lands should receive high priority for protection. their genes move about only through the pro- cesses of seed dispersal and pollen transport. Therefore, it is not surprising that many plants have narrowly restricted ranges, are locally adapted to conditions within those ranges, and are isolated, often by great distances, from other sites where similar conditions prevail. Although locally endemic plants can often be relatively abundant inside their ranges, their populations are easily jeopardized by habitat alteration (e.g., by all-terrain vehicles) within their narrow distributions. Of Utah's approxi- mately 2600 plant species and 400 named varieties (Albee et al. 1988, Welsh et al. 1993), about 180 (or 7% of species) are currently clas- sified by federal or state agencies as endan- gered, threatened, or sensitive. A majority of these (133, or —74%) definitely or probably occur on BLM lands (Atwood et al. 1991), and a substantial subset of the classified species are narrow endemics. Shultz (1993) provides a useful summary of endemism in the Utah flora. Approximately 240 species, or 10% of all Utah plant species, are endemic to the state. This rate of endemism. E.XAMPLES OF Rare and Endemic Species The design advocated above is based largely on conservation strategies for preserv- ing wide-ranging vertebrate species. Although such strategies can help to ensure the long- term viability of most species within a given region, exclusive reliance on such approaches may well overlook and endanger many locally isolated, rare, and endemic plants and animals. We cannot give a comprehensive treatment of this subject here, but we discuss 3 ta.xonomic groups of organisms for which especially high rates of endemism or existing threats to iso- lated populations present particular manage- ment dilemmas that should be taken into account in wilderness decisions. In most cases, specific habitats must be protected to assure the presei'vation of these species. Plants of Special Concern Unlike the wide-ranging animals discussed above, plants occupy fixed positions; they and 100 Great Basin Natufl\list [Volume 56 the percentage of the flora considered for hst- ing as threatened or endangered, and the per- centage of rare species in the flora are among the highest in the continental United States. The vast majority (86%) of Utah endemics reside in arid and semiarid regions of the state, and 90% are edaphicalK' restricted to fine-textured and/or high pH substrates (limestone, clay, silt, mudstone, and shale) that magnify drought stress. Plant distributions generally appear to respond more to edaphic, topographic, and geologic features of the environment when drought is a factor (Stebbins 1952). Because most endemics live in close proximity to mor- phologically similar species (Albee et al. 1988), these species appear to be mainly neoendemics that have evolved since the last glacial maxi- mum (18,000 yi's BP), or in the Bonneville basin during the past 10,000 >ts. Geographically, endemism of Utah plants is highest in the Canyonlands Phytogeographic Section of the Colorado Plateau Division of the Intermountain Region (Cronquist et al. 1972, Fig. 3 modified from Shultz et al. 1987). An unusual diversity of substrates occurs here, and these substrates are more apt to be exposed, rather than coxered with alhnium as in other areas of semiarid Utah (Welsh et al. 1993). Thus, fully 50% of Utah's 240 rare and endemic plant species occur on the Colorado Plateau, whereas just 15% occur in the Great Basin, 11% in the Mojave Desert, and 10% in the Uinta Desert (Welsh 1978, Shultz 1993). About half of Utah's endemics belong to just 5 genera that are both common and physiologi- cally adapted to aridity (total Utah species and percent endemics, in parentheses): Astragalus, Fabaceae (114, 36.8%), Penstetnon, Scrophulari- aceae (106, 26.4%), Cnjptantha, Boraginaceae (61, 36.1%), Eriogoniim, Polygonaceae (60, 23.3%), and Erigeron, Asteraceae (54, 24.1%; Welsh et al. 1975, Welsh 1978, Shultz 1993). Because most of the state's endemic plants are restricted to particular geologic formations, and because multiple endemics often occur on the same formation, groups of endemics gen- erally can be protected simultaneously by safe- guarding those soil formations and surround- ing areas. Two regions where large nimibers of endemics stand to benefit from wilderness protection of BLM lands are the Uinta Basin and the San Rafael Swell and surrounding San Rafael Desert (Fig. 3, Table 1; M. Windham personal communication). No fewer than 15 plant species are endemic to the region in and around the proposed wilderness area (PWA) near the White River south of Vernal (UWC 1990), and most of these are confined to the Parachute and Evacuation Creek members of the Green River Shale formation. Another dozen endemics occur in a diversity of habi- tats in and around the San Rafael Swell. Here the most important habitat is a beige (rather than red) Moenkopi formation, spatially iso- lated from other Moenkopi outcrops and un- usual in its soil chemistiy. A few endemics also occur on the younger Carmel and Summer- ville formations surrounding the core of the swell, especially between Muddy Creek and Crack Canyon (S. Welsh personal communica- tion). Wilderness designation in these 2 regions (the San Rafael PWA and the White River PWA of the Uinta Basin [Fig. 3]; see UWC 1990) could afford significant protection to some of Utah's endemic plants. South and east of the San Rafael, in the Dirty Dexil PWA (UWC 1990), are the distinctixe flora of the Orange Cliffs region (Fig. 3) and some additional nar- row endemics deserving protection in the Main and South forks of Happv Canvon (Shultz et al. 1987). The Moenkopi formation is also important as a substrate for endemics elsewhere in semi- arid Utah. Two federalK' listed endangered species, Arctomecon limnilis (the dwarf bear- claw poppy) and Pediocactiis sileri (a cactus), and several other species are endemic to par- ticular Moenkopi outcrops in southwestern Utah. Wherever possible, the boundaries of wilderness areas and other protected areas should encompass these specialized habitats. Bees and Wasps in the San Rafael Desert Because of their capacit) for directed mo\'e- ments, animals are less likely than plants to exhibit high rates of endemism. Nexertheless, since insects often tend to be host- or habitat- specific (e.g., in pollinators, herbixores, or sub- strate-specific ground nesters), endemism can often be high in insect taxa. Bees and wasps (order H\menoptera) are examples of such insects. Here, as elsewhere, bees and preda- tor)' wasps are especialh' di\'erse in arid regions (Michener 1979). The state supports a mini- mum of 950 species of native bees (roughly 25% of the total number of species known 1996] Wilderness Selection eor Biodin i.Ksin 101 SweU San Rafael Desert ^Orange CUffe Fig. 3. Satellite image of Utah showing the positions of the San Rafael Swell, the San Rafael Desert, and tlic Orange Cliffs, all within the Canyonlands Phytogeographic Section, ontlined in bold. The arrow in the Uinta Basin shows the approximate position of the White River PWA (Utah Wilderness Coalition 1990). from America north of Mexico), and 50 of the Utah species are currently inidescribed (T. Griswold, K Parker, and V. Tepedino personal communication). Many areas, especially in the southern part of the state, have not been explored intensively and undoubtedly harbor many additional undescribed species. Bees and plants often show comparable geo- graphic patterns in diversity and endemism (Neff and Simpson 1993), and many of the areas currently under consideration for wilderness designation in Utah are centers of endemism for both groups. Although we lack extensive in- formation on bees of the Canyonlands Section (Fig. 3), where endemism is highest for plants (see above), intensive collecting in that small part known as the San Rafael Desert has yielded a total of 316 species of bees, 42 of 102 Great Basin Natur.\list [Volume 56 Table 1. Plants endeinic to the 2 areas with the higliest eiuleinisiii on Utah BLM lands. Endemics of tlie sontheni Uinta Basi Endei the San Haiael Swel Aqtiilegia barnebtji Miniz (Ranunciilaceae) Asfragdiiis eqiiisolensis N'eese 6c Welsh (Fahaceae) A. hainiltonii C. Porter A. hitosii.s Jones A. saiiriniLs Barnehx Cirsiiiin hanwhiji Johnst. (^Asteraceae) Cryptcmtlid hantchiji Johnst. ( Boraginaceae) C grdliainii Johnst. CtjDioptcri.s (liiclu'siicn.sis Jones (Apiaceae) Pensteinoii floiccifiii Neese & \M'lsh (Scrophulariaceae) P. goodhcliii .\. Holmgren P. grahainii Keck SchoencniinlH' argilhicea (\\'elsh & Atwood) Rollins (Brassicaceae) S. .suff'ruti'.sceus (Rollins) Welsh 6c ChatterK' Sclerocactiis glaitciis (K. Schnm.) L. Benson Astragalus rafaclciisis Jones (Fabaceae) Cnjpfaiitha crciitzfclclii Welsh ( Boraginaceae) C. Johnstoiui Higgins C. joiu'siaiui (Pa\son) Pa\son Erigeroii inaquirci Cronquist (Asteraceae) Loinatiiimjiinceiiiii Banieh\ 6c N. Holmgren (Apiaceae) Lijgoclcsmia entrada Welsh 6c (iooilrieh (Asteraceae) Pcdiocactiis dcspaiiiii Welsh 6c C.ot)drieh (Cactaceae) Pcitsteinoit inarnisii (Keck) \. Holmgren (Scropluilariaceae) ScliOi'iicraiidH' harncbt/i (Welsh 6c .\tA\ood) Rollins (Brassicaceae) 'ndinuiii dtninpsoiiii Atwood 6c Welsh (Portiilacaceae) Tt>uiisciidi(i aprica Welsh 6c Re\eal (Asteraceae) which are presentK undeseribed (T. Griswold, E Paikei; and \^ Tepedino personal conniuuii- cation). Thus, 33% of the state's total species count, and 84% of Utah s undeseribed (but catalogued) species, are endemic to a region comprising just 2.0% ot the states land area. Fin-thermore, a signiticant portion of this tauna (24%) occin\s onI\' on the Colorado Plateau. The remainder of the Cainonlands Ph\togeo- graphic Section, in which the San Rafael Desert is embedded, is likeK" to be equalK di\ erse and to ha\ e as man> new species. Other hymenopteran groups, such as the aculeate \\asps, also are highly di\ erse in the San Rafael Desert (T. Gris\\ old, E Parker, and V Tepedino personal connnimication). For ex- ample, with a total of 22 species there, the cir- cinnglobal genus Fhildntluia is more di\"erse in the San Rafael Desert than an\A\'here else in North America, and probably the world. These predatoiy "digger wasps " nest in the soil and ma\ ha\ e di\"ersified in response to the \ aried substrates present in this desert. ClearK, des- ignation of wilderness in the San Rafael region (see UWC 1990) could afford significant pro- tection to an area of \er>- high endemism and di\ersity for the order H\nienoptera. Bees and wasps are among the most benefi- cial insects. Predaton' and parasitic wasps help to control populations of pest species (e.g., grasshoppers, aphids, etc.) below outbreak densities. An estimated 67% of flowering plants depend on insects (primariK' bees) for pollen transfer and sexual reproduction (Axlerod 1960), and the welfare of nian\ plant species in semiarid Utah assuredh depends on their relationships with bees. Eor example, a rare species of Pcrdita, found in Utah only at the BeeHi\e Dome site southeast of St. George, pollinates the rare and endangered dwarf bear- claw popp\' (^. Tepedino personal commimica- tion). Bees that ha\e specialized b\ collecting pollen onl\ from flow ers of a particular plant family, or exen from a single genus within a famih; are termed oligoleges. Such bees tend to be most common in arid regions (Neft and Simpson 1993) and generalK" are regarded as being closely adapted to the phenolog>' and floral traits of the plants on \\ hich the)' spe- cialize. Such adaptations tend to make them superior pollinators. Scjuash bees and squash flowers are examples of such a co-adapted pair in the Americas (Tepedino 1981). Some oligo- leges ma> one da\' proxe to be useful as crop pollinators. The legume specialist Osniia san- rafachic. a nati\ e ot the San Rafael Desert, has been inxestigated as a potential pollinator of alfalfa {Mcdicdgo sativa L.), an important for- age crop (Parker 1985, 1986). Man> of the species of the San Rafael Desert appear to be oligoleges. A brief list of some of the unde- seribed and recentK" described bee species and their host plants is pro\ ided in Table 2. These entries were chosen only to illustrate the \ariety of plant taxa upon which nati\e bees specialize. Nati\ e and Endemic Fishes Freshwater ecosxstems are natinal habitat "islands ; as sutli. thcii- long-tcnn isolation b\ 1996] WiiJ^EKNEss Selection for Biodivkksitv 103 Tahi.K 2. Pollen piffereiices for represcnlativf oli.ujolL'ctif liccs in llic San lialacl Dl'S(m1 (data Cioni 'I'. Crisvvold, F Farkt-r and V. Tepedinc) personal connnunication). Plant family Plant genus/species Bee species Asteraceae Boraginaceae Eiipliorbiaceae Fabaceae Loasaceae Onagraceae Papaveraceae Polenioniaeeae Scn)})liu!ariaceae Hcliiiiitliiis (inoinolu.s W'l/ctliid Kccil specie intei'vening terrestrial liabitats, or by unsuitable aquatie habitats, often promotes loeal speeial- ization, evoliitionaiy diversification, and endem- ism in aquatic organisms. Seven centers of endemism are recognized for fishes of western North America (Miller 1959), and Utah includes substantial portions of 2 of these centers, the Bonneville Basin and the Colorado River Basin. Collectively, 28 fish species are native to these basins (Smith 1978), and 27 are extant. Because of their limited distributions, en- demic species are easily endangered by both habitat alterations and introductions of nonna- tive competitors and predators. Seven species and subspecies from the Bonneville and Col- orado basins are now federally listed as endan- gered (U.S. Fish and Wildlife Ser-vice 1993). A further 11 species and subspecies are consid- ered by fisheiy specialists to be endangered, threatened, or of special concern in Utah (War- ren and Burr 1994). The decline of native fishes has been associated with both water- shed development (e.g., reservoirs, irrigation diversions, channelization, floodplain drainage) and the introduction of alien species. Conservation of endemic fish populations has been especially successful when much of the watershed has been protected (Williams 1991), but adherence to strict legal definitions of wilderness often precludes such wide- spread protection. In Utah, opportunities for protecting entire watersheds are limited to relatively small drainage systems extending from stream headwaters in mountain ranges of the Bonneville Basin to diy or saline lake beds at lower elevations. A particularly important case is in the Deep Creek Range, where the Bonneville cutthroat ti'out {Oncorhtjnclms clarki iitali), once thought to be extinct (Behnke 1992), survives in populations in Trout Creek and Birch Creek within the Deep Creek PWA (UWC 1990). Where protection of whole watersheds is not possible, wilderness that includes key habi- tats may help to stabilize declining populations of native fishes, preclude new listings and draft- ings of recovery plans, and promote recoveries and delistings. This should be the case most often for fishes living in headwater streams protected by natural and artificial downstream barriers from unintended invasions of alien cold-water species. For example, habitat in the upper Book Cliffs-Desolation Canyon PWA may support the Colorado River cutthroat trout {Oncorhyncluis clarki plenriticu.s), considered the rarest of the cutthroat taxa (Behnke and Zani 1976) and federally listed as a categon^ 2 species (Kerchner 1995). Although the region has not been surveyed for this subspecies, native populations occur in streams entering the Duschesne River from the north (Shiozawa and Evans 1994) and have recently been found in streams of the western Book Cliffs, closer to Price and Soldier Summit (Shiozawa and Evans unpublished data). Given these obsei-vations, it is likely that streams flowing into the Book Cliffs-Desolation Canyon PWA will also con- tain this subspecies. In relatively large downstream systems (secondary and tertiaiy streams), key habitats include floodplain wetlands, among the first habitats to be lost due to human activities. Although wetlands have been viewed tradi- tionally either as breeding sources for insect 104 Great Basin Naturalist [Volume 56 pests or as waterfowl production sites, periodic or continuous connection to rivers renders them important appendages to lotic systems. Densities of aquatic invertebrates are signifi- cantly higher in wetlands than in main river channels, over 100-fold in some cases (Wolz and Shiozawa 1995, Mabey and Shiozawa unpublished data). Floodplain wetlands can therefore serve as important nurseiy grounds for laival and immature native fishes. The loss of wetlands may be a significant factor endangering sexeral native fishes in the Colorado River (Tyus and Karp 1989). Fishes native to the larger streams and rivers of the Colorado River Basin are predominantly min- nows (Cyprinidae) and suckers (Catostomidae) that have evolved in isolation, are adapted to unique local conditions of this drainage (e.g., heav\' silt loads and wide fluctuations in dis- charge and temperature), and are the most moiphologicalK' distinct fishes in North Amer- ica (Hubbs 1940, 1941, Deacon and Minckley 1974, Minckley et al. 1986). Four of these native species, the Colorado squawfish {Pty- chocheilus lucius), the humpback chub {Gila cijpha), the bonytail chub {Gila elegans), and the razorback sucker {Xyrauchen texamis), are now federally listed as endangered. The decline of both the bluehead sucker {Catostomus [Pan- tosfeii.s] discobolus) and the flannelmouth sucker {Catostomus latipinnis) within the main stems of the Colorado and Green rivers may result in their listings as threatened, especially if popu- lations in tributaiy streams are not stabilized. Several of these species occur in areas under consideration for wilderness status. Both the Price River, in the Book Cliffs-Desolation Can- yon PWA, and the San Rafael River, in the San Rafael PWA, have populations of roundtail chub, flannelmoudi sucker, and bluehead sucker Bluehead sucker are also known from the Dirty Devil and Muddy Creek drainages (Smith 1966), and both flannelmouth sucker and round- tail chub are likely to occur there. Wilderness designation could broaden the protected ranges of several of these species by stabilizing wet- land habitats in the Dirty Devil, San Rafael, and Book Cliffs-Desolation Canyon PWAs. Although the Virgin River drainage is also part of the Colorado River Basin, it has a unique fish fauna that appears to have evoKed in isolation from populations in other parts of the basin. The Virgin River spinedace {Lepi- domeda mollispinus). the woundfin {Plagoptenis argentissimus), and the Virgin River chub {Gila robusta seminuda) are endemic to this system. Two additional species, the flannelmouth sucker and the desert sucker {Catostomus clarki), have evolved very slender caudal peduncles, possibly as a response to occasional high flows in the Virgin River (Smith 1966). The health of this unique fish fauna already is cause for concern. Two of the endemics, the woundfin and the Virgin River chub, are feder- ally listed as endangered. Although the desert sucker occurs in Arizona, Nevada, and New Mexico, this species merits special concern in Utah (Utah Division of Wildlife Resources [UDWR] 1992), where it is limited to the Virgin Ri\'er drainage. Loss of either this species or the flannelmouth sucker from the Virgin River system would eliminate only a subset of their existing populations and is unlikely to move either species to endangered status. However, the uniqueness of these populations (Smith 1966) may warrant their designation as sepa- rate subspecies. This, toge flier wifli the concern now e\'idenced for the flannelmouth sucker throughout its range, could easily translate into candidacy for listing if existing populations are not protected. Concern for native fishes of the Virgin River drainage has already constrained water devel- opment ill Washington Count); Utali. An>' actions that would help presene the integrit}' of ripar- ian habitat and stream channels would also reduce stress for these fishes. Since the integ- rit\' of riparian habitats is best maintained over large areas, wilderness designation in PWAs of the Beaver Dam slope and the greater Zion area would sei"ve this purpose. Finally, protection of Utah s rare and en- dangered fishes would likeK also afford signif- icant protection to other aquatic organisms, for example, Utah's diverse communities of aquatic insects. Reciprocally, the maintenance of high species diversity in stream insect com- munities is critical to assuring a continuous food supply to fishes in rivers with wide sea- sonal and annual fluctuations in flow rates. Mayflies (Ephemeroptera) are among the best- studied stream insects in Utah, and 16-18 genera (22-24 species) are known from warm water tributaries of the Colorado Rixer sxsteni (G. Edmunds personal commimication). Con- struction of reservoirs on these rivers has iilready inundated many river miles and altered flow rates, sediment loads, and downstream 1996] Wilderness Selection for Biodiversity 105 teniperutures. Mayflies mid other aquatic insects are highly sensitive to all these variables. Unnatinalh' constant temperatures in tailwaters beneath dams can lead to depauperate com- munities of ma) flies and other stream insects, for example, below Flaming Gorge Reservoir (Edmunds 1994, 1995). (Four mayfly genera from this area of extremely high natural diver- sity have not been collected since the dam was built.) Habitats rich in mayflies and other aquatic insects, and most in need of protection from future impoundments, include the Green River from the Colorado border to Ouray, Utah, and the Colorado River from the Colo- rado border to Moab, Utah. Relatively warm sections of the Duchesne, Uintah, White, Escalante, Virgin, and Santa Clara rivers would also be sensitive to manipulations of stream flows. Examples of Biologically Important Sites on BLM Lands The floras and faunas in different parts of Utah have unique evolutionaiy histories deter- mined by the geography and topography of the lands they inhabit. In this section, we dis- cuss 4 such sites in the context of important scientific criteria (outlined above) for wilder- ness site selection. We also review various sci- entific and educational values of these same sites. Book Cliffs and the Tavaputs Plateau For several reasons, the Book Cliffs and Tavaputs Plateau areas, along both sides of the Green River, are critical for the long-term con- servation of biological diversity in Utah. This region contains some of the largest remaining roadless areas on BLM lands in Utah (Fig. 1) and therefore provides important habitat for sensitive species with large area requirements. It includes broad elevational gradients with the potential to protect a wide range of natural communities and to maintain crucial routes for seasonal wildlife migration between high and low elevation. Furthermore, it constitutes a vital dispersal coiridor linking the Uinta moun- tains to the north and the Colorado Plateau to the south. Because of both the high habitat diversity and the central location of the Book Cliffs- Tavaputs region, the biota is unusually diverse and compositionally unique, and includes many species at their distributional limits. Among reptiles and amphibians, for example, the Great Basin spadefoot toad {Scaphiojms intennon- Uinus), the western whiptail lizard {Cnemi- dopJwrm fi^ris), and possibly the rubber boa {Charina bottae) reach their eastern distribu- tional limits here. Three additional species, the longnose leopard lizard {Gmnhclia wis- lizenii), the collared lizard {Crotaplujtus col- laris), and possibly the plateau striped whip- tail {Cnemidoplwrns velox) are represented here by "edge" populations at the periphery of their respective ranges. Other species, such as the northern leopard frog {Rana pipiens), east- ern fence lizard {Sceloporus undnlatiis). Great Plains ratsnake {Elephe guttata), and the Utah milk snake {Lampropeltis triangidum), have their westernmost limits in this region (Steb- bins 1985, unpublished BYU museum records). While none of these species is federally listed as threatened or endangered, a few are so listed by the state (UDWR 1992). Moreover, geographically peripheral populations such as these are particularly important as dynamic foci of evolutionaiy change (e.g.. Brown 1995, Lesica and Allendorf 1995). The Book Cliffs-Tavaputs region also sup- ports a rich mammalian fiiuna. Although our knowledge is far from complete, the area con- tains at least 62 native species, including a rel- atively stable population of black bear {Ursus americanus; H. Black personal communication). Recent fieldwork has resulted in records for 6 species previously unreported from the region (D. Rogers personal communication); these include Merriam's shrew {Sorex merriami), dwarf shrew (S. nanus), water shrew (S. palus- ths), big fi-ee-tailed bat {Nijctinomops macrotis), northern flying squirrel {Glaucomys sabrimis), and western jumping mouse {Zapiis princeps). Of these species, S. merriami, S. nanus, and N. macrotis appear to be rare throughout their known distributions. More fieldwork is likely to pioduce additional records for this region. Isolated Desert Mountain Ranges The isolated mountain ranges in Utah's Great Basin and Colorado Deserts are extremely important biologically because of their role in maintaining critical ecological and evolution- ary processes. Because of their broad eleva- tional gradients, extending from high peaks to desert valley floors, these ranges support a wider variety of habitats and a greater diver- sity of species than do areas of comparable 106 Great Basin Naturalist [Volume 56 size but less elevational relief. This eharacter- istic also enables them to support the seasonal migrations of animals ranging from large ungu- lates to small passerine birds. Furthermore, these mountain ranges have outstanding sei- entific value because they represent cool and mesic habitat islands in an otherwise warm, arid landscape. Their natural communities have developed through intermittent periods of extreme isolation (Grayson 1993). Coupled with the great geological diversity of the region, this isolation has led to the formation of unique plant assemblages, often including rare local endemics (Albee et al. 1988, Welsh et al. 1993). By illustrating how populations and communities of habitat islands are modified through colonization and extinction, these mountain ranges have played a major role in the development of theories of geographical ecology and biogeography (Brown 1971, 1995, Grayson 1993, E. Rickart in preparation). Portions of several isolated mountain ranges are represented within PWAs on BLM lands (UWC 1990). Such ranges include the Henry Mountains of the Colorado Plateau and the Deep Creek, Fish Springs, House, and New- foundland ranges of Utah's west deserts (Fig. 1). As the most isolated range in Utah, the Newfoundland Mountains in Box Elder County are especially distinctive. At 2129 m above sea level, Desert Peak and a considerable area of surrounding uplands would have existed as an island throughout the histoiy of ancient Lake Bonneville. Currently, the range forms a 154 + km- island of arid to semiarid vegetation immersed in a salt playa sea. No doubt salt marshes have covered the present salt flats periodically as the lake has advanced or receded in response to glacial and interglacial climates. The range has therefore been an eco- logical island throughout nearly 2 million years of Pleistocene and Quateniaiy time. Given such long isolation, these mountains have much to teach scientists about the persistence, local extinction, vagility, and evolutionaiy dynamics of a variety of animal and plant species that either live there now or have lived there in the past. In Utah and elsewhere in tlie inteniioun- tain region, knowledge of these topics will be important in the future as land managers tiy to anticipate plant and animal responses to the increasing fragmentation and isolation of nat- ural habitats within the human-dominated landscape (Brown 1995). Mojave Desert in Southwestern Utah Washington County includes Utah's only representative of the Mojave Desert, a warm desert commonly recognized by biogeogra- phers as lying between the Great Basin Desert to the north and the Sonoran Desert to the south (Shreve 1942, Jaeger 1957, Rowlands et al. 1982, MacMahon 1986). The Mojave Desert is physically part of the Basin and Range Geo- logical Province, but it is characterized by rel- atively low elevation over most of its area (600 to 1500 m above sea level) and by both limited precipitation (100-275 mm annually in most places) and warm summers (35°-40°C mean maxima for July; see MacMahon 1986). The uniqueness of the physical environment of the Mojave is reflected in its biota. Characteristic plants include the Joshua tree {Yucca brevifolia), creosote bush {Larrea tridentoto), white bur- sage {Ambrosia diimosa), brittle bush {Encelia farinosa), and several species of saltbush {Atri- plex). Of these, the Joshua tree can be consid- ered endemic, and if the distribution of this species is used to define the boundaries of the Mojave Desert, then the desert covers a sub- stantial portion of southeastern California, the southern cone of Nevada, the northwestern and west central parts of Arizona, and the ex- treme southwestern corner of Utah. Judicious designation of new wilderness areas in this corner of the state could help to safe- guard the many components of Utah's biologi- cal diversity that are endemic to the Mojave Desert and the associated Virgin Mountains of northwestern Arizona and adjacent Nevada. Figure 4 details land ownership in this region of Washington County. Because so much of this land is already in the public domain, there is opportunity' for biodiversity conservation with minimal disruption of economic activity. Protected areas include Zion National Park, a sul)stantial wilderness in the Pine Valley Mountains of the Dixie National Forest (no. 1 in Fig. 4), the Upper Virgin River Desert W'ild- life Management Area (or DWMA, a reserve for the desert tortoise, Gophcrus a :=:3 -2-. i a - ^ a, a; 1; _£ .2 U a _2 Z ^ OJ . o > O -73 N 0 5 CQ o ^ OJ <; -f= ^ ^ Q in '•^ ^ u- u Hi a 5^ > ;5 o 2 ii " 5. ^ 1^ n , t*- Oj E a ^ v: (; w. *; O c ji: >. 2^ p ^ , _j t£ t; .5J (M ^ ~ — ' •£ p « < "3 — 1 ^iS S ,^ u -r > .^ — , - 5 C -t- = -S Ji t = >_/ _c i- """ c ^ - C -J ~ ir. SI _ ^' >- X - ■- St c< i\ »* _^-c — Ph 1^ ^ . "" r^ ir f^ 1 r- ■^ b ci r^ '.fch ^r? « O > ^^ rf^ s 2 ^— O "t- -— y? ^ E^z ^ 1; 1^ (J 5 ^ _aj aj ^ "5 EC sr hr j_ > .^ .^ Uh ^ ij .*- 108 Great Basin Naturalist [Volume 56 too high in elevation and/or too far to the northeast to include many Mojave Desert species. The Upper Virgin River DWMA will protect lower elevation communities and will include some Mojave Desert taxa. However, many Mojave Desert species in Utah do not extend northeast of the Beaver Dam Moun- tains, and existing protected areas on the Beaver Dam slope are relatively small and iso- lated from each other (Fig. 4). By virtue of both size and location, 2 PWAs, the Beaver Dam Wash and Joshua Tree units (nos. 3a and 3b, respectively, in Fig. 4; see UWC 1990), could make important contributions to biodiversity consei-vation in Utah. Together these 2 units cover a range of elevations, include several distinctive plant communities not represented in the Upper Virgin River DWMA, and are close enough to one another and to the exist- ing protected areas to serve as stepping stones for animal movement. We illustrate the conservation value of these 2 PWAs dirough an example. The heipeto- fauna of the Mojave Desert includes 3 anu- rans, 1 tortoise, 16 lizards, 18 snakes, and about 28 additional species whose distributions are peripheral but extend into this desert along one of its edges (Stewart 1994). The portion of this fauna ranging into Utah includes 2 anu- rans, the turtle, and 13 squamates (5 lizards and 8 snakes). Their distributions across exist- ing or proposed protected areas are summa- rized in Table 3. Of this total, the relict leop- ard frog {Rana onca) apparently is extinct in Utah (Platz 1984, Jennings and Hayes 1994) and therefore absent from all existing and pro- posed protected areas in Washington County The other anuran confined to this part of Utah is the southwestern toad {Bnfo microscaphus). It is known to exist with certainty in several areas and is likely widespread throughout the region where appropriate acjuatic habitats exist (Table 3). The desert tortoise {Gophcnis agassizii) has been studied extensively over the past decade and intermittently for a much longer period of time (Woodbury and Hardy 1948, Bur\' and Germano 1994, Grover and DeRilco 1995). While Utah populations have apparently de- clined in the Beaver Dam slope area, they persist at high densities north of St. George (data summarized in Bury and Germano 1994) and are now protected in the Virgin River DWMA. Protection of the proposed Joshua Tree and Beaver Dam Wash wilderness areas would thus provide an economical way to aug- ment consei-vation of tortoise populations con- fined to the south-facing slopes of the Beaver Dam Mountains. Of the 13 squamate reptiles listed in Table 3, nine are confined to either the Mojave habitats proper (sites 3a, 3b, 4, and 5 in Fig. 4) or to tliese sites plus the Upper Virgin River DWMA (sites 2a and 2b in Fig. 4). Four species have more extensixe distributions because they are also recorded from Zion National Park. Among the 9 squamates with restricted distributions, the lizards Helodenna suspecfum and Xantusia vigilis and the snakes Crotalus cerastes and Leptotyphlops humilis may occur at all 5 Mojave sites, although this needs to be confirmed through additional fieldwork. Xantusia vigilis also occurs further east in isolated populations in Garfield and San Juan counties, and previ- ous molecular studies by Bezy and Sites (1987) show deep genetic divisions among many iso- lated populations. Many of these isolates would qualify as full species, following the criteria of Davis and Nixon (1992), but the specific status of the isolated Utah populations remains un- known. The lizard Callisaurus draconoides occurs with certainty in the upper Virgin River DWMA (in Snow Canyon State Park), Beaver Dam Wash PWA, and Lytic Ranch Preserve (sites 2a, 3a, and 5 in Fig. 4). The iguana {Dip- sosaiiriis dorsalis) is known confidenth- from onh' the lower Beaver Dam Wash PWA, although it may occur at low densities in the other 3 Mojave sites. Among the snakes, Cro- talus scutulatus is confined to the 4 strict Mojave Desert areas, and C. mitchellii is known with certaint)' from onh' the higher elevation Moja\'e sites (3b and 4, although the other 2 locations are possible). Based on a new snake record for Utah, Phyllorhynchus decurtatus is known from a specimen (BYU 45605) taken on 11 July 1995, ca 1.5 mi N of the Utah-Arizona border along the Beaver Dam slope road. Based on this record, the species likeK occurs in the Beaxer Dam Wash and Joshua Tree areas (3a and 3b), w hich are similar in \egetative struc- ture to the collecting site, and possibly at the other Mojaxe Desert sites as well. Regardless of exact distributions, all 9 squamate species with the most restricted distributions would benefit by wilderness designation of the pro- posed Beaver Dam Wash and Joshua Tree units {IJW'C 1990); and for 7 species (C. draconoides. 1996] Wilderness Selection for Biodiversity 109 Table 3. Distribution of amphibians and reptiles restricted to southwestern Utah, relative to existing protected areas and Beaver Dam Wash and Joshua Tree units of proposed BLM wilderness iucluded in H.R. 1500. The areas numbered are shown in Figure 4''. The proposed Red Mountain and Cottonwood Canyon wilderness areas (UWC 1990) are not illustrated because thc\- are lariielx (Red Mountain) or entireK' (Cottonwood ('anyon) contained within the Upper Virgin River DWMA. Bea\'cr Dam Dixie N.E Upper Virgin Wash Joshua Tree Bea\'cr Dam L>tle Zion National Wilderness RiN'cr DWMA Wildcnicss Wilderness Wilderness Ranch Taxon Park (1) (2A. 2B) (3A) (3B) (4) (5) An LIRA Rami onca — — — — — — — Biifo microscaplnis + -(?) -(?) + -(?) -(?) + Testudines Gopheriis a^assizii — — + + + + + Squamata Callisaunis draconoides — — + + V 9 + Coleomjx variegatits + -(?) + + + (?) + (?) + (?) Dipsosourus dorsalis — — — + ? ? ? Helodenna sitspectinn — — + + ? •? 4- Xantusia vigilis — — + (?) + + +■(?) 4- Crotahis cerastes — — + + + + (?) 4- Crotalus mitchellii — — — ? + + ? Crotahis scutiihitus — — — + + + 4- Leptotijphlups hiimiUs — — + + ? ? +(?) Masticophis flagellwn + -(?) + + + + 4- PhijUorhijnchiis decurtatus — — — + (?) + (?) + (?) + (?) Sonora semiannulata + -(?) + (?) + + (?) + (?) + (?) Trimorphodnn hisciitatits + — C^) + + + + + •'Distributions were inferred from localih' records available in research collections of California Academy of Sciences; M. L. Bean Lite Sciences Museum, Brigham Young University, Prove. Utah; Museum of Vertebrate Zoology, University of California, Berkeley; Utah .Museum of Natural History, University of Utah, Salt Lake City. Species listed as present ( + ) if they (1) e.xist as museum voucher specimens, (2) have been documented photographicalK' but not collected because of threatened or endangered status, or (3) have been collected near a protected area and are known to occupy the appropriate habitat. For example, Stewart (1994) summarized distributions of all Mojave Desert amphibians and reptiles on the basis of their occurrence in distinct habitat t>pes, and we used these data as an indication of the likely presence of a species in an area if not actually documented. Doubts about any occurrences are indicated b\' (?). D. dorsalis, the 3 species of Crotahis, L. species (Furlow and AiTnijo-Prewitt 1995, Lesica Jmmilis, and P. decurtatus), diese 2 PWAs would and Allendorf 1995, Lomolino and Channell constitute the largest blocks of protected area 1995). Designation of the Beaver Dam Wash in the Utah portions of their distributions. and Joshua Tree PWAs as wilderness would The biological significance of the Mojave provide an extremely economical, proactive Desert region could be illustrated with com- conservation strategy for many species, parable examples involving native birds, small mammals, and vascular plants; literally scores IMPACT OF ROADS ON PLANT of species are restricted to the low-elevation and Animal Communities Joshua tree habitats on the southwestern slopes of the Beaver Dam Mountains (see Behle et al. By definition under the 1964 Wilderness 1985, Albee et al. 1988, and Zeveloff 1988 for Act, wilderness areas must be large (at least recent species compilations). Although most 5000 acres) and roadless. Because even some areon tlie periphery of tlieir ranges, it is increas- remote and pristine areas contain primitive ingly apparent that such peripheral popula- roads or tracks, roadlessness is often an issue tions are critical to maintaining genetic diver- in debates over wilderness designation. Envi- sity and to ensuring the long-term survival of ronmentalists tend to argue that the existence 110 Great Basin Naturalist [Volume 56 of minor roads or dirt tracks is not contradic- toiy to wilderness, but that no new roads should be built. Wilderness opponents respond that any road, no matter how primitive, disqualifies PWAs for wilderness status. Decision makers may be pressured to make exceptions to allow new roads and water development within wilderness boundaries. Here, we review the objective evidence bearing on the importance of roadlessness from a purely biological per- spective. We deal with the effects of roads on animals and plants independently. Effects of Roads on Animals Roads affect wildlife in many ways, both direct and indirect. Among the more com- monly reported adverse impacts of roads on animal populations are road mortalities, animal avoidance of roads, isolation of populations by roads acting as barriers to animal movement, reductions in natural habitats, increased poach- ing, and elevated erosion leading to siltation of aquatic habitats. On Utah BLM lands, large mammals such as bighorn sheep [Ovis cana- densis), black bear, and river otter are gener- ally intolerant of human disturbance and activ- ities. These and other mammals are known also to avoid habitat adjacent to roads (Oxle>' et al. 1974, Rost and Bailey 1979, Mader 1984, Witmer and Calesta 1985, Van Dyke et al. 1986) and can therefore be displaced by the presence of roads. Historically, humans in western North America have also persecuted a number of contemporaiy or former occupants of BLM lands; such species include Golden and Bald Eagles, gray wolf, and grizzly bear (Bortolotti 1984, Mech 1995). In Utah, the in- cidence of poaching is considerably higher in regions adjacent to roads than in roadless areas (W. Woody, UDWR, personal communication). The negative effects of roads on wildlife can generally be ameliorated by closing the roads to traffic. Road mortality and the advance of habitat alteration along roads should stop en- tirely, and poaching should be sharply cur- tailed. For larger animals, roads would likely cease to act as barriers to animal movement and gene flow. However, this might not be true for some smaller species, whose moxements are more restricted generally. Significant ero- sion and siltation of aquatic habitats might be reduced only slightly. Siltation can be an impor- tant consideration, for example, on the Aquar- ius Plateau, where reductions (by as much as 1/2) in the depths of some naturally shallow lakes have already increased winter fish kills. Finally, if efforts were made to reintroduce some of the large mammals considered above, these efforts might be greatly facilitated by the protection of large blocks of roadless lands that experience minimal human intrusion. In siunmar)', if tra\'el on minor roads and tracks were to be permanently restricted, most but not all of the negative effects on wildlife would likely be ameliorated. Similar reasoning would suggest that the effects of any new un- paved minor roads or tracks might be minimal if the roads were used briefly and sporadically, e.g., to cany communications equipment. Effects of Roads on Plant Communities The most compelling argument for large roadless areas is probal)l> the protection of plant communities from disturbances that can even- tually transform whole ecosystems. Through both direct and indirect effects, roads tend to disiTipt nati\e communities of both microphytes and macrophytes. Increased off-road vehicle traffic in roaded areas directh' harms ciyptobi- otic soil crusts, which play a key role in main- taining healthy ecosystems in semiarid and arid lands, and kills or injures plants and per- haps soil-nesting insects like bees and wasps. Indirect effects include the introduction of nonnative pest plants, which have gradually replaced many native species and drastically altered features of certain habitats. The eco- system-wide effects of these exotics are well illustrated b>' Asian tamarisk {Tainarix cliincn- sis), which has channelized rivers and streams throughout the Colorado drainage and thereby altered the characteristics (flow regimes, tem- peratures, and sediment loads) of both aquatic and riparian habitats to flie detiiment of num- erous native fishes, insects, birds, mammals, and plants (Loope et al. 1988, Sudbrock 1993). Below, we elaborate on the direct and indirect effects of roads on plant communities and on the maintenance of both biodiversity and nat- ural networks of interactions in Utah's native ecosystems. Threats to cryptobiotic soils. — Across Utah's arid rangelands, a collection of cyano- bacteria, algao, lichens, and mosses form micro- phytic or cr> ptobiotic crusts on soil surfaces. In pristine plant communities tliese cnists often account for at least as much soil surface cover as do \ ascular plants. The cnptoph>'tes provide 1996] Wilderness Selection for Biodiversity 111 a number of \'akiable ecosystem services (re- viewed in Harper and Marl^Ie 1988, West 1990, and Johansen 1993), including stabiliza- tion of soils against wind and water erosion, enhancement of water retention and infiltra- tion (Brotherson and Rushforth 1983, Harpei" and St. Clair 1985, Haiper and Marble 1988), and nitrogen fixation by autotrophic bacteria, including both free-living and symbiotic cyano- bacteria (e.g., Snyder and WuUstein 1973, West and Skujins 1977, Klubek and Skujins 1980, Terry and Burns 1987). Their contribution to the nitrogen economy of these arid ecosystems is substantive. In southern Utah grasslands and cold deserts dominated by pinyon pine and juniper, nitrogen fixation by crusts is demon- strably the dominant source of nitrogen for vascular plants (Evans and Ehleringer 1993). The greater soil moisture and fertility associ- ated with biotic crusts have been shown to result in higher tissue nutrient levels (Belnap and Harper 1995 and references therein), higher seedling sui^vivorship in associated vas- cular plants (St. Clair et al. 1984, Harper and St. Clair 1985, Belnap 1994), and greater (a) floristic diversity (Kleiner and Harper 1972). Herbivores and other consumers may benefit indirectly from the enhanced nutrient status of these ecosystems (Haiper and Pendleton 1993, Belnap and Haiper 1995). Growing recognition of the importance of ciyptobiotic crusts to ecosystem processes has led to concern about the impact of disturbance by recreational users and nonnative grazers on such surfaces (Anderson et al. 1982, Johansen et al. 1984, Terry and Burns 1987, Cole 1991, Evans and Ehleringer 1993, Belnap et al. 1994, Belnap 1995). On most semiarid Utah lands, a single pass of an off-road vehicle will reduce nitrogen fixation by cyanobacteria and increase wind and water erosion of surface soils (Williams et al. 1995). Estimates of time to full recoveiy of disturbed biotic ciTists (includ- ing niti-ogen-fixing capacity) range up to 50 years in the Great Basin or 100 years on the Colorado Plateau (J. Belnap personal communication). The full biological and economic conse- quences of disturbing biotic crusts remain to be quantified. However, in semiarid ecosys- tems where plant productivity is limited by availability of water and nitrogen, even small reductions in these resources can be expected to diminish primary productivity to the detri- ment of both the producers themselves and the many consumers depending directly or in- directly on thes(> pioducers for food. Haiper and Pendleton (1993) have suggested that destruc- tion of soil crusts, and associated changes in forage quality, may be related to a decline in the health of desert tortoise populations in southwestern Utah (Grover and DeFalco 1995). If that suggestion is supported by empirical evidence in the Riture, then destniction of cmsts may account in part for the ~$1() million cost (to date, T. Esque personal communication) of the Desert Tortoise Recovery Program. Roads as corridors for invasions of introduced species. — Possibly the greatest adverse impact of roads on biological commu- nities in Utah is the aggravation of invasions of aggressive weeds along road corridors, where disturbance from road construction has elimi- nated native competitors. These introduced plants now form the dominant cover on many arid and semiarid landscapes in western North America and are widespread in Utah (Mack 1981, Morrow and Stahlman 1984, Young et al. 1987, papers in McArthur et al. 1990 and Monsen and Kitchen 1994). Habitat degrada- tion by nonnative, congregating grazers un- doubtedly aided the initial spread of brome grasses (genus Bromus) and other European or Asian annuals into native habitats, including grasslands previously dominated by caespitose or tussock grasses (Young and Evans 1971, Loope 1976, Mack 1981, 1989, Billings 1990, 1994). Brome grasses (red brome [B. riibens], Japanese brome [B. japoniciis], downy brome [B. mollis], ripgut brome [B. diandrus], and especially cheat grass [B. tectonim]) have gready increased fire frequenc)' (from an average of 60-110 yr to <5 yr in sagebrush steppe), as well as altered the pattern and dynamics of fires (e.g., Whisenant 1990). Invaded lands suf- fer declining productivity (Stewart and Young 1939) and watershed damage (Buckhouse 1985) and become drastically depleted in both native plant species and cnptobiotic soil crusts (Young and Evans 1978, Whisenant 1990, Billings 1990, 1994, Rosentreter 1994; Fig. 5). Treatments to restore these lands often involve introductions of still other exotics (e.g., Agropyron cristatum, Kochia prostrata; see contributions to McAi-thur et al. 1990 and Monsen and Kitchen 1994). The influx of invading weedy annuals has profound effects on genetic, species, and eco- system diversity, although such effects remain poorly documented. In some parts of Utah, 112 Great Basin Naturalist [Volume 56 Harner Quadrats (1.0 YdO 124 Quadrats Total species No. of natives No. aliens/quadrat Fig. 5. Relationship of both total species richness, and numbers of native species per quadrat, to the number of individuals of introduced species per quadrat; plotted from raw data in Harner and Harper (1973). Data are from sagebrush-grasslands on private and BLM foothill lands in Salt Lake, Davis, and Tooele counties. brome grasses form virtual monocultures, en- tirely replacing native communities, especially in wet years (e.g., Pellant and Hall 1994, and authors' observations). In other western states brome grass invasions threaten state or feder- ally listed plant species (Rosentreter 1994, California Native Plant Society, personal com- munication). Effects of habitat conversion radiate upward through the food chain, and adverse effects have been documented on pronghorn {Antilocapra americana) and deer (Pellant 1990, Roberts 1994), small vertebrate prey of eagles and other raptors (Kochert and Pellant 1986, Nydegger and Smith 1986), native birds (Dobler 1994), and insects (Fielding and Brusven 1994). As summarized by Billings (1994), exotic annual grasses could constitute a genuine threat to the existence of large inte- grated ecosystems that have existed since the Pleistocene in the relatively arid lands between the Rocky Moimtains and Sierra Nevada. These operational ecosystems could disappear o\ er large areas of thousands of square kilometers. A very high priority for future ecological work in Utah will be to determine the extent to which the remote BLM lands being consid- ered for wilderness status might serve as ref- uges for native flora and fauna. Seeds of brome grass, dispersed by animal vectors, certainly travel over long distances and into wilderness areas. However, lanre roadless areas with low circumference-to-area ratios might protect arid and semiarid western ecosystems against whole- sale habitat conversion. Exotic weeds tend to invade native plant communities mainly along roadsides, railroad right-of-ways, and other highly disturbed sites (Forcella and Harvey 1983, Hunter 1990, literature cited in Billings 1990 and 1994; see also Bergelson et al. 1993). Favorably wet drainage ditches provide inroads to new habitat, and invaders spread outward from the ditches during particularly wet years. vUthough systematic suneys of nonnatives do not presently exist for PWAs (and are sorely needed), there is evidence that invasions of exotic weeds may be prevented by restricting access on existing roads. Thus, of the replicate roadsides studied by Hunter (1990), introduced species (including not only brome grasses but Erodium cicutarium, Salsohi spp., and Sisijm- J)riiiin altissiinuin) dominated all but the one that had been closed to traffic and left undis- turbed for many years prior to censusing. The effects of roads on plant communities appear to differ importantly from those on ani- mal communities. Construction of new roads, especially those with drainage ditches, may hasten long-term and permanent changes to local floras, and these changes may eventually have markedly adverse effects on whole eco- systems. Existing dirt tracks are probably less threatening to plant communities; although moisture conditions on the tracks may be as favorable here as in drainage ditches, soil com- paction appears to retard growth of most plants. Given the costliness of aggressive fire sup- pression (e.g., Vail 1994) and habitat restoration measures (see I'eports in McArthur et al. 1990 and Monsen and Kitchen 1994), the most eco- nomical strategy for prexenting the spread of introduced grasses to areas that are still rela- tively pristine ma>' be to maintain their road- less character This also would proxide oppor- tunities for investigating the effects of roads (or lack thereof) on the advance of exotic plants on arid lands in Utah. Conclusions Wilderness serves man\ purposes, and its designation inxoK es man\' and \aried consid- erations. The technical issues and evidence presented here demonstrate that BLM w ilder- ness lands can play a major and perhaps pre- dominant role in safeguarding genetic, species, and ecosx'stem dixersitx' across much of arid 1996] Wilderness Selection fok Biodiversi-h- 113 Utah. Over the lont:; term, large, eontii^iuous networks of wilderness and other protected lands can provide sanctuaiy for populations of animals with large area requirements, and can help maintain natural processes and interac- tions that sustain healthy biotic communities. In many situations, wilderness designation can pro\'ide low-cost protection for rare and en- dangered species. BLM lands in geographi- cally diverse regions of Utah all offer unique ecological, scientific, and educational values. To an extent so far unmeasured, wilderness lands may protect native ecosystems from wholesale transformation by invasions of exotic species. Clearly, if biological considerations are taken into account in wilderness decisions, wilderness can play a critical role in the long- term presei^vation of Utah's biological heritage. Acknowledgments Jayne Belnap, Phyllis Coley, Donald Duff Sharon Emerson, Donald Feener, Bruce How- lett, Thomas Kursar, and Samuel Rushforth contributed to an earlier, nontechnical version of this paper. In addition to these individuals and the authors, Patricia Berger, Lynn Bohs, Rex Gates, Steven Clark, Susan Fairbanks, Jer- ran Flinders, Sarah George, James Harris, Richard Hildreth, Carl Marti, Brian Maurer, Norman Negus, Duke Rogers, Jon Seger, John Sperry, Richard Tolman, Delbert Wiens, Michael Windham, and Samuel Zeveloff en- dorsed the earlier paper. Klancy de Nevers and Jon Seger rescued corrupted computer files. T. Griswold, F Parker, and V. Tepedino allowed us to use their unpublished data (Table 2 and section on bees). J. Belnap and L. Shultz commented on a previous draft of the manu- script, and Garla Garrison helped us obtain BLM maps. Literature Cited Albee, B. J., L. M. Schultz, and S. Goodrich. 1988. Atlas of the vascular plants of Utah. Occasional Pub- lications No. 7, Utah Museum of Natural History, University of Utah, Salt Lake City. 670 pp. Anderson, D. C., K. T. Harper, and S. R. Rushforth. 1982. Recoveiy of ci"\i)toganiic soil crusts from graz- ing on Utah winter ranges. Journal of Range Man- agement .35: 35.5-3.59. Armbruster, E, and R. L\nde. 1993. A population viability analysis for Mrican elephant {Loxodonta africana): How big should reserves be? Conservation Biology 7: 602-610. Aiwoon, !).. J. lloMAND, R. Bolander, B. Kiunkun, D. E. House, L. Armstkonc, K. Thorne, and L. England. 1991. Utah threatened, endangered, and .sensitive plant field guide. A.Xl.EROD, D. I. i960. The evolution of flowering plants. Pages 227-305 in S. Tax, editor. Evolution after Dar- win. Volume 1, The evolution ol" life. University of Chicago Press, Chicago. BABHrrr, B. 1995. Science: opening tlie iic^xt cluiptcr of consei-vation history. Science 267: 1954-19.56. Behle, W. H., E. D. Sorensen, and C. M. White. 1985. Utah birds: a revised checklist. Occasional Publica- tions No. 4, Utah Mu,seum of Natural Histon', Uni- versity of Utah, Salt Lake City Behnke, R. J. 1992. Native trout of western North Amer- ica. American Fisheries Society Monograph 6. 275 pp. Behnke R. J., and M. Zarn. 1976. Biology and manage- ment of threatened and endangered western trouts. USDA Forest Sei-vice, Technical Report RM-GTR- 28. Fort Collins, CO. Belnap, J. 1994. Potential value of cyanobacterial inocula- tion in revegetation efforts. Pages 179-185 in S. B. Monsen and S. G. Kitchen, editors, Proceedings — Ecology and Management of Annual Rangelands. Report INT-GTR-313, Ogden, UT . 1995. Soil surface disturbances; their role in accelerating desertification. Environmental Moni- toring and Assessment 37: 1-17. Belnap, J., and K. T. Harper. 1995. The influence of ciyp- tobiotic soil crusts on elemental content of tissue in two desert seed plants. Arid Soil Research and Rehabilitation 9: 107-115. Belnap, J., K. T Harper, and S. D. Warren. 1994. Sur- face disturbance of ciyptobiotic soil crusts: nitroge- nase activity, chlorophyll content, and chlorophyll degradation. Arid Soil Research and Reliabilitation 8: 1-8. Belovsky, G. E. 1987. Extinction models and mammalian persistence. Pages 35-57 in M. E. Soule, editor, Viable populations for consei"vation. Cambridge Uni- versity Press, Cambridge, UK. Belovsky, G. E., J. A. Bissonette, R. D. Dueser, T. C. Edwards, Jr., C. M. Luecke, M. E. Ritchie, J. B. Slade, and E H. W.-^gner. 1994. Management of small populations: concepts affecting the recovery of endangered species. Wildlife Societ\' Bulletin 22: 307-316. Bergelson, J., J. A. Newman, and E. M. Floresrou.x. 1993. Rates of weed spread in spatially heteroge- neous environments. Ecology 74: 999-1011. Bezy, R. L., and J. W. Sites, Jr. 1987. A preliminaiy stud\ of allozyme evolution in the lizard family Xantusi- idae (Reptilia, Sauria). Heipetologica 43: 280-292. Billings, W. D. 1990. Bromus tectoruni, a biotic cause of ecosystem impoverishment in the Great Basin. Pages 301-322 in G. M. Woodell, editor. The earth in transition: patterns and processes of biotic impover- ishment. Cambridge University Press, Cambridge, UK. . 1994. Ecological impacts of cheatgrass and resul- tant fire on ecosystems in the western Great Basin. Pages 22-30 in S. B. Monsen and S. G. Kitclien, editors, Proceedings — Ecology and Management of Annual Rangelands, USDA Forest Service, Technical Report INT-GTR-313. Ogden, UT 114 Great Basin Natupialist [Volume 56 BoLGER, D. T, A. C. Alberts, and M. E. Soule. 1991. Occurrence patterns of bird species in habitat frag- ments: sampling, extinction, and nested species sub- sets. American Naturalist 1.37: 1.55-166. BORTOLOTTI, G. R. 1984. Trap and poison mortalit>' of Golden and Bald Eagles. Journal of Wildlife Man- agement 48: 117.3-1179. Brotherson, J. D., AND S. R. Rlshforth. 1983. Influ- ence of cr}'ptogamic ciiists on moisture relationships of soils in Navajo National Monument, Arizona. Great Basin Natiu'alist 43: 73-78. Brown, J. H. 1971. Mammals on mountaintops: nonequi- librium insular biogeographv'. American Naturalist 105: 467-478. . 1995. Macroecolog>'. Universit\ of Cliicago Press, Chicago. Brown, J. H., and E. J. Heske. 1990. Control of a desert grassland transition by a keystone rodent guild. Sci- ence 250: 1705-1707. Buckhouse, J. C. 1985. Effects of fire on rangeland water- sheds. Pages 58-60 in K. Sanders and J. Durham, edi- tors. Proceedings of a Symposium: Rangeland Fire Effects. U.S. Department of the Interior, Bureau of Land Management, Boise, ID. BiRY. R. B., .\SD D. J. Germano, editors. 1994. Biology- of North American tortoises. USDI Fish and Wild- life Research Report ISSN 1040-2411, No. 13. 204 pp. C.-^se, T. J., and M. L. Cody. 1988. Testing theories of island biogeography. American Scientist 7.5: 402—411. Christensen, N. L., et al. 1989. Inteqireting the Yellow- stone fires of 1988. Bioscience 39: 678-685. Cole, D. N. 1991. Trampling disturbance and recovery of cryptogamic soil crusts in Grand Canyon National Park. Great Basin Naturalist 50: 321-325. Cronouist, a., a. H. Holmgren, N. H. Holmgren, and J. L. Reveal. 1972. Intermountain flora, \blume 1. Hafiier Publishers, New York. Cross, S. P 1985. Responses of small manmials to forest perturbations. Pages 269-275 in Riparian ecosys- tems and their management: reconciling conflicting uses. USDA Forest Sewice, Technical Report RM- GTR-120. Fort Collins, CO. Davidson, D. W, R. S. Inouye, and J. H. Brown. 1984. Granivory in a desert ecosystem: experimental evi- dence for indirect facilitation of ants by rodents. Ecology' 65: 1780-1786. D.A\ is, J. I., AND K. C. NI.XON. 1992. Populations, genetic variation, and the delimitation of phylogenetic species. Systematic Biolog\- 41: 421—435. De.\CON, J. E., AND W. L. MiNCKLEY. 1974. Desert fishes. Pages 385-488 in G. W Brown, editor Desert biol- ogy. Volume II. Academic Press, New York. DEPARTMENT OF THE INTERIOR (BUREAU OF L.\ND MaN./VGE- MENt). 1994. Ecosystem management in the BLM: from concept to commitment. U.S. Government Printing Office (573-183), VV^ishington, DC. 16 pp. Diamond, J. M. 1984. "Normal" extinctions of isolated populations. Pages 191-246 in M. H. Nitecki, cditoi; Extinctions. University of Chicago Press, Chicago. Dobler, E C. 1994. Washington state shrub-steppe eco- system studies with emphasis on the relationship between nongame birds and shrub and grass coxcr densities. Pages 149-161 in S. B. Monsen and S. C;. Kitchen, editors, Procecding.s — Ecology and Man- agement of Annual Rangelands, USDA Forest Ser- vice, Technical Report INT-GTR-313. Ogden, VT. Durr.\nt, S. D. 1950. Mammals of Utah: taxonomy and distribution. Museum of Natural Histoiy, University of Kansas Publications 6. Edmunds, G. F 1994. Tail water insects. The Landing Net (Sinnmer): 6-7. . 1995. Tail water insects; re\isited. The Landing Net (Fall): 4, 11. Evans, R. D., and J. R. Ehleringer. 1993. A break in the nitrogen cycle in aridlands? E\idence from S'^N of soils. Oecologia 94: 314-317. Fielding, D. J., and M. A. Brusven. 1994. Grasshopper community responses to shrub loss, annual grass- lands, and crested wheatgrass seedings: manage- ment implications. Pages 162-166 //) S. B. Monsen and S. G. Kitchen, editors. Proceedings — Ecology and Management of Annual Rangelands, USDA For- est Senice, Technical Report INT-GTR-313. Ogden, UT Forcella, F, and S. J. Har\ey. 1983. Eurasian weed infestation in western Montana in relation to vegeta- tion and disturbance. Madroiio 30: 102-109. Frankel, O. H., and M. E. Soule. 1981. Consen ation and e\oIution. Cambridge Universit\' Press, Cam- bridge, UK. FuRLOW, F B., AND T. Armijo-Prewitt. 1995. Peripheral populations and range collapse. Conservation Biol- ogy 9: 1.345. Gr.\YSON, D. K. 1993. The desert's past: a natural prehis- ton' of the Great Basin. Smithsonian Institution Press, Washington, DC. Grover, M. C, and L. a. DeFalco. 1995. Desert tortoise {Gophenis agassizii): status-of-knowledge outline with references. USDA Forest Service. Technical Report INT-GTR-316. Ogden, UT 134 pp. Grumbine, R. E. 1994. What is ecosystem management':' Consei-vation Biologv' 8: 27-38. Harner, R. F, and K. T. Harper. 1973. Mineral composi- tion of grassland species of the eastern Great Basin in relation to stand producti\it\'. Canadian Journal of Botany 51: 2037-2046. Harper, K. T. 1986. Historical environments. Pages 51-63 ()! W. L. D Azevedo, editor. Handbook of North American Indians, \blume 11: Great Basin. Smith- sonian Institution, Washington, DC. Harper, K. T, and J. R. M.\rble. 1988. A role for non\ as- cular plants in management of arid and semiarid rangelands. Pages 13.5-169 i)i P. T. Tneller, editor. Vegetation science applications for rangeland analy- sis and management. Khiwer .\cademic Publisher, Dordrecht. Harper, K. T, and R. L. Pendleton. 1993. Cyanobacte- ria and cyanolichens: Can they enhance a\ailabilit\' of essential minerals for higher plants? Great Basin Naturalist 53: 59-72. Harper, K. T, and L. L. St Cl.\ir. 1985. Ci-yptogamic soil cnists on arid and semiarid rangelands in Utah: effects on seedling establishment and soil stability. Final Report Buieau of Land Management, Utah State Office, Salt Lake Cit\. Hahkis, L. D. 1984. The fragmented forest: island bio- geography theory and the preser\'ation of biotic di\ersity. l'ni\'ersit\' of Chicago Press, Chicago. Heanky. 1-. H. 19S4. Mammalian species richness on islands on the Simda Shelf Southeast Asia. Oecolo- gia (Berlin) 61: 11-17. H()Li:)F\. P B., R. Stone, \\'. White, G. Somermlle, D. Di KK \\. Gkhvais, and S. Gloss. 1974. Threatened 1996] Wll.nKKNKSS SkI.KC'I'ION I'OH B|()I)I\ khsht 115 fishes of I'tali. Procet'dintis ol tlu' I'tali Academy ol Science, Arts, and Ix-tters 51: 4(i-(i5. Ik BBS, C. L. 1940. Speciation of I'ishes. Anicrican Natu- ralist 74: 198-211. , 1941. The relation of hydrological conditions to speciation in fishes. Pages 182-195 in J. (J. Need- liani, editor, .\ S\nip()siinn on Hydrol)iolog\. Uni\er- sit\ ol \\'isconsin Press, Madison. Ill \II:k, H. 1990. Recent increases in Hroiiius on [\\r i\e\ada Test Site. Pages 22-25 in E. D. McArtlun; E. M. Ronine>, S. D. Smith, R T. Tueller, editors. Proceedings — S\inposiuni on Cheatgrass Invasion, Shrub Die-Off, and Other Aspects of Shrub Riology and Management, USDA Forest Ser\ice, Technical Report INT-GTR-276. Ogden, UT lUCN. 1978. Categories, objectives and criteria lor pro- tected areas. Merges, Switzerland. J.\EGER, E. C. 1957. The North American deserts. Stan- ford University Press, Palo Alto, CA. 308 pp. Jennings, M. R., and M. R Hayes. 1994. Decline of native ranid frogs in the desert Southwest. Pages 183-211 in R R. Rrovvn and J. W. Wright, editors, Heipetol- ogy of the North American deserts. Southwestern Herpetological Society Special Publication No. 5. Van Nuys, CA. 311 pp. JOHANSEN, J. R. 1993. Ciyptogamic crusts ot semiarid and arid lands of North America. Journal of Ph\cology 29; 140-147. JOHANSEN, J. R., L. L. St. Claih, R. L. Webb, and G. T. Nebeker. 1984. Recovery patterns of cryptogamic soil crusts in desert rangelands following fire distur- bance. Bryologist 87: 238-243. Johnson, R. R., L. T Haight, and J. M. Simpson. 1977. Endangered species vs. endangered habitats: a con- cept. Pages 68-79 in Importance, presei'vation and management of riparian habitat. USDA Forest Ser- vice, Technical Report RM-GTR-43. Fort Collins, CO. Ji lander, O. 1962. Range management in relation to mule deer habitat and herd productivity in Utah. Journal of Range Management 15: 278-281. K.\v, C. E. 1990. Yellowstone s northern elk herd: a critical evaluation of the "natural regulation paradigm. Unpublished doctoral dissertation, Utah State Uni- versity, Logan. Kershner, J. L. 1995. Ronneville cutthroat trout. Pages 28-35 in M. K. Young, editor, Consenation assess- ment for inland cutthroat trout. USDA Forest Ser- vice, Technical Report RM-GTR-256. Fort Collins, CO. Kleiner, E. F, .\nd K. T Harper. 1972. Environment and community organization in grasslands of Canyon- lands National Park. Ecolog\- 53: 299-309. Klubek, B., and J. Ski JINS. 1980. Heterotrophic nitrogen fi.xation in arid soils crusts. Soil Biology and Bio- chemistiy 12: 229-236. Knopf, F L. 1985. Significance of riparian vegetation to breeding birds across an altitudinal cline. Pages 105-111 in Riparian ecosystems and their manage- ment: reconciling conflicting uses. USDA Forest Ser- vice, Technical Report RM-GTR-120. Fort Collins, CO. Kochert, M. N., and M. Pellant. 1986. Multiple use in the Snake River Birds of Prev Area. Rangelands 8: 217-220. Kushlan, J. A. 1979. Design and management of conti- nental wildlife resei-ves: lessons from the Everglades. Biological Conservation 15: 281-290. Lesiga, P, and E W. Ai.LENDORE 1995. WIk-h are periph- eral populations valuable for conservation? Conser- \ation Biology 9: 753-760. LoMOLiNO, M. v., and R. CllANNEl.L. 1995. Splendid iso- lation: patterns of the geographic range colla])se in endangered mannnals. |onrna! of .Mainmalogv 76: 335-347. LooPE, L. L.. P C;. Sanghe/., P W. Tarr, VV. L. L(X)PE, AND R. L. Anderson. 1988. Biological invasions of arid land nature reserves. Biological Conservation 44:95-118. LooPE, W. L. 1976. Relationships of \egetation to envi- ronment, Canyonlands National Park, Utah. Unpub- lished doctoral dissertation, Utah State Uni\t'rsit\, Logan. NfvGk, R. N. 1981. The invasion oi' Bronnis tectorum L. into western North America: an ecological chronicle. Agro-Ecosystems 7: 145-165 . 1989. Temperate grasslands vulnerable to plant invasions: characteristics and conseciucnccs. Pages 155-179 ()i J. A. Drake, H. A. Mooncy, F di Castri, R. H. Groves, F J. Kruger, M. Rejmanek, and M. Williamson, editors. Biological invasions. John Wiley and Sons, New York. MacKinnon, J., K. MacKinnon, G. Child, and J. Tiior- SELL. 1986. Managing protected areas in the Tropics. lUCN, Gland, Switzerland. MacMahon, J. A. 1986. Deserts. Audubon Society Nature Guides. Alfred Knopf New York. 638 pp. Mader, H. J. 1984. Animal habitat isolation by roads and agricultural fields. Biological Conservation 29: 81-96. McArthur, E. D., E. M. Romney, S. D. Smith, R T TuELLER, editors. 1990. Proceedings — Symposium on Cheatgrass Invasion, Shrub Die-Off and Other Aspects of Shrub Biology and Management. USDA Forest Service, Technical Report INT-GTR-276. Ogden, UT Mech, L. D. 1995. The challenge and opportunitv- of recovering wolf populations. Conservation Biologv 9: 270-274. Meffe, G. K., and C. R. Carroll. 1994. Principles of con- sei^vation biology. Sinauer, Sunderland, fvIA. 600 pp. MiCHENER, C. D. 1979. Biogeography of the bees. Annals of the Missouri Botanical Garden 66: 277-347. Miller, R. R. 1959. Origin and affinities of the freshwater fish faima of western North America. Pages 187-222 in C. L. Ilubbs, editor. Zoogeographv'. .\j\AS Publi- cation 51. . 1961. Man and the changing fish fauna of the American Southwest. Papers, Michigan .'\cademy of Science, Arts, and Letters 46: 365-404. MiNCKLEY, W. L., AND J. E. Deacon. 1968. Southwestern fishes and the enigma of "endangered species. Sci- ence 159: 1424-1432. MiNCKLEY, W. L. AND J. E. DEACON, EDITORS. 1990. BatUes against extinction: native fish management in the American West. University of Arizona Press, Tucson. MiNCKLEY, W L., D. A. Hendrickson, and C. E. Bond. 1986. Geography of western North American fresh- water fishes: description and relationships to intra- continental tectonism. Pages 519-613 in C. H. Hocutt and E. O. Wiley, editors. The zoogeography of North American freshwater fishes. John Wiley & Sons, New York. MoNSEN, S. B., and S. G. Kitchen, editors. 1994. Pro- ceedings— Ecology and Management of Annual 116 Great Basin Naturalist [Volume 56 Rangelands. USDA Forest Service, Technical Report INT-GTR-313. Ogden, UT. Morrow, L. A., and R W. Stahlman. 1984. The histor\' and distribution of downy brome {Bromiis tectonim) in North America. Weed Science 32 (Supplement 1): 2-7. MoYLE, R B., AND J. E. Williams. 1990. Biodiversity loss in the temperate zone: decline of the native fish fauna of California. Conservation Biology 4: 275-2S4. National Research Council. 1978. Conservation of germplasm resources: an imperative. National Acad- emy of Sciences, Washington, DC. Neff, J. L., AND B. B. Simpson. 1993. Bees, pollination systems and plant diversity. Pages 143-167 /)! J. LaSalle and I. E. Gauld, editors, Hymenoptera and biodiversity. C.A.B. International, Wallingford, UK. Newmark, W D. 1985. Legal and biotic boundaries of western North American national parks: a problem of congruence. Biological Consewation 33: 197-208. . 1987. A land-bridge island perspective on mam- malian extinctions in western North American parks. Nature 325: 430-432. . 1991. Tropical forest fragmentation and the local e.xtinction of understor\' birds in the Eastern Usam- bara Mountains, Tanzania. Conservation Biologv 5: 67-78. . 1995. Extinction of mammal populations in west- ern North American national parks. Conservation Biology 9: 512-526. Noss, R. E 1992. The wildlands project: land consei^vation strategy. Pages 10-25 in The wildlands project. Wild Earth Special Issue, Cenozoic Society. Nydegger, N. C, and G. W. Smith. 1986.' Pages 152-156 in D. E. McArthur and B. L. Welch, editors. Pro- ceedings— Symposium on the Biology o{ Artemisia and Chnjsothainnus, USDA Forest Service, Techni- cal Report INT-GTR-200. Ogden, UT. OxLEY, D. J., M. B. Fenton, and G. R. C.\r.\iodv. 1974. The effects of roads on populations of small mam- mals. Journal of Applied Ecology 11: 51-59. Parker, E D. 1985. A candidate legume pollinator Osinia sanrafaclae Parker journal of Apiculture Research 24: 132-136. . 1986. Field studies with Osmia sanrafaelae Parker, a pollinator of alfalfa (Hymenoptera: Megachilidae). Journal of Economic Entomology 79: 384—386. Patterson, B. D. 1984. Mammalian extinction and bio- geography in the southern Rocky Mountains. Pages 247-293 in M. H. Nitecki, editor. Extinctions. Uni- versity of Chicago Press, Chicago. Pellant, M. 1990. The cheatgrass-wildfire cycle — are diere any solutions? Pages 11-18 in E. D. McArthur, E. M. Romney, S. D. Smith, P T. Tueller, editors. Proceedings — Symposium on Cheatgrass Invasion, Shrub Die-Off and Other Aspects of Shrub Biology and Management, USDA Forest Service, Technical Report INT-GTR-276. Ogden, UT. Pellant, M., and C. Hall. 1994. Distribution of two exotic grasses on intermountain rangelands. Pages 109-112 in S. B. Monsen and S. G. Kitchen, editors. Proceedings — Ecology and Management of Annual Rangelands, USDA Forest Senice, Technical Report INT-GTR-313. Ogden, UT. Pickett, S. T. A., and J. N. Thompson. 1978. Patch dynamics and the design of nature reserves. Biologi- cal Consei^vation 13: 27-37. Pimm, S. L., H. L. Jones, and J. Diamond. 1988. On the risk of extinction. American Naturalist 132: 757-785. Pl.\tz, J. 1984. Status report for Rana onca Cope. Pre- pared for Office of Endangered Species, U.S. Fish and Wildlife Sei^vice, Albuquerque, NM. 27 pp. Primac:k, R. B. 1993. Essentials of conservation biology Sinauer, Sunderland, MA. 561 pp. FIaven, P R. 1992. The nature and value of biodiversity. Pages 1-18 in Global biodiversity strategy: guide- lines for action to save, study, and use earth s biotic wealth sustainably and equitablv. WRI, IL'CN, UNEP Reid, W v., and K. R. Miller. 1989. Keeping options alive: the scientific basis of consenang biodiversity. World Resources Institute, Washington, DC. RiCKART, E. A. In preparation. Elevational diversity gradi- ents, biogeography, and the structure of montane mammal communities in the intermountain region. Journal of Biogeography: In preparation. Roberts, T. C, Jr. 1994. Resource impacts of cheatgrass and wildfires on public lands and livestock grazing. Pages 167-169 in S. B. Monsen and S. G. Kitchen, editors. Proceedings — Ecology and Management of Annual Rangelands, USDA Forest Service, Technical Report INT-GTR-313. Ogden, UT. Rosentreter, R. 1994. Displacement of rare plants by e.xotic grasses. Pages 170-175 in S. B. Monsen and S. G. Kitchen, editors. Proceedings — Ecology and Management of Annual Rangelands, USDA Forest Service, Technical Report INT-GTR-313. Ogden, UT. RosT, G. R., and J. A. Bailey. 1979. Distribution of mule deer and elk in relation to roads. Journal of Wildlife Management 43: 634-641. Rowlands, P, H. Johnson, E. Ritter, and A. Endo. 1982. The Mojave Desert. Pages 103-162 in G. L. Bender, editor. Reference handbook on the deserts of North America. Greenwood Press, Westwood, CT. 594 pp. Salwasser, H., C. Schonewald-Co.\, and R. Baker. 1987. The role of interagency cooperation in manag- ing viable populations. Pages 159-173 in M. E. Soule, editor, Viable populations for consen'ation. Cambridge University' Press, Cambridge, UK. Samson, D. A., T. E. Philippi, and D. W. Da\ idson. 1992. Granivon' and competition as determinants of annual plant diversity in the Chihuahuan Desert. Oikos 65: 61-80. Saunders, D. A., R. J. Hobbs, and C. R. M.\rgules. 1991. Biological consequences of ecosystem fragmenta- tion: a review. Consenation Biolog\ 5: 18-32. Schonewald-Co.x, C. M. 1983. Guidelines to manage- ment: a beginning attempt. Pages 414—445 in C. M. Schonewald, S. M. Chambers, B. MacBnde, and, L. Thomas, editors. Genetics and consenation. Benja- min Cummings, Menio Park, CA. Shaffer, M. L. 1981. Minimum population size for species conservation. Bioscience 31: 131-134 Shiozavva, D. K., and R. P Evans. 1994. The use of DNA to identify geographical isolation in trout stocks. Pages 125-131 in R. Barnhart, B. Shake, and R. H. Hamre, editors. Wild trout \. wild trout in the 21st century. SiiREVE, E 1942. The desert vegetation of North America. Botanical Reviews 8: 195-246. ShuLTZ, L. M. 1993. Patterns of endemisni in the Utah flora. Pages 249-263 in R. Si\inski and K. Lightfoot, 1996] Wilderness Selection for Biodiversiit 117 editors. Southwestern rare and endangered plants. New Mexico Department of ForestiT and Resources Conservation Division, Miscellaneous Publication No. 2. Santa Fe, NM. Shultz, L. M., E. E. Neely, and J. S. Tuhy. 1987. Flora of the Orange Cliffs of Utah. Great Basin Naturalist 47: 287-298.' Smith, G. R. 1966. Distrilnition and evolution of the North American catostomid fishes of the subgenus Pantosteus, genus Catostoiitiis. Miscellaneous Publi- cations of the Museum of Zoology, University of Michigan, No. 129. . 1978. Biogeography of intermountain fishes. Inter- mountain Biogeography: a Symposium. Great Basin Naturalist Memoirs 2: 17-42. Snyder, J. M., ,^nd L. H. VVlllstein. 1973. The role of desert cryptogams in nitrogen fixation. American Midland Naturalist 90: 257-265. SoULE, M. E., EDITOR. 1987. Viable populations for con- sen'ation. Cambridge LIniversitv Press, Cambridge, UK. ' SouLE, M. E., D. T. Bolger, A. C. Alberts, R. Sauvajot, J. Wright, M. Sorice, and S. Hill. 1988. Recon- structed dynamics of rapid extinctions of chaparral- requiring birds on urban habitat islands. Consen'a- tion Biolog)' 2: 75-92. SouLE, M. E., and B. a. Wilco.x. 1980. Conservation biology: an evolutionaiy-ecological perspective. Sin- auer, Sunderland, MA. St. Clair, L. L., B. L. Webb, J. R. Johansen, and G. T. Nebeker. 1984. Cryptogamic soil crusts: enhance- ment of seedling establishment in disturbed and undisturbed areas. Reclamation and Revegetation Research 3: 129-136. Stebbins, G. L., Jr. 1952. Aridity as a stimulus to plant evolution. American Naturalist 86: 33-44. Stebbins, R. C. 1985. A field guide to western reptiles and amphibians. Houghton Mifflin Co., Boston, MA. Stevens, C. C. 1992. The elevational gradient in altitudi- nal range: an extension of Rapoport's latitudinal rule to altitude. American Naturalist 140: 893-911. Stewart, G. R. 1994. An overview of the Mohave Desert and its herpetofauna. Pages 55-69 in P. R. Brown and J. W Wright, editors, Herpetology of the North American deserts. Southwestern Heipetological Soci- ety Special Publication No. 5. Van Nuys, CA. 311 pp. Stewart, G., and A. E. Young. 1939. The hazard of basing permanent grazing capacity on Bromus tectonim. American Society of Agronomy Journal 31: 1002-1015. Sudbrock, a. 1993. Tlimarisk control. I. Fighting back: an overview of the invasion, and a low-impact way of fighting it. Restoration and Management Notes 11: 31-34. Tepedino, V. J. 1981. The pollination efficiency of the squash bee {Pepomipis pniinosa) and the honeybee {Apis mellifera) on summer scjuash {Citcurbifa pepo). Journal of the Kansas Entomological Societv 54: 359-377. Terborgh, J., AND B. Winter. 1980. Some causes of extinction. Pages 119-134 ;'/! M. E. Soule and B. A. Wilcox, editors, Conservation biology. Sinauer, Sun- derland, MA. Terry, R. E., and S. J. Burns. 1987. Nitrogen fi.xation in cryptogamic soil crusts as affected by disturbance. Pages 369-372 in Proceedings — Pinyon-Juniper Conference, USDA Forest Service, Technical Report INT-GTR-215. Ogden, UT Tyus, H. M., AM) C. A. K\HH 1989. Habitat use and streamtlow needs ol rare and endangered fishes, Yampa River, Colorado. U.S. Fish and Wildlife Ser- vice Biological Report 89(14). 27 pp. U.S. Fish and Wildlife Servige. 1993. Endangered and threatened wildlife and plants, 50 CFR 17.11 and 17.12, August 23, 1993. Division of Endangered Species, U.S. Fish and Wildlife Semce, Washington, DC. Utah Division of Wildlife Resourges. 1992. Native Utah wildlife species of special concern. Draft docu- ment. Utah Division of Wildlife Resources, Salt Lake City. 31 pp. Utah Wilderness Coalition (UWC). 1990. Wilderness at the edge. Peregrine Smith Books, Layton, UT. Vail, D. 1994. Symposium introduction: management of semiarid rangelands — impacts of annual weeds on resource values. Pages 3-4 in S. B. Monsen and S. G. Kitchen, editors, Proceedings — Ecology and Man- agement of Annual Rangelands, USDA Forest Ser- vice, Technical Report INT-GTR-313. Ogden, UT Van Dyke, E G., R. H. Brogke, H. G. Shaw, B. B. Agker- man, T. E Hemker, and E G. Lindzey. 1986. Reac- tions of mountain lions to logging and human activ- ity Journal of Wildlife Management 50: 95-102. Wagner, E H., R. Foresta, R. B. Gill, D. R. MgCul- LOUGH, M. R. Pelton, W. E Porter, and gonsulta- TION ON LAW and poligy BY J. L. S.\x. 1995. Wildlife policies in the U.S. National Parks. Island Press. 242 pp. Warren, M. L., and B. M. Burr. 1994. Status of fieshwater fishes of the United States: ovendew of an imperiled fauna. Fisheries 19: 6-18. Welsh, S. L. 1978. Endangered and dueatened plants of Utali, a reevaluation. Great Basin Naturalist 38: 1-18. Welsh, S. L., N. D. Atwood, S. Goodrigh, and L. C. Higgins. 1993. A Utah flora. 2nd edition. Print Ser- vices, Brigham Young University, Provo, UT. Welsh, S. L., N. D. Atwood, and J. L. Reveal. 1975. Endangered, threatened, extinct, endemic and rare or restricted LJtah vascular plants. Great Basin Natu- ralist 35: 327-326. West, N. E. 1990. Stiiicture and hmction of microphytic soil crusts in wildland ecosystems of arid to semiarid regions. Advances in Ecological Research 20: 179-223. West, N. E., and J. Skujins. 1977. The nitrogen cycle in North American cold-winter semi-desert ecosys- tems. Oecologia Plantarum 12: 45—53. Whisenant, S. G. 1990. Changing fire frequencies on Idaho's Snake River Plains: ecological and manage- ment implications. Pages 4-10 in E. D. McArthur, E. M. Romney, S. D. Smith, P T. Tueller, editors. Proceedings — Symposium on Cheatgrass Invasion, Shrub Die-Off, and Other Aspects of Shrub Biology and Management, USDA Forest Service, Technical Report INT-GTR-276. Ogden, UT Whittaker, R. H. 1972. Evolution and measurement of species diversity. Ta.xon 21: 213-251. Wilgove, D. S., C. H. MgLellan, and A. E Dobson. 1986. Habitat fragmentation in the temperate zone. Pages 237-256 in M. E. Soule, editor, Consei-vation biology: the science of scarcity and diversity. Sinauer Associates, Sunderland, MA. WiLGO.X, B. A., AND D. D. Murphy. 1985. Conservation strategy: the effects of fragmentation on extinction. American Naturalist 125: 879-887. 118 Great Basin Naturalist [Volume 56 Williams, J. D., J. E Dobrowolski, N. E. West, and D. A. Gillette. 1995. Microphytic cnist influence on wind erosion. Transactions of the American Society of Agricultural Engineers 38: 131-137. Williams, J. E. 1991. Preserves and refuges for native western fishes; histoiy and management. Pages 171-189 in W. L. Minckley and J. E. Deacon, edi- tors. Battle against extinction, native fish manage- ment in the American West. University of Arizona Press, Tucson. 517 pp. Willis, E. O. 1974. Populations and local extinctions of birds on Barro Colorado Island, Panama. Ecological Monographs 44: 153-169. Wilson, E. O. 1988. Biodiversity. National Academy Press, Washington, DC. WiTMER, G. W, AND D. S. Calesta. 1985. Effect of forest roads on habitat use by Roosevelt elk. Northwest Science 59; 122-125. WOLZ, E. R., and D. K. Shiozawa. 1995. Soft sediment benthic macroinvertebrate communities of the Green River at the Ouray National Wildlife Refuge, Uinta County, Utah. Great Basin Naturalist 55; 213-224. Woodbury, A. M., and R. Hardy. 1948. Studies on the desert tortoise, Gopherus agassizii. Ecological Mono- graphs 18; 145-200. Young, J. A., and R. A. Evans. 1971. Invasion of medusa- head into the Great Basin. Weed Science 18: 89-97. . 1978. Population dynamics after wildfires in sage- brush grasslands. Journal of Range Management 31: 283-289. Young, J. A., R. A. Evans, and B. L. K.ay 1987. Cheatgrass. Rangelands 9; 266-270. Zeveloff, S. I. 1988. Mammals of the Intemiountain West. University of Utah Press, Salt Lake Cit>. Received 27 September 1995 Accepted 21 March 1996 CrcMt Basin Naturalist 56(2), © 1996, pp. 119-12« NUTRIENT DISTRIBUTION IN QUERCUS GAMBELIl STANDS IN CENTRAL UTAH A. R. Tiedeinann' and VV. P. Claiy- Abstract. — Gambel oak {Qucrciis gambchi Nutt.) is increasingly recognized as a valuable fuelwood throughout Ari- zona, Colorado, New Mexico, and Utah. Knowledge of the distribution of nutrients among biotic and abiotic coinpo- iicnts is an important step in developing prescriptions for managing these stands for sustainable productivity. Eight Q. gamhclii stands were sampled for concentrations (%) and accumulations (kg ha"l) of total nitrogen (N), phosphorus (P), sulfur (S), calcium (Ca), magnesium (Mg), potassiimi (K), and sodiimi (Na) among aboveground and helowground biomass components and the upper 30 cm of soil. Highest concentrations of N, R and S occuned in oak Itaxes, underston' leaves, and the forest floor layer. Generally, highest concentrations of Ca, Mg, K, and Na occurred in the soil. The greatest proportion of the total capital of individual nutrients was contained in the soil (82%-99%). Above- ground components of li\e biomass, standing and down-dead, and forest floor contained 10%, 14%, and 8%, respec- tixely, of total capitals of N, V, and S. The forest floor had the largest accumulation (63%) of total nutrients (N, R S, Ca, Mg, K, and Na) of live and dead aboveground components. Nutrient accumulation in live biomass was heavily weighted to the belowground component. The dense system of roots, rhizomes, and lignotubers comprising 56%' of total biomass contained 62% of the total accumulation of nutrients in live biomass. Low levels of total P in the soil and accumulation of 14% of the ecosystem total of P in aboveground biomass compo- nents suggest the need for a better understanding of the role of P in productivity of these stands in development of pre- scriptions for management of residues after harvest. Key words: nutrient cycling, soil nutrients, nitrogen, phosphorus, sulfur cations, Quercus gambelii, Utah. Gambel oak {Quercus ^^amhelii Nutt.) is found as a small shruh or large tree on about 3.8 million ha in Colorado, Arizona, New Mexico, and Utah. It is a clonal species that sprouts readily after harvest or other distur- bance from a dense belowground system of lignotubers and rhizomes (Tiedemann et al. 1987). The lignotubers are similar to those found on Eucalyptus (Carrodus and Blake 1970). Rhizomes (belowground stems) are also common in oaks (Muller 1951). With increasing demands for fuelwood throughout its range, Q. gambelii is coming under close scrutiny for its initial value as a fuelwood source and for continued fuelwood production potential (Wagstaff 1984, Claiy and Tiedemann 1992). The density of the wood, its superior heat-yielding qualities compared with softwoods (Barger and Ffolliott 1972), and its sprouting nature (Tiedemann et al. 1987) make this species ideal for fuelwood management. In the development of management strate- gies for sustainable productivity' of Q. gambelii, an important step is to determine the manner in which nutrients are distributed among the abiotic and biotic components of the system. This information will help develop manage- ment guidelines so that harvest activities do not deplete nutrients to the extent that future site productivity may be jeopardized. Our objectives were to determine the con- centrations and total amounts of major plant nutrients — nitrogen (N), phosphorus (P), sul- fur (S), calcium (Ca), potassium (K), magne- sium (Mg), and sodium (Na) — in live and dead Q. gambelii biomass components and in soil, understory, and forest floor of a representative portion of the Q. gambelii ecosystem in central Utah; and to relate findings to similar studies in other hardwood stands. This study was a companion to a study of biomass distrilnition (Clan' and Tiedemann 1986). Study Areas and Methods Eight Q. gambelii stands (plots) were selected near Ephraim in central Utah. The stands were on slopes with gradients from 5% to 40%. Soils 'Pacific Northwest Research Station, 1401 Gekeler, La Grande, OR 97850. 2lnterinountain Research Station, 316 East Mvrtle Street, Boise, ID 83702. 119 120 Great Basin Naturalist [Volume 56 are Typic Calcixerolls formed on alluvium and colluvium derived from limestone, sandstone, and shale (Swenson et al. 1981). Soils are cob- bly loams in the surface 50 cm and very stony clay loams in the substratum to depths of 150 cm. Elevations of the 8 stands range from 2089 to 2480 m. Average annual precipitation ranges from 36 to 51 cm, and the annual frost- free period is 90 to 110 d (Swenson et al. 1981). Plot sizes varied in approximately inverse proportion to tree stem density (Clary and Tiedemann 1986). We attempted to obtain a sample of the range of stand densities and stem heights. A 3 X 3-m plot was used for the densest stand (34,444 stems/ha), a 10 X 10-m plot for the least dense stand (5000 stems/ha). Mean ages of stems ranged from 37 to 109 yr (Claiy and Tiedemann 1986). At each plot, all live stems were counted and numbered, and 5 were selected at random for measurement of height, diameter, biomass, and nutrient concentration. Sample stems were cut about 4 cm above the ground, parti- tioned into 60-cm sections, and weighed in the field. Live and dead branches and leaves were removed. A 10-cm portion of each bole section was placed in a plastic bag, sealed, and returned to the laboratoiy for determination of moisture content and nutrient concentrations. Live branches, dead branches, and leaves from each tree were bagged, retuined to the labora- toiy, and oven-dried at 70 °C to constant weight. After weighing, a sample was taken from each component for analysis of nutrient concentra- tion. Standing dead trees were counted on each plot, and 5 were randomly selected to be cut and weighed in the field. A section was taken from each including any attached branches for determination of moisture and nutrient concentrations. Understoiy biomass — including Q. gainbc- lii < 1 m, other shrubs, herbaceous plants, for- est floor and down and dead oak — was sam- pled on three 1-m- subplots randomly located within each plot, except plot 8, where only 1 subplot was sampled. Plot 8 was sampled at a different time from plots 1-7, with the main objective of excavation to determine charac- teristics of the underground system (Tiede- mann et al. 1987). We inadvertently collected only 1 subplot for determination of understoiy biomass, forest floor, down and dead oak, and soil. On all subplots, forest floor was collected to mineral soil. No separation into litter (L), fermentation (F), and humus (H) layers was made. Hence, the forest floor includes plant detritus accumulated above mineral soil in- cluding down and dead oak <0.5 cm. All sam- ples were oven-dried at 70 °C and weighed to determine mass per unit area (kg ha~^) of the forest floor. Weight of down and dead oak >0.5 cm was assigned to the categoiy of down and dead oak trees. A small sample of each compo- nent from each l-m^ plot was used for nutri- ent analysis. Forest floor samples contained some soil as a result of wind deposition and the fact that sampling results in collection of a small amount of soil from the forest floor/soil interface. Therefore, weights of forest floor samples were adjusted for content of soil by determining weight loss on combustion of small samples in a muffle furnace at 900 °C. Combustion of organic materials results in a small amount of mineral ash residue of 5 g per 100 g of forest floor (Tiedemann 1987b). We adjusted forest floor weights by this amount. Soil volume weight (bulk density) was determined by collecting a 15- to 20-cm-diam- eter sample to a depth of 30 cm at each of the subplots after vegetation was harvested and the forest floor sampled. This was the maxi- mum depth feasible to collect without using mechanized digging apparatus because of the increased rocks, cobbles, roots, and rhizomes at greater depths. The soil hole was lined with plastic and the xolume determined by measur- ing the quantit)' of water to the nearest 10 niL required to fill the hole. Soil was oven-dried at 70 °C, weighed, and retained for nutrient analysis. This method of bulk density determi- nation compares favorably with the paraffin clod technique (Howard and Singer 1981). One plot (plot 8) was hydraulicalK' exca- xated to a depth of 1 m by use of a hydraulic pump capable of supplying 114 L/min (Tiede- mann et al. 1987). All roots, rhizomes, and lig- notubers were removed and transported to the laboratoiy for drying, dissecting, weighing, and nutrient analysis. Weight of roots at depths > 1 m was estimated from taper-weight relationships established for the first 1 m of vertical roots. A composite sample of the roots (<1.0 cm, 1.0-2.5 cm, and >2.5 cm) and rhi- zomes was taken for nutrient analysis. The proportion of each component in the sample was weighted on the basis of its proportion of total weight. 1996] Nutrient Distribition in Quercus gambelii 121 Eacli 10-cni bole portion was separated into 8 equal radial segments. One of these from each portion \\'as further separated into heartwood, sapwood, and bark. Samples from each radial segment were then composited for each tree prior to analysis. All vegetation sam- ples were ground to 0.25-mm fineness in preparation for analysis of nutrient concentra- tion. Soil samples were sieved through a 2- nnn mesh screen and ground to 0.125-mm fineness prior to analysis. All samples were analyzed for total N by Kjeldalil digestion followed by titrimetric deter- mination of distilled ammonium (Bremner 1965); for total P by sulfuric acid-selenium digestion (Parkinson and Allen 1975) followed b> molybdenum blue detemiination of P (Olsen and Dean 1965); for total S by the procedure of Tiedemann and Anderson (1971); and for total cations Ca, Mg, Na, and K by atomic absolution spectroscopy (Jones and Isaac 1969) on the sulfuric acid-selenium digest used for total P Mass per unit area (kg ha"^) of individual plot values for each individual biomass com- ponent of trees (leaves, live branches, standing dead, etc.) from the study of Claiy and Tiede- mann (1986) were used to convert concentra- tions of individual nutrients to mass per unit area (kg ha'^). In the biomass determination (Clary and Tiedemann 1986), stems were not partitioned into bark, heartwood, and sapwood. We determined the percentage by weight of these 3 components for each bole and con- verted weights to kg \ra~^ for each plot using values from Clary and Tiedemann (1986). These values were then multiplied by concentrations of individual nutrients for determination of mass per unit area (kg ha~^) content of nutri- ents. Mass per unit area (kg ha"^) values for understory vegetation, down-dead oak, and the forest floor were multiplied by concentra- tion values for individual nutrients to deter- mine mass per unit area of each nutrient. Bulk density of the upper 30 cm of soil (minus par- ticles >2 mm) was used to develop mass per unit area (kg har^) values for soil so we could convert nutrient concentration values to mass of indi\'idual nutrients per hectare. Mass per unit area values of Quercus roots, rhizomes, and lignotubers in the upper 1 m of the exca- vated plot plus the extrapolation of larger (>2.5 cm) vertical roots to their extinction point was used to convert concentration values of nutrients to a kg ha"^ basis. Extrapolation was based on application of taper-weight rela- tionships for each root. For purposes of data presentation, nutrient contents (kg ha~l) of individual aboveground biomass components were grouped into three categories: (1) aboveground live overstory and understory vegetation; (2) standing and down- dead that includes standing dead trees, dead branches on live trees, down and dead trees, and dead branches on the ground >0.5 cm; and (3) the forest floor that includes all plant detritus above mineral soil except for Quercus branches >0.5 cm. Analysis of variance in a randomized com- plete block design with the 8 individual plots as blocks was used to determine differences in concentration among aboveground biomass components for each nutrient constituent (Steel and Torrie 1960). Biomass component was the main effect term in the analysis. Val- ues for the 5 individual trees and for the 3 for- est floor and understory subplots in each of the 8 plots (blocks) were pooled, and the means were used in the analysis of variance. Statistical comparison with underground biomass com- ponents was not possible because this was de- termined on only 1 plot. Where the F-test was significant, differences among individual bio- mass components were determined using the LSD test (Carmer and Swanson 1971). Signifi- cant differences are expressed at P < 0.01. No statistical tests were applied to kg har^ nutri- ent content data because individual compo- nents were summed to provide more inclusive groupings. For example, live aboveground bio- mass includes oak leaves, live branches, heart- wood, sapwood, bark, and understory leaves and stems. Results and Discussion Nutrient Concentrations There were no significant differences in con- centrations of nutrients in biomass (F < 0.01) among plots (blocks) for any nutrient con- stituent except Ca. Differences among bio- mass components were highly significant for eveiy nutrient constituent. Nitrogen concentrations in the forest floor and in Quercus leaves were significantly higher than in any other component (Table 1). Under- story leaves were significantly lower in N con- centration than the forest floor or Quercus 122 Great Basin Naturalist [Volume 56 T.ABLE I. Concentration (percent) of nutrient constituents in biotic and abiotic components of Qiiercus gambelii Nutrient Leaves Live branches Heartwood N LSD 0.01 1.57 = 0.08 0.56 0.15 P LSD 0.01 0.21 - 0.024 0.03 0.003 S LSD 0.01 0.08 = 0.014 0.03 0.03 Ca LSD 0.01 0.91 = 0.30 0.90 0.17 Mg LSD 0.01 0.35 = 0.29 0.16 0.02 K LSD 0.01 0.68 = 0.18 0.36 0.33 Na LSD 0.01 0.04 = 0.007 0.01 0.01 Sapwood Bark Dead Standing liranches dead trees 0.27 0.02 0.02 0.17 0.04 0.15 0.002 0.62 0.02 0.04 1.55 0.20 0.32 0.01 0.55 0.02 0.04 0.98 0.14 0.26 0.01 0.35 0.01 0.04 1.00 0.08 0.21 0.01 "Comparisons anions ahoveiiiouiHl 1)Ioiikiss coiiipiiiUTits miK. leaves. We did not obsei^ve increases in N con- centration of the forest floor that usually accompany decomposition, mineralization, and leaching of other constituents from the fallen overstory leaves (Bocock 1963, Gosz et al. 1973). In a litter bag study Klemmedson (1992) measured a 60% increase in N concentration in Q. gambelii leaves in the litter layer over a 750-d time span. Differences between our observations and those of Klemmedson were probably because we report comparisons be- tween Quercus leaves and the entire forest floor, whereas his comparisons were for the lit- ter layer only Lowest concentrations of N were observed in the heartwood. Standing dead and down-dead trees were both higher in N con- centrations than were heartwood and sapwood of living stems. This probably resulted from selective decomposition and loss of other ele- ments causing an increase in the concentration of N in standing dead and down-dead trees. Concentration of N in the upper 30 cm ot soil (0.42) was greater than would be expected for this site. According to Jenny (1941), the normal range of soil N for semiarid sites is 0.10%-0.25% for the surlace 10 cm. The high content of N in these soils can prol)ably be attributed to 2 principal factors: (1) the high clay content is conducive to retention of high levels of organic N (Klenmiedson and Jenn\' 1966, Millar et al. 1966); and (2) the extraordi- nary accumulation of forest floor (37,348 kg ha~l) at this site (Claiy and Tiedemann 1986) provides a continuous supply of N to the soil through decomposition and leaching. Leaves of understory plants (0.27%), Quer- cus leaves (0.21%), and forest floor (0.12%) had highest concentrations of R Differences among these 3 components were significant. Reduced concentration of P in the forest floor compared to Quercus leaves corresponded to obsei^va- tions of Klemmedson (1992). Concentration of P in Q. gambelii leaves at the surface of the forest floor began to decrease shortly after deposit and declined steadily for 500 d to about 60% of original concentration. Concen- tration then leveled off for the remaining 250 d of the experiment. Our lowest levels of P occurred in the heartwood (0.003%). Although there were some significant diiferences among other biomass components, the acttial differ- ences were slight and probably of little biolog- ical significance. Total P in soil (0.02%) was substantially below normal levels, which are 0.09%-0.13% for soils of the United States (Parker et al. 1946). Concentrations of S were greatest in forest floor (0.12%) and understory leaves (0.11%), and there was no significant difference between these 2 components. However, S concentra- tion in both was significantly higher than in Quercus leaves. Lowest S concentrations in abo\ eground components were in the sapwood and heartwood. Our comparisons of N and S 1996] Nutrient Distribution in Quercus gamkfaai 123 (■ci)s\steni,s in cfiital Utah. Understoiy Understory Down-dead I'orfst leaves stems trees lloor Roots and rlii/.onies Lif^notuhers Soil 1.46 0.54 0.43 1.66 0.44 0.33 0.42 0.27 0.0.5 0.01 0.12 0.03 0.02 0.02 0.11 0.04 0.06 0.12 0.04 0.03 0.04 0.98 0.61 0,76 2.67 0.97 1.15 1.29 0.40 0.1" 0.11 1.15 0.14 0.09 1.92 1.14 0.64 0.0' 0.43 0.21 0.14 0.87 0.008 0.02 0.005 0.06 0.02 0.008 0.08 levels in the forest floor with Quercus leaves presented an anomaly. We would expect S comparisons between forest floor and Quercus leaves to be similar to those for N, because S is a companion to N in several amino acids (Allaway and Thompson 1966, Coleman 1966). Klemmedson's (1992) observations bear this out because both N and S concentrations in Quercus leaves increased about 60% over a 750-d period after deposition at the surface of the forest floor. However, when we compared Quercus leaves and the entire forest floor, it appeared that N and S responded differently over the long periods required for develop- ment of the forest floor. Nitrogen concentra- tion tended to remain constant and S concen- tration increased over time. Mineralization of S in deeper layers of the forest floor may pro- ceed more slowly than mineralization of N, thereby resulting in an increase in S concen- tration. Products of decomposition for N may also be more mobile than those for S. Total S concentration in soil (0.04%) was in the middle of the range reported for U.S. soils, 0.01-0.06 (Burns 1968). The ratio of N:S of 10:1 in soil indicates that the S level is great enough that N will be efficiently utilized for the formation of plant proteins (Black 1968, Burns 1968). Concentrations of the 4 measured cations, Ca, Mg, K, and Na, were generally higher in the soil than in any plant component. Excep- tions were higher concentrations of Ca in the forest floor and in the bark of Quercus trees and K in understoiy leaves. Calcium concentrations in the forest floor layer were more than 2.5 times greater than Quercus leaves. The content of Ca in bark was nearly 10 times greater than heartwood or sap- wood. Quercus leaves, live branches, dead branches, standing dead trees, and down-dead trees were all comparable in Ca concentration. Magnesium concentrations in biomass com- ponents were highest in the forest floor layer — approximately 3 times greater than in Quercus and understoiy leaves. In contrast to Ca patterns, Mg concentrations in live branches and standing dead and down-dead trees were significantly lower than in Quercus leaves. Understoiy leaves were significantly higher in K concentration (1.14%) than were Quercus leaves (0.68%) or understory stems (0.64%). Potassium concentrations were about equal for live branches, heartwood, and bark, and about half the concentration found in Quercus leaves. Concentration of K in forest floor was substan- tially lower than in Quercus leaves and may reflect the ease with which K is leached from the forest floor relative to the other cations (Attiwill 1968). Highest concentrations of Na occurred in Quercus leaves and in the forest floor Differ- ences among other biomass components were 124 Great Basin Naturalist [Volume 56 minor, even though some were statisticalK' significant. Comparisons of cation knels in Quercus leaves with levels in the forest floor were vari- able between our study and results of the lit- ter bag study of Klemmedson (1992). We showed significantly greater Ca and Mg in the forest floor than in Quercus leaves. Klemmed- son (1992) found similar increases in Ca in Quercus leaves over 750 d. However, Mg con- centration in his study declined to about 80% of the level in fresh leaves over the 750-d study. Differences in K concentration that we found between Quercus leaves and the forest floor were not nearly as great as the decline in K concentration over time in the litter layer measured by Klemmedson (1992). Potassium concentration in Quercus leaves declined about 70% in 500 d and then stabilized to the end of the 750-d study. Differences between Klemmedson's observations and ours were probably a result of the fact that he studied changes in nutrient concentration in the litter layer and our comparisons were with the entire forest floor There is little information on the concen- trations of nutrients in biomass components in western hardwood stands. There are 2 appar- ent reasons for this. Compared with the east- ern United States, the area occupied by stands of hardwood species in the West is minor Therefore, until recently, western hardwoods have not been viewed as an economically im- portant resource; rather, they were considered weed species because they were assumed to compete with marketable coniferous trees or with understory forage-producing species. With emerging demands for fuelwood and new markets for unique woods for furniture, there is increased awareness of the value of western hardwoods and, especially, Q. gainbelii (Wagstaff 1984, Claiy and Tiedemann 1992). Nutrient concentrations of leaves agreed closely with those reported by Klemmedson (1992) for Q. gambelii in northern Arizona. Bartos and Johnston (1978) determined the concentrations and proportions of indi\ idual nutrients in the various components of 3 clones o( Populus tronuloides Vlich.x. (cjuaking aspen) trees in Utah and Wyoming but did not consider the forest floor, understoiy, and down- dead components of the nutrient pool. Con- centrations of N in the various tree compo- nents of Q. grnnhclii and P. freiiiuloides were comparable except for higher concentrations of N (2.5%) in leaves of P. tremuloides; concen- trations of F, K, and Ca were similar for all tree components. Sodium concentrations were gen- erally greater in Q. gambelii than in P. tremu- loides. Concentrations of N, I^ and S in live aboveground biomass of Q. gambelii were comparable to those reported for Q. robur in Russia (Rodin and Bazilevich 1967) and in Belgium (Duvigneaud and Denaeyer-De Smet 1970). Concentrations of N in forest floor and dead branches also were comparable to values for southern and eastern U.S. Quercus stands (Lang and Forman 1978). Concentrations of cations in our study did not agree as well with those presented in the literature as for N, I^ and S. For example, Q. gambelii forest floor concentrations of K and Mg were 3 and 8 times greater than those reported for Q. robur Calcium concentrations in Q. gambelii were substantialK' greater than those observed in other studies in forest floor live branches, dead branches, standing dead trees, and down-dead trees. Distribution of Nutrient Capital Among Components Comparisons of nutrient distribution be- tween above- and belowground components must be considered from the perspective that our soil sampling was restricted to the upper 30 cm because of rock and the massixe under- ground structures of Q. gainbelii. The actual zone of rooting and nutrient acquisition was undoubtedly much greater than the area we sampled. Therefore, our estimates of the pro- portions of nutrients in aboveground compo- nents were likely to be higher than if the entire rooting zone had been sampled. Also, the kg ha~^ estimates were for the area of the actual clone sampled. Clones ofQ. gambelii do not occupy the entire area of the sites on which they occur Most studies take into account the high- and l()w-densit\ areas of tree occupancy in determining nutrient distri- bution. Therefore, in making projections to an areal basis, the actual area occupied b>' Q. gambelii clones must be considered. The greatest proportion of total nutrient capital sampled was contained in the soil (Table 2). Of the total capitals of individual nutrients, 82%-99% were contained in the soil. Aboveground accumulations of indixidual nutrients in Ii\c' biomass, standinii and down- 1996] Nutrient DisiiiiBi iion in Qiercvs gambelii 125 Tablk 2. Distiil)iiti()n of nutrients anions liiomass, foR'st lloor, standinii plus dowii-ilcad, and soil components of {). iciinhi'Iii stands. Live-' Standing Live'' abo\e- plus Total below- Total'l ground down- Forest'' above- groniid live Total' Nutrient biomass dead floor ground biomass biomass Soil-' capital Nitrogen (kg ha"') 245 140 654 1039 270 515 9500 10810 % of total abovegroiuid 24 13 63 % of total capital 10 2 88 Phosphorus (kg ha"^) 19 4 48 71 19 38 410 500 % of total alioxeground 27 5 68 % of total capital 14 4 82 Sulfur (kg ha-i) 19 13 46 78 22 41 946 1046 % of total aboveground 24 17 59 % of total capital 8 2 90 Calcium (kg ha-^) 334 303 1167 1804 924 1258 28844 31571 % of total abovegroimd 18 17 65 % of total capital 6 3 91 Magnesium (kg ha"l) 62 35 381 478 63 125 42485 43023 % of total abo\'egroimd 13 7 80 % of total capital 1 <1 99 Potassium (kg ha~l) 201 72 144 417 116 317 20268 20801 % of total aboveground 48 18 34 % of total capital 2 <1 98 Sodium (kg ha"') 7 4 20 31 7 14 1765 1804 % of total aboveground 22 13 65 % of total capital 2 <1 98 Total (kg ha-l) 887 571 2460 3918 1421 2308 % of total abovegroimd 23 14 63 9( of total in living biomass 3S 62 "Includes living aboveground overstory and understory vegetation. "Includes all forest floor layers above mineral soil. 'Includes roots, rhizomes, and lignotubers in the upper 100 cm of soil. "Standing crop plus belowground biomass. ■^Upper 30 cm of soil. ■Standing crop plus standing and down-dead plus forest floor plus belowground biomass plus soil. dead, and forest floor ranged from 31 kg ha~^ for Na to 1804 kg ha-l foj. q.^ Proportions of total capitals of N, E and S in aboveground components were highest with 10%, 14%, and 8%, respectively. The proportion of N (the most widely reported nutrient) in aboveground components (10%) was comparable to that described for other semiarid and temperate forest and woodland ecosystems (Klennnedson 1975, Brown 1977, Tiedemann 1987a). The forest floor was the most important aboveground reservoir of nutrients with 63% of the total accumulation above ground. Accu- mulations of individual nutrients in the forest floor ranged from 20 to 1167 kg ha'^ and con- stituted 34%-80% of the aboxeground capitals. 126 Great Basin Naturalist [Volume 56 Total nutrient content of the forest floor in our Q. gambelii clones (2460 kg ha"^) substantially exceeded the range described by Lang and Forman (1978) in their summary for U.S. Quercus forests (206 kg har^ [Yount 1975] to 1462 kg ha-l [Gosz et al. 1976]). Greater accu- mulation of Ca in the forest floor layer (1167 kg ha~^) compared with that reported b>' other obsei-vers (98-400 kg ha~^; Lang and Forman 1978) accounted for much of the difference in total accumulation of nutrient elements in Q. gambelii compared with other Quercus stands. Also, forest floor biomass accumulation in our Q. gambelii stands (37,348 kg ha"^; Claiy and Tiedemann 1986) was near the upper limit (46,800 kg ha"l) of that presented for U.S. Quercus forests (Lang and Forman 1978). The massive belowground system of ligno- tubers, rhizomes, and roots comprised 56% of the total biomass of Q. gambelii (Clary and Tiedemann 1986) and contained <\% to 4% of the total of the capitals of individual nutrients. However, relative to the total nutrient accu- mulation in live biomass, the live belowground component was an important storage area con- taining 37%-74% of the individual nutrient accumulations. The proportion of total nutri- ents in belowground biomass (61%) substan- tially exceeded the range for deciduous forests worldwide (30%-40%) summarized by Rodin and Bazilevich (1967). This finding supported the conclusions of Chattaway (1958), Robbins et al. (1966), and Blake and Canodus (1970) that storage of nutrients is an important fonction of belowground components such as lignotubers. Total content of nutrients in the entire organic component (total live and dead above- ground and belowground biomass) of our Q. gambelii stands (5339 kg ha~l) was in the mid- dle of the range for deciduous forests world- wide (2000-7500 kg ha-l) summarized by Rodin and Bazilevich (1967). Similarly, total nutrient content of live biomass (2308 kg ha"^) was comparable to values for oak forests in Russia (2600-3400 kg ha-l; Ro^Iji-^ .j,^j Bazile- vich 1967). Worldwide, leaves usualh' constitute 8%'-10% of the store of mineral elements in plant bio- mass (Rodin and Bazilevich 1967). Mineral element accumulation in Q. gambelii leaves and understoiy leaves (245 kg ha-l; not shown in Table 2) comprised 11% of the total mineral content of live biomass and was within the rel- atively constant, narrow range of 200-300 kg ha-l normally found in leaves reported by Rodin and Bazilevich (1967). Conclusions Gambel oak appears to be unique from other deciduous forests in the accumulation of nutrients in the forest floor and in below- ground biomass components. Both were major areas of nutrient accumulation. The leaves, in contrast, were a minor storage area. Accumulation of nutrients in aboveground living and dead components expressed as a proportion of total site nutrients was similar to that reported for other semiarid and temper- ate forest habitats. The quantity of N, the most commonly measured nutrient stored in the forest floor, also agreed well with this litera- ture. It should be noted that had we been able to sample a larger proportion of the total root- ing zone, the proportion of the total nutrient capital aboveground would likely have been smaller. Low levels of P in the upper 30 cm of soil suggest that this element may limit productiv- ity of Q. gambelii. Because of potential limita- tions in the soil, accumulation of 14% (71 kg ha-l) Qf jIjp jqj-^i ecosystem P in aboveground living and dead components, we suggest cau- tion in the wa)' the forest floor and residues are managed. Fuelwood harvest followed by removal of residues by broadcast burning could cause large losses of P, depending on degree of consumption of organic matter and fire temperatures (Covington and DeBano 1988, DeBano 1988). This loss may reach 60% (of 71 kg ha~l) if fuels are totalK' consumed (Raison et al. 1985). However, such losses need to be weighed against changes in P availability that result from burning. In his summaiy of plant- and litter-contained nutrients, DeBano (1988) indicated that fire-induced increases in P avail- abilit>' decline and reach pre-fire levels within 1 yr DeBano and Klopatek (1988) showed that inorganic P is released In prescribed burning but is quickly immobilized and ma\' not be readily availalile for plant growth. Although there are also substantial accu- mulations of N and S in aboN egroimd biomass and these are sensitive to losses from vola- tilization (Knight 1966, Tiedemann 1987b), they are not limiting in the soil and quantities are likcK sutticient to replenish losses. 1996] Nutrient Distribution in Quercus gambeui 127 Fertilizer amendment with P may warrant consideration as a means of impro\in^ {). gain- helii productivity after liarvest. This decision, however, should be based on soil tests to determine the a\ ailabilit\ of E Literature Cited Aluawav, W. H., and J. E Thompson. 1966. Sulfur in the nutrition of plants and animals. Soil Science 101: 240-247. Attiwill, P M. 1968. The loss of elements from decom- posing litter Ecology 49: 142-145. Barger, R. L., and R E Fkolliott. 1972. The physical characteristics and utilization of major woodland tree species in Arizona. USDA Forest Service Research Paper RM-83. 80 pp. Rocky Mountain For- est and Range E.xperiment Station, Fort Collins, CO. Bartos, D. L., and R. S. Johnston. 1978. Biomass and nutrient content of quaking aspen at two sites in the western United States. Forest Science 24: 273-280. Black, C. A. 1968. Soil-plant relationships. 2nd edition. John Wiley and Sons, Inc., New York. 792 pp. Blake, T. J., and B. B. Carrodus. 1970. Studies on the lignotuhers oi Eucalyptus obliqua tHeri. II. Endoge- nous inhibitor levels correlated with apical domi- nance. New Phytologist 69: 1073-1079. Bocock, K. L. 1963. Changes in the amounts of diy matter nitrogen, carbon, and energy in decomposing wood- land leaf litter in relation to the activities of soil fauna. Journal of Ecology 52: 27.3-284. Bremner, J. M. 1965. Total nitrogen. Pages 595-624 in C. A. Black, D. D. Evans, L. E. Ensminger et al., editors. Methods of soil analysis. Part 2, Chemical and microbiological properties. Agronomy 9, Ameri- can Societ\' of Agronomy, Inc., Madison, WI. Brown, T. H. 1977. Nutrient distribution of two eastern Washington forested sites. Unpublished masters the- sis, Washington State University, Pullman. Ill pp. Burns, G. R. 1968. O.xidation of sulfur in soils. Sulphur Institute Technical Bulletin 1. 41 pp. Carmer, S. C, and M. R. Swanson. 1971. Detection of differences between means: a Monte Carlo study of five pairwise multiple comparison procedures. Agronomy Journal 63: 940-945. Carrodus, D. D., and T J. Blake. 1970. Studies on the lig- notuber oi Eucalyptus obliqua EHeri. I. The nature of the lignotuber New Ph\tologist 69: 1069-1072. Chattawav; M. M. 1958. Bud development and lignotu- ber formation in eucalypts. Australian Journal of Botany 6: 103-115. Cl.\rv, W. P, and a. R. Tiedemann. 1986. Distribution of biomass within small tree and shrub form Quercus f^ainhclii. Forest Science 32: 232-242. . 1992. Ecology and values of Cambel oak wood- lands. Pages 87-95 in P E Ffolliott et al., technical coordinators. Ecology and management of oak and associated woodlands: perspectives in the south- western United States and northern Mexico, 27-30 April 1992, Sierra Vista, AZ. USDA Forest Sewice, General Technical Report RM-218. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. Coleman, R. 1966. The importance of sulfur as a plant nutrient in world crop production. Soil Science 101: 230-240. CoviNCTON, W. W, AND L. E DkBano. 1988. Effects of fire on pinyon-jimiper soils. Pages 78-86 in J. S. Krammes, technical coordinator. Effects of fire man- agement of southwestern natural resources. USDA Forest Service, General Technical Report RM-191. Rocky Moimtain Forest and Range Experiment Sta- tion, Fort Collins, CO. DeBano, L. E 1988. Effects of fire on the soil resource in Arizona chaparral. Pages 65-77 in J. S. Krammes, technical coordinator. Effects of fire management of southwestern natural resoiux'es. USDA Forest Service, General Technical Report RM-191. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. DeBano, L. E, and J. M. Klopatek. 1988. Phosphorus dynamics of pinyon-juniper soils following simulated bmning. Soil Science Society of America Journal 52: 271-277. Duvigneaud, R, and S. Denaever-De Smet. 1970. Bio- logical cycling of minerals in temperate deciduous forests. Pages 199-225 in D. E. Reichle, editor, Analysis of temperate forest ecosystems. Springer- Verlag, Berlin, Heidelberg, Germany. Gosz, J. R., G. E. Likens, and F H. Bormann. 1973. Nutrient release from decomposing leaf and branch litter in the Hubbard Brook Forest, New Hamp- shire. Ecological Monograph 43: 17.3-191. . 1976. Organic matter and nutrient dynamics of the forest and forest floor in the Hubbard Brook For- est. Oecologia (Berlin) 22: 305-320. Howard, R. E, and M. J. Singer. 1981. Measuring forest soil bulk density using irregular hole, paraffin clod, and air permeability. Forest Science 27: 316-322. Jenny, H. 1941. Factors of soil formation. McGraw-Hill Book Co, Inc., New York. 281 pp. Jones, J. G., Jr., and R. A. Isa,\c. 1969. Comparative ele- mental analysis of plant tissue by spark emission and atomic absorption spectroscopy. Agronomy Journal 61: 393-394. Klemmedson, J. O. 1975. Nitrogen and carbon regimes in an ecosystem of young dense ponderosa pine in Ari- zona. Forest Science 21: 163-168. . 1992. Decomposition and nutrient release from mixtures of Cambel oak and ponderosa pine leaf lit- ter Forest Ecology' and Management 47: 349-361. Klemmedson, J. O., and H. Jenny. 1966. Nitrogen avail- ability in California soils in relation to precipitation and parent material. Soil Science 102: 215-222. Knight, H. 1966. Loss of nitrogen from the forest floor by burning. Forestry Chronicle (June): 149-152. L\NG, G. E., AND R. T. T. FoRMAN. 1978. Detrital dynamics in a mature oak forest; Hutcheson Memorial Forest, New Jersey. Ecology 59: 580-595. Millar, C. E., L. M. Turk, and H. O. Foth. 1966. Funda- mentals of soil science. John Wiley and Sons, Inc., New York. 491pp. Muller, C. H. 1951. The significance of vegetation repro- duction in Quercus. Madrono 11: 129-137. Olsen, S. R., and L. a. Dean. 1965. Phosphorus. Pages 10.35-1049 in C. A. Black, D. D. Evans, L. E. Ens- minger et al, editors. Methods of soil analysis. Part 2, Chemical and microbiological properties. Agron- omy 9, American Societ>' of Agronomy, Inc., Madi- son, WI. 128 Great Basin Naturalist [Volume 56 Parker, E W, J. R. Adams, K. G. Clark et al. 1946. Eer- tilizers and lime in the United States: resources, pro- duction, marketing, and use. USDA Miscellaneous Publication 586. Washington, DC. 94 pp. Parkinson, J. S., and S. E. Allen. 1975. A wet oxidation procedure suitable for the determination of nitrogen and mineral elements in biological material. Com- munications in Soil Science and Plant Analysis 6; 1-11. Raison, R. J., R K. Khanna, and R V. Woods. 1985. Mechanisms of element transfer to the atmosphere during vegetation fires. Canadian Joinuial of Forest Research 15: 132-140. RoBBiNS, W W, T. E. Weier, and C. R. Stocking. 1966. Botany: an introduction to plant science. John Wiley and Sons, Inc. 614 pp. Rodin, L. E., and N. I. Bazilevich. 1967. Production and mineral cycling in teiTestrial vegetation. [English translation by G. E. Eogg, editor] Oliver and Boyd. Edinburgh-London. 288 pp. Steel, R. G. D., and J. H. Torrie. 1960. Principles and procedures of statistics. McGraw-Hill, New York. 481 pp. SwENSON, J. L., D. Beckstrand, D. T. Erickson et al. 1981. Soil sui-vey of Sanpete Valley area, Utah. USDA Soil Consei"vation Sei'vice, and U.S. Department of the Interior Bureau of Land Management, in coop- eration with Utah Agriculture E.xperiment Station and Utah State Division of Wildlife Resources. 179 pp. Tiedemann, a. R. 1987a. Nutrient accumulations in pinyon-juniper ecosystems — managing for future site productivity. Pages 352-359 in R. L. Everett, compiler. Proceedings — Pinyon Juniper Conference, Reno, NV, 13-16 Januaiy 1986. USDA Eorest Ser- vice, General Technical Report INT-215. Intermoun- tain Research Station, Ogden, L^T. . 1987b. Combustion losses of sulfur from forest foliage and litter Forest Science 33: 216-223. Tiedemann, A. R., and T. D. Anderson. 1971. Rapid analy- sis of total sulfur in soils and plant material. Plant and Soil 35: 197- 200. Tiedemann, A. R., W. P Cl^ry, and R. J. Barbour. 1987. Underground systems of Gambel oak {Qiiercus gam- belii) in central LItah. American Journal of Botany 47: 1065-1071. Wagstafe E J. 1984. Economic considerations in use and management of Gambel oak for fuelwood. USDA For- est Sendee, General Technical Report INT-165. 8 pp. Intermountain Forest and Range E.xperiment Sta- tion, Ogden, UT. YOUNT, J. D. 1975. Forest-floor nutrient d\'namics in soutliem Appalachian hardwood and white pine plan- tation ecosystems. Pages 598-608 in F. G. Howell, J. B. Gently, and M. H. Smith, editors. Mineral cycling in southeasteiTi ecosystems. Energy Research and Development Administration Symposium series. Technical Information Center Springfield, VA. Received 17 October 1995 Accepted 11 March 1996 Great Basin Naturalist 56(2), © UMi p|i. 129-134 COMPARISON OF TWO ROADSIDE SURVEY PROCEDURES FOR DWARF MISTLETOES ON THE SAWTOOTH NATIONAL FOREST IDAHO RolH'rt L. Mathiasm', jaiiirs T. Iloirman-, Jolin C. Cuyon'^ and Linda L. Wadlci^Ir' Abstract. — Two roadside surveys were condueted tor dwarl inistlctoes parasitizing lodgepole pine and Donglas-lir on the Sawtooth National Forest, Idaho. One sur\e\ used xariahle-radins plots loeated less than 150 in From roads. The 2nd survey used variable-radius plots established at 2()()-in intewals along 16()0-m transects run peipendieular to the same roads. Estimates of the incidence (percentage of trees infected and percentage of plots infested) and severit\- (aver- age dwarf mistletoe rating) for both lodgepole pine and Douglas-fir dwari' mistletoes were not significantly different for the 2 suney methods. These findings are further evidence that roadside-plot surveys and transect-plot surveys con- ducted awa\- from roads pro\'ide similar estimates of the incidence of dwarf mistletoes for large forested areas. Key ivords: dwaij mistletoes, surveys, lodgepole pine, Douglas-fir. Dwarf mistletoes {Arccuilxohiwn spp.) are damaging disease agents in many western forests (Hawksworth and Wiens 1995). In the Intermountain West lodgepole pine {Pinus con- toiia Dongl. ex Loud.) and Douglas-fir [Pseudo- tsiigo inenziesii [Mirb.] Franco) are the most commonly infected trees (Hawksworth and Wiens 1972, 1995, Hoffman 1979). Each of these hosts is parasitized by a different dwarf mistletoe: lodgepole pine dwarf mistletoe (A. americanwn Nutt. ex Engelm.) and Douglas- fir dwarf mistletoe (A. douglasii Engelm.). Severe infection by these parasites is often associated with tree mortality, reduced growth and cone production, tree deformity, and pre- disposition to attack by other diseases and/or insects (Hawksworth and Wiens 1995). There- fore, resource managers in many private, state, and federal land-management agencies imple- ment management activities designed to reduce the damage associated with dwarf mistletoes. Because information on the incidence and severity of these pathogens is required by resource managers for making decisions regard- ing dwarf mistletoe management, surveys are commonly conducted in designated manage- ment units (stands) and over larger areas, such as national forests. Surveys of dwarf mistletoe infection over large areas frequently combine roadside recon- naissance information with data collected using variable-radius or fixed-area plots located near roads (roadside-plot surveys) for estimating the incidence (percent of trees or plots in- fected) and severity (intensity of infection in individual trees; Hawksworth 1956, 1958, Hawksworth and Lusher 1956, Andrews and Daniels 1960, Graham 1960, 1964, Dooling 1978, Hoffman 1979, Johnson et al. 1980, Johnson et al. 1981, Hoffman and Hobbs 1985, Merrill et al. 1985, Maffei and Beatty 1988). Roadside reconnaissance surveys consist of driving roads at slow speed and recording \isual estimates of dwarf mistletoe infection within a short distance from the roadside, usualK 20 m. Dwarf mistletoe incidence is estimated b\ determining the ratio of the number of kilo- meters surveyed adjacent to infected trees to the total kilometers surveyed adjacent to stands predominated by host trees (Dooling 1978). Roadside-plot surveys involve locating plots near roads at specific intervals and collecting tree data including species, diameter, height, age, and mistletoe severity on each plot. Dwarf mistletoe incidence has typically been repre- sented by the percentage of plots infested with mistletoe, rather than the percentage of trees infected in all plots (Dooling 1978). Roadside surveys have the benefit of allow- ing large areas to be surveyed rapidly and ^Idaho Department of Lands, Box 670, Coeur d'Alene, ID 83816. ^Forest Pest Management, USDA Forest Service, 1750 Front Street, Boise, ID 83702. ^Forest Pest Management, USDA Forest Service, 4746 South 1900 East, Ogden, UT 84401. ■tUSDA Forest Service, 524 25th Street, Ogden, UT 84403. 129 130 Great Basin \atlii\li.st [Volume 56 inexpensive!)'. In addition, roadside sunexs concentrate efforts in areas that are accessible and more likely to be considered for manage- ment actions. Concerns about the relialiilit\ of roadside survey methods are primarily related to the bias that may be encountered by sam- pling mistletoe incidence and se\'erity near roads because roads are t\picall\' constructed according to topographic features (in drainages or along ridgetops) rather than randomly or systematically located throughout the suney area. Since there is exidence that d\\ arf mistle- toe distiibution is related to topograph)" ( Hawks - worth 1959, 1968), these concerns need to be considered when conducting dwarf mistletoe suneys oxer large forested areas. Because few sun e)s ha\e compared data collected from roadside reconnaissance or roadside-plot sun'e)'s with data collected from more intensixe, random or s) stematic sun e) s for dwarf mistletoes over large areas (Haw ks- worth 1956, 1958, Johnson et al. 1981, Merrill et al. 1985), we initiated this study to compare dwarf mistletoe incidence and sexerity esti- mates obtained from roadside-plot surve)s xvith those from transect-plot surx'e)S that sampled areas at greater distances from roads. We sune)ed 3 distiicts of the Saxxtooth National Forest, Idaho, because this national forest is representatix'e of forests in the Intennountain West xvhere lodgepole pine and Douglas-fir are the predominant tree species and dxxarf mistletoes are common (Hoffman 1979, Hoff- man and Hobbs 1985). Methods We used a roadside-plot sun e) and a tran- sect-plot suney to collect dxx^arf mistletoe in- cidence and sexerit)- data in 3 adjacent distiicts (Ketchum Ranger District, Fairfield Ranger District, and Saxvtooth National Recreation Area) of the Sawtooth National Forest, Idaho, in 1990. We surveyed each district b)' arbitrar- ily selecting a major road S) stem in each toxxn- ship containing > 10 sections of f ederall) man- aged land. ToxxTiships xvitli no roads or with fexv roads were not sampled. Road systems were chosen before fieldxvork began, and adjustments were made in the field onK xvhen selected I'oad systems xvere closed or impassable. Roadside-plot Suney Field crexvs arbitraril) chose a starting ref- erence point on each selected road sxstem. Starting reference points xxere landmarks that could easily be relocated such as a bridge, stream crossing, or road junction. Crexxs droxe a distance of 800 m from the starting reference point toxvard the center of each toxxnship. Thex' then selected a compass bearing peipen- dicular to the right-hand side of the road and located an end point 120 m from the road. Three 20 basal area factor x ariable-radius plots (iDoint samples; Aver)' and Burkhart 1983) xvere established 40 m from this end point at com- pass bearings of 240°, 120°, and 0° from the compass bearing used to locate the end point. Crexvs then droxe another 800 m doxxn the road and established a 2nd cluster of 3 xari- able-radius plots using the same procedure. For each plot tree the folloxving information xvas recorded: plot number, species, diameter at 1.37 m aboxeground (nearest 0.25 cm), sta- tus (live or dead), and dxvarf mistletoe rating (DMR, 6-class system; Haxvksxvorth 1977). If a plot did not contain trees, it xvas recorded as nonstocked. Transect-plot Suney A 1600-m (approximately 1-mi) transect perpendicular to the road xxas run along the same compass bearing used for establishing the 1st set of roadside plots (800 m from the starting reference point) in each toxxiiship sur- X exed. A 20 basal area factor xariable-radius plot xxas located exen* 200 m along each tran- sect for a total of 8 plots. Infonnation recorded for plot trees xvas the same as aboxe. Analyses The incidence of each species of dxxarf mistletoe (percentage of trees infected) xx'as calculated for each set of roadside plots (up to 6 plots) and each set of transect plots (up to 8 plots) for each toxvnship. Incidence xvas calcu- lated on a per-hectare basis b) multipKing b)' pei-hectare conxersion factors based on 2.54- cm-diameter classes for 20 basal area factor xariable-radius plots (Axen and Burkhart 1983). ^\'eighted dxxarf mistletoe ratings xvere calcu- lated b) multipKing the DMR of each tree by the per-hectare conx ersion factors also. These xveighted xalues xvere used to calculate the mean percentage of trees infected and mean dxvarf mistletoe rating for each suney proce- dure in each toxxnship on a per-hectare basis. These x'alues xx'ere then used to calculate the 1996] Roadside Suhneys eoh Dwarf Mistletoes 131 percentage of trees infected and a mean DMR for each tree species and survey method. Data from townships wliere the surveys chd not sample at least 3 Douglas-lir or lodgepole pine for each of the survey procedures were not included in the analyses. Only living trees were used in the anaKses for calculating mean DMR because it was not aK\a\s possible to accurateh' assign a DMR to dead trees. Inci- dence xalues were calculated for 9 townships for lodgepole pine and for 17 townships for Douglas-fir. The roadside-plot sur\'ey sampled a total of 206 lodgepole pine and 357 Douglas- fir in 46 and 75 plots, respectixeK. The tran- sect-plot survey sampled 171 lodgepole pine and 342 Douglas-fir in 42 and 87 plots, respecti\ el\. A one-way analysis of variance (ANOVA, P > 0.05) was used to determine if the mean values for incidence and severity were significantK different between the 2 survey procediues. Percentages were conxerted using arcsin transformations before ANOVA analyses were perfomied (Snedecor and Cochran 1989). To compare our results with those of other dwarf mistletoe sui-veys, we determined inci- dence of both dwarf mistletoes for both suiA/ey procedures by calculating the percentage of plots infested. If a plot had at least 1 infected tree, it was considered infested. This method of reporting dwarf mistletoe incidence has been applied in the majority of roadside-plot surxeys conducted for dwarf mistletoes in the western United States. Results AND Disc L >>|()N Mean diameters for trees sampled using each survey method were approximately the same for lodgepole pine and Douglas-fir (Table 1). Sampled tree diameters were clearly skewed toward larger trees (Table 1) because both sur\e\' methods used variable-radius plots that sample large trees more often than small trees (Avery and Burkhart 1983). Because both survey methods sampled trees in the same way, the suney results should be comparable. However, it is probable that the percentage of infected trees and mean DMR would have been lower for both lodgepole pine and Douglas-fir had more small trees been sampled because small trees are typically less often and less severelv infected (Parmeter 1978). Estimates of incidence for Douglas-lir dwaif mistletoe using the 2 survey methods were within 3% of each other based on the percent- age of trees infected (Table 2). Estimates of Douglas-fir dwarf mistletoe severity were sim- ilar also. The differences between Douglas-fir dwarf mistletoe incidence and severity for the 2 sui-vey methods were not statistically signifi- cant. The differences between estimates of the incidence and severity of lodgepole pine dwarf mistletoe for the 2 sunex' methods were larger than for Douglas-fir dwarf mistletoe (Table 3). However, the differences were not significant. Therefore, the 2 survey methods Table L Distribution of lodgepole pine and Douglas-fir sampled by diameter classes for the roadside-plot and transect- plot sur\ e\s on the Saw tooth \ational Forest, Idaho. Lod,£ ;epole pine Douglas- fir Roadside -plot Transect-plot Roadside -plot Transect-plot Diameter Mean Mean Mean Mean class diameter diameter diameter diameter (cm) (cm) N (cm) N (cm) N (cm) N 2-13 9.1 47 S.6 39 10.6 17 9.1 10 14-25 19.6 108 19.8 92 20.6 90 20.1 108 26-38 29.2 39 30.0 34 31.8 98 32.0 99 39-51 41.9 7 42.9 5 43.7 85 44.7 60 52-64 60.7 5 51.1 1 56.6 32 57.4 24 >64 — •' — — — 96.5 35 89.4 41 Total 20.8 206 20,1 171 39.5 .357 40.9 342 ••No trees sampled in this size tla 132 Great Basin Naturalist [Volume 56 Table 2. Incidence and seventy of Doiiglas-fir dwarf mistletoe estimated from roadside-plot and transect-plot sui-veys on the Sawtooth National Forest, Idaho. Incidence Se\erit\ Snrvey method Mean percent infected'' 95% mean confidence limit Mean DM Rl' 95% mean confidence limit Roadside-plot Transect-plot 28.4^- 25.8 11.0-15.8 10.0-41.5 0.9^- 0.8 0.2-1.5 0.2-1.4 ''Based on the percentage of individual trees infrctecl im a per-lifctart- l)asis ''Dwaii mistletoe rating (Hawksworth 1977) ^'Means in this cohiinn are not significantK- different; one-way A.NO\'A. P > 0.05. Table 3. Incidence and severit)' of lodgepole pine dwarf mistletoe estimated from roadside-plot and transect-plot sur- ie\'s on the Sawtooth National Forest, Idaho. Incidence Se\'e ■it\' Sin-vey method Mean percent infected'' 95% mean confidence limit Mean DMR'^ 95% mean confidence limit Roadside-plot Transect-plot 48.5^- 55.7 29.4-67.5 35.3-76.1 1.2'' 1.6 0.6-1.8 1.1-2.1 ctare basis '^ Dwarf mistletoe rating (Hawksworth 1977) ''Means in this column are not significantly different; one-way ANOVA, P > 0.0.5. provided equivalent estimates of dwarf mistle- toe incidence, based on the percentage of trees infected, and severity for both dwarf mistletoes. Dwarf mistletoe incidence based on the percentage of plots infested is presented in Table 4. Both survey methods provided esti- mates that were within 2% of each other for both dwarf mistletoes. Calculating dwarf mistle- toe incidence based on the percentage of plots infested greatly increases the estimates of dwarf mistletoe incidence when compared to the incidence based on the percentage of trees infected because it requires only 1 infected ti^ee for a plot to be treated as infested. Lodgepole pine dwarf mistletoe is one of the most widely distributed dwarf mistletoes in the western United States (Hawksworth and Wiens 1995). The incidence of this mistle- toe, based on the percentage of plots infested, has varied between approximately 40% and 70% for the majority of national forests sur- veyed, and averages about 50% (Hawksworth 1958, Graham 1960, 1964, Johnson et al. 1980, 1981, Hoffinan and Hobbs 1985). The incidence of lodgepole pine dwarf mistletoe, based on the percentage of plots infested estimated from our surveys in the Sawtooth National Forest (approximately 80%), is higher than for most national forests surveyed thus far. An earlier dwarf mistletoe suivey of the Sawtootli National Forest (Hoffman and Hobbs 1985) reported the incidence of lodgepole pine dwarf mistle- toe as 71%. However, that survey did not in- clude the Sawtooth National Recreation Area, the district in which we detected a veiy high incidence of lodgepole pine dwarf mistletoe (83%). Therefore, the Sawtooth National For- est probably does have a higher incidence of lodgepole pine dwarf mistletoe than many other western national forests. An earlier estimate of the incidence of Douglas-fir dwarf mistletoe, based on the per- centage of plots infested, for the Sawtooth National Forest was 53% (Hoffman 1979). Although that sune)' sampled onl\ the south- ern districts of the Sawtooth National Forest and did not include the districts we surveyed, our estimate for Douglas-fir dwarf mistletoe, based on the percentage of plots infested, is approximately the same (almost 50%). Our findings provide additional evidence that estimates of incidence and severity of dwarf mistletoes irsing roadside-plot surveys 1996] Roadside Surveys for Dwarf Mistletoes 133 Table 4. Incidence of Donglas-fir and lodf^epole pine dwarf mistletoes based on the percentage of plots infested estimated from roadside-plot and transect-plot surveys on the Sawtooth National Forest, Idaho. Douglas-fir dwarf mistletoe Lodgepole pine dwarf mistletoe Surve\ method Plots Percent infested Plots Percent infested Roadside-plot 75 47 46 80 Transect-plot 87 48 42 78 approximate those of similar surve>'s con- ducted away from roads. Hawksworth (1956) reported similar results based on a more in- tensive comparison of roadside-plot and tran- sect-plot surveys for dwarf mistletoes on the Mescalero Apache Indian Reservation, New Mexico. Partridge and Canfield (1980) com- pared the incidence of several forest pests in southern Idaho estimated using roadside-plot surveys and plots randomly located in areas without roads. They reported no discernible differences between the incidence of the pests detected (including dwarf mistletoes) for the 2 sur\'ey procedures. Because this study and oth- ers indicate that roadside-plot surveys provide similar estimates of dwarf mistletoe incidence to surveys conducted away from roads, we recommend that resource managers continue to use roadside-plot surveys for estimating dwarf mistletoe incidence for national forests or other large forested areas. However, because these surveys sample only a small fraction of the survey area, they will provide only rough estimates of the incidence and severity of dwarf mistletoes. Acknowledgments We appreciate the field assistance provided by Al Dymerski, Valerie DeBlander, Carl Koprowski, and Lia Spiegel. Reviews of the original manuscript by Ralph Williams, Greg Filip, and Catherine Parks are appreciated also. Literature Cited Andrews, S. R., and J. E Daniels. 1960. A survey of dwarfmistletoes in Arizona and New Mexico. USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Paper 49. 17 pp. Avery, T. E., and H. E. Burkhart. 1983. Forest measure- ments. McGraw-Hill Book Company, New York. 331 pp. DooLLNC, O. J. 1978. Survey methoils to determme the distribution and intensity of dwarf mistletoe. Pages 36-44 in Proceedings of the Symposium on Dwarf Mistletoe Control Through Forest .Management. USDA Forest Service, General Technical Report PSVV-31. Ghaham, D R 1960. Surveys expose dwarfmistletoe prob- lem in Inland Empire. Western Conservation J(jur- nal 17; 56-58. • 1964. Dwarfmistletoe survey in western Mon- tana. USDA Forest Sei-vice Research Note I NT- 14. 7 pp. Hawksworth, E G. 1956. Region 3 dwarhnistlet(je survey, progress report on the 1954-55 held work [mimeo- graphed]. USDA Forest Service Special Report, Rocky Mountain Forest and Range Experiment Sta- tion, Fort Collins, CO. 5 pp. . 1958. Survey of lodgepole pine dwarftnistletoe on the Roosevelt, Medicine Bow, and Bighorn National Forests. USDA Forest Sei-vice, Rocky Moimtain For- est and Range Experiment Station, Paper 35. 13 pp. . 19.59. Distribution of dwarfmistletoes in relation to topography on the Mescalero Apache Reserva- tion, New Mexico. Journal of Forestn' 57: 919-922. . 1968. Ponderosa pine dwarf mistletoe in relation to topography and soils on the Manitou Experimen- tal Forest, Colorado. USDA Forest Service Research Note RM-107. 4 pp. . 1977. The 6-class dwarf mistletoe rating system. USDA Forest Sendee Research Note RM-48. 7 pp. Hawksworth, F G., and A. A. Lusher. 1956. Dwarf- mistletoe sui^vey of the Mescalero Apache Indian Reservation, New Mexico. Journal of Forestr)' 54: 384-390. Hawksworth, F G., and D. Wiens. 1972. Biologx- and classification of dwarf mistletoes {Arceuthohium). USDA Forest Service Agricultural Handbook 401. 234 pp. . 1995. Dwarf mistletoes: biology, patholog\', and systematics. USDA Forest Service Agricultural Hand- book 709. 410 pp. Hoffman, J. T. 1979. Dwarf misdetoe loss assessment sur- ve\- in Region 4, 1978. USDA Forest Senice, Inter- mountain Region, Forest Pest Management Report R.4.79.4. 12 pp. Hoffman, J. T, and L. Hobbs. 1985. Lodgepole pine dwarf mistletoe survey in the Intermountain Region. Plant Disease 69: 429^31. JcjHNSON, D. W., E G. Hawksworth, and D. B. Drum- mond. 1980. 1979 dwarf mistletoe loss assessment survey on national forest lands in Colorado. USDA Forest Service, Forest Pest Management, Methods Application Group Report 80-6. 18 pp. Johnson, D. W., F G. Hawksworth, and D. B. Drum- mond. 1981. Yield loss of lodgepole pine stands to dwarf mistletoe in Colorado and Wyoming national forests. Plant Disease 65: 437-438. Maffei, H. M., and J. S. BE.^rrt'. 1988. Changes in the incidence of dwarf mistletoe over 30 years in the Southwest. Pages 80-90 in Proceedings of the 36th Western International Forest Disease Work Confer- ence, Park City, UT, 19-23 September 1988. Merrill, L. M., E G. Hawksworth., and D. W. John- son. 1985. Evaluation of a roadside survey proce- dure for dwarf mistletoe on ponderosa pine in Col- orado. Plant Disease 69: 572-573. 134 Great Basin Naturalist [Volume 56 Parmeter, J. R. 1978. Forest stand cKnaiiiics and ecologi- cal factors in relation to dwarf mistletoe spread, impact, and control. Pages 16-30 in Proceedings of the Symposium on Dwarf Mistletoe Control Through Forest Management. USDA Forest Sewice, General Technical Report PSW-31. Partridge, A. D., and E. R. Canfield. 1980. Frequency and damage by torest-tree pests in southern Idalio. Universit)' ol Idaho, Moscow; Forestry, W'ildland, and Range Experiment Station Note 34. 7 pp. Snedecor, G. W, and W. G. Cochran. 1989. Statistical methods. Iowa State University Press, Ames. Received 13 February 1995 Accepted 23 October 1995 Great Basin Naturalist 56(2), © 1996, pp. 135-141 EFFECTS OF DOUGLAS-FIR FOLIAGE AGE CLASS ON WESTERN SPRUCE BUDWORM OVIPOSITION CHOICE AND LARVAL PERFORMANCE KimherK' A. Dockls', Karen M. Clanc>-, Kathryn J. Lcyva'^, David Greenherg\ and Peter W. Price'^ Abstract. — The western spruce hudworm {Churistoiiciira occidcnfali.s Freeman) prefers to feed on llusliin^ hiids and current-year needles ot Douglas-tir {Pseudotsuga menziesii [Mirh.] Franco). Budworin lan'ae will not t\picail\- con- sume older age classes of needles unless all current-year foliage is depleted. We tested the following null hypotheses: (1) budworm lanae can feed on foliage with a wide range of ((ualities (i.e., cuirent-year versus 1-, 2-, or 3-vear-old needles) without measurable effects on fitness; and (2) budworm adults do not show any oviposition preference linked to the age of the foliage they fed on as larvae. We used both laborator\ and field experiments. There was strong evidence to sup- port rejection of hypothesis 1. Budworm larvae had greater survival from the 4th instar to pupal stage when they fed on current-year foliage (43%-52% sui-vival) versus older age classes of foliage (0-25% survival). Pupae from current-year foliage were also heavier than pupae from > 1-year-old foliage. There was weak evidence to support rejecting hypothesis 2; budworm adults that had fed on current-year or 3-year-old foliage as larvae preferred to oxijiosit on current-year foliage. Similar conclusions were drawn from the laboratoiy and field experiments. Keij words: Choristoneura occidentalis, western spruce hiidwonii. oviposition preference, needle age, foliar qualitij. eruptive species. The western spruce budworm {Choris- toneura occidentalis Freeman) is a major defo- liator of Douglas -fir {Pseudotsuga menziesii [Mirb.] Franco) trees in western North Amer- ica (Fellin and Dewey 1982, Wulf and Gates 1987, Clancy et al. 1988). Budworm lan^ae pre- fer to feed on the flushing buds and current- year needles of their host trees. However, if all current-year foliage is depleted, larvae will feed on older needles (Fellin and Dewey 1982, Talerico 1983, Blake and Wagner 1986). Previ- ous experiments by Talerico (1983) and Blake and Wagner (1986) show older foliage is sub- optimal, resulting in reduced fecundit\; higher mortality rates, and impaired dexelopment. When budworm larvae are forced to feed on only mature foliage, they have reduced growth, lower pupal weights, and decreased sunaval, or they may not sui-vive at all (Blake and Wag- ner 1986). Variations in host foliage quality may influ- ence the feeding and oviposition behavior of the western spruce budworm (Clancy et al. 1988). Differences in levels of foliar nutrients, water content, needle toughness, etc., between cunent-year and older (> 1-year-old) age classes of needles impact the budworm's fecimdit); growth rate, and survivorship (Mattson and Scriber 1987, Clancy et al. 1988, Clancv- 1991b, 1991c), and may influence female oviposition choices. However the budwonn s oligophagous feed- ing behavior and eruptive population dynamics suggest it is unlikely that there is a tight link- age between female oviposition preference and larval performance (Price et al. 1990). Female moths do not determine where their offspring will feed. Budworm adults lay eggs on mature foliage in late summer (Furniss and Carolin 1977, Brookes et al. 1987); upon hatch- ing, the 1st instars (which do not feed) dis- perse to sheltered locations (e.g., beneath bark scales), where they spin a hibernaculum and ovei-winter When larvae emerge from their hibernacula the following spring, the>- disperse again (typically on silken threads) to find appro- priate food sources. The budworm's life his- tory suggests that neither adults nor larvae actively select host foliage based on differ- ences in nutritional quality among individual host trees. Instead, larvae passively disperse from their ovenvintering sites and ma> land 'Dt-partiiient ol Forestn-, Northern Arizona Universih; Box 1.5018, Flagstaff, A2. 86011. 2Rock>- Mountain Forest and Range Experiment Station, USDA Forest Senice Rcsearcli. 2.500 S. I'inc Kn ^Department of Biological Sciences, Northern Arizona University', Box .5640, Flagstaflf, .^Z 8601 1. li:)rivc. Flagstaff, . 1/86001 . 135 136 Great Basin Naturalist [Volume 56 on acceptable food sources. Once larvae are on a host tree, they search for expanding cur- rent-year l>uds and needles. If suitable foliage is not available, larvae can disperse horizon- tally or vertically within and between tree crowns and stands, but dispersal invariably results in significant losses; whether dispers- ing larvae live or die depends largely on whether they find hospitable sites (Brookes et al. 1987). Therefore, the ability to utilize a broad range of foliage qualities would be advantageous for budworm sui"vival. This study was designed to compare results from laboratory and field tests of the null hypotheses that (1) budworm larvae can feed on foliage with a wide range of qualities (i.e., current-year versus 1-, 2-, or 3-year-old nee- dles) without measurable effects on fitness; and (2) budworm adults do not show any oviposition preference linked to the age of the foliage they fed on as lai"vae. Furthermore, we wanted to determine if conclusions drawn from laboratory versus field experiments were similar. This is important because many previ- ous studies conducted with budworms and other forest defoliators have used clipped foliage without knowing the effects this may have on foliar nutrition or host defenses. By conducting parallel experiments using intact and excised foliage from the same trees, we were able to evaluate the importance of changes in foliar quality that may be associ- ated with bagging lan'ae on intact branches in the field versus feeding lai^vae excised foliage in the laboratoiy Study Area and Organisms The study area is located at Little Springs (elevation 2560 m), 16 km north of Flagstaff, Arizona, within the Coconino National Forest. The site is a high-elevation, mixed-conifer for- est, with Douglas-fir as the primaiy host species and with a recent history of western spruce budworm infestation. The western spruce budworm has a imix'ol- tine life cycle. Adults are present from Jidy to August, with mating typically occurring within 24 h of eclosion. Eggs are laid soon after mat- ing; females lay between 25 and 40 eggs per egg mass (Brookes et al. 1987). Eggs hatch in about 10 d; after dispersing to sheltered loca- tions and spinning a hibernaculum, larvae molt into 2nd instars and ovei'winter. In spring, the 2nd instars emerge fi-om diapause, feed through the 6th instar, and pupate in late June or early July. Adults eclose within 10 d. Our laboratoiy population of nondiapausing western spruce budworm differs in tliat there is no ovei'win- tering stage. Methods Field Experiment To determine larval performance in the field, we selected and tagged 50 Douglas-fir trees of various sizes and ages on 1-2 June 1993. All tagged trees had abundant foliage in the lower crown. Sleeve bags made of fine mesh screen were placed over 4 branches on each of the 50 trees, and each bagged branch was randomly assigned to a foliage age class (current-year, 1-year-, 2-year-, or 3-year-old needles). We removed by hand all needles that were not of the appropriate age class. Any wild budworms present on the bagged branches were also removed. On 4 and 8 June, two 4th or early 5th instar budworm larvae from oin- laboratory culture were placed on foliage inside each bagged branch (a total of 400 lan'ae were used); this constituted the parental (P^) generation. We have established that budworm larvae from our laboratory culture have rates of sur\'ival and reproduction equivalent to wild bud- worms when reared on Douglas -fir foliage in the field (Le\'va et al. 1995). Bags were closed with string or duct tape at each end. We exam- ined the bagged branches on 20 June to deter- mine if sufficient foliage remained for comple- tion of larval development. Pupae were not obsei"ved at this time. Budworm lanae remained in the field imtil about half of them had pupated, and then the bagged branches were clipped, placed inside large plastic bags, and transported to the labo- ratoiy. Pupae were weighed (to the nearest 0. 1 mg), sorted into trays according to treatment (foliage age class) and sex, and then refriger- ated at 10 °C until we obtained 10 males and 10 females from the same treatment. Larvae that had not pupated were placed in labeled petri dislies lined with moist filter paper. Dou- glas-fir foliage of the appropriate age class was provided for them to feed on until they pupated; this foliage was collected at random from tagged Douglas-fir trees at the study site. Foliage was replaced ever\ 2-3 d to ensure 1996] Effects of Fouack Ack Class on Budworms 137 freshness. Petri dishes were ehecked each Mon- day, Wednesday, and Friday to remove and weigh new pupae (tliest^ were also sorted and refrigerated). When 10 pairs of male and female pupae were a\'ailable from a treatment, they were placed in a l)rown paper mating hag; oviposi- tion preference tests for both field and lahora- tor\' experiments were conducted in the labo- iator\. 13i"o\\'n paper bags provided appropri- ate lighting conditions both for mating (which occurs from 2000 to 2300 h in nature), when only safety lights were on in the laboratory at night, and for oviposition (which normallv occurs the da>' following mating), when all lights were on during the day (i.e., bags are not opaque). Bags were checked every other day until 5 or 6 moths emerged; then liranches of freshly clipped Douglas-fir foliage were added for oviposition substrate. Once foliage was added, moths were allowed to mate and oviposit for 7-8 d. After oviposition occurred, Douglas- fir branches were removed and inspected for egg masses. These Fj egg masses were sorted according to treatment (foliage age class) to de- temiine if female moths showed a preference for ovipositing on a particular age class of foliage. The Fj egg masses collected were surface- sterilized with formalin and placed into labeled cups containing an artificial diet nutritionally similar to Douglas-fir foliage (Clancy 1991a). Fj lai-vae were reared on the diet until the 4th or early 5th instar stage, after which they were placed in labeled petri dishes lined with moist filter paper. Douglas-fir foliage of the same age class that their parents consumed was pro- vided for them to feed on until they pupated; this foliage was collected at random fiom tagged Douglas-fir trees at the study site. These lar- vae were not placed in the field because it was too late in the season for conditions suitable for budwomi dexelopment. Foliage was replaced eveiy other day to ensure freshness. Fj larvae were reared on foliage within petri dishes until they pupated; pupae were handled in the same manner as in the first generation. Laboratoiy Experiment This study was conducted to determine if laboratory experiments using excised foliage would yield results similar to those from field experiments using intact foliage. The experi- ment was started 24 June 1993. Douglas-fir foliage used in this experiment was collected from the same 50 trees we used for the field experiment. Four hundred 4th instar bud- worms were placed on excised foliage in petri dishes lined with moist filter paper, 2 larvae per dish. Fift\' petri dishes were used per foliage age class treatment (current-year, 1- year-, 2-year-, and 3-year-old needles), corre- sponding to the 50 trees used in the field experiment. Needles of the appropriate age class were left attached to the stem to prevent desiccation of foliage. Foliage was replaced ever>' 2-3 d to ensure freshness. Petri dishes were labeled according to the tree number and foliage age class. If a single larva or both lai-vae in each petri dish died before pupation, they were replaced with new lanae from our lal:)oratoiy culture. Othenvdse, we used the same procedures for the laboratory experiment as for the field experiment. Results Effects of Foliage Age Class on Pj Survival and Pupal Weight Budworm larvae that consumed current- year needles of Douglas-fir in the field experi- ment had higher sumval rates from 4th instar to pupal stage (43% sui-vival) compared to lar- vae that fed on 1-year-old (2% survival), 2- year-old (1% survival), or 3-year-old (0% sur- vival) needles (Fig. lA) {y} =^130.19, df = 3, P < 0.001, n = 400). We believe that many of the larvae bagged on the branches with only > 1-year-old needles to feed on escaped from the mesh bag enclosures, so it may be more appropriate to refer to this response as "per- cent lai-vae accounted for" rather than "percent lai-val sui-vival.' Lai-vae from older foliage age class treatments were more likely to escape because budwonn lan'ae tend to disperse when suitable food is not available, and our bags were not so tightly sealed that lanae could not wriggle out through small openings along the seams or closures at the ends. The age class of foliage ingested had a simi- lar effect on sunival from 4th instar to pupal stage in the laboratory experiment (Fig. IB) (X2 = 59.46, df = 3, P < 0.001, n = 727). Approximately 52% of larvae that consumed current-year needles survived. Survixal was 25% for larvae consuming 1-year-old needles, 18% for larvae feeding on 2-year-old needles, and 20% for larvae eating 3-year-old needles. 138 Great Basin Naturalist [Volume 56 Current-yr 1-yr-old 2-yr-old 3-yr-old Current-yr 1-yr-old 2-yr-old 3-yr-old Foliage Age Class Fig. 1. Percentage of Pj 4th instar western spruce bud- worms surviving to the pupal stage when reared on cur- rent-, 1-, 2-, and 3-year-old Douglas-fir needles for the (A) field experiment and (B) laboratory experiment. %- tests showed that sui-vival varied among the foliage age classes for both the field (P < 0.001) and laboratoiy (P < 0.001) experiments. Numbers above the bars indicate sample sizes, i.e., number of budworm lai^vae used per treatment. Foliage age class did not have a significant effect on pupal masses for the field experi- ment (F = 1.97, df = 2,41, P = 0.152; Fig. 2). This inability to detect differences among foliage age classes can be attributed to the veiy small sample sizes (n = 0-2) for > 1 -year- old foliage. As expected, female pupae were heavier than male pupae (F = 20.39, dl = 1,41, P < 0.001). There were detectable differences in pupal masses among different foliage age classes for the laboratory experiment (F = 36.47, df = 3,182, P < Oi)01; Fig. 3). Larvae consuming current-year foliage became nuich heavier pupae than larvae feeding on > 1 -year-old foliage. Once again, female pupae were bigger than male pupae (F = 14.70, df = 1,182, P < 0.001). Effects of Foliage Age Class on Oviposition Preference of Fy Females Sample sizes for the field experiment were not large enough for data analysis (n = 2 egg masses), but a contingency table analysis of data from the laboratory experiment indicated 03 Q. 3 Q. UJ '\'a et al. 1995). Nonethe- less, the distribution of Fj egg masses indi- cated that moths reared as Uwvae on current- year or 3-year-old foliage laid more of their F^ egg masses on current->'ear needles than on older age classes of needles. Effects of Foliage Age Class on F| Sin\i\al and Pupal Weight Only 2 Fj egg masses were produced from the field experiment. This precluded analyz- ing data on sur\ i\ al or pupal weights for this experiment. For the laborator) experiment, we found a significant difference in survival from 4th instar to pupal stage between larvae reared on cinrent-\'ear (83.3% sundval) versus 3-year- old (3().87r surNixal) needles {%- = 11.78, df = 1996] Effects of Foliage Ace Class on Budworms 139 a 3 Q. UJ W CN +1 Current yr 1 -yrold 2vr-old 3-vr-old Foliage Age Class Fig. 3. Mean (± 2 s, or =95% confidence interval) Pj male (■) and female (D) pupal weight for larvae reared from the 4th instar to pupation on foliage of different age classes in the laboratory experiment. ANOVA tests showed that foliage age class had a significant effect on pupal weight {P < 0.001), as did sex {P < 0.001). 1, P = 0.0006, n = 44). Sample sizes were 0 for 1-year- and 2-year-old needles. This result was consistent with results for the Pj genera- tion in that survival was higher for larvae reared on current-year foliage than on 3-year- old needles. However, F^ pupal masses were equivalent for pupae from current-year and 3-year-old foliage (F = 1.14, df = 1,19, P = 0.299). As before, female pupae were larger than male pupae (F = 6.01, df = 1,19, P = 0.024). Discussion Although the western spruce budworms life history and population dynamics suggest that larvae shoidd be able to utilize a broad range of foliage qualities, our results confirm previous studies that indicate the budworm is not well adapted to feeding on 1 -year-old or older needles (Ttilerico 1983, Blake and Wag- ner 1986). Thus, we must reject hypothesis 1 and conclude that the budworm cannot feed on foliage with as wide a range in qualities as is found in current-year versus > 1-year-old needles without measurable effects on fitness. We found that whether budworm larvae were feeding on bagged foliage in the field or on excised foliage in the laboratory, larval survival (Fig. 1) and pupal masses (Figs. 2, 3) declined for larvae feeding on > 1 -year-old needles compared to larvae feeding on current-year foliage. The same patterns in relation to the effects of foliage age on sm-vival were evident for both the P^ and Fj generations of the labo- ratory experiment. It is well established that the nutritional quality of Douglas-fir needles declines rapidly as current-vear needles age (Clancy et al. 1988, 1995). Furthermore, Clancy et al. (1995) point out that the general pattern for l-ycar or older needles of conifers is typically an extension of the seasonal trends for nutrient concentration changes in cur- rent-year needles. Needle toughness and fiber content also increase as foliage matures, thus making older needles less suitable food for the budworm. On the other hand, the fact that budworm lar- vae could survive at all when reared on older age classes of needles may indicate that their nutritional niche is indeed broad, as suggested b>' Price et al. (1990) and Lex^a et al. (1995). We fotmd less-convincing evidence to sup- port rejection of hypothesis 2, but nonetheless we conclude that budworm adults may show an oviposition preference that is linked to the age of the foliage they fed on as lanae (Table 1). Budworm females that fed as larvae on cm- rent-year foliage laid more egg masses on cur- rent-year needles than on older needles; females reared on 3-year-old foliage also laid more egg masses on current-year needles. This result is suiprising because the budworm typically ovi- posits on mature foliage (Brookes et al. 1987, Price et al. 1990). However, many egg masses from our experiment were very small and arc most likely aberrant (Leyva et al. 1995); thus, we suspect this may not represent normal oviposition behavior for the budworm. If we remove these veiy small egg masses from the data set, most egg masses were laid on 3-year- old needles, indicating the budworm does indeed prefer to oviposit on mature foliage. An alternative explanation ma\- be that female moths distribute egg masses randomly across age classes of needles available. Current-year needles represent a small proportion of total needles present under natural conditions, and they may be nearly absent when defoliation is heavy. Thus, our result could be an artifact of providing an atypical distribution of needle age classes for oviposition substrate. 140 Great Basin Naturalist [Volume 56 Table 1. Distribution of Fj egg masses of the western spruce budworm laid on current-, 1-, 2-, or .3-year-old needles of Douglas -fir, in relation to the age class of the foliage on which the Pj moths were reared'*. Age class of foliage on which Fj egg masses were laid Age class of foliage on which Pj female was reared'' Cmrent-vear l-\ ear-old .'3-\ear-old Current-year 1-year-old 2-year-old 3-year-old 27 6 5 11 14 0 1 0 ''Data are the niimber of Fj egg niaises lioni the laboiatoiy experiment. The distiiluitioii ol egg masst.s \\ as examined for ri)w-column dependence nsing a 2 x 2 contingency talile (current-year versus > 1-year-old foliage; older foliage age classes weie combined in order to meet requirements for minimum expected cell frequencies); Yates-corrected X" = .5.72.5, df = 1, P = 0.017, n = 64. ''Pi budwonu larvae were reared on foliage of different age classes from the 4th instar to pupation. Conclusions from the field versus labora- tory experiments were similar in regard to the effects of foliage age class on larval survival and pupal masses. This indicates that foliar qual- ity does not change dramaticall)' when foliage is excised (at least not over a 2-3 d period), nor does intact foliage on bagged branches change markedly in temis of a local or systemic- induced response to budworm defoliation. It is noteworthy that pupae from the cuirent- year foliage treatment in the field experiment were heavier than equivalent pupae from the laboratory experiment (Figs. 2, 3). This differ- ence may well be related to the 15-19 d delay between the start of the field experiment (ini- tiated on 4 and 8 June 1993) and the begin- ning of the laboratory study (initiated on 24 June 1993), with the concomitant decline in nutritional quality of the expanding current- year foliage. Alternatively, current-year needles remaining on bagged branches could have acted as a local nutrient sink in the absence of competing older needles, which were removed. Thus, lai-vae feeding on bagged branches with current-year foliage may have benefited from this improved nutritional quality, whereas lar- vae that were fed clipped foliage in the labora- tory experiment would not have received this nutritional boost. Survival rates were higher overall in the laboratoiy experiment than in the field experi- ment (Fig. 1). We attribute this difference to a combination of factors. For example, lai^vae on bagged branches in the field were exposed to some predation since the bags were not per- fect barriers. Also, some budworm lai-vae un- doubtedly escaped from the bags, and weather could have played a role in the lower survival of lai-vae in the field. Acknowledgments We thank Daniel Huebner and Kenneth Dodds for assistance in collecting data and conducting experiments. We also thank D. Leatherman and an anonymous reviewer for their comments on an earlier draft of this paper. Our western spruce budworm colony was started with egg masses obtained from the Canadian Forest Service s Pest Management Institute in Sault Ste. Marie, Ontario. This re- search was funded in part by the Rock>' Moun- tain Forest and Range Experiment Station, Cooperative Agreement 28-C3-708 between KMC and PWE Literature Cited Blake, E. A., and M. R. Wagner. 1986. Foliage age as a factor in food utilization by the western spruce bud- worm, Chorisfoneuro occidenfalis. Great Basin Natu- ralist 46: 169-172. Brookes, M. H., J. J. Colbert, R. G. Mitchell, and R. W. Stark, editors. 1987. Western spruce budworm. USDA Forest Service Technical Bulletin 1694. Clancy, K. M. 1991a. Multiple-generation bioassa> for investigating western spruce budworm (Lepidoptera; Tortricidae) nutritional ecology. En\ ironmental Ento- molog)' 20; 136.3-1374. . 19911). Douglas-fir nutrients and terpenes as potential factors intluencing western spruce budwomi defoliation. Pages 123-133 in Y. N. Bai-anchikov, W. J. Mattson, F Hain, and T. L. Payne, editors. Forest insect guilds: patterns of interaction with host trees. USDA Forest Sen ice, General Technical Report NE- 1,53. . 1991c. Western spruce biidworni response to dif- ferent moisture levels in artificial diets. Forest Ecol- ogy and Management 39: 223-235. Clancy, K. M., M. R. W.agner, and R B. Reich. 1995. Ecophysiology and insect herbivory. Pages 125-180 in W K. Smith and T. M. Hinckley, editors, Eco- physiology of coniferous forests. Academic Press, New York. 19961 Effects ok Foliage Ace Class on Budworms 141 Clancy, K. NL, M. R. \V.u;m:i;, am) R. W. Tims. 198' plants; water, nitrogen, fiber, and mineral considerations. Pages 105-146 in F. Slansky, Jr., and J. Rodriguez, editors, Nutritional ecology of insects, mites, spiders, and related inxertebrates. John VV'ilev and Sons, New York. l^RicE, R VV, N. Cobb, T R Crak., c. U. J-imn andls, j. K. ITAMI, S. MOPPER, AND R. W. Preszler. 1990. Insect herbivore population dynamics on trees and shrubs; new apiiroaches relevant to latent and eruptive species and life table development. Pages 1-38 i» E. A. Rernays, editor, Insect-plant interactions. Vol- ume 2. CRC Press, Roca Raton, FL. Talerico, R. L. 1983. Summaiy of life histoiy and hosts (jf the spruce budworms. Pages 1-4 in Proceedings; Forest Defoliator- Host Interactions; a Comparison Between Gypsy Moth and Spruce Rudworm, USDA Forest Service, General Technical ReiMirt \E-85. Wl'LF, N. W, and R. G. C.\tes. 1987. Site and stand char- acteristics. Pages 89-115 in M. H. Brooks, R. VV. Campbell, J. J. Colbert, R. G. Mitchell, and R. W. Stark, editors. Western spruce budworni. USDA Forest Service Technical Bulletin 1694. Received 20 October 1995 Accepted 27 February 1996 Great Basin Naturalist 56(2), © 1996, pp. 142-149 TRYPANOPLASMA ATRARIA SR N. (KINETOPLASTIDA: BODONIDAE) IN FISHES FROM THE SEVIER RIVER DRAINAGE, UTAH J. Stephen Cranne\'' and Ricliard A. Heckmann^ Abstract. — A total of 181 fishes l)elon^ing to 10 species were captured near Richfield, Utah, and examined for para- sites. A new species of hemotlagellate, Tnjpanoplasma afraria sp. n., was ohsei-ved in 3 species: Utah chub {Gila atraria [Girard]), redside shiner (Richardsoniits balteatus [Richardson]), and speckled dace (Rhinichthijs oscuhis [Girard]). Seven other species of fishes examined in the study area were negative for T. atraria sp. n. The salmonid leech, Piscicola salmositica (Meyer), collected in the same area harbored developmental stages of Tnjpanoplasma, suggesting a possible leech vector for the hemoHagellate. Characteristics of Tnjpanoplasma atraria sp. n. place it near T salmositica, but the new species is twice as large. Key words: Tnpanoplasma atraria n. sp., blood parasites, Gila atrari;i,_//s/i parasites. Tnjpanoplasvm is a biflagellated protozoan found in the blood of freshwater fishes in the United States. It has caused significant mor- tahty in rainbow trout {Oncorhiinchus inykiss [Walbauni]) and king sahnon (O. t.shauytscha [Walbaum]) under hatchen' conditions (Becker and Katz 1966, Wales and Wolf 1995). This genus has also been described from the blood of marine fish (Strout 1965). Another name for the blood biflagellate of salmonids described above is Cryptobia. There are differing opinions on the use of the two genera, Cryptobia and Tryp(inoplas)na, but these differences have been recently clarified by Lom and Dykova (1992). The genus Cryptobia was first proposed by Leidy (1846) for biflagellated protozoans occur- ring as parasites in the seminal vesicles of snails. Chalachnikow (1888) was the first to record the parasite in the blood of fishes, observing it in freshwater loaches in Russia. Uaveran and Mesnil (1901) established the genus Trypano- plasma for a biflagellated blood parasite from freshwater fishes in France. In 1909, Crawley stated that Cryptobia from snails and Trypano- plasina iiom fishes were moiphologically iden- tical, and that Cryptobia had taxonomic prior- ity. In defending the creation of the genus Try- panoplastna, Laveran and Mesnil (1912) argued that morphological similarities were not suffi- cient criteria for maintaining a single genus when strong biological differences, such as method of infection, were evident. The para- sites in snails were transferred directh dminir copulation, while a leech vector was necessar\^ to transfer the flagellate from the blood of one fish to another Putz (1970) submitted that comparative biological studies between simi- lar morphological types are necessary for a correct ta.xonomic classification. Use of the genus Cryptobia has, in most cases, emerged as the popular choice, and Trypanoplasina is generally recognized as a synonym. Recently, Lom and Dykova (1992) used Trypanoplasma for biflagellated blood-inhibiting parasites of fishes in which a leech vector is involved. Thus, we adopted the classification scheme used by Lom and Dykova (1992). Four species of Trypanoplasma from the blood of fresh\\'ater fishes ha\'e been reported in North America. Mavor (1915) found T. borreli in a moribund white sucker {Catostoinus com- inersoni [Lacepe]) from Lake Hiu-on. The iden- tification of T. borreli was based on similarities with the species initialK described by Laxeran and Mesnil (1901). Katz (1951) recorded C. ( = Trypanoplas)na) .sabuositica from silver salmon (O. kisiitcli [Walbaum]) and C. ( = Try- panoplas)na) lynclii from cottids in the state of Washington. Subsequent transmission studies showed C. hjnchi to be a synonym of C. salmo- sitica (Becker and Katz i965a). Laird (1961) described C. ( = Trypan()plasma) <;urni'yonim from northern pike {Esox luciiis [Linnaeus]) and from 2 salmonids: lake whitefish {Core- ' 61: 1242-1250. . 1984. Detection of infection and susceptibilib,' of different Pacific salmon stocks {Oncorhynchiis spp.) to the haemoflagellate Cryptohia sahnositica. Journal of Parasitolog)- 70: 273-278. BuRRESON, E. M. 1982. The life cycle of Trypanoplasma hiillocki (Zoomastigophorea: Kinetoplastida). Journal of Protozoology 29: 72-77. Chalachnikow, A. P 1888. Recherches sur les parasites du sang. Arkliives Veterinan' Science, St. Petersburg 1: 65. Cope, O. B. 1958. Incidence of external parasites on cut- throat trout in Yellowstone Lake. Proceedings of the Utah Academy of Science, Arts, and Letters 35: 95-100. Crawley, H. 1909. Studies on blood and blood parasites. II. The priority of Cryptohia Leidy, 1846, over Try- panoplasma Laveran and Mesnil, 1901. U.S. Depart- ment of Agriculture, Bulletin of the Bureau of Ani- mal Industry 119: 16-20. Heckmann, R. A. 1971. Parasites of cutthroat trout from 'i'ellowstone Lake, Wyoming. Progressive Fish Cid- turist 33: 103-106. Hoffman, G. L. 1967. Parasites of North American fresh- water fishes. University of California Press, Berkeley and Los Angeles. 486 pp. Humason, G. L. 1967. Animal tissue technicjues. Freeman and Compan\', San Francisco. 596 pp. K-VTZ, M. 1951. Two new hemoflagellates (genus Cryptohia) from some western Washington teleosts. Journal of Parasitolog>' 37: 245-250. Khan, R. \. 1991. Further obsenations on Cryptohia dahli (Mastigophorea: Kinetoplastida) parasitizing marine fish. Journal of Protozoologx 38: 326-329. Khan, R. A., and E. R. Noble. 1972. Ta.\onomy, preva- lence, and specificity of Cryptohia dahli (Mobius) (Mastigophora: Bodonidae) in liunpfish, Cyclopterus himptis. Journal of the Fisheries Research Board of Canada 29: 1291-1294. L\1RD, M. 1961, Parasites from northern Canada. II. Haematozoa of fishes. Canadian Journal of Zoolog\- 39:541-548. Laveran, A., and F Mesnil. 1901. Sur les flagelles a mem- l)rane oudulaute des poissons (genres Trypanosoma 1996] Tnri'ANOPLASMA ATllMilA BLOOD PaRASITE OF FiSH 149 Fniln- ct Tn/i>(i)ioplasiiui n. gen.). Tran.saction.s of the French Acaclcnu' of Science 133: 670-675. . 1912. Tnpano.sonie.s et Tnpanosomiases. 2n(l edi- tion. Xhusson et Cie, Editenrs, Pari.s. 999 pp. Leidy, E 1846. Descriptions of a new genns and species of Entozoa. Proceedings of the National Academy of Science, l^hiladelpliia 3: 100-101. LOM, J. 1979. BiologN of the trypanosonies and trxpano- plasnis of fish. Pages 269-337 in W. II. H. Eninsden and D. A. Evans, editors. Biology of the Kinetoplas- tida. Volmne 2. .'\cademic Press, New York. LoM, J., ,\ND I. DYKt)\A. 1992. Protozoan parasites of fishes. Dexelopnients in acjiiacnhnre and fisheries science, N'oknne 26. Elsevier Pnhlishers, Anistersckun. 31.5 pp. M.A\OK, J. W. 1915. On the occnrrence of a tn'panoplasm, probal)l\ Tnjpanophisina boireli Laveran et Mesnil, in the blood of the common sncker, Catostornus com- iiicrsonii. Jom^nal of Parasitology' 2: 1—6. McD.'VNiEL, D. W. 1970. Personal communication. U.S. Department of Interior Fisheries. Spring\'ille, Utah. Noble, E. R. 1968. The flagellate Cnjptobia in two species of deep sea fishes from the eastern Pacific. Journal of Parasitolog>' 54: 720-724. PUTZ, R. E. 1970. Biological studies on the hemoflagel- lates (Kinetoplastida: Cnptobiidae) Cnjptohki catarac- tae sp. n. and Cnjptobia saJinositica Katz, 1951. Un- published doctoral dissertation, Fordham University, New York. 98 pp. . 1972a. Cnjptobia catamctae sp.n. (Kinetoplastida: Cryptobiidae), a hemoflagellate of some cyprinid fishes of West Virginia. Proceedings ul tlit Helminthological Society of Washington 39: 18-22. 1972b. Biological studies on the heni()flagellate> Cnjptobia cafaractae and Cnjptobia salmositica. Technical Papers, Bureau of Sport Fisheries and Wildlife 63: 1-25. Stkout, R. G. 1965. A new hemoflagellate (genus Cnjpto- bia) from marine fishes of northern .New England. Journal of Parasitology 51: 654-659. SWEZV, (). 1919. The occurrence q{ Tnj])anoplasma as an ectoparasite. Transactions of the American .Micro- scopical Society 38: 20-2-1. W.ALES, J. H., AND H. WoLE 1955. Three protozoan dis- eases of trout in California. California Fish and Game 41: 183-187. Wenkich, D. II. 1931. A tnpanoplasm on the gills of caip from the Schnvlkill Hiver |ourn:il oii^nasilologv 18; 133. Woo, P T. K. 1987. Cryptobia and ciyptobiosis in fishes. Advances in Parasitology 26; 199-237. Woo, P T. K., A\n Wkiinekt, S. D. 1983. Direct transmis- sion of a hemoflagellate, Cnjptobia sabnositica (Kinetoplastida: Bodonina), between rainbow trout under laboraton conditions. Journal of Protozoology 30: 334-337. Received 1 September 1995 Accepted 19 Januanj 1996 Great Basin Naturalist 56(2). © 1996, pp. 150-156 GEOGRAPHICAL REVIEW OF THE HISTORICAL AND CURRENT STATUS OF OSPREYS {PANDION HALIAETUS) IN UTAH Clark S. Monson^ Abstract. — Small numbers of Ospreys {Pandion haliaetits) are known to have nested historically in Utali. A precise baseline figure is unavailable, but the 19th-century Osprey population in Utah probably consisted of at least 15 breeding pairs scattered in 4 geographic regions. Human persecution is believed to have caused the abandonment of nesting ter- ritories along the Wasatch Front and in the western Uinta Mountains b\' 1900 and 1960, respectively. Osprey popula- tions in the southern plateaus and Green River areas, however, began increasing in the late 1970s. Sexeral recent nesting attempts and numerous summer sightings at nontraditional and abandoned historical sites in Utah suggest the Osprey is also expanding its range in Utah. High productivib.' for local pairs and long-range dispersal from more northerly Osprey populations are discussed as sources for the current surge in Utah's Osprey population, which now consists of approxi- mately 35 breeding pairs. Key words: Osprey. Pandion haliaetus, raptor Flaming Gorge Reservoir, dispersal. The Osprey {Pandion haliaetus) is one of the most wideh' distributed species of raptors dur- ing the breeding period. The extent of its cos- mopohtan range is exceeded by only 2 other raptors: the Peregrine Falcon {Falco peregri- mis) (Cade 1982, del Hoyo et al. 1994) and Barn Owl {Tyto alba) (Marti 1985, Eckert and Karalus 1987). Despite the Ospreys broad geographic distribution, local populations occur in fragmented and low densities in much of the species' range (Bent 1937, Palmer 1988, del Hoyo et al. 1994). This scenario holds true for most of the intermountain region of the west- ern United States (Kenny 1986, Johnsgard 1990). In Utah, Osprey distribution has been particularly limited. Recently, however, several personal summer obsen^ations of Ospreys over 140 km from known breeding pairs prompted an investigation into the possible occurrence of Ospreys at other nontraditional Utah localities. A sui^vey of indi\'iduals from the U.S. Forest Sei^vice, Utah Division of Wildlife Resources, Utah State Parks, and other persons familiar with Osprey ecology was conducted dining 1994-95. The survey revealed man\' Osprey sightings and several nesting attempts between 1 June and 15 August at numerous lakes, res- ei"voirs, and rivers from nearly eveiy region of the state since 1990. These sightings represent the first widespread effort by Ospre>'s to expand their range in Utah. This paper reviews histor- ical Osprey breeding territories in Utah, sub- secj[uent population declines, and current Ospre>' population and range expansion in Utah. ' Geographic History of the Osprey in Utah Nesting Ospreys have been reported from 4 geographical areas of Utah (Fig. 1): the Wasatch Front, Uinta Mountains, southern pla- teaus, and Green River (Table 1). Accounts of early ornithologists, naturalists, and egg col- lectors indicate the Osprey was a regular sum- mer resident and breeder in Utali. Allen (1872) found them along the Great Salt Lake marshes west of Ogden, and Henshaw (1874) saw them at Utah Lake near Proxo. Neither discussed nest obsenations in these areas, but R. G. Bee (unpublished ornithological notes) mentioned that Ospreys formerly nested along the shores and tributaries of Utah Lake (Fig. 1; Table 1, region A). Other records were for the Uinta Nh)imtains (Fig. 1; Table 1, region B). J. D. Da>'nes (unpub- lished ornithological notes) described the repeated use of an Osprex nest from 1915 to 1938 on the Weber River, 20 km east of Oak- ley, Sunniiit Count)'. Also, Ha\\\ard (1931) ^Department ofGeograpli), Brinhain Voiini; Uiiivursity, Frovo, UT 84(302-.5.526. 150 1996] Osim{i:y Status in Utah 151 Fig. 1. Known historical distribution of nesting Ospre>s in Utah: A, Wasatch Front: B. western Uinta Mountains: C, southern plateaus; D, Green River. surveyed (8 July-21 August 1930) birds in the western Uintas where Summit, Duchesne, and Wasatch counties converge. While not giving actual locations, he said a few Ospreys nested in the Mirror and Tryol [sic] lakes region. Bee and Hutchings (1942) specifically cite Mirror and Trial lakes as having Osprey nests. Twomey (1942: 382) visited an occupied nest bet\veen 16 and 20 July 1932 at the north end of Mirror Lake, "Wisatch County." This nest was actually 152 Great Basin Naturalist [Volume 56 Historical pairs Cu rent pairs Unknown 1 5-8 0 2-4 8-11 6-8 20-25 Table 1. Nesting populations of Ospreys in Utali. Region A: Wasatch Front B; Uinta Mountins C; Southern plateaus D: Green River in Duchesne County and perhaps the same nest Hayward et al. (1976) referred to when they hsted Wasatch County as a former nest- ing area. R. G. Bee (unpubhshed ornithological notes) also cited single pairs at Fish, Scout, and Lily lakes in the western Uintas. On 23 May 1945, Bee recorded that a game warden in Duchesne informed him of 2 pairs at Moon Lake and another pair at an unidentified Uinta lake. Other early observers of Osprey nests in Utah include Wolfe and Cottam (Hayward et al. 1976), who, along with Bee and Hutchings (1942), saw Ospreys nest at Fish Lake (not the Uinta Mountains lake with the same name), Sevier County, beginning in 1928 (Fig. 1; Table 1, region C). On 18 July 1936, R. G. Bee (un- published ornithological notes) visited the Fish Lake nest. A local rancher told him Ospreys had used that particular nesting site for at least 20 years. Behle et al. (1958) noted a pair of Ospreys in southwestern Utah at Navajo Lake, Kane County, on 17 and 18 June 1950 (Fig. 1; Table 1, region C). This particular territory (and an additional site at nearby Panguitch Lake, Garfield Count)') has been used regularly since Behle's discovery (Eyre and Paul 1973, Salt Lake Tribune, 13 August 1978, Walters 1981, Anonymous 1989). Ospreys also nested along the Green River, northeastern Utah (Fig. 1; Table 1, region D). On 23 and 24 July 1959, C. M. White and C. Bosley (White and Behle 1960) located 2 nests along this river in Horseshoe Canyon, Daggett County. Both nests contained 2 young. An additional nest was discovered on the Green River in Uintah County by M. Horton in June 1974 (Behle 1981). White (1969) suggested the total population of Ospreys nesting along the Green River probably consisted of 6-8 pairs. Status Historical Events Although numerous records of Ospreys nest- ing in Utah exist, these birds have apparently undergone 2 separate declines. The 1st decline involved Ospreys nesting along the Wasatch Front (Fig. 1). During winter 1848-49, depreda- tions upon livestock, poultry, and grain led to a much-publicized contest to kill the 'wasters and destroyers ' (Arrington 1958: 51). Hundreds of mammalian predators and thousands of raptors were killed (Arrington 1958). Ospreys would have been on their southeiTi wintering grounds during this assault on local predators, but the incident suggests that early pioneers in Utah treated all carnivores and birds of prey with contempt. Other similar hunts followed, and 40 years later the Utah Legislatiu'e implemented a law awarding bounties for the killing of preda- tors (Rawley 1985). Rewards were available for several species of fish-eating birds including Ospreys. The destiiiction that this bounty in- flicted upon fish-eating birds in the name of "consenation" was significant and is vi\'idly described by Pritchett et al. (1981). The attitudes of early residents toward pred- ators, coupled with laws encouraging their destruction, may have led to the Ospreys extirpation from the Wasatch Front (Fig. 1; Table 1, region A) around the turn of the cen- tury. In 1935, R. G. Bee (unpublished ornitho- logical notes) speculated that himian persecu- tion caused the abandonment of Osprey nests near Utah Lake. Bee did not record when these Ospreys disappeared, but his manner of re- flection on their absence suggests the loss occurred well before his 1935 notation. The 2nd period of Osprey decline occurred in the western Uinta Mountains (Fig. 1). The Uinta Mountain nests that Da\nes, Bee, Hay- ward, and Twomey reported \\'ere observed before, but apparently not after, the 1950s and 1960s when Osprey colonies along the eastern seaboard were decimated b\' organochlorine compounds (Palmer 1988, Poole 1989). .\ldiough the impact of synthetic agricultural biocides upon Ospreys in Utah is unknown, Ospreys in other areas of the \\ estern United States were generally less affected In enxironmental con- taminants than eastern populations (Poole 1989, White 1994). Another possible reason for the decline of Ospreys in the Uinta Moimtains is indiscrimi- 1996] OsPREY Status in Utah 153 nate shooting resulting from a hostile attitude by local residents (Bee unpublished ornitho- logical notes, Twomey 1942). Man> Ospreys were formerly shot at northern Utah fish hatcheries during spring migrations (White 1969, Ha\'\vard et al. 1976), and some of these casualties could ha\e been local breeders. HayAvard et al. (1976: 66) recorded the Osprey was "formerly a sparse but regular summer resident in Utah; now greatK' reduced in numbers and considered to be rare and endangered." They preface their discussion of birds in Utah by stating they have included all records concerning rare species in the state. However, they cited Ospreys only in the west- ern Uintas and records for Fish Lake in Sevier County. They did not include information on the 1 or 2 pairs nesting in the Navajo Lake- Panguitch Lake area, southern Utah, or the pairs at Flaming Gorge Resen^oir, northeast- ern Utah. In the same year (1976) that Hay- ward et al. (1976: 66) described the Osprey as "gi-eatly reduced," more Osprevs (6 pairs) nested at Flaming Gorge Reservoir (Wagner 1977, Salt Lake Tribune, 13 August 1978) than had been recorded in any particular year in the western Uinta Mountains. Current Events Flaming Gorge Dam on the Green River was completed in 1964 and created a narrow, 150-km-long reservoir on the Utah-Wyoming border The Osprey population here remained relatively stable until the late 1970s and 1980s when an increase was noted (Behle 1981). Crawley and White (1989) found 21 pairs and 1 trio of Ospreys at Flaming Gorge in 1989. Of these, 15 pairs succeeded in fledging 37 young. Osprey numbers at Fish Lake in Sevier County increased from 2 pairs in 1989 (Anony- mous 1989) to 6 in 1993 (B. Lowiy, U.S. Forest Service, personal communication). Addition- ally, 1 or 2 pairs now nest 3 km away at John- son Valley Reservoir (E Wagner personal com- munication). Other current Osprey nest sites at traditional waters in Utah include 2 pairs in the Panguitch Lake-Navajo Lake area of southern Utah (Anonymous 1989). In 1990 a pair of Ospreys nested at Tropic Reservoir, Garfield County (Sorensen 1990). This site is 20 km east of region C (Fig. 1, Table 1) and should be regarded as a geographical extension of that area. In 1994 a pair of Ospreys constructed a nest near the Midway fish hatcheiy, Wisatch County (Fig. 2A). In 1995 a 2nd pair built a nest 2 km away at Deer Creek Reservoir on a 5-m-high artificial platform erected for Ospreys (Fig. 2B). Deer Creek Reservoir and the adjacent Midway fish hatch- en' have been fre(juented by Ospreys during spring migrations for many years (Behle and Perry 1975). Additional Osprey nesting attempts in 1995 include 1 pair at Jordanelle Resen'oir, Wasatch C^ounty (Fig. 2C), and another pair near Highland, Utah County (Fig. 2D). Incu- bation behavior at the latter site was observed for approximately 2 wk before strong winds destroyed the nest. This site was possibK- the first Osprey nest along the Wasatch P>()nt in 80-100 yr The origin of Ospreys colonizing new waters in Utah is currently unknown, but their reluc- tance to disperse more than 125 km from their natal sites is well documented (Henny 1986, Poole 1989). Reproduction for nests at Flam- ing Gorge Reservoir is generalK' high (Craw- ley and White 1989), and considering the Ospreys pronounced philopatr>', one might expect that Ospreys at new locations in Utah derive from this local population. While high productivity has augmented the Osprey popu- lation on Flaming Gorge, the frequency with which Ospreys are being witnessed in Utah is too great to be the sole result of dispersal from tliat resei"voir Moreover, if Flaming Gorge were the primaiy source of Ospreys pioneering new waters in Utah, one would expect lakes and rivers near that resei"voir to be the initial areas of range expansion. This has not been the case. A more plausible source of Ospreys attempt- ing to colonize nontraditional (and abandoned historical) waters in Utah is fi^om spring migi-ants stopping short of their natal territories farther north. Osprey populations in Idaho and Wyo- ming number in the hundreds of pairs (Henny 1986, Poole 1989), and Osprey counts made at several migration points in the West ha\e bur- geoned since 1983 (Hoffman et al. 1992). Fur- thermore, migrating subadult Ospreys are knowTi to linger sometimes and even remain at pro- ductive foraging sites south of their traditional breeding grounds (Swenson 1981, Poole 1989). These lingering individuals may represent young adults without an established histon- of breeding elsewhere. If more nordierly populations constitute the primary source of Ospreys currently pioneer- ing nontraditional waters in Utah, this long- 154 Great Basin Naturalist [Volume 56 Fig. 2. A. Ospiev nest, Midway fish hatchery; B, ()spre>' nesting plationn and nest. Deer Creek Resenoir; C, Osprey >st, Jordanelle Reservoir; D, Osprey nest and inenbating adnlt near Highhmd, Utali. 1996] OsPHEY Status in Utah 155 distance dispersal is a recent phenomenon and possibK indicates a satnrated hreedint!; popn- lation in the northern Intennonntain West. A cnrrent, qnantitati\e evahiation ot Osprc)' pop- nlations in Idaho and Wyominu and extensive handine; efforts in these states conld help determine if this speculation is correct. Until snch a project is nndertaken, the ori,' le\'els, the difference in feeding rates between species was proportionally higher (10%). At high turbidity levels (> 20 NTU) trout predation rates were relatively insensitive to prey size. However, shiner continued to consume more, larger pre>' at the highest turbidibi' levels. These results indicate that Lahontan redside shiner may be superior to Lahontan cutthroat trout as zooplankton predators at high turbidity levels, and may explain the recent success of shiner in Summit Lake. Key words: Daphnia, Laliontan cutthroat front, Oncorlniichus clarki henshawi, Laliontan redside shiner Hichardso- nius egregius, plankfivorij, predation, size selectivity, turbidity. The Lahontan cutthroat tiout {Oncorhynchus clarki henshawi) is an inland subspecies endemic to die physiographic Laliontan basin in noilhern Nevada, eastern California, and southern Ore- gon. These ti^out were once widespread through- out the basins of Pleistocene Lake Lahontan (USFWS 1995). Currently, they occupy < 1% of their former lacustrine range and 11% of their former stream habitat within the native range (USFWS 1995). Listed as endangered in 1970, the fish was subsequently reclassified as threatened in 1975. This facilitated manage- ment and peniiitted regulated angling (USFWS 1995). Summit Lake is located in the Summit Lake Paiute Indian Reservation in northwestern Humboldt County Nevada (41 °N latitude 119°W longitude), at an elevation of 1828 m. Formed by a landslide about 20,000 years ago, Summit Lake is relatively shallow (maximum depth 12 m) and has historically been subject to high turbidity levels during summer months from suspended algae and silt (LaRivers 1962). It contains the most secure remaining lacus- trine population of Lahontan cutthroat trout, and no other salmonids occur in the basin (Cowan and Blake 1989, Valeska 1989). Other lacustrine populations are either maintained by artificial stocking or are subject to higher levels of harvest and disturbance. Conserx'ation of this population is compelling, and it has been identified as important for recover)' of the subspecies (USFWS 1995). Cutthroat trout spawning runs at Summit Lake have generally declined since the late 1970s (Cowan and Blake 1989). Collection of roe during the 1960s and 1970s and excessive loss of spawning habitat in Mahogany Creek from livestock overgrazing (Cowan and Blake 1989, Vinyard and Winzeler 1993) ha\e been blamed. However, coinciding with the decline in trout, Lahontan redside shiner {Richardso- nius egregius) also increased in abundance in the lake, suggesting a competition effect. Redside shiner are native to the Great Basin, but they do not occur naturally in Sum- mit Lake. Origins of the present shiner popu- lation in the lake are unknown, but the>' have been used frequently as live bait. Lahontan redside shiner feed on drift in streams and are zooplanktivorous in lakes (Vinyard and Winzeler 1993). Laboratory observations suggest they 1 Dcpartim-nt of Biology, Universih' of Nevada. Reno, Nevada 895.57-001.5. 157 158 Great Basin Naturalist [Volume 56 may also prey on larval trout (Vinyard and Winzeler 1993). Analysis of stomach contents suggests that Lahontan cutthroat trout and Lahontan redside shiner probably consume similar foods both in Summit Lake and in Mahogany Creek, the primaiy spawning tribu- taiy for trout from Summit Lake (Vinyard and Winzeler 1993). Both species consume drift in the stream, and mostly amphipods in Summit Lake (Cowan and Blake 1989). In contrast, simi- larly large Lahontan cutthroat trout in Pyramid Lake are piscivorous (USFWS 1995). Because most fish species depend on vision to locate prey (Hobson 1979, Guthrie 1986), it is possi- ble that high turbidit)' in Summit Lake limits the visibility of prey and impedes the ability of trout to catch redside shiner and other large prey. Our experiments compared the relationships of feeding rate, turbidit)', and prey size for Lahontan cutthroat trout and Lahontan red- side shiner, with the primary focus being to examine the relative performance of both species under various turbidity le\'els. Methods and Materials Lahontan redside shiner were captured fiom Mahogany Creek, Humboldt County, Nevada, and transported to the University of Nevada. Lahontan cutthroat trout from the current Pyramid Lake stock were acquired from the Lahontan National Fish Hatcheiy, Gardner- ville, Nevada. Although the historical origins of the existing Pyramid Lake stock are mixed. Summit Lake fish were heavily planted into Pyramid Lake for a number of years, and they likely constitute the dominant component of the population (USFWS 1995). Fish were housed in 19-L tanks and acclimated to local water conditions for at least 3 wk prior to experiments. Experiments were conducted in a secluded section of a greenhouse at the University of Nevada. The experimental protocol was simi- lar to that emploved b\' Vinvard and Winzeler (1993) and Li et'al. (1985). Visual isolaticm of experimental tanks was ensured by opaque black polyethylene sheeting (10 mil, 2.5 m high), which enclosed all sides of the experi- mental area and controlled external light. Temperatures ranged between 12 °C and 17 °(' during the experiments, and diel variation never exceeded 4°C, a range easih' tolerated b\' both species. Lighting was provided by a bank of three 56-watt fluorescent tubes con- trolled by an automatic timer (10L:14D). Light intensity' at the water surface averaged 93 jnE m~- S~^. An airstone in the center of each of four 38- L aquaria provided aeration and kept turbidity in suspension. Turbidity (nephelo- metric turbidity units, NTU) was measured with an HF Instruments Model DRT 15 tur- bidimeter Six turbidity levels (3.5, 6, 10, 20, 22, and 25 NTU) were produced using sus- pensions of bentonite. Bentonite concentra- tions (mg/L) were significantly correlated vdth measured turbidity (NTU = 2.583 + 0.162 B, r- = 0.99). This material is nontoxic and remains in suspension for long periods. Feeding rates were determined for fish ex- posed to single-sized groups ofDaphnia magna at each turbidit\' level. Laborator)'-reared D. magna were sorted into 3 size groups using a dissecting microscope: 1.7 mm, 2.2 mm, and 3.0 mm (top of head to base of tail spine, ± 0.3 mm). Before each feeding trial, a single fish was placed into each experimental tank and allowed to acclimate for 24 h. A group of 200 Daphnia were introduced into the tank and the fish allowed to feed for 2 h. Fish were then removed and the water and remaining prey siphoned through a 363-micron mesh net. Prey retained on the net were counted to deter- mine consumption rates. This procedure was repeated for each of the 3 prey size classes and 6 turbidity levels with 4 fish from each species, yielding a total of 144 feeding trials. Fish used in the feeding trials ranged from 70 mm to 93 mm SL. Analysis of variance and lin- ear regression were used to assess the effects of fish species, prey size, and turbidity level on predation rates. Results An analysis of overall predation rates for both fish species consuming all prey sizes (Figs, la, lb) indicates that feeding rates varied inversely with turbidit)' (multiple regression, F = 1894, P < 0.001) and between fish species (F = 28.4, P < 0.001), and that larger prey gener- ally were consumed at greater rates (F = 38.3, P < 0.001). Significant results were observed for both the species* NTU and species '^daph- nia size interaction terms, indicating that the 2 (ish species differ in their responses to these 2 \ariables. Lahontan redside shiner consumed 1996] TiHiJii:)! IT Effects on Fish Feeding 159 significantly more prey than l^ahontan cut- throat trout. At the lowest turhidit) level (3.5 NTU), approximately 909^ of all prey were consumed l)> both fish species. Ilcmever, even small increases in turbidity reduced predation rates. This decrease in predation with tmbid- ity was strongK linear and there was no indi- cation of a minimum xalue ha\'ini2; been reached b> 25 NTU. At that turbidity level, predation rates declined bv approximately 80% for trout (Fig. la) and by 60% to 80% for shiner (Fig. lb), depending on prey size. Predation rates for trout were significantly affected by prey size and turbidit\' (multiple regression F = 2.67, P = 0.009 for prey size; F = 35.1, P < 0.001 for turbidit>). Similar results were obsened for shiner (multiple regression F = 6.54, P < 0.001 for prev size; F = 27.15, P < 0.001 for turbidit> ). At higher turbiditx levels, differences in performance of the 2 fish species became most apparent. At turbidity levels of 20 NTU or more, prey of all sizes were consumed at virtually equal rates by Lahontan cutthroat trout (Fig. la). In contrast, Lahontan redside shiner showed increasing predation on 3-mm prey relative to the smaller sizes at high turbidity levels (Fig. lb), and shiner showed the greatest differ- ences in predation rates between prey of dif- ferent size at the highest turbidity levels. Lahontan cutthroat trout exhibited the oppo- site trend, with greater differences in preda- tion rates between prey of different sizes at low turbidity levels. Discussion Foraging behaxior and efficiency are affected by local visibility. Many workers have demon- strated reduced effectiveness by visual preda- tors at elevated turbiditv (Vinyard and O Brien 1976, Li et al. 1985, Barrett et al. 1992, Gregon- and Northcote 1993). Sigler et al. (1984) found that chronic high turbidity impedes growth and increases mortality of steelhead (O. mykiss) and coho salmon (O. kisiitch). Evidence sug- gests that high turbidit\' or low light intensity reduces predator selectivity because relative differences in prey-detection distance for dif- ferent sizes of prev are reduced (Vinvard and O'Brien 1976, Gregoiy and Northcote 1993). Gregory and Northcote (1993) observed log- linear declines in reactive distance with increased turbidity in chinook salmon (O. tshaii'ijtscha). Our results demonstrate that turbidity reduces predation rates for all prey sizes for both Lahontan redside shiner and Lahontan cutthroat trout. Larger prey were generally consumed with greater frequency, although this frequency \'aries with turbiditx' and fish species. The effect of prey size was most con- sistent for Lahontan redside shiner. These fish consumed more large (3.0 nun) pre\ at all tnr- bidit> levels than did Lahontan cutthroat trout (Figs, la, lb). In contrast, prey size had little effect on the relative numbers of prey of each size consumed by trout at turbiditv levels of 20 NTU or above'(Fig. la). Redside shiner also consumed more pre\ oi all 3 sizes combined over all turbidity levels. For all prey sizes combined, shiner consimied approximately 3% more prey than Lahontan cutthroat trout at low turbidity levels and approximately 10% more at high levels (Figs. la, lb). Angradi and Griffith (1990) found pre- dation by rainbow trout (O. tnyki.s.s) to be more selective for large prey in clear water, \\hereas selectivity was reduced in elevated turbidit\'. Similar effects on prey selection under reduced visibility conditions have been obser\ed in bluegill sunfish {Lepoinis macrochirus). Under low-light conditions bluegill sunfish consumed fewer zooplankton but proportionally more large individuals (Miner and Stein 1993). Neither trout nor shiner have been shown explicitly to possess adaptations that might en- hance their effectiveness as foragers in turbid waters. However, fish that feed nocturnally, such as walleye {Stizostedion vitreiim), may perform equally well in either clear or turbid waters (Vandenbyllaardt et al. 1991). \\yie\'e have higher densities of retinal cells and also develop scotopic vision earlier in life in comparison to salmonids (Vandenbyllaardt et al. 1991, Borgstrom et al. 1992, Ilurber and Rylander 1992). Such species-specific fac- tors may contribute to differences in \ isual performance. Behavioral responses offish to turbidit> ma\ also affect their feeding abilities or rates. In laboratoiy experiments, golden shiner {Nuteini- gonus cnjsoleucas) showed increased flight responses with increased turbidity (Chiasson 1993). Juvenile chinook salmon apparenth experienced reduced predation from piscivo- rous birds and fishes at elevated turbidit> le\'- els (Gregory 1993). During our experiments, redside shiner were observed to search faster 160 Great Basin Naturalist [Volume 56 100 3.0 mm LAHONTAN CUTTHROAT TROUT r^ i (Oncorhynchus clarki henshawi) 10 15 20 TURBIDITY (NTU) 25 30 100 80 z liJ 2 60 I- §40 oc UJ °-20 2 mm LAHONTAN REDSIDE SHINER (RIchardsonius egreglus) 10 15 20 TURBIDITY (NTU) 25 30 Fig. 1. Mean percent prey consumed in relation to turbiditx. Upper pane! (a) shows results from feeding trials with Lahontan cutthroat trout {Oncorlujnchus clarki henshawi), and lower panel (b) shows results from Lahontan redside shiner (Ricliarclsoniiis egrcgiits). Four fish of each species were exposed to prey of a single size for 2-h feeding trials. DapJuiia mapui prey sizes are as indicated. Vertical bars indicate 1 standard deviation. and more widely at higher turbidity. Elevated turbidity may have provided greater visual i.solation and promoted greater mobility b>' predators as suggested by Confer et al. (1978) and Gradall and Swenson (1982). Increased activity may have compensated for reduced visual effectiveness, resulting in larger search volumes for shiner than for trout. In a study of brook trout {Salvclinus fontinalis) and creek chub {Seinotihis atroinaciilatiis), Ciradall and Swenson (1982) found creek chub to be less affected by turbidity than brook trout. They suggested such differential effects may explain local disparities in fish density. High turbidit}' in Summit Lake may decrease reactive distance and search \ olinne unequally for shiner and trout. This ma\' differentially reduce the probabilitx of successful prey cap- ture and could produce altered pre\' selection patterns under different turbidity' conditions. Although our results are generally similar to those shown for other fishes (Vinyard and O'Brien 1976, Berg and Northcote 1985, Li et al. 1985), we document higliK significant differences between potentially competing fish species. Because Lahontan cutthroat trout and Lahontan redside shiner consume the same prey in Siunmit Lake, competition tor food 1996J TiKBiDiiT Effects on Fish Ffkdinc Ibl may exist. Our results sut^^est that iu elevated turbidit)' eouditious Lahontau redside shiner nia> be a better competitor for food than Lahontan cutthroat trout. A factor contribut- inu; to the success of Lahontan redside shiner in Summit Lake ma> be that their predation rates are higher than those of cutthroat trout at elevated turbidity levels. Ackno\vled(;ments We thank Larn' Marchant of the Lahontan National Fish Hatchery and Alice Winzeler for providing fish, and Louis Christensen for assistance in setting up the experimental appa- ratus. We thank R. S. Gregory and 2 anony- mous reviewers for helpful suggestions for this manuscript. The University of Nevada Depart- ment of Biolog\' undergraduate tliesis commit- tee facilitated completion of this project. Literature Cited Angradi, T. R., and J. S. Griffith. 1990. Diel feeding chronology and diet selection of rainbow trout {Oncorhynchiis mijkiss) in the Henry's Fork of the Snake River, Idaho. Canadian Journal of Fisheries and Aquatic Sciences 47: 199-209. Barrett, J. C, G. D. Grossman, and J. Rosenfeld. 1992. Turbidity-induced changes in reactive distance of rainbow trout. Transactions of the American Fish- eries Society 121: 437-443. Berg, L., and T. G. Northcote. 1985. Changes in territo- rial, gill-flaring, and feeding behavior in juvenile coho salmon (Oncorhijnchits kisiitch) lollowing short term pulses of suspended sediment. Canadian Jour- nal of Fisheries and Aquatic Sciences 42: 1410-1417. Borgstrom, R., a. Brabrand, and J. T. Solheim. 1992. Effects of siltation on resource utilization and dynam- ics of allopatric brown trout, Salmo tnittci, in a reser- voir Environmental Biology of Fishes .34: 247-255. Chiasson, a. 1993. The effect of suspended sediments on ninespine stickleback, Piingitiiis pungitiii.s, and golden shiner, Noteinigonits chrysoleiiccis, in a current of varying velocitv. Environmental Biologv of Fishes 37; 283-295. Confer, J. L., G. L. Howick, M. H. Corzette, S. L. K.\mer, S. Fitzgibbon, and R. Landesberg. 1978. Visual predation by planktivores. Oikos 31: 27-37. Cowan, W, and R. Blake. 1989. Fisheries management services contract #CTH50913089, annual report. Report to Summit Lake Paiute Tribe. 31 pp. Gradall, K. S., and VV. A. Swenson. 1982. Responses of brook trout and creek chubs to turbidity. Transac- tions of the American Fisheries Society 111: 392-395. Gregory, R. S. 1993. Effects of turbidity on the predator avoidance behavior of juvenile chinook salmon (Oncorhynclni.s Ishawytsclm). Canadian Journal of Fisheries and Aejuatic Sciences 50: 241-24(i. Gregory, R. S., and T. G. Northcote. 1993. Surface, planktonic, and benthic foraging by juvenile C:hi- nook salmon (Oncorhynchus tsliauylscha) in turbid laboratory conditions. Canadian Journal of Fisheries and Aquatic Sciences 50: 2.3.3-240. Guthrie, D. M. 1986. Role of vision in fish behavior. Pages 75-113 in T. J. Pitcher, editor, The behavior of teleost fishes. Groom Helm, [..ondon. HoBSON, E. S. 1979. Interactions between piscivorous fishes and their prey Pages 231-242 /;i R. H. Stoud and H. Clepper, editors, Predator-i)re\' systems in fisheries management. Sport Fishing Institute, Wash- ington, DC. HUBER, R., AND M. K. Rylandeh. 1992. yuautitative his- tological study of the optic nerve in species of min- nows (Cyprinidae. Teleostei) inhabiting clear and turbid water Brain Behavior and Evolution 40: 2.50-255. LaRivers, I. 1962. Fishes and fisheries of Nevada. .Nevada State Fish and Game Commission, Reno. 782 pp. Li, K. T, J. K. Wetterer, and N. G. Hairston, Jr. 1985. Fish size, visual resolution, and pre\- selectivit^'. Ecology 66: 1729-1735. Miner, J. G., .\nd R. A. Stein. 1993. Interactiv e inlluences of turbiditv' and light on larval bluegill (Lepomis inacrachirus) foraging. Canadian Journal of Fisheries and Aquatic Sciences 50: 781-788. Sigler, E W. and J. W. SiGLER. 1987. Fishes of the Great Basin: a natural histon; University of Nevada Press, Reno. 425 pp. SiGLER, J. W, T C. BjORNN, AND E H. EVEREST 1984. Effects of chronic tiu^bidity on density- and growth of steel- heads and coho salmon. Transactions of the .Ameri- can Fisheries Society 113: 142-150. U.S. Fish and Wildlife Service. 1995. Lahontan cut- throat trout, Oncorhynchus chirki henshaui. recov- eiy plan. Portland, OR. Valeska, J. R 1989. Sunnnit Lake lacustrine studv tech- niques. Unpublished manuscript. Summit Liike Paiute Tribe, Winnemucca, NV Vandenbylla.\rdt L., E J. Ward, C. R. Brakev eli. and D. B. McIntyre. 1991. Relationships between tur- bidity, piscivoiy, and development of the retina in juvenile walleyes. Transactions of the .American Fisheries Society 120: 382-390. Vinyard, G. L., and W. J. O'Brien. 1976. Effects of light and turbidity on the reactive-distance of bluegill {Lepomis inacrochirns). Canadian Journal of Fish- eries and Aquatic Sciences 33: 284.5-2849. Vinyard, G. L, and A. L. Winzeler. 1993. Results of investigations at Summit Lake. Report to Summit Lake Paiute Tribe. 62 pp. Received 12 June 1995 Accepted 19 January 1996 Great Basin Naturalist 56(2), © 1996, pp. 162-166 POGONOMYRMEX OWYHEEI NEST SITE DENSITY AND SIZE ON A MINIMALLY IMPACTED SITE IN CENTRAL OREGON Peter T. Soule' and Paul A. Knapp- Abstract. — Little is known about the basic characteristics of the western hanester ant {Pogonoimjrmex oictjheei) in the absence of anthropogenic disturbances. We examined the role of P. oictjhcei as an agent of disturbance in an area of semiarid \ egetation in central Oregon known as the Horse Ridge Research Natural Area (HRRNA) that has been largely free of livestock grazing and other significant anthropogenic influences for over 23 yr. We determined densit\' and size characteristics of nest sites and estimated total area cleared by P. owyheei activities on HRRNA. From random sampling of twenty-five 0.04-ha plots we found a mean nest density /standard eiror of 1.6 (±0.16) nests/0.04 ha. Mean area cleared per nest site was 4.8 m-, which results in an estimated banen area of 46,080 m- on the 240-ha HRRNA. Comparing our findings to others on P. owyheei and P. occidentalism we foimd nest densit} and mean cleared area to be in the middle range of reported obsenations under a \ariet> of land-use influences. The literature suggests that moderate disturbance ma\' increase nest site densit>', but little relationship exists between distinbance histon' and mean size of nest sites. Key words: Pogonom\rmex ow>heei, western harvester (uits. nest density, nest size, vegetation clearing. Western hai^vester ants are a major compo- nent of arid rangeland ecosystems in the United States. Because of the combined effects of seed predation, seed dispersal, and vegetation re- moval, har\'ester ants are "keystone species," meaning their effects on vegetation structure and dynamics exceed expectations given their density and biomass (Holldobler and Wilson 1990: 616). The most visible impact of har- vester ant activities is vegetation clearing around their nest sites. Although the size of the cleared area, or disc, varies, Pogonomijnnex harvester ants have the capacity to cut annual plants surrounding their nest sites at rates of over 200 million plants/ha/yr (Clark and Comanor 1975). While much of the plant bio- mass cut is not consumed by the ants, it reduces the total volume availal)le for con- sumption b)' livestock and other grazers (Willard and Crowell 1965). Range managers have viewed Pogunonnjnnex as pests that need to be controlled, giving the ant both economic and ecological importance in arid rangelands (Wight and Nichols 1966, Cole 1968). Because of the paucity of undistiubed areas in the semiarid West, little is known about the basic characteristic of P. owyheei nest sites in the absence of anthropogenic disturbances. The primaiy objectives of this stud) are (1) to determine the densit\ and size characteristics of P. owyheei nest sites and (2) to estimate the total area denuded b\' clearing and foraging acti\'ities of P. owyheei within a largely undis- turbed semiarid ecosNstem. Study Area The Horse Ridge Research Natural Area (HRRNA) is a 240-ha exclosure 31 km south- east of Bend, Oregon, managed by the Prine- \'ille District, Bureau of Land Management (BLM). The natural area was established in 1967, and a surrounding fence was completed in 1974. The exclosure ranges from 1250 to 1430 m elevation over rolling topography of Columbia Basalts (Anonymous 1972). Direct human impacts on the site are minimal as there is only occasional use by hunters and naturalists, and fire suppression is not active (HaKorson 1991, R. Halvorson personal com- munication 1995). The fence has kept the area free of livestock grazing since 1974, but before its establishment the area apparently received minimal domestic animal grazing pressure because of a lack of a permanent water source to attract animals (Anon\'mous 1972) and the distance from well-traveled public roads (Baldwin 1974). AdditionalK, the abundance on HRRNA of threadlea\ ed sedge (Carex fiU- Jolia). a species that has been shown to decline 'ncp;irtniciiliir(;ciii;rapli> .mrI I'l.iiiiiiim. Aiip.ihuliiaii State L iii\rrsit\, lioone. NC 28fi07 ^Dcpartiiii'iil 111 f ;ccii;rapliy, Ceoiyia Slalr UiuMMsit), Allantu. C..\ :H)M)^. 162 199(i] r. OwiiiEEi Nest Densih and Size 163 because of o\'ergrazing; in the central Oregon sagebrush steppe, and the absence of clieat- grass {Broinus tectonim) suggest a minimally disturbed site (Anon>'nious 1972, personal ob- servation 1995). Vegetation on IIHRNA is classified as the western juniper/big sagebrush/threadleaved sedge community {Jiiniperiis occidentalisi Artemisia tridciitata/Carcx filijolia) (Franklin and D\'rness 1988). Less common but present species are bluebunch wheatgrass {Agropyron spicatwn), Idaho fescue {Festuca idahoensis), junegrass {Koeleria cristata), and horsebrush {Tctradijniia glahruia) (Anonymous 1972). HRRNA climate is dominated by winter precipitation. Over half the annual 31 cm falls as snow. Mean temperatures at Bend range from -0.6°C in Januaiy to 17.7°C in Juh' (Karl et al. 1990). Soils on our study plots are entirely within the Stookmoor-Wesbutte complex soil series (USDA-NRCS in press). This soil series is found on approximately 85% of HRRNA. A t> pical soil profile is represented by a surface layer of mixed ash and loamy material approxi- mately 15 cm thick, and a pale brown, sandy loam subsoil 46 cm deep overlying bedrock. Percentage of organic matter in the topsoil is l%-2% and 0.5%-2% in the subsoil (USDA- NRCS in press). Besides P. owyheei, there is disturbance pressure on HRRNA from grazing activities of herbivores and granivores such as Rock-)' Moun- tain mule deer {Odocoileus hemioniis Jitnniomis), badger {Taxidea taxus), and cottontail rabbits {Syhilagus niif(dli) (Gashwiler 1972, personal obsei-vation 1995). BLM records on HRRNA report no outbreaks of intense herbivory or episodes of pathogens causing severe plant losses in the last 20 years (R. Halvorson per- sonal communication 1995). Methods In roughly the center of HRRNA, a 19.6-ha permanent grid was established by (kislnviler (1977) for use in an ecological study in 1972. Stations on the 12x12 grid are marked !)>• re- bar stakes and spaced 40.2 m apart. Using this grid, we randomly selected 25 stations and established 0.04-ha circular plots from the lebar- marked center points for a total sample area of 1 lia. We tallied and measured each acti\'c P. owyheei nest site within each plot. We placed line transects over the center of each nest site and measured the cleared disc area in north- south antl east-west directions. The edge of each disc was determined i)y the intersection of any perennial with the N-S or E-W transect lines. Results There were 40 active P. owyheei nest sites in our 1-ha sample. We found nest sites on 23 of the 25 circular plots, and the maxinumi number of nest sites was 3 pei- 0.04 ha. Mean nest density/standard error was 1.6 (±0.16) nests/0.04 ha. Characteristics of the cleared discs are shown in Table 1. Assuming a circu- lar shape, the mean area cleared per nest site is 4.8 m^. Factoring in the nest density results in an estimated barren area of 192 m^/ha, or 1.92% of the total land area of the permanent grid. If the influences oi P. owyheei are consis- tent throughout the 240-ha HRRNA, then ant foraging and plant cutting surrounding a total of 9600 nest sites should leave approximateK 46,080 m- of barren land on the 240-ha site. DlSCUSSI(3N The premise of this article is to pro\ ide information on P. owyheei nest site densitx Table 1. Characteristics of P owyheei nest sites on HRRNA. Standard Discs Mean Median M, aximum M inimum de\iation .V (cm) (cm) (cm) (cm) (cm) N-S diameter 241.1 207.5 740 60 144.4 40 E-W diameter 254.6 220.0 670 68 1.56.S 40 164 Great Basin Naturalist [Volume 56 Table 2. Pogonomyrtnex oivyheei and P. occidentalis nest ren area due to P. ouyheci and P. occidentalis aeti\ ities repor site densities and mean size of nest site and estimated Iiar- ted in the literature. Source Pugonomynnex Nest site State species densit\7h;i Nest site \ sr/,e 111 Ill- Estimated barren Stiidv' site area % disturbance Dominant \egetation SiiarpandBan(19«)) Idah( ouyheei-' Sharp and Barr (I960) Idaho owijheei^ Sharp and Barr (1960) Idaho owyheei'^ Willard and Crowell (1965) Oregon oivyheei Wight and Nichols (1966) Wyoming occidentalism Rogers and Lavigne (1974) Colorado occidentalis Rogers etal. (1972) Rogers et al. (1972) Rogers et al. (1972) Colorado occidentalis Colorado occidentalis Colorado occidentalis Clark and Comanor (1975) Nevada occidentalis Sneva (1979) Sneva (1979) Sneva (1979) Oregon Oregon vyheei vyheei Oregon oicyhcei Coffin and Lauenrotli (198S) Colorado occidentalis CofFin and Lauenroth (1990) Colorado occidentalis Nowaket al. (1990) Idaho owyheei Nowak et al. (1990) Idahi owyheei 40 0.8 6.0 "misused/ depleted" Atriplex nuttallii / Halogeton glomeratus 9 1.3 3.7 "vigorous stand" Atriplex nuttallii 12 nr'' nr not discussed Ariplex confertifoliei 49-74 22.5 11-17 not discussed Bromiis tectorwn nr 65.7 m- lightK' grazed'^ Atriplex nuttallii 23 1.2 0.3 ungrazed for 30 years Buchloe dactyloides 1 Bouteloua gracilis 28 0.7 nr lightly grazed Buchloe dactyloides 1 Bouteloua gracilis 31 0.4 nr moderate grazing Buchloe dactyloides 1 Bouteloua gracilis 3 0.6 0.02 lieaNy grazing Buchloe dactyloides / Bouteloua gracilis 30-13 2.4-15.9 nr varied — lightly grazed / recent burns Artemisia tridentata / Agropyron desertorum 32 9.3 3.0 grazed pasture/ no intensit\- specified Artemisia tridentata 1 Agropyron spicatum 1 Stipa thurheriana 1 80 0.9 0.7 lightly grazed/ brush control 10 \T prior to stud)' killed 95% of plants Artemisia tridentata 1 Agropyron spicatum 1 Stipa thurheriana 1 Bromus tectorwn 57 1.5 0.8 lightly grazed/ brush control 22 yr prior to study killed 95% of plants Artemisia tridentata 1 Agropyron spicatum 1 Stipa thurheriana 1 Bromus tectorum 25 1.4 nr moderateK grazed Bouteloua gracilis 31 1.2 nr lightK grazed Bouteloua gracilis nr 3.5 nr no grazing or fire in 30+ yr Artemisia tridentata 1 Oryzopsis hymenoides 11 r 5.3 nr burned 5 yr prior to sample, then ungrazed Artemisia tridentata 1 Onjzopsis hymenoides "Identified as occidciilalis. Imt in the kiumii ranne of i)» |//i( ci ''Not reporteci ^All references to grazing refer to grazing of cattle or other li\'estocl<. "P. owyheei was considered to be part of P occidentalis until 19.50. and cleared disc size in an undisturbed area. Much of the information on areas cleared b\' Pogonomijrmex hai^vester ants relates to stud\' sites with vaiying degrees of disturbance his- toiy However, few studies examine the role of P. owyheei and P. oceidentalis as agents of plant removal in undisturbed environments. In our study we briefly compare results of plant removal in undisturbed areas with those results presented elsewhere. Our nest site density of 40/ha is in the approximate middle range of reported obser- vations under a variety of land-use influences (Table 2). Disturbance may sene to inciease the nest site densities at any given site up to a point. For example, Rogers and Lavigne (1974: 995) found an increase in nest site density under "light" and "moderate" grazing, but shaiply re- duced densities under "heavy" grazing. Findings of Sharp and Barr (1960) and Sne\ a (1979) also suggest increases in nest site densit\' are asso- ciated with distiubance (Table 2). Across the range of P. owyheei and P. oecidentalis, nest site densities are likeK controlled b\ a suite of factors (soils, \egetation composition, climate, disturbance histor}) acting synergistically. Increases in nest site density in grazed areas probabK result from alterations of the d\nam- ics of competition between plant species that in turn modifv seed densit\ distributions 1996] P. OwYHEEi Nest Density and Size 165 (HolldoMer and Wilson 1990). On their study- site in southern Arizona, tor example, Da\id- son et al. (1984) found that haivester ant popu- lations began to decrease approxiniateK 2 > r after rodent populations were intentionally reduced. Da\idson et al. (1984: 1780) con- eluded that rodent removal led to a "differen- tial increase ' in large-seeded annuals because of the cessation of granivory, and this in turn precipitated the competiti\e displacement of small-seeded species that were the ant s pri- man' food source. Although other studies have used larger sample sizes to determine nest density (e.g.. Coffin and Lauenroth [1988] used a 2.5-ha sample), we believe our nest site density is a reasonable estimate for HRRNA because (1) the study site is consistent in regard to soils and \'egetation, and has only minor topo- graphic variability; (2) our standard error per sample for nest density is small, suggesting lit- tle variability within our study area; and (3) research from studies on other Pogonotnyrmex species has shown that soil texture can affect nesting location (e.g., Johnson 1992, DeMers 1993), and that a uniform dispersion of ant colonies develops regardless of spatial scale examined (Wiernasz and Cole 1995). There appears to be little relationship between dis- turbance history and mean size of nest sites {Table 2). Sneva (1979) has speculated that while there may be great variability in nest site density and disc area, the potential avail- able forage per nest site generally remains consistent, suggesting that vegetation cover and species composition can affect disc size. Soil characteristics also impact disc size, with a tendency for colonies to expand horizontally in shallow soils (Sneva 1979). Therefore, disc size may be largely linked to the amount of vegetation cover, plant species composition, and soil depth, and less influenced by distur- bance than is nest density. Literature Cited Anonymous. 1972. Horse Ridge Re.searcli Natural Area. Federal Research Natural Areas in Oregon and Washington. Pacific Northwest Forest and Range Experiment Station, Portland, OR. Baldwin, E. M. 1974. Evaluation of Horse Ridge Natural Area, Deschutes County, region for eligibility for registered natural landmark designation. Supple- mental Report, University of Oregon, Eugene. Clark, W. H., and P L. Comanor. 1975. Removal of annual plants from the desert ecosystem by western harvester ants, Pononoimjrrnex occideutalis. En\ iron- mental Entomology 4: 52-56. Coi-i in, D. P, and W. K. LAL'liNKOTli. 1988. The ell'ects of disturbance size and frcciuenc)' on a sliortgrass plant conunuiiity. Ecology 69: 1609-1617. • 1990. Vegetation associated with nest sites ol' western harvester ants {Po^oiiomi/nncx accUlcntalis Cresson) in a semiarid grassland. American Midland Naturalist 123; 226-2.35. Cole, A. C, Jh. 1968. Po^,()ii(>inijnii('x harxesler ants: a study of the genus in North America. The University oi ieunessee Press, Kno.wille. 222 pp. Davidson, D. W, R. S. iNouvt:, and J. H. Brown. 1984. Granivoiy in a desert ecosystem: experimental evi- dence for indirect facilitation of ants b\ rodents. Ecology 65: 1780-1786. DeMers, M. N. 1993. Roadside ditches as corridors for range expansion of the western harvester ant (fogo- nomi/nncx occidcnidlis Cresson). LaiidscaiK' Ecolog\' 8: 93-102. Fr\nklin, J. F, AND C. T. DVRNESS. 1988. Natural vegeta- tion of Oregon and Wiishington. Oregon State Uni- versity Press, Cor\'allis. 452 pp. Gashwiler, J. S. 1972. List of birds, manunals, and plants found on Horse Ridge. Letter (12/27) to Paul W. Arrasmisth, Prineville, OR, BLM. . 1977. Bird populations in four \egetational t>pes in central Oregon. Special Scientific Report — Wildlife No. 205. United States Department of Agriculture, Fish and Wildlife Service, Washington, DC. Halvorson, R. 1991. Natural ignition (fire) in Horse Ridge ACEC/RNA, concerns and challenges. Correspon- dence (9/16) to ADM-Resomce Serx'ices, Princv ille. OR, BLM. HOLLDOBLER, B., AND E. O. WiLsoN. 1990. The ants. The Belknap Press, Cambridge, MA. 732 pp. Johnson, R. A. 1992. Soil texture as an influence on the distribution of the desert seed-hanester ants Pofioii- omynnex rugosus and Messor pergandei. Oecologia 89: 118-124. K.\RL, T R., C. N. Williams, Jr., F T Qhnlan, and T A. BODEN. 1990. United States Historical Climatolog\' Network (HCN) serial temperature and precipita- tion data. Carbon Dioxide Information Analysis Center, Oak Ridge, TN. 379 pp. NowAK, R. S., C. L. NowAK, T. DeRocher, N. Cole, and M. A. Jones. 1990. Prevalence of Orysopi.s hijincnokh's near harvester ant mounds: indirect facilitation b\ ants. Oikos 58: 190-198. Rogers, L. E., and R. J. Lavigne. 1974. Environmental effects of western hai-vester ants in the sliortgrass plains ecosystem. Environmental Entomolog\- 3: 994-997. Rogers, L. E., R, J. Lwigne, and J. L. Miller. 1972. Bioenergetics of the western harvester ant in the shortgrass plains ecosystem. Environmental Ento- mology 1: 763-768. Sharp, L. E., and W F Barr. 1960. Prcliminan investiga- tions of hanester ants on southern Idaho range- lands. Journal of Range Management 13: 131-134. Sneva, F A. 1979. The western harvester ants: their den- sity and hill size in relation to herbaceous productiv- ity and big sagebrush coven Journal of Range Man- agement 32: 46-47. USDA-Natural Resources Conservation Service. In press. Upper Deschutes River Area, Oregon Soil Sur- vey. USDA-NRCS, Washington, DC. 166 Great Basin Naturalist [Volume 56 WIERNASZ, D. C, AND B. J. CoLE. 1995. Spatial distribu- Willard. J. R. AND H. H. CRO^VELL. 1965. Biological tion of Pogonomyrrnex occidentdis: recnutment. mor- activities of tl.e harvester .ni, Togonomynnex ou..jhe.^ tality and overdispersion. Journal of Animal Ecology in central Oregon. Journal of Economic Entomology 64: 519-527. 58: 484-489. Wight J. R., and J. T Nichols. 1966. Effects of harvester . , ,o m 7 loo^t * 1 „„ p J,c.o„ of a saUb,.^ c„,„„,u„«y, ,o,„™l r'^'^S* 'Z of Range Management 19: 69-/1. ^^ Great Basin Xatiiralist 56(2), © 1996, pp. 167-171 FIELD MEASUREMENTS OF ALKALINITY FROM LAKES IN THE UINTA MOUNTAINS, UTAH, 1956-1991 Dennis I). Austin' Abstract. — Data follcitid hom alpine lakes in tl:r L inla Monntains dnrinu; I'islieiy siint-ys In' tlie Utali Division of Wildlife Resources indicate alkalinit\ has decreased in some drainages since tlie mid 195()s. implications for continued monitoring, as well as environmental and recreational values, are discussed. Key uords: alkdliiiify. acid pirripitdtioii, dlpiiic lakes, tcaler (jiialitij, I'iiili. Alpine lakes in the Uinta Mountains have the lowest total alkalinity of all suriaee waters in Utah (EPA 1982). The low alkalinity is due to the Piecanibrian roek geologie origin eom- posed primarily of nietamorphic quartzite, ph\llite, and diamictite. Because of low alkalin- it>; these lakes are sensitive to acid precipita- tion, which may affect long-term water quality, fish, and other a(|uatic organisms. The Utah Division of Wildlife Resources has measured alkalinity in many of these lakes since 1956. The purpose of this paper is to document the changes in alkalinit\' between 1956 and 1991 in the Uinta Mountains by drainage. Methods Water from lakes in the Uinta Mountains, Utah, was sampled and measured for alkalinity from 1956 to 1991 b>' the Utah Division of Wildlife Resources in conjunction with the fisheries sun^eys. Data were collected during summers on selected lakes within 16 of the 18 major drainages (Fig. 1) and during 3 desig- nated sampling periods: mid 1950s-early 1960s (period 1), 1970s-early 1980s (period 2), and mid 1980s-early 1990s (period 3). All alkalin- it>' data were collected in the field using col- orimetric methods and converted to mg/L. In period 1, tests were made using methylpuqDle indicator and titrating with 0.02 N sulfuric acid. Alkalinity' titrations were made at stepwise increments of 5.0 mg/L (per drop). In periods 2 and 3, tests were made with Hach (Hach Company, PO Box 389, Loveland, CO 80539) Water Ecology Kits, model AL-36B. Alkalinity titrations were made at increments of 6.8 mg/L (per drop) in all drainages, except during period 2 in Rock Creek, Duchesne, and Provo River drainages, and during period 3 in Hock Creek, Burnt Creek, and Sheep-Carter C'reek drainages when the increments were 17.1 mg/L (per drop). The effects of 3 weaknesses in the available data — the lack of data sets from all drainages during all 3 periods, the 3 levels of sensitivity in the alkalinity measure- ments, and the differences in sample sizes — are unknown and suggest interpretive caution of the results. The significance lc\el was set at P < 0.05 for the 3 comparisons of statistical testing. To test for changes over all drainages, mean alkalinity among drainages was compared between periods using ANOVA for imequal sample sizes (Sokal and Rohlf 1981). To test for changes in alkalinit\ within drain- ages, data were compared between periods. T tests of the mean were used when data from 2 periods were available, and ANOVAs for un- equal sample sizes were used when 3 periods of data were available. To test for changes in alkalinit)' within drainages for the same sampled lakes, I com- pared data between periods. T tests for paired comparisons were used when 2 periods of data were available, and ANOVAs for equal sample sizes when 3 periods of data were a\ ailable. Results Mean alkalinity among drainages signifi- cand\' decreased (F < 0.05) between all 3 ' Dupiirtimiit of Rangf Scit-nce. i:tali State Uni\'ersit>', Logan, UT 84322-,5230. Present addres.s: 43 .South 700 East. H>Tuin. L'T 84319. 167 168 Great Basin Naturalist [Volume 56 SALT LAKE CITY US 189 SOUTH US 80 NORTH U S 40^ U S 30 KAMAS • DUCHESNE RIVER TABIONA^y J ROCK CREEKS! BEAR RIVErIN^ UTAH WYOMING U-150 EVANSTON DUCHESNE LAKE FORK ^BLACKS FORK J YELLOWSTONE RIVER .SMITHS FORK> *•>:.*•;•:• %•;;.•.*. :.\\\\\\v gSWIFT CREEK| HENRYS FORK ^DRY GUlCHll;.- ^^^^^^^^^^^^ :••.•.*••-••-••••• EEAVER CREEK ^ ( \'.V UINTA RIVER,*, j^——- ' ' • « % < WHITEROCKS/ KAPOINT/ •WHITEROCKS RIVER ^""^^^ ^^^^ '/ ^ SHEEP CREEK ASHLEY CREEK CARTER 5 ^5^ CREEK 1^ •ft 'MOUNTAIN VIEW LONETREE RED( CLOUD LOOPl ROAD U.S. HIGHWAY 30 EAST MANILA I VERNAL U.S. HIGHWAY 40 EAST U-44 DUTCH JOHN FLAMING GORGE RESERVOIR! \ I NORTH Fig. 1. General location and major clrainatjes in the Uinta Mountains. Utah. 1996] Uinta Lakes Alk.\linity 169 periods (Table 1) from 33 niti/L in period 1, to 23 mg/L in period 2, to 17 nie;/L in period 3. (Standard de\iations for all means listed in Tiible 1 are available from the anthor) Alkalinity within individnal drainai^es sit:;- nifieantK deereased in the Dnehesne Kixcr and l^r()\() Hi\er drainages between all 3 peri- ods. Alkalinit>' decreased between periods 1 or 2, and period 3 in the Rock Creek, Weber River, and \Vhiteroeks River drainages. No change in alkalinity was fonnd between periods 1, 2, and 3 in the Bear River drainage. Similarly, no change in alkalinity between periods 2 and 3 was fonnd from Beaver Creek, Blacks Fork, Smiths Fork, and Henns Fork drainages. Due to lack of data, no additional comparisons could be made. Changes in alkalinity within drainages, where data from the same lakes were a\'ailable be- tween 2 periods, were variable. Alkalinity did not change in die Rock Creek drainage between periods 1 and 3, or in the Weber River drain- age between periods 1 and 2. Alkalinity also showed no change in the Rock Creek, Bear Rixer, Blacks Fork, Smiths Fork, or Whiterocks River drainages between periods 2 and 3. However, alkalinity decreased in the Duchesne River and Provo River drainages between periods 1 and 2, 2 and 3, and 1 and 3. Alkalin- ity also decreased in the Weber River and Bear River drainages between periods 1 and 3, but increased in the Henrys Fork and Beaver Creek drainages between periods 2 and 3. No additional comparisons could be made. Alkalinity in the Duchesne River and Provo River drainages where the same lakes were sampled during all 3 periods was significantly (F < 0.05) different between all 3 periods for both drainages. Mean alkalinity values using the combined data from these 2 drainages decreased from 37 mg/L in period 1 to 22 in period 2, to 6 in period 3. No comparisons over the 3 periods could be made from the other drainages. Discussion Negative effects of acid precipitation on aquatic ecosystems have been well documented, particularly in Europe and eastern North America (Haines 1981a). Acid precipitation can have negative impacts on water chemistiy and quality, algae, bacteria, invertel)rates, amphibians, fish, waterfowl, and aquatic vege- tation (Rough and Wilson 1977, Fl^A 1979, Haines 1981b, Kretser et al. 1983), and result in a general reduction in biodiversity (Fryer 1980). Alkalinity and pll are directly related in maintaining acjuatic ecosystems; and as alka- linity decreases, lakes become increasingly sus- ceptible to acidification (Haines 1981a). Acidi- fication rates were reported by Dillian et al. (1987) for 2 Canadian lakes as 2 ue(|/L/yr be- tween 1979 and 1985 with a 3-fold decrease in alkalinity accompanied b\' a 0.2 pH decrease. Decreases in alkalinity have been reported in Colorado. In die Colorado Rockies, 64 lakes were compared between 1938-1960 and 1979 with a mean decrease between periods (1938- 1960 vs. 1979) of 17% alkalinity (Lewis 1982). In the Mt. Zirkel Wilderness Area, Turk and Campbell (1987) reported an approximate loss of buffering capacity of < 10% in most lakes they sui-veyed. Data from this study indicate a 50% de- crease in alkalinity since the 1950s, with the rate of decrease about 0.5 mg/L/yr in the Uinta Mountains. At this rate of decrease, studies extended for only a few years would likely show no change in alkalinit\. Contrary to our results, 2 previous studies conducted in Utah indicated no effects of acidi- fication. In a snowmelt stud\' of the Wasatch Mountains, Messer et al. (1982) reported a mean snowmelt pH of 6.17 and concluded that enough buffering capacity was retained in the snowpack to neutralize acid ecjuivalents from air pollution. In a report from the Utah Techni- cal Advisory Committee (m acid deposition, Ellis (1986) concluded that although lakes and streams in the Uinta Mountains are ven' sen- sitive to acidification, no evidence was found that demonstrated acidification had occurred. The lack of acidification was based primarih on data collected in the Mirror Lake water- shed during 1983-1986. Both studies sug- gested windblown particulates from the Great Salt Lake Desert were sufficient to buffer acid deposition. Decreased alkalinity from alpine lakes sam- pled by the Utah Division of Wildlife Resources in the Uinta Mountains over 35 years indi- cated a slow decline in alkalinity; particularly in the Provo River and Duchesne River drain- ages. Unaltered, this decline may eventually result in deterioration of the aquatic ecosys- tem and, sul)sequently, recreational values. 170 Great Basin Naturalist [Volume 56 Table 1. Mean total alkalinib.' (mg/L) by drainage from alpine lakes in the Uinta Mountains, Utah. C Dnibine d data from all sampled lake Phase n = IS Period 1 Period 2 Period 3 Drainage Year n mg/L Year 11 mg/L Year n mg/L Rock Creek 1956 9 311" 1973 3 34" 1983 54 21b Duchesne River 1956 30 35'' 1979 1 241' 1985 34 8^ Provo River 1956 23 36" 1979 20 231' 1986 54 8^- Weber River 1956 27 35" 1983 3 22"1' 1987 16 12'' Bear Ri\'er 1956 5 30 1982 26 17 1989 30 21 Blacks Fork ND^ — — 1982 22 26 1989 21 25 Smiths Fork ND — — 1983 24 16 1990 20 15 Henr\'s Fork ND — — 1984 22 15 1990 21 24 Beaver Creek ND — — 1984 23 20 1991 17 29 Biunt Fork ND — — ND — — 1984 11 17 Sheep/Carter ND — — ND — — 1984 32 19 Creeks Ashley Creek ND — — ND — — 1988 21 13 Whiterocks River ND — — 1976 2 34" 1985 43 14'' Uinta River ND — — ND — — ND — — Di-y Gulch ND — — ND — — 1987 12 14 Yellowstone River ND — — ND — — 1986 24 14 Lake Fork River ND — — ND — — ND — — Swift Creek ND — — ND — — 1987 17 14 Total/Mean — 94 33'' — 152 23'- — 427 17^- In umbers witli difierent letters across rows wen ^NDC = No data from common lakes available. 3ND = No alkalinit\' data collected. sitjnilkantK dittercnt. F < 0.0.5. Additional sampling is essential to monitor and document alkalinit\' and potential acidifi- cation of Uinta Mountain lakes. Acknowledgments This report was funded, in part, by the Utah State Division of Wildlife Resources, Pittman- Robertson, Federal Aid Project W-105-R. Spe- cial thanks is given to Jeriy D. Weichman for his careful reviews of this manuscript. Literature Cited DiLLiAN, D. J., R. A. Reid, and E. de Grosbois. 1987. The rate of acidification of aquatic ecosystems in Ontario, Canada. Nature 329: 45-49. Ellis, M. T. 1986. Acid deposition in Utah. Utah Depart- ment of Health, Salt Lake Cit>'. EPA. 1979. Research summary: acid rain. U.S. En\iron- mental Protection Agency Publication 600/8-79-028. Washington, DC. EPA. 1982. Total alkalinity' of smface waters. U.S. En\ iron- mental Protection Agenc\ Publication 6()0/D-82- 333. Corvallis, OH. Fryer, G. 1980. Acidity and species diversity in iresh water crustacean faunas. Freshwater Biolog\ 10: 41-45. Haines, T. A. 1981a. Acidic precipitation and its conse- quences for acjuatic ecosystems: a review. Transac- tions of the American Fisheries Society 110: 669-707. . 1981b. Waterfowl and their habitat: threatened by- acid rain? Pages 177-190 //; 4th International Water- fowl Symposium. KrETSER, W a., J. R. COLQLHOUN, .\ND M. H. Pfeiffer. 1983. Acid rain and the Adirondack sportfishen. Conservationist 37: 22-29. Lewis, W M., Jr. 1982. Changes in pH and buffering capacity of lakes in the Colorado Rockies. Limnology and Oceanography 27: 167-172. Messer, J. J., L. Slezak, and C. I. Life 1982. Potential for acid snowmelt in the Wasatch Mountains. Water Quality Series UWRL/Q-82/06. Water Research LaboratoiT, Utali State Uni\ersity, Logan. PouGH, F H., AND R. E. Wilson. 1977. Acid precipitation and reproduction success of AmhijsfoiiKi salaman- ders. Water, Air, and Soil Pollution 7: 307-316. SoiCAL, R. R., AND F J. ROHLF 1981. Biometn. 2nd edi- tion. W. H. Freeman and Company, New York. Turk, J. T., and D. H. Campbell. 1987. Estimates of acid- ification of lakes in the Mt. Zirkel Wilderness .\rea, Colorado. Water Resources Research 23: 1757-1761. Received 22 March 1995 Accepted 30 Decouher 1995 1996] Uinta Lakes Aljwvlinitv 171 Tablk 1. Contiiuifd. Dat; from the saiiic lakes Data ti oni the same lakes sampled during 2 periods saTiipl L'd during all 3 jieriotls 1 2 1 3 2 3 1 2 3 11 inii/L niU/L /( ni,ti/L ni.iVl. n niK/L mg/L II mg/L mg/L iiiK/L NDC- — — 8 30 24 3 34 17 NDC 6 35'' 231' 21 36" 5'' 6 23" 61' 6 33" 23'- 6'- 14 37'' 211' 14 38" 71, 15 22" 5I' 10 41" 21'' 6^- 2 33 7 8 33^' 10'' NDC — — NDC NDC — — 5 30" I3I' 17 20 20 NDC _ _ NDC — — NDC — — 18 24 26 NDC NDC — — NDC — ■ — 13 11 12 NDC NDC — — NDC — — 14 12" 171, NDC NDC — — NDC — — 17 22" 29'' NDC NDC — — NDC — — NDC — NDC NDC — — NDC — — NDC — — NDC — — — NDC NDC NDC _ _ NDC NDC — — NDC — — 2 34 14 NDC NDC — — NDC — — NDC — — NDC NDC — — NDC — — NDC — NDC NDC — — NDC — — NDC — — NDC NDC — — NDC — — NDC — — NDC NDC — — NDC — — NDC — — NDC 22 35^' 171- 56 33" 12'' 88 23 17 16 37" 22'^ 6^ Great Basin Naturalist 56(2), © 1996, pp. 172-176 DENSITY, BIOMASS, AND DIVERSITY OF GRASSHOPPERS (ORTHOPTERA: ACRIDIDAE) IN A CALIFORNIA NATIVE GRASSLAND Eric E. Porterl-, Richard A. Redakl, and H. Elizabedi Braker^ Abstract. — A native California perennial grassland \\'as sampled for grasshopper populations. The grassland is man- aged for the presei-vation of the native perennial bunchgrass, Nassella pulchra Hitch. Grasshopper densit)', biomass, diversity, and richness were measured from July 1993 to October 1994. Average density of all grasshoppers was 2.30 hoppers/m- (0.66 s) for 1994 (June through August). Overall forage consumed for 1994 was 140 kg/ha, suggesting diat grasshopper populations e.xist at economically damaging levels. Grasshoppers do not appear in the grasslands until late spring, after annual grasses have set seed. Biomass of grasshoppers peaks in July when adults are predominant. Both grasshopper density and biomass were higher in 1993 than in 1994, and a total of 5 species were found throughout the stud). Mclanophis sangiiinipes Fabricus dominated the acridid communities and accoimted for more than 95% of the indi\iduals. Key words: Nassella pulchra, Melanoplus sanguinipes, Califonuii native grassland, density, diversity, grasshopper herbivory, Acrididae. California's native perennial bunchgrass communities have been reduced to less than 1% of their original range (Heady 1977), with much of this loss attributable directly to the development of agricultural and urban areas (Huenneke 1989). Additionally, most undevel- oped patches of native grasslands have con- verted to grasslands dominated by annual grasses native to the Mediterranean region (Jackson 1985). Factors leading to the success of these Mediterranean species are not com- pletely understood; however, heavy grazing pressure has been implicated as a major foctor that favors these more ruderal annual species (Burcham 1957). In their pristine state, before the arrival of European settlers, California's grasslands had light grazing pressure (Wagner 1989). Removal of major anthropogenic distur- bances such as grazing and fire does not lead to the recoveiy of native perennial grasslands (White 1967, Keeley 1981). Most investigators now agree that the annual grass species should be considered naturalized, and a return to the pristine disturbance pattern will not lead to reestablishment of native grasslands (Headv 1977). Joern (1989) suggests that through differen- tial herbivoiy upon the perennial grasses (rela- tive to annuals), grasshoppers may have con- tributed to the establishment of exotic annual grasses in California's native grasslands. Grass- hopper herbivory is presumed to be greatest in simimer months when annual grasses already have set their seed and prior to germination in the fall (Joern 1989). Therefore, only perennial grasses and summer forbs are susceptible to damage by grasshopper herbivoiy. Fintheniiore, many grasshopper species exhibit preference for perennial grasses in the field (Capinera and Sechrist 1982). Joern (1989) suggests that this phenology-based, selective damage could re- duce the competitive abilit\' of native perennial grasses against naturalized annuals. There are few data available to support or refute Joern's (1989) hypothesis beyond basic surveys of grasshoppers throughout the state (Strohecker et al. 1968). No population or communit\ -level studies are axailable for Cali- fornia s grasshoppers in California natixe peren- nial grasslands (e.g., population densit>', species abundance, and biomass estimations). The ob- jective of this study was to describe the grass- hopper communitx' found in a representatixe remnant stand of nati\e perennial grassland over a period of 2 seasons. These data will provide information necessary to understand the role of grasshoppers in California's grass- lands and shoidd lead to more informed deci- sions for grassland conservation managers. 'Department of Entomology, University of Clalifornia at Riverside. Ri\erside, C;.\ 92.521. ^Please address all correspondence to this author ■■^Departnu'Tit of Biolc)t;\, Occidental C^ollcKc, Los .\ngcles, C:A 90(1.| 1 172 1996] GllASSllOlTKUS OF A CaLIFOKNIA GRASSLAND 173 Methods ani:> Matkhials Study Area The stucK" WHS condiieted in the Santa Rosa Phitean Ee()l(),uical Reserve (SRPER), located 10 km west of Mnrrieta, C'ahfornia. The site is acti\el\' managed b\ The Natuie C()nser\'ane>' for the restoration and preser\'ation of its rare habitats. Tlie resene covers 2800 ha and con- tains abont 1200 ha of nati\e perennial grass- lands amongst oak woodlands, coastal sage scrnb, and chaparral. Six sites were established within the perennial grasslands. These sites were burned in June ot 1992 as a management practice to retard annual grass establishment. Grazing has been excluded from all sites since at least 1990 (R. Wells, SRPER reserve man- ager, personal communication). Purple needle grass {Nassella pulchra A. Hitch.) is the most abundant native grass in the reserve. Common exotic annual grasses include slender wild oats {Avena barbata Link) and red brome {Brotmis laevipes rubens Labill). Common forbs include annual bursage {AinJ)ro- sia acanthicarpa Hook.), doveweed [Ereino- carpus setigenis H.), and filaree {Erodium cicutarhnn EHer.; Lathrop and Thorne 1985). Grasshopper Sampling Six tiansects were arbitrarily placed through- out the perennial grassland areas representing maximum topographic and vegetational het- erogeneity. Each transect measured 200 m long b)' 20 m wide. Grasshopper density was determined with twenty 0.25-m^ hoops (Onsager and Henry 1977, Thompson 1987). Hoops were placed along each transect at 10- m intervals. Density was determined monthly beginning in July 1993. Grasshopper days (GHD) and forage consumption estimates were determined following Onsager (1984). GHD is a measure of total grasshoppers found per m- for a given year. Forage consumption is an estimate of the yearly forage consumption of grasshoppers based on estimated daily con- sumption (0.65 times body weight) and GHD. Biomass-days were calculated using the same formula for GHD replacing grasshoppers/m- with g/m-. When possible (density >0.5 grass- hoppers/m-), 100 individual grasshoppers were collected from each site and frozen immediately. These collections were taken directly following density counts and were made in August and October 1993; June, JuK, and August 1994. Grasshoppers were identi- fied to species and weighed to the nearest mg. Identifications provided the proportion of adults (p.,) in each sample. Gixen total deiisit)' (d), adult densit\ (dj was calculated with the following formula: da = (d) * (Pa). Species diversity was measured using the Shannon-Weiner index (Pielou 1977). Feeding category designations follow Capinera and Sechrist (1982) and Otte (1981; graminixorons, forbivorous, or mi.xed). Identification of nymphid stages is difficult, and damaging feeding does not occur until the 4th instar (Onsager 1984). Therefore, where possible, adult grasshopper data are analyzed separately from total grass- hopper data. Results Species of grasshoppers collected arc listed with subfamily and known feeding prefer- ences (Table 1). Average grasshopper density for the 1994 season (June-August) was 2.30 grasshoppers/m^ (Table 1). Density measure- ment began too late in the season to estimate an average for 1993. A total 198 GHD were determined for 1994, leading to an estimated 140 kg/ha of forage consumed. Density esti- mates of zero grasshoppers/m^ were found from November 1993 through May 1994. Densit>' peaked in June for 1994 at 2.9 grasshoppers/ m^. This peak in density was dominated by immature stages (Fig. lA). Density measures were higher in 1993 than in 1994 for all paired sample dates in July, August, and October (f = 4.69, df = 1, 20; F = 0.041). Biomass peaked in July when most grasshoppers were in the adult stage (Fig. IB). Biomass days for 1994 totaled 13.2 g-d/m^ (Table 1). Peak biomass (August) was higher in 1993 than in 1994 (t = 2.43; F = 0.036). The Shannon-Weiner diversity index, includ- ing adults and nymphs combined, averaged 0.140 over the 5 collection dates (Fig. 2A). The peak in adult diversity, in June 1994, repre- sents only 6 individuals of 2 species. Combined adult and nvniph species richness averaged 3.4 for the sampled dates. Highest species richness was found in August. In total, 5 grass- hopper species were found in these sites for the collection dates. 174 Great Basin Naturalist [Volume 56 Table 1. Species of grasshoppers collected in 1993-94 on the SRPER with known feeding t\pes (Capinera and Sechrist 1982, Otte 1981) and appearance, % composition, GHD, hioniass-days, average density; and weight as calcn- lated for the sampling period. GHD^ A\erage A\erage Sub- Feeding % (grasshopper Bioniass- densitv adult Species famih'' type'' Appearance composition days/m-) days^' (#/m2)^- weight (nig) Cammila peUucida Sciidder O G Mav-Aug 0.2 0.36 0.076 <().01 72 Mchiiioplits aridiis Scudder M n/a Jun-Aug 0.3 0.50 0.031 <0.01 90 Melanopltis sansitiiiipes Fabriciis M M Mav-Oct 97.5 193.47 12.754 2.24 132 Menncria bivittata Ser\ille G G Mav-Oct 1.6 3.18 0.375 0.03 81 Psolessa fcxcina Scudder G G Jun-Oct 0.5 0.89 0.121 0.01 61 Total. 198 13.240 2.30 131 ''O = Oedipodiiiae, .\l = .Melanoplinae, G = Goniphocerinat- "G = grass, M = nii.xed, N/A = not available ^GHD, biomass-da\'s. and average densit\' for 1994 season onl\ Discussion Grasshopper populations in the SRPER appear in late May or early June. Grasshopper biomass peaks, and the most severe herbivoiy occurs, in July and August. By this time, annual grasses have already died and their seeds are buried and protected from above- ground herbivory (Savelle and Heady 1970). Grasshopper densities decrease dramatically after August, and few are present by October. Annual grasses are triggered to germinate after the first fall rains (Heady 1958). By the time these rains arrive, grasshopper densities are near 0; therefore, both the mature annual grasses and dieir seedlings escape serious grass- hopper herbivory. One species, Melanopliis sanguinipes, accounted for over 95% of grasshoppers found in SRPER (Table 1). This species commonly damages crops and rangelands throughout North America (Hewitt 1977). Intense out- breaks are common and can remove up to 92% of aboveground vegetation (Nerney 1966, Hil- bert and Logan 1981). Melanopliis sanguinipes is classified as a mixed feeder and may prefer grasses or forbs depending upon the area sam- pled. If M. sanguinipes feeds extensively on grasses within the SRPER, given its phenol- ogy, it will damage perennial grasses more than annual grasses. Overall forage consinnp- tion was 140 kg/ha in 1994, which is an eco- nomically damaging level according to Onsager (1984). However, using M. sanguinipes in a shortgrass prairie communit\, Quinn et al. (1993) found significant reductions in grass biomass only at grasshopper densities equiva- lent to 845 GHD or greater. We found only 198 GHD in 1994, suggesting that densities during these years may not greatly affect grass- land plant community dynamics according to Quinn (1993). The climate of Galifornia s grasslands makes comparison with other North American stud- ies difficult. Most comparable studies examine tallgrass, mixed-grass, and shortgrass prairies east of the Sierra Nevada. The dn; hot summers that characterize Galifornia s Mediterranean climate severeK' limit growth (Risser et al. 1981). Regrowth following summer herbivory is similarly limited. Glearly, only perennial grasses are susceptible to herbi\'ory at this time, and the actual effect of such herbivoiy in CalifoiTiia grasslands is undocumented. There- fore, specific studies on the effect of grass- hopper herbivorx on nati\e Galifornia grass- lands must be conducted to predict the level and t>'pe of infestation, if an\, that ma>' fa\or annual grasses o\'er nati\e perennial species. Galifornia's grasshopper fauna is rich, com- pared to other North American regions, with over 120 species (Joern 1989). About half of these grasshopper species arc considered rangeland species. Southern Galifornia has the richest grasshopper fauna of any region in Gal- ifornia (Strohecker 1968). Still, richness for the SRPER \\'as vcr\' low^ compared to other studies of North American grasslands. Few studies report dixersit)' indexes, but richness has been measured for other natixe North American grasslands. Richness is nearlv^ alwavs 1996] GaVSSHOFPERS OF A CALIFORNIA GlUSSLAND 175 a Nymphs ■ Adults ■ Adults D Adults and Nymphs I '-' I ■ I ■ 1 1 Fig. 1. (A) Grasshopper density for all sampling dates and (B) biomass for dates collected (a = proportion of adults vs. nymphs unknown for July 1993). higher than the vahie of 5 species determined here (e.g., Joern 1982, Evans 1988). We feel it is hkely that the grasshopper community is particularly species depauperate due to isola- tion of the SRPER grasslands (MacArthur and Wilson 1967). Frequent burning of the grass- lands may also help explain the low diversity found in the SRPER. On the other hand, burned sites contain more even species com- positions than unburned grasslands and con- tain species not found in unbunied sites (Porter 1995). Therefore, we feel that to presence the diversity of not only grasshoppers but presum- ably many arthropods, birds, plants, and other taxa, it may be necessary to preserve larger tracts of native grasslands. Furthermore, the effects of grasshopper herbivoiy in these grass- lands must be accounted for in a well-rounded consenation effort. Acknowledgments We thank James Bethke, Brian Cabrera, Miriam Cooperband, Kim Hammond, Mari- anne van Laarhoven, Carl Matthies, and James Nichols for participating in the grasshopper collections. Helpful reviews by James Bethke, Fig. 2. Diversity (A) and species richness (B) for all dates in wliich collections were made. Timothy Paine, Anthony Joern, and an anony- mous reviewer are also greatly appreciated. We also thank Robin Wells and Cedra Shapiro of The Nature Conservancy for their assis- tance as well as the use of the reserve. This work was partially supported by Academic Senate grants to R. A. Redak. Literature Cited BuRCHAM, L. T. 1957. California rangi-laiid. California Division of Forestry, Sacramento. Capinera, J. L., AND T. S. Sechrist. 1982. Grasshoppers (Acrididae) of Colorado: identification, biolo,g\ and management. Colorado State Universitv- Experiment Station, Fort Collins. Bulletin 5.S4S. Evans, E. W. 1988. Community' dynamics of prairie grass- hoppers subjected to periodic fire: predictable tra- jectories or random walks in time? Gikos 52: 28.'3-292. Heady, H. F 1958. Vegetational changes in the annual grassland t>'pe. Ecology 39: 402—416. ^. 1977. Valley grassland. Pages 491-574 //! M. G. Barbour and J. Major, editors, Terrestrial vegetation of California. Wiley-lnterscience, New York. Hewitt, G. B. 1977. Review of forage losses caused by rangeland grasshoppers. USDA Miscellaneous Publi- cation 1348. 24 pp. HiLBERT, D. VV., AND J. A. LoGAN. 1981. A review of the population biolog>- of the migratory grasshopper, 176 Great Basin Naturalist [Volume 56 Melanophis sanguinipes. Colorado State University Experiment Station, Fort Collins. Bulletin 577S. HUENNEKE, L. E 1989. Distributions and regional pat- terns of California grasslands. Pages 1-12 in L. E Heuenneke and H. Mooney, editors. Grassland stnie- ture and function. Kluwer Academic Publishers, Dordrecht. Jackson, L. E. 1985. Ecological origins of California's Medi- terranean grasses. Journal of Biogeography 12: 349-361. JOERN, A. 1982. Distributions, densities and relative abun- dances of grasshoppers (Orthoptera: Acrididae) in a sandhills prairie. Prairie Naturalist 14: 37-45. . 1989. Insect herbivorv' in the transition to Califor- nia annual grasslands: did grasshoppers deliver the coup de grass? Pages 117-134 in L. E Huenneke and H. Mooney, editors, Grassland structure and function. Kluwer Academic Publishers, Dordrecht. Keeled; J. E. 1981. Reproductive cycles and fire regimes. Pages 231-277 in H. A. Mooney et al., editors. Pro- ceedings of the Conference Eire Regimes and Eco- system Properties. USDA Eorest Service, General Technical Report WO-26. Lathrop, E. W, and R. F. Thorne. 1985. A flora of the Santa Rosa Plateau. Southern California Botanists, Special Publication 1. MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, NJ. Nerney, N. J. 1966. Interrelated effects of grasshoppers and management practices on shortgrass rangeland. USDA Agricultural Research Sei^vice Special Report Z192: 1-14. Onsager, J. A. 1984. A method for estimating economic injury levels for control of rangeland grasshoppers with malathion and carbaryl. Journal of Range Man- agement 37: 200-203. Onsager, J. A., and J. E. Henry. 1977. A method for esti- mating the densit\- of rangeland grasshoppers (Orfiop- tera: Acrididae) in experimental plots. Acrida: 6: 231-237. Otte, D. 1981. The Nortli American grasshoppers. Volumes 1 and 2. Hai^vard University Press. PlELOU, E. C. 1977. Mathematical ecology. Wiley, New York and London. Porter, E. E. 1995. The grasshoppers of a California native grassland: a description of the community and its ecological importance. Unpublished master's the- sis. University of California, Riverside. QuiNN, M. A., ET AL. 1993. Effect of grasshopper (Ortliop- tera: Acrididae) density and plant composition on growth and destruction of grasses. Environmental Entomology 22: 993-1002. RISSER, P G., ET AL. 1981. The true prairie ecosystem. Hutchinson Ross Publishing Co., Stroudsburg, PA. 557 pp. Sa\elle, G. D., and H. E Heady. 1970. Mediterranean annual species: their responses to defoliation. Procla- mations of the 11th International Grassland Congress 548-551. Strohecker, H. E, et al. 1968. The grasshoppers of Cal- ifornia (Orthoptera: Acridoidea). Bulletin of the Cali- fornia Insect Survey No. 10. Uni\ersity of California Press. Thompson, D. C. 1987. Sampling rangeland grasshoppers. Pages 219-233 in J. L. Capinera, editor. Integrated pest management on rangelands: a shortgrass prairie perspective. Westview Press. Wagner, E H. 1989. Grazers, past and present. Pages 151-162 in L. E Huenneke and H. Mooney, editors. Grassland stnicture and function. Kluwer Academic Publishers, Dordrecht. White, K. L. 1967. Native bunchgrass (Stipa pulchra) on Hastings Reser\'ation, California. Ecology 48: 949-955. Received 10 ]ulij 1995 Accepted 19 October 1995 Great Basin Naturalist 56(2), © 1996, pp. 177-179 SUMMER NOCTURNAL ROOST SITES OF BLUE GROUSE IN NORTHEASTERN OREGON Kenneth J. Popper', Eric C. Pelrenl-, and John A. Crawford' Key words: Blue Gnntsf, DL'iidragapiis ohsciinis, nocturnal, Oregon, roost. Avian habitat studies frequently focus on diurnal habitat use because of ease of obsei'va- tion and high le\'els of activity associated with breeding and foraging. Nocturnal habitat use may be critical for all birds but has received far less attention. Thus, there is a need to bet- ter understand nocturnal habitat use, espe- cially by crepuscular and diurnal birds, and factors that may contribute to this use. Blue Grouse {Dendra^opiis obscunis) are associated primarily with true fir {Abies spp.) and Douglas-fir {Pseudotsuga menziesii) forests in mountainous regions of western North Amer- ica (Johnsgard 1983). Breeding season habitat associations often include nonforested and shrub or steppe regions. These birds are diur- nal with increased activity in the morning and evening hours. Pekins et al. (1991) determined that both diurnal and nocturnal winter roosts of Blue Grouse were located in conifers. Blue Grouse shifted from eating conifer needles in winter to groimd-layer vegetation in summer and fall in northeastern Oregon (Crawford et al. 1986). Blue Grouse summer habitat studies have dealt with diurnal activities (Mussehl 1963, Bendell and Elliot 1966, Zwickel 1975), but nocturnal obsen^ations are minimal. John- son (1929) witnessed a brood fly into a tree, apparently to roost overnight, and Blackford (1958, 1963) observed > 3 adult males flying into "roost trees" in spring, where they pre- sumably stayed overnight. Blackford (1963) also obser\'ed a male displaying on the ground approximately 1 h after dark. Zwickel (1992) suggested that ground roosting may occur, particularly on breeding ranges where trees are unavailable or before chicks are able to fly. In the course of monitoring radio-equipped Blue Grouse during summer, we identified 20 independent nocturnal roost sites. Our objec- tive here is to describe these roost sites. Study Area and Methods The study area is located in northeastern Oregon, 30 km north of Enteiprise in the Wal- lowa-Whitman National Forest in Wallowa County. Elevation ranges from 900 to 1500 m, with ridge slopes as great as 35°. North-facing slopes are dominated by stands of Douglas-fir and ponderosa pine {Pinus pondero.sa), and common shrubs are mallow ninebark [Phy.so- carpiis malvaceiis), snowberry {Symphoricar- pos albus), and big huckleberry {Vaccinium membranacenm). Bunchgrass meadows, pre- dominantly bluebunch wheatgrass {Agropyron spicatiim) and Idaho fescue {Festucu idahoen- sis), occur on south-facing slopes. Cattle graze parts of the area during summer months, resulting in variable grass cover. Grouse were captured in walk-in traps and fitted with poncho- or necklace-moimted radio transmitters, 15 to 18 g (Advanced Telemetr> Systems, Inc., Isanti, MN, and Telemetiy Sys- tems, Inc., Mequon, WI), from June through August 1993. Radio-equipped juvenile birds were > 500 g, capable of flight, and > 1 mon of age. Each radio-equipped bird was located at night once between 5 July and 3 August 1993. In addition to radio telemetry, a spotlight was used to verify- the location of the bird. The exact roost site was identified by the presence of fi-esh fecal droppings. When 2 or more grouse were observed roosting together (<10 m apart) only 1 roost site was counted for use in analyses to ensure independence of locations. 1 Department of Fisheries ujui Wildlife, Nasli Hall 104, Oregon State University, Corvallis, OR 97331. "Address all correspondence to Eric C. Pelren. 177 178 Great Basin Naturalist [Volume 56 Results and Discussion Twenty-five radio-equipped Blue Grouse and 38 birds without radios were located at 20 independent nocturnal roost sites (Table 1). The radio-equipped birds consisted of 12 adults and 13 juveniles; sexes and ages of the other birds were unknown. All roost sites were on the ground. Males usually roosted alone, whereas hens and juveniles frequently roosted together. Sixteen of 20 independent roosts, including birds of all sex and age groups, were in grass of a relatively consistent height; the others were in forbs {n = 2) and shrubs {n = 2). Twenty-three of 25 radio-equipped birds were within 50 m of potentially useful roost trees. An adult female and a juvenile female roosted 75 and 100 m from trees, respectively, both easy flight distances for grouse. Adult males usually roosted closer to trees than other birds. During daytime, radio-equipped birds were seldom located in trees (<1% of 614 obsei-vations, July-August 1991 through 1993; E. Pelren unpublished data). However, almost all birds flushed during the day landed in trees, and conifer needles were found in crops of birds taken from the study area in August and September 1981 and 1982 (Crawford et al. 1986). Crawford et al. also found plants such as prickly lettuce [Lactuca serriola), yellow salsify [Tragopogon diibius), wild buckwheat (Eriogoniun spp.), and snowbeny {Syniphori- carpos albiis), as well as short-homed grasshop- pers {Acrididae) in at least 30% of 145 Blue Grouse crops in this area. Douglas-fir needles were found in only 16% of the crops. This greater use of ground-cover forage and inver- tebrates corresponded with observed diurnal and nocturnal use of ground habitat by Blue Grouse in summer. Blackford (1963) suggested that selection of roosting sites may result from foliage preference and feeding habits. Motion sensors on grouse transmitters indicated that some birds continued foraging on moonlit nights, which implied that benefits of feeding outweighed energy loss associated with move- ment or increased risk of predation. Pekins et al. (1991) suggested Blue Grouse selection of conifers as roosts in winter may be based primarily on thermal properties of the sites. Higher temperatures during summer make thermal considerations less relevant to survival than during winter. The lowest tem- perature we noted at a nocturnal roost site was 4°C, well above the lower critical temperature of-10°C to -15 °C (Pekins 1988). Hines (1986) found that 96% of juvenile and adult Blue Grouse mortalities were the result of predation. In winter. Blue Grouse in trees may be less conspicuous or available to predators than those on the ground (Bergerud and Gratson 1988), and Pekins (1988) obsewed snow roosting only occasionally, after heavy snowstorms. However, lack of snow and in- creased presence of grasses, forbs, and shrubs in summer, along with cryptic coloration of Blue Grouse, provide ground-layer camou- flage superior to that available in winter. Food availabilit)' ma\' outweigh any increased risk of predation and account for use of nocturnal ground roosts by Blue Grouse in summer where selection of ground roosts occurs. Table 1. Characteristics of 20 Blue Grouse nocturnal roost sites, northeastern Oregon, JuK-August 1993. Adult male Adult female [uxenile male |u\enile female No. of roost sites 6 6 3(8a) 5 No. of other birds 1 16b 7 9^ Plant cover at roost Grass 4 6 3(8*) 3 Forb 1 0 0 1 Shrub 1 0 0 1 Plant height (m) at roost Median 0.50 0.45 0.50" 0.75 Range 0.25-1.20 0.25-1.00 0.30-0.75^' 0.30-1. .30 Distance (m) to potential roost tree Median 4.5 37.5 50.0^' 20.0 Range 1.0-40.0 15.0-75.0 3.0-75.0^' 5.0-100.0 •'Includes data for 5 radio-eiiiiippcd juvenile males that were with radio-equipped adult or ju\ enile fe "Does not include 2 radio-ecinipjied jusenile males that were with radio-e(]nipped adult lemales- "•Does not include.'} radio-equi|)iic-d juM-nile m.ilcs ih.it were \mIIi radio-cciuipped jumiuIc lemales. 1996] Notes 179 Acknowledgments This research was conducted as part of a Bkie Grouse winter ecology study funded by the U.S. Forest Service and Oregon Depart- ment of Fish and Wildlife. We thank R. L. Jarvis for assistance during the development of this paper. This is Oregon Agricultural E.xperiment Station Technical Paper 10,673. Literature Cited Bendell, J. F, AND P VV. Elliot. 1966. Habitat selection in Blue Grouse. Condor 68; 431-466. Bergerud, a. T, .\.\d M. W. Gratson. 1988. Sui-vival and breeding strategies of grouse. Pages 47.3-.57.5 /';( A. T. Bergeiiid and M. W. Gratson, editors. Adaptive strate- gies and population ecology of Northern Grouse. Uni\'ersity of Minnesota Press, Minneapolis. Bl.\ckford, J. L. 19.58. Territoriality and breeding beliav- ior of a population of Blue Grouse in Montana. Con- dor 60; 145-1.58. . 1963. Fiu^ther obsei"vations on the breeding be- ha\ior of a Blue Grouse population in Montana. Condor 65; 485-513. Cr.\\vford, J. A., W. V. Dyke, S. M. Meyers, and T. E Haensly. 1986. Ftill diet of Blue Grouse in Oregon. Great Basin Naturalist 46: 123-127. HiNES, J. E. 1986. Recruitment of young in a declining population of Blue Grouse. Unpublished dissertation, University of Alberta, Edmonton. 256 pp. JoHNSCARD, i^ A. 1983. The grouse of the world. Univer- sity of Nebraska Press, Lincoln. 413 pp. Johnson, R. A. 1929. Sununer notes on the Sooty Grouse of Mount Rainier Auk 46; 291-293. MUSSEHL, T. W. 1963. Blue Grouse brood cover selection and land-u.se implications. Journal oi' Wildlife Man- agement 27; .547-555. Pkkins, P J. 1988. Winter ecological energetics of Blue Grouse. Unpublishi'd dissertation, Utah State Univer- sity, Logan. 141 pp. Pekins, P J., E G. LiNDZEY, AND J. A. Gessa.man. 1991. Physical characteristics of Blue Grouse winter use- trees and roost sites. Great Basin Naturalist 51; 244-248. ZwiCKEL, E C. 1975. Nesting i)arameters ol Blue Grouse and their relexance to poiiiilatious. (londor 77: 423-430. . 1992. Blue Grouse. In: A. Poole, P Stetteuheim, and E Gill, editors. The birds of Nordi .'\merica. No. 15. The Academ>' of Natural Sciences, Philadelphia. The American Ornithologists Union, Washington, DC. 28 pp. Received 4 Augmt 1995 Accepted 23 October 1995 Great Basin Naturalist 56(2), © 1996, pp. 180-182 OOCHORISTICA SCELOPORI (CESTODA: LINSTOWIIDAE) IN A GRASSLAND POPULATION OF THE BUNCH GRASS LIZARD, SCELOPORUS SCALARIS (PHRYNOSOMATIDAE), FROM ARIZONA Stephen R. Goldberg^, Charles R. Bursey^, Chris T. McAllister3, Hobart M. Smith^, and Quynh A. Truong^ Key words: Sceloporus scalaris, bunch gross lizard, Phnjnosomatklae, Oochoristica scelopori, Cestoda, Arizona. The bunch grass Hzard {Sceloporus scalaris Wiegmann, 1828) is known from tlie Huachuca, Dragoon, Santa Rita, and Chiricahua moun- tains of Arizona, the Animas Mountains of New Mexico, and in the Sierra Mache Occidental and Sierra del Nido of Mexico, usually above 1830 m, but a few isolated valley populations occur as low as 1200 m (Stebbins 1985). To our knowledge, the only report of helminths of this species was a study of a high-elevation (2438- 2560 m) Chiricahua Mountain population of Sceloporus scalaris slevini by Goldberg and Bursey (1992a). The puipose of our note is to report on a helminthological examination of a low-elevation (ca 1524 m) grassland popula- tion of S. scalaris slevini Smith, 1937 from Ari- zona, and to compare our findings with those of Goldberg and Bursey (1992a). We examined 51 S. scalaris slevini (mean snout-vent length 51 ± 3.4 mm [s], range 40-55 mm) collected (mostly b\' hand, a few by dust shot) on the Sonoita Plain, elevation ca 1524 m (3r39'N, lir32'W), in the vicinity of Elgin, Santa Cruz County, Arizona. Specimens were deposited in the University of Colorado, Museum of Natural Historv, Boulder, Colorado as UCM 57259-57282; 57284-57286; 57289- 57292; 57295-57298; 57300-57305; 57307-57310; 57313-57316; 57318-57319. UCM 57318-57319 were collected 20 August 1989; others were collected 12-19 July 1990. The abdomen was opened, and the esopha- gus, stomach, and small and large intestines were removed from the carcass. Each organ was slit longitudinally and examined under a dissecting microscope. The liver and body cavitv were also examined. Each helminth was identified using a glycerol wet mount. Repre- sentative cestodes were stained with hema- toxylin and mounted in balsam for further ex- amination. Voucher specimens were deposited in the U.S. National Parasite Collection, Beltsville, Maryland 20705 (USNPC 85053). Terminology' use is in accordance with Margo- hsetal. (1982). Only 1 helminth was found, the cestode Oochoristica scelopori Voge and Fox 1950. Prevalence of infection was 10% (5 of 51); mean intensity = 1.2 ± 0.45 [.s], range 1-2. In the only other investigation of helminths of S. scalaris, Goldberg and Bursey (1992a) reported finding tetrathyridia of the cestode Mesocestoides sp. (prevalence 8%) and lan'ae of the nematode Physaloptera sp. (prevalence 3%). That study was done on a coniferous for- est high-elevation population (approximately 2500 m) in the Chiricahua Mountains, whereas the current study considered a low-elevation population (ca 1524 m) on the Sonoita Plain, located ca 126 km SE of the Chiricahua Moun- tains study site. Although both populations harbored mutually exclusive helminth faunas, additional work on larger S. scalaris samples from these sites will be required to determine the constancy of these differences. Oochoristica scelopori is a common cestode of North American lizards and has been found in 14 other North American phrxnosomatid lizards (Table 1). In addition, Anuein (1951) and Telford (1964) reported finding O. scelo- pori in the xantusiids, Xantusia henshawi, X. riversiana, and X. vigilis. Measurements of various structiues of these cestodes were strik- ingly different from the measurements as given 'Department of Biology; Whittier College, Whittier, CA 90608. Address correspondence to this autlior ^Department of Biology, Pennsylvania State University, Slienango VUlley Campus. Sharon, PA 16146. •'Department of Biology, Te.\as VVesleyan University, 1201 Wesleyan, Fori Worth, T.\ 76105-1536. -•EPG Biology, University of Colorado, Boulder, CO 80,309-0334. 180 1996] Notes 181 Table 1. Definiti\t^ hosts oi'OocIioristicd scclopori in North America. Host L<)L'aht\ Prevalence Keierence Crotaph ijtus colhiris Cahlornia Gamhelki wislizenii California Sceloponis clarkii Arizona S. graciosus California California Idaho Idalio Utah S. jarrovii Arizona Arizona Arizona S. imigisfer Arizona Texas S. occidentalis California California Idaho Oregon Utah S. olicaccus Texas S. orcutti California S. poinsettii Texas S. scalaris Arizona S. iindiihitiis Arizona Uma inornata California U. nofata California Urosauriis graciosus California 100% (1/1) 40% (2/5) 5% (1/20) not given 10% (7/71) 22% (2/9) 1% (1/118) 5% (1/22) 10% (47/489) 3% (1/31) 5% (15/302) (?/3) 6% (1/17) 20% (13/65) 23% (27/116) 11% (2/19) 33% (20/60) 9% (1/11) 3% (2/61) 22% (16/74) 30% (3/10) 10% (5/51) 6% (3/48) 7% (1/15) 42% (10/24) 6% (2/34) Telford 1970 Telford 1970 Goldberg etal. 1994 Voge and l-bx 1950 Telford 1970 VVaitz 1961 Lyon 1986 Pearce and Tanner 1973 Goldberg and Bnrsey 1990 Goldberg and Bursey 1992b Goldberg et al. 1995a Walker and Mathias 1973 Goldberg et al. 1995b Voge and Fox 1950 Telford 1970 Lyon 1986 White and Knapp 1979 Pearce and Tanner 1973 Goldberg et al. 1995b Goldberg and Bursey 1991 Goldberg et al. 1993 this paper Goldberg et al. 1994 Telford 1970 Telford 1970 Telford 1970 in the original description of O. scelopori by Voge and Fox (1950). Amrein (1951) reported the average length of 25 mature cestodes from X. henshawi and X. vigilis to be 15.82 mm; the cestodes from X. riversiana measured 33—37 mm. Telford (1964) indicated his cestode spec- imens from .xantusiid lizards were less than 45 mm. Both Amrein and Telford identified these cestodes as O. scelopori. Bursey and Goldberg (1992) found Amrein's measurements of ces- todes from X. henshawi and X. vigilis to approx- imate the measurements of O. bezyi, whereas Telford's measurements of cestodes from X. riversiana approximated measurements of O. islandensis and suggested that X. henshawi, X. riversiana, and X. vigilis be removed from the host list of O. scelopori, leaving only phrynoso- matid lizards as hosts for O. scelopori. Literature Cited Amrein, Y. U. 1951. The intestinal entozoa of the night lizards of California and their mode of transmission. Unpublished doctoral dissertation, University of Cali- fornia, Los Angeles. 162 pp. Bursey, C. R., and S. R. Goldberg. 1992. Oochoristica islandensis n. sp. (Cestoda: Linstowiidae) from the island night lizard, Xanfttsia riversiana (Sauria: Xan- tusiidae). Transactions of the American Microscopi- cal Societ\' 111: 302-313. Goldberg, S. R., and C. R. Bursey. 1990. Gastrointestinal helminths of the Yarrow spiny lizard, Sceloporus jar- rovii jarrovii Cope. American Midland Naturalist 124: 360-365. . 1991. Intestinal helminths of the granite spiny lizard [Sceloporus orcutti). Journal of Wildlife Dis- eases 27: 355-357. . 1992a. Helminths of the bunch grass lizard, Sceloporus scalaris slevini (Iguanidae). Journal of the Helminthological Societ\' of Washington 59: 130-131. . 1992b. Prevalence of the nematode Spauligodon giganticus (Oxyurida: Phaiyngodonidae) in neonatal Yarrow's spiny lizards, Sceloporus jarrovii (Sauria: Iguanidae). Journal of Parasitolog\' 78: 539-541. Goldberg, S. R., C. R. Bursey, and R. L. Bezy. 1995a. Helminths of isolated montane populations of Yan'ow's spiny lizard, Sceloporus jarrovii (Phnnosomatidae). Southwestern Naturalist 40: 330-333. Goldberg, S. R., C. R. Bursey, .\nd C. T. McAllister. 1995b. Gastrointestinal helminths of nine species of Sceloporus lizards (Phr>nosomatidae) from Texas. Journal of the Helminthological Society of Washing- ton 62: 188-196. Goldberg, S. R., C. R. Bursey, and R. Tawtl. 1993. Gas- trointestinal helminths of the crevice spiny lizard, Sceloporus poinsettii (Phr\nosomatidae). Journal of the Helminthological Society of Wishington 60: 263-265. . 1994. Gastrointestinal helminths of Scel(>i)orus lizards from Arizona. Journal of the Helminthologi- cal Society of Washington 61: 73-83. Lyon, R. E. 1986. Helminth parasites of six lizard species from southern Idaho. Proceedings of the Helmintho- logical Societ>- of Washington 53: 291-293. 182 Great Basin Naturalist [Volume 56 Margolis, L., G. W. Esch, J. C. Holmes, A. M. Kuris, and G. A. SCHAD. 1982. The use of ecological terms in parasitology (report of an ad hoc committee of the American Society' of Parasitologists). Journal of Para- sitology 68: 131-133. Pearce, R. C., and W. W. Tanner. 1973. Helminths of Sceloporus lizards in the Great Basin and Upper Colorado Plateau of Utah. Great Basin Naturalist 33: 1-18. Stebbins, R. C. 1985. A field guide to western reptiles and amphibians. Houghton Mifflin Companx', Boston. 336 pp. Telford, S. R. 1964. A comparative study of endopara- sitism among some southern California lizard popu- lations. Unpublished doctoral dissertation. Univer- sity of California, Los Angeles. 260 pp. . 1970. A comparative study of endoparasitism among some southern California lizard populations. American Midland Naturalist 83: 516-554. VoGE, M., and W. Fox. 1950. A new anoplocephalid ces- tode, Oochoristica scelopori n. sp., from the Pacific fence lizard, Sceloporus occidentalis occidentalis. Transactions of the American Microscopical Society 69: 236-242. VVaitz, J. A. 1961. Parasites of Idaho reptiles. Journal of Parasitology 47: 51. Walker, K. A., and D. V. Matthlas. 1973. Helminths of some northern Arizona lizards. Proceedings of the Helmin thole )gical Society- of Washington 40; 168-169. White II, R. L., xsd S. E. Knapp. 1979. Helminth para- sites of sceloporine (Iguanidae) lizards from central Oregon. Proceedings of the Helminthological Soci- ety of Wiishington 46: 270-272. Recewed 25 July 1995 Accepted 31 October 1995 Great Basin Naturalist 56(2), © 1996, pp. 1S3-185 POCKET GOPHERS DAiMAGE SALTCEDAR {TAMARIX RAMOSISSIMA) ROOTS Sara j. Maniiiiiu', iiriaii L. Casliorc', and Joseph M. Szewczak^ Kcij uord.s: sdltcidar. Tamari.x VMwosissitmi. pockti ^(>i)lier. Thoiiioiins hottac, tuiiiari.sk. Owens Valley, imusivc i)lanl, exotic plant. Saltcedar {Tamarix ramosissima Ledeb., Tamaricaceae) is an invasive, exotic woody slirul) natixe to Asia (Bauni 1978, Hickman 1993) that has colonized extensive areas tliroiighout the western United States (Robin- son 1965, Brotherson and Winkel 1986). Saltcedar possesses many characteristics that render it a nuisance plant (Brotherson and Winkel 1986), and because it has been viewed as a threat to native vegetation communities, researchers have examined its ecology (Car- man and Brotherson 1982, Brotherson and Winkel 1986, Shafroth et al. 1995), water con- sumption (Robinson 1958, van Hylckama 1970, Davenport et al. 1982, Bureau of Reclamation 1992), and cost of control efforts (Brotherson and Field 1987, Neill 1990, Barrows 1993). It is known to inhibit flows in creeks and springs (Robinson 1965, Rowlands 1990); thus, its spread has been detrimental not only to native vegetation but also to native wetland and aquatic fauna (Neill 1983). Altliougli efforts are under wa\' in the United States to develop biocontrol agents using insects that occur on saltcedar in its native range (DeLoach 1990), to date there have been no reports of native herbivores, insects, or diseases causing saltcedar mortality. Herein we report the first known mortality caused by native mammals on saltcedar. Our discovery occurred in Owens Valley, California. Water has been exported from Owens Valley — located in the rain shadow created by the Sierra Nevada range directly to its west — since 1913. Alteration of natural water flows created conditions favorable to the spread of saltcedar (Cashore 1985, Babb 1987). During the winter of 1995, when foliage was absent from saltcedar, we obsei-ved that a few plants within a young, even-aged stand were dead. Some of the plants were leaning over, supported by neighboring plants. Upon inspection, we obsened that dead plant tap- roots had been gnawed apart approximateK' 10 cm beneath the soil surface. Teeth marks were clearly visible on the tapered stumps. In addi- tion, prolific gopher tunneling was exident within and around the saltcedar stand, and excavated dirt mounds were located near the dead saltcedar. Examination of growth rings of plants within the stand showed the saltcedar plants to be 7 years old in 1995. In early April 1995, when saltcedar w as just beginning to break bud, we revisited the site to quantify the extent of animal damage and to capture and identify the species tunneling at the site. We examined plants by working from one end of the stand toward the center Ever>' saltcedar plant in approximately 1/2 of the stand was sampled, for a total of 545 plants. Height was measured, and then plants were tugged to detect the degree of below-ground damage. If tugged plants freely exited the soil and had no attached live roots, the damage was scored as fatal. All of these plants appeared dead, no resprouting was evident, and each had a chewed taproot stump, the diameter of which was measured and recorded. If tugged plants could be pulled from the ground easily, but still had live laterals above the chewed taproot, they were noted as sustaining severe damage. In these instances, diameter of the largest chewed root was measured. Tpically, these plants had many dead, but a few li\ ing, branches. If tugged plants felt loose, but could 'liiNci County Water Department, 163 May Street, Bishop CA 93514. -Deep Springs College, Dyer NV 89(110. 183 184 Great Basin Naturalist [Volume 56 not be easily pulled from the soil, they were scored as sustaining minimum damage. If tugged plants were tightly rooted in the soil, we assumed no root damage. The majority of branches on plants in both these categories appeared alive. Results of gopher damage are listed in Table 1. Nearly 23% of the plants sampled had experienced some degree of gopher damage; of these, 7.0% were dead as a result of gophers, 5.3% had been severely affected, and 10.6% had been minimally affected. The diameter of gopher-chewed roots ranged from 11 mm to 55 mm and averaged 27.7 mm. Gopher damage appeared to affect plant height; analysis of variance revealed significant height differences between plants in the 4 categories of damage (F = 4.463, P = 0.004, df = 3). However, saltcedar plants not dam- aged by gophers tended to be only slightly taller than plants sustaining gopher damage (Table 1), suggesting that gopher damage had been relatively recent. The study area was searched for evidence of active gopher mounds. Early in the evening, 7 active mounds were excavated, and Sherman live-traps baited with seeds and fresh plant material were placed at the tunnel level. These traps were then covered with soil, using local materials to prevent cave-ins at the trap en- trance. Trapping was done under the provision of a scientific collector's permit issued by the California Department of Fish and Game. Traps were checked the following morning shortly after sunrise. From tlie 7 traps set in active gopher tunnels, 1 valley pocket gopher {Thomoniys hottae) (Ingles 1965) was captured. Two other traps were found packed with soil, presumably by gophers. The 4 remaining traps showed no obvious sign of gopher activity. These data are the first reported evidence of a native species, Thomotnijs hottae, inducing mortalit}^ in the exotic Tamarix ramosissima. The proximity of a saltcedar stand to gopher habitat may increase its susceptibility to gopher damage. At our site, gopher mounds appeared more extensive in the alkali meadow immedi- ately adjacent to the saltcedar stand than in the stand itself We subsequentK' made obser- vations at other even-aged stands of saltcedar that occur adjacent to alkali meadows at other locations in Owens Valley and in Deep Springs Table 1. E.xtent of gopher damage within a stand of saltcedar plants in the Owens Valley. Gopher damage # plants % of Avg. ht. total (cm) None 420 77.0 128.9 33.3 Minimum 58 10.6 120.3 30.5 Severe 29 5.3 116.2 32.6 Fatal 38 7.0 113.4 25.3 All totii 545 100.0 126.: 32.8 Valley. Again we found gopher damage, so the phenomenon is not isolated to this single stand. In general, the influence of fossorial ani- mals on plant communities has received rela- tively little research attention (Andersen 1987). Although gophers may kill or slow the growth of saltcedar, their long-term effects on stand size and vigor or on saltcedar establishment in the meadow remain unknown. Other re- searchers have found that pocket gophers cause significant woody plant mortality in a variety of plant communities (Crouch 1971, Marsh and Steele 1992, Cox and Hunt 1994, Ferguson and Adams 1994), and Huntly and Inouye (1988) and Cantor and Whitman (1989) reported that tree encroachment into mead- ows was significantly slowed when gophers were present in meadows. However, given the vigorous growth of saltcedar in general, gopher damage may merely thin the stand, allowing the remaining individuals to continue unabated. Literature Cited Andersen, D. C. 1987. Below-ground herbi\or>- in nat- ural commimities: a review emphasizing fossorial animals. Quarterly Review of BiologN' 62: 261-286. Babb, D. E. 1987. Report on the saltcedar control stud>-. Unpublished report prepared for the Invo/Los Ange- les Technical Group. In\ o Coimt) Water Department, Bishop, CA. Barrows, C. W. 1993. Tamarisk control II: a success stoiy Restoration and Management Notes 11: 35-38. Baum, B. R. 1978. The genus Tamarix. Israel Academ\ of Science and Humanities. 209 pp. Brotherson, J. D., and D. Field. 1987. Tamarix: impacts of a successful weed. Rangelands 9: 110-112. Brotherson, J. D., and V. Winkel. 1986. Habitat rela- tionships of saltcedar {Tamarix ramosissima) in cen- tral Utah. Great Basin Naturalist 46: 53,5-541. Bl real oe RECLANl.vriON. 1992. Vegetation management study: Lower Colorado Ri\er Phase I report. Lower Colorado Region, Boulder Cit\-, NV. 103 pp. Cantor, L. F, and T. G. Wihtnl\n. 1989. Importance of belowgroimd herbi\'ory: pocket gophers may limit aspen to rock outcrop refugia. Ecology 70: 962-970. 1996] Notes 185 Carman, J. G., and J. D. Brotfiehson. 1982. Comparison of sites infested and not infested with saltcedar {Tainarix pcntaiidra) and Russian olive (Elea^iuis angustifolia). Weed Seienee 30: 360-364. Cashore, B. 1985. Saltcedar in the Owens Valley. Unpub- lished report, Inyo Coinit\ Water Department, Bishop, CA. 32 pp. Cox, G. W, AND J. Hunt. 1994. Pocket gopher herhivory and mortality of oeotillo on stream terrace, bajada, and hillside sites in the Colorado Desert, southern California. Southwestern Naturalist 39; 364-370. Crouch, G. L. 1971. Susceptibility' of ponderosa, Jeffrey, and lodgepole pines to pocket gophers. Northwest Science 45: 252-256. Davenport, D. C, P E. Martin, and R. M. Hagan. 1982. Evapotranspiration from riparian vegetation: water relations and irreco\erable losses for saltcedar Joiu- nal of Soil and Water Conservation 37: 233-236. DeLoach, C. J. 1990. Prospects for biological control of saltcedar (Tamarix spp.) in riparian habitats of the southwestern United States. Pages 307-314 in E. S. Delfosse, editor. Proceedings of the 7th Interna- tional Symposium on Biological Control of Weeds, 6-11 March 1988, Rome, Italy Ferguson, D. E., and D. L. Adams. 1994. Effects of pocket gophers, bracken fern, and western cone- flower on survival and growth of planted conifers. Northwest Science 68: 241-249. Hickman, J. C, editor 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley 1400 pp. HUNTLY, N., and R. Inouye. 1988. Pocket gophers in ecosystems: patterns and mechanisms. Bioscience 38: 786-793. Ingles, L. G. 1965. Mammals of the Pacific states. Stan- ford University Press, Stanford, CA. 508 pp. Marsh, R. E., and R. W Steele. 1992. Pocket gophers. Pages 205-230 in H. C. Black, editor, Silvicultural approaches to animal damage management in Pacific Nortliwest forests. USDA Forest Service, General 'i'echnical Report PNW-GTR-287. Nkill, W. M. 1983. The tamarisk invasion of desert ripar- ian areas. Desert Protective Council, Educational Bulletin 83-4. • 1990. Control of tamarisk by cut-stump herbicide treatments. Pages 91-98 in M. R. Kunzmann, R. R. Johnson and P S. Bennett, editors, Tamarisk Control in Southwestern United States: Proceedings of the Tamarisk Conference, 2-3 September 1987. Special Ik-port 9, University' of Arizona, Cooperative National Park Resources Studies Unit, Tucson. Robinson, T. W 1958. Phreatophytes. U.S. Geological Sur- vey Water Supply Paper 1423. 84 pp. . 1965. Introduction, spread, and aerial extent of saltcedar (Taniarix) in the western states. U.S. Geo- logical Survey Professional Paper 491-A. Rowlands, R G. 1990. Histoiy and treatment of the saltcedar problem in Death Valley National Monu- ment. Pages 46-56 mi M. R. Kunzmann, R. R. John- son, and P S. Bennett, editors. Tamarisk Control in Southwestern United States: Proceedings of the Tamarisk Conference, 2-3 September 1987. Special Report 9, University of Arizona, Cooperative National Park Resources Studies Unit, Tucson. Shafroth, P B., J. M. Friedman, and L. S. Ischinger. 1995. Effects of salinity' on establishment of Popiihis freinonfii (cottonwood) and Tamarix ramosissima (saltcedar) in southwestern United States. Great Basin Naturalist 55: 58-65. van Hylckama, T. E. a. 1970. Water use by salt cedar Water Resources Research 6: 728-735. Received 11 June 1995 Accepted 19 January 1996 Great Basin Naturalist 56(2), © 1996, pp. 186-187 SALTCEDAR {TAMARIX RAMOSISSIMA), AN UNCOMMON HOST FOR DESERT MISTLETOE {PHORADENDRON CALIFORNICUM) Sandra L. Haigh^ Key words: Phoradendron californicuni, Taniarix ramosissima, inisflctoc, saltcedai; host, parasite. The genus Tatnorix (saltcedar) contains approximately 54 species of phreatophytic plants whose origins are in Europe, Asia, and Africa. Several members of the genus were introduced into the United States in the early 1800s, mainly as ornamental plants. Approxi- mately 8 species have since escaped cultiva- tion and have become naturalized to varying degrees (Baum 1967). Tainarix ramosissima Ledeb. has become established in ripaiian areas throughout the West and Southwest, where it has proven to be an aggressive invader that eventually displaces native vegetation. Desert mistletoe {Phoradendron californiciim Nutt.) is a native parasitic plant that grows on several species of riparian plant hosts. Its range includes southern Nevada, southwestern Utah, southeastern California, southwestern Arizona, and northern Baja California, Sonora, and Sinaloa (Benson and Darrow 1981). Previously published information on hosts for desert mistletoe include Blumer (1910), Shreve and Wiggins (1964), Walters (1976), Daniel and Buttenvick (1992), and Overton (1992), none of whom mentions T. ramosissima. Holland et al. (1977) and Benson and Darrow (1981) state that "saltcedar" and "the introduced tamarisks" are possible hosts, while Munz and Keck (1965) and McDougall (1973) list Tamarix but men- tion no particular species. Cohan et al. (1978) state that P. californiciim does not occur in saltcedar This paper describes 2 occurrences of P. californiciim on T. ramosissima in south- ern Nevada. I found the 1st parasite and host specimen on 27 June 1995 at Hiko Springs in Clark County, Nevada, approximately 11 km west of Laugh'lin along State Highway 163 (3,894,000 N 711,650 E) at an elevation of 605 m (Fig. 1). A 2nd specimen was found on this host tree on 16 October 1995. Voucher specimens from 1 parasite and host are deposited in the Depart- ment of Biological Sciences herbarium. Uni- versity of Nevada, Las Vegas, accession num- ber 38971. The host tree was growing in a canyon approximately 2 m from a small, flowing stream on quartz monzonite-derived soil. The first mistletoe clump measured 33 cm long X 32 cm high X 14 cm wide and was growing on the southwest side of a branch 2.1 m above the ground. The branch to which the mistletoe was attached measured 5.2 cm in diameter and 16.2 cm in circumference. The length of the branch from trunk to point of mistletoe attachment was 2.1 m. The trunk base of the 5-m-high saltcedar measured 8 cm in diameter and 29 cm in circumference, which would indicate an age of approximately 24 yr (based on average value of California and Arizona sites as reported by Smith 1989). The 2nd mistletoe also faced southwest and was located on the main trunk of the tree .9 m above the ground. It was a newly sprouted plant that con- sisted of only 12 stems, the longest of which measured 4 cm. Both mistletoes and the host tree appeared to be healthy, actively growing specimens. The parasites were young plants and were a more vivid green than other mis- tletoes in the area. Sex of the mistletoes could not be determined. Other hosts for P. californiciim at this site include catclaw acacia [Acacia greggii), honey mesciuite (Prosopis glandidosa), and creosote bush {Larrea tridentata). Although many other Tamarix trees occur here, none ha\'e been in- fected by mistletoe. Desert mistletoe is usually spread from host to host by birds, which ingest the seeds and later defecate them onto a branch. Two bird species that occur frequently at this 'Departnient of BioloKical Sciences. 4505 Manlaml Parkuay, Box 454004. Las Veijas, NV 89154-4004. 186 1996] Notes 187 Fig. 1. Parasite Fhoradendron calijornicwn growing on host plant Tainarix rainosissima. site and have been seen feeding on misdetoe and perching in saltcedar are die Phainopepla [Phainopepla nitens) and Northern Mocking- bird {Miiniis polyglotfos) (personal obsei-vation). Acknowledgments I wish to thank Wesley Niles for help with identification of specimens and review of the manuscript, and Delbert Wiens who provided information on mistletoe hosts. This project was funded by a research grant provided by the Harry Reid Center for Environmental Studies, Las Vegas, Nevada. LiTERAiuHK Cited Halm, B. 1{. 1967. Introduci'd and iiatiiralizrd tamarisks in the United States and Canada Clliiuaricaeeae). Baileya 15: 19-25. Bi'NSON, L., .AND R. A. Dahhow. 1981. Trees and shnihs of the sonthvvesteni deserts. University of Arizona Press, Tucson. 416 pp. Blumer, J. C. 1910. Mistletoe in tlie Sonthwest. Plant World 13; 240-246. Cohan, D. R., B. W. Anderson, .•vnd R. II Ohmaut. 1978. Avian population responses to salt cedar along the Lower C:olorado River. Pages 371-382 in R. R. John- son and J. E McCormick, editors, Strategies for pro- tection and management of floodplain wetlands and other riparian ecosystems. USDA Forest Service, General Technical Report WO-12. Daniel, T. E, and M. L. Buttervvick. 1992. Flora of .south mountains of south-central Arizona. Desert Plants 10(3): 99-119. Holland, J. S., R. K. Grater, and D. H. Huntzinger. 1977. Flowering plants of the Lake Mead region. Southwest Parks and Monuments Association. Popu- lar Series No. 23. 49 pp. McDouGALL, W. B. 1973. Seed plants of northern Ari- zona. The Museum of Northern Arizona, Flagstaff. 594 pp. MUNZ, P A., and D. D. Keck. 1965. A California flora. University' of California Press, Berkeley. 1681 pp. 0\ ERTON, J. M. 1992. Host specialization in desert mistle- toe Phoradendnm califonucuiii. Bulletin of the Eco- logical Society of America 73: 293. Shreve, F, and I. L. Wiggins. 1964. Vegetation and flora of the Sonoran Desert. Volume I. Stanford Univer- sity Press, Palo Alto, CA. 840 pp. Smith, S. D. 1989. The ecology of saltcedar (Tamahx chi- nensis) in Death Valley National Moniunent and Lake Mead National Recreation Area: an assessment of techniques and monitoring for saltcedar control in the park system. Contribution CPSU/UNLV 041/03, National Park Service/University of Ne\ada, Las Vegas. 65 pp. Walters, J. W. 1976. A guide to misdetoes of Arizona and New Mexico. USDA Forest Service, Southwestern Region, Forest Insect and Disease Management. 7 pp. Received 6 November 1995 Accepted 4 March 1996 Great Basin Naturalist 56(2), © 1996, pp. 188-189 BOOK REVIEW Wild Plants of the Pueblo Province. Explor- ing Ancient and Enduring Uses. William W. Dunniire and Gail D. Tieniey. Foreword by Gary Paul Nabhan. Museum of New Mexico'Press, Santa Fe, NM. 1995. 290 pp. $19.95, softback. This book immediately appears field wor- thy and feels good in the hands. And that's simply judging the book by its cover! Once opened there is much to praise about this text. The authors have succeeded in putting together a wondei-fully interesting and well-wiitten field guide for the lay person as well as a useful ref- erence for serious students and professionals interested in ethnobotany of the Southwest. Within 9 chapters of text, an illustrated section involving about 73 plants, and an extensive chart summarizing plant uses, the reader learns of the ecology, representative flora, ethnobotany, and cultural history of the Pueblo Province. The original intent of the book was to provide a guide to commonly seen plants of Bandelier National Monument and the Pajarito Plateau in central New Mexico, and a discussion of the plants' prehistoric and recent uses. The authors have surpassed this goal. The 9 chapters reveal a cohesive and inter- esting histoiy of the people, plants, and land itself Ample information provides the reader insight as to how these elements interact and what the consequences of those interactions have been and continue to be. It is easy not only to move through the spatial and geo- graphical regions, but to enjoy a voyage in time as well and feel as if you were there. Line drawings, photographs, and maps lend addi- tional interest to the text. Although there is a great deal of information given about vegeta- tive zones, human history, and other topics, the authors have retained the importance of plants by referencing particular species wher- ever appropriate. The chapter on indicator species is particularly interesting and useful. This is a subject that few field guides address, and yet it is so easily applied and can be obsen'ed in the field when adequate informa- tion is provided. The main focus of the text is the center section that includes photographs and descriptions of 73 plants. Line drawings accompany each plant treated. The technical descriptions are somewhat brief, but the illus- trations provide enough detail that field iden- tification can be made easily in most cases. Perhaps one of the most valuable sections is the annotated plant list included at the end of the book. In an easy-to-read fonnat, a great deal of infomiation is concisely summarized for over 300 plants. The chart is subdivided into 7 gen- eral categories of plant use (i.e., food and bev- erage, medicine, constiiiction, etc.), with infor- mation given on how each plant is used by specific pueblos. The chart is well referenced and includes original citations for every use. A brief, yet well-organized analysis of the changes in plant utilization that occurred with the Spanish colonization in the Southwest is provided in chapter 3. The authors take a very complex histoiy and present it in the context of plant ecology. It provides an informative view of the ecological consec|uences of the collision of cultures. Gontemporary culture, plant use, and ecological modification are also included in this text. Two chapters provide insightful information on current cultural and ecological issues. Throughout the text, and reflected in the annotated plant list as well, the authors have attempted to treat religious and ceremonial plant uses with appropriate respect. An added benefit of the book is the authors' personal association with indi\'iduals in different Pueblo tribes. Their sense of respect and honor for these cultures is felt throughout the book. Our only complaint relating to this text is the lack of references citing specific informa- tion. It is quite difficult to identif\' references for much of the information included within the text, with the exception of the chapter dis- cussing indicator species. A bibliography with 188 1996] Book Review 189 145 references is included at the end of the book, but it is difficult to relate these refer- ences to particular chapters and specific infor- mation. This omission weakens the usefulness of the book as a reference lor serious students. It may be that the authors consciously omitted citations in an effort to allow the text to flow more easily, but it is a constant frustration when one is interested in identifying sources. A list of suggested reading is included at the end of each chapter, but no reference is given to original sources that support specific facts. In the preface, the authors do mention many sources that contribute in a general way. Overall, this book is one that should be included in a field book box, on the bookcase as a reference for plants and their uses by cul- tures of the Southwest, and in a travel file as it gives suggestions for specific hikes located in the Pueblo Province. For anyone interested in plant ecology, taxonomy, ethnobotany, cultural anthropology, or simply those with a general love for the Southwest, this book is highly rec- ommended. It is well written, informative, and aesthetically delightful. Renee Van Buren Kimberly Hamblin Hart Department of Botany and Range Science Brigham Young University Provo, UT 84602 University of Nevada, Reno, Department of Anthropology, Historic Preservation, Biological Resources Research Center, Divi- sion OF Continuing Education The University of Nevada, Reno, offers continuing education training courses in heritage resources management. Courses are designed for professionals in cul- tural and natural heritage management positions in the public and private sectors. The program is conducted in cooperation with the Advisory Council on Historic Preservation, the Bureau of Land Management, the National Park Service, and the U.S. Forest Service. The following information is offered on one of the upcoming courses. Ecosystem Management 30-31 May 1996 Reno, Nevada 9:00 a.m. -4:00 p.m. Fee: $250 Registration deadline: 2 May 1996 Instructor: Peter F. Brussard, Ph.D., is chairman of the biology department at the University of Nevada, Reno, and director of the Nevada Biodiversity Initiative, housed in the Biological Resources Research Center. Brussard is a founding mem- ber and past president of the Society of Conservation Biology and recognized as a leading authority in conservation biology and ecology. Ecosystem management is only recently beginning to be understood and used. This course will address the scientific basis for ecosystem management as well as the steps for managing areas so that biological diversity and ecosystem services remain conserved while human needs are also met. Ecosystem management focuses on systems as a whole rather than simply on the parts and involves the public in setting management goals. It represents a shift from linear comprehensive manage- ment to adaptive management. For further information, phone 1-702-784-4046 or fax 1-702-784-4801; to register, call 1-800-233-8929. INFORMATION FOR AUTHORS The Great Basin Naturalist welcomes previously unpublished manuscripts pertaining to the biologi- cal natural history of western North America. Preference will be given to concise manuscripts of up to 12,000 words. Simple species lists are dis- couraged. SUBMIT MANUSCRIPTS to Richard VV. Baumann, Editor, Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young University, Provo, UT 84602-0200. A cover letter accompanying the man- uscript must include phone number(s) of the author submitting the manuscript, and FAX number and E-mail address when applicable; the letter must also provide information describing the extent to which data, text, or illustrations have been used in other papers or books that are published, in press, submitted, or soon to be submitted elsewhere. Authors should adhere to the following guidelines; manuscripts not so prepared may be returned for revision. MANUSCRIPT PREPARATION. In general, the Great Basin Naturalist follows recommendations in Scientific Style and Format: The CBE Manual for Authors, Editors, and Publishers, 6th edition (Council of Biology Editors, Inc., II South LaSalle Street, Suite 1400,' Chicago, IL 60603, USA; phone 312-201-0101; FAX 312-201-0214). We do, however, differ in our ti'eatment of entries in Literature Cited. Authors may consult Vol. 51, No. 2 of this journal for specific instructions on format; these instruc- tions. Guidelines for Manuscripts Submitted TO THE Great Basin Naturalist, are printed at the back of the issue. Also, check the most recent issue of the Great Basin Naturalist for changes. TYPE AND DOUBLE SPACE all materials, including literature cited, table headings, and figure legends. Avoid hyphenated words at the right-hand margins. Underline words to be printed in italics. Use stan- dard bond (22x28 cm), leaving 2.5-cm margins on all sides. SUBMIT 3 COPIES of the manuscript and the origi- nal on a 5.25- or 3.5-inch disk utilizing WordPerfect 5.1 or above. Number all pages and assemble each copy separately: title page, abstract and key words, text, acknowledgments, literature cited, appendices, tables, figure legends, figures. TITLE PAGE includes an infomiative title no longer than 15 words, names and addresses of authors, a running head of fewer than 40 letters and spaces, footnotes to indicate change of address and author to whom correspondence should be addressed if other than the first author. ABSTRACT states the puipose, methods, results, and conclusions of the research. It is followed by 6-12 key words, listed in order of decreasing importance, to be used for indexing. TEXT has centered main headings printed in all capital letters; second-level headings are centered in upper- and lowercase letters; third-level head- ings begin paragraphs. VOUCHER SPECIMENS. Authors are encouraged to designate, properly prepare, label, and deposit high-(iuality voucher specimens and cultures docu- menting their research in an established permanent collection, and to cite the repository in publication. REFERENCES IN THE TEXT are cited by author and date: e.g., Martin (1989) or (Martin 1989). Multiple citations should be separated by commas and listed in chronological order. Use "et al." after name of first author for citations having more than two authors. ACKNOWLEDGMENTS, under a centered main heading, include special publication numbers when appropriate. LITERATURE CITED, also under a centered main heading, lists references alphabetically in the fol- lowing formats: Mack, G. D., and L. D. Flake. 1980. Habitat rela- tionships of waterfowl broods on South Dakota stock ponds. Journal of Wildlife Management 44: 695-700. Sousa, W P 1985. Disturbance and patch dynamics on rocky intertidal shores. Pages 101-124 in S. T. A. Pickett and P S. White, eds.. The ecolo- gy of natural disturbance and patch dynamics. Academic Press, New York. Coulson, R. N., and J. A. Witter. 1984. Forest ento- mology: ecology and management. John Wiley and Sons, Inc., New York. 669 pp. TABLES are double spaced on separate sheets and designed to fit the width of either a single column or a page. Use lowercase letters to indicate foot- notes. PHOTOCOPIES OF FIGURES are submitted initially with the manuscript; editors may suggest changes. Lettering on figures should be large enough to withstand reduction to one- or two-column width. Originals must be no larger than 22x28 cm. NOTES. If the manuscript would be more appro- priate as a short communication or note, follow the above instructions but do not include an abstract. A CHARGE of $50 per page is made for articles published; the rate for individual subscribers will be $35 per page. However, manuscripts with com- plex tables and/or numerous half-tones will be assessed an additional charge. Reprints may be pur- chased at the time of publication (an order form is sent with the proofs). FINAL CHECK: • Cover letter explaining any duplication of information and providing phone number(s), FAX number, and E-mail address • 3 copies of the manuscript and WordPerfect diskette • Conformity with instructions • Photocopies of illustrations GREAT BASIN NATURALIST (ISSN 001 7-361 4) Vol. 56, No. 2, April 1996 CONTENTS Articles Selecting wilderness areas to consei^ve Utah's biological diversity Diane W. Davidson, William D. Newmark, Jack W. Sites, Jr., Dennis K. Shiozawa, Eric A. Rickart, Kimball T. HaqDcr, and Robert B. Keiter 95 Nutrient distribution in Quercus gambelii stands in central Utah A. R. Tiedemann and W. E Clar\' 1 1 9 Comparsion of two roadside survey procedures for dwarf mistletoes on the Saw- tooth National Forest, Idaho Robert L. Mathiasen, James T. Hoffman, John C. Guyon, and Linda L. Wadleigh 1 29 Effects of Douglas-fir foliage age class on western spruce budworm oviposition choice and lai^val performance Kimberly A. Dodds, Karen M. Clancy, Kathryn J. Le>'\'a, David Greenberg, and Peter W. Price 1 35 Trypanoplasma atraria sp. n. (Kinetoplastida: Bodonidae) in fishes from the Sevier River drainage, Utah J. Stephen Cranney and Richard A. Heckmann 1 42 Geographical review of the historical and current status of Ospreys {Pandion haliaetus) in Utah Glark S. Monson 1 50 Effects of turbidity on feeding rates of Lahontan cutthroat trout {Oncorhynchus clarki henshawi) and Lahontan redside shiner {Richardsonius egregius) . . . Gary L. Vinyard and Andy C. Yuan 1 57 Pogonomyrmex owyheei nest site density and size on a minimally impacted site in central Oregon Peter T. Soule and Paul A. Knapp 1 62 Field measurements of alkalinity from lakes in the Uinta Mountains, Utah, 1956-1991 Dennis D. Austin 1 67 Density, biomass, and diversity of grasshoppers (Orthoptera: Acrididae) in a Cal- ifornia native grassland Eric E. Porter, Richard A. Redak, and H. Elizabeth Braker 1 72 Notes Summer nocturnal roost sites of Blue Grouse in northeastern Oregon Kenneth J. Popper, Eric C. Pelren, and John A. Crawford 177 Oochoristica scelopori (Cestoda: Linstowiidae) in a grassland population of the bunch grass lizard, Sceloporus scalaris (Phrynosomatidae), from Arizona . . . Stephen R. Goldberg, Charles R. Bursey, Chris T. McAllister, Hobart M. Smith, and Quynh A. Truong 1 80 Pocket gophers damage saltcedar {Tamarix ramosissinia) roots . . . Sara J. Manning, Brian L. Cashore, and Joseph M. Szewczak 1 83 Saltcedar {Tamarix ramosissima), an uncommon host for desert mistletoe {Phora- dendron calif ornicum) Sandra L. Haigh 1 86 Book Review Wild plants of the Pueblo Province. Exploring ancient and enduring uses. William W. Dunmire ami Gail D. Tienwy Renee Van Buren and Kimberly Hamblin Hart 1 88 H E Sep U 5 1996 HARVARD UNIVERSITY GREAT BASIN NATURALIST VOLUME 56 Ne 3 — JULY 1996 BRIGHAM YOUNG UNIVERSITY GREAT BASIN NATURALIST Editor Assistant Editor Richard W. Baumann Nathan M. Smith 290 MLBM 190 MLBM PO Box 20200 PO Box 26879 Brigham Young University Brigham Young University Provo, UT 84602-0200 Provo, UT 84602-6879 801-378-5053 801-378-6688 FAX 801-378-3733 E-mail: NMS@HBLL1.BYU.EDU Associate Editors Michael A. Bovvers Paul C. Marsh Blandy Experimental Farm, University of Center for Environmental Studies, Arizona Virginia, Box 175, Boyce, VA 22620 State University, Tempe, AZ 85287 J. R. Callahan Stanley D. Smith Museum of Southwestern Biology, University of Department of Biology New Mexico, Albuquerque, NM University of Nevada-Las Vegas Mailing address: Box 3140, Hemet, CA 92546 Las Vegas, NV 89154-4004 Jeffrey J. Johansen Paul T. Tueller Department of Biology, John CaiToll University Department of Environmental Resource Sciences University Heights, OH 44118 University of Nevada-Reno, 1000 Vallev Road BOBISCKONDRATIEFF Reno. NV 89512 Department of Entomology, Colorado State Robert C. Whitmore University, Fort Collins, CO 80523 Division of Forestry, Box 6125, West Virginia University, Morgantown, WV 26506-6125 Editorial Board. Jerran T. Flinders, Chairman, Botany and Range Science; Duke S. Rogers, Zoology; Wilford M. Hess, Botany and Range Science; Richard R. Tolman, Zoology. All are at Brigham Young University. Ex Officio Editorial Board members include Steven L. Taylor, College of Biology and Agriculture; H. Duane Smith, Director Monte L. Bean Life Science Museum; Richard W. Baumann, Editor, Great Basin Naturalist. The Great Basin Natwalist, founded in 1939, is published quarterly by Brigham Young University. Unpublished manuscripts that further our biological understanding of the Great Basin and surrounding areas in western North America are accepted for publication. Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1996 are $25 for individual sub- scribers ($30 outside the United States) and $50 for institutions. The price of single issues is $12. All back issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other busi- ness should be directed to the Editor Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young University, Provo, UT 84602-0200. Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin Naturalist through a continuing exchange of scholarly publications should contact the Exchange Librarian, 6385 HBLL, PO Box 26889, Brigham Young University, Provo, UT 84602-6889. Editorial Production Staff JoAnne Abel Technical Editor Jan Spencer Assistant to the Editor Copyright © 1996 hv Brigham Young University ISSN 0017-3614 Official publication date: 26 July 1996 7-96 750 19016 The Great Basin Naturalist Published at Pkovo, Utah, by Brigham Young University ISSN 0017-3614 Volume 56 31 July 1996 No. 3 Great Basin Naturalist 56(3), © 1996, pp. 191-196 BIOGEOGRAPHIC SIGNIFICANCE OF LOW-ELEVATION RECORDS FOR NEOTOMA CINEREA FROM THE NORTHERN BONNEVILLE BASIN, UTAH Donald K. Grayson^ Stephanie D. Livingston^, Eric Rickart'^, and Monson W. Shaver III^ Abstract. — The existence of low-elevation populations of Neotoma cinerea in the northern Bonneville Basin shows either that these mammals can survive many thousands of years in xeric habitats or that the\' can mo\'e across xeric low- lands far more readih' than has been appreciated, or both. Current models of Great Basin small mammal biogeography are far too static to encompass properly the interaction of the wide range of geographical and biological variabilit>' that has produced the modem distribution of those mammals that have, for several decades, been treated as "montane" within the Great Basin. Key words: Great Basin, biogeography, island biogeography, Neotoma cinerea, mammals. Ever since J. H. Brown's insightful analyses coherence that has been assigned to it. Here, of Great Basin small mammal biogeography we add to that growing body and call for a (Brown 1971, 1978, see also Lomolino et al. more dynamic view of Great Basin small mani- 1989), biogeographers have treated the bushy- mal historic biogeography. tailed woodrat (Neotoma cinerea) as a member of an assemblage of small mammals that is cur- Neotoma cinerea ON rently isolated on Great Basin mountains. The Homestead Knoll, Utah composition of this assemblage is of particular importance because it has been used to gener- Located a few km west and south of Great ate and test hypotheses about the past and Salt Lake in north central Utah, the Lakeside fiiture of Great Basin "montane" mammals (e.g.. Mountains are formed fiom a complex of north- Grayson 1987, 1993, Patterson 1990, Cutler trending hills, ridges, knolls, and small moun- 1991, McDonald and Brown 1992, Murphy and tains (Fig. 1). The northwestern-most spur of Weiss 1992, Grayson and Livingston 1993). this complex is Homestead Knoll, a low (maxi- However, there is a growing body of data that mum elevation 1615 m), rocky promontory suggests that this group of mammals lacks the that is devoid of active springs and permanent i Burke .Vlemorial Museum. Box .353010, Universih' of Washington, Seattle. WA 9819.5. ^Quaternary Sciences Center, Desert Research Institute, Box 60220, Reno, NV 8950(1 3Utah Museum of Natural History, Universit\' of Utah, Salt Lake City, UT 84112. ■»Utah Geological Sui^ey 2363 South FooothiU Drive, Salt Lake City, UT 84109. 191 192 Great Basin Naturalist [Volume 56 Fig. 1. Location of Homestead Cave within the northern Bonneville Basin. streams, and that is separated from other parts of the Lakeside group by valleys whose maxi- mum elevations do not exceed 1465 m. The barren playa of Pleistocene Lake Bon- neville is located to the immediate west and northwest of Homestead Knoll. Vegetation of the knoll is dominated by shrubs and grasses, aldiough there are a few scattered Utali junipers {Juniperiis osteospenua) on its highest reaches. Most prominent among the shrubs are Atriplex confertijolia, Tetradijinia spinosa, and Tetradymia glabrata. Artemisia tridentata is present along seasonally moist drainages, while Aiii'inisia spincscens, ChrysntJiainniis sp., and Sarcuhatus vcnniciilatiis are present but un- common above the flanks of die knoll. Artemisia nova occurs on those flanks as does Ceratoidcs lanata, while S. vermieiilatiis becomes increas- ingly abundant as the \alley bottoms are approached. We made no attempt to identify the grasses that form the understoiy beneath the shrubs, but cheatgrass {Bromiis teetormn and, perhaps, B. ruhe)is) is extremely abundant on the flats beneath the knoll. Homestead Knoll is dotted b\' a number of caves, one of which. Homestead Cave, sits on the northwestern edge of the knoll at an eleva- tion of 1406 m (Fig. 2). Approximately 11 m wide and 6 m high at its mouth, this 25-m- deep cave has, since 1992, been the focus of interdisciplinaiy paleoecological work funded by the Department of Defense. With D. B. Madsen of the Utah Geological Surve\', 3 auUiors of this paper (DKG, SDL, and MWS) have been involved with the excavation and analysis of a deep sequence of vertebrate remains from this site. To provide background data for the anaKsis of the mammalian compo- nent of the excax ated fauna, we conducted a brief (270 trap-night) small mammal suney in the vicinit\' of Homestead Ca\'e in June 1995. With 1 exception, the residts of this sui^vey were quite predictable. Trapping success was low, with 3 species — Dipodomijs ordii (3 indi- viduals), Peromyseus immieulatiis (11 individu- als), and Neotoma lepida (6 iudi\ iduals) — com- prising nearly the entire trapped assemblage. The 1 exception, howexer, was remarkable: we 1996] LOVV-ELEVAIION NeOTOMA CINEREA 193 Fig. 2. Location of Homestead Cave (white anow) on Homestead Knoll; the prominent tenaces represent Proxo, post- Provo regressive, and Stansbuiy beaclies left by the waters of Pleistocene Lake Boniiex ille. took a single Neotoma cinerea from the back of Homestead Cave itself Because this individual was live-trapped and released, we cannot report its age or sex or provide standard measurements. Even though we do not have a voucher specimen, we do have an excellent videotape of the animal (taken by MWS and on file at the Utah Geological Sui-vey), and there is no doubt as to the identi- fication of the individual. Vegetation in the immediate vicinity of Homestead Cave departs from the Homestead Knoll vegetation that we have described in only 1 major way: the mouth of the cave sup- ports a luxuriant growth of Rihes cercum immediately beneath the dripline. It would be surprising if this shrub were not heavily uti- lized by both Neotoma cinerea, taken at the rear of the cave, and Neotoma lepida, taken at the front. Other Low-elevation Northern Bonneville Basin Records FOR Neotoma cinerea Our discover)' of Neotoma cinerea on Home- stead Knoll led us to search the mammal col- lection at the Utah Museum of Natural His- toiy, University' of Utah, for additional records of this species from other low-elevation set- tings in the northern Bonneville Basin. We were (juite successful in this search: (a) Locomotive Springs: The only pre\ iousK published low-elevation record for Ne()t())iia cinerea for the northern Bonneville Basin was provided by Durrant (1952:348; UU 5048) as having been taken in October 1947 fi-om "State- house, Locomotive Springs, 5500 ft. [1676 in]." However, we are unable to determine the loca- tion of "Statehouse" and are othenvise hesitant to accept this record because of the substantial difference between the actual elevation of Locomotive Springs (1283 m) and the reported elevation of "Statehouse" (1676 m). Given the well-watered nature of Locomoti\e Springs, the record might be accurate, but it is in need of verification. Locomotive Springs is approxi- mately 60 km north of Homestead Knoll. (b) Lakeside Mountains: A\\ adult male Neo- toma cinerea (UU 14374) was collected "5 mi. E Lakeside, 4600 ft. [1402 m]" in June 1957. This distance and direction fi-om Lakeside, how- ever, describe a point in the Great Salt Lake. If 194 Great Basin Naturalist [Volume 56 the actual direction were southeast, the speci- men could have come fi-om Cave Ridge on the eastern edge of the Lakeside Mountains, approximately 10 km east of Homestead Knoll. (c) Newfoundland Mountains: A series of three juvenile Neotoina cinerea (UU 9995, 9996, 9998) were collected in June 1951 from an unspecified site at the north end of the Newfoundland Mountains. The collector s field notes do not provide the elevation of the site but do indicate that the specimens came from an area of granite cliffs with a plant communitv' that included Juniperus and Tetrodyinia. The north end of the Ne\\^oundland Mountains is approximately 40 km west-northwest of Home- stead Knoll. (d) Cedar Mountains: There are records for Neotoina cinerea from 2 separate locations in the southern Cedar Mountains: 4 from the Cane Springs area (elevation 1768 m; UU 26340, 27297, 27299, and 27301-2, collected between October 1952 and Januaiy 1953), and 1 from the "south end Cedar Mtn., 4850 ft. [1478 m]." This last specimen is reported to have been caught in a garage, suggesting that it may have come from near Dugway. Although these specimens come from no closer than 95 km to the south of Homestead Cave, we men- tion them because they establish the likelihood that Neotoma cinerea occurs in suitable habitat throughout the Cedar Range. BlOGEOGRAPHIC CONSIDERATIONS Although Neotoma cinerea has frequently been treated as being isolated on Great Basin mountains (Brown 1971, 1978, Grayson 1993), these records demonstrate that bushy-tailed woodrats can and do exist at low elevations in arid contexts within at least the northern Bon- neville Basin. How, one must wonder, did Neotoma cinerea come to occupy such arid, low-elevation settings as the Newfoundland Mountains (maximum elevation 2130 m) and isolated knolls on the Lakeside Mountains (maximum elevation 2020 m)? It is well established that during the late Pleistocene, bushy-tailed woodrats were far more widely distributed within the Great Basin than they are today, occupying low-elevation settings where they are no longer found (Gray- son 1988, 1993). As a result, it is reasonable to speculate that these animals were also wide- spread in this part of the northern Bonneville Basin during those years. We can, however, do much more than speculate about the histoiy of N. cinerea in the Homestead Knoll area. Widi a maximum elevation of 1615 m. Home- stead Knoll was covered by the waters of Pleis- tocene Lake Bonneville 14,500 years B.E, when Pleistocene Lake Bonneville was at its high (see Figure 2). Obviously, Homestead Knoll must have received its woodrats after this time, but when this occurred is not clear. Between 14,500 and at least 14,200 years B.P, when Lake Bonneville stood at the Provo level, Home- stead Knoll was an island of approximately 770 acres. Not until Lake Bonneville fell to a local elevation of 1463 m did this island become connected to the main body of the Lakeside Mountains. Once this occurred. Homestead Knoll became part of the faunal mainland and would have been open to overland colonization by terrestrial mammals. Unfortunately, we do not know when the lake fell to this level. However, we do have direct evidence from Homestead Cave con- cerning the regional history of Neotoma cinerea. E.xcavations in this cave have provided a rich, stratified sequence of vertebrate remains, the mammals of which are being identified and analyzed by one of the authors (DKG). To date, a substantial sample of mammal specimens from the 4 lowest Homestead Cave strata has been identified (37,381 specimens). All 4 assemblages contain both N. cinerea and N. lepida, but the ratio of N. cinerea to N. lepida xaries dramatically through time. In stratum I, which dates to between ca 11,300 and 10,000 years B.P, bush\'-tailed woodrats make up 99.38% of the Neotoma fauna. In sub- sequent strata, however, they decline steadily in abundance; by stratum IV (ca 8200-7200 years B.P), N. cinerea comprises only 4.74% of the Neotoma assemblage (Fig. 3). Similarh; N. cinerea contributes 23.97% of die total number of identified mammalian specimens in stratum I, a number that declines to 1.01% in stratum IV (Fig. 4). The Homestead Cave fauna thus documents that N. cinerea was present in the Homestead Knoll area by 11,300 years B.P and remained a common species in the small mammal fauna through much of the EarK Ilolocene. After ca 8200 years B.P, however, N. lepida became the ovenvhelmingly dominant member of the genus, and N. cinerea became localK rare. Since mam- mals from later strata within Homestead Cave 1996] Low-elevation Neotoma cim:ria 195 N = 1933 00 99.38 - N = 1564 90^ 86.13 80- 70- 60- N = 564 50- 46.81 40- 30- s o o 20- 10- § O o s 5 N = 4392 g 4.74 ° ■D -f ■0 |!-:;;ii.K ■.;?;!;] "D 1 II III IV ST RATL M Fig. 3. Changing contribution of N. cinerea to the Neotoma {N. cinerea phis N. lepida) fauna. Homestead Cave strata I-IV (N = total number of Neotoma specimens identified to the species level, including those identified as N. of cinerea and N. cf lepida). have not yet been completely identified, we do not know whether N. cinerea sui-vived the veiy xeric Middle Holocene (ca 7500-5000 years B.E) here. Currently there are 2 options for explaining the modern existence of N. cinerea on Home- stead Knoll. First, animals living here today may be direct descendants of the initial woodrat colonizers of the knoll, colonizers that arrived sometime between 14,500 and 11,300 years B.E If so, the population has survived even though its numbers dropped precipitously toward the end of the Early Holocene (ca 8200-7200 years B.E), and presumably fell even further during the heart of the Middle Holocene. Assuming that N. cinerea does not now survive in the val- leys diat separate Homestead Knoll fi-om nearby uplands, and that it has not been able to sur- vive in those valleys since at least 7000 years B.E, then this population has existed on an iso- lated upland a few thousand acres in extent for a minimum of some 7 millennia. The other, and certainly more likely, option is that Neotoma cinerea has not been isolated on Homestead Knoll for this entire period of time, that populations on the knoll have been augmented by immigrants from elsewhere, and that any local extinctions of N. cinerea on the knoll have been followed by recolonizations from nearby populations. Indeed, it is even possible that the current representatives of the species colonized Homestead Knoll during the mid-1980s, a time of extraordinarily high pre- cipitation in the northern Great Basin (Amow and Stephens 1990). 30- N - 1912 25- 23.97 20- 15- 10- I 5- i ? 'u 1347 83 N - 1 1 264 ,09 N - 208 I 1.01 w Fig. 4. Changing contribution of Neotoma cinerea to the total number of identified mammalian specimens (\1SP) per stratum at Homestead Cave, strata I-IV. Implications The discovery o( Neotoma cinerea on Home- stead Knoll does not simply represent an unex- pected natural historical tidbit. Our discoven- documents either that populations of Neotoma cinerea within the Great Basin can find suffi- cient refuge in low-elevation, xeric habitats to survive for many thousands of years, or that this species can move across xeric lowlands far more readily than has been appreciated, or both. Indeed, insofar as bushy-tailed woodrats are more effective colonizers than has been realized, an effective parallel may exist in the yellow-nosed cotton rat {Sigmodon ochrogna- thus), a "montane" mammal of the Southwest that has apparently expanded its range across low-elevation valleys dining the past 50 years (Davis and Dunford 1987; see also Davis and Callahan [1992] on Microtus mexicanus). Elsewhere, Grayson and Livingston (1993) have noted that Sylvilagus nuttallii can cross valley bottoms in at least parts of the Great Basin. Now, it seems that N. cinerea can sur- vive in habitats that are anything but montane. This fact leads us to suggest that the nested- ness of Great Basin mammal faunas (sensu Fat- terson and Atmar 1986, Fatterson 1987, 1990) might reflect a combination of extinction histo- ries and colonization abilities. In addition, die Homestead Knoll record for N. cinerea takes its place alongside other recent data docu- menting that current models of Great Basin small mammal biogeography are far too static to encompass properly the wide range of geo- graphical and biological variability that has produced the modern distribution of those 196 Great Basin Natur.\list [Volume 56 mammals that, for several decades, have been treated as "montane" within the Great Basin (e.g., Gravson 1993, Grayson and Livingston 1993, Lawlor 1995, Rickart 1995). In the South- west, modern montane mammal distributions have clearly been determined by a complex combination of Holocene extinctions and colonizations (e.g., Davis and Dunford 1987, Lomohno et al. 1989, Davis and Ctilldian 1992). It now appears that the situation in the Great Basin is quite similar Acknowledgments The research reported here was supported by a grant fi-om the U.S. Department of Defense Legacy Program (Project #0304843028X728, "Paleoenvironmental Change on Hill Air Force Base and Dugway Proving Grounds"). Our thanks to D. B. Madsen for assistance at all stages of this project and to R. S. Thompson for confinning plant identifications. We also thank R. Davis, D. B. Madsen, B. D. Patterson, and T A. Vaughan for helpful comments on a draft of this paper Literature Gited Arnow, T, and D. Stephens. 1990. Hydrologic character- istics of the Great Salt Lake, Utah: 1847-1986. U.S. Geological Siii"vey Water-SuppIy Paper 2332. Brown, J. H. 1971. Mammals on mountaintops: nonequi- libriiiin insular biogeography. American Naturalist 10.5:467-178. . 1978. The theoiy of insular biogeograph\- and the distribution ot boreal birds and mammals. Pages 209-227 in K. T Harper and J. L. Reveal, editors, Intermountain biogeography: a symposium. Great Basin Naturalist Memoirs 2. Cutler, A. 1991. Nested faunas and extinction in frag- mented habitats. Consei^vation Biology 5: 49(i-.50.5. Davis, R., and J. R. Callahan. 1992. Post-Pleistocene dis- persal in the Mexican vole (Microfiis inexicanus): an example of an apparent trend in the distribution of southwestern mammals. (Jreat Basin Naturalist .52: 262-268. Davis, R., and C. Dunfokd. 1987. An example of contem- porary colonization of montane islands by small, non- flying mammals in the American Southwest. Ameri- can Naturalist 129: 398-406. Durrant, S. D. 1952. Mammals of Utah: taxonom\' and distribution. Universit> of Kansas Publications, Museum of Natural Histoiy 6. Grayson, D. K. 1987. The biogeographic histoiy of small mammals in the Great Basin: observations on the last 20,000 years. Journal of Mammalogy 68: 359-375. . 1988. Danger Ca\'e, Last Supper Cave, Hanging Rock Shelter; the faimas. American Museum of Nat- ural Histoiy Anthropological Papers 66. . 1993. The deserts' past: a natural prehiston' of the Great Basin. Smithsonian Institution Press, Washing- ton, DC. Gr.\yson, D. K., and S. D. Livinc.ston. 1993. Missing mammals on Great Basin mountains: Holocene extinc- tions and inadequate knowledge. Conservation Biol- ogy' 7: 527-532. Lawlor, T. 1995. Biogeography of Great Basin mannnals: paradigm lost. Unpublished manuscript on file at the Department of Biological Sciences, Hiunboldt State University, Aicata, CA. LoMOLiNO, M. v., J. H. Brown, and R. Dams. 1989. Island biogeography of montane forest mammals in the American Southwest. Ecology 70; 180-194. McDonald, K. A., and J. H. Brown. 1992. Using mon- tane mammals to model extinctions due to global change. Consen'ation Biolog\' 6; 409—115. Mlhphy, D. D., and Weiss, S. B. 1992. Effects of climate change on biological diversity in western North America: species losses and mechanisms. Pages 355-368 in R. L. Peters and T. E. Lovejoy, editors. Global warming and biological di\'ersit\-. Yale Univ er- sity Press, New Haven, CT. PATfERSON, B. D. 1987. The principle of nested subsets and its implications for biological consenation. Con- sen ation Biologv' 1: 32.3-^334. . 1990. On the temporal development of nested subset patterns of species composition. Oikos 59: 330-342. Patterson, B. D., and W. Atmar. 1986. Nested subsets and the structure of insular mammalian faimas and archipelagos. Biological lournal of the Linnean Soci- ety' 28; 65-82. Rickart, E. A. 1995. Ele\ational diversity gradients, bio- geography', and the structure of montane mammal communities in the intermountain region. Unpub- lished manuscript on file at the Utah Museum of Natural Histon, Unixersity of Utah, Salt Lake City. Received 15 Novcmlycr 1995 Accepted 20 March 1996 (Jrt-at Basin Naturalist 56(3), © 199(i, pp. 197-204 SYNOPSIS OF THE MOSSES OF WYOMING PM. Eckel' Abstkact. — A it'\ isrd list ol the mosses oi tlic Stale ol Wyoming is prescntccl. Ik'corck'd are 315 species and \arieties. Kci/ words: hn/dplit/tcs. \Vy()inin<:,. rortcr. Rocky Moiinlai)t.s. citcrklisl. flora. Publication of the mosses of Wyoming began with A\en Nelson (1900) listing 119 species. His collections were made essentially in the Laramie and Medicine Bow ranges of Carbon and Albany counties, and his specimens are now at the Rock)' Mountain Herbarium (RM). Nelson s specimens were determined in large part by Professor John M. Holzinger, of the State Normal School at Winona, Minnesota. Aven Nelson s and Elias Nelson's Wyoming col- lections were distributed under printed labels as "Plants of Wyoming from the Rocky Moun- tain Herbarium" and "Plants of Yellowstone National Park from the Rock)' Mountain Herb- ariiun,' citing Holzinger and a few others as determiners. However, because they were issued without serial numbers, showing only the collectors' numbers together with other data, they do not constitute true exsiccatae (Sayre 1971). In 1935 Cedric Lambert Porter published a valuable checklist of the mosses of the State of Wyoming, citing 215 species and varieties. In this publication Porter mentioned a paper by Dwight C. Smiley on the mosses of Yellow- stone National Park, which Porter cited again two years later as "A Key to the Mosses of Yel- lowstone National Park ' [unpublished] (Porter 1937). Porter included Smiley's names but apparently did not examine his specimens. The substance of a dissertation written by Porter at the University of Washington in 1937 was the development of a useful key to the hepatic and bryophyte taxa of Wyoming. It includes additional county records, a few addi- tional species, and references to Yellowstone National Park, again apparently citing Smiley's unpublished material. On a recent visit to the Yellowstone National Park Herbarium (YE LLC), I was able to locate and borrow for stud) some of Mr Smile\' s specimens. Several were deter- mined 1)\ R. S. Williams. No checklist focusing on the moss flora of Yellowstone National Park has as vet been pub- lished. The 357 or so moss specimens curated at YE LLC all derive from the park. The fol- lowing are collectors and dates of collecting activity: Dwight Smiley, 1932; H. S. Conard, 1948; Winona Welch, 1951; Eula Whitehouse, 1951; EKa Lawton, 1953. No biographical data exist on Mr. Smiley at YELLO (Whipple per- sonal commimication). I have no other record that these collectors worked in Wyoming, but my use of various herbaria for this checklist is not exliaustive. Incidentally, there are 11 pack- ets of livei^worts at YELLO, and no representa- tion of Spljagnuin. Recent collectors in the state include the late Fredrick Hermann, Holmes Rolston, and William Weber, all of whose Wyoming speci- mens are distributed in various herbaria, espe- cially COLO. Steven Churchill (1979, 1982), John Spence (1985), and Alvin L. Medina (1994) collected additional taxa. Two taxa listed here as new to the state were recently collected by Joseph Elliott {ScorpidiiDn scorpioides and Cinclidium stygium). General references to ex- tensive collections made in the state are given by Lawton (1971). The purpose of this paper is to present an up-to-date list of the mosses of the State of Wyoming, incorporating reports published since Porter's manuscripts as well as additional unpublished infonnation. There are 315 species and varieties in the present list. Some idea of the degree of representation of the flora of Wyoming comprising this list may be inferred from a glance at similar checklists for other 'Clinton Herliarium. BufSilo \Iiis£'uni of Science, BnflTalo, NY 14214. 197 198 Great Basin Natumlist [Volume 56 ai-eas. It is probabK to be expected that the arid intermountain states, such as Utah and Nevada, will have a more depauperate flora; Spence (1988) recorded 342 species for the entire Intermountain West. The following counts are for species and varieties (except Utah which is species only): Arizona 381 (Johnsen, no date), Colorado 292 (Weber 1973), Idaho 257 (McCleaiy and Green 1971), Mon- tana 358 (Eversman and Shaip 1980), Nevada 165 (Lawton 1958, Lavin 1981), Oregon 441 (Christ)- et al. 1982), Utiili 256 (Flowers 1973). There appears to be no checklist for the state of Washington. New \brk State, which is said to have a diverse moss flora, has 503 species and varieties (Ketchledge 1980). A striking comparison is to the oceanic island of New- foundland, which boasts a moss flora of 445 species (Brassard 1983) and which is geoph>s- iographicalK rather plain compared to tlie geo- moiphic extremes and di\ersit>' of \\yoming in tlie Central Rock> Mountains. The following checklist is based largely on a rexiew of specimens housed in the Rock> Mountain Herbarium. Additional herbaria were contacted in instances of taxa reported in die literature but witli no representation at RM, and nimierous new records ha\e been added from field collections b>" m>self and others. I have attempted to cite at least one reliable specimen of each taxon by giving the abbreviation of tlie herbarium at which the specimen is located. Additional tiixa are added from Porter's 1937 dissertation if die\' did not occur in his previ- ous publication. If a herbarium designation is noted below for a ta.xon, no further reference to the literature is given. Although sexeral ref- erences to the same species throughout the lit- eratiue ma\ ha\e been cited, only one citation is presented for names for which specimens ha\e not been seen. Nomenclature other than Spliagninn follow s Anderson et al. (1990). Sphagninn nomencla- ture follows Anderson (1990). Families are gi\en alphabetically, genera alphabeticalK widiin families, and species iilpha- beticalK within genera. Checklist of Mosses of Wyomlxg Amblysteciaceae CaUu'rTa (RM. WTU) Palustriclla decipiens (De Not.) Ochyra (RM. WTU) Pseudoculliergon turgescens (T. Jens.) Loeske (COLO, RM) Sanionia uncinata (Hedw.) Loeske (NY^ RM) Sanncnthypnuni sannentosum (Wahlenb.) Tuom. &: T. Kop. (RING. COLO, RM) Scorpidiuni scorpioides (Hedw.) Limpr. (BUE) Wanistoifia cxannulaia (Schimp. in BSG) Loeske \ar. exannulata (CSU. RM) Warnstoifia jhiitans (Hedw.) Loeske (COLO, RM) .\\DREAEACE.\E Andrcaca ntpesfris Hedw. (YELLO) AUL.\C0M-\UCE.\E Aulacomnium androgynwn (Hedw.) Schwaegr. (TENN) Aulacouiniuni pahistre (Hedw.) Schwaegr. (BUE RM> TENN) Bartfl\miace.\.e Anacolia nienziesii (Turn.) Par (WTU) Bartramia ithyphylla Brid. (RM) Philonotis fontana (Hedw.) Brid. \-dv. fontana (BUE R.M) \ar americana (Dism.) Flow. (BUF RM) \ar cacspitosa (Jun) Schimp. (BUE RM) \ ar putnila (Tum.) Brid. (RM) Amblystegium serpens (Hedw.) Schimp. \ar serpens (NY. RM) va.r.juratzka)utin (Schimp.) Ran & Hen-. (RM) Amblystegium cariuin (Hedw.) Lindb. (NY. RM) BtvU:inTHECI.\CE.\E Brachythccium acutum (Mitt.) Sull. (RM) Brachythcciuni (dhican.s (Hedw.) Schimp. in BSG (RM) 19961 Mosses of Wyoming 199 Brdclu/tluciiiin colliniim (Sclik'icli. c.v (J. MiR'll.) Scliiiiip. //( BS(;(BUFNY, RM) Bracltythecititn enjthrorrhizon Sciiimp. in BSU (RM) Bracliijtlu'ciitmfendh'h (Sull.) Jaeg. (RM) Braclujtlu'ciumfri^idum (C. Miiell.) Besch. (NY, US) Brachythecitim leiij'-^ii Grout (NY, RM, US) Brachytlu'ciwn nelsonii Cioiif "''SU, NY, RM) Brachjtheciwn ocdipodiwn (Mitt I Tae?. (COLO, RM, US) Brachijthcciuin rivularc Sdiinip. in BSG (NY, RM, US) Bracliythcciiitn ndalndum (lli'tlw.) Schimp. in BSG (NY) Brachytlteciuni salebrosuin (Wch. & Mohr) Scluin]). in BSG (BUR RM, US) Brachytlu'ciwn starkci (Brid.) Schimp. in BSG (Porter 1935) Brachythcciian twnidwn (C.J. Hartin.) Kincll). (COLO, NY, RM, US) Bnichythcciwn velutinum (Hedw.) Schimp. in BSG var. velutinum (Spence 1985) var. venustum (De Not.) Arc. (BUR RM) CiniphyUum cirroswn (Schvvaegr. in Schiiltes) Grout (NY) Eurhynchium ureganum (Sull.) Jaeg. (Spence 1985) Eurhynchium pulchcUum (Hedw.) Jenn. (RM) Homidothccium acneuin (.Mitt.) Lawt. (RM, US) Honmlotheciwn nevadense (Lesq.) Ren. & Card. (US) Homcdothecium pinnatifidum (Sull. & Lesq.) Lawt. (BUF, RM) Tonientypnwn nitens (Hedw.) Loeske (COLO, RM) BRVACEAE Brywn alsis australasiac (Grev. & Hook.) Robins. (BING, BUF) PFERVGY-NANDRACEAE Hetcrocladium diinori>lium (Brid.) Schimp. in BSG (BUF) Myurclla julacca (Schwaegr) Schimji. in BSG (RM) RllVTIDIACEAE Rlujtidiuin rugosiim (Hedw.) Kiiidb. (WTU) Selk;eriaceae Blindia acuta (Hedw.) Brnch & Schimp. in BSG (COLO, CSU, RM) Seligcria campylopoda Kindb. in Macoun & Kindb. (ALT\, BUERM) Sphagnaceae Si)hagnum angustifolium (C. Jens, t'.v Russ.) C. Jens, in Tolf (COLO, RM) Sphagnutn annulatum H. Lindb. ex Wanist. (BING, RM) Sphagnum contoiium Scliultz (BING) Sphagnuni find)riatum W'ils. in Wils. & Hook. f. (COLO, RM) Sphag)iU)nfuscum (Schimp.) Klinggr. (RM) Sphagnum platyphyllum (Lindb. ex Braithw.) Sull. t'.v Wamst. (RM) Sphagnum russowii Wamst. (RM) Sphagnum squarrosum Crome (RM) Sphagnum subsecundum Nees in Sturm (BING) Sphagnum teres (Schimp.) Aongstr in Hartni. (BING) Sphagnum uarnstoifii Russ. (RM, COLO) Spl.achn.\ceae SjylachiHun splmcricum Hedw. (Y'ELLO) Tayloria acuminata Hornsch. (Cram and Anderson 1981) Tayloria ligulata (Dicks.) Lindb. (COLO, RM) Tayloria serrata (Hedw.) Bruch & Schinip. in BSG (NY) Tetr.\phii:)aceae Tetraphis pcllucida Hedw. (RM) Thuiduceae Abictinclla ahiclina (Hedw.) Eleisch. (BUE RM) Tl.MMI.ACEAE Timmia austriaca Hedw. (BUE RM) 202 Great Basin Naturalist [Volume 56 Timmia megapolitana Hedw. var. megapoUtana (NY, RM) var. bavarica (Hessl.) Brid. (BUF; RM) Problematic Taxa Brachythecimn campestre (C. Muell.) Schimp. in BSG. Old Faithful, Yellowstone National Park, Smiley, according to Porter (1934, 1935). Porter's citation is apparently not based on a specimen but on Dwight Smiley's checklist. Since this is primarily a taxon of the eastern United States, it should probably be excluded from the state flora. Specimen not seen. Brachythecium oxycladon (Brid.) Jaeg. This taxon, typical of the eastern United States, is based by Porter (1934, 1935) on a citation by Smiley. No coiTcsponding specimens were seen. Funaria flavicans Michx. No specimens were seen of this taxon reported by Porter (1935). As it appears to be a species of the eastern region of the United States (Crum and Anderson 1981), it is of doubtful occurrence in Wyoming. Mniiim honiiim Hedw. (Porter 1935). This is a taxon of the eastern montane region of North America and the Piedmont (Crum and Ander- son 1981) and not likely to occur in Wyoming. No specimen seen. Platydictya confervoides (Brid.) Ciimi. Cited by Porter (1935) as a doubtful determination; it cannot be located in the herbaria consulted. Rocomitriumfosciculare (Hedw.) Brid. Porter (1935) did not see a specimen of this species, reported by Nelson (1900). Spence's citation for Teton County (1985) refers to Porter's doubtful citation. No specimens were seen by the present author Excluded Taxa Brachythecium calcarewn Kindb. A specimen of Smiley's o{ Brachytheciwn flexicaule, now B. calcarewn, from Yellowstone National Park and cited by Porter (1935) was determined as a depauperate specimen o( B. fiigichnn. Brachythecium glareosum (Br) B. & S. Lake, Yellowstone National Park (Smiley) (Porter 1934, 1935). Specimens of Smiley and other collectors at YELLO were variously Brachy- thecium salehrosum, B. leihergii, B. albicans. and B. frigidum. BreuteUa mohriana (C. Muell.) Broth., Car- bon Co. (Porter 1937). E.xcluded from North America (Anderson et al. 1990). Brotherella recurvans (Michx.) Fleisch., Lincoln Gulch, Albany County (Aven Nelson 2628). "The material is scanty, and Prof. Holzinger, who identified it, expressed a doubt as to the correctness of the determination' (Porter 1935). The specimen with Holzinger's opinion is at RM. Bryum canariense Brid. (Porter 1935). This is a species of the West Coast and not likely to occur in Wyoming. The specimen cited by Porter (Nelson 7814) curated at RM and US seems to be Bryum caespiticium Hedw. Encalypta streptocarpa Hedw. (Porter 1930, 1935). Excluded from North America (Ander- son et al. 1990). Gymnostomwn calcarewn Nees & Homsch. in Nees et al. (Porter 1937. "Washakie Co."). The specimen from WTU of Porter, Sept. 9, 1935, No. 2094, "On limestone boulders in a shady canyon' from the Ten Sleep Canyon in the Big Horn Mts, Washakie, Co., has been determined to be Gymnostomwn aeruginoswn by R. Zander. The specimen in section shows a ventral costal epidermis, two stereid bands, a central strand in the stem. The capsules were young and so rather ovoid. Homalothecium lutescens (Hedw.) Robins. Based on a citation by Porter (1937) and proba- bly the specimen: Yellowstone National Park Nelson No. 6041 (RM) appears to me to be Homalothecium aeneum instead. Hypnum callichrown Funck. ex Brid., Evan- ston, Uinta County (Aven Nelson 4128, in part: "The identity of this plant is doubtful, " Porter 1935; Uinta Co., Porter 1937). The specimen Nelson 4128 appears to be Hypnum lindbergii. Macrocoma sullivantii (C. Muell.) Grout (BUF). This record is due to a labeling error (Vitt 1981; D. Vitt, in litt). in the Orthotri- chaceae Boreali — Americanae Exsiccatae Fas- ciculus HI Nos. 21-30. The label issued with this species name, number 30, should be num- ber 27, Orthotrichum rupestre. The Macro- coma specimen originated in North Carolina, the Oiihotrichuiu from Yellowstone National Park (J. A. Christie, in litt.). Meesia triquetra (Richt.) Aongstr. reported by Cooper and Andrus (1994) is Oncophorus wahlenhergii. Orthotrichwn spcciosum Nees in Sturm (Porter 1935). Specimens at YELLO and RM were determinable as O. laevigatutn. Orthotrichum pallens Bruch ex Brid. var. parvum Vent., "Yellowstone National Park" 1996] Mosses of Wyoming 203 (Flowers 1973). Excluded from North America (Anderson et al. 1990). Physcomitrium pyrifonne (Hedw.) Hampe, cited by Porter (1937) for Crook Co., is proba- bly a specimen collected by Marion Ownbey from that count>' (No. 556a, TENN) and deter- mined by Porter as P. turhinatwn, but which, upon examination, is P. hookeri. P](i(i.ion}nhiiu ajfinc (Bland, ex Fimck) T. Kop. (Porter 1935). Excluded from North America (Anderson et al. 1990). PlagiotJwciiitn cavifoUiim (Brid.) Iwats. Porter (1935) based this name on a specimen of Elias Nelson (5242), which is Isoptcrygiopsis pulchella. Sphagnmn capillifolium (Ehrh.) Hedw. (Porter 1935). Porters specimens 1198 and 1199 collected in 1932 identified as S. capilli- folium had been redetermined by R. Andrus as S. nissowii (Andrus in lift.). Sphagnum )najus (Russ.) C. Jens. (Porter 1935, CiTun 1984). Taxa collected from Wyom- ing and identified as S. majus have all been S. annulatum, according to Andrus {in lift.), who states that the nearest sites would be in British Columbia, central Alberta, and Minnesota. Sphagnum palustre L. (COLO, RM). This species has been found only along the West Coast by Andrus {in lift). Sphagnum recurvum R Beauv. (COLO, RM). Material of this species from the interior of the United States is referable to S. angustijolium, according to Andrus {in lift.). Sphagnum recur- vum is an eastern coastal plain species. Tortula princeps De Not. Reports by Porter from Carbon and Crook counties (1935) were based on Nelson 2818 and 5034 at RM, and a specimen (RM) by Hennann (No. 17844), which were redetermined by R. Zander as Tortula ruralis. Weissia controversa Hedw. (Porter 1935). All citations for this species appear to be based on specimens of Dwight C. Smiley, deposited at YELLO. All 3 specimens seen were Dicrano- weisia crispula. Acknowledgments The author gratefulK' acknowledges the help of the late Mason Hale for his early support of this project, Holmes Rolston, and the late Fred Hermann. The assistance of Ronald Hartman of the Rock\' Mountain Herbarium, who pro- vided me with a copy of Porter's dissertation and access to field collections, and Jennifer Wipple of Yellowstone is acknowledged. Taxo- nomic assistance was provided by Richard Andrus, Teny Mcintosh, Howard Crinii, John Spence, Ronald Pursell, Norton Miller, Dale Vitt, and Richard Zander. The treatment of Sphagnum was considcral)ly improved by infor- mation provided by Dr Richard Andrus. Valu- able specimens were sent to me by Janice R McKee, Nancy Kastning-Culp, William Buck, Holmes Rolston, William Reese, and Joseph Elliott. This project was sponsored b\' the Sav- age Fund of the Buffalo Museum of Science and is dedicated to the memory of the late Elva Lawton. Literature Cited Anderson, L. E. 1990. A checklist of Sphagnum in North America north ofMe.xico. Bi^ologist 93: 500-501. Anderson, L. E., H. A. Crum, and W. R. Buck. 1990. List of the mosses of North America north of Mexico. Biyologist 93: 448-199. Brassard, Guy R. 1983. Checklist of the mosses of the island of Newfoimclland, Canada. Biyologist 86: 54—63. Christy, J. A., J. H. Lvford, and D. H. Wagner. 1982. Checklist of Oregon mosses. Biyologist 85: 22-36. Churchill, S. E 1979. Mosses of the Great Plains III. Additions to Nebraska and the Black Mills of St^uth Dakota and Wyoming. Biyologist 82: 72-75. . 1982. Mosses of the Great Plains VIII. Additions. Biyologist 85: 218-221. Cooper, David J., and Richard E. Andrus. 1994. Pat- terns of vegetation and water chemistry in peatlands of the west-central Wind River Range, W\'oming, U.S.A. Canadian Jonrnal of Botany 72: 1586-1597. Crum, H. A. 1984 Sphagnopsida, Sphagnaceae. North American flora, series II; part 11. New York Botanical Garden. 180 pp. Crum, H. A., and L. E. Anderson. 1981. Mosses of east- em Nortli America. Columbia University Press, NY. EvERSMAN, S., AND A. J. Sharr 1980. First checklist of Montana mosses. Proceedings of the Montana Acad- emy of Science 39: 12-24. Flowers, S. A. Holmgren, editor. 1973. Mosses, Utah and the West. Brigham Young University Press, Provo, UT HoRTON, D. G. 1983. A revision of the Encahptaceae. II. Journal of the Hattori Botanical Laboratoiy 54: 353-532. Ireland, R. R. 1982. Moss flora of tlie Maritime Provinces. National Museums of Canada, Ottawa. JOHNSEN, Ardith B. [Undated]. Keys to the mosses of Ari- zona. Research Paper 14. Museum of Northern Ari- zona, Flagstaff. Ketchledge, E. H. 1980. Revised checkhst of the mosses of New York State. In: R. S. Mitchell, editor. Flora of New York State. New Y'ork State Museum Bulletin 440. Albany, NY Lavin, M. 1981. New records for the moss flora of Nevada. Bryologist 84: 93-94. Lawton, E. 1958. Mosses of Nevada. Bryologist 61: 314-334. 204 Great Basin Naturalist [Volume 56 . 1971. Moss flora of the Pacific Nortliwest. Journal of the Hattori Botanical Laljoraton, Nichinan, Japan. McCleary, J. A., AND V. V. Gree.x. 1971. A checklist of Idalio mosses. Biyologist 74: 175-180. Medina, Alvix L. 1994. Lichens and biyophytes of the Rochelle Hills, Campbell County, Wyoming. Evansia 1: 121-130. Nelson, Aven. 1900. The ci-yptogams of Wyoming. 10th annual report of the W\oming E.xperiment Station, Laramie. 38 pp. Porter, C. L. 1930. Fruiting plants of Encalypta contorta. Bi-yologist 34; 93. . 1934. The moss genus Brachythecium in Wyoming. University of Wyoming Publications in Science. Botany 1(9): 235-241. . 1935. Bnophytes of Wyoming. Part II. Hepaticae (concluded) and Musci. Bryologist 38: 101-114 . 1937. The biyophytes of Wyoming. Unpublished doctoral dissertation. University of Washington, Seatde. Sayre, G. 1971. Ciyptogamae e.xsiccatae, an annotated bib- liograph) of published e.xsiccatae of Algae, Lichenes, Hepaticae, and Musci. Memoirs of the New York Botanical Garden 19(2): 175-176; 214-215. Shaw, A. J. 1981. A taxonomic revision of the propagulifer- ous species of Pohlia (Musci) in North America. Jour- nal of the Hattori Botanical Laboratoiy 50; 1-81. Spence, J. R. 1985. Checklist of the mosses of Grand Teton National Park and Teton County, Wyoming. Great Basin Naturalist 45; 124-126. . 1988. Checklist of the mosses of the Intermoun- tain West, USA. Great Basin Naturalist 48; 394-101. ViTT, D. H. 1973. A revision of the genus Orthotrichtim. Br>'ophytorum Bibliotheca, J. Cramer Verlag, Vaduz. , editor. 1981. Orthotrichaceae Boreali — Ameri- eanae exsiccatae fasciculus III, Nos. 21-30. Univer- sity of Alberta, Edmonton, Canada. . 1991. Rediscovei-y of Orfhotrichwn holziiii'eri: its moipholog)' and habitat in western North America. Biyologist 94: 77-79. Weber, W. 1973. Guide to the mosses of Colorado. Insti- tute of Airtic and Alpine Research Occasional Paper 6. Universit}' of Colorado, Den\'er Wynne, E E. 1943. Range extensions of mosses in western North America. Biyologist 46: 149-155. Received 21 December 1995 Accepted 27 March 1996 Great Basin Naturalist 56(3), © 1996, pp. 205-210 VARIATION IN BITTERBRUSH {PURSHIA TRIDENTATA PURSH) CRUDE PROTEIN IN SOUTHWESTERN MONTANA Carl L. Wanibolti, W. Wyatt Fraas^, and Michael R. Krisina^ Abstract. — The objective of this study was to clcterniine iCcrudc protein \aries sij^iiilicaiitlN duriiiti; late suiimier and midwinter among stands of hitterbrush {Piirshia trklentata Piush) in southwestern Montana. A secondaiy objective was to determine if leaves, when present, contribute significant additional protein in the region. Nine sites with different en\ironniental conditions and witliin a radius of 14.5 km were studied. Bitterl)rush leaves and leaders collected in August 1990 and 1991 and FebruaiT 1991 were used for crude protein and leaf-to-leader ratio determinations. Crude protein difTered (F < 0.001) among sites for both leaves and leaders on individual collection dates. Crude protein in lea\es was nearly twice the level found in leaders. Because few leaves were present in Februaiy, they increased crude protein in total foliage by only 0..3% over twigs alone. Feliruan' crude protein levels averaged 6.8% for total foliage, which is below the estimated requirement for wintering deer. Kcij words: Purshia tridentata, hitterhrush, crude protein, winter range, big game nutrition, Montana. Protein is one of the most important nutri- ents for wintering ungulates (Dietz 1972). Welch et al. (1983) estimated that winter crude protein levels o{ Purshia trklentata Pursh (hit- terbrush) are not high enough to meet ungu- late requirements, but postulated that protein content might vaiy with populations of bitter- brush. Differences in hitterbrush protein con- tent between sites have been noted (Giunta et al. 1978), although not between local habitat types (Morton 1976). Slausen and Ward (1986) found no difference in crude protein among 3 Colorado accessions in a common garden, but Welch et al. (1983) found differences in a com- mon garden test with plants from a wider geo- graphical area. No differences in nutrient con- tent have been found at varying browse levels of hitterbrush plants (Dietz et al. 1962, Shep- herd 1971). Crude protein levels were higher when winter leaves were present (Dietz et al. 1962), but winter leaf presence varies between populations of hitterbrush (Welch et al. 1983). Our objective was to determine if crude pro- tein varies significantly during late summer and midwinter among stands of hitterbrush in soutli- western Montana. Secondarily, we wished to determine if hitterbrush leaves in our region contribute significant additional crude protein quantities when present. Methods Study Sites Nine study sites were chosen primarily to represent hitterbrush stands from a range of environmental conditions (Table 1). This in- cluded burned sites and hitterbrush sites pro- tected from browsing. All study sites were located within a radius of 14.5 km near Butte and Anaconda in southwestern Montana. Long- term climatic records were available for the general study area from the Anaconda weather station at 1700 m elevation. Annual precipita- tion at Anaconda averages 340 mm, with 47% received between April and July (NOAA 1991). Vegetation types at all but 3 sites (burn, unburn, and High Rye) were serai stages of the bitterbrush-bluebunch wheatgrass {Agropijron spicatum Pursh) habitat type (Mueggler and Stewart 1980). The dominant shrub was hitter- brush, but understoiy vegetation was regressed (Fraas et al. 1992) on the other 6 sites from the described potential climax composition (Youtie et al. 1988). The Butte site at Maude S Canyon, near Butte, Montana, was selected because it re- ceives no ungulate browsing. The plant com- munity consisted of bitterbmsh, Centaurea mac- ulosa Lam. (spotted knapweed), Ribes cereum Dougl. (squaw currant), and Rosa woodsii Lindl. (Woods rose). 'Department of Animal and Range Sciences, Montana State University, Bozeman, MT 59717 2Montana Fish, Wildlife, and Parks. Butte, MT 59701. 205 206 Great Basin Naturalist [Volume 56 Table 1. Topographic characteristics of the 9 study sites. Data from the last 4 sites were obtained from Guenther (1989). Site Elevation Slope Aspect (m) (%) (degrees) Butte 1730 26 234 Cattle e.xclosiue 1830 16 188 Cattle + deer 1820 10 190 Bum 2010 21 220 Unburn 2010 24 180 Powerline 1640 16 85 Willow Creek 1780 31 110 Railroad Gulch 1650 32 115 High R\e 1940 38 120 At Diy Cottonwood Creek in the Deerlodge district of the Deerlodge National Forest, a livestock e.xclosiue with deer-only use was studied and known as the cattle exclosure site. Near the exclosure, a bitterbrush stand was studied and known as the cattle + deer site because it sustained both cattle and mule deer browsing. These 2 sites have a scattered over- story of Pseudotsuga menziesii [Mirb.] Franco (Douglas-fir). A high number of native peren- nial forbs occurred in the understory on these sites. Two sites were selected to gauge the impacts of burning bitterbrush in southwestern Mon- tana. The 2 sites (bum, unburn) were situated on either side of the burn line on the south flank of Steep Mountain, 8 km northwest of Butte, in the Butte District of the Deerlodge National Forest. The plant community on these 2 sites was a bitterbrush-mountain big sage- brush {Artemisia thdentata Nutt. ssp. vaseijana [Rydb.] Beetle)-bluebunch wheatgrass associa- tion intermediate to the big sagebrush-blue- bunch wheatgrass and bitterbrush-bluebunch wheatgrass habitat types of Mueggler and Stewart (1980). The prescribed burn was con- ducted 3 November 1981 after a year's rest fi'om livestock grazing to increase fuel loads. Live- stock use resumed 15 September 1982. When sampled for protein content, bitterbrush on the burned site was significantly lower in canopy cover (F > 0.01), flower production (P > O.lj, and seed production (F > 0.1) than on the un- bunied site (Fraas et al. 1992). Four sites were located on the Moimt Hag- gin Wildlife Management Area (MHWMA), owned and managed In Montana Fish, Wildlife, and Parks. The Powerline site was on a slope 50 m above a perennial stream on the north- east edge of the MHWMA big game winter range. The plant community consisted of bitter- biTish and spotted knapweed. The WiUow Creek site was near the top of a grassy ridge 150 m above Willow Creek. This site supported a rel- atively large amount o{Elijmus cinereus Scribn. & Merr. (basin wild lye), along with other perennial grasses and bitterbrush. This area was used as winter range b\' mule deer, elk, and moose. The Railroad Gulch site was also on the deer and elk winter range. This site occupied a midslope position 30 m above an intermittent stream, where the plant commu- nity consisted of bitterbrush and spotted knap- weed. The High Rye site was 1500 m higher in elevation than the other MHWMA sites and appeared to receive the greatest snowpack. The plant community on die High Rye site was typical of the bitterbrush-rough fescue {Fes- tiica scahreUa Torrey ex Hook.) habitat type (Mueggler and Stewart 1980) with those species currently dominant. Guenther (1989) found the least amount of big game use at this location among the 4 MHWMA sites. The MHWMA study sites received insignificant levels of live- stock grazing. Sampling and Analysis Leaves and leaders (current-year stem growth minus leaves) were collected at each study site from 10 randomly selected plants for crude protein analysis on each of the sampling dates. The same plants were sampled to deter- mine leaf-to-leader ratios. Material was col- lected the 1st week of August prior to or at seed set in 1990 and 1991. This was estimated to be the period of minimum soluble carbohy- drate content for bitterbrush plants (Menke and Trlica 1981). Material was also collected on 12 Februaiy 1991, when mule deer were con- centrated on these sites. Plant material was oven-dried at 60 °C for 48 h and weight of diy matter determined. Leaves were separated from leaders and weighed separateK' to deter- mine leaf-to-leader ratios on a percent dry matter basis. Leaves and leaders were then ground to approximateK 1 mm diameter in a grinder (Janke & Kunkel kg, t\pe AlO). Kjel- dahl (nitrogen) analyses were used to arrive at crude protein contents. Winter crude protein values were calculated with a weighted aver- age of winter leaf and leader protein levels. This allowed comparison with other studies (Dietz et al. 1962, Welch et al. 1983). 1996] Variation in Buterbrusii Crude Protein 207 Soil samples were obtained at a depth of 15 cm below the surface fiom a soil pit in each study plot. Because soils at most sites con- tained a large rock fiaction, it was necessan' to sample at the relatixeK shallow depth of 15 cm to standardize sampling. The Montana State Unixersity Soil Test Laboratory performed organic matter determinations and total Kjel- dahl nitrogen analyses on all non-MHWMA samples. Texture was determined by both the h> drometer and Bouyucous mechanical analy- sis methods. Soil pH was determined in 1 part soil to 2 parts water extractions. Topographic information was also recorded at each site. Aspect was determined by taking a compass bearing from the major slope. Slope was mea- sured widi a clinometer. Elevation was deter- mined from USGS topographic maps. The information from MHWMA sites was derived from Guenther (1989). A one-way ANOVA, with site as the factor, was conducted for each sampling date and pro- tein source combination (Snedecor and Coch- ran 1989). This was done with the knowledge that protein sources (leaves or leaders) con- tained veiy different levels of crude protein within each sampling date. Site was also the factor in an ANOVA for percent leafiness for the Febiaiaiy 1991 sample. The least significant difference (LSD) method (F < 0.05) protected by a prior F-test (F < 0.05) was used for com- paring treatment means (Snedecor and Cochran 1989). Results and Discussion Crude protein levels differed (F < 0.001) among sites within each protein source and collection date combination (Fig. 1). Thus, we rejected the hypothesis that crude protein val- ues are equal during August and Februaiy among local stands of bitterbnjsh. Crude pro- tein in the leaves, when averaged over all sites, varied with a 13% to 10% decline from August 1990 to FebiTjaiy 1991 and subsequent increase to 15% by August 1991. Crude protein in the leaders for these 3 dates was 7.1%, 6.5%, and 7.2%, respectively, when averaged over all sites. These crude protein levels generally agreed with previous reports for bitterbrush throughout its range (Dietz et al. 1962, Bay- oumi and Smith 1976, Morton 1976, Tiede- mann 1983, Welch et al. 1983). Protein levels also differed (F < 0.001) among the 3 collection dates (Fig. 1). When all sites were pooled, August leaf protein increased 11% between years (F < 0.05) and February leaf protein decreased 21% from August levels (F < 0.001). Leader crude protein did not vary significantly between years but was higher in August 1991 than during the previous Febni- aiy (F < 0.05). The unbrowsed Butte site rated highest in crude protein (Fig. 1) for 3 of the 6 measure- ments, although none was significantly higher than the next lower site. When the Butte site was compared to the aggregated crude protein levels of the other 8 sites, it was significantly (F < 0.05) higher for both leaves and leaders in August 1990, but did not differ from browsed sites in Febiaiar)' or August 1991. Thus, it does not appear that browsing affects crude protein levels. Protein values for the 4 MHWMA sites were lower for August 1990 leaves (F < 0.07) and leaders (F < 0.01) than for other sites and col- lectively rated lowest for 4 of the 6 measure- ments. These site differences were not expected fi-om Morton's (1976) work, but were supported by that of Giunta et al. (1978) and Welch et al. (1983). Bitterbrush crude protein levels on the deer + cattle site were 1% higher (F < 0.05) than on the adjacent cattle exclosure site for August 1990 leaves (Fig. 1). Other protein levels did not differ significantly between these 2 sites. Although a difference in use might thus seem to affect protein levels on these sites, the unbrowsed Butte site had higher protein levels than browsed sites in August 1990 (F < 0.05) and no difference in February or August 1991. Related to these site and possible popula- tion (Alderfer 1977) differences are soil differ- ences. Soil samples from shrub interspaces (Table 2) contained 49% more soil nitrogen at the Butte site than at the burn and unburn sites and 78% more than at the cattle exclosure and deer + cattle sites. Bayoumi and Smith (1976) found a positive response of bitterbrush protein levels to fertilization with nitrogen, although Tiedemann (1983) found slighdy neg- ative to no response to fertilization. However, most desert shmbs accumulate nutrients under their canopies, and the surrounding interspaces have low nutrient content (Garcia-Moya and McKell 1970, Tiedemann and Klemmedson 1973), conditions that we did not sample. 208 Great Basin Naturalist [Volume 56 Leaves Rajiroad Gulch Willow. Creek Powerlirle Un&urn/ BurrT CAitjepfid Deer jCatlle Exclosure Aug. 90 Aug. Leaders ajkoa^ Gulch ow Creek rliPie Burn V Cqltle^hd Deer _Cattle Exclosure BuUe_Z Aug. 90 Feb. 91 Aug. 91 Fig. 1. Average percent crude protein in I)itterhnish Iea\es and leaders found in August 1990, Februaiy 1991, and Augirst 1991. Protein values within eacli protein source and collection date with similar lowercase letters are not signifi- cantK' diflercnt (LSD, P > 0.05). Insufficient leaf material was available for statistical analysis in Febniar>' 1991. 1996] Vauiatiun in Buterbuush Crude Piu)ii:inj 209 Tablk 2. Soil charactt-ristics for stuch' areas, including pll, organic matter (OM), total Kjeldahl nitrogen (N), percent sand, silt, and clay, and textiiral class. Soil nitrogen was sampled at onl\' 5 sites. Data from the last 4 sites were oI)tained from Cinenther (1989). Site pll OM N Sand Silt C:lay 'le.xtural class (%) (%) (%) (%) (%) Butte 5.6 2,6 0.11 63 24 13 sandy loam Cattle e.xclosure 5.7 1.0 0.06 80 12 8 loamy sand Cattle + deer 5.7 1.0 0.06 80 12 8 loamy sand Burn 6.3 1.1 0.07 67 23 10 sandy loam Unburn 6.3 1.1 0.07 67 23 10 sand}' loam Powerline 5.8 1.4 — 65 15 20 sandy loam Willow Creek 5.2 3.6 — 65 18 17 sandy loam Railroad Gulch 5.7 1.0 — 72 18 10 sandy loam High R> e 6.7 2.8 — 69 15 16 sand) loam Protein levels at our study sites were therefore not necessarily related to soil nitrogen levels. Although most leaves had fallen by Febru- aiy, all sites contained plants that had retained some leaves at that time. Bitterbrush phenol- ogy seems to vaiy more by season and climate than b\' ecot>pe (Shaw and Monsen 1983). Most leaves are deciduous, dropping in response to moisture stress in late summer or fall (Shaw and Monsen 1983), but some small leaves over- winter on some populations (Alderfer 1977). Dietz et al. (1962) alluded to the high protein level of leaves in winter but did not quantify those levels. Welch et al. (1983) reported that winter leafiness (presumably, weight of leaves compared with weight of stems) of plants from Idaho, Colorado, Utah, and California ranged from 5.9% to 15.5%, while combined leaf and leader crude protein ranged from 5.9% to 7.9%. These ranges are similar to values found for these Montana sites: leafiness (percent weight of leaves per weight of stems) of 1.5% to 15.8% and combined cnide protein of 6.1% to 7.6% (Table 3). Because so few leaves were present in Febiaiaiy (Table 3), crude protein in total foliage increased by only 0.3% over twigs alone for all sites. Although we concluded that leaves contain significantly more crude protein than leaders on our study sites, leaf scarcity during winter in our region prevents total (leaf and leader) crude protein from meeting deer requirements (Welch et al. 1983). The February crude pro- tein levels for total foliage averaged 6.8% across sites, which were below the estimated neces- saiy direshold of 8.9% for wintering deer (Welch et al. 1983). However, September through November protein levels might have been higher, as many plants retained leaves through that period. Guenther (1989) reported that deer pellets from the MHWMA sites contained large amounts of Rocky Mountain juniper {Jiiniperus scopiilonou Sarg.) and Oregon grape {Bcrheris repens Lindl.). Protein values for small winter samples of Oregon grape and juniper from the Willow Creek site were 8.4% and 6.9%, re- spectively. These values are below those re- ported by Welch et al. (1983) and, like bitter- brush, are also below what they considered to be the necessary threshold of 8.9% crude pro- tein for wintering deer. Hamlin and Mackie (1989) suggested that mule deer have more need for high-quality forage in the fall, while building energy reserves, than in the winter. Bitterbrush in southwestern Montana may supply this needed level of nutrients in the fall, as we observed delayed leaf-fall on wind-pro- tected bitterbiTish plants in late November 1990, but we did not sample plants at that time. Restoration efforts for ungulate winter ranges capable of maintaining bitterbrush ma\ benefit through consideration of our results. We have found that bitterbiiish populations of even a localized ecotype, such as we studied, should not be expected to attain the same lev- els of crude protein over different environmen- tal conditions that will var>' between sites. Revegetation of bitterbrush ranges will involve consideration for obtaining the best possible plant materials. Our evidence indicates that plant characteristics, other than protein con- tent, should likely be of primar\' concern as protein can be expected to van' b\' site condi- tions regardless of plant material. However, it appears that consideration should be made of bitterbmsh genot>q3es that maintain a high per- centage of leaves into the winter These geno- types may provide a higher level of crude pro- tein that is desirable for \vintering ungulates. 210 Great Basin Naturalist [Volume 56 Table 3. Winter cnide protein content (percent) of bit- terbrush leaves and leaders combined and percent leafi- ness (percent weight of leaves per weight of stems) for study sites sampled Februaiy 1991. Column entries with similar letters are not significantly different (LSD, P < 0.05). Site Crude protein Leafiness {%) (%) Butte 7.fr' I3.(t'1' Cattle e.xclosure 6.4"! 6.5'l Catde + deer 6.6^' 9.9^- Bum 6.7l'^- 9.6«l Unburn 7.6« 10.5'^^ Powerline 6.1^ 8.5"' Willow Creek 7.lah 8.1"! Railroad Gulch 7.2^> 15.8" High Rye 6. Id 1.5^ Literature Cited Alderfer, J. M. 1977. A ta.\on()mic study of bitterbrush {Piirshia tridentata [Pursh] DC.) in Oregon. Unpub- lished master's thesis, Oregon State University, Cor- vallis. B.4,YOUMi, M. A., AND A. D. Smith. 1976. Response of big game winter range vegetation to fertilization. Journal of Range Management 29: 44-48. DiETZ, D. R. 1972. Nutritive value of shrubs. Pages 289-302 in C. M. McKell, J. P Blaisdell. and J. R. Goodin, editors, Wildland shrubs — tlieir biology and utilization. USDA Forest Sei-vice, General Technical Report INT-1, Ogden, UT. DiETZ, D. R., R. H. Udall, and L. E. Yeager. 1962. Chemical composition and digestibility' by mule deer of selected forage species. Cache la Poudre Range, Colorado. Colorado Game and Fish Department Technical Publication 14. Fraas, W. W, C. L. Wambolt, and M. R. Frisina. 1992. Prescribed fire effects on a bitterlinish-mountain big sagebrush-bluebimch wheatgrass community. Pages 212-216 in W. P Claiy, E. D. McArthur, D. Bedunah, and C. L. Wambolt, compilers, Proceedings of the Symposium on Ecolog\' and Management ot Riparian Shnib Communities. USDA Forest Senice, General Technical Report INT-289, Ogden, UT. Garcia-Moya, E., and C. M. McKell. 1970. Contribution of shrubs to the nitrogen economy of a desert-wash plant community. Ecology 51: 81-88. Giunta, B. C, R. Stevens, K. R. Jorcensen, and A. P Plummer. 1978. Antelope bitterbrush: an important wildland shrub. Utah Division of Wildlife Research Publication 78-12. GUENTHER, G. E. 1989. Ecological relationships of bitter- brush communities on the Mount Haggin Wildlife Management Area. Montana Department of Fish, Wildlife, and Parks and Montana State University Department of Animal and Range Sciences, Boze- Hamlin, K. L., and R. J. Mackie. 1989. Mule deer in the Missouri River Breaks, Montana. Montana Depart- ment of Fish, Wildlife, and Parks. Menke, J. W, and M. J. Trlica. 1981. Carbohydrate reserve, phenology, and growth cycles of nine Col- orado range species. Journal of Range Management 34: 269-277. Morton, M. A. 1976. Forage relationships of mule deer in the Bridger Mountains, Montana; nutritional values of important mule deer winter forage plants in the Bridger Mountains, Montana. Unpublished master's thesis, Montana State University, Bozeman. Mueggler, W, and W. L. Stewart. 1980. Grassland and shnibland habitat types of western Montana. USDA Forest Service, Intermountain Forest and Range Experiment Station, General Technical Report INT- 66, Ogden, UT. NOAA. 1991. Climatological data, Montana, 84-94(1-13). National Climatic Data Center, Asheville, NC. Shaw, N. L., and S. B. Monsen. 1983. Phenology and growth habits of nine antelope bitterbrush, desert bitterbrush, Stansbur\' cliffrose, and Apache-plume accessions. Pages 55-69 in A. R. Tiedemann, and K. L. Johnson, compilers. Proceedings of the Research and Management of Bitterbrush and Cliffrose in Western North America. USDA Forest Sen'ice, Gen- eral Technical Report INT-152. Shepherd, H. R. 1971. Effects of clipping on key browse species in southwestern Colorado. Colorado Game, Fish, and Parks Di\ ision, GFP-R-T-28, Denver, CO. Slausen, W L., and R. T. Ward. 1986. Ecogenetic pat- terns of four shrub species in semi-arid communities of northwestern Colorado. Southwestern Naturalist 31: 319-329. Snedecor, G. W, and W G. Cochr.\n. 1989. Statistical methods. Iowa State Universit>- Press, Ames. Tiedemann, A. R. 1983. Response of bitterbrush and asso- ciated plant species to broadcast nitrogen, phospho- rus, and sulfur fertilization. Pages 240-253 in A. R. Tiedemann, and K. L. Johnson, compilers. Proceed- ings of the Research and Management of Bitterbrush and Cliffrose in Western North America. USDA For- est Sei-vice, General Technical Report INT-152. Tiedemann, A. R., and J. O. Klemmedson. 1973. Effect of mesquite on physical and chemical properties of die soil. Journal of Range Management 26: 27-29. Welch, B. L., S. B. Monsen, .and N. L. Sh.aw. 1983. Nutriti\e \alue of antelope and desert bitterbrush, Stansbur\ cliffrose, and Apache-plume. Pages 173-185 in A. R. Tiedemann, and K. L. Johnson, compilers. Proceedings of the Research and Manage- ment of Cliffrose in Western North America. USDA Forest Service, General Technical Report INT-152. YouTiE, B. A., B. Griefith, .\nd J. M. Peek. 1988. Succes- sional patterns in bitterbrush habitat t\pes in north- central Washington. Journal of Range Management 41: 122-126. Received 1 November 1995 Accepted 1 Maij 1996 CiL-at Basin Naturalist 56(3), © 199fi, pp. 211-224 DAM-FORMING CACTI AND NITROGEN ENRICHMENT IN A PINON-JUNIPER WOODLAND IN NORTHWESTERN ARIZONA M()ll\ rhoinas Ihsrll' and Charles C. (irier^ Abstract. — In a pinon-junipcr woodland in nortliwi'stcrn Aii/ona, Lonncctcd basal cladodcs of a prickK pt-ar cactus {Opuntio Jittoralis var. martiniaiw) ionn check dams (hat cause deposition ol N-rich detritus in interspaces otherwise lackinu litter Seventy-eight percent of connected hasal eladodes measured in transects grew at an angle (w ith respect to till' slope contour) < 45° — an orientation facilitating tleposition of flood-horne itehris. Soil total N was significantly greater {F < 0.01) and organic C was greater, hut not significantly, a!)o\'e cactus dams compared to helow cactus dams. Soil total N and organic C both above and below cactus dams were significantly greater {P — 0.0001) compared to adjacent interspaces. Soil total N and organic C above cactus dams were equal to areas beneath canopies (tree and shrub combined). Net NO3 (0-5 em depth) above cactus dams was significantly greater (P = 0.0001) than below cactus dams, at interspaces, and beneath canopies. Net NH4 (0-5 cm soil depth) above cactus dams was significantly greater {P < 0.01) than below cactus dams and interspaces, and was greater (but not significantly) than beneath canopies. At 5-10 cm soil depth, differences in net NH4 and net NO3 between sampling locations were not significant except for the difference in net NO3 above and below cactus dams {P < 0.05). The litter layer above cactus dams had twice as much total N (P < 0.01) as the litter layer beneath canopies (tree and shrub combined); differences in net mineralized N were not significant between litter layers. Over the course of a single rainy season, detritus depth behind cactus dams increased up to 23 cm, with a mean increase of 4.3 cm (sj — 0.625, P = 0.0001). Key words: prickhj pear cactus, nitrogen enrichment, growth habit, soil characteristics, check dams, detriltis, runoff, bulk density, total nitrogen, organic carbon, mineral nitrogen, pinon-jtmiper woodlands, islands of fertility. The growtli habit of Opuntia littoralis var. martiniana (L. Benson) L. Benson consists of connected basal eladodes growing across wood- land slopes roughly along the contour Clad- odes in contact with the ground sprout adven- titious roots and become anchored. Sequentially anchored eladodes fiuiction as check dams dur- ing runoff events, causing deposition of flood- borne detritus including surface soil, animal feces, and litter of piiion pines, juniper, and oak. Piiion-juniper woodlands occupy at least 17 X 10*^ ha in the western U.S., with widespread distribution in Colorado, New Mexico, Arizona, eastern California, Nevada, and Utah (West 1988). These woodlands fall between mesic conditions that support closed-forest canopies and arid conditions in which plants are widely spaced. Compared with forests of wetter envi- ronments, pinon-juniper woodlands have low biomass, leaf area, and primary productivity (Crier et al. 1992). Woodland structure varies but can generally be described as single trees and shrubs and clumps of trees and shrubs sur- rounded by a network of interspaces (Lanner 1981). Litter occurs in patches due to the non- contiguous canopy cover, and soil N distribu- tion corresponds to litter and canopy distribu- tion (DeBano and Klopatek 1987, Tiedemann 1987). In mixed-species stands, patches may be mosaics of different litter components. Interspace and canopy area soils usually dif- fer in characteristics such as concentrations of nutrients, pH, bulk density, soil water, and in numbers and species of resident microorgan- isms and microarthropods (Everett and Shar- row 1985, Klopatek 1987, Klopatek and Klo- patek 1987), although there are exceptions to this generalization (DeBano et al. 1987). Soil organic matter and nutrients are concentrated near the soil surface (West and Klemmedson 1978, Lyons and Gifford 1980, DeBano and Klopatek 1987), and runoff from storms can cany considerable amoimts of detritus rich in organic matter and N (Fletcher et al. 1978). Objectives of this study were (1) to charac- terize the angle of growth (relative to slope con- tour) of connected basal eladodes of Opuntia littoralis van martiniana, (2) to compare litter and soil properties above and below cactus 'Department of Forest Resources, College of Natural Resources, Utah State University, Logan, UT 84321. -Department of Forest Science, Colorado State Universit\; Fort Collins, CO 8052.3. 211 212 Great Basin Naturalist [Volume 56 Table 1. Sites of measurement of angle (relative to the slope contour) of connected basal cladodes. (Samples for soil comparisons were taken only in the Hualapai Mountains [see Table 2].) Transect Soil Piiion pine length parent Elevation Aspect Slope cover Location (m) material (m) (%) (%) Cerbat Mountains'' 401 granite 1930 S-SSE 10-25 10-30 Hualapai Mountains'' 110 granite 1524 N 15-45 30-40 Music Mountains^ 302 vesicular basalt 1712 E 30 40-70 «27 km NW of Kingman, AZ (Iat.35''27', long.ll4°09'; T24NR18\VS23nw). 'n2 km SE of Kingman, AZ {lat..35''08', long.ll3°5.5'; T20NR16WSlsw). ^^^53 bii NE of Kingman, AZ (kit..35°41', long.ll3°49'; T27NR36WS.36ne). dams, and (3) to compare litter and soil proper- ties above and below cactus dams with inter- spaces and areas beneath canopies. Methods Two distinct physiographic provinces come together in northwestern Arizona: southeast, west, and north of Kingman, Arizona, is the Basin and Range Province, characterized by north-trending fault-block mountain ranges separated by broad desert valleys; the Col- orado Plateau lies to the east. This area is the interface of 3 deserts as well as a physiographic interface. North of Kingman is the Great Basin Desert, west is the Mojave Desert, and south- west is the Sonoran Desert. The climate of northwestern Arizona is semiarid (Sellers and Hill 1974). Precipitation is bimodal, occuning mostly in winter and summer months, with more rainfall during winter than summer. Summer rain sometimes occurs as intense thundershowers (Sellers and Hill 1974). We first observed dam-forming cacti in the Hualapai Mountains (rising to over 2438 m, 12 km southeast of Kingman, Arizona) in the course of data collection for studies of piiion- juniper woodland productivity'. We subse(iuently visited 2 nearby ranges (the Cerbat Mountains [over 2133 m at highest point] 29 km northwest of Kingman, and the Music Mountains [over 2011 m] 53 km northeast of Kingman) and found dam-forming cacti in these locations. To characterize die angle of growth of connected basal cladodes of prickly pear cacti (our 1st ol)- jective), we took angle measurements in July 1991 on all cacti intercepting straight-line tran- sects in the 3 mountain ranges (Table 1). Start- ing points of line transects were randomK' lo- cated, and direction of transects was along ran- dom azimuths. A total of 233 angle measure- ments were recorded. Sequentially connected Fig. 1. Mcasurcineut of angle ol growth ol couuectetl basal cladodes with respect to the slofX' contour. Point of origin ndicated In' solid cladode. 1996] Dam-fohming Cacti and Nithogen Enrichment 213 basal cladodes with series ranging from 0.4 ni to 2.5 m in length were nieasnred with an engineer's acljnstable triangle as shown in Fig- ure 1: a direction of growth parallel to slope contour was 0° while a direction of growth per- pendicular to slope contour either upslope or downslopc, was 90°. Soil and Litter Sampling Site description. — We restricted litter and soil sampling to 1 of the 3 transect locations (the Hualapai Mountains, 12 km southeast of Kingman [Table 2, Fig. 2]), to minimize con- foimding factors such as different soil types, site histories, and land-management practices. About 40% of the study site is open interspaces (combined data [unpublished] from eighteen 2 X 2-m plots using Daubenmire s [1968] cover- age classes, and from 12 permanently marked 25-m-long line transects using methods de- scribed in Meeuwig and Budy [1981]). Inter- spaces are mostly bare soil and rock surface, with 3% grass cover (mostly Boiiteloua gracilis [H.B.K.] Lag. ex Steudel and B. curtipendiila [Michx.] Torr.) and traces of litter, herbs, and ciyptogams. Shrubs, mostly scrub oak {Qiier- ciis iurbinella Greene), cover about 30% of the study area. Pinon pines {Piniis monophylla war. falhix [Little] Silba) cover about 36% of the area and Jiiniperiis osteosperma (Torr) Little about 4%. The added cover of vegetation com- ponents is greater than total vegetation cover due to the presence of different vertical layers of shrub and tree canopies and aggregation of vegetation in clumps. Trees ranged in age from seedlings to about 260 w (estimated fiom annuiil ring counts of cores [unpublished data]). Age estimates are approximate due to occurrence of false rings in wood of pifion pines and junipers. Size range of soil surface patches covered by cacti and associated litter accumulations was estimated by measuring eveiy cactus dam on a 25 X 25-m plot. We recorded length, widtli, and circumference for each cactus dam and associated litter accumulation (32 total). The area of soil surface covered by cactus dams and litter was calculated as the area of a circle plus 1/2 the difference between the area of a rectan- gle and a circle. Soil and litter sampling approach. — Sampling was stratified by woodland micro- habitats: above cactus dams, below cactus dams, interspaces, and beneath canopies. We took Tahlk 2. (^Iiaractoristic'S of litter and soil saniplinf^ site ill a pinoii-jiniipcr woodland in tlic Hualapai Mountains of noitliwt'stern Arizona. Records (1967-1991) of licensed livestock grazing show year-round grazing of cattle and horses with year-to-year variation in season of heaviest use and in number of animals (USDA BLM 1991). Elevation: 1524 ni (5000 ft) % slope-; 20-40 Aspect; north Soil parent material; granite Soil texture: sandy-loam Soil classification'': Barkerville Series\ loamy, mixed, mesic, shallow\ Udorthentic liaplnstolls Other soil characteristics'": Al horizon 10 cm deep, 39% coarse frag. pH surface soil interspace — 6.5 pH surface soil under canopy — 8.0 non-calcareous throughout Species and % cover"^: Pimi.s monophylla suhsp.jallax Jiinipcrus osteosperma Qtiereiis fitrbinella Yueca hacatta Opimtia littoralis var. iiuiiiiniana RJiiis trilohata Ceanothiis greggii Canotiu holocantha Bouteloua gracilis Gutierrizia sarothrae 36.0% «f 5.7 4.0% «f 1.6 30.0% Sy 6.2 4.0% Sy LI L9% Sy 1.1 0.7% AY 0.4 0.4% «r 0.2 0.3% «r 0.3 2.9% «r 1.3 <1% "Richmond and Richardson (1974). ''Unpublished data, this study. 'Two methods were used to estimate cover: For all species, estimates were made on eighteen 2-m2 plots according to coverage class ratings (Daubenmire 1968). Tree and shrub cover were estimated on 12 permanent 25-m line tran- sects as % Cover = [(25*.3.14/Transect Length)] [Sum of crown diameters] (Meeuwig and Budy 1981). Values reported here for trees and shrubs are aver- ages of both methods, and standard eiTors are from pooled variances. paired samples 10.2 cm above (litter present) and below (little to no litter present) cactus axes to compare soil properties above and below cactus dams. We took additional sam- ples from bare interspaces and from areas beneath tree and shrub canopies to compare these areas with the areas above and below cactus dam. Interspaces were considered to lie beyond the influence of canopies and associ- ated litter and beyond the influence of cactus dams and associated litter. Vegetation and litter were scant to absent in interspaces. Beneath tree and shrub canopy, sampling included pinon pines, scrub oaks, junipers, and occa- sionally mixed-species canopies roughly in pro- portion to the presence of these components (as estimated by percent canopy cover) on the site (Table 2). The sampling location beneath canopies was at 2/3 canopy radius out from the stem or clump center Litter of Yucca haccata 214 Great Basin Naturalist [Volume 56 Fig. 2. Soil and litter sampling area in the Hiialapai Mountains of northwestern Arizona. The contour intenal is 12.2 in (40 ft). Enlarged from U.S. Geological Sunex', Rattlesnake Hill. Arizona Quadrangle. Torr. and a few other species was occasionally (though rarely) present in litter samples along with litter of the dominant species. With the exception of hulk density samples, soil and lit- ter samples were composited within microhah- itat strata by combining equal umubers of equal- sized indix'idual samples. Compositing followed guidelines in Peterson and Cabin (1986) and was suitable for the present study since we were not examining variation within nucrohab- itats. As pointed out b\' Crepin and Johnson (1993), composite sampling can be used in 1996] Dam-I()Kmi\g Cacti and Nitrogen Enrichment 215 conjunction with stratification: i.e., the hmd- scape can be cli\ ided into meaningful units and good averages of soil properties obtained b\ compositing samples within each unit. All soil and litter sampling was conducted in Jul\ 1991. Bulk density. — Bulk density was deter- mined by the exca\ation method (Blake and Hartge 1986). T\vent>'-t\vo paired samples weie taken 10.2 cm al)o\e an'' below cactus axes, 10 samples were taken fiom '"'^erspaces, and 10 were taken from beneath tree and shrub canopies. Soil was e.\ca\"ated with a bulb planter (diameter 5.5 cm at cutting edge), creating a hole 7 cm deep. A thin, tough plastic bag was placed in tlie hole, filled witli water, and then emptied into a graduated c>'linder to deter- mine hole \olume. Extracted soil was dried at 105° C and weighed, resulting in a weight-to- \ olume measurement. Total N, total organic C, .\nd soil tex- ture.— Thirt)' pairs of soil cores (mineral soil surface to 7 cm deep) were extracted with a bulb planter (diameter 5.5 cm at the cutting edge) adjacent to cactus axes (10.2 cm above and below cactus axes), 30 li-om beneath cano- pies, and 30 from interspaces. Samples were taken near each of the 6 satellite plots estab- lished for the net mineralization stud\ (see below). Litter (all litter from surface to mineral soil) was retained for detemiination of total N. Samples were air-dried and stored in paper wrappers. Soil samples originalK' taken for determining bulk density (see above) were added to these soil samples for a total of 51 samples from each side of cactus dams, 40 samples from beneath canopies, and 40 from interspaces. One of the 22 paired bulk densit) samples was lost and could not be included. Samples were combined to create compos- ites: above cactus dams 51 samples of soil were composited to make 3 samples of soil, and 30 samples of litter w ere combined to make 3 lit- ter samples. Below cactus dams (no litter pre- sent) 51 samples of soil were composited to make 3 soil samples. Beneath canopies 40 soil samples were composited to make 3 samples of soil, and 30 litter samples were composited to make 3 litter samples. From interspace areas (no litter present) 40 samples were composited to make 3 samples of soil. AnaKsis was b>- Utah State Universit}' Soils Testing Lab following the Kjeldahl method (Bremner and Mulvane\ 1982) to detennine percent total N, the Walk- le\-Black method (Nelson and Sommers 1982) for percent organic C, and methods described b\ Gee and Bander (1986) for particle-size analysis. Net mineralized N.— The total amount of N liberated from organic matter is "gross min- eralization"; the quantity remaining after micro- bial immobilization is "net mineralization" (Car- lyle 1986). Net mineral N, the N available for plant uptake, is an index of soil fertilitv'. To com- pare soil N fertilit)' among woodland sites, net mineral N was assessed by laboratorx' aerobic incubations (Binkle\- and Vitousek 1989). Seven pemianent plots were created on the study site, the 1st plot serving as a central point fiom which 6 satellite plots were created, each 32 m fiom tlie central point at 60° inteniils beginning with a random azimuth. Because of topography, 1 plot was relocated 32 m from the center of a satellite plot. From each plot center 8 cacti (0.5 to 5 m IroiP center) were selected at 45° intervals beginnmg with a random azimuth, for a total sample of 56 cacti. Paired soil samples were taken 10.2 cm from cactus axes on all 7 permanent plots be- ginning li-om the easternmost cactus and mov- ing clock-wise. Samples were composited com- bining 4 individual samples into 1 composite sample. Compositing and field processing (see below) were perfomied immediatcK- upon the extraction of 4 cores. For example, on the 1st plot 4 cores 10.2 cm above cactus axes in the 90° -270° hemisphere of the plot were taken, composited, and field processed before the next 4 cores were drawn. This ensured pro- cessing fresh soil. Fourteen composite sample pairs were prepared. At approximateh" midpoints of the six 32-m lines creating satellite plots, 2 samples were taken beneath canopies (piiion pines sampled most heavily followed by scrub oak, mi.xed- species canopies, and juniper) and 2 fiom inter- spaces. Composites of 4 individual samples were prepared and field processing completed immediateK' as each set of 4 cores was drawoi. Three composite samples were prepared. Samples were taken with a 2-cm-diameter soil corer to a depth of 10 cm. Preparation of samples for anabsis followed methods outlined in Vitousek et al. (1982): In the field cores were divided into 3 components (litter layer, top 5 cm of mineral soil, and mineral soil between 5 and 10 cm soil depth) and composited. Com- posite soil samples were sieved through a 2- 216 Great Basin Naturalist [Volume 56 30-39 40-49 50-59 60-69 70-79 80-90 degrees Fig. 3. Angle of growth of connected basal cladodes with respect to slope contour Zero degrees is a direction of growth parallel to the slope contoin-; 90 degrees is a direc- tion of growth peipendicular to the slope. Table 3. Size distribution of cactus dams and associated litter accumulations on a 25 X 25-m plot. The area mea- sured was the soil surface co\'ercd b\ cactus dams and associated litter Size class (m2) 0.05 0.1-1.0 1.1-2.0 2.1-3.0 3.1-4.0 10.30 Number of cactus dams 1 16 6 6 2 1 mm screen; litter was not sieved. Subsamples were sealed in bags for detemiination of mois- ture content, while a 2nd subsample of approx- imately 10 g was placed in 100 ml 1 N KCl adjusted with HCl to pH 2.5 with phenylmer- curic acetate (PMA) added as a presei-vative. Solutions were refrigerated, transported to the laboratory, mixed frequently for 4 d, then allowed to settle for 48 h. After settling, the solution was removed with a pipette, and NH4"^ and NO3 were determined at Bilby Research Facilit>' at Northern Arizona Univer- sity using methods described by Keeney and Nelson (1982). The remainder of composited field samples (after removal of the above 2 subsamples) was transported to the laboratory and incubated aerobically following procedures in Vitousek et al. (1982): Soils were wetted to approximately field moisture capacity (assessed visually), placed in plastic-covered cups, each of which had a small air hole, and kept in a dark, moist chamber at a constant temperature of 22° C. During an 8-wk incubation period, samples received distilled water (applied as a fine mist to the surface with no mixing) as needed to maintain an approximately constant moisture content. So as not to disturb incubating sam- ples, moisture content was assessed by visible soil color easily observable through the clear plastic incubation cups. At the end of 8 wk, subsamples (approxi- mately 10 g) of incubated samples weie taken for determination of moisture content, and subsamples of approximately 10 g were placed in the KCl solution described above. These solutions were shipped to the soils testing lab- oratoiy at Utah State Universit)' for detemiina- tion of NH4+ and NO3" (U.S. EPA 1983). Change in deposit depth. — Depth of de- posits above cactus dams (i.e., above con- nected basal cladodes) was measured before (Jul)') and after (September) the rainy season of 1991 on 6 of the 7 plots designated for net min- eralization sampling (see above). Two sampling points could not be relocated at the end of the rainy season, making a total sample size of 46 cactus dams (i.e., 6 plots, 8 cacti per plot, minus 2). Depth was measured from base to top of deposits in the area of greatest accumulation. Statistical Analysis A heterogeneity chi-square analysis followed by a chi-square anabsis (Zar 1984) was per- formed with the 3 data sets of angle of cactus growth from the 3 mountain ranges. Soil and litter analyses. — Tests of nor- mality were performed for each data set (above cactus dams, below cactus dams, interspaces, and beneath canopies) of each soil and litter characteristic sampled. A paired t test (»= = 0.05) was used to compare means of soil char- acteristics abo\ e and below cactus dams, and to compare the depth of deposits at cactus dams l)efore and after the rainy season. An anabsis of variance F-test (oc = 0.05) for unbal- anced sample sizes (the GLM procedure in SAS software [SAS 1985]) was used to compare sample means of soil abo\ e and below cactus dams with beneath canop\ and interspace sam- ple means. Plots of residuals were generated to assess equalit) of variance. Significant differ- ences between means were separated and lanktxi using a nniltiple comparison method 1996] Dam-fokminc Cacti and Nitroc;en Enrichment 21' Table 4. Results of paired t tests comparing sample means of soil characteristics above and below cactus dams, and comparing depth of detritus above cactus dams between early July and mid-September A minus B refers to tlu' value alio\e cactus tlams minus the \alue below. N Pail ed /-test statistics .Utnbute Mean difference ■''■.V / P A minus B Hulk di'nsit> (g/ml) (natural log) 0-7 cm depth 22 -0.222 0.117 1.901 0.0711 Total N (%) 0-7 cm depth 3 0.043 0.003 13.000 0.0059 Organic C (%) 0-7 cm depth 3 0.933 0.231 4.035 0.0563 Net mineralized N (ug/g) 0-5 cm depth NH4 14 9.621 2.353 4.088 0.0013 0-5 cm depth NO3" 14 41.979 5.398 7.776 0.0001 5-10 cm depth NH4'' 12 2.050 2.063 0.993 0.3418 5-10 cm depth NO3" 12 7.550 2.672 2.826 0.0165 Change in depth (cm) of detritus aIio\'e cactus dams (natural log) 0-7 cm depth 46 1.240'' 0.124 10.038 0.0001 '^Mean difference of September detritus depth ininu.s JiiK detritus depth. (REGWF) cited as being compatible with the overall analysis of variance F-test (SAS 1985). A t test (oc = 0.05) was used to compare sam- ple means of total N and net mineralized N fi'om the litter above cactus dams with the lit- ter layer beneath woodland canopies. Results The pattern of angle of connected basal cladodes with respect to slope contour was similar in the 3 mountain ranges sampled; data were pooled based on results of a heterogene- ity chi-square analysis. Analysis of pooled data (X" = 85.4, P < 0.001) indicated that orienta- tion of connected basal cladodes of Opuntia lit- toralis var mai-tiniana was nonrandom: growth was most frequently parallel to the woodland slope contour (Fig. 3). The size range of cactus dams and associated litter on a 25 X 25-m plot at the Hualapai Mountains study area is given in Table 3. Soil and litter analyses. — The null hy- pothesis for normality was not rejected for most of the data sets; however, total N data were nonnormal and were not normally dis- tributed when transformed with standard transformations. Therefore, results of total N analyses should be interpreted with caution. Residual plots indicated equality of \'ariance assumptions were reasonable. Bulk density above and below cactus dams was not significantly different at P = 0.05 (Table 4). Bulk density was significantly lower (F = 0.0001) in soil deposits above cactus dams, below cactus dams, and beneath tree canopies, compared to soil from interspaces (Table 5, Fig. 4). Soil above and below cactus dams was also lower in bulk density than soil beneath tree canopies, although this difference was not significant at F = 0.05. There was little differ- ence in soil texture among the 4 microhabitats (Tible 5). Soil total N above cactus dams was greater (F < 0.01) than below cactus dams {Table 4). Organic C was not significantly different (F = 0.05) above cactus dams compared to below cactus dams. Soil total N and organic C were 2-3 times greater (F = 0.0001 in both cases) in soil above and below cactus dams than in interspace soil (Table 5, Fig. 4). Soil total N and 218 Great Basin Naturalist [Volume 56 E A. Bulk density F = 9.29 p = 0.0001 r T I X'''- above below interspaces cactus dams cactus dams beneath canopies B. Total F = 35.77 p = 0.0001 ""d&'T" above below cactus dams cactus dams interspaces beneath canopies C. Organic C above cactus dams below cactus dams interspaces beneath canopies Fig. 4. Comparisons of soil characteristics above cactus dams, helow cactus dams, beneath canopies (tree and shrub combined), and in bare interspaces. organic C above cactus dams were equal to areas beneatli canopies. Below cactiis dams, soil total N was significantly lower than beneath canopies, and organic C was not significantly different compared to beneath canopies. While soil organic C and soil total N differed among woodland locations, die C:N ratio was similar between locations (Table 5). Net mineral NH4^ and NOg" at 0-5 cm depth were significantly greater (F = 0.001 and P = 0.0001) above cactus dams compared to below (Table 4). At 5-10 cm depth net min- eral NO3 was significanth' greater (F = 0.0165) abo\'e cactus dams compared to below. Net mineral N in soil 0-5 cm deep above cac- tus dams was over 3 times that in interspace Dam-i'()kminc Cacti and Nithocen Emuchment 219 D. NH4"^ 0-5 cm depth ■3 16 14 12 10 8 6 4 2 0 F^T^^rn F = 4.93 p = 0.0067 above cactus dams below cactus dams interspaces beneath canopies 3 E. NO" 0-5 cm depth 90 80 70 H 60 50 40 30 H 20 10 0 F = 11.21 p = 0.0001 p-SfWiife^'fe^^^ below cactus dams interspaces beneath canopies F. NH4"^ 5-10 cm depth O) 5 - 4 - 3 - 2 - 1 0 F = 0.74 p = 0.5374 above below interspaces cactus dams cactus dams beneath canopies G. NO3" 5-10 cm depth 16 n 14 - 12 - 10 - 8 - 6 - 4 - 2 - 0 - ,-^ iiii isjiiiiiSiiiiiS F = 0.97 p = 0.4219 0: =3 r CO 0 z Bil „._ I. : .! „ above below interspaces beneath cactus dams cactus dams canopies Fig. 4. Continued. 220 Great Basin NatuR'\list [Volume 56 Table 5. Comparison of sample means of soil characteristics at 4 woodland microhahitats (above and below cactus clams, interspaces, and beneath canopies) at the Hualapai Mountains site. Superscript letters separate means signifi- cantly different at °<: = 0.05. For te.xture, s = sand, si = silt, and cl — cla\. Samples are composites except for bulk den- sity'. N = sample size and is followed in parentheses by the number of individual samples that were composited. Attribute Location Above cactus dams Below cactus dams Interspaces Beneath canopies sandy loam sandv loam sandv loam sandv loam s 65.0 s 63.0 s 63.7 s 62.7 si 25.0 si 26.0 si 27.6 si 28.3 cl 9.0 cl 11.0 cl 8.7 cl 9.0 n = 3 (51) n = 3 (51) n = 3 (40) n = 3 (40) X = 0.99b X = 1.13b X = 1.85" X = 1.40b ±0.11 ±0.92 ±0.10 ±0.16 n = 22 n =22 /) = 10 n = 10 .T = 0.16^' X = 0.12b .V = 0.06^' .V = 0.17" ± 0.006 ± 0.003 ± 0.007 ±0.16 J! =3(51) n = 3 (51) n = 3 (40) n = 3 (40) X = 3.9'' .V = 3.0'' X = 1.5b .V = 3.6-' ±0.24 ± 0.07 ±0.04 ±0.26 n = 3 (51) n = 3 (51) n = 3 (40) n = 3 (40) Soil te.xture % separates kilk density (g/ml) Total N (%) 0-7 cm depth Organic C (%) 0-7 cm depth C:N ratio Net mineralized N (iig/g) 0-5 cm depth NO," 5-10 cm depth NH4^ Nor 24.4 25.0 25.0 21.2 X = 14.8" X = 5.3b .V = 2.0b .V = 11.5" ±2.6 ±1.6 ±0.2 ± 1.3 X = 88.7" .V = 46.7bc X = 27. l^' X = 54.2b ±6.9 ± 5.9 ±2.6 ±6.4 n = 14 (56) n = 14 (56) n = 3 (12) n =3(12) X = 5.3 .V = 3.2 x =0.9 X = 4.5 ±1.1 ±1.8 ±0.97 ±1.5 .f = 31.6" X = 24.0b X = 16,6 .V = 30.9 ± 4.0 ±4.4 ± 3.0 ±8.2 u = 14 (.56) n = 14 (.56) /! =3(12) n = 3 (12) soil and almost twice that in soil beneath tree canopies (Table 5, Fig. 4). Net mineral N below cactus dams was greater than in interspaces, but the difference was not statistically signifi- cant. Litter accumulated at cactus dams had total N (0.74%) over twice as high as litter beneatli tree and shrub canopies (0.32%) {t = -8.4, P = 0.01). NH4^ and NO^" in the litter layer were greater beneath canopies than above cactus dams, but not significantly (Table 6, Fig. 5). From early July to mid- September, depth of detritus behind cactus dams increased signifi- cantly (P = 0.0001) from -2 cm to -1-23 cm, widi an average of 4-4.3 cm (.v^ 0.625; Fig. 6). Discussion The similarit) of soil te.xture abo\e cactus dams, below cactus dams, beneath tree and shrub canopies, and in interspaces agrees with findings of Schlesinger et al. (19(S9) that desert soils receiving overland flow and adjacent soils deprived of overland flow were similar in fine material or cla\ content. The effects of cactus dams and associated litter and detritus deposits on bulk density, total N, organic C, and net mineralized N of nearby soil were expected based on a mnnber of studies in shrub lands and woodlands documenting islands of fertility, i.e., localized areas of nutrient enrichment 1996] Dam-fouminc Cacti and Nith(x;i:n Enhichmi<:nt 221 I'aiu.K 6. Comparison of total \ and ni-t niinerali/.fcl N in tlit' littt-r la\t'r hcMU'atli tanopics with liltiT acciininlations above cattns dams. N = sample size and is followed in pari-ntheses by tlie miniber of individnal samples that were eomposited. A1)()\( Beueatl cactus d lUlS canopies .Attribute Mean .S^T Mean ■""■.v Total N (%) 0.737 n = 3 (30) 0.047 0.320 n = 3 (30) 0.015 Net mineralized N (ug/g) NH4 60.233 n = 2^' (12) 15.018 84.550 n = 121' (48) 11.250 NO3" -2.108 n =2(12) 6.410 40.050 n = 12 (48) .34.950 -8.4275 1.2929 1.1865 0.0095 0.2426 0.4398 ''Three composites were prepared; however, initial (before incubation) net mineral N values were not obtained lor 1 sample ''Values before incubation were not obtained for 2 of the original 14 composited samples. (Garcia-Moya and McKell 1970, Tiedemann and Klemniedson 1973, Baith and Klemmed- son 1978, Baith 1980, Doescher et al. 1984, Exerett et al. 1986, Garner and Steinberger 1989, Schlesinger et al. 1990). Deposits at cactus dams ofOpiintia littoraUs var. inariiniana raised soil total N from 0.06% (interspace soil) to 0.16% above connected basal cladodes and to 0.12% below (Table 4). Nitrogen enrichment and soil amelioration associated with deposits at cactus dams may increase cactus productivity. Nobel et al. (1987) observed that while annual aboveground pro- ductivity of prickly pear cacti can be higli imdcr optimal conditions, cacti productivity is often limited by low levels of soil N (Nobel et al. Litter total N and net mineral N z .g 120 100 80 60 - 40 20 -20 p = 0.0095 total N mi cactus dams canopies p = 0.2426 ^i^ net NH/ p = 0.4398 net NO, Fig. 5. Nitrogen in the litter accumulated above cactus dams compared with the litter layer beneath canopie.s (trees and shrubs combined). 222 Great Basin Naturalist [Volume 56 Q. September mean depth 1 1 .4 cm July mean depth 7.1 cm deposition at individual cactus dams Fig. 6. Deposition at cactus dams during 1 season of summer thundershovvers; depths of detritus accumulations at 46 cactus dams in July and in September. 1987, Nobel 1989). Increased productivity in desert prickly pear cacti is positively coirelated with both number of new cladodes produced and cladode size (Nobel et al. 1987). VVe do not know if similar patterns occur in woodland species of prickly pear. Additionally, Nobel (1988) describes a tendency for "daughter" clad- odes to replicate the orientation of "mother cladodes and points out that if a particular direction of growth is favorable, it may be per- petuated. This happens because favorably ori- ented cladodes are expected to be more pro- ductive than other cladodes and produce more and larger similarly oriented cladodes. This may be occurring in dam-forming cacti, but it was not investigated in this study. Cactus dams lower soil bulk density and enrich patches of woodland interspace with organic matter, total N, and net mineral N, sug- gesting that they may play roles in nutrient cycling and other ecosystem processes. Some possible functions of cactus dams are to (1) in- crease woodland detritus storage, (2) increase the rate of N turnover, (3) mitigate nutrient loss in interspace areas, (4) reduce soil erosion and dampen effects of disturbances, (5) provide seedbeds, and (6) provide habitat for other organisms. Acknowledgments We thank Ron Lanner and Helga Van Miegroet for advice, encouragement, and assis- tance. The senior author thanks Chuck Crier for accepting an unconventional student and for sharing his abundant cajoleiy and scientific acumen. Literature Cited Barth, R. C. 1980. Influence of pin\ on pine trees on soil chemical and ph\sical properties. Soil Science Soci- ety' of America Journal 44: 112-114. Barth, R. C, and J. O. Klemmedson. 1978. Shrub- induced spatial patterns of dn' matter, nitrogen, and organic carbon. Soil Science Society of America Jour- nal 42: 804-809. BiNKLEY, D., and E Vitousek. 1989. Soil nutrient avail- ability. Pages 75-96 in R. W. Pearc\-, J. R. Ehleringer, H. A. Mooney, and P W. Rundel, editors, Plant physi- ological ecolog>'. Field methods and instrumentation. Chapman and Hall, London/New York. Blake, G. R.. and K. H. Hartce. 1986. Bulk density Pages 363-375 in A. Klute, editor. Methods of soil analysis, part 1: physiciil and mineralogical methods. 2nd edition. Soil Science Society of America, Inc., Madison, WL Bremner, J. M., AND C. S Ml LVANEV. 1982. Nitrogen — total. Pages 595-694 in A. L. Page et al., editors. Methods of soil analysis, part 2. 2nd edition. Agron- omy Monographs 9. American Societ}' of Agronomy and Soil Science Societ\' of America, Madison, Wl. 1996] Dam-forming Cacti and Nitrogen Enrichment 223 Carlyle, J. C. 1986. Nitrogen cycling in lorestccl ecosys- tems. Forestiy Abstracts 47: 30(5-336. Crepin, J., A.ND R. L. J()iiN.S()N. 1993. Soil .sampling for en- vironmental assessment. Pages 5-18 ;/i M. \\. Carter, editor. Soil sampling and methods of analysis. Lewis Pnblishers, Boca Raton/Ann Arbor/London/ Tokyo. D.\UBENMIRE, R. 196S. Plant conminnities: a textbook of plant s\necolog>. Harper &: Row, New York. DeBano, L. E, and J. M. Keop.\tek. 1987. Effect of man- agement on nutrient dynamics in sonthwesteni pinxon juniper woodlands. Pages 157-160 in C. A. Troendle, M. R. Kanfmann, R. II. Hamre, and R. P Winokur, coordinators. Management of subalpine forests: build- ing on 50 \'ears of research. General Technical Report RM-149. Rock> Mountain Research Station, Fort Collins, CO. DeBano, L. E, H. M. Perry, and S. T Overby^ 1987. Effects of fiielwood harvesting and slash burning on biomass and nutrient relationships in a pinyon-juniper stand. Pages 382-386 in R. L. Everett, compiler. Proceed- ings— pinyon-juniper conference. General Technical Report INT-215. Intermountain Research Station, Ogden, UT Doescher, R S., R. E Miller, and A. H. Winward. 1984. Soil chemical patterns under eastern Oregon plant communities dominated by big sagebrush. Soil Sci- ence Society- of America Jovmial 48: 659-663. Everett, R. L., and S. H. Sharrow. 1985. Soil water and temperature in hai"vested and nonhai"vested pinyon- juniper stands. Forest Service Research Paper INT- 342. Intermoimtain Research Station, Ogden, UT. Everett, R. L., S. H. Sharrow, and D. Than. 1986. Soil nutrient distribution imder and adjacent to singleleaf pin>'on crowiis. Soil Science Society' of America Jour- nal 50: 788-792. Fletcher, J. E., D. L. Sorensen, and D. B. Porcella. 1978. Erosional transfer of nitrogen in desert ecosys- tems. Pages 171-181 in N. E. West and J. Skujins, editors. Nitrogen in desert ecosystems. Dowden, Hutchinson & Ross, Inc., Stroudsburg, PA. Garcia-Moya, E., and C. M. McKell. 1970. Contribution of shrubs to the nitrogen economy of a desert-wash plant community. Ecology 51: 81-88. Garner, W, and Y. Steinberger. 1989. A proposed mech- anism for the formation of 'Fertile Islands' in the desert ecosystem. Journal of Arid Environment 16: 257-262. Gee, G. W, and J. W. Balider. 1986. Particle-size analysis. Pages 383-411 in A. Klute, editor. Methods of soil analysis, part 1: physical and mineralogical methods. 2nd edition. Soil Science Society of America, Inc., Madison, WI. Grier, C. C, K. J. Elliott, and D. G. McCullough. 1992. Biomass distribution and productivity of Finns edulis-Juniperus monospenna woodlands of north- central Arizona. Forest Ecologv and Management 50: 331-350. Keeney, D. R., and D. W. Nelson. 1982. Nitrogen — inor- ganic foiTns. Pages 643-698 in A. L. Page et al., edi- tors. Methods of soil analysis, part 2. 2nd edition. Agronomy Monographs 9. American Society of Agron- omy and Soil Science Society of America, Madison, WI. Klopatek, J. M. 1987. Nutrient patterns and succession in pinyon-juniper ecosystems of northern Arizona. Pages 391-396 in R. L. Everett, compiler, Proceedings — pinyon-juniper conference. General Technical Report INT-215. Inteimountain Research Station, Ogden, UT. Klopaiek, C. C, and J. M. Klop.viek. 1987. Mycorrhizae, microbes and nutrient cycling processes in pinyon- juniper systems. Pages 360-364 in R. L. Everett, compiler. Proceedings — pinyon-juniper conference. General Technical Report INT-215. Intermountain Research Station, Ogden, UT. Lanner, R. M. 1981. The pinon pine: a natural and cul- tural history. University of Nevada Press, Reno. Lyon,s, S. M., and G. E GlKEORD. 1980. Impact of incre- mental smface soil depths on plant production, tran- spiration ratios, and nitrogen mineralization rates. Journal of Range Management 33: 189-196. Meeuwig, R. O., and J. D. Budy. 1981. Point and line intersect sampling in pinyon-juniper woodlands. General Technical Report INT-104. Intermountain Research Station, Ogden, UT. Nelson, D. W, and L. E. Sommers. 1982. Total carbon, organic carbon, and organic matter. Pages 539-579 in A. L. Page et al., editors. Methods of soil analysis, part 2. 2nd edition. Agronomy Monographs 9. Ameri- can Society of Agronomy and Soil Science Societ)' of America, Madison, WI. Nobel, P S. 1988. Environmental biology of agaves and cacti. Cambridge University' Press, Cambridge. . 1989. A nutrient inde.x quantifying productivity' of agaves and cacti. Journal of Applied Ecology 26: 635-645. Nobel, P S., C. E. Russell, P Felker, G. C. Medin.^, and E. ACUNA. 1987. Nutrient relations and productivity of prickly pear cacti. Agronomy Journal 79: 550-555. Petersen, R. G., and L. D. Calvin. 1986. Sampling. Pages 33-51 in A. Klute, editor. Methods of soil analysis, part 1: physical and mineralogical methods. 2nd edi- tion. Soil Science Society of America, Inc., Madison, WI. Richmond, D. L., and M. L. Richardson. 1974. General soil map and inteqoretations, Mohave County, Ari- zona. U.S. Department of Agriculture, Soil Conserva- tion Sei'vice. SAS. 1985. SAS user's guide: statistics, version 5 edition. SAS Institute, Caiy, NC. SCHLESINGER, W H., P J. FONTEYN, AND W A. REINERS. 1989. Effects of overland flow on plant water rela- tions, erosion, and soil water percolation on a Mojave Desert landscape. Soil Science Society of America Journal 53: 1567-1572. SCHLESINGER, W H., J. E REYNOLDS, G. L. CUNNINGHA.M, L. E HUENNEDE, W M J.'VRRELL, R. A. VIRGINIA, AND W. G. Whitford. 1990. Biological feedbacks in global desertification. Science 247: 1043-1048. Sellers, W. D., and R. H. Hill, editors. 1974. Arizona climate, 1931-1972. University of Arizona Press, Tucson. Tiedemann, a. R. 1987. Nutrient accumulations in pinyon- juniper ecosystems — managing for future site pro- ductivity. Pages 352-359 in R. L. Everett, compiler. Proceedings — pinyon-juniper conference. General Technical Report INT-215. Intermountain Research Station, Ogden, UT Tiedemann, A. R., and J. O. Klemmedson. 1973. Effect of mesquite on physical and chemical properties of the soil. Joumal of Range Management 26: 27-29. 224 Great Basin Naturalist [Volume 56 USDA Bureau of Land Management. 199L Record of licensed livestock, Hualapai Ph. allotment, 1967-199L Kingman Resoince Area, Kingman, AZ. U.S. EPA. 1983. Methods for chemical analysis of water and wastes. EPA-600/4-79-020, revised March 1983, Method 350.1, Method 353.2. ViTOUSEK, P. M., J. R. Gosz, C. C. Grier, J. M. Melillo, AND W. A. Reiners. 1982. A comparative analysis of potential nitrification and nitrate mobility in forest ecosystems. Ecological Monographs 52: 155-177. West, N. E. 1988. Intermountain deserts, shiiib steppes, and woodlands. Pages 209-230 in M. G. Barbour and W. D. Billings, editors. North American terrestrial vegetation. Cambridge University Press, Cambridge. West, N. E., and J. O. Klemmedson. 1978. Stmctural dis- tribution of nitrogen in desert ecosystems. Pages 1-16 in N. E. West and J. Skujins, editors. Nitrogen in desert ecosystems. Dowden, Hutchinson & Ross, Inc., Stroudsburg, PA. Zar, J. H. 1984. Biostatistical analysis. 2nd edition. Pren- tice-Hall, Inc., Englewood Cliffs, NJ. Received 20 April 1995 Accepted 14 March 1996 Great Basin Naturalist 56(3), © 1996, pp. 225-236 DISTRIBUTION AND ECOLOGICAL CHARACTERISTICS OF LEWISIA LONGIPETALA (PIPER) CLAY, A HIGH-ALTITUDE ENDEMIC PLANT Anne S. Halford'-^ and Robert S. Nowak'-'^ Abstract. — Lewisia longipetala (Piper) Clay is a high-altitude endemic Iduiul in llic northern Sierra Nevada. The characteristics of 12 sites with L. longipetala, which represent all known populations, were studied to define habitat requirements of the species. Meso- and microscale characteristics of the habitat were examined, including characteris- tics of the associated plant communit>'. Average plant size and plant density of L. longipetala were also determined for each population. Similar measurements were made on 6 populations of Lewisia pijgniaeu (A. Cray) Robinson, a more common Lewisia. Populations of L. longipetala that had larger plants and higher plant density were associated with gen- tK' sloped, north-facing sites that were near large, persistent snowbanks and had low vegetative cover. Plant species associated with populations of L. longipetala were similar among the 12 sites and were indicative of mcsic, rocky alpine sites. These t\pes of plant commimities found near persistent snowbanks are often termed snow-bed vegetation. In con- trast, L. pijgnmea was found to be less site specific. Lewisia pygmaea was foimd adjacent to or interspersed with L. longipetala at 5 sites, but was found in areas associated with a higher percentage of herbaceous cover and a wider vari- ety' of species. This integration of ecological and commmiity information for L. longi}H'tala populations contributes to the interim management and long-term monitoring of this species by providing needed information concerning its habitat and en\ ironmental specificity. Key words: Lewisia longipetala, Lewisia pygmaea, site cluiracteristies, snow-bed vegetation, alpine, endemic, plant size, plant density. The recent implementation of programs to preserve rare plant taxa indicates the elevated concern for effective and long-term steward- ship of sensitive species (Sutter 1986). One of the initial steps toward the protection of rare plants is to document their occurrences (Utter and Hurst 1990). Mountain ranges are typi- cally rich in endemics (Major 1989), and within the Sierra Nevada they comprise a high per- centage of the flora (Stebbins and Major 1965). Factors that characterize the species habitat are inferred from the species' geographic dis- tribution and often suggest environmentally imposed limitations on the distribution of sen- sitive plant taxa (Baskin and Baskin 1988, Hutchings 1991, Nelson and Haiper 1991). For example, some limitations that influence en- demic plants within alpine environments are snowbank depth and duration (Komarkova 1975, Webber et al. 1976) and levels of disturbance to root systems fi-om needle ice (Fitzgerald et al. 1990). To help ensure the survival of rare plant species, habitat and biological information should be integrated with long-term monitor- ing programs (Sutter 1986, Baskin and Baskin 1988, Hutchings 1991). Species within the genus Lewisia (Portula- caceae) are well known in horticulture (Elliot 1966, Mathew 1989). However, little informa- tion exists regarding these species in their native environments. Only 4 species within the genus Lewisia have relatively wide distribu- tions: Lewisia pygmaea (A. Gray) Robinson, L. nevadensis (A. Gray) Robinson, L. triphyUa (S. Watson) Robinson, and L. rediviva Fursh. The remaining 15 species have considerably smaller distributions, and 9 that occur in California are listed by the U.S. Fish and Wildlife Ser\ace as candidates for threatened or endangered status. Leivisia longipetala (Piper) Clay is a federal candidate 2 species, which implies that data on identifiable threats are insufficient to support federal listing as threatened or endangered (Skinner and Pavlik 1994). Lewisia longipetala is an endemic species with limited distribution that the California Native Plant Society classi- fies as a category 1 B species, which is a cate- gory for rare, threatened, or endangered plants within California. Lewisia longipetala popula- tions are fairly remote, and most exist in U.S. Forest Service wilderness areas. Although L. 'Department ot Environmental and Resource Sciences, Mail Stop 199, University of Nevada at Reno, Reno, NV 89.5.57. ^Present address: Bureau of Land Management. 785 N. Main St. Suite ?. Bishop, CA 93514. ■'Author to whom reprint requests should be submitted. 225 226 Great Basin Naturalist [Volume 56 longipetala populations are not an immediate management concern, one population (Basin Peak) is on private land, and mining claims within close proximity of the site pose a poten- tial threat. FurtheiTnore, the potential also exists for activation of mining claims within wilder- ness areas as well as increased ski area devel- opment within the vicinity of the other L. longi- petala populations. The first specimen of L. longipetala was col- lected by J. G. Lemmon in 1875 in the moun- tains west of TiTickee, California. In 1913, Piper described L. longipetala as Oreohroma longi- petalum, an intermediate between L. pygmaea and L. oppositifolia (S. Watson) Robinson. Later descriptions (Munz 1959) placed L. longipetala as a subspecies of L. pygmaea. More recently, L. longipetala was again recognized as a dis- tinct species (Dempster 1993), a distinction supported b>' moiphological as well as chromo- somal differences between L. longipetala and L. pygmaea (Stebbins 1968, Halford 1992). Lewisia longipetala (Fig. 1) is an herbaceous perennial with a basal tuft of green, linear leaves. An individual plant produces numerous scapes, 30-60 mm long, each bearing 1-3 pale pink flowers with petals 11-20 nmi long. The two sepals are distinctly fuchsia in color, 4-10 Fiy;. 1. Line drawing ni Lcicisia loii^iix-tdld (Piper) Cla\ slum iiiii lirow tli liahil. I99(ij Distribution and Ecolocy oi^ L. longipetala 227 mm long, and coiispicuoush i!;laii(liiIai"-(l(Mitate (Elliot 1966, Mathew 19S9). In contrast, inllo- rescences of L. i)i/30%) with west-, southwest-, or soudieast-facing slopes. Regression analyses between plant density and 9 site characteris- tics did not yield 1 "best" model but rather 2 models that had similar adjusted R^ values (Table 3). For both models, slope was a signifi- cant dependent variable, and plant density was inversely correlated with slope (i.e., as slope increased, plant density decreased). Surface water and surface rock cover were significant dependent variables in 1 model, and plant density was positively correlated with both of these dependent variables. In the 2nd model, total vegetative cover was a significant depen- dent variable, and L. longipetala density was inversely correlated with vegetative cover. Populations with the largest plants generally were also those with the highest plant density (Table 1). Hc-gression analyses between clump diameter and 9 site characteristics yielded a single model that all forward and backward stepwise regressions converged upon (Table 3). Mean plant diameter from each population was inversely correlated with distance from the nearest uphill snowbank. The value of the regression coefficient for surface litter cover was significantly different from zero at the 6% probability level rather than the 5% level, and plant size was inversely correlated with the amount of surface litter cover. Classification and ordination of die floristic data corroborated these results (Fig. 3). Four site characteristics were found to be significantly correlated with DCA axis 1 scores (Table 4), and these site characteristics are shown in Fig. 3A as vectors that indicate the directional increase of slope, surface rock cover, bare ground cover, and total vegetative coven rVVINSPAN classified the 12 populations into 3 groups (Fig. 3A), and the species groupings associated with the TWINSPAN population groupings are shown in Fig. 3B. The Basin Peak populations had higher vegetative cover, whereas the other populations had higher rock cover (Fig. 3A). These populations that are associated with increasing rock cover contain species such as Antennaria media, Cassiope nieHensiana, and Kalmia polifolia var. micropy- Ua that are indicative of such environments (Fig. 3B). L. longipetala populations at Granite Chief, Top Lake, Mt. Price 2, Mt. Price 3, and Table 1. Descriptive site attributes for 12 Lcicisia longipetala populations, ordered from north to south. Mean ± stan- dard error ot plant diameter from 20 randomly selected plants, as well as plant density, is gi\en for each population. Parent Elevation Slope Plant diameter Plant density Population material (m) Aspect m (cm) (# per 0.5 ha) Basin Peak 1 Basalt 2800 NNE 2-8 9.7 ± 0.3 185 Basin Peak 2 Basalt 2840 NNE 2-8 3.9 ±0.1 10 Pole Creek 1 Basalt 2733 NNE 2-6 13.0 ±0.9 >.500 Pole Creek 2 Basalt 2733 NNE 2-6 8.2 ± 0.4 >.500 Granite Chief Granite 2800 N >.30 6.5 ±0.1 135 Dick s Lake Granite 3033 NNE 2-10 8.6 ±0.4 >.500 Top Lake Granite 2866 W >30 4.2 ± 0.2 12 Mt. Price 3 Granite 31.33 wsw >30 6.3 ± 0.4 35 Mt. Price 1 Granite 3200 SSE 2-8 3.4 ±0.2 40 Keith's Dome Granite 2800 NNE 2-8 10.8 ±0.4 >,500 Mt. Price 2 Granite 2966 ssw >30 8.3 ±0.3 30 Pyramid Peak Granite 2787 WNW >30 4.1 ±0.2 25 230 Great Basin Naturalist [Volume 56 Table 2. Mean percent cover for species found within Lewisia longipetahi populations and the number of L. longipetala populations that contained that species. Species are listed from highest to lowest cover. Hickman (1993) was used as the authorit\' for all species. Letter codes used in Figures .3 and 4 are gi\'cn in brackets for each species. Species Species code Mean cover #of pop. Carex scopuloriim Holm, var bracteosa (L. Bailey) E Hemi. Antennaria media E. Greene Juncus mertensiamis Bong. Erigeron peregrinus (Pursh) E. Greene Lupinus hreweri A. Gray Lewisia pijgmaea (A. Gray) Robinson Lewisia longipetala (Piper) Clay Arnica mollis Hook. Mimulus guttatus DG. Salix artica Pallus Aster alpigemis (Toirey & A. Gray) A. Gray ssp. andcrsonnii (A. Gray) M. Peck Calijptridium umheUatum (Torrey) E. Greene Phletim alpimim L. Juncus drummondii E. Meyer Sibhaldia procumbens L. Dodccatheon alpinum (A. Gray) E. Greene Cassiupe mertensiana (Bong.) Don Kabnia poIifoUa Wangenh. ssp. microphijlla (Hook.) Galder & Roy Taylor Ltjcopodiuin sp. Mi)nubis primuloidcs Benth. Foa wheeleri Vasey Polygonum bistortoides Pursh Eriogonum incamim (Torrey & A. Gray) Penstemon rijdbergii Nelson ssp. oreocharis (E. Greene) N. Holmgren PhyUodoce breweri (A. Gray) Ma.xim. Anemone drummondii S. Watson Poa secimda J.S. Presl ssp. secunda Sedum roseum (L.) Scop. ssp. integrifolium (Raf ) Hulten [Gascb] 6.2 9 [Anme] 5.1 9 Qume] 4.2 4 [Erpe] 2.9 2 [Lubr] 2.6 1 [Lepy] 1.9 5 [Lelo] 1.7 12 [Anno] 1.4 1 [Migu] 1.2 3 [Saar] 1.2 3 [Asala] 1.0 5 [Gaum] 0.9 4 [Phal] 0.9 2 [Judr] 0.8 3 [Sipr] 0.8 4 [Doal] 0.7 2 [Came] 0.7 2 [Kapom] 0.5 2 [Lycsp] 0.5 5 [Mipr] 0.5 3 [Powh] 0.3 1 [Pobi] 0.3 2 [Erin] 0.2 1 [Peryo] 0.2 1 [Phbr] 0.2 2 [Andr] 0.1 1 [Poses] 0.1 1 [Seroi] 0.1 1 Pyramid Peak were situated iu cracks on steep granitic sU^bs, and one of the most common species found associated with these sites was rock sedum {Sedum roseum ssp. integrifolium). NJMDS results differed slightly from the TWINSPAN classification by the separation of the Top Lake population from all other popula- tions and by a change in association of the Basin Peak 2 population from the group of 3 Basin Peak 1 plots to the group of 7 plots to the right of Basin Peak 2. In the 2-dimensional space defined by DCA axes 1 and 2, Basin Peak 2 appears to be transitional in its floristic composition between the Basin Peak 1 plots and this group of 7 populations. Species that contributed to these different classifications of the Basin Peak populations were Phleum alpin- um and Lupinus breweri: P. alpinwn was within Table 3. Multiple regression results between each of 2 dependent xariables (plant density' and plant diameter") and the set of 9 site characteristics for 12 Lewisia longipetala populations. Dependent variable Regression statistics adj. R- Variable statistics Model variables CoefRcient P Slope -18.5 <0.01 Surface water cover 11.2 <0.01 Sm-face lock co\er 6.9 0.01 Slope -16.6 <0.01 Total vegetati\e co\'er -6.5 <0.()I Snowbank distance -0.07 <0.01 Surface litter cover -0.24 0.06 Plant density Plant diameter 12 12 12 0.81 <0.01 0.81 0.53 <0.01 0.01 1996] DlSIRlliU IKJN AND EcoUKiV OF L. LONGIPETALA 231 400 300 200 100 CN < A Vegetation cover Bareground cover ick's Lk Q^eith's Dome PoleCkl O OPoleCk2 BasinPk2 0; ^^ O ^Pyramid Pk Mt Price 3 Rock cover > 100 200 300 400 DCA Axis 1 500 600 Fig. 3. A, Population ordinations generated by DCA for L. longipetala; B, species groupings associated with the popula- tions. For both graphs, circled groups were determined from TWINSPAN dench-ograms; broken hues indicate NMDS groupings. Letter codes for each species are given in Table 2. 232 Great Basin Naturalist [Volume 56 Table 4. Conelation coefficients generated from a rota- tional correlation program for all Lewisia lon^ipctala and Lewisia pijginaea DCA axis 1 scores. Variables with an * are significant at the 0.05 level. Species Vniables Correlation coefficients Le w is id longipetala Elevation 0.24 *Slope -0.58 Aspect -0.49 Snowbank distance 0.41 *Bare ground cover 0.66 Litter cover 0.36 *Surface rock cover 0.71 Surface water cover 0.11 *Total vegetative cover 0.67 Lewisia pygimiea Elevation -0.53 Slope 0.17 Aspect -0.14 Snowbank distance 0.38 *Bare ground cover 0.92 * Litter cover 0.92 Surface rock cover 0.34 Surface water cover 0.26 *Total vegetative cover 0.78 the sampled transects of both Basin Peak 1 and 2 populations but not within any of the other populations; on the other hand, L. breweri was only within the Basin Peak 1 populations. In general, it should be noted that classification inferences based solely on location within the 2-dimensional space defined by any 2 DCA axes can be misleading: for example, Mt. Price 1 and Mt. Price 2 are close together in the 2- dimensional space defined by DCA axes 1 and 2, but they do not classify into the same group in either TWINSPAN or NMDS because they are on different planes in the 3-dimensional space defined by the addition of a 3rd axis. Lewisia pygmaea Lewisia pygmaea grew in areas where total vegetative cover was greater than that where L. longipetala was found. The strongest evi- dence for this difference in site characteristics was from the 5 sites where L. longipetala and L. pygmaea coexisted in proximit)' to each other: Basin Peak 1, Basin Peak 2, Granite Chief, Dick's Lake, and Keith s Dome. Total vegeta- tive cover for areas with L. pygmaea averaged 60.4 (s^: 5.9), which was 55% greater than the mean cover of .39.0 {sj: 12.3) for areas with L. longijH'tala: tliis difference was significant at P < 0.10 (paired t test, 4 d.f, P = 0.067). This large difference in vegetative cover persisted even when all populations were considered: for all the known L. longipetala populations, mean vegetative cover was 31.8 (s^: 5.8); for the 6 L. pygmaea populations used in this study, mean vegetative cover was 53.7 {s^- 8.3). Although this difference was significant (2- sample t test, 16 d.f, P = 0.046), note that our original selection of L. pygmaea populations was not designed to be a random sample of all L. pygmaea populations and thus extrapolation to all L. pygmaea populations is not statistically justified. TWINSPAN results for L. pygmaea popula- tions grouped the Basin Peak populations sep- arately from other populations (Fig. 4A), but the environmental site attributes that were sig- nificantly correlated with DCA axis 1 scores differed between L. pygmaea and L. longi- petala (Table 4). Litter cover was a significant site attribute for L. pygmaea, but slope and surface rock cover were not. The vegetative cover vector increased toward the Basin Peak population, which suggested that these popula- tions contained a greater herbaceous compo- nent. The species indicative of such areas include Erigeron peregrinus, Sali.x artica, and Arniea mollis (Fig. 4B). The bare ground vector also increased toward the Basin Peak stands. Although the concomitant increases in bare ground and vegetative cover may seem contra- dictoiy, surface rock cover tended to decrease toward Basin Peak. Thus, smface rock was replaced by vegetation and bare ground (i.e., inorganic soil) along these vectors. High litter cover was commonK' associated with the Gran- ite Chief and Keith s Dome populations. Piute Pass is an area historically thought to contain L. longipetala. However, only L. pyg- nmea indixiduals were verified at this site. The area is south of Yosemite National Park, Cali- fornia, which makes it the southernmost site surveyed for L. longipetala. Environmental attributes of Piute Pass are similar to those of other L. longipetala and L. pygmaea popula- tions, except for some differences in species composition, nameK the relative preponder- ance of Dodeeatlieon Jeffreyi. Discussion Site chaiacteristics that were most highly associated with the occurrence of L. longi- petala and also correlated with plant size and 1996] DlSTKIBUTION AND EcXiUKiY OK L. LONGIPETALA 233 400 300 A Litter cover Vegetation cover Bareground cover -300 -TOO 00 200 300 400 500 DCA Axis 1 Fig. 4. A, Population ordinations generated by DCA for L. pygmaea; B, species groupings associated with the popula- tions. For both graphs, circled groups were determined from TWINSPAN dendrograms. Letter codes for each species are given in Table 2, except Doje = Dodecatheon jeffreiji. 234 Great Basin Naturalist [Volume 56 density were proximity of snowbanks, steep- ness of slope, slope aspect, and cover of vege- tation, surface rock, and surface water. These inferences are supported by inspections of the site characteristics and by statistical analyses. The L. longipetala populations with higher densit\' were found on gently sloping sites with a northern exposure that were close to snow- banks and had low vegetative cover of all species. For example, Pole Creek 1 and 2, Keitli s Dome, and Dick's Lake have populations that exceeded 500 individuals, whereas Basin Peak, which overall had the most herbaceous cover of any of the other populations, had much lower plant density. Plant density of L. longi- petala populations increased with increased cover of surface water and rock, but decreased with total vegetative cover and slope steep- ness. Plant size, as measured by clump diame- ter, increased with decreased distance from snowbanks and decreased litter cover Further- more, at Basin Peak 1 as well as other sites of L. longipetala populations, plants that were more distant fi'om snowbanks or that were on south-facing slopes were more water stressed (Halford 1992). Site characteristics that are associated with more vigorous L. longipetala populations are indicative of areas that receive high snowpack accumulations. In alpine environments plant communities whose occurrences are influ- enced by geomoiphological characteristics that favor high snowpack accumulations are often termed snow-bed vegetation (Billings and Bliss 1959, Kuramoto and Bliss 1970, Canaday and Fonda 1974, Tomaselli 1991). Some species in the Siena Nevada that Major and Taylor (1977) commonly found associated with areas of high snowpack accumulations and that often occur in mesic depressions with low vegetative cover are Fhijllodoce hreweri, Cassiopc inciiensiana, Kaltnia polijolia van niicrophylla, Fhleiun alpinum, Mimulus primuloides, and M. gutta- tus. Additional species that occm- in mesic to even hydric habitats include Antennaria media, Sibbaldia procumhens, Dodecatheon alpinwn, and Sedwn roseum (Major and Taylor 1977). These species were tilso associated with L. longi- petala populations. Conversely, species that are more frequently associated with xeric sites, such as Lupinus hreweri and Juncus drwn- mondii (Chabot and Billings 1971, Nachlinger 1985), were less fre(|uently associated with L. longipetala populations. The restriction of some species to sites with low vegetative cover may be related to reduced interspecific competition (Ostler et al. 1982). For example, competition partially accounts for the reduced growth of Talinum calcaricwn, a highly restricted rock outcrop species of the Portulacaceae family, in herbaceous sites domi- nated by Poa pratensis (Ware 1991). Viable populations of the endangered Furbish's louse- wort {Pedicularis furbishiae) occur on mesic, rocky sites that experience intermediate distur- bances from hydrological processes, which remove potential competitors (Menges 1990). Potentilla robbinsiana, an endemic from New Hampshire's White Mountains, also requires rocky mesic sites that are moderately dis- turbed, in this case by frost heaving that limits other species (Fitzgerald et al. 1990). The lower densities and smaller L. longipetala plants in areas \\dth high vegetative cover and high soil organic matter (Halford 1992) suggest that interspecific competition may also restrict this species, but specific studies need to be con- ducted to explicitly test this mechanism. The environmental site characteristics of L. pijgmaea are broader than those of L. longi- petala. Populations of L. pygmaea have been documented in dense herbaceous meadows, cracks in steep rocks, and open gravely depres- sions (Elliot 1966, Major and Taylor 1977). In our study plants of L. pygmaea were found adjacent to 3 and interspersed with 2 of the 12 L. longipetala populations, which suggests that L. pygmaea and L. longipetala can grow in sim- ilar environments. However, an important dif- ference between the 2 species is that L. pyg- maea was found in areas with more herbaceous cover The less pronounced site specificity ex- hibited by L. pygmaea parallels other widely distributed, mesic alpine species, whereas the relative restriction of L. longipetala to more open sites is similar to other restricted plant taxa (Fitzgerald et al. 1990, Menges 1990). The potential threats to L. longipetala are not imminent at this time but include both sto- chastic and anthropogenic processes. Climatic events such as periodic droughts that reduce snowpack accmnulations as well as potential in- creases in interspecific competition may signif- icantly reduce die viability' of L. longipetala pop- ulations, especially those that already have low densities of individuals. Human activities may also ha\'e significant impacts. For example, if slopes above populations are altered by mining 1996] DiS I lUBUTION AND EcOLOCiY OF L. LONGIPETALA 235 activity or ski area cle\ t^lopment, the displace- ment of substrate could alter the topograph) and hence hydrology of the site through changes in snow acciunulation and melt water rimoff To enhance the long-term viability of this endemic species, primary management goals should include (1) monitoring of L. lonciisis + , O. tuhcrctilata, and Thrassis iHin(lorae + ), 1 X^iok (Ixodes sculptu.s + ), and an eyeworni (Nematoda: Hhalxliti.s ()rhitalis* + , also 1st records from Sciuridae); S. /;. ctulcmiciis was host to a louse species {Neohaematopimis la('iiiisciilii.s + ). 5 flea ta.\a (RJuidmopsijUu sp. + , (). t. tubcrculata, Tlirassis f. fr(incm + , T. f. b(irnesi + , and T. f. rockwoodi), and a mite {Aiidn)l(i('lai).s J(ilircnht)lzi + ). S))enni)pliilus hrimneus had fewer known ectoparasite species than other congeners. Although all of their parasites had many other hosts, S. h. endeinicus and S. h. bntnneiis shared only a single parasite species in common, whereas all but one of their eetoparasites also occurred on the closely related Townsend's ground squirrel (S. townsendii). The proportion of parasitized individuals and the para- site loads per individual were significantly lower in S. b. bninneus, which lives in small, isolated populations, than in S. b. cndemlcus, which has larger, less fragmented populations, suggesting a relationship between host population structure, parasite loads, and parasite species diversity. All but one of the flea species have been linked to plague transmission. Keij words: ground squirrels, eetoparasites, Spermophilus brnnneus, Idaho. Tlie Idaho ground squirrel {Spermophilus bninneus) is one of die rarest and, until recently, least known North American mammals (Sher- man 1989, Yensen 1991, Yensen and Sherman in press). This endemic species inhabits a 125 X 90-km area in west central Idaho, but it actually occupies only a small fraction of this limited range (Yensen 1991). Despite the species' restricted geographic distribution, there are 2 allopatric subspecies that are mor- phologically and genetically differentiated and possibly have reached species-level separation (Yensen 1991, Gill and Yensen 1992, Gavin et al. submitted). Spermophilus b. brunneus occurs in montane meadows surrounded by coniferous forests at elevations of 1035 to 1550 m in Adams and Val- ley counties (Yensen 1991). As of 1995, only 18 of the 28 known populations remained, and only one of these contained >100 animals. The majority of the sites were within an area of 22 X 9 km and totaled <300 ha of occupied habi- tat (T A. Gavin, E W. Sherman, and E. Yensen unpublished data). Fire supression began in the area about 100 yr ago. Subsequent succession and expansion of forests has filled in many of the natural meadows in the range of S. /;. brunneus (Truksa and Yensen 1990), eliminating habitat. The remaining populations are presently isolated from each other by the encroachment of coni- fers into meadows and by competition with Columbian ground squirrels (Yensen and Sher- man in press). Today, there is apparently little or no gene flow among populations. Allozyme analyses of 55 protein loci in 12 populations (Gavin et al. submitted) indicated that the pro- portion of polymoiphic loci was 11.5%-19.2% and heterozygosity values were 0.041-0.080. Fj.f was 0.317, implying that there is genetic differentiation among populations despite their geographic proximity and the apparent recency of their separation. In 1993 the total number of individual S. b. brunneus was 1000-1200, but the number fell to 600-800 in 1994 and 1995 (T A. Gavin, E W. Sherman, E. Yensen per- sonal observation). Spermophihis b. endemicus occurs in rolling foothills at elexations of 670 to 975 m in Gem, Payette, and Washington counties (Yensen 1991). It is patchily distributed throughout its range of 75 X 30 km. Although censuses of S. /;. endemicus populations have not been made, its total population is apparently much larger than that of S. h. brunneus. The area occupied, esti- mates of population densities, and the amount iMuseum of Natural History, Albertson College, Caldwell, ID 83605. ^University of Idaho, Pamia Research and Extension Center, Pamia, ID 836fi(). ■^Section of Neurobiology and Behavior. Cornell Universih', Ithaca, NY 14853. 237 238 Great Basin Naturalist [Volume 56 of remaining habitat are more than 2 orders of magnitude greater than for S. b. brimneus (E. Yensen personal obsei'vation). Parasites of S. brunneiis have not been pre- viously sui^veyed. The only prior records (Baird and Saunders 1992) were 2 flea species, Oro- psijUa t. tuberciilata and Thrassis francisi rock- woodi, collected from specimens now referred to S. b. endemicus (Yensen 1991). We were interested in how ectoparasite diversit>' and density are affected by reduction in size and isolation of host populations. According to epidemiological models (Ander- son and May 1979, May and Anderson 1979), the number of contacts between hosts and in- fective stages of parasites determines the rate at which adult parasites are acquired. Mean parasite load should equal growth rate of the population divided by mortality from the dis- ease. Thus, as population growth slows, para- site load per individual should drop. At veiy low host population densities, there may be too few contacts even to maintain ectoparasite popu- lations. Thus, we predicted that S. b. brimneus should have fewer ectoparasite species and fewer ectoparasites per individual than con- generic, more widely distributed western ground squirrels {Spennophilus spp.). We also predicted that due to its fragmented popula- tion structure and smaller population sizes, S. b. brimneus should have fewer ectoparasite species than S. b. endemicus. Because of questions about the taxonomic similarity of S. /;. brunneiis and S. b. endemi- cus, we also wished to leani if they had similar ectoparasites, and how similar their ectopara- sites were to those of other western ground squirrels. Further, because of the limited geo- graphic range and low number of small popu- lations, both subspecies of S. brunneiis would be vulnerable to extirpation liy an epizootic such as plague. Thus, it was important to learn if their ectoparasites were species involved in plague transmission. Methods From 1980 to 1990, specimens of S. brunneiis were collected for a taxonomic study (Yensen 1991). To minimize negative impacts on small populations, a mean of 0.5 individuals/site/yr of S. brunneus was collected. Squirrels were killed by shooting or by live-trapping and injecting nembutol into the heart. ImmediateK post- mortem, squirrels were placed individually in plastic bags; fleas, ticks, lice, and larger mites were collected with forceps or a camel's hair brush moistened with 70% ethanol as they left the host. Squinels were not examined under a dissecting microscope, so smaller mites were not collected; eyes were not examined for eye- worms. From 1987 to 1994, S. b. brunneus were live -trapped for demographic and behavioral studies (Sherman 1989, and ongoing). They were hand-held and parasites were picked off with forceps; because the animals were not anesthetized, all of the smaller and some of the larger ectoparasites may not have been seen. Eyes were checked for eyeworms by pulling back the upper lid; specimens were removed from the cornea of the eye with a cotton swab moistened with sterile water. All parasites were placed in 70% ethanol. In addition, 21 S. b. endemicus were live-trapped at Sand Hollow, Payette Count)', Idaho, in 1994 and examined for eyewomis. Collected specimens of S. brunneus were prepared as standard museum study skins and skiills and deposited in the Albertson College Museum of Natural Histoiy (ACMNH), Cald- well, Idaho, and the National Museum of Nat- ural Histoiy (USNM); diey are identified below by museum number Specimens of ectopara- sites were sent to appropriate specialists for identification and deposited in the entomologi- cal collections at the University of Idaho, Moscow, and ACMNH. Differences in parasite loads between individuals and subspecies were analyzed with hand-calculated Mann-Whitney C^-tests and chi-square tests, as appropriate. Results We examined 29 freshly collected individu- als of S. b. brimneus and 53 of S. b. endemicus for ectoparasites. These represent 43% of the 192 museum specimens of this species known to us (Yensen 1991, plus 4 additiouiil specimens). AdditionalK', we opportunisticalK' collected ecto- parasitic arthropods from 12 lixe-trapped indi- \ iduals of S. b. brunneus and eyeworms from another 36; we examined 21 S. b. endemicus for eyeworms. We collected 6 ectoparasite species from Spennophilus b. brunneus: 4 fleas, 1 tick, and 1 nematode (Table 1). We collected 7 taxa of ectoparasites from S. 1). endemicus: 5 fleas, 1 louse, and 1 mite. 1996] SpERMOPHILUS BRUNNEUS Ec rOPARASITES 239 Tahlk 1. Parasites (if S. hninneits that also occur on some other species of western j^round squin-els (subgenus Sper- inopliilti.'i). S\ inhols: * = knowni priniaiA' host(s); + = records, possibly accidental on host; - = no records in references bi'iow''. Host This study Lit erature reco -ds'> Parasite Sbb Sbe Sto Sec Sbl Sar Seh- Sri Swa Sp>' \ACE \c()l}(inntit(>f)iiiti.s lacriiisciiluti - + + + - + + + - + V\a:\s \c(>i)siilhi inopiiia + - + + + + + + - _ Oropst/Ild icialiocn.sis + - + * * * * + _ * O. t. tiiberculata + + + + + + _ + + _ RhadiiiopsijUa s. scctilis - ? + - - - - - + _ Thrassis f. barnesi - + + - - + + - _ _ T. f. francisi - + * - + + - _ _ _ T.f. rockwoodi - + + - * - + _ _ _ T. p. pandorae + - + + * * * + + - i'lCKS Ixodes scidptus + - + + + + + + - - Mrn:s Aiidroladups jahrcnholzi - + + + - + + + - + Nematoda Rliahdiiis orhitalis + - - - - - - - - - •'lYom records in Hulibard (1947), Burgess (1955), Stark (1970), Hilton and Mahrt (1971), Wliitaker and Wilson (1974), Holekamp (1983), Lewis at al. (1988), Baird and Saunders (1992), Baird (unpul)lished), and this study. ''Host acronyms: Sbb = Spennophtlus h. bnmnetis, Sbe = S. b. endemictK:, Sto = S. "totvnsendii" (sensu latu), Sec = S. columbianus, Sbl = S. beklingi, Sar = S. iinniittis, Sel = S. clcgans. Sri = S. richardsonii, Swa = S. washingtoni. Spy — S. pamjii. 'Confused in the literature with S. richardsonii. The records here are those that unambiguously refer to this species, and the total for S. richardsonii may include a few parasites of this species. The proportion of parasitized individuals in the 2 subspecies was strikingly different. We found ectoparasites on 37 of 53 (70%) S. h. endemicus but on only 8 of 29 (28%) S. b. briin- neus collected (x^ = 13.4, d.f = 1, P < 0.001). Parasitized individuals of S. b. brunneus had 1-3 species of ectoparasites each {X = 1.75, n = 8), and parasitized individuals of S. b. endemicus had 1-4 species of ectoparasites (X = 1.59, n = 37). This difference was not signif- icant (Uj = 154, P > 0.5). However, there was a significant difference between subspecies in the parasite load of parasitized individuals. Fleas were the only common group of ectopar- asites of both ground squirrel taxa. There were 4.1 fleas per parasitized individual in S. b. brunneus and 7.8 in S. b. endemicus (U^ = 95.5, P < 0.05). Annotated List of Ectoparasites In the ectoparasite species accounts below, letters and numbers in brackets refer to the number of male and female fleas, e.g., [1 m, 2 f], or to conversions of original collecting data to latitude, longitude, and metric units. Anopleura: Haematopinidae Neohaemafopinus laeviiisculus (Grube) We found this louse on S. b. endemicus in the following locations: 11 mi [18 km] N Emmett, Gem Co., T8N, R2W, Sec. 13 [44°02'N, 116°31'W, 830 m elev], 21 February 1982 (ACMNH 222), 28 February 1982 (ACMNH 226, 227, 236, 237, 238); 0.1 mi E Payette Co. line, 12.6 mi [20 km] N Emmett, Gem Co., T8N, R2W, Sec. 12 [44°03'N, 116°32'W, 810 m elev], 28 February 1982 (ACMNH 224); Weiser Cove, Washington Co. [44°13'N, 116°44'W 715 m elev], 7 March 1982 (ACMNH 228, 229, 230); lower Mann Creek, 2.5 mi [4 km] N jet. Weiser River Road, Washington Co. [44°16'N, 116°51'W, 720 m elev], 14 March 1982 (ACMNH 231, 240, 242, 243, 244). This louse occurs fi-om Eurasia east to Alaska and the Northwest Territories, and south through western United States to Mexico; it is apparently a species complex (K. C. Emerson personal communication). Lice of this complex have been collected from many ground squir- rels (Eurasian Spermophilus major, S. citellus, S. pygmaeus, S. undulatus, and North American 240 Great Basin Naturalist [Volume 56 S. beecheyi, S. armatus, S. beldingi, S. colum- bianus, S. parryii, S. townsendii, S. washing- toni, and Ammospennophilus leucunis), as well as marmots [Marmota flaviventris), chipmunks {Tamias niiniiniis), pocket mice {PerognatJiiis parvus), and deer mice {Peromijscus manicida- tus; Raybum et al. 1975, Shaw and Hood 1975, records fiom National Museum of Natural His- toiy). Although N. kieviiiscidus is the most com- mon louse species taken fi-om ground squirrels in Idaho (C. R. Baird personal communication, K. C. Emerson personal communication), S. b. endemicus is a new host record. Siphonaptera: Hystrichopsyllidae Neopsyllo inopina Rothschild We collected 8 individuals of this flea from S. b. brunneus in the following locations: Lick Creek, Adams Co., T19N, R3W, Sec. 14 [44°59'N, 116°40'W, 1290 m elev.], 17 April 1983 (ACMNH 305 [1 m, 2 f], ACMNH 306 [1 f]); 1 mi [1.6 km] NE Bear Guard Station, Adams Co. [45°05'N, 116°37'W, 1480 m], 2 June 1988 (ACMNH 518 [1 f]); and Price Valley [45°01'N, 116°26°W, 1270 m elev.], 3 June 1981 (ACMNH 209 [1 m, 1 f], ACMNH 210 [If]). This flea occurs from British Columbia south to Oregon and Nevada and east to Saskatchewan, North Dakota, and Utah (Lewis et al. 1988). It has been collected from other western ground squirrels of subgenus Sper- mophdiis (Table 1) and from badger [Taxidea taxiis) dens (Lewis et al. 1988, Baird and Saun- ders 1992); S. b. brunneus is a new host record. RJiadinopsylla sp. We collected 1 female specimen of this flea genus from S. b. endemicus. UnfortunateK', it could not be identified to species. The locality' was Diy Creek Road, Payette Co., 1.4 mi [2.2 km] E Litde Willow Creek, T9N, R2W, Sec. 18 [44°07'N, 116°37'W, 815 m elev.], 26 February 1983 (ACMNH 318 [1 f], reported in Baird and Saunders 1992). The flea is most likely R. s. secfdis, which occurs in many western states on deer mice {Peronujscus sp.) and ground squirrels, includ- ing S. townsendii and S. washingtoni (Lewis et al. 1988, Baird and Saunders 1992). Rhadino- psylla are uncommon fleas and have popula- tion peaks in the colder months (Lewis et al. 1988). This is the 1st record of any Rhadino- psylla species Iroiii N. brunneus. Siphonaptera: CeratophyUidae Oropsylla idahoensis (Baker) This flea species was collected on S. b. brunneus at the following locations: Price Val- ley [45°01'N, 116°26°W, 1270 m elev.], 3 June 1981 (ACMNH 209 [3 fj); and OX Ranch 1-2 km S, 1-2 km E Bear, Adams Co. [45°00'N, 116°39'W, 1340 m elev.] (live-trapping collec- tions). Oropsylla idahoensis occurs from Alaska to New Mexico and is one of the most common fleas of ground squirrels in the Rocky Moun- tains and westward. Hosts include other west- ern ground sc^uirrels of subgenus Sj}ermophihis (Table 1), golden-mantled ground squirrels (S. lateralis), and marmots {Marmota sp.; Lewis et al. 1988, Baird and Saunders 1992); S. b. brun- neus is a new host record. Oropsylla tuberculata tuberculata (Baker) This was the most common flea on both S. b. brunneus and S. b. endemicus, occurring at nearly all locations fi'om which we collected ectoparasites. We found O. t. tuberculata on S. b. brimneus at the following localities: Price Vallev [45°01'N, 116°26°W, 1270 m elev.], 3 June 1981 (ACMNH 209 [1 m], ACMNH 210 [1 m]); MiH Creek summit, 5 km N Hornet Guard Station, Adams Co., T18N, R3W, Sec. 25, 4500' elev. [44°53'N, 116°39'W, 1370 m], 2 June 1985 (ACMNH 510 [2 m, 3 f], ACMNH 512 [2 m, 3 f]); Lick Creek, Adams Co., T19N, R3W, Sec. 14 [44°54'N, 116°40'W, 1290 m elev.], 17 April 1983 (ACMNH 305 [4 m, 3 f], ACMNH 306 [1 fj); Round Vallev, Vallev Co. [44°21'N, 116°00'W, 1460 m elev.], 18 Mav 1985 (ACMNH 315 [1 f]). Records from S. b. endemicus are as follows: Sucker Cr 11 mi [18 km] N Emmett, Gem Co., T8N, R2W, Sec. 13 [44°02'N, 116°31'W, 830 m elev.], 21 FebruaiT 1982 (ACMNH 221, 222, 223), 28 Februarv 1982 (ACMNH 225, 226, 227), 3 May 1987 (ACMNH 544 [1 m]); 0.1 mi E Payette Co. line, 12.6 mi [20 km] N Emmett, Gem' Co., T8N, R2W, Sec. 12 [44°03'N, 116°32'W, 810 m], 28 Februaiy 1982 (ACMNH 224, 236, 237, 238; reported in Baird and Saunders 1992); Diy Creek Road, 1.4 mi [2.2 km] E Little Willow Creek, Payette Co., T4N, R2W, Sec. 18 [44°07'N, 116°37'W, 815 m elcN.], 20 FebruaiT 1983 (ACMNH 318 [10 m, 13 f]), 26 Februaiy 1983 (ACMNH 317 [8 m, 3 f]); Weiser Coxe', \\ashington Co. [44°13'N, 116°44'W, 715 m ele\.], 7 March 1982 1996] Sl'EliMOl'lUlA 'S BIWNNEUS ECTO PARASITES 241 (ACMNH 228, 229, 230); lower Mann Civck. 2.5 mi [4 km] N jet. Weiser Ki\ er Road, Wash- ington Co. [44°13'N, llCrsrW, 720 m elev.], 14 Mareh 1982 (ACMNH 231, 232, 233, 240, 242, 243, 244); Washington Co., lower Mann Creek, 3.3 mi [5.3 km] N jet. Weiser River Road [44°17'N, 116°51'W,"730 m elev.], 14 Nlareh 1982 (ACMNH 239). This is a \er)' common flea in most of the western United States and western Canadian provinces (Baird and Saunders 1992). Hosts include other \\'estem ground squirrels of sub- genus Spennophihis (Table 1), antelope ground squirrels {A)ninospennophilus Icuciirus), wood- rats {Neototna sp.), and badgers (Lewis et al. 1988, Baird and Saunders 1992). It was previ- ously recorded from S. hriinneus by Baird and Saunders (1992). Thrassis pandorac pandorae Jellison We iound 1 specimen of this flea on S. b. briinnens at Lick Creek, Adams Co., T19N, R3W, Sec. 14 [44°54'N, 116°40'W 1290 m elev.], 17 April 1983 (ACMNH 305 [1 m]). This flea is distributed from Washington to California and east to Colorado (Stark 1970). It is found most fi^equently on Sperrnophilus artna- tiis, S. beldingu and S. elegans (= richardsonii in Stark 1970), but also occurs on S. cohiinbiamis, S. elegons (Table 1), and a variety of other rodents, lagomorphs, and carnivores (Stark 1970). S. b. bninneus is a new host record. Thrassis francisi barnesi Stark We found this flea on S. b. endemicus at Sucker Cr. 11 mi [18 km] N Emmett, Gem Co., T8N, R2W, Sec. 13 [44°02'N, 116°31'W, 830 m elev.], 31 May 1981 (ACMNH 220 [3 m, 4 f]), 3 May 1987 (ACMNH 540 [4 m, 3f], ACMNH 541 [2 m, 1 f], ACMNH 542 [1 m, 6 f], ACMNH 543 [4 m, 7 f], ACMNH 544 [1 m, 1 f], ACMNH 545 [2 m, 1 f], ACMNH 547 [4 m, 9 f], ACMNH 548 [1 f], ACMNH 549 [3 m, 7 f]); 7 mi [11 km] N Emmett, Gem Co., T7N, RIW Sec. 5 [43°58'N, 116°29'W, 920 m elev], 23 May 1987 (ACMNH 546 [4 m, 2 f]); Sand Hollow, 5.6 km N, 5.0 km E Payette, Payette Co., T9N, R4W, Sec. 7 [44°08'N, 116°51'W, 750 m elev.], 30 March 1989 (USNM 565927 [3 m, 2 f]). This flea occurs north of the Snake River in western Idaho, and on both sides of the river in eastern Idaho and south into central Utah and eastern Nevada (Stark 1970). Its most connnon hosts are S. annafiis and S. elegans, rather than S. townsendii mollis, the usual host oi' T. f francisi. Stark (1970) felt that host asso- ciations ma\ separate the 2 subspecies of T. francisi, although the 2 lleas appeared to inter- grade in eastern Nevada. S. b. endemicus is a new host record. Thrassis francisi fnnicisi (Fox) We collected 14 indi\iduals of this Ilea from S. b. endemicus at 1 locality: Dry Creek Road, 1.4 mi [2.2 km] E Little Willow Creek, Pavette Co., T4N, R2W, Sec. 18 [44°07'N, 116°37'W, 815 m elev.], 26 February 1983 (ACMNH 318 [1 m, 5 f], SM2 [2 m, 3 f]), 24 Februar^' 1986 (ACMNH 920 [2 m, If]). This flea is known from the Great Basin desert of eastern Oregon, Idaho south of the Snake River, eastern Nevada, Utah, and parts of Wyoming. It occins primarily on S. town- sendii, but the white-tailed prairie dog {Cyno- mys leucurus) is the usual host in Wyoming (Stark 1970). There are incidental records from several species of ground squirrels (Table 1), marmots, and deer mice (Stark 1970). Our records are the 1st from any host north of the Snake River in Idaho (Stark 1970, Lewis et al. 1988, Baird and Saunders 1992); S. /;. endemi- cus is a new host record. Thrassis francisi rockwoodi Hubbard Two males of this flea were collected from S. b. endemicus at a single locality: Sucker Creek, 11 mi [18 km] N Emmett, Gem Co., T8N, R2W, Sec. 13 [44°02'N, 116°31'W, 830 m elev.], 21 Februaiy 1982 (ACMNH 223), 28 February 1982 (ACMNH 227 [2 m]; reported in Baird and Saunders 1992). This subspecies has been recorded liom east- ern Oregon, northwestern Nevada, and north- ern California, where it occurs most commonly on S. beldingi, although collections ha\'e been made from S. townsendii (Stark 1970, Lewis et al. 1988). Acarina: I.xodidae Ixodes sculptus Neumann We collected specimens of this tick fiom S. b. brunneus at 1 localitv: OX Ranch 1-2 km S, 1-2 km E Bear, Adams Co. [45°00'N, 116°39'W, 1340 m elev.] (live-trapping collections). This widespread tick occurs from western Canada south to California and Texas and east across the Great Plains. It occurs on several 242 Great Basin Naturalist [Volume 56 western ground squirrels of the subgenus SpermopJuhis (Table 1), prairie clogs {Cynomijs sp.), marmots, voles {Microtus sp.), pikas [Ochotona sp.), gophers {Thomonit/s sp.), jump- ing mice {Zapiis sp.), domestic animals, and various carnivores (Doss et al. 1974). S. h. hrimneus is a new host record. Acarina: Laelapidae Androlaelaps fahrenholzi (Berlese) We collected 8 specimens of this mite from S. b. endemicus at the following localities: Sucker Cr 11 mi [18 km] N Emmett, Gem Co., T8N, R2W, Sec. 13 [44°02'N, 116°31'W, 830 m elev.], 21 February 1982 (ACMNH 227 [4 f, 2 deutonymphs]); lower Mann Creek, 2.5 mi [4 km] N jet. Weiser River Road, Washington Go. [44°16'N, 116°5rW, 720 m elev.], 14 March 1982 (ACMNH 233 [2 f]). This mite is widespread in Eurasia, North America (Whitaker 1979), and Central America (Strandtmann 1949). It occurs on a wide vari- ety of mammals, including marsupials {Didel- phis sp.), insectivores, bats, several families of rodents, lagomorphs, carni\'ores, and birds (Strandtmann 1949, Whitaker and Wilson 1974, Raybum et al. 1975). Opossums, insectivores, and rodents are the primaiy hosts, but A. fahrenholzi has the least host specificity and widest geographic range of any North Ameri- can ectoparasitic mite (Whitaker 1979). These are the 1st records from S. hrimneus. Nematoda: Rhabditidae Rliahditis {Pelodera) orhitalis Sudhaus and Schulte We obsen^ed this parasitic eyeworm only in live-trapped S. h. hrunneiis from OX Ranch 1-2 km S, 1-2 km E Bear, Adams Co. [45°00'N, 116°39'W, 1340 m elev.]. All specimens were collected in April and May 1990 to 1994. We found them in 1 eye or both eyes of yearling and adult S. h. bninneiis. The number per eye varied from 0 to 1272. The museum specimens were not checked for eyeworms. In 1994, T A. Gavin and E W Sher- man examined 21 live-trapped S. h. endemicus from Sand Hollow, Payette Co., and found no eyeworms. This eyeworm has been reported previously from Eurasian and North American voles and lemmings {Microtus spp., Lemmus trimucroiui- tus, Dicrostomjx groenlandicus, Pitinn/s suhter- raneus, Arvicola terrestris, and Clelhrio)U)nu/s spp.), mice {Apodemus spp. and Mus muscii- his), and rats {Rattus norregicus; Poinar 1965, Kinsella 1967, Cliff et al. 1978, Hominick and Aston 1981, Schulte 1989). S. h. hrunneus is a new host record, the 1st record of any RJiabdi- tis from Sciuridae, and also the 1st record of R. orbitaUs from Idaho. Epizootics In 11 field seasons (April-June) of work with S. b. hrunneus, we found only 2 dead indi- viduals, and none were obsei^ved sick or dying. While a number of populations have declined (T. A. Gavin, P W Shemian, and E. Yensen per- sonal obsei"vation), mortality occuiTcd while die animals were in hibernation rather than during the active season. The most serious population declines were estimated to be around 50% in 1 yr, rather than the 95%-100% active season mortality typically associated with plague (Lechleitner et al. 1968, Payor 1985). Although numbers of fleas on indixidual squirrels were relatively low, especially in S. b. hrunneus, all flea species we collected are important in plague epidemiolog)' in other hosts (Pratt and Stark 1973) and could potentially play a role in an Idaho epizootic. Discussion Collections of ectoparasites from S. hrun- neus have resulted in new state records for the flea TJirassis francisi rockwoodi and the eye- worm Rhahdiiis orhitaUs, plus 9 new host records. Because there have been no previous studies of S. hrunneus, the new records are hardly surprising. However, the records of Thrassis f. francisi and T. f rockwoodi on S. h. endemicus were unexpected. Thrassis f har- nesi occurs north of the Snake River in the Snake Ri\'er Plain (Stark 1970) and is the sub- species of Tlvassis francisi that would be ex- pected to occur in the range of S. b. endemicus. Instead, we found TJirassis f francisi. which is common in S. toicnsendii mollis south of the Snake River, and T f rockwoodi, for which the nearest locality' is from Oregon across die Snake River (Stark 1970), a major biogeographic bar- rier in southern Idaho (Da\ is 1939). This inter- esting situation merits fmtlier stud\'. With the exception of eyeworms, ectopara- sites of S. hrunneus are all known from multi- ple other species of ground stjuirrels (Table 1). Thus it is curious that S. h. hrunneus and S'. I). 1996] SrEHMormus bhuxneus Ectopaiusites 243 endemicus shared only a single ectoparasite, OropsijUci t. tiiberadata, a widespread Ilea tonnd on at least 4 other species of ground s(|uirrels. By contrast, die geographicalK and taxononii- cally close (Nadler et al. 1984) S. tuwii.sendii has all but one of die ectoparasite species found on both S. ])riiii)U'ii.s subspecies. However, Sper- mophdiis townsciidii is now recognized (Ih)ff- niann et al. 1993) as a complex of 3 closely related sibling species with different kary- otypes, and it was not always clear to us from the literature (Table 1) which parasites were associated with which host. Consequently, we have treated S. towiisendii as a single entity herein. There are several possible explanations for the lack of shared ectoparasites between S. h. brunneiis and S. b. endemicus: (1) they are geo- graphically separated, and their ranges are inhabited b)' different ectoparasites; (2) they occur in different habitats and therefore have different ectoparasites; (3) pelage differences between them may be different "microhabitats" for ectoparasites; (4) possibly the formerly shared ectoparasites on one or the other sub- species have been lost via a founder event, due to population structure, or because of popula- tion bottlenecks; and (5) we did not adequately sample all ectoparasites on either subspecies. Among these hypotheses, (5) is the least inter- esting evolutionarily, and (4) is the most inter- esting. Most western ground squirrel species are allopatric or parapatric; thus, there is little pos- siliility of direct transmission of ectoparasites among them. Historically, the 2 subspecies of S. brunneiis were separated by 19 km, 250 m in elevation, and a habitat change from arid shiTib- steppe vegetation to montane meadows (Yensen 1991). At present, the nearest extant popula- tions are separated by 48 km. Because S. town- sendii is allopatiic to S. brunneiis, occurs in non- montane habitats, and has all ectoparasites found on both subspecies of S. brunneiis, differences in geography (hypothesis 1) and habitats (2) are unlikely to be the sole explanations for the dif- ferences in ectoparasites between S. I), brun- neiis and S. b. endemicus. There are significant differences in pelage length between S. b. brunneiis and S. b. endem- icus (Yensen 1991). Interestingly, the pelage of S. townsendii is intermediate in length between the 2 S. brunneiis subspecies (E. Yensen un- published data). There also appear to be differ- ences in hair density and diameter, although these were not quantified b>' Yensen (1991). Possibly S. townsendii is inhabitable by the entire set of ectoparasites, and each subspecies ol S. brunneiis is a suitable host for about half the set. Thus, pelage differences (hypothesis 3) aic a possible explanation for the lack of over- laj) in ectoparasite species between 2 veiy close relatives, but it would not explain the dif- ferences in parasite loads or the low percent- ages of nonparasitized individuals. Anderson and May (1979) argued that para- site infestations should be sensitive to host population structure (hypothesis 4). As popula- tion size declines and populations become more isolated, the probability of parasite species loss should increase. Our data were consistent with this pattern: the proportion of parasitized S. b. brunneiis was significantly lower than that of S. b. endemicus; the former has smaller, more isolated populations. The isolated S. b. brunneiis populations would also retard exchange of ectoparasites among populations. Thus, there might be sto- chastic losses of parasite populations with low probabilit)' of recolonization (Anderson and May 1979). The differences in incidence of parasites between S. b. brunneus and S. b. endemicus are consistent with this inteipretation. The low density and wide dispersion of in- dividuals within S. b. brunneus populations at a site (E. Yensen and E W. Sherman personal observation) may also retard direct transfer of ectoparasites, and, consequently, S. /;. brunneus populations may not be able to support large ectoparasite populations. The low incidence of parasitism in Idaho ground squirrels thus appears to be related to population stioicture. Because we did not examine ground squir- rels under a microscope, we do not suppose that all ectoparasites were collected (hypothe- sis 5). However, there was no systematic bias in the sampling that would account for the dif- ferences in the proportion of parasitized ani- mals and parasite load differences between S. b. brunneus and S. b. endemicus. The low propor- tion of parasitized S. /;. brunneus (28%) and S. b. endemicus (70%) in this study may have been partially because our collecting techniques missed smaller ectoparasites. However, the same techniciues were used for both subspecies; therefore, the sampling differences between them should reflect real differences in parasite load. Thus, with the number of animals and 244 Great Basin Naturalist [Volume 56 localities sampled, the low overlap in lists of parasites is striking. Further, the low proportion of S. briinneus with ectoparasites (55%), especially in S. b. brunneus, is atypical of Spennophilus. For ex- ample, Hilton and Mahrt (1971) found that in Alberta 100% of S. cohitnbiamis and S. frank- linii and 92% of S. richardsonii had ectopara- sites. We were collecting S. townsendii and S. columbianus at the same time as S. brunneus and were impressed by the much higher para- site loads on those species. Although we did not obsei^ve plague in S. brunneus during this study, it does occur in southwestern Idaho. Serum samples positive for Yersina pestis, the plague bacterium, were reported from S. townsendii during a major ground squirrel die-off in 1941-42 in Ada, Canyon, and Payette counties, immediately south of the range of S. b. brunneus (Hubbard 1947, Link 1955). In 1975-1977, positive anti- body titers to plague were found in 72%-91% of badgers in the Snake River Birds of Prey Area, 50 km south of the range of S. b. endemi- cus (Messick et al. 1983). Badgers are impor- tant predators of ground squirrels. Eight of 9 dead Townsend s ground squirrels examined by Messick et al. (1983) were positive for Y. pestis. The plague bacterium has been detected in other species of Spenno))hilus in all 5 Idaho counties where S. brunneus populations exist, but until 1995 no S. brunneus had been exam- ined (Idaho Department of Health and Welfare personal obsenation). In April 1995, T. A. Gavin found a dead S. b. brunneus at the OX Ranch and sent it to the Wyoming State Veterinaiy Laboratoiy (Laramie) where it was assigned case #95W3914. The carcass was found to be nega- tive for Y. pestis (E. Williams persontil comment). Nonetheless, in tlie event of a plague epizootic, local populations of S. brunneus could easily be decimated. With only a small number of popu- lations remaining, plague could jeopardize the sui'vival of both subspecies of S. brunneus. Note Added in Press Six hibemacula of S. b. brunneus were exca- vated in spring 1995 (Yensen and Sherman unpublished data). Nests recovered from the hibernacula were placed in plastic bags in the field, taken to the laboratory, and then placed in Berlese funnels; small invertebrates were collected in 70% ethanol. Onl>' the fleas have been identified to date, but we can now add the following records: Neopsylla inopina Adams Co., 1.5 km N, 1.5 km E Bear Guard Station, 28 April 1995 [6 m, 7 f]; Adams Co., Steve's Creek, 2 km S, 2 km E Bear, 15 April 1995 [8 m, 7 f]; Adams Co., mouth of Cold Springs Creek, 14 May 1995 [1 m, 1 f]. Oropsijlla idahoensis Adams Co., 1.5 km N, 1.5 km E Bear Guard Station, 28 April 1995 [1 m, 2 f]; Adams Co., Steve's Creek, 2 km S, 2 km E Bear, 15 April 1995 [4 m, 2 f]; Adams Co., 3 km S Bear, 16 April 1995 [1 f]. Oropsijlla tuberculata tuberculata Adams Co., 1.5 km N, 1.5 km E Bear Guard Station, 28 April 1995 [18 m, 16 f]; Adams Co., Steve's Creek, 2 km S, 2 km E Bear, 15 April 1995 [20 m, 21 f]; Adams Co., moutli of Cold Springs Creek, 14 May 1995 [3 f]. Thrassis pandorae pandorae Adams Co., 1.5 km N, 1.5 km E Bear Guard Station, 28 April 1995 [28 m, 31 f]; Adams Co., Steve's Creek, 2 km S, 2 km E Bear, 15 April 1995 [8 m, 15 f]; Adams Co., 3 km S Bear, 16 April 1995 [1 m]. Cat(dlagia sp., prob. descipiens Adams Co., 1.5 km N, 1.5 km E Bear Guard Station, 28 April 1995 [1 f]. Foxella ignota Adams Co., Steve's Creek, 2 km S, 2 km E Bear, 15 April 1995 [4 m, 3 f]. Spennopliilus b. brunneus is a new host record for Catallagia sp. and Foxella ignota. Catallagia deeipiens is widely distributed in the western United States and is usualK found on deer mice (Baird and Saunders 1992). Foxella ignota is commonly foimd on pocket gophers in the northern Rock^' Mountains (Hubbard 1947). These new records also indicate that differ- ent sets of ectoparasites occur on S. b. brun- neus and S. /;. endeniicus, thus corroborating the earlier results. The same 4 flea species were again found associated with S. b. brun- neus, and neither Catallagia nor Foxella is kiiown from S. b. endemicus. 1996] SPERMOrillLUS BHl'S'NEUS ECTOPAIUSITES 245 Acknowledgments We thank Elizabeth J. Dyni, Elizabeth Domingiie, Thomas A. Cia\in, D. Brad llam- mond, Da\id O Neill, Daniel A. Stevens, and William E Lanrance for assistanee in the field. Robert E. Lewis (Iowa State Universit)), Richai'd B. Eads and Eduardo Campos (Centers for Disease Control), and Elizabeth Doniinyue (Cornell Universit) ) kindly identified the fleas; Nixon Wilson (University of Northern Iowa), J. E. Keirans (Rock>' Mountain Laboratory), and JoAnn Tenorio (Bishop Museum) identified the ticks and mites; K. C. Emerson (National Mu- seum of Natiual History) identified the lice; Susan E. Wade (Cornell University) identified the nematode eyeworms; and Amy Doerger- Fields (University of Wyoming) tested the spec- imen for plague. Financial support was pro- vided by the National Science Foundation (DEB-9225081), National Geographic Society (grant #3485-86 to P W Sherman and E. Yensen), George C. ("Tim ) Hixon, and Univer- sity of Idaho Agricultural Experiment Station. John and Jeanne Dyer and Tim Hixon pro- vided encouragement, housing, and access to research sites. We thank Sherilyn Robison and Rita Colwell for helpful discussions, and William H. Clark, Eric Eldredge, James Munger, John O. Whitaker, Jr., and an anonymous referee for constructive comments on an earlier version of the manuscript. Literature Cited Anderson, R. M., and R. M. May. 1979. Population biol- ogy of infectious diseases: part I. Nature 280: 361-367. Baird, C. R., and R. C. Saunders. 1992. An annotated checklist of the fleas of Idaho. University of Idaho, College of Agriculture, Research Bulletin 148. 34 pp. Burgess, G. D. 195.5. Arthropod ectoparasites of Richard- son's ground squirrel. Joiunal of Parasitology 35: 325-352. Cliff, G. M., R. C. Anderson, and E E Mallory. 1978. Dauerlai-vae of Pclodera stronglyoides (Schneider, 1860) (Nematoda: Rhabditidae) in the conjimctival sacs of lemmings. Canadian Journal of Zoology 56: 2117-2121. D.wis, W. B. 1939. The Recent mammals of Idaho. Ca\ton Printers, Ltd., Caldwell, ID. 400 pp. Doss, M. A., M. M. Farr, K. E Roach, and G. Anastos. 1974. Inde.x-catalogue of medical and veterinaiy zool- ogy. Special Publication No. 3. Ticks and tick-borne diseases. I. Genera and species of ticks, part 2. Gen- era H-N. U.S. Department of Agriculture, Agricul- tural Research Service, Washington, DC. 593 pp. Gavin, T. A., P VV. Sherman, E. Yensen, and B. May Pop- ulation stracture and gene flow among disjunct pop- ulations of Idaho ground squirrels {Spermophilus l)niniH'tis), with reference to otlu-r species of SjU'r- iiiDpliilits. Submitted. Cii.L, A. E., and E. Yensen. 1992. Biocliemical dilferenti- ation in the Idaho ground s(|uirrel, Siwrinophilus hninnciis (Rodentia: Sciuridae). (weat Basin Natural- ist ,52: 1.5.5-159. Hilton, D. E J., and J. L. Mahrt. 1971. Ectoparasites from three species of Spennopliilus (Rodentia: Sciuri- dae) in Alberta. C'anadian [ournal of ZoologN 49: 1,501-1504. Hoffmann, R. S., G. G. Anderson, R. W. Th(jrinc;ton, Jr., and L. R. Heaney 1993. Family Sciuridae. Pages 419-465 in D. E. Wilson and D. M. Reeder, editors, Mammal species of the world. 2nd edition. Smith- sonian Institution Press, Washington and London. 1206 pp. Holekamp, K. E. 1983. Pro.ximal mechanisms of natal dis- persal in Bclding's ground scjuirrel {Spenni)philu.s heldmgi hcldin' colonies have been found in cavities of both live and dead trees in California (Rainey et al. in press). Despite these records, a clear understanding of siKer-haired bat roosts and roost habitat is still lacking. To better understand the roost requirements of siKer-haired bats, we investigated roost selection b\ the silver-haired bat in the Black Hills of South Dakota. Although forests in this region have been intensu ely managed for tim- ber (Boldt and Van Deusen 1974), silver-haired bats are relatively abundant compared to the 9 other bat species present in the region (Matt- son 1994). Although Mattson (1994) captured twice as many males as females, pregnant or lactating females were not uncommon. Our goal was to characterize roost selection by silver- haired bats in terms of attributes potentially affected by cunent forestry practices. Study Area Our study area is located in the southern Black Hills of South Dakota near the town of Custer (43°46'N, 103°35'W). Most of the study area is in the Black Hills National Forest and occurs at elevations from 1360 to 1985 m asl. The topograph)- of the area varies from rolling highlands with parklike valleys to narrow, steep canyons with rock\' ridge tops. The climate of the Black Hills differs from the surrounding semiarid plains in that it is moister and less subject to temperature extremes. Average maxi- mum temperature at Custer in JuK is about 23° C, while mean annual precipitation is 457 mm. The forests of the area are dominated by pure stands of ponderosa pine (Pimis ponderosa). Small stands of quaking aspen (Populus tremu- loides) precede ponderosa pine on disturbed sites. Paper birch (Betida papyhfera) grows in small clusters in more mesic sites, whereas 'Department of Zoology and Physiolog\, L'ni\ersit> of Wyoming, Laramie, V\T S2071-.3166. ^Present address: PIC Technologies, Inc., .309 South 4th Street, Suite 201, Laramie. \\T 82070. 241 248 Great Basin Naturalist [Volume 56 Rocky Mountain juniper {Junipenis scopiiloruin) grows on diy ridges. The forests of the Black Hills have been managed for timber production since logging first began in the 1870s. During the past 100 yr, most areas have been cut once, and many have experienced multiple partial cuts (Alexan- der 1987). In all, nearly 12 X 10^^ m'^ of timber has been removed. Only a few small scattered stands of unharvested forest remain (Boldt and Van Deusen 1974). Although clearcutting was once the primary means of harvest, shelter- wood cutting, a method using a series of cuts, is now standard. We delineated two 10.1 X 10.1-km study sites in areas in which we located silver-haired bat roosts. The Jewel Cave Study Site encom- passes Jewel Cave National Monument and adjacent areas of the Black Hills National For- est. The Hazelrodt Study Site is located south- east of Custer on national forest land and Custer State Park. Much of the Hazelrodt Study Site burned during a fire in 1990 that covered over 5670 ha. Materials and Methods Capture and Tracking Techniques Silver-haired bats were captured using mist nets set above small ponds and streams be- tween 25 June and 4 August 1994. We deter- mined the sex and reproductive condition for all captured bats using external features (Racey 1988). Bats were classified as adult or juvenile based on fusion of the epiphyseal-diaphyseal suture of the finger bones (Anthony 1988). We attached 0.7-g radio transmitters (model BD-2B, Holohil Systems Ltd., Woodlawn, Ontario) to 4 adult males and 12 adult females. After fur had been trimmed from the bats, transmitters were attached to the area between the shoulder blades using a cyanoaciy late -based glue (Fing'rs, Camarillo, CA). Bats to which transmitters were affixed weighed 11-14 g, so that transmitters represented 5-6.4% of body mass, slightly over the 5% maximum recom- mended by Aldridge and Brigham (1988). We did not use any other marking technique to identify individuals. Hand-held, 3-element yagi antennas and portable receivers (model TR-2, Telonics, Mesa, AZ) were used to track bats to roost trees. If we were unable to determine where in the tree the bat was roosting, or whether it was alone or with others, we returned to the tree before dusk to watch and count bats leaving the site. We attempted to approach the tree quietly to reduce disturbance. We used a bat detector (Bat Box III, Stag Electronics, St. Agnes, Eng- land) to listen for echolocation calls. These, along with body size and flight pattern, were used to confirm that bats in a given roost were only silver-haired bats. Roost Measurements We located 18 roost trees in the Jewel Cave Study Site and 21 in the Hazelrodt Study Site. When possible, the type of roost (i.e., wood- pecker cavity, crevice, loose bark, etc.) was recorded. Each roost tree was classified as being used by either a maternity aggregation or solitary bats. Maternity roosts, located by tracking pregnant and lactating females, always contained 6 or more bats. Solitan' roosts con- tained only a single bat and were located by tracking males or females that did not appear pregnant or lactating or were post-lactating. We categorized the aspect of the roost exit as northeast (0-89°), southeast (90-179°), south- west (180-269°), or northwest (270-359°). Each roost tree was identified to species and its height and diameter at breast height (dbh) measured. We placed each roost tree into 1 of 7 decay stages; decay stage 1 included live trees with intact bark and branches, whereas decay stage 7 included dead trees beginning to decompose with broken tops and no loose bark (Thomas et al. 1979). Plot Measurements Within a 5-m-radius (78-m-) circular plot centered at each roost tree, we measured aver- age tree size, total basal area, and snag densit)'. Trees were defined as standing woody stems >1.5 m in height and >10 cm dbh. We also recorded whether disturbance b\' fire or log- ging had taken place in each plot. Disturbance by fire was considered to have occurred if there was any charred woody material in the plot, and disturbance by logging was noted if we obserxed any saw cuts on wood\' material in the plot. To compare characteristics of roost site plots with Hie sunounding areas, we located four 5- m-radius neighborhood plots for each roost plot and recorded the same information as for roost plots. We located the center of the neigh- borhood plots b\- pacing 100 m from the roost 1996] Sil\eh-iiaihl;d Bat K(x:)sts 249 tree in each of the cardinal directions (north, south, east, west) and then pacing an adchtional 30 ni in a randomly selected direction. We measured elexation and distance to the nearest source of Water loi' each loost tiee using topographic maps (7.5 minute series, US(iS, Denver, CO). For comparison, we randoniK located a point in the Jewel C^ave StucK Site or Hazelrodt Study Site lor each roost tree found in that site. To examine roost site selection on a larger scale, we calculated the number of snags in all neighborhood plots to estimate snag den- sity for the stud)' site generalK. This estimation was made by dividing the total ninnber of snags in the 156 neighborhood plots by their total area. The fire in the Hazelrodt Study Site in- flated snag densities in this area. To remove the influence of fire, we calculated snag densities witliin the study sites by removing the 77 neigh- borhood plots that had been disturbed b\' fire. Analysis Chi-square tests for goodness-of-fit (Jelinski 1991) were used to compare obsei"ved with ex- pected roost aspects and tree decay stages by roost t)^pe (maternity vs. solitaiy). For the latter test, because of small sample size, we pooled the roost trees into 3 decay stage categories: stage 1-3, stage 4, and stage 5-7. To compare continuous attributes between roost plots and neighborhood plots, we sub- tracted attribute means for the 4 neighborhood plots from corresponding means for the roost plots. So, each roost plot was compared only to its 4 neighborhood plots. We tested the null hypotheses that the mean differences did not differ from 0 using paired t tests. Chi-square tests for homogeneity (Jelinski 1991) were used to compare obsei-ved with expected distiu-bances at roost plots. Expected disturbances were based on the proportion of neighborhood plots that had burned or been logged. We used 2-sample t tests to compare the means for elevation and distance to nearest water for roost sites and random sites. To avoid type I errors that may result fi-om using a number of inferential statis- tical tests with the same predictor variable, we arbitrarily set oc = 0.025. Results Roost Attributes We radio-tracked 16 bats for a mean of 8 d (range: 1-20) and located 39 roosts, all of which occurred in trees. Nine adult females were tracked to 10 trees that were used by maternity aggregations averaging 22.2 ± 4.9 (.v^) indi\iduals (range: 6-55). Three other females and 4 adult males were tracked to 25 roost tiees, none of which were used by mater- nity aggregations. Three of the females that originally used maternity aggregations were lal(>r followed to 4 trees where they roosted alone. Maternity roosts were found exclusively in tree cavities, pinmarily those created l)y wood- peckers (Picidae). Cavity opc-nings were 7.5-10 cm in diameter. Solitary bats roosted under loose bark {n = 15), in a tree crack or crevice {n = 5), or in a woodpecker cavity (n = 1). We could not determine the specific roost location for 8 trees. These trees were placed in the soli- tary category because bats tracked to these 8 trees were always observed roosting alone at other trees. Maternity roosts were 10.2 ± 1.5 m (range: 3.1-13.8) aboveground. The height of measured solitaiy roosts averaged 3.4 ± 0.5 m (range: 0.9-8.9). Cavity openings of maternit>' roosts and solitaiy bat roosts were found more frequently on the south side of tree boles over other aspects (x^ = 15.8, d.f = 3, P = 0.001). Of 39 roost trees, 38 (97%) were ponderosa pine and 1 (3%) was aspen. Of 508 trees on neighborhood plots, 483 (95%) were ponderosa pine and 25 (5%) were other species: aspen, juniper, and paper birch. The 10 trees used by maternity aggregations of silver-haired bats ranged from decay stage 2 to 7 (median = 5). The 29 trees used by solitar)' bats varied from tree decay stage 3 to 7 (median = 4). Trees in neighborhood plots ranged from decay stage 1 to 7 (median = 1). Bats in maternity aggrega- tions selected roost trees in significantly differ- ent decay stages than solitaiy roosting bats (x^ = 10.2, d.f = 2, P = 0.0062; Fig. 1). Roost trees averaged 14.2 ± 0.9 m (range: 3.7-24.1) in total height, and 39 ± 2 cm dbh (range: 13-63). They averaged 17 ± 2 cm larger in dbh than neigh- borhood trees. The 10 maternity roost trees averaged 44 ± 4 cm dbh (range: 29-62), 24 ± 4 cm larger than neighborhood trees. The 29 soli- taiy roost trees averaged 37 ± 2 cm dbh (range: 12-55), 15 ± 3 cm larger than neighborhood trees. Matemitv and solitaiT roost trees did not differ in diameter {t = 1.64,'? = 0.12). The 9 bats found in maternity aggregations returned to the same roost tree for a mean of 8 d (range: 1-21). We tracked 1 bat fi-om a tree containing a maternity aggregation of 55 bats 250 Great Basin Naturalist [Volume 56 1-3 4 5-7 Tree Decay Stage Categories Available (n = 132) \^7/\ Aggregation (n = 10) ^^ Solitary (n=29) Fig. 1. Percentages of trees in each tree decay stage categoi^ used by maternity aggregations and solitaiy roosting silver- haired bats, and available trees in tlie Black Hills, South Dakota, June-August 1994. to a 2nd tree with a maternity aggregation of 44 bats about 440 m away. The following eve- ning no bats were observed exiting from the 1st roost tree, but it is not clear how many bats from the 1st roost tree moved to the 2nd tree with the bat we were tracking. We tracked 10 bats that used solitaiy roosts to a mean of 3 solitaiy roost trees (range: 1-6). For the most part, these bats switched trees daily. However, on 5 occasions solitary bats used the same tree on consecutive days. Three of the 7 solitaiy roosting bats that we followed to multiple trees returned at least once to trees they had used several days before. Solitary roosting bats traveled a mean of 405 ±93.7 m (n = 13) between successive roost trees. Radio- tracked bats traveled a mean of 2060 ± 440 m (n = 12) from the captme point to their first roost tree, significantly farther (/ = 3.67, P = 0.004) than the distance between successive roost trees. Plot Attributes Roost plots had 1.7 ± 0.6 more live trees {t = 3.09, ? = 0.004) than neighborhood plots. Live and dead trees on roost plots were 6.5 ± 1.7 cm larger in dbh on average than those on neighborhood plots {t = 3.77, P = 0.0006). Roost plots also had basal areas of both lixe and dead standing trees that were 14.07 ± 3.46 cm^/m^ greater {t = 4.06, P = 0.0002) than neighborhood plots. Neither fire disturbance ()^2 = 0 005, d.f = 1, P = 0.94) nor logging disturbance (x^ = 2.72, d.f = 1, P = 0.099) dif- fered between roost and neighborhood plots. Maternity and solitan' plots did not differ in the attributes studied (Table 1). Roost trees tended to be located higher in ele\ation than random points (/ = 1.67, P = 0.10). Roost sites were significantlv farther from water than random points {t = 2.78, P = 0.007). Using all 156 neighborhood plots, we calcu- lated snag densit)' for the area to be 117 snags/ ha. After removing 77 neighborhood plots that were disturbed by fire, we recalculated snag densities to be 21 snags/ha. Discussion Roosts used by maternity aggregations dif- fered from those used b\' solitaiA sil\ er-haired 1996J SlLVER-llAlHED BaI' RoOS TS 251 Table 1. Conipadson l)(>t\vfeii solitan aTuI iiiatfniit\ roosi plot attributes in tlit- Black Mills, South Dakota, Jiiiu'- Aiigust 1994. Attribute Solitary (;i = 29) Maternitv in = 10) r \ali Li\ e trees (no./plot) Snags (no./plot) Mean tree dbh (cm) Total basal area (cni^/m-) 4.5 ± 0.6 5.2 ± 0.7 0.72 0.47 2.1 ±0.5 2.2 ± 0.5 0.15 0.88 26.7 ± 3.2 27.5 ± 1.9 0.22 0.83 17.8 ±1.3 25.3 ± 4.7 1.54 0.13 bats. Mateniih' aggregations always used a hol- low ca\ity within a tree bole. Usually these cavities were created by woodpeckers, likely haiiy woodpeckers {Picoides villosus) or black- backed woodpeckers {P. arcticus), based on the size of the openings (Terres 1980). Although rare in the Black Hills (Black Hills National Forest 1989), Lewis' woodpeckers {Melanerpes lewis), northern flickers {Colaptes aiiratiis), or three-toed woodpeckers {Picoides tridactyhis) may have excavated some of the cavities. Soli- taiy roosts were located under loose bark or in a natural crack or crevice in the tree bole. Only once did a solitary bat use a woodpecker cavity. Although silver-haired bats are cryptically col- ored, they were never observed roosting openly on a tree trunk or limb, or in foliage. This be- havior differs from other cryptically colored, tree-roosting bats (e.g., Lasiiirns spp.), which tend to roost among tree foliage (Shump and Shump 1982a, 1982b). Roosts required by ma- ternity aggregations may limit silver-haired bat abundance; clearly trees with cavities are less available than are those without. Reproductive females seem to require roosts that provide a relatively enclosed and unexposed space for protecting young from predators or maintain- ing the necessary thermal environment. Cavity openings of maternity roosts and solitary bat roosts occurred more frequently than expected on the south side of tree boles. We hypothesize that these roosts are warmer than sites facing north because of insolation and that these differences result in energetic savings, providing more energy for growth and development (McNab 1982). Reller (1972) has shown that several species of woodpeckers ori- ent their nest cavity openings southwesterly for warming by the sun and/or ventilation by the wind. However, it is unclear whether bat use of cavities with south-facing entrances reflects the selections of bats or woodpeckers. SiKer-haired bats roosted exclusively in trees during the summer Although all but one of the roosts were located in ponderosa pine trees, the dominance of ponderosa pine in our study area prevented us from testing for tree species preference. The wide geographic distribution of silver-haired bats relative to that of pon- derosa pine and the use by silver-haired bats of both coniferous and deciduous roost trees in other parts of their range (Novakowski 1956, Parsons et al. 1986, Barclay et al. 1988, Camp- bell et al. in press, Rainey et al. in press) sug- gest that these bats select for the structure of the roost itself rather than for a particular tree species. As for other tree-roosting bats (Tide- mann and Flavel 1987), it is imlikely that tree species is important to silver-haired bats ex- cept that at the local level 1 species ma\ tend to have preferred attributes. Roost trees were standing, dead, and larger than average in diameter. The single living tree selected as a roost was dying (stage 2) and missing its top; it also had many dead limbs and several woodpecker holes high in the bole. There was an obsei-ved difference in tree decay stage between roost trees used b)' maternity aggregations and solitary bats. Solitaiy roosting bats frequently used trees in decay stage 4, which are characterized by the presence of loose bark. Alternatively, mateniit>' roosts tended to be found in older, more decomposed trees (decay stages 5-7), trees that are more com- monly used by excavating woodpeckers (Thomas et al. 1979). Although the importance of snags as roost sites in other forest types remains in question, large snags appear to be important resources for silver-haired bats in ponderosa pine forests. 252 Great Basin Naturalist [Volume 56 Clearly, solitan' roosting silver-haired bats switch roosts regularly. This lack of fidelity may be related to the abundant nature of potential roosts (Brigham 1991) or a predator-avoidance strategy (Kunz 1982b). Because they will return to roost trees used several days previously and these roosts are often close together, solitary bats may use a series of trees in the same area and thus maintain a level of site familiarity. Conversely, maternity aggregations tend to re- main in the same roosts for longer periods. This may be related to the less abundant nature of tree cavities and the importance of retaining roosts that are suitable for raising offspring. At least some of the mateniit\' aggregations appear to swatch roosts during the reproductix e period. The reason for this is not clear, although it may involve predator or ectoparasite avoidance (Lewis 1995). We expected bats to select roosts relatively close to water bodies, minimizing energetic costs of moving between roosting areas and areas potentially used for drinking and forag- ing. Although trees were al^undant in the study sites, bats traveled an average of >2 km fi"om point of capture to their 1st roost tree, and sig- nificantly farther from water than expected randomly. This seems to support other avail- a])le evidence for insectivorous bats in that roost site location is not strongly influenced by commuting costs (Fenton et al. 1985, Brigham 1991). Roost sites located farther from water tlian random points appear puzzling but may rep- resent the large number of roost trees located along hill or ridge tops, sites with potentially higher snag densities. Silver-haired bat roost trees were foimd at sites that differed fi^om nearby areas in a num- ber of attributes. Roost plots differed in having more, large trees and hence a higher total basal area than suiTounding plots. Roost trees located in areas that are ideal for tree growth or are logged infi-ecjuently might explain why the roost plots have more, larger trees. Undoubtedly, snags are important in pro- viding roost sites for silver-haired bats in the Black Hills. As suitable roosts are critical resources for bat sui-vival (Kimz 1982b), snag availability likely influences the distribution and abimdance of this species. Forest stands containing silver-haired bat roosts had snag densities ol 21 snags/ha, a value much higher than current management objectives. These densities were even higher in the Hazelrodt Study Site, an area with a large number of fire- killed trees. How fire suppression and logging practices have affected the number of snags in the Black Hills remains unclear; however, early photographs suggest that many forested areas were more open with many standing dead trees (Knight 1994). Because snags are used for nests or roosts by a large number of vertebrate species (Thomas et al. 1979), reduced snag densities may increase interspecific competi- tion. We hypothesize that forest management practices that reduce snag densities will lead to declines in local silver-haired bat populations. Acknowledgments Funding was proN'ided b>' the University of W\'oming-National Park Senice Research Cen- ter; the National Park Ser\'ice, Rocky Moun- tain Region; and the National Biological Sur- vey, Midcontinent Ecological Research Center. We especially appreciate support from Kate Cannon and the staff of Jewel Cave National Monimient, from Mike Bogan of the National Biological Sui^vey, and from Joel Tigner, Oscar Maitinez, and Alice Lippacher of the Black Hills National Forest. Mike Bjelland and Jay Grant provided valuable field assistance. We thank Thomas H. Kmiz and R. \hirk Brigham for com- menting on an earlier draft of the manuscript. Literature Cited Aldrioge, H. D. J. N., and R. M. Bkicham. 1988. Load earning and nianeu\erabilit> in an insectivorous hat; a test of tlie .5% "rule of radio-telemetn; Jonrnal of MammalogN' 69: 379-;382. Alexander, R. R. 1987. SiKieultural systems, cutting methods, and cultmal practices for Black Hills pon- derosa pine. US DA Forest Senice, General Technical Report RM-139. Rock\ Mountain Forest and Range Experiment Station, Fort Collins, CO. 32 pp. Anthony, E. L. R 1988. Age determination in hats. Pages 47-.58 in T. H. Kunz, editor. Ecological and hehav- ioral methods for the study of bats. Smithsonian Institution Press, Washington, DC. Barbour, R. W., and W. H. Da\IS. 1969. Bats of .\nierica. University^ of Kentucky, Le.xington. 286 pp. Barclay, R. M. R., E A. Faure, and D. R. Farr. 1988. Roosting behavior and roost selection by migrating sil\ er-haired bats {Lasionycteri.s iioctivagans). Journal of Manmialogy 69: 821-825. Blvck Hills National Forest 1989. Black Hills National Forest: checklist of birds. USDA Forest Senice. Boldt, C. E., and J. L. Van Del sen. 1974. Silviculture of ponderosa pine in the Black Hills: the status of our knowledge. USDA Forest Service, Research Paper RM-124. Rocky Mountain Forest and Range Experi- ment Station, Fort Collins, CO. 1996] Silm:k-iiaired Bat Roosts 253 Bhic.ham, R. M. 1991. Flexibility in tbniging and roostinj^ hehavioin- b\- the I)ig brown bat (Ei)t('siciis fiisctis). Canadian Journal ofZoolou;\ 69: 117-121. Campbell, L. A., J. C. IIai,li-:t, and M. A. O'C^onneli,. In pre.ss. Consenation of bat.s in managed fore.sts: use ol loost.s by Lasionyctcris n()rtiva' hoar\' and silver- haired bats in Oregon. Murrelet 69: 21-24. Hacev, P a. 1988. Reproductive assessment in bats. Pages 31-45 in T. H. Kunz, editor. Ecological and behav- ioral methods for the study of bats. Smithsonian Institution l^ress, Washington, DC. Rainev, W. E., J. Sh'ehek, and R. M. Mii.i.ek. In press. Colonial maternity roosts of the silver-haired bat {Lasionycteris noctivagam) in northern California forests. American Midland Naturalist. RelleR, a. W. 1972. Aspects of behavioral ecology of lii'd- headed and lied-bellied Woodpeckers. American Midland Naturalist 88: 270-290. Shump, K. a., and a. U. Shumr 1982a. Lasiurits borealis. Mammalian Species 183. American Society of Mam- malogists, New York. 6 pp. . 19H2h. Las iunis cine re us. Mammalian Species 185. American Society' of Mammalogists, New York. 5 pp. Tehres, J. K. 1980. The Audubon Society: encyclopedia of North American birds. Wings Books, Avenel, NJ. 1109 pp. Thomas, D. W 1988. The distribution of bats in different ages of Douglas-fir forests. Journal of Wildlife Man- agement 52: 619-626. Thomas, J. W, R. G. Anderson, C. Maser, and E. L. Bull. 1979. Snags. Pages 60-77 in J. W. Thomas, editor. Wildlife habitats in managed forests; the Blue Moun- tains of Oregon and Washington. US DA Forest Ser- vice, Agricultural Handbook 553. TiDEMANN, C. R., AND S. C. Fl.\\el. 1987. Factors affect- ing the choice of diuiTial roost site by tree-hole bats (Microchiroptera) in south-eastern Australia. Aus- tralian Wildlife Research 14: 459-473. Received 13 December 1995 Accepted 1 7 May 1996 Great Basin Naturalist 56(3), © 1996, pp. 25-J-260 PERCEPTIONS OF UTAH ALFALFA GROWERS ABOUT WILDLIFE DAMAGE TO THEIR HAY CROPS: IMPLICATIONS FOR MANAGING WILDLIFE ON PRIVATE LAND Teny A. Messmer^ and Sue Schroeder- Abstract. — We conducted a survey of Utali alfalfa (Mcdicago sativa) growers in 1993 to identify- wildlife damage problems to hay crops. Such surveys can provide wildlife managers with important insights regarding landowners' wildlife damage management concerns and needs. Pocket gophers (Thomonnjs spp.) and mule deer [Odecoileus hcmkmus) were perceived by growers as causing the most damage. Respondents reported a total annual loss of $350,000 or $24.79/ha (2.8% of the total crop value) because of wildlife damage in alfalfa crops. Decreased hay quantity' was the most fiequently cited problem caused by wildlife. Compensation and incentive programs were preferred over assistance and information programs for managing wildlife damage in alfalfa crops. Key words: wildlife damage perceptions, alfalfa growers, wildlife damage management, wildlife manageinent. Alfalfa is an important livestock forage. In 1994 over 58 million tons of alfalfa ha\' were harvested in the U.S. on 9,802,400 ha of pri- vately owned land. This represents over 40% of the hay hai^vested as livestock forage (National Agricultural Statistics Sei"vice 1995). Alfalfa hay is the most important cash crop grown in Utah. In 1994 Utali farmers harvested 2,205,000 tons of alfalfa on 210,000 ha of pri- vately owned land. This crop was worth $158 million (Gneiting 1994). Rodents, lagomoiphs, ungulates, and water- fowl can impact alfalfa production (Piper 1909, Sauer 1978, Luce et al. 1981, Dunn et al. 1982, Packam and ConnolK' 1992, Austin and Urness 1993, Conover 1994). Big game grazing of alfalfa during the growing season creates conflicts be- tween growers and wildlife managers (Austin and Umess 1993). Conflicts also may arise between landown- ers and wildlife managers because of differing perceptions about the extent of wildlife damage in cultivated crops. Farmers ma\' feel that wild- life managers are unaware of the extent of crop losses caused by wildlife and hence are insen- sitive to their needs (Decker et al. 1984, Conover and Decker 1991). Crop owners con- cerns about wildlife damage strongly affect how the agricultiual conmnniity will respond to environmental issues and whether federal or state wildlife programs aimed at maintaining or improving wildlife habitat on private property will succeed (Conover 1994). There is consensus among professionals working for federal and state wildlife and agri- cultural agencies that wildlife damage reduces the profitability of U.S. agriculture (Conover and Decker 1991). Professionals agree tliat wild- life depredation has increased over time but disagree over the seriousness of the impact. Although the actual costs associated with wild- life depredation are difficult to estimate and can differ on each farm or ranch and crop t^pe (Tebaldi and Anderson 1982, Austin and Umess 1987a, 1987b, 1989, 1993, Lewis and O'Brien 1990), landowners have demonstrated an abil- ity' to accurately assess crop losses caused by wildlife (Decker et al. 1984, Conover 1994, Mch'or and Conover 1994a). Crop losses and potential future losses caused h\\ or related to, the presence of wildlife must be assessed to determine if control is warranted (Rennison and Buckle 1988). Several Great Basin states including Utah, Wyoming, Colorado, New Mexico, Nevada, Idaho, and Arizona have enacted laws to com- pensate crop owTiers for wildlife -caused dam- age (Musgra\'e and Stein 1993). These actions have been initiated largely in response to con- stituent concerns oxer the economic impact of depredating wildlife, particularh' big game, in cultivated crops. ' l)(_ixntiiii-iil ol Hslu-i ii's ami WiUllilr, Utali StaU- Lni\ersit\. Lo.uan, UT S4322-.5210. -OepailiiKMil of Foicsl Resources, Utah State Llniveisit\', Logan, UT 84322-.5215. 254 1996] VViLDLiFK Damage to Alfalfa 255 Crop owTiers in lltali ma\' destroy depredat- ing big game animals if the animals are not removed by the Utah Division of WildHfe Resources (UDWR) within 72 h of notification (Chapter 183, Utah Code 1993a). Utah crop owners also may receive monetary compensa- tion for damage caused by big game animals (Chapter 307, Utah Code 1994b) and ring- necked pheasants {Pluisianiis colcliiciis: Chap- ter 46, Utali Code 1971). We surveyed Utah alfalfa growers to deter- mine their perceptions regarding wildlife dam- age to hay crops. Such surveys can provide wildlife managers with important information regarding landowner wildlife damage manage- ment needs and concerns (Conover 1994). Methods We sunexed 334 alfalfa growers (4% of all alfalfa growers in Utah) whose names were on the Utah Department of Agriculture's (UDA) 1993 Hay List. The UDA maintains this list to provide information to individuals who contact the department about purchasing alfalfa hay in Utah. The UDA updates this list each Januaiy We included a 2-page wildlife damage sur- vey in a UDA mailing sent to the growers. In addition to the survey, growers received a cover letter, the UDAs questionnaire, and a business reply envelope. The cover letter stated that if no response was received within 30 d, the grower's name would be removed fi"om the hay list. A follow-up letter was sent to nonrespon- dents 3 wk after the initial mailing. Those fail- ing to respond to the 2nd mailing were removed from the hay list. The sui-vey contained questions about the growers' experiences witli wildlife in their alfalfa crops. Growers were asked to identify wildlife species causing damage to hay crops, type of damage, their annual monetary loss from wild- life damage, specific damage control techniques employed on their fanii to control wildlife dam- age, whether they received any type of damage compensation or assistance, who they contacted for assistance and information, and what type of information and programs they found most useful in managing wildlife damage. Further, growers were asked to rate on a scale of 0 to 5 (0 = no cost through 5 = high cost) relative losses caused by different wildlife species to their alfalfa crops and the costs associated with common management practices used on their farms and ranches. Responses were stratified and analyzed bv the number of hectares in alfalfa (0— fO, 41-80, 81-200, 201-400, and >4()0) and type of oper- ation (inigated or diyland). Levere's tests were us(>d to determine ('(jualit}' of variances by types and sizes of alfalfa operation (SPSS 1995). We assumed that alfalfa growers on the hay list have the same values and perceptions as the population of Utah alfalfa growers, 'lb deter- mine if the hay list was statistically representa- tive of Utah alfalfa growers, we compared the mean alfalfa farm size and regional distribu- tions of farms on the hay list with acreage cate- gories reported by the UDA for all Utah alfalfa farms (Gneiting 1994) using a Kruskal-Wallis one-way analysis of variance. Differences in these tests were considered significant if P < 0.05. Results Alfalfa Production One hundred sixty-four completed ques- tionnaires (49.1%) were returned, of which 150 (91%) were useable for analysis. Sun'e)' respon- dents reported growing 16,867 ha of alfalfa, of which 14,391 ha (85%) was irrigated and 2486 ha (15%) was dryland alfalfa. Irrigated alfalfa farms ranged in size from 5 to 1062 ha. Dr>'- land alfalfa farms ranged in size from 3 to 320 ha. All farms were family owned and operated. Since the UDA hay list is relatively dynam- ic, it contains infoniiation regarding the grower's mailing address, telephone number, and inter- est in selling alfalfa hay, but not the size and type of operation. Information on alfalfa opera- tions was obtained through the survey; thus, we were unable to determine if there were any significant differences between respondents and nonrespondents. Although the responses received consti- tuted 2% of all Utah alfalfli growers {N = 7600), our sample was representative of the popula- tion based on mean farm size {H = 7.0; 7 df; P = 0.001) and regional distribution. Utah alfalfa acreage percentages reported by the UDA for northern, central, eastern, and southern regions were 30%, 31%, 19%, and 20%, respectively (Gneiting 1994). Regional alfalfa acreage per- centages for our sample were northern 27%, central 34%, eastern 21%, and southern 18%. 256 Great Basin Naturalist [Volume 56 Wildlife Species Present in Utah Alfalfa Fields Respondents reported 20 different species of wildlife were present in their alfalfa fields. Pocket gophers and mule deer were the most abundant, being reported present on 124 (82.7%) and 120 (80.0%) farms, respectively. Other wildlife species reported by farmers as common in alfalfa fields included jackrabbits {Lepus spp.; n = 89, 59.3%), ground squirrels {Spermophihis spp.; n = 83, 55.3%), prairie dogs {Cynomys spp.; n = 69, 46.0%), waterfowl {Anatidae\ n = 66, 44.0%), elk {Cervus elaphus; n = 62, 41.3%), pronghorn {Antilocopro ameri- cana; n = 54, 36.0%), and voles [Microtus spp.; n = 50, 33.3%). Wildlife species reported by farmers as being less common in alfalfa fields included marmots {Marmoto flaviventtis), bad- gers {Tax idea taxiis), red foxes {Vulpes vidpes), sandhill cranes {Gnis canadensis), Canada geese {Branta canadensis), cottontail rabbits {Syvda- giis spp.), deer mice {Pewmyscus manicukitus), raccoons {Procyon lotor), ring-necked pheas- ants, and muski"ats {Ondatra zibethica). Monetary Losses Caused by Wildlife One hundred nine growers (72%) reported losing $350,000 (a^ = $3242, .s- = 526) be- cause of wildlife damage in their alfalfa fields. Monetaiy losses averaged $24.79/ha. The average dollar loss reported by respon- dents who grew only iiri gated alfalfa was $3016 {n = 86, Sy = 554). Respondents who grew both irrigated and dnland alfalfa reported an aver- age loss of $4388 {n = 21, .sy = 1525). Those who grew only dryland alfalfa reported an average foss of $3750 (n = 2, 5- = 250). The highest losses per/ha were reported by respondents who grew both irrigated and dr>^- land alfalfa ($42 ha). Respondents who grew only irrigated or dryland alfalfa reported losses per/ha of $19 and $28, respectiveK'. Growers with irrigated alfalfa farms >200 ha in size reported significantly higher mone- taiy losses than operations <200 ha in size (F = 15.5; 1,103 df; P < 0.001). Although the average monetary loss reported by larger alfalfa farms was $5078 (n = 50) compared to $1639 for smaller farms {n = 55), the average loss per/ha was higher on smaller ($37) tlian larger farms ($21; F = 24.9; 1,103 df; P < 0.001). Growers reported no significant difference in damage losses by size for irrigated/drvland alfalfa farms (F = 0.4; 1,26 df ■ P = 0.52). Respondents with alfalfa farms >80 ha re- ported that rodents (F = 7.9; 1,107 df; P = 0.006) and ungulates (F = 18.2; 1,107 df; P < 0.001) caused higher monetaiy losses when compared to smaller farms (<80 ha). No signif- icant diflferences in monetaiy losses due to water- fowl were detected bv alfalfa farm size (F = 0.006; 1,107 df;P = 0.940). Relative Costs of Wildlife Damage in Alfalfa Fields Respondents ranked on a scale of 0-5 (0 = no cost through 5 = high cost) the relative damage costs associated with common wildlife species reported in their alfalfa fields as fol- lows: mule deer (2.9), pocket gophers (2.4), elk (1.6), prairie dogs (1.4), ground squirrels (1.4), jackrabbits (1.3), waterfowl (1.0), pronghorn (0.7), and meadow voles (0.9). Respondents with irrigated alfalfa farms >200 ha reported that elk (F = 7.9; 1,56 df; P = .007) and prong- horn (F = 7.5; 1,48 df; P = .008) caused signif- icantly greater cost-related problems than on smaller farms (<200 ha). Respondents with diyland alfalfa farms >200 ha reported greater significant cost-related problems caused by jackrabbits (F = 14.1; 1,20 df' P = 0.001) and mule deer (F = 8.5; 1,28 df; P = 0.007) than on smaller farms (<200 ha). Sun'ey respondents indicated that alfalfa production problems dif- fered b)' specific wildlife species (Talile 1). Farm and Ranch Management Practice Comparisons Respondents ranked on a scale of 0—5 (0 = no cost through 5 = high cost) the relatixe cost of the 7 farm management practices as follows: irrigation (3.8), fertilization (3.4), weed control (2.9), insect control (2.6), fencing (2.3), big game control (2.0), and rodent/rabbit control (1.9). Fertilization, weed control, and irrigation were used on 82%, 81%, and 80% of the farms, respectively. Big game and rodent/rabbit con- trol were used by 71% and 38% of the respon- dents, respectively. Respondents also reported employing several techni(|ues to control wildlife damage in alfalfa fields (Table 2). Based on sizes and types ol alfalfa operations, the only significant cost differences reported In' man- agement practices were for irrigation on farms >200 ha (F = 5.0; 1,124 df; P = 0.03). 1996] WiLDLiKK Damage to Alfalfa 257 Table L Percentage of all respondents {N = 150) reporting problems caused by a specific wildlife species in Utah alfalfa fields in 1993 and a breakdown of that percentage into subcategories based on the most severe tvpe of problem caused. cies i^ejiorting i\'rcent age identilying ; speci iic problem i-s most severe Wildlife spi" Hay Hay Equipmen Increased causing dan age problems (juality quantity damage costs Pocket gophers 68.7 14.0 20.7 26.0 8.0 Ground sciu nel 33.3 4.0 10.7 15.3 3.3 \bles 10.7 2.7 6.7 1.3 0.0 Jackrabbits 32.8 2.7 28.7 0.7 0.7 Prairie tlogs 23.3 0.7 8.0 13.3 1.3 Elk 20.0 6.0 12.7 1.3 0.0 Mule deer 64.0 8.7 54.0 1.3 0.0 Antelope 9.3 1.3 8.0 0.0 0.0 Waterfowl 17.3 2.7 14.7 0.0 0.0 Wildlife Damage Management Assistance Programs Fourteen respondents (9%) reported receiv- ing compensation for wildlife damage in their alialfa fields. Of these, 12 received compensa- tion for damage caused by mule deer. Another 48 (31%) indicated they received some type of technical assistance to control wildlife damage. Most of this assistance (75%) was provided to control damage caused by mule deer. One hundred twenty-two respondents (80%) reported seeking either information or assis- tance in dealing with wildlife depredation prob- lems. Conservation officers were cited by 53 growers (43%) as being their primary contact for infoniiation or assistance. Countv' agents and UDWR biologists ranked 2nd (22%) and 3rd (18%), respectively. Other sources of informa- tion in order of decreasing importance were other landowners (7%), farm and ranch stores (5%), and UDA agricultural representatives (3%). Respondents preferred compensation and incentive programs (42%) to other types of pro- grams to manage damage caused by wildlife in alfalfa fields. Research (17%), field demonstra- tions (13%), workshops (13%), facts sheets (13%), and videos (14%) were rated nearly equal in usefulness. Discussion Relationship of Perceived Damage Costs to Wildlife Management Surveys can be cost-effective means of assess- ing the magnitude and economic impact of wild- life depredation (Crabb et al. 1986). Unfortu- nately, due to the cost and time associated with conducting reliable surveys, many wildlife agencies are unable to perform this work on a regular basis. Our experience suggests that wildlife agencies should consider using state agriculture department hay lists to conduct benchmark sui^veys to identify wildlife damage management concerns and needs. Most states maintain hay lists (R. Parker, personal commu- nication, UDA, 1995). Our results summarize perceived losses. The relationship between perceived and actual losses is unclear and probably difficult to esti- mate (Conover 1994). This relationship depends in part on how conspicuous the damage appears and which wildlife species causes the damage (Wakeley and Mitchell 1981, Decker et al. 1984, Mclvor and Conover 1994b). Most respondents reported problems with pocket gophers and mule deer. Other species commonly causing problems included jackrab- bits, ground squirrels, prairie dogs, waterfowl, elk, pronghorn, and meadow voles. Conover (1994) also found that these species, in particu- lar deer, were perceived to cause most damage to agricultural crops in the U.S. Based on statewide averages, in 1993 Utah alfalfa growers harvested 10.5 tons/ha with a market value of $71.66 a ton. Survey respon- dents produced 177,104 tons of alfalfa on 16,867 ha having a total value of $12,691,000. The $350,000 loss reported due to wildlife rep- resents 2.8% of the crop value. Expanding this to the total value of alfalfa produced in Utah during 1993 results in a total perceived loss of $4.4 million. This is 9 times the amount the Utah State Legislature annually appropriates ($500,000) to reimljurse crop owner depreda- tion claims and expenses (Chapter 307, Utah Code 1994b). 258 Great Basin Naturalist [Volume 56 Table 2. Percentage of all respondents {N = 150) using a specific technique to control damage caused by wildlife species in Utah alfalfa fields in 1993 and a breakdowTi of that percentage into subcategories based on the most effective technique used. Percentage identif\ ing a specific techniiiue as being most efPectixe Using Wildlife species damage control Shooting/ Poison causing damage techniques (%) Trapping hunting baits Fumigants Cultural Fencing Hazing Pocket gopher Ground squinel Voles Jackiabbits Prairie dogs Elk Mule deer Antelope Waterfowl 41.7 6.7 0.0 33.0 2.0 0.0 0.0 0.0 45.4 4.7 17.3 22.0 0.0 0.7 0.0 0.7 13.3 2.0 2.7 7.3 0.0 1.3 0.0 0.0 39.3 0.0 36.0 2.7 0.0 U.O 0.7 0.0 24.0 2.0 12.7 7.3 1.3 0.0 0.7 0.0 21.3 0.0 12.0 0.0 0.7 0.0 7.3 1.3 46.7 0.0 22.7 0.0 0.7 0.0 16.0 7.3 9.7 0.0 4.0 0.0 0.7 0.0 --> - 1.3 16.7 0.0 13.3 0.7 0.7 0.0 0.0 2.0 Utah Code authoiizes the UDWR to imme- diately pay any approved damage claims < $500. Claims or total amounts of claims sub- mitted by a claimant in the fiscal year that are >$50() are not paid until the total amount of approved claims for the fiscal )'ear is deter- mined. If the amount claimed exceeds the appropriation, the per claimant amounts paid in excess of $500 are prorated. The current appropriation falls short of satisfying wildlife damage compensation claims and expenses (R. Valentine, personiil communication, UDWR, 1996). If 13^'^ of Utah alfalfa growers {n = 1000) submitted appro\ed claims of $500, their claims would deplete the annual appropriation. Although the alflilfa growers we suneyed pre- ferred compensation and incentive payments over other types of wildlife damage manage- ment programs, only 9% had e\er receixed an\' financial support. In the United States, 2.1 million farmers control 400 million ha of our 937 million ha land base. Their actions largely influence the qualit)' and quantit) of the existing wildlife habitat base (Cierard 1995). Landowners per- ceptions and concerns about wildlife damage are important because they influence their atti- tudes and behavior toward wildlife. Conover (1994) suggested that wildlife damage has reached levels that discourage prixate land- owners from managing for wildlife on their property'. Our results suggest that Utah alfalfa growers also perceive wildlife damage in alfalfa fields as a serious concern. Although wildlife professionals working for federal or state w ild- life and agricultural agencies believe that wild- life damage has increased in the last 30 \ r, our sun'ey results reinforce Conoxer and Decker's (1991) suggestion that programs necessar\' to adequately address crop owner conceiTis have not yet been implemented. Role of State Agencies in ResoKing Wildlife Damage Management Concerns State wildlife management agencies are responsible for managing damage caused by big game, upland game, and waterfowl (Mus- grave and Stein 1993). State agriculture de- partments administer and enforce pesticide control legislation that regulates the safe and proper use of pesticides for vertebrate pest damage. Because of tliis role, agiiculture depart- ments have jurisdiction oxer the control of unprotected wildlife species (xertebrate pests). In Utah these include pocket gophers, field mice, muskrats, ground squirrels, jackrabbits, raccoons, skunks, red fox, and coyotes. The UDWR recognizes that prixate lands xvithin Utah proxide habitat for xvildlife and that under some circumstances xxildlife may cause economic losses to the landoxxnier. With this understanding, the UDWR cooperates xxidi the UDA and the U.S. Department of Agricul- ture Animal Plant Health and Inspection Ser- vice/Animal Damage Control (ADC) program to conduct predator, bird, and rodent control actix ities and compensate landox\ners for cer- tain losses caused by xxildlife using funds appropriated by the legislatine. In 1994 the Utah legislature enacted an alteinatixe compensation program that alloxx's landoxxners to receix e pennits to han est antler- less animals as mitigation for damage caused 1996] WlLDLII i: llWIACli TO Al.KALFA 259 b\- big game (Cliapter 176, Utah Code 1994a). In 1995 the UDW'R Southern Region issued >12()() mitigation permits, of which 50% were filled, in 1996 both the number ol tags issued and number ol animals har\ ested declined as landowners lost interest in the program (N. McKee, personal connnunication, UIDWR, 1996). To better address landowners concerns gi\'en fiscal and legal constraints, we suggest that agencies and organizations responsible for managing wildlife resources and w ildlife dam- age on Utah agricultural lands collaborate to develop strategies that allow profitable agricul- ture and wildlife to coexist. Utah's posted hunt- ing unit (PHU; Chapter 288, Utah Code 1993b) and \\ ildlife habitat authorization (WHA) pro- grams (Chapter 75, Utah Code 1995) may offer additional mechanisms to achieve this goal. The Posted Hunting Unit Program The UDWR also recognizes that wildlife can be a significant benefit to the landowner The PHU program provides landowaiers with mon- etary incentives, through an allocation of hunt- ing pennits, to include wildlife (small game, waterfowl, and big game) in farm and ranch management plans. Landowners who partici- pate in the program are required to improve wildlife habitat but are ineligible to receive compensation for crop losses caused by wildlife. The most successfiil of UtiilVs PHU programs involves big game animals. In 1994, 47 big game PHU programs, encompassing over 400,000 ha of private land, proxided additional economic returns for hundreds of landowners and hunting experiences for thousands of hunters. Current program guidelines limit par- ticipation to landowners or landowner groups who own at least 4000 ha (Chapter 288, Utah Code 1993b). The size limitation was estab- lished to create more manageable herd units. In our sui-vey, respondents reported that big game animals caused the greatest damage. We suggest that big game PHU guidelines be modified to accommodate farm or ranch units <4000 ha in size. This modification would pro- N'ide the stimulus necessary to alleviate many crop owners' wildlife damage concerns and provide an additional incentive to include wild- life in farm and ranch management plans. In addition, we suggest that big game PHU oper- ators be encouraged to incoiporate provisions in their wildlife management plans to compen- sate smaller nonparticipating landowners adja- cent to their operation for crop damage- caused by big game animals. The Wildlife Habitat Authorization Program The WHA program rc(|uires persons 14 yr ol age or older to purchase a wildlife habitat authorization prior to purchasing certain hunt- ing or fishing licenses or permits. The funds generated from this authorization arc placed into a restricted account to be used lor wildlife habitat improvements. Several odier Croat Basin states operate similar programs designed to generate funds to do habitat work. We reconunend that state wildlife agencies consider using habitat funds to implement and evaluate enhancement projects and programs on public and private land that are designed specifically to reduce big game depredation on pri\'ate land. Habitat funds could be used to es- tablish big game lure crops, situate interceptor strips, or modify migration corridors as a means of abating localized depredation problems. Wildlife Damage Education Needs Crop owners also need additional informa- tion on techniques used to manage wildlife damage. Several respondents reported using fumigants and poison baits to control damage caused by ungulates, lagomorphs, and birds. These practices are illegal, as no products are currently registered in the U.S. to control dam- age caused by these species. We recommend that state wildlife agencies, agriculture departments, and federal ADC pro- grams cooperate in the development of public outreach, extension education, and research activities intended to inform crop owners about techniques that can l)e used to manage wildlife damage. These programs also should provide information on consei^vation technologies, non- lethal strategies, and opportunities that can be used to control wildlife damage and benefit wildlife resources while maintaining or enhanc- ing agricultural profitability. In conclusion, previous studies conducted in the Great Basin focused on evaluating the effects of big game depredation (Tebaldi and Anderson 1982, Austin and Urness 1987a, 19871), 1989, 1993) and sandhill cranes (Mclvor and Conover 1994b) on agricultmal production. Our study adds to this research b\' providing important insights regarding crop owners' 260 Great Basin Naturalist [Volume 56 perceptions about wildlife damage and their needs and preferences in managing damage. Our results suggest that Utah alfalfa grow- ers perceive wildlife damage as a serious con- cern. This concern should be shared by wild- life managers. In addition to informing landowners of their concern over wildlife damage, wildlife man- agers should demonstrate it by addressing potentials for increasing damage on private lands when developing wildlife habitat man- agement plans (Conover 1994). Wildlife man- agers also should incoiporate strategies in man- agement plans to benefit wildlife and reduce depredation potentials on private land. Acknowledgments We acknowledge R. Parker for assistance in distributing the sui"vey. We thank M. Conover, S. Barras, and A. Hall for reviewing earlier drafts of this manuscript. Literature Cited Austin, D. A., and R J. Urness. 1987a. Consumption of fresh alfalfa hay by mule deer and elk. Great Basin Naturalist 47: 100-102. . 1987h. Guidelines for evaluating annual crop losses due to depredating big game. Utah Division of Wildlife Resources, Publication 87-5. 42 pp. . 1989. Evaluating production losses from mule deer depredation in apple orchards. Wildlife Society Bulletin 17: 161-165. . 1993. Evaluating production losses from mule deer depredation in alfalfa fields. Wildlife Society Bulletin 21:397-401. Conover, M. R. 1994. Perceptions of grass-roots leaders of the agricultiual community about wildlife damage on their farms and ranches. Wildlife Societv Bulletin 22: 94-100. Conover, M. R., and D. J. Decker. 1991. Wildlife dam- age to crops: perceptions of agricultural and wildlife professionals in 1957 and 1987. Wildlife Society- Bul- letin 19: 46-52. Crabb, a. C, T. P Salmon, and R. E. Marsh. 1986. Sunex s as an approach to gathering animal damage informa- tion. Pages 2-4 in Vertebrate pest control and man- agement materials. American Society for Testing and Materials STP 974, Philadelphia, Pa! 12 pp. Decker, D. J., G. E Mattfieed, and T. L. Brown. 1984. Influence of deer damage on farmers' perceptions of deer population trends: important implications for managers. Proceedings of the First Eastern Wildlife Damage Control Conference 1: 191-195. Dunn, J. P, J. A. Chapman, and R. E. Marsh. 1982. Jack- rabbits. Pages 124-145 in Wild manuuals of North America: biology, management and economics. John Hopkins University Press, Baltimore, MD. 1147 pp. Ger\RD, R W. 1995. Agricultural practices, farm iDolic\, and the conservation ol biological (li\t'rsit\. Biological Science Report 4. U.S. Department of Interior, National Biological Sei-vice, Washington, DC 28 pp. Gneiting, D. J. 1994. Utah agricultural statistics. Utah Agricultural Statistics Service and the Utah Depart- ment of Agriculture, Salt Lake Cit\'. 138 pp. Lewis, S. R., and J. M. O'Brien. 1990. Survey of rodent and rabbit damage to alfalfa hay in Nevada. Pages 166-117 in Proceedings of the 14th Vertebrate Pest Conference. University of California, Davis. 320 pp. Luce, D. G., R. M. Case, and J. L. Stubbendieck. 1981. Damage to alfalf;i fields by plains pocket gophers. Journal of Wildlife Management 45: 258-260. McIvoR, D. E., AND M. R. Conover. 1994a. Perceptions of fanners and non-famiers toward management of prob- lem wildlife. Wildlife Societ>' Bulletin 22: 211-219. . 1994b. Impact of Greater Sandhill Cranes forag- ing on corn and barley crops. Agriculture, Ecosys- tems, and Environment 49: 233-237. MusGRAVE, R. S., A.ND M. A. Stein. 1993. State wildlife laws handbook. Center for W''ildlife Law, Institute of Public Law, University of New Mexico, Albuquerque. 840 pp. National Agricultural Statistics Service. 1995. 1994 crop statistics. U.S. Department of Agriculture, Wash- ington, DC. Pack-am, C. J., and G. Connolly. 1992. Control methods research priorities for animal damage control. Pages 12-16 in Proceedings of the 15th Vertebrate Pest Conference. University of California, Davis. 300 pp. Piper, S. E. 1909. The Nevada mouse plague of 1907-08. Farmers Bulletin 352: 1-23. Rennison, B. D., and a. P Buckle. 1988. Methods for estimating the losses caused in rice and other crops by rodents. Pages 69-80 in Rodent pest management. CRC Press Inc., Boca Raton, FL. 238 pp. Sauer, W. C. 1978. Control of the Oregon ground squiirel. Pages 99-109 in Proceedings of the 7th Vertebrate Pest Conference. University of California, Davis. 323 pp. Statistical Programs for Social Sciences. 1995. Micro- soft Windows Release 5.0. Microsoft Corporation, Redmond, WA. Teb.aldi, a., and C. C. Anderson. 1982. Effects of deer use on winter wheat and alfalfa production. WVoming Fish and Game Department. Job Final Report FW-3- R-26. 78 pp. Utah Code. 1971. Chapter 46. Section 23-17-5. Damages for destroyed crops — Limitations — Appeal. . 1993a. Chapter 183. Section 23-16-3. Damage to cultivated crops by big game animals — Notice to di\ ision — Crop owmer authorized to kill animals. . 1993b. Chapter 288. Section 23-23-1. Posted hunt- ing units. . 1994a. Chapter 176. Section 23-16-3.5. Damages to livestock forage, fences, or irrigation equipment on private land. . 1994b. Chapter 307. Section 23-16-4. Damages for destroyed crops — Limitations — .\ppraisal. . 1995. Cliapter 75. Section 23-19-42. Wildlife habi- tat autliorization. Wakelev, J. S., AND R. C. Mitchell. 1981. Blackbird dam- age to ripening field corn in PennsvKania. Wildlife Society Bulletin 9: 52-55. Received 11 Septonber 1995 Accepted 10 April 1996 Great Basin Naturalist 56(3), © I99(i, iiji, 261-266 SPATIAL RELATIONSHIPS AMONG YOUNG CERCOCARPUS LEDIFOLIUS (CURLLEAF MOUNTAIN MAHOGANY) Brad W. Scliultzl, Kohin J. Tausch^, ami Paul T lueller^ AusTKACT. — Tilis stiul\ anaK/.t'd spatial location patterns of Cercncarpiis ledifoliiis Nutt. (cnrllcai nionnlain niali()gan>) plants, classified as current-year seedling, estahlislied seedling, juvenile, and inmiatuie indi\ iduals, at a cen- tral Nevada stncK site. Most current-year seedlings were located in mahogany stands in wliicli large, niatiue individuals had the greatest ahundance. These stands had greater litter cover and a thicker layer of litter than areas with few cur- rent-\'ear seedlings. Most estahlislied young Cercocarpiis were located in adjacent Artemisia tridentata ssp. vaseyana (mountain big sagebrush) communities, or in infrequent canopy gaps between relatively few large, mature Cercocarpus. We discuss potential roles of plant litter, root growth characteristics, nurse plants, and herbivoiy in the establishment and renewal oi Cercociirpiis connmmities. Key words: Cercocarpus, litter iiuniiitaiii inalio^dny, seedlin 1 yvM' ol age; 2-7 mm hasal diameter; smooth hark; ma\ he u]) to 30 cm tall; S or i.. .^' leaxcs. Young plants >7 nun hasal diameter; smooth hark; plants to HO cm tali. ^oung plants >1.25 cm hasal diameter; smooth hark; plants to 1.5 m tall. Cracked hark; 1.5-3.0 m tall; crown broadened; ma\' be nuiltistemmed from base; not suppressed by adjacent larger mountain mahogany plants. Cracked bark; wide full crown; few dead branches; may have several stems from base; >3 m tall. Cracked bark; ma\' be multistemmed; numerous dead branches; ma)' be >3 m tall; frequentK suppressed by adjacent larger mountain mahogany plants. and had the same aspect throughout its length, and (4) all transects located in the same stand were 40 m or more apart. Table 3 describes the elevation, slope, and aspect of each transect. Cercocarpiis in the Shoshone Range are largely restricted to the Fo.xmount soil seiies (Carol Jett personal communication), which is a gravelly loam (specifically, a Loamy- skeletal, mixed Topic Cryboroll). This soil is well drained and moder- ately permeable. Depth to a paralithic contact averages 60-100 cm. All transects were located such that 20 m occurred in the Cercocarpiis stand and 20 m in the adjacent Artemisia community. Each tran- sect was divided into forty 1 X 1-m quadrats. Every Cercocarpiis rooted in each quadrat was classified by maturity class. For Cercocarpiis in established seedling, juvenile, and immature maturity classes, we determined whether the plant was rooted under the protective canopy of a live or dead shrub. Distribution of current-year seedling, estab- lished seedling, juvenile, and immature Cerco- carpiis was summarized for 10 classification categories (populations). These were (1) the number of Cercocarpiis in current-year seed- ling, established seedling, juvenile, and imma- ture maturity classes rooted in either the Cer- Elevation Slope Aspect Transect (ill) (%) (degrees) 1 2688 41 80 2 2688 41 80 3 2688 41 80 4 2400 29 290 5 2758 34 0 6 2758 34 0 FT 2758 25 168 cocarpiis connnunity or the adjacent Arteiimia community, and (2) the number of established seedling, juvenile, and inmiatiue Cercocarpiis rooted under and not under the canopy of a live or dead shrub. The Wilcoxon signed rank test was used to determine if there was a sig- nificant difference in the distribution of indi- viduals in the Cercocarpiis and Artemisia com- munities, respectively, for each maturity class. The significance level is P < 0.05 unless other- wise noted. Results Cunent-year Cercocaqms seedlings were not distributed evenly between Cercocarpiis stands and adjacent Artemisia communities (Table 4). Significantly more current-year seedlings were rooted in the Cercocarpiis community. At least 81% of established seedling, juve- nile, and immature Cercocarpiis were rooted in the adjacent Artemisia community (Table 4). For established seedling and juvenile maturity classes the difference in spatial distribution was significant; the significance level for immature Cercocarpus was P < 0.06. More established seedling, juvenile, and immature Cercocarpiis were rooted under the protective canopy of a live or dead shrub than in the open (Table 5). Only 1 transect had more plants without a protective canopy, but the sig- nificance level was P < 0. 10. Discussion Spatial distribution of current-year Cerco- carpiis seedlings and established young Cerco- carpiis had an inverse relationship (Tables 1, 4). Current-year seedlings were most abundant in Cercocarpiis stands dominated by large, ma- ture Cercocarpiis and least abundant in ddyd- cent Artemisia communities. Young, established 264 Great Basin Naturalist [Volume 56 Table 4. Number of cunent-year seedling, establislied seedling, juxenile, and immature mahogany rooted in Cercocar- piis (CER) stands dominated by mature individuals, and in adjacent Artem/sifl (ART) communities. Within each maturity class, total values between community types with different letters are significantly different (P < 0.05). Current- ■vear Established seedling seedli ng Juveni le I nun ature Transect CER ART CER ART CER ART CER ART 1 20 0 1 11 1 5 0 1 2 72 15 1 15 3 3 0 5 3 75 53 0 16 0 6 5 6 4 31 39 0 2 0 7 0 4 5 .337 25 0 11 0 19 0 0 6 .506 28 1 11 0 4 0 0 7 33 0 1 9 0 2 0 5 Total 1074a 160b 4a 75b 4a 46b 5a 21ai Percent 87 13 5 95 8 92 19 81 ISiRiiificantly different at P < 0.06. Cercocarpus were virtually absent from mature Cercocarpus stands but had a greater abun- dance in adjacent Artemisia communities (Tables 1, 4). Young Cercocarpus were also abundant in stands with low Cercocarpus crown cover or relatively few large Cercocarpus (Table 1). The low densitv' of current-year seedlings in adja- cent Artemisia communities (Table 4) has 2 possible interpretations: (1) viable Cercocarpus seeds were not dispersed into the Artemisia community, or (2) germination of Cercocarpus seed was impaired. Because data about seed densities are lacking, a definitive conclusion cannot be made. Cercocarpus seed, however, is primarily wind dispersed (US DA 1948); there- fore, it is unlikely that few seeds were present in the Artemisia community, particularly since all data were collected within 20 m of the Cer- cocarpus stands. Most likely, over 85% fewer Cercocarpus seedlings were in the Artemisia community (Table 4) because seed germination was substantially lower than in the Cercocar- pus stands. The inverse relationship for distribution of current-year seedlings and established young Cercocarpus indicates that locations with a high abundance of current-year seedlings are not necessarily locations with the best seedling sui"vival. Populations peipetuate when seedlings survive and advance into successively older niatiuity classes, eventually producing new seedlings. The pattern for spatial distribution of current-year seedling, established seedling, juvenile, and immature Cercocarpus deri\ed from this sttidy and that conducted by Schult/ et al. (1990, 1991) indicates that 4 factors may influence sui-vival of current-year seedlings as well as plants in the youngest maturity classes: (1) presence or thickness of plant litter, (2) root growth characteristics, (3) presence of nurse plants, and (4) herbivoiy. Moderate levels of litter can favor seed ger- mination and seedling establishment by de- creasing soil temperature and increasing soil moisture (Evans and Young 1970). Thick litter, however, can reduce seedling establishment and survival by preventing or restricting contact between soil and seed or soil and root (Fowler 1986). High litter cover (Table 1) and a thick la\er of litter (personal obser\'ation) were common in Cercocarpus stands in the Shoshone Range. Litter cover and litter thickness were not mea- sured in adjacent Artemisia communities; how- ever, litter cover in high-ele\ation (>2200 m) Artemisia communities ranges from 15% to 50% (Tueller and Eckert 1987). Extensive and deep litter in Cercocarpus stands may promote seed germination but decrease seedling sur- vival because roots from Cercocarpus seedlings seldom make contact with the mineral soil. Less litter in the Artemisia community' may re- duce Cercocarpus seed geniiination but enhance stuAi\'al of seeds that germinate. Root growth characteristics ma\ pla> an important role. Rapid root giowth that current-year Cerco- carpus seedlings experience (Dedy 1975) should enhance sin-\ i\ orship of Cercocarpus seedlings during seasonal drought, a common phenome- non in the Great Basin. Root systems that undergo rapid elongation should be able to fol- low a retreating zone of soil moistiue (down- ward) better than root systems that elongate slowly. We excavated several Cercocarpus seed- lings rooted in thick plant litter and found that root growth was extensive (20+ cm) but not 1996] Cercocahi'us Ki:c;i:Ni:iivrioN 265 Table 5. The number of established seecHin^, juvenile, and ininiatiue Ccrrocarpiis rooted under and not under another sliiiib or sluiib skeleton. Siuniliianee le\t'l is /' < 0.10. Transeet Hooted ui uler Not rooted under 1 16 3 2 2.3 4 3 20 13 4 S 5 5 6 24 6 9 / 7 15 2 Total y7a 58b Peixi-ntaue 03 37 d()\\n\\'arcl toward or into the mineral soil. Root growth was largely lateral. Following ger- mination in early spring, available moistnre in both mineral soil and plant litter is probably high, sinee cool temperatures and abundant precipitation are common (Houghton et al. 1975). Because moisture is not limiting early in the growing season, root growth probably fol- lows the path of least resistance. When thick litter resides on top of mineral soil, the path of least resistance would be laterally through the litter, not downward through the mineral soil. The loamy soil that Cercocorpiis stands inhabit undoubtedly stores and retains more water than plant litter does, and thus should desic- cate more slowly. If thick plant litter prevents or retards roots of current-year Cercocarpiis seedlings from reaching or penetrating moist mineral soil, seedling mortality should be high when litter desiccates rapidly later in the sum- mer. We obsen'ed high mortality for current- year Cercocarpiis seedlings in August in Cerco- carpiis stands with thick accumulations of lit- ter. Less litter on Peavine Mountain (Table 1) and in the Artemisia community (see Tueller and Eckert 1987) may enable root systems of Cercocarpiis seedlings at these locations to grow downward into mineral soil immediately following germination. This should increase survivorship of current-year seedlings, which may account (at least partially) for the greater abundance of established seedling, juvenile, and immature Cercocarpiis on sites with less surface litter. Herbivory may also play a role in seedling survival. Current-year Cercocarpiis seedlings have an average leaf surface area of only 4 cm^ (Dealy 1975), which herbivores can easily con- sume. Herbivory can adversely affect estab- lishment of woody species (Marquis 1974, McAuliffc 1986), including Cercocar^ms (Scheldt and Tisdale 1970). The presence of protective ninsc plants, therefore, may be important for regeneration ol Cercocarpiis seedlings. Cercocarpiis stands in the Shoshone Range had a mean shrub canopy cover of 11% (Schultz et al. 1990). Total shrub canopy cover was not measured in adjacent Artemisia communities; however, it generally ranges from 41% to 50% (Tueller and Eckert 1987). Thus, shrub cover in adjacent Artemisia communities is 3.5 to 4 times greater than that in Cercocarpiis stands. Since more established seedling, juvenile, and immature Cercocarpiis were rooted imder a shrub or shrub skeleton than not (Table 5), the difference in shrub canopy cover between Cer- cocarpiis stands and adjacent Artemisia com- mimities may influence survival of cuiTcnt-year seedlings, established seedlings, juvenile, and immature Cercocarpiis. Artemisia and other short-statured shrubs may serve as nurse plants and protect small Cercocarpiis (including cur- rent-year seedlings) from herbivores until their photosynthetic surface is large enough to cope with frequent browsing. Since shrub cover is low in Cercocarpiis stands, more young Cerco- carpiis are probably exposed to herbivores than in Artemisia communities. This may help explain the near absence of young Cercocarpiis in Cer- cocarpus stands and their greater abundance in adjacent Artemisia communities. Conclusions Abundance of current-year Cercocarpiis seedlings is greatest in Cercocarpiis stands that have high Cercocarpiis canopy cover, large mean Cercocarpiis crown volume, and an extensive layer of plant litter. These stand attributes also result in a low density of plants in established seedling, juvenile, and immature maturity classes. Established young Cercocarpiis are most abundant where gaps occur in the Cerco- carpiis canopy, or in adjacent Artemisia com- munities. Survival of current-year seedlings appears best at locations that permit roots of seedlings to make contact with mineral soil. Survival of current-year seedlings and progres- sion of individuals from established seedling maturity class into successively older maturity classes appear to be enhanced by the presence of a shrub canopy that protects small Cercocar- piis from herbivores. 266 Great Basin Naturalist [Volume 56 Literature Cited Austin, D. A., and E J. Urness. 1980. Response of curlleaf mountain mahogany to piiining treatments in north- em Utali. Journal of Range Management 33: 27.5-277. Davis, J. N. 1976. Eeological investigations in Cercocarpus ledifoliiis Nutt. communities of Utah. Unpubhshed master's thesis, Brigham Young University', Prove, UT. Davis, J. N., and J. D. Brotmerson. 1991. Ecological rela- tionships of curlleaf mountain mahogany {Cercocar- pus ledifolitis Nutt.) communities in Utah and impli- cations for management. Great Basin Naturalist 51: 153-166. Dealy, J. E. 1975. Ecology of curlleaf mountain mahogany {Cercocarpus ledifolius Nutt.) in eastern Oregon and adjacent areas. Unpublished doctoral dissertation, Oregon State University, Coi^vallis. Duncan, E. 1975. The ecology of curlleaf mountain mahog- any in southwestern Montana with special reference to mule deer Unpublished master's thesis, Montana State University, Bozeman. Evans, R. A., and J. A. Young. 1970. Plant litter and the establishment of alien annual weeds in rangeland communities. Weed Science 18: 697-703. Fowler, N. L. 1986. Microsite requirements for germina- tion and establishment of three grass species. Ameri- can Midland Naturalist 115: 131-145. Holmoren, R. C. 1954. A comparison of browse species for the revegetation of big-game winter ranges of southwestern Idaho. USDA, Intermountain Forest and Range E.xperiment Station, Researcli Paper 33. HosKiNS, L. W., and R D. Dalke. 1955. Winter browse on the Pocatello big game range in southeastern Idaho. Journal of Wildlife Management 19: 215-225. Houghton, J. G., C. M. Sakamoto, and R. O. Gifford. 1975. Nevada's weatlier and climate. Nevada Bureau of Mines and Geology, Special Publication 2. LlACOS, L. G., and E. C. Nord. 1961. Cercocarpus seed dor- mancy yields to acid and thiorea. Journal of Range Management 14: 317-320. Marquis, D. A. 1974. The impact of deer browsing on Allegheny hardwood vegetation. USDA, Forest Ser- vice, Research Paper NE-308. McAuLlFFE, J. R. 1986. Herbivore limited establishment of a Sonoran Desert tree, Cercidium microphijUum. Ecology 67: 276-280. Ormiston, J. H. 1978. Response of curlleaf mountain mahogany to top piiming in southwest Montana. Pro- ceedings of the First International Range Congress, Denver, CO. Plummer, A. R, R. L. Gensen, and H. D. Stapley. 1957. Job completion report for game forage revegetation project W-82-R-2. Utah State Department of Fish and Game. Plummer, A. R, D. R. Christensen, and S. B. Monsen. 1968. Restoring big game range in Utah. Utah Divi- sion of Fish and Game, Publication 69-3. Scheldt, R. S. 1969. Ecologx' and utilization of curl-leaf mountain mahogany in Idaho. Unpublished master's thesis. University' of Idaho, Moscow. Scheldt, R. S., .4ND E. W. Tisdale. 1970. Ecology' and uti- lization of curlleaf mountain mahogany in Idaho. Uni- \'ersity of Idaho, Forest, Wildlife, and Range E.xperi- ment Station, Note 15. SCHULTZ, B. W 1987. Ecology of curlleaf mountain mahogany {Cercocarpus ledifolius) in western and central Nevada: population stnicture and dynamics. Unpublished master's thesis. University of Nevada, Reno. Ill pp. Schultz, B. W, R. J. Tausch, and R T Tueller. 1991. Size, age, and density relationships in curlleaf moun- tain mahogany {Cercocarpus ledifolius) populations in western and central Nevada: competitive implica- tions. Great Basin Naturalist 51: 183-191. Schultz, B. W, P T Tueller, and R J. Tausch. 1990. Ecology of mountain mahogany {Cercocarpus ledi- folius) in western and central Nevada: community and population structure. Journal of Range Manage- ment 43: 13-20. Smith, A. D. 1950. Feeding deer on browse species during winter. Journal of Range Management 3: 130-132. Smith, A. D., and R. L. Hubrard. 1954. Preference rat- ings for winter deer forages from northern Utah ranges based on browsing time and forage consumed. Journal of Range Management 7: 262-265. Tueller, R T, and R. E. Eckert 1987. Big sagebrush {Artemisia tridentata vaseyana) and longleaf snow- beny {Symphoricarpos oreophilus) plant associations in northeastern Nevada. Great Basin Naturalist 47: 117-131. USDA. 1948. Cercocarpus H. B. K. mountain mahogany. Pages 132-133 in Woody plant seed manual. USDA Miscellaneous Publication 654. Young, J. A., R. A. Evans, and D. L. Neal. 1978. Treat- ment of curlleaf cercocaipus seeds to enhance germi- nation. Joimial of \\'ildlife Management 42: 614—620. Received 17 May 1995 Accepted 25 March 1996 Great Basin Naturalist 56(3), © 1996, pp. 267-271 POTENTIAL FOR CONTROLLING THE SPREAD OF CENTAUREA MACULOSA WITH GRASS COMPEL ITION John L. Liiul' Mountains. It is oiten found in plant communities dominated by Psc'uchmn'pu-ria spicatum or Fcstuca idahoeims, hut it rareK invades roadsides dominated !)>• Bnnniis biennis Leyss. Ahove^round hiomass of the 3 j^rass species j^rown in nii.xture with Cciitaiirea was compared to growdi in monoculture at a range of nitrogen input levels. The results suggest that Brointi.s is capable of suppressing the gnw'th of Centaurca with the degree of sui:)pression increasing with increasing nitrogen lexels. The 2 nati\e grasses had no impact on Centaurca under tlie controlled en\ ironment conditions of this study. Keij words: annpctition, weed control. Centaurca maculosa, Bromus inermis, Agropyron spicatimi, Festuca idahoen- sis, exotic plants. Centaurca maculosa Lam. (spotted knap- weed) is a major weed associated with spring wet-simimer diy areas of the northern Rocky Mountains (Forcella and Harvey 198L Tyser and Key 1988, Weaver et al 1989). Centaurca dominates waste places, invades disturbed rangeland, and sometimes invades undisturbed range (Tyser and Key 1988). In contrast, it rarely invades roadsides dominated by Bromus inermis Leyss. (Weaver et ak 1989). This suggests that it may be exchided fi-om waste pkices that are planted to Bromus before Centaurca invades. Alternatively, because planting exotics violates the charge of national park managers, one may ask whether Centaurca might also be excluded from disturbed areas by planting native grasses that naturally dominate either relatively dry {Pseudorocgneria spicatum [Pursh] Scribner and Smith = Agropyron spicatum) or more moist {Festuca idahoensis Elmer) foothill habitats. Weed suppression may be accomplished by (1) preempting resources with more competi- tive plant species or (2) using biocontrols or herbicides that selectively increase weed mor- talit\', decrease vigor, or prevent reproduction (Lindquist et al. 1995). This study considers management of Centaurca maculosa by compe- tition rather than by common herbicide and biocontrol methods. This approach deserves attention because it may be less expensive and more effective than herbicides in the long term. Our objective was to measure the competi- tive ability of 3 grass species against Centaurca in 2-way interaction experiments in sand cul- ture. Mixture and monoculture treatments were tested for 12 wk at 5 positions on a nitro- gen gradient to determine whether competi- tive relations were influenced by differences in nitrogen availability. A plant's ability to com- pete is related to its growth rate or ability to gain biomass relative to associated species (Haiper 1977). We compared aboveground bio- mass of each species grown in mixture with Centaurca to its growth in monoculture. Materials and Methods The rhizomatous exotic pasture grass Bromus inermis Leyss. and 2 native bunchgrasses nor- mally dominating relatively diy foothills {Pseu- dorocgneria spicatum) or moister grasslands immediately above and below the conifer zone {Festuca idahoensis) were grown in 2-species mixtures (replacement series) with C. maculosa. Experiments consisted of 3 competition treat- ments (monocultures of both grass and Centau- rca, and 50:50 mixture) combined with 5 nitro- gen addition treatments. Each treatment com- bination had 10 replicates. Within each experi- ment, pots were arranged in a completely ran- domized design on a greenhouse bench and rotated weekly to minimize position effects. Each experiment was subject to different light 'Department of Biology and Department ot Plant, Soil, and Environmental Sciences, Montana State University, Bozeman, MT.5971" -Present addres.s: Department of Agronomy, Universit\' of Nebraska. Lincoln, NE 68.58.3-091.'5. 267 268 Great Basin Naturalist [Volume 56 conditions because of its position in the green- house. A square planting pattern was used with 4 plants spaced 5 cm apart. In each pot in the mixture treatments, plants of the same species were located on the diagonal. Seeds were planted at a depth of 1.0 cm in 1000-cm'^ pots filled with coarse washed sand. Pots were watered daily for 1 wk to allow seedling establishment. Excess seedlings were thinned and remaining seedlings allowed to grow for an additional week prior to the addi- tion of nutrients. The basic nutrient solution was balanced with respect to all essential nutri- ents but could be varied to allow the establish- ment of nitrogen levels from 0%, 1%, 10%, 30%, and 100% of a standard level (Machlis and Torrey 1956). Sufficient nutrient solution (200 ml) was applied to satinate the pot twice weekly and water (200 ml) was added once each week. Regular watering with nutrient solution and alternate washing with tap water held the soil solution near the applied level and pre\'ented any concentration of the soil solution due to evapoti-anspiration. Experiments were conducted during March, April, and May 1988, when greenhouse temperatures ranged from 14° to 32° C (25° C mean). Twelve weeks after emergence, plants in each pot were clipped at the soil surface, sepa- rated by species, dried at 45° C for 5 d, and weighed. Nonlinear regression procedin-es (SAS 1988, Gauss-Newton least squares estimation method) were used to fit a rectangular hyperbola equa- tion [1] (Cousens 1985) to mixture and mono- culture data for each species: /, • N ;i + /. • N [1] calculated to determine whether competitive relationships varied across relative nitrogen levels. RCI is calculated as RGI = (B, I^mix)/B, [2] A, where B,^^,^,-,^ and B,^^j^ are the aboveground diy biomass (g plant"^) for a species grown in mono- culture and mixture, respectively. A negative RCI value indicates that the species performs better in mixture than in monoculture. RCI may be the best measure for determining species displacement under competitive conditions across a resource gradient (Grace 1995). Analy- sis of variance was used to test for differences in RCI within a species across nitrogen treat- ments. Student's t was used to compare RCI between species at each nitrogen addition level. Results A hvperbolic relationship between individ- ual plant biomass and relative nitrogen level was found in all mixtures and monocultures (Figs, la-f ). Estimates of /,• (biomass at inter- cept) differed between mixtures and monocul- tures only for Centaurea grown in mixture with Broiniis (Table 1). Estimates of A,- (maximum biomass) differed for Bronuis and Centaurea (Tiible 1). Relative competition intensity was signifi- cantly negative for Broimis at all nitrogen addi- tion levels; it varied from negati\'e values at low nitrogen to positixe \'alues at higher nitro- gen levels for Centaurea in competition with Bronius (Fig. 2). However, RCI did not differ from zero in the experiments where P. spiea- tum and E idahoensis were in competition w ith Centaurea (data not shown). Discussion where B — aboveground dry biomass (g plant~l), Aj = maximum aboveground biomass of species / (g plant"^), N = relative nitrogen addition level, and /, = biomass of species / as relative nitrogen addition level approaches zero. To determine the relative success of Centau- rea in competition with each grass species, estimates ol A,- and Ij were compared between mixtures and monocultures using the extra sum of sc^uares procedure (Ratkowsky 1983, Lindcjuist et al. 1996). In addition, relative^ competition intensity (RCI; Grace 1995) was Growth response of Bromus to nitrogen was greater in mixture with Centaurea than in monoculture, as indicated by the regression lines (Fig. la) and the negative RCI values across all nitrogen addition levels (Fig. 2). In contrast, growth response of Centaurea was lower in mixtme with Bromus than in mono- culture (Fig. 1). The increase in Centaurea RCI at high relative nitrogen le\el indicates that Bronuis is a better competitor in the high nitiogen treatments (Fig. 2). Results suggest that Bnnnus is capable of suppressing the 1996] CoMFEiiTivE Suppression of Centaurea 269 CQ 0.0 0.2 0.4 0.6 0.8 1.0 i3 CL OS E o 03 CX3 O "3 o nj S d 0.0 0.2 0.4 0.6 0.8 1.0 1.5 Q. o> 1.2 CO CO TO F 0.9 o CQ 0.6 E 3 ^ 0.3 o a to 0.0 CT3 1.5 Q. -S 1.2 CO CO CO F 0.9 O CQ nfi cn on o "3 0.3 o (t! 0.0 0.0 0.2 0.4 0.6 0.8 1.0 O 0.0 0.2 0.4 0.6 0.8 1.0 10 C O -c 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 ca 3.0 Q. O) 2.5 c/) CO 2.0 F o m l.b to 1.0 u n 0,5 o 0.0 o 0.0 0.2 0.4 0.6 0.8 1.0 Relative Nitrogen Addition Level Fig. 1. Plot of obsei-ved (o) and predicted (•) abovegroiind dn' biomass plant"' on relative nitrogen addition level when grown in monoculture ( ) and mi.xture ( ): a, Bnvniis grown in monoculture and in mixture with Centaurea; b, Centaurea grown in monoculture and in mixture with Bronius; c, Pscudoroegneria grown in monoculture and in mix- ture with Centaurea; d, Centaurea grown in monoculture and in mixture with Pseudoroegneria; e, Centaurea grown in monoculture and in mixture with Fe.stuea; f, Festuca grown in monoculture and in mixture with Centaurea. 270 Great Basin Naturalist [Volume 56 Table 1. Estimates of parameter values followed by asymptotic standard errors for maximum aboveground biomass (g plant~l) (A), biomass as relative nitrogen level approaches zero (/), and the coefficient of determination (r-) obtained from fitting equation [1] to monoculture and mixture data of each species. Variation in / and A between competition treatments was tested using the extra sum of squares principle, with P value indicating the significance level for the comparison of parameter (Coeff ) values from the monoculture and mixture regressions. Species Coeff Competition treatment Monoculture Mixture P value Broiiius A I 1-2 Centaiirea with A Bromiis / 1-2 Pscudoroef^neria A / r2 Centaiirea with A Pseiidoroegneria I r2 Festuca A I r2 Centaiirea with A Festuca I 1.884(0.11.5) 8.769(1.141) 0.90 1.306(0.082) 10.621(1.861) 0.85 1.289(0.237) 1.758(0.323) 0.80 0.636(0.048) 6.770(1.639) 0.77 0.595(0.029) 16.747(3.892) 0.83 1.262(0.114) 6.418(1.295) 0.77 3.819(0..361) 12.893(2.197) 0.84 0.438(0.0,56) 45.717(35.47) 0.14 0.827(0.087) 2.490(0.442) 0.82 0.805(0.107) 4.617(1.454) 0.60 0.491(0.056) 23.921(14.925) 0.39 1.522(0.172) 10.061(2.870) 0.66 < 0.000 0.283 < 0.000 0.023 0.081 0.410 0.307 0.665 0.303 0.857 0.514 0.535 growth of Centaiirea, the degree of suppres- sion increasing with increasing nitrogen levels. Growth response of Pseudoroegneria and Festuca to nitrogen when growing in mixture witli Centaiirea did not differ fi-om their response in monoculture. Likewise, growth response of Centaiirea did not differ between monoculture and mixtures with Pseudoroegneria or Festuca. Therefore, these results suggest that these native grasses are not likely to increase or sup- press growth of Centaiirea, regardless of nitro- gen addition level. This result is contrary to the obsened invasion of Centaiirea into com- munities dominated by these grasses. One ex- planation may be that disturbance (especially grazing) in the field creates gaps in the grass communit\' where Centaiirea can establish itself even though it is not a superior competitor for resources. Competitive interactions were greater be- tween each grass species and Centaiirea at the high end of the nitrogen gradient. This may be a function of rapid growth. Thus, in nitrogen- rich environments fast-growing plants may rapidly occupy space and usurp resources to the exclusion of slow-growing species (Grime 1979, Radosevich and Holt 1984). Similar com- petitive effects may be expected to occur on other soil resource gradients, assuming adapta- tions for ac(}uisition of nitrogen and other mobile nutrients, as well as water, are similar (Grime 1979, Fitter and Hay 1987). In addition, one may hypothesize, based on the resource ratio theory (Tilman 1982), that Bronius is a superior competitor for nutrients other than nitrogen relative to Centaiirea. By increasing nitrogen, both species should be limited by essential nutrients other than nitrogen, and the species with the lowest R* (the superior competitior) for the other nutrients should displace the species with the higher R* for the same nutri- ents (Tilman 1990). The abilit) of Bromiis to out-compete Cen- taiirea in nutrient culture provides one expla- nation for the obsei'ved population dynamics of Centaiirea in the field. Roadsides seeded with Bromiis are rareh in\'aded In Centaiirea (Weaver et al. 1989). Both field and laboraton' observa- tions suggest that disturbed sites seeded simul- taneousK' with Centaiirea and the exotic, Bro- miis, will be dominated b\ Bromiis. The effec- tiveness of Bromiis in suppressing Centaiirea may be increased with fertilization. Further- more, it may be expected that established Bro- miis plants will suppress the growth oi Centaii- rea seedlings. The results of this study suggest that at the seedling stage Bromus may be used to competiti\cl\' exclude Centaiirea. This method of weed management merits trial in the field. On the other hand, the regional dominants. 1996] CoMiuniTivK Sli'I'hkssion of Cemavrea 271 -1.5 0 0.01 0.1 0.3 1.0 Relative Nitrogen Addition Level Fig. 2. Relative competition intensity (RCI) of Broinii.s and Ccntawea across 5 relative nitrogen addition levels. Let- ters above bars indicate whether RCI varies (Duncan's multiple range test, F < 0.05) within species across nitrogen le\el. An asterisk indicates that RCI differs (P < 0.05) among species at that nitrogen level. Pseiidoroegneria and Festuca, probably would not sufficienth' suppress Centaiirea to decrease the potential for invasion. Advantages of the competitive method over herbicides and biocontrol treatments used to manage Centaurea are its long duration and low environmental impact. Given these advan- tages, exclusion of Centaurea with Bromiis merits trial in environments where the danger of invasion exists. Literature Cited COLSEXS, R. 1985. A simple model relating yield loss to weed density. Anuids of Applied Biology- 107: 239-252. Fitter, A. H., .^nd R. K. M. Hay. 1987. Environmental physiology of plants. 2nd edition. Academic Press Inc., San Diego, CA. 423 pp. FORCELLA, E, AND S. Harvey. 1981. Migration and distri- bution of 100 alien weeds in NW USA, 1881-1980. Herbarium, Montana State University, Bozeman. 117 pp. Gr\ce, J. B. 1995. On the measurement of plant competi- tion intensit}'. Ecology' 76; 305-308. Grime, J. E 1979. Plant strategies and vegetation pro- cesses. John Wiley and Sons, London. 222 pp. Harper, J. L. 1977. Population biology of plants. Academic Press Inc., San Diego, CA. 892 pp. LlNDQUIST, J. L., B. D. M.\.\\\ ELL, D. D. BlULER, AND J. L. GuN,soLUS. 1995. Modeling the population dynamics and economics of velvetleaf (Abutilon flicophrasti) control in a com {Zea //ia|/.s)-soybean (Glycine max) rotation. Weed Science 43: 269-275. LlNDQUIST, J. L., D. A. MoRTENSEN, S. A. Clay, R. Schmenk, J. J. Kells, K. Howatt, and R Westr\. 1996. Stability of corn-velvetleaf interference relationships. Weed Science 44; 309-313. Machlis, L., and J. ToRREY 1956. Plants in action. Free- man, San Francisco, CA. 282 pp. Radosevich, R. S., and J. S. Holt 1984. Weed ecology: implication for vegetation management. John Wiley & Sons Inc., New York. 265 pp. Ratkowsky, D. a. 1983. Nonlinear regression modeling: a unified practical approach. Marcel Dekker, Inc., New York. 276 pp. SAS. 1988. SAS/STAT user's guide, release 6.03. SAS Institute, Caiy, NC. 1028 pp. Tilman, D. 1982. Resource competition and comnumity stiiicture. Princeton University Press, Princeton, NJ. . 1990. Mechanisms of plant competition for nutri- ents: the elements of a predictive theory of competi- tion. Pages 117-141 in J. B. Grace and D. Tilman, editors. Perspectives on plant competition. Academic Press, Inc., New York. Tyser, R., and C. Key 1988. Spotted knapweed in natural area fescue grasslands; an ecological assessment. Northwest Science 62; 151-160. Weaver, T, D. Gustaf.son, J. Lichthardt, and B. Woods. 1989. Distribution of e.xotic plants in the Northern Rocky Mountains by environmental type and distur- bance condition. MSU Biology Report 41, Bozeman, MT 91 pp. Received 26 September 1995 Accepted 29 April 1996 Great Basin Naturalist 56(3), © 1996, pp. 272-275 INDICATORS OF RED SQUIRREL {TAMIASCIURUS HUDSONICUS) ABUNDANCE IN THE WHITEBARK PINE ZONE David J. Mattsonl and Daniel E Reinhart^ Abstract. — We investigated occupied squinel middens and s(|uinel sightings and vocalizations as indicators of red squiiTel (Tamiasciunis hudsonicus) abundance in the high-elevation whitebark pine {Pimis alhicauUs) zone. Data were col- lected 1984-1989 fioni line transects located on 2 study sites in the Yellowstone ecosystem. We evaluated the perfomiance of each measure on the basis of precision and biological considerations. We concluded that, of the 3 measures, active mid- dens were the best indicator of red sc|uirrel abundance. We also obsei^ved that the density of active middens dropped by 48%-66% between 1987 and 1989, following a severe drought and extensive wildfires that burned one of the study sites during 1988. Key words: transect. Fourier series, midden, voccdization, sigJitinf;, wildfire. Whitebark pine {Piniis alhicauUs) seeds are an important bear food that affects the sui^vival and fecundity of grizzly bears {Ursiis arctos) in the Yellowstone ecosystem. Use of pine seeds by grizzlies is almost entirely contingent upon the availabilit)' of cones cached in middens (i.e., larder hoards) by red squirrels {Tamiasciunis hudsonicus). Management of whitebark pine habitats for grizzlies has thus become contin- gent upon management of red squirrel popula- tions (Mattson and Reinhart 1994). We studied red squirrels in the whitebark pine zone using data collected from line tran- sects. Because these data included counts of middens, animals, and vocalizations, we were able to evaluate the relative efficacy of these 3 indicators of sc|uinel presence. We were inter- ested in identifying a "well-behaved" and rele- vant indicator of density to facilitate our inves- tigation of relationships between squirrel abundance and environmental factors such as midden use by grizzly bears. We were also interested in providing managers with an approach they could use to indicate squirrel abundance, short of using intensive methods that relied upon marked animals. Study Area Om- study area consisted of 2 sites, one located on the Mt. Washburn massif in north central Yellowstone National Park (44°47'N), and the other near Cooke City, Montana, immediately northeast of the park (45°00'N). These sites spanned the whitebark pine zone, from 2360 to 2870 m elevation. The whitebark pine zone borders upper timberline and is accordingly cold (average annual temperatures <0°C), often wind\', and subject to deep (1-2 m) winter snow accumulations (Weaver 1990). Materl^ls and Methods Broad study objectixes affected our transect design. We mapped the study area by habitat type-cover type strata based upon ground- truthed interpretation of 1:20,000 aerial pho- tography. The result was a fine-scale mosaic, with individual map poKgons (forest stands) sometimes as small as indi\ idual S(]uirrel terri- tories. To minimize effects of edge between different habitat types, we placed transects so as to maximize the number of right-angle inter- sections with stand l)oimdaries as well as the amount of intersection with stand interiors. Because of this consideration and because for- est and meadow were \ariousl\ intermixed, transect lines were of imequal length. We surveyed transects in the same order each year, beginning after 10 August and end- ing prior to 28 September Two obseivers walked pennanenth' marked transect lines, with one ob- sener primariK responsible for obsen ations and the other primarily responsible for recording iNalKMKil Bi(,li,i;ii;il Scr-vicv, Dcparlinciit nl Fish ami Wikllilc Ucmhuvcs, liiiivc-rsit) cit Idalio, Moscow. ID 83843. 2U.S. NalioTial Park Srnacc, Bcsoiinc MaiiaKrinnit, Ycllousloiic \atioiial Park, WV 82190. 272 1996J Red Squikhkl Indicatoks 273 data and keeping on line. At least one of llie observers (the junior author) was the same dining all years at both sites. Voealizations and obsened scjuirrels or middens (both aeti\'e and inactive, b\^ criteria of Finlc)' [1969]) were recorded along with their estimated perpen- dicular distances from the transect line. Indi- \ idual cone caches were not considered to be a "midden and were easily distinguished fiom these larger, more peniianent features. We used die computer program TRANSECT (Burnham et al. 1980) to estimate densities. Individual transects constituted sample units for densit) calculations. As recommended by Burnham et al. (1980), we used the Fourier series, with 1-4 terms, to estimate distance-to- line probability detection functions (g[x]). The distance at which we specified the limits of detectabilit}' lor oiu" measures (i.e., the cut-point) exerted considerable influence on the fit of the Fourier function to the observed detection dis- tribution. Accordingly, we varied cut-points to achieve the best fit to each year- or site-spe- cific data set. Because data were collected from only 35 transects on the Mt. Washburn massif during 1984 and from 15 transects in the Cooke City area during 1984 and 1989 (com- pared to 57 and 21 transects, respectively, for all other years), we also calculated densities solely from these original 35 and 15 transects for all years so as to allow comparison with results from 1984. Results We sampled the study area 5 yr, 1984-1987 and 1989. During 1988 wildfires Immed 562,000 ha of the Yellowstone area, including 52% of the Mt. Washburn transects (47% severely). Transects on the Mt. Washburn area totaled 18.9 km during 1984 (mean transect length [A'] = 539 ± 245 m[s]) and 29.8 km during the remaining 4 yr (X = 523 ± 258 m). Similarly, during 1984 and 1989 transects on the Cooke City area totaled 16.4 km (X = 1091 ± 427 m) and 21.1 km during the remaining years (X = 1005 ± 405 m). We recorded 124 squirrel sight- ings, 641 vocalizations, and 300 active middens on die Mt. Washburn study site and 54 sight- ings, 528 vocalizations, and 201 active middens on the Cooke Cit>' study site during the 5 study years. The small number of sightings from the Cooke Cit>' site prevented us from estimating annual densities from this measure for this area. Total distance-to-line frecjuencN' distribu- tions for each of the 3 measures did not differ between the Mt. Washburn and Cooke City study sites (Mantel-Haenszel x^ for ordinal categories, P = 0.51, 0.61, and 0.35 for active middens, vocalizations, and sightings, respec- tively). Perpendicular distributions of sightings and active middens peaked in the nearest (<10 m) distance categoiy, although the distribution of sightings more closely resembled a negative exponential and the distribution of middens a negative sigmoidal function. The majorit>' (65% and 78%, respectively, by year and study site) of both these distributions were adequately fit ix^ test, P > 0.10) by a single-term Fourier fimction. Distributions of vocalizations peaked in the 2nd (11-20 m) distance categor)' and were characteristically (94%) fit by a 2- or higher-term Fourier function. In 3 (18%) in- stances we could not achieve an adequate fit by any model. Relationships among annual density esti- mates from the 3 measures were varied (Fig. 1). On the Mt. Washburn site, mean sighting and vocalization densities were weakly corre- lated (r = 0.722), but tended to have overlap- ping 95% confidence inten'als. Onl\' 2 of 9 con- fidence intei-vals for the obsewed estimates (all years, for both the 1984 and inclusive samples) did not contain the line describing perfect cor- respondence (Fig. Id). In all but a single in- stance (Cooke City, 1984), mean midden densi- ties were greater than mean densities of the other 2 measures and were more strongly cor- related with sightings than vocalizations (r = 0.981 versus r = 0.831, respectively, for tran- sects 1-57, Mt. Washburn; Fig. Ic). However, in this case, only two of nine 95% confidence intei-vals for midden and sighting densities in- cluded the possibility of perfect correspondence. Conclusions From these results we concluded diat densi- ties calculated from active middens were more useful than densities calculated from the odier 2 measures for indicating red squirrel abun- dance. Our conclusion followed from the greater apparent detectabilit}' of middens compared to the squinels themselves, the consistency with which a single-term Fourier function described 274 (a) MT. WASHBURN Great Basin Naturalist (b) COOKE CITY [Volume 56 1.20 .00 0.20 0.40 0.60 0.80 1.00 1.20 DENSITY OF VOCALIZATIONS (n/lia) Q 0.80 > 0.60 0.00 0.20 0.40 0.60 0.80 1.00 1.20 DENSITY OF VOCALIZATIONS (n/lia) (c) MT. WASHBURN (d) MT. WASHBURN 0.80 > 0.60 0.00 0.00 0.20 0.40 0.60 0.80 1.00 1.20 DENSITY OF SIGHTINGS (n/ha) 0.80 N 0.60 - > 0.40 - 0.00 0.20 0.40 0.60 0.80 DENSITY OF SIGHTINGS (ii/lia) Fig. 1. Relation.ship.s heKveen annual e,stiniate.s of den.sity for acti\t^ middens compared to \ ocalization.s. (a) for Mt. Wa.shbiirn and (h) for Cooke Cih; (c) sightings compared to active middens for Mt. Washburn, and (d) sightings compared to vocahzations, Mt. Washbimi, 1984— I9(S7 and 19S9. Enor bars correspond to 95% confidence inten'als, soHd circles to results from all transects, and open circles to results from the fewer transects established and first sun e>ed in 1984. Diag- onals represent perfect correspondence between estimates. the probability detection distiibiition for mid- dens, and the resulting consistently smaller standard errors for the density estimates. In addition, scatter plots showed that active mid- den densities tended to be >() when sighting and vocalization deiTsities were not. B\ impli- cation, vocalization and especially sighting densities were more likely to underestimate true squirrel densities; i.e., at the same time that active middens clearly indicated the pres- ence of squirrels, sightings and \ocalizations could suggest there were none. Because red squiiTel middens ai^e nonmobile, often numerous, relatixeK' easily obsei^ved, and t\picall> associated w ith onl\ one s(iuirrel (Kil- ham 1954, M. Smith 1968, Wolff and Zasada 1975, Vahle and Patton 1983), the\- are logical indicators ol s(iuirrel abundance. Furthermore, the\' do not suffer from sampling problems associated with w eadier, season, and time of da\' 1996] Rkd SnuiUREL Indicators 275 T.XBLK 1. E,stiniated mean (n lur') and standard error (.s^) Tor dcMisities of active middens oti the Mt. Washburn and tlookc City study sites, 1984-1987 and 1989, percent coefficient of'variations for animal variation 1984-1987, and percent decline in densit>' i'rom 1987 to 1989. Results are given for the transects estahlished and sui-veyed during 1984 (1-.35 and 1-15) and for the larijer sample of transects sin\e\'ed during ail other vears, except for 1989, in the Cooke Cit\' area (1-57 and 1-21). Mt. Wash! )nrn (^ooki ■ City Tians. [ 1-35 Trans. 1-57 Trans. 1 1-15 Trans. 1-21 Year .Mean [s^) Mean (,v^) Mean (.v.v) Mean (.s-) 1984 0.447 (0.083) — 0.428 (0.077) 1985 0.557 (0.110) 0.632 (0.084) 0.682 (0.116) 0.635 (0.095) 1986 0.219 (0.042) 0.426 (0.078) 0.548 (0.126) 0.540 (0.098) 1987 0.453 (0.093) 0.838 (0.143) 0.544 (0.103) 0.790 (0.170) 1989 0.234 (0.062) 0.285 (0.098) 0.262 (0.137) — C\ 1984- -1987 35.4 32.6 18.9 19.2 % decline 1987- -1989 48.3 66.0 51.8 — to the same extent as do sightings and vt)cahza- tions (cf. C. Smith 1968, Pauls 1978, Ferron et al. 1986). These expectations were corroborated b\' our analysis. Middens also have a direct tie to management of resources, such as bears, that are of common concern in this zone. Densities of active middens in our study area averaged between 0.2 and 0.8 ha~^ and on both study sites were lowest during 1989, following the drought and wildfires of 1988 (Table 1). Although annual variation tended to be greater on the Mt. Washburn site compared to the Cooke Cit>' site, this difference was not statistically significant (d.f = 4/4, F = 1.31, P > 0.5). Both sites exhibited similar annual pat- terns of variation, including relatively low den- sities during 1984 and 1986 and a substantial decline in active midden densities between 1987 and 1989. Acknowledgments This study was funded by the U.S. National Park Service through the Interagency Grizzly Bear Study Team. M. Hubbard, D. Campopi- ano, D. Dunbar III, and G. Green provided invaluable field assistance. We also thank L. S. Mills, J. Peek, R. G. Wright, D. Johnson, C. Smith, C. Benkman, and an anonymous re- viewer for their thoughtful reviews of this manuscript and its earlier versions. Literature Gited BuRNHAM, K. E, D. R. Anderson, and J. L. Laake. 1980. Estimation of density fiom line transect sampling of biologiciil populations. Wildlife Monographs 72; 1-202. Ferron, J., J. R Ouellet, and Y. Lemay. 1986. Spring and summer time budgets and feeding behaviom- of tlie red squinel {Tainia.sciiirws liud.sonicits). Canadian Jom- nal of Zoology 64: 385-.391. FiNLEY, R. B., Jr. 1969. Cone caches and middens (A'Taini- ascitirus in the Rocky Mountain region. Uni\ersit> of Kansas Museum of Natural Ilistoiy Miscellaneous Publications 51: 233-273. KiLHAM, L. 1954. Territorial beha\iour of red scjuirrel. Journal of Mammalogx' 35: 252-253. Mattson, D. J., and D. E Reinhart. 1994. Bear use of whitebark pine seeds in Noitli America. Pages 212-220 in W. C. Schmidt and F-K. Holtmeier, compilers, Proceedings — International Workshop on Subalpine Stone Pines and Their Environments: the Status of Oiu' Knowledge. U.S. Forest Sendee, General Tech- nical Report INT-GTR-309. Pauls, R. W. 1978. Behavioural strategies rele\ant to the energy economy of the red squinel {Tamiasciurus huchonicus). Canadian Journal of Zoolog\- 56: 1519-1525. Smith, C. C. 1968. The adaptive nature of social organiza- tion in the genus of tree squinels Tainiasciiinus. Eco- logical Monographs 38: 31-63. Smith, M. C. 1968. Red squinel responses to spnice cone failure in interior Alaska. Journal of Wildlife Manage- ment 32: 305-317. Vahle, J. R., AND D. R. Patton. 1983. Red scjuinel co\er requirements in Arizona mixed conifer forests. Jour- nal of Forestiy 81: 14-15, 22. Wea\'ER, T. 1990. Climates of subalpine pine woodlands. Pages 72-79 in W. C. Schmidt and K. J. McDonald, compilers, Proceedings — S\mposium on Whitebark Pine Ecosystems: Ecology and Management of a High-mountain Resource. U.S. Forest Sei"vice, Gen- eral Technical Report INT-270. Wolff, J. O., and J. C. Zasada. 1975. Red squirrel re- sponse to clearcut and shelteiAvood systems in interior Alaska. U.S. Forest Service, Research Note PNW- 255. Received 9 October 1995 Accepted 8 May 1996 Great Basin Naturalist 56(3), © 1996, pp. 276-278 THERMAL CHARACTERISTICS OF MOUNTAIN LION DENS Vernon C. Bleichl-, Becky M. Piercel-2, Jeffiey L. Davisl, and Vicki L. Davisl Abstract. — We used radiotelemetn,^ and searched with a trained hound to locate the dens of 3 recentK- parturient mountain hons [Felis concolor). These dens were located in dense riparian vegetation along the same stream in die bot- tom ot a steep canyon. We monitored the circadian temperatin-es of 2 dens at 1-h intenals and compared tliem to ambient temperatures recorded simultaneously. We found mountain lion dens to effectively moderate high ambient temperatures, but these dens failed to pro\ide a themial advantage at the lowest ambient temperatures recorded in this in\estigation. We conclude that mountain lion dens pro\ide effecti\ e protection fi-om thermal maxima for \oung, immobile kittens. Key words: Felis concolor. mountain lion, temperature. California, den. behavior Female mountain lions (Fclis concolor) select protected locations in which to bear young (Shaw 1989:7, Beier et al. 1995), but little infor- mation is available on den site characteristics for this elusive felid. Here, we describe some characteristics of 3 dens used by different females and their litters and quantify' the ther- mal characteristics of 2 of those dens. Description of Study Area Our study area is located in Mono Co., Cali- fornia, approximatelv 35 km NW of Bishop (118°25'W, 37°20'N), Inyo Co., California. This area is on the western edge of the Great Basin, immediately east of the crest of the Sierra Nevada. The dominant x'egetation t\pe in the general area is sagebrush {A)ie)nisia tri- dentata) scrub with pinyon pine {Finns niono- phijUa) forest at higher elevations. Dense vege- tation, dominated b>' willows {Salix spp.) and wild rose {Rosa spp.), occurs along the major water courses. Methods During August and September 1994 and 1995, telemetry indicated that several adult females in our investigation of mountain lion ecolog)' had restricted their daiK movements. These females returned repeatedh' to the same locations, suggesting that they had established natal dens (Beier et al. 1995). We searched these 3 areas and, after detecting vocalizations of neonatal mountain lions, we used a trained hound (Bruce 1918) to locate the dens and kit- tens. We estimated the ages of these kittens according to criteria summarized by Anderson (1983:43) and Currier (1983). We examined the thermal characteristics of the dens b>' placing a recording thennograph (model RTM, R\'an Instruments, Inc., Kirk- land, WA) on the floor of each den and an identical instrument on the ground < 100 m away, on a north exposure supporting sage- brush and pinyon pine. Because of the shinibs and trees present on these north-facing slopes, thermographs were not exposed directK* to the sun for most of each dax; Hourlx' temperatures were recorded at den 2 fiom 4 September to 4 October 1994, and at den 3 from 11 August to 16 September 1995; we did not haxe access to thermographs during the period that den 1 was actixe. We made ocular estimates of tree height and canop)' closure, as well as horizontal cover, at each den. We used analysis of xariance and anahsis of covariance to explore the effects of da\' and time on temperature, simple linear regression to examine the relationship between da>" of the study and daiK' temperature, and t tests to compare den temperatures with ambient tem- peratures (Zar 1984). Results Three dens containing kittens were located along the Owens Ri\ er: den 1 contained 1 male and 1 female; den 2 contained 3 males and 1 female; den 3 contained 2 males and 1 female. IC.ililoTnunrpartiiic-iilorKisharKlCainr, tOTW. l.iiK-St,, Hi>.li(.p, CA 9:5511. -histitvitc <)( Antic liii)l,ii;\ and D.parlini-iil cif Hii)l(ii;> and Wildlilf. liiiver.sity of Alaska, Fairbanks. AK 99775. 276 1996] Mountain Lion Di:ns 277 We estimated tlie kittens at dens 1 and 3 to he < 20 da\s of age, and those at den 2 to he < 10 dax's old. All 3 dens were located in dense groves of willows that ranged in height to approxiniateK 4 ni. Wild rose was ahnndant at all 3 sites, and each den was located < 50 ni from the river. Canop\ closme at each den was nearly 100%, and direct simlight did not reach the suhstrate during am of our midda\' visits {n = 2, den 1; ;i = 5, den 2; n = 3, den 3). Horizontal coxer at each location was sufficientK dense that, e\'en while standing, we were totally obscured from each other s xiew at < 3 m. The substrate of all 3 dens was littered with deciduous leaves as well as tree trunks, branches, twigs, and bark. We were able to reach the kittens only b\' craw ling into the dense vegetation present at each site. We found significant differences between the dens in mean ambient temperature (-t j^,^ 9 = 13.18 ± 9.94 [sY C; .vj^,, 3 = 20.47 ± 10.61° C; F = 202.584, df = 1, 1630, P < 0.001), mean den temperature (-V^jg„ 0 = 6.01 ± 5.77° C; .va^„ 3 = 15.22 ± 7.08° C; F = 807.949, df = 1, 1630, P < 0.001), and mean dail>' tempera- ture differential (ambient temperature - den temperature; .v j^,^ .7 = 7.16 ± 6.05° C; .v^j^,-, 3 = 5.25 ± 6.09° C; F = 40.224, df = 1, 1630, P < 0.001). At den 2, there was a significant effect of day on ambient temperature (F = 3.814, df = 30, 713, P < 0.001), den temperature (F = 3.191, df = 30, 713, P < 0.001), and tempera- ture differential (F = 4.320, df = 30, 713, P < 0.001). At den 3, however, there was no such effect on ambient temperature (F = 0.421, df = 36, 851, P = 0.999), den temperature (F = 0.535, df = 36, 851, P =0.989), or temperature differential (F = 0.488, df = 36, 851, P = 0.995). As the study progressed, there was a significant decline at den 2 in ambient temper- ature (r = -0.340, P < 0.001), den temperature (r = -0.112, P < 0.001), and temperature dif- ferential (r = -0.228, P < 0.001); lesser de- clines in ambient temperature (r = -0.100, P < 0.001), den temperature (r = -0.051, P = 0.001), and temperature differential (r = -0.048, P < 0.001) occun-ed at den 3. At both dens, there was significant diel vari- ation in ambient temperature (den 2: F = 103.382, df = 23, 720, P < 0.001; den 3: F = 618.443, df = 23, 864) and den temperature (den 2: F = 91.008, df = 23, 720, P < 0.001; den 3: F = 431.275, df = 23. 864, P < 0.001). When date was used as a covariate to control lor dail\ solar radiation, the mean temperatine differential also \aried on an hourly basis at both dens (den 2: F = 112.271, df = 23, 719, P < 0.001; den 3: F = 329.936, df = 23, 863, P < 0.001). HourK' ambient temperatures were greater than corresponding den temperatures at both locations (den 2: t = 32.285, df = 743, P < 0.001; den 3: t = 25.662, df = 887, P < 0.001); this difference was especially pro- nounced at high ambient temperatures (>31° C [HAT]; Fig. 1). At HAT, the temperature dif- ferential at den 2 (x = 21.92 ± 4.49° C) was >3 times that at moderate ambient tempera- tures (< 31° C [MAT]; X = 6.03 ± 4.46° C), and the temperature differential at den 3 {x = 13.56 ± 8.37° C) at HAT was nearly 5 times that at MAT (x = 2.87 ± 3.50° C). At den 2, the mean range of daih' ambient temperatures (x = 28.96 ± 7.81° C) was nearly double that of daily den temperatures (x = 15.79 ± 5.26° C) (/ = 15.83, df = 30, P < 0.001). Similarly at den 3 the mean range of daily ambient temper- atures (.Y = 32.54 ± 3.71° C) was >1.5 times that of daiK' den temperatures (x = 20.89 ± 3.50° C; t = 15.24, df = 36, P < 0.001). For both locations combined, den temperatures were less than ambient temperatures for all but 2 (<0.2%) of the paired hourly obsenations. Discussion These mountain lion dens effectiveh' mod- erated high ambient temperatures, consistent with the hypothesis of Shaw (1989) that dens play an important role in protecting young, defenseless kittens from thermal maxima. At HAT, mean temperature differentials were 3-5 times greater than at MAT (Fig. 1). There were significant effects of time of day (both dens) and da)' length (den 2) on temperature differ- ential and, hence, the moderating influence of the dens. Nevertheless, den temperatures were less variable than were ambient tempera- tures. We found no evidence that these dens provided a thermal adxantage (i.e., den tem- peratures greater than ambient temperatures) at the minimum ambient temperatures we recorded; dens may, however, provide protec- tion for kittens when temperatures fall below those that we encountered. Few descriptions of mountain lion dens are available, but females max- select caves, rocky areas, or dense thickets in which to bear young 278 Great Basin Naturalist [Volume 56 10 20 30 40 AMBIENT TEMPERATURE (C) 50 Fig. 1. Mean temperature differential (ambient - den) is more than 3 times greater at high ambient temperatures (>31° C) than at moderate ambient temperatures (<31° C; data fi^om dens 2 and 3 combined, F — I, 241.07, df = 1, 1630, P < 0.001). Mountain Hon dens in dense vegetation effectively moderate extreme high temperatures and afford young, helpless kittens protection from ambient ma.xima, consistent with the hypothesis of Shaw (1989). (Bruce 1918, Young and Goldman 1946, McBride 1976, Russell 1978, Shaw 1989). We hv^podie- size that thermal characteristics var\' among types of dens, and that mountain lions inhabit- ing particular environments select den sites based, in part, on the thermal advantage(s) they provide. In an area with a warm, Mediterranean cli- mate, Beier et al. (1995) reported 2 dens that were located in a small canyon with very heavy cover of "brush," similar to those we investi- gated. Dens located in thick, woody vegetation may conceal young that are vulnerable to pre- dation, but they also provide protection for kit- tens from extreme temperatures associated with direct insolation. Such locations provide important thermal benefits for kittens at high ambient temperatures, and a more stable ther- mal environment than exists outside the den throughout the range of ambient temperatures we recorded. Movements bv kittens for ther- moregulator)' puiposes miglit be lessened under these circumstances. Fewer movements by kit- tens may decrease the probabilit\' of discovery by potential predators, thereb>' enhancing the survival of young, defenseless mountain lions. Acknowledgments We thank S. Parmenter, D. Becker, and C. Milliron for lending us the recording themio- graphs, N. G. Andrew and E Beier for critical comments on an early draft of the manuscript, and M. W Oehler Sr, for assistance in the field. Financial support was provided by the Mule Deer Foundation, University of California White Mountain Research Station, CalifoiTiia Depart- ment of Fish and Game (CDFG), National Rifle Association, Safari Club International, and the Fish and Game Advisoiy Committees of Inyo and Mono counties. This is a contribution fi-om the CDFG Deer Herd Management Plan Implementation Program. Literature Cited Anderson, A. E. 1983. A critical review of literature on puma {Felis concolor). Colorado Division of Wildlife Special Report 54: 1-91. Beier, R, D. Choate, and R. H. Barretf. 1995. Move- ment patterns of mountain lions during different behaviors. Journal of Mammalog\' 76; 1056-1070. Bruce, J. C. 1918. Lioness tracked to lair. California Fish and Game 4: 1.52-153. Currier, M. J. E 1983. Felis concolor Mammalian Species 200; 1-7. McBride, R. T. 1976. The status and ecologx- of the moun- tain lion Felis concolor stanleijana of the Te.\as-Mex- ico border. Unpublished master s thesis, Sul Ross State University, i\]pine, TX. Russell, K. R. 1978. Mountain lion. Pages 207-225 in J. L. Schmidt and D. L. Gilbert, editors. Big game of North America: ecologx' and management. Stackpole Books, Harrisburg, PA. Sh.w, H. 1989. Soul among lions. Johnson Publisliing Com- pany, Boulder, CO. Young, S. R, and E. A. Goldman. 1946. The puma, myste- rious American cat. American Wildlife Institute, Wash- ington, DC. Zar, J. H. 1984. Biostatistic;iI an;Jysis. Prentice-Hall, Engle- wood Cliffs, NJ. Received 30 December 1995 Accepted 5 April 1996 Great Basin Naturalist 56(3), © 1996, pp. 279-2S0 JAMES WILLIAM BEE 1913-1996 W'iliner \\. TaniUT^ Jiunt's W. Bee James W. Bee, professor of zoolog)' and emer- itus, University of Kansas, Lawrence, Kansas, died at Seatde', Washington, 18 April 1996. He was bora 25 September 1913 in Provo, Utah. His family, including parents, Robert G. and Mary Culbertson Bee, and brother and sister. Max and Mary, were residents of Provo, Utah, where they receixed their earK' education. It was fiom this setting in Utah Valley that James was introduced at an early age to the sciences of archaeolog)' and ornithology by his father, who loved natural histoiy and the little-known histoiy of Utah Valley, its lake, and its early inhabitants. As a youth and young man, he accompanied his father on man\' collecting trips that resulted in assembling artifacts of the past. These archaeological finds provided valuable infonua- tion pertaining lo IncUan winter camps, sum- mer camping areas, and burial groimds, and an insight into tlie role of Utah Lake and the sur- roimding mountains as providers of aliundant fish and game. Their travels near this lake and in the moun- tains brought them in contact with numerous birds. Each spring flocks of birds entered the valley — some remained and others moved on. This phenomenon stimulated a great interest, so much so that James, i.is father and various friends became amature ornithologists. Their ornithological work encompassed life histoiy studies, obsei^vation of arrivals in the spring, and investigation of nests and nesting. Ulti- mately, this interest in birds lead to the assem- bly and preparation of eggs for those species nesting in the valleys and mountains of central Utah. Thus was bora a naturalist whose contribu- tions are invaluable and most of which could not now be assembled. The archaeological col- lections are presently at the Museum of Peo- ples and Cultures, Brigham Young University. James and his father contributed 812 sets of bird eggs and 1 12 single eggs representing 234 species. James contributed 7918 mammal, 245 bird, and 504 amphibian and reptile specimens to the M. L. Bean Life Science Museum, also at Brigliam Young University'. In die Bean Museum Libraiy are field records, 27 volumes from James and 20 xolumes of his father s, all well documented and done with great care. Tliese were written in the field as the data were obtained and represent field records of a time when some pristine conditions still existed. James entered Brigham Young Unixersity in 1932 and received his B.A. degree in 1937. World War II interrupted his studies for the M.A., but this he finished in 1947. As an under- graduate, he became interested in and re- searched mammals. Thus his master's research 'M. L. Bi'.m Museum, Brigham Young Universih, Provo, UT S4602. 279 280 Great Basin Naturalist [Volume 56 was the mammals of Utah County. While in the Armed Forces (1941-1946), he was trained as a hospital administrator and sewed as a sergeant- major, organizing 50 key men as a cadre to establish a new hospital. He supei-vised several new hospital departments and for a year and a half sei-ved in field hospitals for airborne units in India, Assam, Burma, and China. During these years he met Annette E Malseed, R.N. They were married 15 October 1945 in Kun- ming, Yeaman, China. In September 1948 James entered the Uni- versity of Kansas to continue his research in mammalogy, with a desire to complete his study on the genus Microtis. He completed his studies at KU and spent a summer at Friday Harbor, Washington. He was a noted field zoologist and spent many years collecting re- search material and field data for the Museum of Natural Histoiy at the University of Kansas. Students doing research in vertebrate zoology at Brigham Young University or at the Univer- sity of Kansas will find numerous specimen tags labeled "collected b>' James Bee." After 37 years he retired from KU and built a new home on Lopez Island, Washington. James and Annette were the parents of three children: James Robert, Annette Christine Kenagy, and Mary Pauline Bee Kaufman. It was my pleasure to have spent several summer field trips with James. A highlight was the summer of 1939 when we studied the ver- tebrates of western Utah County. During this time we prepared and assembled museum specimens; of importance to me was finding a nesting colony of the western skink and secur- ing additional specimens of Hypsigalena. We both participated in the new discoveries, and it was obvious that Jim was at his best in prepar- ing precise field data. I learned much from him that sinnmer and appreciated his dedication to a complete understanding of the natinal world we were investigating. James had a very likeable personalit) that was reflected in his family, which he held in high esteem. Publications Bee, James W. 1957 Biological sui'vey of the Virgin Islands National Park. Procedures, obsei^vations and recommen- dations. 33 pages, 2 maps. Submitted to National Park Senice. 1958 Birds foimd in the Aictic Slope of noitliem Alaska. Universitv' of Kansas Publications, Museum of Natiu-al History 10(5): 163-211. 2 plates, 1 figure in text. 1970 Vasco M. Tanner — a diversified career. Great Basin Naturalist 30: 216-217. 1994 Rough Grouse siting. The Trumpeter, San Juan Audubon Societ>' 14(4):3. Bee, James W., Dumitru Murariu, and Robert S. hoffmana. 1980 Histolog\' and histochemistiy of specified integu- mentaiy glands in eight species of North Ameri- can shrews (Mammalia, Insectivora). Travau.v du Museum d' Histoire Katurelle "Grigor Antipas " (Bucharest, Romania) 22: 547-569. Bee, James W, and E. R. Hall 1951 An instance of coyote-dog hybridization. Trans- actions of the Kansas Academy of Science 54(1): 73-77. 4 figures. 1 table. 1954 Occurrence of the harbor porpoise at Point Barrow, Alaska. Journal of Mammalog\- 35(1): 122-123. 1956 Mammals of northern Alaska on the Aictic Slope. University of Kansas, Museum of Natural His- tory, Miscellanceous Publications 8: 1-309. 4 plates, 122 figures, 51 tables. 1960 The red fig-eating bat, Steuoderma rufum Des- marert, found alive in the West Indies. Mam- malia, Museum d'Histoire Naturelle, Paris 14(1): 67-75. 2 figures in text. Bee, James W, and Howard Le\ enson Bald Eagle use of Kansas Ri\er riparian habitat in northeastern Kansas. Kansas Ornithological Society Bulletin 3(4): 28-33. Great Basin Naturalist 56(3). © 1996, i^p. 281-282 BROOK STICKLEBACK (CULAEA INCONSTANS [KIRTLAND 1841]), A NEW ADDITION TO THE UPPER COLORADO RIVER BASIN FISH FAUNA TiTiiotln Moddc' and (I. Bruce llaiucs' Key words: hrook stichlclHich. raiiiic cxtciisioii. noiiiialirc. Brook stickleback {Ciilaea incoii.stdiis) is a small gasterosteid fish native to Arctic and Atlantic drainages in North America. The species natixe range extends west from Nova Scotia to British Columbia and south horn the Northwest Territories to southern Ohio drain- ages, including the Mississippi-Missouri River above the confluence of the Illinois River (Scott and Grossman 1973). llubbs and Lagler (1958) reported brook stickleback from the Illinois River in Illinois and the Missouri River in Kansas. Historical accounts exist of relictual populations in the Platte River system, but Cross (1967) noted its absence from Kansas. An isolated, and presumably relict, population occurs in the Canadian River drainage of New Mexico (Koster 1957). Brook stickleback has been collected outside its native range in Alabama (Boschung 1992), Kentucky (Burr and Warren 1986), Tennessee (Etnier and Starnes 1993), the Rio Crande River drainage in New Mexico (Sublette et al. 1990), Colorado (Zuck- erman and Behnke 1986), and the Klamath River, California (Peter Moyle, University of California, Davis, personal communication). Between luly and October 1995 we col- lected 5 brook stickleback from the middle Green River, Uintah County, Utah, the 1st rec- ord for the species in Utah (catalog number LFL 24871, Lai-val Fish Laboratory, Colorado State University). Brook stickleback was first reported elsewhere in the upper Colorado River drainage in 3 small tributaries of the Elk River (South, Coleman, and Deep creeks) in northwestern Colorado in 1993 (Jake Bennett, Colorado Division of Wildlife, personal com- munication). One brook stickleback juvenile, 27 mm total length (TL), was collected 18 July 1995 in a (juatrefoil light trap at the outlet of Old Charley Wash, river kilometer (RK) 402 on the Green River (RK measured from the conflu- ence of the Green and Coloiado rivers). Four adult fish, (41, 46, 48, 54 mm TL) were col- lected between 1 October and 12 October 1995 from Old Charley Wash, a wetland on the Oura>' National Wildlife Refuge that connects to the Green River during high spring flows. Fish were collected when the wetland was drained (Modde in press); all were found in low or no velocity habitats. Tyus et al. (1982) cited the establishment of 42 nonnative fishes in the upper Colorado River compared to 13 native species. Brook stickle- back is an additional transplanted species, prob- ably the result of human introduction rather than a natural range extension. Brook stickle- back introductions elsewhere in the United States were presumably through bait bucket transfers or contaminated game fish stockings (Zuckerman and Behnke 1986, Sublette et al. 1990). Literature Cited Boschung, H. T. 1992. Catalogue of freshwater and marine fishes of Alabama. Alabama Museum of Natural His- toi-v' Bulletin 14:1-266. Burr, B. M., and M. L. Warren, Jr.. 1986. A distrilni- tional atlas of Kentucky' fishes. Kentucky Nature Pre- serves Commission Scientific and Technical .Series, No. 4. Cross, E B. 1967. Handbook of fishes of Kansas. State Bio- logical Sun'ey and the University' of Kansas Museum of Natural History, Lawi-ence. Etnier, D. A., and W. C. Starnes. 1993. The fishes of Ten- nessee. The University of Tennessee Press, Kno.wille. HUBBS, C. L., and K. E L.'\gler. 1958. Eishes of the Creat Lakes region. The Universit\- of Michigan Press, Ann Arbor 'Colorado River Fish Project, U.S. Fish and Wildlife Service. 266 West 100 North, Suite #2, Vernal, UT 84078. 281 282 Great Basin Natur.\list [Volume 56 KOSTER, W. J. 1957. Guide to the fishes of New Mexico. University of New Mexico Press, Albuquerque. MODDE, T. In press. Juvenile razorback sucker in a man- aged wetland adjacent to the Green River. Great Basin Naturalist. Scott, W. B., .\nd E. J. Crossm.an. 1973. Fishes of Canada. Fisheries Research Board of Canada Bulletin 184. Sublette, J. E., M. D. Hatch, and M. Sublette. 1990. The fishes of New Mexico. University- of New Mex- ico Press, Albuquerque. Tvus, H. M., B. D. BuRDiCK, R. A. Valdez, C. M. H.aynes, T. A. Lytle, and C. R. Berry. 1982. Fishes of the Upper Colorado River Basin: distiibution, abundance. and status. Pages 12-70 in W. H. Miller, H. M. Tyus, and C. A. Carlson, editors. Fishes of the Upper Col- orado River System: present and future. American Fisheries Society, Bethesda, MD. ZUCKER\L\N, L. D., AND R. J. Behnke. 1986. Introduced fish in the San Luis Valley, Colorado. Pages 435^52 171 R. H. Stroud, editor Role offish culture in fishery management. American Fisheries Societv, Bethesda, MD. Received 4 January 1996 Accepted 12 April 1996 Great Basin Naturalist 56(3), © 1996, p. 282 ERRATA Correction to: Sutherland, Steven D., and Robert K. Vickeiy, Jr. 1993. On the relative importance of flower color, shape, and nectar rewards in attracting pollinators to Mimuhis. Great Basin Naturahst 53: 107-117. The article states: "Hummingbirds are com- monly said to have evolved a preference for red or orange-red flowers, " citing (1) K. A. Grant, 1966, A h\q3othesis concerning tlie preva- lence of red coloration in California humming- bird flowers, American Naturalist 100: 85-98; and (2) K. A. Grant and V. Grant, 1968, Hum- mingbirds and their flowers, Columbia Uni- versity Press, New York, 115 pp., among other references. In fact, the Grants point out just the opposite, i.e., that experimental investiga- tion shows that hummingbirds ha\ e not evolved a preference for red or any other color. Actu- ally, Sutherland and Vickeiy's article comes to this conclusion also. Great Basin Naturalist 56(3), O 1996, pp. 283-285 BOOK REVIEW Snakes of Utah. Douglas C. Cox and Wilmcr W. Tanner; Mark Fhilbrick, photography. Monte L. Bean Life Science Museum, Brig- ham Young Universit), Provo, UT. 1996. $17.95 softcover. Snakes of Utah, anticipated for some time, is finalK' avaikible for distribution. This book- let (92 total pages) includes all known species and subspecies of snakes found in the state, with brief descriptions, habits, and habitats, along with colored photographs of each. While most people will likeK shudder at the thought of snakes, especially while viewing photographs, the enthusiast will recognize the value of the illustrations and other published information. Generalh; the booklet is written in nonscien- tific language, but it also includes some scien- tific notations. For instance, scientific names and autliorities of the 33 species and subspecies, along with common names, are included for each. Of interest (perhaps only to the special- ist) is the fact that only 2 binomials are found among all Utah snakes; 31 are trinomials. It might be concluded that, because of subspeci- ation, onh' 27 kinds of snakes are found in Utah. To the general public, a night snake is a night snake, a garter snake is a garter snake, and a rattlesnake is a rattlesnake. Heipetologists have named subspecies for practically all snakes, compounding one's knowledge of these animals. Technically, where closely related subspecies show sympatric distribution, there should be intergradation between the 2 t\'pes. Most indi- viduals using this booklet will probably not recognize differences between related sub- species found especially in these sympatric regions. If intergrades are not present, then these should be elevated to species and not kept as subspecies. Little infomiation is found in the booklet on intergradation of characteristics. An important contribution of this booklet is the colored photographs. While not captioned, most photographs are obxious because they are shown on the page opposite the name and other information on that snake. This publication would be more useful if a caption were shown by the other photographs throughout the text, e.g., the photo opposite page 1 and those shown on pages 3, 4, 5, 8. The herpetologist will probably recognize these without caption, but, as stated, it's likely these specialists will not be the primaiy users of the te.xt. Identification of snakes by these photographs may not be obvi- ous to most readers. Most photos show colors and patterns of snakes, but a few, such as the full view of the Upper Basin garter snake on page 59, do not show these identifiable fea- tures. It's interesting that the only snake not represented by a photo of the entire body is the Sonoran lyre snake on page 67. One won- ders why. Perhaps it's because this snake is "considered to be rare. " However, the Dixie College Natural Science Museum contains records of 7 specimens, 2 having been found in what is now considered "downtown" St. George, 1 specimen as recently as 1980. It seems likely that widi a little eflFort, one of diese "rai'c " snakes might have been found. The photo of the Utah blind snake on page 17 is a surprise. Of the several dozen blind snakes observed by this writer, representing localities from the Red Cliffs Recreation Area near Leeds, Washing- ton County, to the extreme northwest corner of Arizona, not 1 specimen even approached this dark phase. They have all been a pale tan color, frequently showing a suffusion of pink. Another important contribution of this book- let is the distribution maps included with each species along with the general and sometimes specific distribution of the snake within the state. While it is difficult to show accurac\- on a small map, some maps are erroneous. For instance, the distribution of the Painted Desert glossy snake is "in the extreme southeastern sector of the state, adjacent to northeastern Arizona" (page 40). The map, however, shows it is found more south central than southeastern. An inconsistency from text to map is also obsen^ed with the California king snake (page 46). If this snake occurs "from the southwest corner east to the Colorado River," wh\' does 283 284 Great Basin Naturalist [Volume 56 the distributional map extend considerabh' beyond the Colorado Ri\ er along the San Juan River? Nothing in the text is speculative of a range extension. The maps of the Utah moun- tain king snake (page 48) and the Utah milk snake (page 50) do not accurate!)' depict their known distributions in Washington Count)'. On page 60, of the western blackneck garter snake, the text states "its northernmost habitat is associated with streams ... in the regions of southeastern Utah. The map shows its distri- bution into east central Utah. Reference is made to a ground snake having been collected in Carbon Count)', far from its known range, and this area is shown on the map. Might this specimen have been one that escaped or was released from captivity? (Reports haxe been made of indixiduals transporting this snake from the St. George area, where it is common, to elsewhere in the state.) There is speculation that the Utah blackhead snake "may occur fur- ther north in Emery and Carbon Counties." (The proposed expansion is not shown on the map.) Wh)' might it not, then, be found in Wa)'ne Coimt)' and perhaps even San Juan and Grand counties? If the midget faded rattle- snake is found at Flaming Gorge, why does the map not show distribution in that area? While it would add to the length of the text, it ^vould ha\e been better had the authors given complete distribution ranges for all species and subspecies, rather than just a few. A snake doesn t recognize a political boundan as being its limits! However, it could l)e reasoned, if the distribution extends to the Utah boundaiy the occurrence of that snake would also be in the neighboring state. The full-page map of the state of Utah (page 11) is a good addition to the text. However, with the number of snakes found only in Utah s Mojave Desert, this feature might ha\'e been identified along with the others. In the geo- graphical and ecological descriptions of Utah (pages 9-10), considerable discussion is gi\'en about montane regions, some at high ekna- tions, yet little is written about the low, hot desert or the higher, cold desert, although the authors admit to the richness of reptile fauna, especially in the low, hot desert, the south- western region of the state. In addition to these other features. Snakes of Utah includes both glossar)-, though not inclusive of all technical words used in the text, and index. The writer wonders at the importance of the full page of illustrations (page 13) showing scalation witli so little reference to most of these features in descriptions. Some of these features are referenced; most are not. While full pages of color separate groups of snakes, does this mean that Joshua trees are characteristic of the distribution of the Utah blind snake? Although the illustration on page 18 may be t) pical of the habitat of the rubber boa in Utah, and on page 72 of the habitat of some of the rattlesnakes, does the illustration on page 22 depict the typical distribution of the colubrids? Perhaps these "division pages" were added merely for color; nevertheless, they are attractive. The authors of the booklet include a number of interesting anthropomorphisms, perhaps intentionally. Some of these are noted: (1) In the introduction, the statement is made {page 5) that "the snake employs rocks and biaish to snag the skin and hold it while the snake crawls out. One wonders if the snake does this inten- tionall)'. (2) "Denning is a behavior pattern that provides the snake with an opportunity to come in contact with other snakes of the same species" (page 6). (3) Of the rubber boa, "it will often cling like a bracelet and seem to enjoy it as much as the person" (page 20). (4) The statement is made about the western )'ellow- belly racer (page 28) that "it will attempt to bite if it feels at all threatened." (5) Another exam- ple is that rattlesnakes use the rattle "as a warning de\'ice to intimidate other animals that may harm the snake (page 75). Miscellaneous errors or inconsistencies in narratix e, grammatical or othenvise, are found. The introduction, for instance, discusses tall tales and folklore of the American West. This booklet is, of course, about snakes of one region of the American West, but tall tales and folk- lore— even some of the same stories heard in the American West — are repeated wherexer snakes are found. On pages 4 and 5 the statement is made tliat "the mouth is the most unixersally used weapon emplo)'ed by snakes in self-defense." The emphasis is obvious because the accom- pan)ing text is about self-defense, but snakes use their mouths more often as a means of obtaining food. Also, in the introduction, the statement is made that "these studies and our inu.sciiDi i)ro''s Monte L. Bean Life Science Museum. The complete identification of the museum should ha\'e been made when it was lirst referenced on page 6. It could he pointed out, too, that other schools and nmseums might ha\'e the same purpose — to "lielp them to understand about snakes. While the following is not necessarily in error, it reflects a writing style. On page 12 the following statements are made: "These snakes do not pose any threat to man but they do pro- \ ide a mild venom to help immobilize their prey. Their prey includes worms, insects, frogs, lizards, and small mammals. " In writing, re- peated words and phrases should be avoided in consecutive sentences or within the same sentence. It could better have been written, "to help immobilize their prey, which includes worms, insects. . . . " In the introduction to the tropical wormlike snakes, the statement is made that "they feed on insects and worms, especially termites and ants, found in the soil." The emphasis in this statement suggests that termites and ants are kinds of wonns. This should have been written, "they feed on worms and insects, especially termites and ants. " In reference to the Utah blind snake the statement is made (also on page 15) that Vasco M. Tanner "had seven specimens to examine, and the name is based on No. 662 in the BYU type collection." Name is inappropriately used, although specimen No. 662 might have been published as the t\"pe specimen. One of the most frequently made grammat- ical errors in writing is the inconsistency of singulars and plurals within a sentence. On page 20, this type of error is made. The rubber boa "is a delightful animal to have around their wrist." Inasmuch as their is plural, the plurality ofivhsts must also be used. Reference is made twice (on pages 30 and 44) that the snakes occur on "the margins of deciduous forests." Small groups of deciduous trees may occur in riparian areas or where trees are cultivated, but technically, deciduous forests do not occur in the state of Utah. The redundant statement is made about the western leafnose snake that the rostral scale "looks leaflike. ' An inconsistency is noted about the Utah mountain king snake and the Utah milk snake. Page 48 states: "if a specimen has a white nose, it is most likely a mountain king snake. If, how- ever, it has a black nose, it is probably a milk snake. These characteristics are not completely reliable" (emphasis added). Page 50 states that "the milk snake differs in that it has a black nose." On pages 68 and 70 the habits ol the Mesa Verde night snake and the desert night snake are described as "nocturnal, secretive, and sel- dom seen." Furthermore, it is stated that the former "feeds primarily on the lizard Uta stans- huriana imiformis and other small lizards," while the latter "feeds primarily on the side- blotched lizard Uta stanshuriami stanshiiriami." One wonders about this inasmuch as lizards are primarily diurnal and snakes nocturnal. Of course, snakes could feed at night while lizards are inactive. While reference is made in the booklet about the influence of soil on the ground color of some snakes, there is no mention of this occur- ring in the Mojave Desert sidewinder (page 78). Of the hundreds of sidewinders obsei^ved by the author in the past 50 years, the influ- ence of soil color on the ground color of the snake is most obvious. Despite these criticisms. Snakes of Utah should contribute importantly to our knowl- edge of these reptiles within a limited political region. As noted, the booklet is written for lay- men, and its distribution is more appropriate in national and state parks and monuments than in the scientific community. It is a "must" for backpackers, individuals, and families spend- ing time in tlie out-of-doors where snakes might be encountered. The authors, tlie photographer, and the publisher are to be commended for finally making this booklet available. Andrew H. Barnum Professor Emeritus Dixie College St. George, UT 84770 The Future of Arid Grasslands: Identifying Issues, Finding Solutions 9-1 3 October 1 996, Tucson, Arizona A solution-oriented conference for everyone interested in the future of grasslands in the American Southwest and northern Mexico. This four-day conference will focus on understanding problems facing those grasslands and practical tools for grass- land management, preservation, and restoration. Attendees will be a mix of private and public land managers and owners, scien- tists, representatives of nonprofit groups, and concerned citi- zens. Two full days will be spent in the field studying examples of grassland management in southern Arizona. The other two days will include keynote speakers and panelists as well as small-group discussion and information sessions. The final day will focus on methods for preservation ranging from coordi- nated monitoring systems, land use, and taxation tools to public involvement techniques. Most of the speakers and panelists will be invited, but abstracts are welcome for a few open sessions dealing with grasslands management, interrelationships between grasslands and humans or wildlife, and specific methods for preser- vation, especially success stories. Researchers are encouraged to submit abstracts for poster sessions, which will be incorporated into the program featuring on-the-ground examples of problem solving to protect or restore grasslands. Both successful and unsuccessful ex- amples are sought to illustrate what has and has not worked — and why. The conference is organized by the Audubon Research Ranch and is co- sponsored by numerous government agencies, educational institutions, and non- profit groups. For more information: Grasslands Conference, Tucson Audubon Society 300 E. University #128 Tucson, AZ 85705 or University of Arizona Water Resources Center (520) 792-9591 INFORMATION FOR AUTHORS The Great Basin Naturalist welcomes previously unpublished manuscripts pertaining to the hiologi- cal natural history of western North America. Preference will he given to concise manuscripts of up to 12,000 words. Simple species lists are dis- couraged. SUBMIT MANUSCRIPTS to Richard W. Baumann, Editor Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young University, Provo, UT 84602-0200. An accompanying cover letter must include phone numher(s) of the author submitting the manuscript, and FAX number and E-mail address when applicable; the letter must also pro- \ ide infomiation describing the extent to which data, text, or illustrations have been used in other papers or books that are published, in press, submitted, or soon to be submitted elsewhere. Authors should adhere to the following guidelines; manuscripts not so prepared may be returned for revision. MANUSCRIPT PREPARATION. In general, the Great Basin Naturalist follows recommendations in Scientific Style and Format: The CBE Manual for Authors, Editors, and Publishers, 6th edition (Council of Biology' Editors, Inc., 11 South LaSalle Street, Suite 1400,' Chicago, IL 60603, USA; phone 312-201-0101; FAX 312-201-0214). We do, however, differ in our treatment of entries in Literature Cited. Authors may consult Vol. 51, No. 2 of this journal for specific instructions on format; these instruc- tions. Guidelines for Manuscripts Submitted TO THE Great Basin Naturalist, are printed at the back of the issue. Also, check the most recent issue of the Great Basin Natwalist for changes. TYPE AND DOUBLE SPACE all materials, including literature cited, table headings, and figure legends. Avoid hyphenated words at the right-hand margins. Underline words to be printed in italics. Use stan- dard bond (22x28 cm), leaving 2.5-cm margins on all sides. SUBMIT 3 COPIES of the manuscript (5 copies of fish manuscripts) and the original on a 5.25- or 3.5- inch disk utilizing WordPerfect 5.1 or above. Number all pages and assemble each copy sepa- rately: title page, abstract and key words, text, acknowledgments, literature cited, appendices, tables, figure legends, figures. TITLE PAGE includes an informative title no longer than 15 words, names and addresses of authors, a running head of fewer than 40 letters and spaces, footnotes to indicate change of address and author to whom correspondence should be addressed if other than the first author. ABSTRACT states the purpose, methods, results, and conclusions of the research. It is followed by 6-12 key words, listed in order of decreasing importance, to be used for indexing. TEXT has centered main headings printed in all capital letters; second-level headings are centered in upper- and lowercase letters; third-level head- ings begin paragraphs. VOUCHER SPECIMENS. Authors are encouraged to designate, properly prepare, label, and deposit high-quality voucher specimens and cultures docu- menting their research in an established permanent collection, and to cite the repositoiy in publication. REFERENCES IN THE TEXT are cited by author and date: e.g., Martin (1989) or (Martin 1989). Multiple citations should be separated by commas and listed in chronological order Use "et al. ' after name of first author for citations having more than two authors. ACKNOWLEDGMENTS, under a centered main heading, include special publication numbers when appropriate. LITERATURE CITED, also under a centered main heading, lists references alphabetically in the fol- lowing formats: Mack, G. D., and L. D. Flake. 1980. Habitat rela- tionships of waterfowl broods on South Dakota stock ponds. Journal of Wildlife Management 44: 695-700. Sousa, W R 1985. Disturbance and patch dynamics on rocky intertidal shores. Pages 101-124 in S. T. A. Pickett and P S. White, eds.. The ecolo- gy of natural disturbance and patch dynamics. Academic Press, New York. Coulson, R. N., and J. A. Witter. 1984. Forest ento- mology: ecology and management. John Wiley and Sons, Inc., New York. 669 pp. TABLES are double spaced on separate sheets and designed to fit the width of either a single column or a page. Use lowercase letters to indicate foot- notes. PHOTOCOPIES OF FIGURES are submitted initially with the manuscript; editors may suggest changes. Lettering on figures should be large enough to withstand reduction to one- or two-column width. Originals must be no larger than 22x28 cm. NOTES. If the manuscript would be more appro- priate as a short communication or note, follow the above instructions but do not include an abstract. A CHARGE of $50 per page is made for articles published; the rate for individual subscribers will be $35 per page. However, manuscripts with com- plex tables and/or numerous half-tones will be assessed an additional charge. Reprints may be pur- chased at the time of publication (an order form is sent with the proofs). FINAL CHECK: • Cover letter explaining any duplication of information and providing phone number(s), FAX number, and E-mail address • 3 copies of the manuscript and WordPerfect diskette • Conformity with instructions • Photocopies of illustrations (ISSN 001 7-3614) GREAT BASIN NATURALIST vc 56 no 3 ju.y 1995 CONTENTS Articles Biogeographic significance of low-elevation records for Neotoma cinerea from the northern Bonneville Basin, Utah Donald K. Grayson, Stephanie D. Livingston, Eric Rickart, and Monson W. Shaver III 191 Synopsis of the mosses of Wyoming EM. Eckel 197 Variation in bitterbrush {Purshia tridentata Pursh) crude protein in southwestern Montana Carl L. Wambolt, W. Wyatt Fraas, and Michael R. Frisina 205 Dam-forming cacti and nitrogen enrichment in a piiion-juniper woodland in noilli- western Arizona Molly Thomas Hysell and Charles C. Crier 211 Distribution and ecological characteristics of Lewisia longipetala (Piper) Clay, a high-altitude endemic plant Anne S. Halford and Robert S. Nowak 225 Larger ectoparasites of the Idaho ground squirrel {Spermophilus brunneus) Eric Yensen, Craig R. Baird, and Paul W. Sherman 237 Roost sites of the silver-haired bat [Lasionycteris noctivagans) in the Black Hills, South Dakota . . . Todd A. Mattson, Steven W. Busldrk, and Nancy L. Stanton 247 Perceptions of Utah alfalfa growers about wildlife damage to their hay crops: implications for managing wildlife on private land Terry A. Messmer and Sue Schroeder 254 Spatial relationships among young Cercocarpus ledifolius (curlleaf mountain mahogany) Brad W. Schultz, Robin J. Tausch, and Paul T Tueller 261 Potential for controlling die spread of Centaurea maculosa with grass competition John L. Lindquist, Bi-uce D. Maxwell, and T. Weaver 267 Indicators of red squirrel {Tamiasciurus hudsonicus) abundance in the whitebark pine zone David J. Mattson and Daniel E Reinhart 272 Thermal characteristics of mountain lion dens Vernon C. Bleich, Becky M. Pierce, Jeffrey L. Davis, and Vicki L. Davis 276 James William Bee, 1913-1996 Wilmer W Tanner 279 Note Brook stickleback {Culaea inconstans [Kirtland 1841]), a new addition to the Upper Colorado River Basin fish fauna Timothy Modde and G. Bruce Haines 281 Errata 282 Book Review Snakes of Utah Douglas C. Cox and Wilmer W Tamier .... Andrew H. Barnimi 283 H E l-IBRARY ^^N Qc ;997 GREAT BASII^ ARD SITV NATURALIST VOLUME 56 N^ 4 — OCTOBER 1996 BRIGHAM YOUNG UNIVERSITY GREAT BASIN NATURALIST Editor Richard W. Baumann 290 MLBM PO Box 20200 Brigham Young University Provo, UT 84602-0200 801-378-5053 FAX 801-378-3733 Assistant Editor Nathan M. Smith 190 MLBM PO Box 26879 Brigham Young University Provo, UT 84602-6879 801-378-6688 E-mail: NMS@HBLL1.BYU.EDU Associate Editors Michael A. Bowers Blandy Experimental Farm, Universit>' of Virginia, Box 175, Boyce, VA 22620 J. R. Callahan Museum of Southwestern Biology, University of New Mexico, Albuquerque, NM Mailing address: Box 3140, Hemet, CA 92546 Jeffrey J. Johansen Department of Biology, John Carroll University University Heights, OH 44118 Boris C. Kondratieff Department of Entomology, Colorado State University, Fort Collins, CO 80523 Paul C. Marsh Center for Environmental Studies, Arizona State University, Tempe, AZ 85287 Stanley D. Smith Department of Biology University of Nevada-Las Vegas Las Vegas, NV 89154-4004 Paul T. Tueller Department of Environmental Resource Sciences University of Nevada-Reno, 1000 Valley Road Reno, NV 89512 Robert C. Whitmore Division of Forestry, Box 6125, West Virginia University, Morgantown, \W 26506-6125 Editorial Board. Jerran T. Flinders, Chairman, Botany and Range Science; Duke S. Rogers, Zoology; Wilford M. Hess, Botany and Range Science; Richard R. Tolman, Zoology All are at Brigham Young University. Ex Officio Editorial Board members include Steven L. Taylor, College of Biology and Agriculture; H. Duane Smith, Director, Monte L. Bean Life Science Museum; Richard W. Baumann, Editor, Great Btisin Naturalist. The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University Unpublished manuscripts that fiirther our biological understanding of the Great Basin and surrounding areas in western North America are accepted for publication. Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1996 are $25 for individual sub- scribers ($30 outside the United States) and $50 for institutions. The price of single issues is $12. All back issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other busi- ness should be directed to the Editor, Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young University Provo, UT 84602-0200. Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin Naturalist through a continuing exchange of scholarly publications should contact the E.xchange Librarian, 6385 HBLL, PO Box 26889, Brigham Young University, Provo, UT 84602-6889. Editorial Production Staff JoAnne Abel Technical Editor Jan Spencer Assistant to the Editor Copyright © 1996 by Brigham Young University Official publication date: 21 Novemher 1996 ISSN 0017-3614 11-96 750 20300 The Great Basin Naturalist Published at Provo, Utah, bv Brk.ham Younc; University ISSN 0017-3614 Volume 56 31 October 1996 No. 4 Great Basin Naturalist 56(4), © 1996, pp. 287-29.3 SPECIES-ENVIRONMENT RELATIONSHIPS AMONG FILTER-FEEDING CADDISFLIES (TRICHOPTERA: HYDROPSYCHIDAE) IN ROCKY MOUNTAIN STREAMS Timothy B. Mihuc'l-, G.Wayne MinshalP, and Janet R. Mihuc^ Abstract. — Species-environment relationships were determined for filter-feeding macroinvertehrates from 55 Rocky Mountain stream sites to estahlish species distribution patterns. Species abundance and 20 environmental vari- ables were measured at each site with species-environment relationships determined using canonical correspondence analysis and stepwise multiple regression. Results suggest that the distribution of several taxa was strongly related to upstream-downstream environmental gradients. Arctopstjche grandis abundance increased with stream size (width and depth) and decreased with increasing turbulence (Reynolds number). Brachijcentrus abundance also increased with stream size (depth). Hijdropsyche abundance increased with increasing baseflow. Parapsyche elsis abundance demon- strated negative correlation with depth, Froude number and conductivity. Ta,xa followed previously reported patterns, partitioning habitat according to stream size. Arctopsyche grandis, Brachyccntrus. and Hydwpsyche were found in larger (3rd- to 6th-order) streams, while Parapsyche elsis was obsei^ved in small headwater (1st- and 2nd-order) streams. Other filter-feeding taxa such as Sirnuliwn, Pisidiuin, and ostracods exhibited little or no apparent habitat partitioning among stream sites. Key words: species-environment relationships, filter feeders. Rocky Mountain streams. Benthic macroinvertebrates adapted for re- Minshall 1990, Richardson and xMackay 1991). moving particles fi-om suspension (filter feeders) Many studies have determined filterer associa- are an important component of stream commu- tions with food resources and environmental nities. Distribution patterns and habitat associ- factors such as water velocity or temperature ations among filterers have been well docu- (e.g., Edington 1968, Wallace 1974, Haddock mented, particularly for members of the Tri- 1977, Wiillace and Merritt 1980, Alstad 1982, choptera family Hydropsychidae (e.g.. Decamps Hauer and Stanford 1982, Brims et al. 1987, 1968, Edington and Hildrew 1973, Gordon Osborne and Herricks 1987, Wetmore et. al. and Wallace 1975, Wdlace and Merritt 1980, 1990, Voelz and Ward 1992). Few studies have Ross and Wallace 1982, Tachet et al. 1992) and considered the entire filterer component foimd for lake outlet communities (e.g., Robinson and in natural (unimpounded, unregulated) streams 'Stream Ecolog\ Center, Department of Biologieal Seienees. Idaho State Universit\-, Pocatello, ID 8.3209-8007. ^Present address: Louisiana Cooperative Fisheries and VVildhfe Researeh Unit, School of Forestn, Wikllilr and Fisheries, U)uisiana State University, Baton Rouge, LA 70803. -^Biolog\- Program, 104 Life Sciences Building, Louisiana State University, Baton Rouge, LA 7080.3. 287 288 Great Basin Naturalist [Volume 56 and distribution patterns of filterer species with respect to a wide range of environmental variables (Edington and Hildrew 1973, Gordon and Wallace 1975, Boon 1978, Ross and Wal- lace 1982). Our objective was to assess the dis- tribution patterns of filter feeders in unim- pounded Rocky Mountain, USA, streams to determine relationships with specific environ- mental variables including flow parameters; stream size, depth, and width; benthic organic matter content; slope; water chemistry; peri- phyton biomass; and temperature. While many studies have considered current velocit>', tem- perature, and food relationships, partitioning of habitat by filter feeders in relation to other environmental variables is poorly known. Methods Stream sites encompassed the Rocky Moun- tain region from northern Wyoming to central Idaho, including 22 streams in Yellowstone National Park and 33 in central Idaho. Streams ranged from 1st to 6th order in size (Table 1). All sites were unimpounded and none were located below lake outflows. Yellowstone sites were sampled each August from 1988 to 1992. All other sites were sampled between July and September during the year(s) indicated in Table 1. Sampling methods were routine meth- ods used in stream ecology (e.g., Platts et al. 1983). Briefly, benthic organisms were sam- pled using a surber net (250 micron mesh) in riffle habitat at 5 transects located at 50-m inter- vals along a stream reach (250 m total reach length). Samples were taken to a depth of 10 cm. VIean densitv' for each filterer species with- in each stream reach was used in statistical analyses to determine relationships with physi- cal variables. Physical environmental variables measured at each stream reach included stream order, slope, width, baseflow (1 transect), mean depth (n = 100 random measurements), mean water velocity {n = 100 random measm-ements), mean embeddedness (n = 100 random mea- surements), and mean substrate size (n = 100 random measurements). Reach-scale means for all variables were used in statistical analyses. Width/depth ratio and several hydraulic para- meters (mean Froude number, mean Reynolds number) were calculated from tliese measine- ments. Annual stream temperature range was estimated from annual maximum (estimated as temperature at the time of sampling) and minimum temperature (the freezing point of water). Water chemistry variables included hardness, alkalinity, pH, and specific conduc- tance. Other biotic variables measured at each stream reach were chlorophyll a (n = 5 per site), ash-free diy mass (AFDM) of periphyton (n = 5 per site), biomass/chlorophyll ratio of periphyton (B/C), and benthic organic matter content (BOM; n = 5 per site). This study did not address food resources or food acquisition among filter feeders; therefore sampling of transported and benthic fine particulate mater- ial was not included in sampling protocol. Relationships between species and environ- mental variables were detemiined using canon- ical correspondence analysis (Ter Braak 1986) and stepwise multiple regression. All compar- isons were made on reach-scale data (reach means for all variables). Comparisons reflect spatial differences among sites sampled in 1 season (summer) to detemiine large-scale distri- bution patterns of filter feeders in 1st- through 6th-order streams. Temporal patterns were not considered here. Canonical correspondence analysis (CCA) allows the investigator to inter- pret multiple species responses along a gradi- ent of multiple environmental variables. This analysis provides a useful interpretation of species -environment relationships through the resulting ordination plot. Once species-envi- ronment correlations were identified using CCA, multiple regression analysis was used to further discern relationships between species abundance and environmental variables. Results In the canonical correspondence analysis (Fig. 1) the first ordination axis (.v axis) ex- plained 37.9% of the total species-environment relationship and the second {y axis) an addi- tional 30.7% (Table 2). Results indicate that sexeral environmental variables were impor- tant in explaining variation in species abim- dance across sites (Fig. 1). Arctopsychc ^randis and Hyclrop.syche abundance related directly to increasing baseflow, width, and stream order (Fig. 1). Parapsyche ekis abundance was inverse- ly related to increasing baseflow, width, and stream order. Brachycentrus abundance related primariK' to depth, substrate size, Reynolds number, and annual temperature range (Fig. 1). Si)uiilitim, Pisidiiim, and Ostracoda abundance 1996] Filter-feeding Invehtebiutes in Rocky Mountain Streams 289 T.\HI,K 1. Suminan of tlic 55 stii streaiiis. Sites are ananj^ed by increasing; stream order and increasinjj; depth within each order Stream Sample dates Order Avg Avg Basedow Slope depth (ni) width (m) (m/s) (%) Caclie. YNP 1988-1992 1 O.OK 0,704 0.003 12 E Blacktail Deer, YNP 1988-1992 1 0.13 0.665 0.048 4.7 Twin, YNP 1988-1992 1 0.13 0.643 0.06 10.7 W Blacktail Deer, YNP 1988-1992 1 0.17 0.550 0.043 3.8 Faiiy, YNP 1988-1992 1 0.23 0.307 0.066 1.0 Pioneer, ID 1990 2 0.05 0.342 0.13 6 Dunce, ID 1990,91 2 0.06 0.109 0.07 17 Goat, ID 1990,91 2 0.06 0.089 0.05 18 Cache, YNP 1988-1992 2 0.09 0.764 0.012 10.1 Packhorse, ID 1991 2 0.09 0.413 0.04 4 Castle, ID 1992 2 0.09 0.160 0.03 11.5 Yellow. ID 1992 2 0.09 0.220 0.03 8 Rose, YNP 1988-1992 2 0.10 0.416 0.027 7.8 Sliver ID 1991 2 0.10 0.243 0.04 5 EFWhinistick. ID 1991 2 0.10 0.460 0.02 2 Cache. YNP 1988-1992 2 0.11 0.832 0.012 8.8 Cliff, ID 1988,90.91 2 0.12 0.407 0.18 12 Amphitheater, YNP 1988-1992 2 0.13 1.11 0.146 4.9 Pony, ID 1992 2 0.13 0.380 0.08 13 Iron Springs, YNP 1988-1992 2 0.14 0.237 0.038 13.1 E McCall, ID 1991 2 0.14 0.196 0.05 2 Blacktail Deer, YNP 1988-1992 2 0.15 0.710 0.151 15.2 Fair>; YNP 1988-1992 2 0.18 0.395 0.083 0.26 WF Cave, ID 1990 3 0.05 0.124 0.01 6 Doe, ID 1990 3 0.10 0.316 0.02 16 SF Cache, YNP 1988-1992 3 0.16 1.70 0.195 3.0 Pioneer, ID 1990 3 0.16 0.612 0.31 6 Hellroaring, YNP 1988-1992 3 0.17 1.23 0.32 2.5 McCall, ID 1991 3 0.17 0.196 0.05 2 Pebble, YNP 1988-1992 3 0.18 1.10 0.592 2.5 Cougar, ID 1990,91 3 0.18 0.297 0.10 12 Cache, YNP 1988-1992 3 0.19 4.60 0.475 1.7 Lava, YNP 1988-1992 3 0.24 0.768 0.893 2.1 Iron Springs, YNP 1988-1992 3 0.27 0.587 0.520 1.1 Beaver, ID 1988 3 0.27 0.800 1.17 4 Cache, YNP 1988-1992 4 0.18 2.05 0.67 1.2 Ramey, ID 1988 4 0.18 0.630 0.74 3.5 Boulder, ID 1992 4 0.19 1.23 0.41 2 Hellroaring, YNP 1988-1992 4 0.20 2.61 0.43 1.8 McCall, ID 1991 4 0.22 0.240 0.13 2 Whimstick Main, ID 1991 4 0.23 0.800 0.10 1 WF Rapid, ID 1992 4 0.25 0.930 1.80 3 Lamar YNP 1988-1992 4 0.34 2.87 2.85 .97 Soda Butte, YNP 1988-1992 4 0.35 2.90 3.00 1.3 Indian, ID 1992 5 0.21 1.43 1.31 1.5 Pistol. ID 1992 5 0.33 1.70 1.80 1.8 Rush, ID 1988 5 0.35 1.51 1.61 1 Camas, ID 1992 5 0.38 2.10 2.92 1 Chamberlain, ID 1992 6 0.24 1.69 2.43 3.5 Big Ck @ Coxey, ID 1988 6 0.31 3.42 5.23 1.5 Rapid, ID 1992 6 0.37 1.48 1.11 2.5 Loon, ID 1992 6 0.37 2.91 3.29 1 Big Ck @ Gorge, ID 1988 6 0.37 4.32 8.83 1 Big Ck@ Rush, ID 1988 6 0.45 4.3 8.04 1.5 Salmon Ri\en ID 1992 6 0.48 1.40 5.47 1 290 Great Basin Naturalist [Volume 56 ^ CN «: Ostraco^SL \CHLa P. elsis Axis #1 (X Axis) Fig. 1. Biplot results of canonical correspondence analysis. Environmental variables (circled) are listed in Table 3. Species are plotted using species names. Positive abundance relationships with a given environmental xariable are indi- cated by species that fall close in the ordination plot to the environmental variable. Species that fall on the opposite end of the plot from an environmental variable exhibit a negative relationship with that variable. Species near the center of the plot exhibit little relationship with environmental variables. did not relate to any of the environmental vari- ables in the ordination plot and are not consid- ered further Stepwise multiple regression results indi- cate species-environment relationships similar to those found in the ordination (Table 3). Arc- topsijche grandis abundance was positively correlated with stream depth and width and negatively correlated with turbident flow (Reynolds number). BracJjycentni.s abimdance was positively correlated with stream depth (Table 3). Hydropsyche abundance showed positive correlation with baseflow and negatixe correlation xvith water hardness and substrate size. Parapsyche elsis abundance showed nega- tive correlation with depth, surface turbulence (Froude number), and specific conductance (Table 3). Discussion Our results support the idea that macroin- vertebrate species in streams respond to envi- ronmental conditions in individualistic ways. Each ta.xon was related to a different set of enx'ironmental variables. General relationships xvith environmental variables for A. grandis, Brachycentrus, and Hydropsyche suggest that these ta.xa are adapted for larger rixer systems (3rd-6th order; Fig. 2). Brachycentrus and Hy- dropsyche are usually found in loxxer reaches in river systems (4th-6th order; Edington and Hildrexv 1973, Boon 1978, Hauer and Stanford 1982, Ross and Wallace 1982, Wetmore et. al. 1990). A. grandis is most often found in mid- reaches (3rd-5th order; Alstad 1980, Cuffney and Minshall 1981, Hauer and Stanford 1982).' 1996] Filter-feeding Invertebrates in Rocky Mountain Streams 291 Table 2. Ht'sulls of canonical correspondence analysis. Eigenvalues give the importance of an axis on a scale between 0 and 1. Total inertia is the total variance in the species data. The species-environment correlations scale the strength of the relationship hetwcen species and en\ironment lor the axes. Total Axes 1 2 3 4 inertia Eigenvalues .444 .360 .174 .085 2.50 Species-en vironnieul correlations .S19 .772 .640 .441 Cumulative percentage of \arianci': of species data 17.8 32.1 39.1 42.5 of species-environment relationsh P .37.9 68.6 83.5 90.7 Sum of all canonical eigenvalues 1.172 Among the taxa adapted for large streams, habitat partitioning is apparent in tliis study as in others (Edington and Hildrew 1973, Boon 1978, Alstad 1980, Hauer and Stanford 1982, Ross and Wallace 1982). Taxa exhibited reach- scale macrohabitat preferences with Brachij- centnis distribution related to stream depth, Hydropsijche related primarily to stream flow, and A. grandis to a combination of width, depth, and turbulence. P. elsis was prevalent in headwater stream reaches (Fig. 2), a pattern found in several other studies (Alstad 1980, Hauer and Stanford 1982). Distribution patterns for P. elsis were explained by flow and stream-size variables. Stream temperature may also be an important variable explaining P. elsis distribution patterns (Alstad 1980, Hauer and Stanford 1982). Annual temperature was measured in this study based only on yearly max/min readings, which may not adequately reflect differences in tempera- ture between headwater sites and downstream locations, resulting in the lack of P. elsis pat- terns explained by temperature in our analysis. Also, previous studies that suggest a down- stream temperature gradient as the explanation for P. elsis distribution (Hauer and Stanford 1982) did not consider other variables (e.g., physical and hydrologic variables) that may con- tribute to habitat selection by P. elsis. Multiple factors are probably responsible for P. elsis high abundance in headwater streams, including temperature patterns and hydraulic conditions. Our results agree with published distribution patterns for all 4 taxa and provide evidence for physical factors that are important in determin- ing habitat selection for each taxon (Fig. 2). Habitat preferences demonstrated in this study are for distribution patteiTis among streams at the reach scale. Data were collected within a 250-m reach at each site and expressed as reach means for all variables in order to iden- tify factors affecting large-scale (among site) distribution patterns among taxa. Microhabitat requirements are ultimately responsible for the physiciil habitat selected by filter feeders (Smith- Cuffney and Wallace 1987, Wetmore et. al. 1990), but reach-scale comparisons allow broader scale distribution patterns to be stud- ied. The reach-scale comparisons herein indi- cate general conditions at each site in temis of available macrohabitat. The trends observed in the data indicate animal preferences for a given reach and its associated habitat condi- tions. Differences in reach-scale means among variables may also reflect differences in gen- eral microhabitat conditions available among sites (e.g., slow- or fast-velocit\' microhabitats). Reach-scale means, therefore, can ser\e as a useful integrator of microhabitat conditions in order to facilitate comparisons at larger scales. Evolutionary patterns probably have led to habitat partitioning based on current speed and filtration rate among filter feeders in Rock>' Mountain streams with some taxa adapted for larger streams {Brachycentrus, Hydropsyche, and A. grandis) and some for smaller systems {P elsis; Alstad 1980, 1982). Filter feeders may be a usefijl group to address habitat partition- ing on large spatial scales in streams because many filterer taxa appear to have partitioned habitat at these scales. In this study, stream size (width, depth) and hydraulic parameters (l)aseflow, turbulence) were more important in explaining species-environment relationships than other variables such as water chemistry, periphyton biomass, or benthic organic matter. Our results provide support for the idea that evolutionaiy divergence among benthic macro- invertebrate filterers has resulted in habitat partitioning according to stream size and hydro- logic parameters in Rocky Mountain streams 292 Great Basin Naturalist [Volume 56 Table 3. Summaiy of the stepwise multiple regression results of the 4 most abundant species (dependent variable) versus the 20 environmental variables. Partial con-elation coefficients and p values (parentheses) are shown for each variable. Variables included in the regression model for each species are shown (varial)le included if P < 0.05). Variable acronyms in Figure 1 are shown in parentheses. Environmental Arctopsijche Braclujcciitnis Hiidropsijchc Parap.sijche variable '. 1996] Filter-feeding Invertebrates in Rocky Mountain Streams 293 Acknowledgments Yellowstone National Park streams were sampled as part of a study supported by the University of Wyoming/National Park Service Research Center. Idaho stream sampling was supported by the U.S. Forest Service, Idaho Division of Environmental Quality; and Idaho State University. Numerous members of the Stream Ecology Center at Idaho State Univer- sity aided in the field sampling, including P Dey, P Koestier III, D. E. Lawrence, M. J. Mclntrye, J. N. Minshall, G. C. Mladenka, D. C. Moser, J. S. Nelson, C. T. Robinson, R. L. Van- note, and many others. C. T. Robinson did much of the data compilation for the environ- mental variables that were used in the analy- ses. Data analysis and writing by TBM and JRM were supported in part by Louisiana State University. Literature Cited Alstad, D. N. 1980. Comparative biolog>' of the common Utali Hydropsycliidae. American Midland Natiualist 103; 167-174. ' . 1982. Cunent speed and filtration rate link cad- disfly phylogeny and distributional patterns on a stream gradient. Science 216: 533-534. Boon, E J. 1978. The pre-impoundment distribution of certain Trichoptera larvae in the North Tyne river system, with particular reference to current speed. Hydrobiologia 57: 167-174. Bruns, D. a., a. B. Hale, and G. W. Minshall. 1987. Ecological correlates of species richness in three guilds of lotic macroi7i\'ertebrates. Journal of Fresh- water Ecology 4: 163-176. CUFFNEY, T. E, and G. W. MiNSHALL. 1981. Life histoiy and bionomics of Arctopsyche granclis (Trichoptera) in a central Idaho stream. Holarctic Ecologv 4: 252-262. Decamps, H. 1968. Vicariances ecologiques chez les Tri- chopteres de Pyrenees. Annates de Limnologie 4: 1-50. Edington, J. M. 1968. Habitat preferences in net spinning caddis larvae with special reference to the influence of water velocity. Journal of Animal Ecologv 37: 675-692. Edington, J. M., and A. H. Hildrevv. 1973. Experimental observations relating to the distribution of net-spin- ning Trichoptera in streams. Verhandlungen der Internationalen Vereinigung fuer Limnologie 18: 1549-1558. Gordon, E., and J. B. Wallace. 1975. Distribution of the family Hydropsycliidae (Trichoptera) in the Savannah River basin of North Carolina and Georgia. Hydrobi- ologia 46: 405-123. Haddock, J. D. 1977. The effect of stream cunent velocity on the habitat preference of a net-spinning caddis fly larva, Ihjdropsijche oslari Banks. Pan-Pacific Ento- mologist 53: 169-174. llAi 1., C. A. S., J. A. Stanford, and E R. Hauer. 1992. rhe distribution and abundance of organisms as a consecjucnce of energy balances along n\ultiple envi- ronmental gradients. Oikos 65: 377-390. Hauer, E R., and J. A. Stanford. 1982. Ecological responses of hydropsychid caddisflies to stream regu- lation. Canadian Journal of Fisheries and Aquatic Sci- ences 39: 1235-1242. Heard, S. B., and J. S. Richardson. 1995. Shredder-col- lector facilitation in stream detrital food webs: is there enough evidence? Oikos 72: 359-.366. OsRORNE, L. L., AND E. E. Herricks. 1987. Microhabitat characteristics o'i Hyihopsijche and the importance of body size. Journal of the North American Bentholog- ical Society 6: 115-124. Platts, W. S., W. E Megahan, and G. W. Minshall. 1983. Methods for evaluating stream, riparian, and biotic conditions. U.S. Forest Senice General Technical Report INT- 138. 70 pp. Richardson, J. S., and R. J. Mack.\v. 1991. Lake outlets and the distribution of filter feeders: an assessment of hypotheses. Oikos 62: 370-380. Robinson, C. T, and G. W. Minshall. 1990. Longitudinal development of macroinvertebrate commimities be- low oligotrophic lake outlets. Great Basin Naturalist 50:303-311. Ross, D. H., and J. B. Wall.'\ce. 1982. Factors influencing the longitudinal distribution of larval Hydropsychi- dae in a southern Appalachian stream system. Hydrobiologia 96: 185-199. S.MITH-CUFFNEY, E L., AND J. B. WALLACE. 1987. The influence of microhabitat on availability of drifting invertebrate prey to a net-spinning caddisfly. Fresh- water Biology 17: 91-98. Tachet, H., J. E Peirrot C. Rou.x, and M. Bournaud. 1992. Net-building behaviour of si.x Hydropsyche species (Trichoptera) in relation to ciUTent velocit)' and distribution along the Rhone River. Journal of the North American Benthological Societ>' 11: 350-.365. Ter Bra.\k, C. J. F. 1986. Canonical conespondence analy- sis: a new eigenvector techni(iue for direct gradient analysis. Ecology 67: 1167-1179. VoELZ, N. J., AND J. V. Ward. 1992. Feeding habits and food resources of filter-feeding Trichoptera in a regu- lated mountain stream. Hydrobiologia 231: 187-196. Wallace, J. B. 1974. Food partitioning in net-spinning Tri- choptera lai-vae: Hydropsyche ventdaris, Cheumato- psyche etrona, and Macronema zehrafum. Annals of the Entomological Society of America 68: 463—472. Wallace, J. B., and R. W Merritt. 1980. Filter-feeding ecology of aciuatic insects. Annual Review of Ento- mology^' 25: 103-132. Wetmore, S. H., R. J. Mackay, and R. W Newbury. 1990. Characterization of the hydraulic habitat of Brachy- centrus occidentalism a filter feeding caddisfly. Journal of the North American Benthological Societ)' 9; 157-169. Received 6 December 1995 Accepted 27 June 1996 Great Basin Naturalist 56(4), © 1996, pp. 294-299 STEM GROWTH AND LONGEVITY DYNAMICS FOR SAL/A AR7ZOMCA DORN Vicki L. Taylor^ Kimball T. Harper^, and Leroy L. Mead^ Abstract. — Diameter-age relationships of Salix arizonica (Arizona willow) stems were investigated for 5 populations on the Markagunt, Paunsaugunt, and Sevier plateaus in southern and central Utah. Of the 430 stems studied, none exceeded 26 mm in diameter at ground level (estimated age of 19 \ r). Equations developed for predicting age from stem diameters consistently accounted for over 90% of the obsei-\'ed variation. Slopes of predictive equations were homoge- neous across die 3 sites considered in detail. At 2 sites 46% and 38% of the stems exceeded 10 mm (~7 yr old) diameter at ground level. At a 3rd site, no stems survived to exceed that size. Stem-age profiles at specific sites may thus be usefld for assessing the relatixe favorability of local enxironments for the species. Key words: Arizona icilloic, Salix, stem diameter dendroehi jnviology. sou them Utah. The puipose of this study was to assess stem diameter-age relationships in Salix arizonica (Arizona willow), a species so rare that routine severance of stems for aging cannot be justi- fied. Our objective was to develop a stem-age prediction model based on stem basal diame- ters. Ultimately, we desired to accurately esti- mate stem age at a broad range of ecological situations without sacrificing stems. We also evaluate the possibilit}' of using stem-age pro- files at an array of sites to determine their rela- tive favorabilit)' for growth of S. arizonica. Dendrochronology as a Tool Growth rings of trees and shrubs have been used for many decades for aging stems and dating past climatic events (Douglas 1935, Glock 1937). Growth rings are also used to establish unique sequences of good and poor years that permit dating nonliving tree frag- ments used in prehistoric human structures (Schulman 1956, Fritts 1971, Stockton and Meko 1975, Harper 1979). Ring-width varia- tions are often used to assess differences in the favorability of various environments for the growth of selected species (Ferguson and Humphrey 1959, Fritts 1962, Stockton and Fritts 1973, Fritts 1974). Although these stud- ies have focused mainly on trees (Glock 1955, Argeter and Glock 1965), some have dealt with shrub species (Ferguson 1958, 1959, Ferguson and Humphrey 1959, Brotherson et al. 1984, 1987). Shrub studies have detailed the effects of variations in available moisture on plant growth in specific habitats or provided infor- mation for interpreting archaeological prob- lems. Ring counts have also been used to pre- dict stem diameter-age relationships in predic- tive models for inteipreting site quality for var- ious species or for clarification of successional patterns in vegetation diat includes many woody species (Brotherson et al. 1984, 1987). The Species and Its Distribution Salix arizonica is small. Rarely do stems ex- ceed 1.0 m in height. The species occurs in such dense carpets of other species (both vascular and nonvascular) that reproduction via its tiny, wind-dispersed seeds appears to be uncom- mon. Accordingly, the species apparently per- sists at occupied sites primarily by vegetative reproduction. In the process, what appear to be large clones (as much as 10 m across) may develop. Salix arizonica occurs in 2 disjunct locations in the Intermountain West. The species was first discovered on the White Mountains of east central Arizona by Carl-Eric Granfelt in 1969 (Galeano-Fopp 1988). Robert Dom (1975) used holotype specimens collected by Granfelt to describe the species in 1975. In November 1992, unaware that the species occurred in Utah, the U.S. Fish and Wildlife Service pro- posed S. arizonica for listing as endangered with 'Departriiciil dl HotaiiN ami Haiiiic ScioiKv, HriKliaiii Vouiii; Uiii\cr.sity, Provo, UT 84(i()2. 294 1996] Arizona Willow Stlm CtHowth 295 designation of critical habitat (Atwood 1995). In June 1993 a prcvionsly inisidcntilicd herb- arium specimen of S. arizonica was discovered; it had l)een collected on the "Sevier Forest" (now Dixie National Forest) in 1913. During June 1994, S. arizonica was discovered on the Markagimt I^lateau near Brianhead resort area. Subsequent searching revealed a small popula- tion on the Paunsaugunt Plateau and 2 more farther north on the Sevier Plateau (Mead 1996). Following this "rediscovery" of S. ari- zonica in Utah, USDA Forest Service, USDI Fish and Wildlife Sei-vice, and USDI National Park Service officials cooperated in developing a conservation agreement and strategy that outlines the "actions, costs and skills needed to implement protective measures and research studies needed for the species" (Atwood 1995). As a result of the consei"vation agreement and strategy, which documents long-term plans for consei-vation of S. arizonica, the Fish and Wild- life Service withdrew their proposed rule to list the species as endangered (Arizona Willow Interagency Technical Team 1995). Although die species is locally abundant near Brianhead, its total range is small in both Ari- zona and Utah, and populations rarely include more than a few score plants. This rarity seems related to the plant's preference for an uncom- mon habitat: it grows preferentially on igneous soils in cold, wet sites. In addition, in the White Mountains, management has favored conifers that reduce flow in riparian systems, leading to poor drainage as waterways become peat- choked. Such environments become poorly aer- ated and less suitable habitat for S. arizonica. Heavy use by elk has also adversely affected the species in Arizona (Arizona Willow Intera- gency Technical Team 1995). This study has been confined to the Utah populations of Salix arizonica (Fig. 1), but we have attempted to sample the full range of conditions associated with the species in our study area. Methods and Study Areas The diameter-age data for S. arizonica were collected from 3 populations: 2 on the Cedar City Ranger District and another on the Powell Ranger District, Dixie National Forest (Fig. 1). The Rainbow Meadows, Lowder Creek, and East Fork of the Sevier River populations were chosen because they represent environmen- tally intermediate (Lowder Creek) as well as extreme environmental conditions for S. ari- zonica in Utah. The Rainbow Meadows site occurs on acid soils at near niitximal elevations for the species, while the East Fork of the Sevier River population occurs on alluvium derived from calcareous substrates at the low- est elevation known for the species. Depth of p(>at la>'er was determined at each site by digging pits to expose soil profiles (Mead 1996). At Lowder Creek, Sheepherder Camp, and Sevenmile Creek, depth to water table was determined by opening a hole approximately 1 m deep with a 1.27-cm-diameter pointed rod, then inserting a ().64-cm-diameter wooden dowel into the hole to measure depth to water. This measurement was taken at each plant sam- pled and an average value was computed for each site. Depth to water table at Rainbow Meadows was determined by measuring dis- tance from soil surface to water table surface in a soil pit (Mead 1996). Depth to water table was determined at the East Fork site by mea- suring distance from the soil surface to the sur- face of water running in the creek. This mea- surement was taken at each S. arizonica clone; the mean distance is reported in Table 1. Depth of peat layer and depth to water table are vari- able among the study sites, with the Rainbow Meadows site having the highest water table and greatest peat depth (Table 1). Two otlier populations of S. arizonica are con- sidered in this report. Populations at Sheep- herder Camp, Sevenmile Creek, and Lowder Creek have been sampled to establish stem- diameter profiles based on samples of many randomly chosen stems (154, 104, and 130 stems, respectively, sampled at the 3 foregoing sites). No stems were severed for aging at the Sheepherder or Sevenmile sites. The Rainbow Meadows site is approximately 1.6 km south and slightly east of Brianliead Peak at approximately 3155 m elevation (37°40'N, 112°56'W). Soils are derived from tertiaiy vol- canics with a histosol surface horizon (Mead 1996). The Lowder Cieek population is approx- imately 4 km east and slightly south of Brian- head Peak (37°41'N, 112°48'W). Soil at this site is developed from tertiaiy volcanic mater- ial below an alluvium surface layer (Mead 1996). The East Fork population, approximately 48 km from the Lowder population (37°26'N, 112°2rW), is at the lowest elevation knowai for 296 Great Basin Naturalist [Volume 56 UTAH Fig. 1. A, Rainbow Meadows; B, Lowder Creek; C, Sheepherder Camp; D, Sevenniile Creek; E, East Fork of die Sevier River this species in Utah. This population grows on alluvium from the Claron Limestone Forma- tion with an organic surface horizon (Mead 1996). Commonly associated plants at the sites sam- pled include Salix planifolia. Polygonum bistor- toides, Aconitum cohimbimnim, Carex rnicro- ptera. Geranium richardsonii, Geum macro- phijllwn, and Pedicuhiris groenlamlica (Mead 1996). As Mead (1996) has shown, the relative abundance of these species varies from site to site depending on such variables as soil temper- ature, depth to water table, and soil reaction. Fifteen randomly chosen stems were sam- pled at each site at the Rainbow and Lowder locations. At each site 4 quadrants were estab- lished around randomly chosen points. The stem closest to the random point in each of 5 size-classes was collected in each of 3 quad- rants (the right rear quadrant was not sampled). Stems were severed at ground level using wire cutters or a small hand saw. The diameter- classes sampled were 0-5 mm, 5.1-10 mm, 10.1-15 mm, 15.1-20 mm, and >20 mm at ground level. Thus, 3 stems per size-class were sampled at each site. Due to the low densit)' of S. arizonica at the East Fork site, quadrants were not used. Stems were collected from all S. arizonica clones inside a livestock- grazing exclosure in the stiuK' area. No stems could be found at this site for the >20 mm size-class, so only 12 stems were sampled. Stem samples were labeled, placed in indi- \idual bags, and taken to the lab. Stem bases 1996] Arizona Willow S ilm Growth 297 Taulk L Environmental eonditions at 5 Salix arizonka sites. Water table was taken at all plants sampled wherever soil stoniness permitted insertion of flie dowel to water depth. At East Fork water depth was based on only 16 points beeanse only 16 plants e.xist at that site. The measme ot varianee around water table mean depth is standard error. Site Ele\ation Soilpll Soil temp, (a) Mean depth to Peat depth (m) 50-cm depth (°C) water table (em) (em) Rainbow 3155 5.15 8.3° (September) 5.1 ±NA 32 Lowder 3139 5.79 10° (August) 45.5 ±1.81 0 Sheepherder 3130 5.72 6° (August) 44.4 ±1.60 44 Sevenmile 2789 6.38 10° (August) 10.5 ±1.12 0 East Fork 2536 7.61 16° (July) 46.5 ± 6.89 0 N.\ = not a\ailal)le. Table 2. Regression ecjuations relating stem diameter to age of S'«/j.v arizonica stems taken from 3 different sites. The regression equation for all sites combined is also shown. In the equation the independent variable, X, represents stem diameter (in mm). The symbol Y represents estimated age of any given stem. Site No. of stems Equation fi2 Signiiieance lexel Lowder 15 Y = -0.42 + 0.82X .953 0.01 Rainbow 15 Y - -0.28 + 0.78X .950 0.01 East Fork 12 Y = -1.40 + 0.71X .910 0.01 All 3 sites combined 42 Y = -0.99 + 0,81X .926 0,01 were sectioned diagonally and sanded with fine sandpaper; growth rings were counted twice (once by each of 2 observers) with the aid of a stereoscopic microscope (Brodierson et al. 1987). Diagonally cut surfaces permitted growth rings to be identified widi greater confidence. Sanded surfaces sometimes had to be polished with immersion lens oil to enhance ring visibility. Each growth ring was assumed to represent 1 yr's growth. Linear regression was used to quantify stem diameter-age relationships. Results S. arizonica stems from the 3 sites at which stems were cut and aged ranged in basal diam- eter from 2 to 26 mm and in age from 1 to 19 yr. Stem diameters (mm) were plotted against stem age (yr), and regi"ession equations were generated (Table 2). Slopes for regression equations from the 3 sites were tested for simi- larity using methods described in Snedecor and Cochran (1967) and were found not to dif- fer significantly (F > 0.50). Thus, data from all sites were pooled to produce a single equation (Y = -0.99 + 0.8 IX) for subsequent use in estimating age (Y) from diameter (X) (Fig 2). As a further test of the validity of pooling data from all sites, we used the individual esti- mator equation developed for each site to pre- dict age of willows collected from the other 2 sites (i.e.. Rainbow equation used to test Low- der and East Fork samples, Lowder equation used to test Rainbow and East Fork samples, etc.). These analyses demonstrated that esti- mated ages for any equation-test site combina- tion were always strongly correlated with actual age {R^ always > .90). In these analyses no stems were found to differ fi^om predicted age based on diameter by more than 3 yr, and most stems (>90%) differed by less than 2 \t (Fig. 1). An application of the age-estimator equa- tion is shown in Figure 3. As part of the yearly monitoring program, basal diameters of S. ari- zonica were taken for a large sample of stems at each of 3 sites: Sheepherder Camp, located approximately 8 kin south of Brianhead Peak at 3130 m elevation (37°37'N, 112°56'W); Sev- enmile Creek, 11 km north of Fish Lake in the Fishlake National Forest, Loa Ranger District at 2789 km elevation (38°39'N, 111° 40 'W); and Lowder Creek (described above). At each of these sites, the numbers of stems within each diameter-class were tabulated and are reported as percent of total stems in each size-class. The results (Fig. 3) demonstrate large differences in stem-diameter profiles among the 3 sites. At Sheepherder Camp over 4% of the stems are 298 Great Basin Naturalist [Volume 56 Y = -0.99 + 0.8 IX ^ R' = 0.926 ■ Qp, □ 10 20 Basal Diameter (mm) 30 Fig. 2. Stem basal diameter-age relationships of S. ari- zonica on the Markagunt and Paunsaugunt plateaus of southern Utah. larger than 20 mm diameter at ground level. However, less than 1% of the stems at Lowder Creek exceed that diameter, and at Sevenmile Creek no stems have sui-vived to become 10 mm in diameter. These results suggest that Sheepherder Camp is a more favorable site for growth of the willow than either Lowder Creek or Sevenmile Creek. Alternatively, the results may indicate that willows are less severely browsed by ungulate grazers at Sheepherder Camp than at the other 2 sites. Since ungulate exclosures were not erected at these sites until fall 1994, data are cuirently too limited to dis- tinguish between these alteiTiatives. Discussion The regression equation created from the pooled data of all 3 sites should be useful for predicting ages of S. arizonica from any known Utah location using only basal stem diameters. The equation should be useful for many pro- jects in which stem age is desired but stems cannot be sacrificed. For example, the ability to estimate age of stems accurately from basal diameter may permit scientists studying the species to correlate stem ages and stem-age profiles with site conditions without destroying individual stems. The results of this study demonstrate little variation in stem growth rates for S. arizonica over a wide range of ele\ations and parent materials (Table 1). That result suggests that the species occupies but a narrow range of habitat 100 % of All Stems Stem Diameter Class ■ 0-5 mm B 5.1-10 mm m 10.1-15 mm 0 15.1-20 mm 20.1-25 mm >25mm D Lowder Sheepherder Sevenmile Site Fig. 3. C'oinpar:ili\e stem diameter distrihutioiis foi' sites iur whicii a large, rantloiii in\eiit()r\ oi stem diameters was available. 1996J Arizona Willow Stlm CiuowTii 299 situations witliin its o\CM-all geographic range. Occupied sites almost always appear to have been modified by biological processes that result in peat deposition and development of a rooting zone that is somewhat isolated hom the unaltered geologic substrata at the site. Stem-age profiles should permit managers to identify sites where performance (stem sur- vival and/or reproduction b\' seed oi" rhizome) of the willow is above or below regional aver- ages. Such data would help managers deter- mine whether growth and i^eproduction of the species could be enhanced by reduction of use b)- browsers. To assist managers with such deci- sions, fenced areas that exclude domestic and wild ungulate browsers have been erected at Lowder Creek, Sheepherder Camp, and on the East Fork of the Sevier. An additional exclo- sure will be built at Sevenmile Creek in 1996. The U.S. Forest Sei'vice intends to continue monitoring Salix arizonica populations through- out its range to learn about factors that influ- ence growth, reproduction, and stem sui^vival. Data from grazing exclosures will reveal the extent to which browsing controls stem size and longevity. The extent to which the abiotic environment limits stem growth and seed pro- duction can be more readily separated from the effects of browsing now that animal exclo- sures have been constructed. Acknowledgments We thank Ron Rodriguez for financial and moral support for this project. Julie Tolman provided invaluable assistance in the field and lab. This work was completed in part with hmds provided by the U.S. Forest Service. Literature Cited Argeter, S. R., and W. S. Clock. 1965. An annotated bib- liography of ti-ee growtli and growth rings, 1950-1962. University of Arizona Press, Tucson. 180 pp. Arizona Willow Interagency Technical Team. 1995. Arizona willow consei"vation agreement and strategy. U.S. Forest Sei"vice, Intermountain Region, Ogden, UT; U.S. Forest Service, Southwest Region, Albu- querque, NM; National Park Sei-vice, Rocky Moun- tain Region, Denver, CO; U.S. Fish and Wildlife Ser- vice, Mountain-Prairie Region, Salt Lake City, UT; U.S. Fish and Wildlife Sei-vice, Southwest Region, Albuquerque, NM. Atwood, D. 1995. Where have all the Arizona willows gone? Sego Lily (Newsletter of the Utah Native Plant Society) 18; 3. Brothkhson, J. D., J. G. Carman, and L. A. S/yska. 1984. Steni-dianieter age relationships of Tamxiiix ravwsui- sima in central Utah. Journal oi' Range Management 37: 362-364. BiurniERsoN, J. D., K. P Price, and L. ()"R()Urke. 1987. Age in relationship to stem circumference and stem diameter in cliffrose {Cowania mexicana van stam- huriana) in central Utah. Great Basin Naturalist 47: 334-338. DoRN, R. D. 1975. A systematic stud> oi' Salix section Cor- datae in Nordi America. (Canadian Joinnal of Bolanv 53: 1491-1522. Douglas, A. E. 1935. Cliiuatic cycles and tree growth I: a study of the annual rings of trees in relation to cli- mate and solar activity CJarnegie Institute of Wash- ington, Publication 289. Volume I. Washington, DC. Ferguson, C. W. 1958. (irowth rings in big sagebrush as a possible aid in dating archaeological sites. Pages 210- 211 in A. E. Dittert, Jr., editor. Recent developments in Navajo Project salvage archaeol()g\. El Palacio 65: 201-211. . 1959. Growth rings in wood) shrubs as potential aids in archaeological interpretation. Kiva 25: 24-.30. Ferguson, C. W, and R. R. Himphrey. 1959. Growth rings on big sagebnish reveal rainfall records. Progressive Agriculture in Arizona 11; 3. Fritts, H. C. 1962. The relation of growth ring widths in American beech and white oak to xariations in cli- mate. Tree-Ring Bulletin 25; 2-10. . 1971. Dendroclimatology and dendroecologv'. Quaternan' Research 1: 419-449. . 1974. Relationships of ring widths in arid-site coni- fers to variations in monthly temperature and precip- itation. Ecological Monographs 44: 411—140. Galeano-Popp, R. 1988. Salix arizonica Dom on the Apache- Sitgreaves National Forest: inventon' and habitat study. USDA Forest Service, Apache-Sitgreaves National Forest, Contract 43-8173-8-687. 46 pp. Clock, W. S. 1937. Principles and methods of tree-ring imalysis. Carnegie Institute of Washington, Publication 486. Washington, DC. 100 pp. . 1955. Tree growth II. Growth rings and climate. Botanical Review 21; 73-188. Harper, K. T. 1979. Dendrochronologv' — dating with tree rings. In W M. Hess and R. T. Matheny editors. Sci- ence and religion: toward a more useful dialogue. Volume 1. Paladin House Publishers, Geneva, IL. Mead, L. L. 1996. Habitat characteristics of Arizona willow in southwestern Utah. Unpublished master's thesis, Brigham Yoimg University, Provo, UT. SCHULMAN, E. 1956. Dendroclimatic changes in semi-arid America. University' of Aiizona Press, Tucson. 142 pp. Snedegor, C. W, and W. G. Cochran. 1967. Statistical methods. Iowa State University Press, Ames. 593 pp. Stockton, C. W, and H. C. Fritts. 1973. Ix)ng-term recon- struction of water level changes for Lake Athabaska by analysis of tree rings. Water Resources Bulletin 9: 1006-27. Stockton, C. W, and D. M. Meko. 1975. A long-term his- tory of drought occun^ence in western United States as inferred from tree rings. Weathenvise 28: 245-249. Received 15 March 1996 Accepted 5 July 1996 Great Basin Naturalist 56(4), © 1996, pii. 300-307 SOURCES OF VARIATION IN COUNTS OF MERISTIC FEATURES OF YELLOWSTONE CUTTHROAT TROUT [ONCORHYNCHUS CLARKI BOUVIERI) Carter G. Kru.se^ \Va\ ne A. Huljert^, and Frank J. Raliel- Abstfl\ct. — We determined variability in counts of nieristic featines (pyloric caecae, vertebrae, pelvic fin rays, gill- rakers, basibranchial teeUi, scales above the lateral line, and scales in the lateral series) of Yellowstone cutdiroat trout {Oncorhijnchiis chirki buiivieri) by 3 independent readers, by the same reader on 3 different occasions, and among fish from 12 sampling sites within a 650-kni- watershed. Genetic purit>- of the cutthroat trout was determined b\' elec- trophoretic analysis. Significant differences in nieristic counts were obsen'ed among 3 readers and among sampling sites, but not among 3 occasions b\ a single reader Scale counts were within the reported range for Yellowstone cutthroat trout, but counts of other structures (pyloric caecae, gillrakers, \ertebrae) were as similar to rainbow trout as to Yellowstone cut- throat trout. Meristic counts identified the fish as cutthroat trout; however, variation among readers and sampling sites, as well as within the species, limits their use w hen identih ing geneticalK pure cutthroat trout or assessing possible integi"a- tion with rainbow trout. Key words: meristic counts, Yclloustonc cutthroat trout, lucristic variation, genetics, rainlnnc trout, conservation biology. H>briclizati()n of nati\'e cutthroat trout {OncorJiynchiis clarki) with introduced rainbow trout (O. mykiss) has contributed to the dechne of cutthroat trout in the western United States (Allendorf and Leary 1988, Gresswell 1988, Behnke 1992). An important initial step toward restoration or presentation of native cutthroat trout populations is reliable identification of geneticalK' pure populations (Rinne 1985, Lean et al. 1989). Meristic features, such as fin ra> or \ erte- brae counts, ha\'c been used to identif)' hy- bridization among species of trout. The tech- nic^ue assumes that hybrids are intermediate to parental taxa and haxe increased morphologi- cal variance (Leaiy et al. 1985, 1991, Marnell et al. 1987). This assumption is not alwa\s \ alid and meristic comparisons can proxide mislead- ing taxonomic information (Leaiy et al. 1984, 1985, Cunens et al. 1989). Enxironmental influ- ences and ol)ser\ er error are 2 factors that can lead to variation in meristic counts for a species among sampling sites (CuiTcns et ill. 1989, Leaiy et al. 1991, Hubert and Alexander 1995). Even though more definitive biochemical methods have been dc\ eloped (Lean et al. 1987, 1989, Nielsen 1995), biologists continue to use meristic features to assess genetic purity of cutthroat trout populations (Loudenslager and Gall 1980, Rinne 1985, Behnke 1992). Protein electrophoresis is a reliable method of determining genetic status of trout popula- tions (Marnell et al. 1987, Leary et al. 1989, Nielsen 1995). Electrophoresis provides data on allelic frequencies at genetic loci for differ- ent populations (Avise 1974). Hybridization can be determined when allele frequencies un- usual for a particular species are found at sev- eral diagnostic loci that occur between taxa (A>ala and Powell 1972, Leaiy et al. 1989). For example, Yellowstone cutthroat trout (O. c. hou- vicri) can be differentiated from rainbow ti^out using alleles at 10 diagnostic loci (R. Leary, Uni- \ersit\ of Nh)ntana, personal communication). If this procedure is xalid, managers could save considerable time and money using meris- tic features instead of biochemical analysis to assess genetic purit\- of cutthroat trout. How- ever, unless xariation in meristic counts is min- imal among readers or sampling sites, the use- fulness of meristic features in adecjuateh' assess- ing genetic purit>- will be limited. The objec- ti\es of this stiid\ were to determine \ariabil- it\ in counts of meristic featines (1) among 'us. Ci-olo.nicul Survi'v, Wyoiiiini; ( :ocip.MMlivi- Kisli .uul Wikllilc Ucsrarch liiit. I'liivrrsilv (il WytiMiiivi l,aianiic, WY S2()71-.31fi(i. iTlic Wvcmium,!; Coop- erative Fish and Wildlife Research Unit is joiiilU sii|)porteil In the UMi\ersit\ olWyoriiini;. llic WVoinini; Came ,iml h'isli l)r|i.»lnuMit. Ocpailnu'iil ol'tlu- Inte- rior, and Wildlife Manasenient Instilnte.) ^Department of Zooloj,", and Ph\si<)lo,i;\. I nixcrsitv of \\\(iininu, Laramie, WV N2()71-:iHi(v 300 1996] Variation in Yellowstone Cunii hoax Trout 301 Shoshone National Forest Boundary 15 Greybull River Wood River Cody 77 km/ Meeteetse 34 km Fig. 1. Map of Wyoming showing the location of the Greybull River drainage. Sites where cuttiiroat trout were sam- pled are numbered in reference to Table 1 . independent readers, (2) among counts by a single reader, and (3) among sampling sites within a moderate-sized watershed (650 ki ull River drainage, numlier of fish collected, and sample sizes from each used for meristic counts and analysis. Genetic status indicated by pure Yellowstone cutthroat trout (P) or potential finespotted cutthroat trout hybridization (FSC). Ninnber preceding the stream name conesponds to sites in Figure 1. Stream Number offish collected Allozyme analysis Counted In all readers Counted by single reader 1 Anderson 15 2 Brown 17 3 Chimney 16 4 Cow 16 5 Deer 16 6 Dundee 2 7 Eleanor 19 8 Francs Fork 9 Upper Grevbull 15 10 Lo\ver Grevbull 20 11 Jack 21 12 Mabel 2 13 MFWood 15 14 NF Pickett 15 Picket 17 16 Pinev 17 Red 4 18 SF Anderson 19 SFWood 18 20 Venus 16 21 VVarhouse 18 22 W Timber 23 Wood 21 15 (P) 20 (P) 19 (FSC) 15 (FSC) 19 (P) 15 (FSC) 20 (FSC) 5 10 14 16 15 11 16 7 20 10 2 3 4 4 10 14 18 trout, and rainbow trout have been stocked in the system. Methods Twenty-three streams in the Greybull River drainage were sampled with batteiy baclq^ack electroshockers from June to September 1994. Cutthroat trout were collected from 1 site (12-20 fish) on each of 18 streams. For analysis puqjoses the upper and lower Greybull River sites were considered separately (Table 1). Fish were collected from the midpoint of the length of each stream in which cutthroat trout were found. A sample of eye, liver, and muscle tissue was removed from each fish, wrapped in alu- minum foil, and frozen within 1 h in liquid nitrogen. The remainder of each specimen was preserved in 75% ethyl alcohol. Tissue samples from each fish were individually identified. Frozen tissue samples from 7 of the 18 streams were sent to the Wild Trout and Salmon Genetics Lab (WTSCiL) at the Univer- sit\' of Montana, Missoula, for genetic analysis. The 7 sites were selected to represent fish dis- tribution in the drainage (Tible 1, Fig. 1). Also, they were close to locations where finespotted cutthroat trout and rainbow trout had been pre- viously introduced in the drainage (Wyoming Game and Fish Department records). Protein electrophoresis (Allendorf and Phelps 1980, Leary et al. 1984, Perkins et al. 1993) was per- foiTned to detect each specimen's genetic char- acteristics at 45 loci in muscle, liver, or eye tis- sue. Allele frequencies at 10 diagnostic loci (Table 2) were evaluated to determine hxbridi- zation with rainbow trout. Additionally, the presence of the AK- 1*333 allele was evaluated to detect possible finespotted cutthroat trout hybridization. Seven meristic features were counted on the preserved cutthroat trout: (1) basibranchiiil teeth, (2) anterior gillrakers (upper and lower limb of the first branchial arch), (3) pelvic fin rays, (4) scales in the lateral series, (5) scales above the lateral line, (6) pyloric caecae, and (7) vertebrae (Mamell et d. 1987, Behnke 1992). Three independent readers (all fisheries biolo- gists with training in anatomy and taxonomy of salmonids) counted each meristic structure on the same 50 cutthroat trout (> 150 mm total length) chosen randoniK from 9 of the 18 1996] Variation in Yellowstone Cutthroat Tkolt 303 Tahle 2. Alleles at the 10 diagnostic loci that distin- guish Yellowstone cntthroat trout and rainbow trout along with the tissue needed for each. The most coninion alk'le existing at each loci is listed fust. Char icteristic alleles Locus Ysc: KB'r Tissue SAAT-1* 165 100,0 Liver CK-A2* 84 100 Muscle CK-Cl* 38 100,150,38 Eye mIDHP-1* 75 100 M uscle sIDHP-1* 71 100,114,71,40 Liver sMEP-1* 90,100 100 Muscle sMEP-2* 110 100,75 Liver PEPA-1* 101 100,115 Eye PEPB* 135 100 Eye PGM-1* null lOO.null Muscle streams (Table 1) 3 different times to assess repeatabilit)' and \'ariation of connts within and among indi\'idnal readers. One reader counted tlie 7 meristic features on 125 additional cut- throat trout to determine mean counts for each structure and allow comparison among the 12 sampling sites where > 5 fish were counted (Table 1). The initial count from this reader's original 50 fish was also included in the analy- sis, leading to a sample of 175 cutthroat trout. All counts were done on the right side of each cutthroat trout. Scales in the lateral series were counted 2 scale rows above the lateral line starting at the opercle opening and contin- uing to the insertion of the caudal fin, while scales above the lateral line were counted from the anterior of the dorsal fin on a vertical diag- onal down to the lateral line. Vertebrae were counted during dissection of the fish. Pyloric caecae were enumerated by stretching the stomach and counting caeca ends. Meristic fea- tures were counted under a dissecting micro- scope using 30X magnification and reflected light. Readers practiced the protocol and com- pared results to resolve procedinal differences before initiation of counts. All fish were counted at similar times by each reader with several different cutthroat trout counted between sub- sequent counts. Three-way analysis of variance (ANOVA) was used to assess differences in counts of meristic features among (1) readers, (2) read- ings by individual readers, and (3) sampling sites. The sampling site effect was then con- trolled for and a 2-way ANOVA was used. One-way ANOVA was used to compare counts among readers and sampling sites. Tukey's multiple comparison test was used to make pairwise comparisons if significant differences were found. Statistical analyses were per- formed using SPSS/PC+ (SPSS Inc. 1991). Significance was determined at P < 0.05 for all tests. Results and Discussion Cutthroat trout were present in all 23 study streams. Electrophoretic anaKsis of fish from 7 streams found no genes at diagnostic loci that identify niinliow tiout (Table 2). Because genetic samples were collected from sites most likely to contain rainbow trout alleles (e.g., streams stocked with rainbow trout), we considered all trout in the drainage to be pure cutthroat trout. The AK-1*333 allele is common among fine- spotted cutthroat tiout in the Snake Rivei* drain- age and was detected in 4 of the 7 samples (Table 1). This allele, while not unicjue to fine- spotted cutthroat trout, is rare in Yellowstone cutthroat trout populations outside the Snake River drainage; its presence indicates possible integration with finespotted cutthroat trout. An ANOVA showed no consistent difference in counts for any of the 7 meristic features between fish from sites potentially hybridized with fine- spotted cutthroat trout and those considered pure Yellowstone cutthroat trout. Additionally, Behnke (1992) stated that meristic counts of finespotted and Yellowstone cutthroat trout are indistinguishable, and there is considerable debate as to whether finespotted cutthroat trout are a formal subspecies. Therefore, we did not differentiate between finespotted and Yellowstone cutthroat trout in our analysis. No significant differences among counts by the same reader for any meristic feature were obsei-ved. All 3 readers had high agreement among multiple counts for each structine (Tiible 3). Significant differences in mean counts among different readers were observed for all struc- tures except gillrakers (Tables 4, 5). All 3 read- ers had significantly different mean counts of pyloric caecae, pelvic fin rays, and scales above the lateral line, while at least 1 reader was sig- nificantly different from the other 2 readers in mean counts of vertebrae, basibranchial teeth, and scales in the lateral series. Hubert and Alexander (1995) also found poor agreement 304 Great Basin Naturalist [Volume 56 Table 3. Significance values for differences in mean meristic counts among 3 readers (RDR), 3 readings by individual readers (RUN), and sampling site (SITE). Main effects Interactions Structiue RDR RUN SITE RDRxRUN RDR X SITE RUN X SITE RDR X RUN X SITE Pvloric caecae 0.000 0.903 0.000 1.000 0.000 1.000 1.000 Vertebrae 0.000 0.819 0.061 0.757 0.047 0.997 1.000 Pelvic fin ravs 0.000 0.996 0.012 0.794 0.000 1.000 1.000 Gillrakers 0.765 0.356 0.244 0.352 0.045 0.098 0.051 Basibranchial teeth 0.448 0.945 0.000 0.952 0.323 1.000 1.000 Scales in lateral series 0.000 0.939 0.000 0.989 0.000 1.000 1.000 Scales above lateral line 0.000 0.986 0.000 1.000 0.000 1.000 1.000 Table 4. Significance values for the difference in mean meristic counts among 3 readers (READER) and among 3 readings by individual readers (RUN) at 5 sampling sites. effects Structure Site READER RUN Interaction 0.998 1.000 0.993 0.808 0.S60 1.000 0.932 0.972 0.999 0.998 0.812 0.984 0.618 0.561 0.887 0.918 0.886 0.969 0.849 0.969 0.802 0.924 0.628 0.882 0.880 0.924 0.621 0.435 1 .000 1,000 1.000 1. 000 0.815 0.992 0.871 0.492 0.981 0.881 0.938 0.880 0.878 0.995 0.683 0.902 0.975 0.907 0.889 0.990 0.907 0.886 0.951 0.932 0.860 0.818 0.431 0.535 0.879 0.905 0.975 0.999 0.886 0.973 0.888 0.843 0.712 0.815 0.885 0.885 0.644 0.694 Pvloric caecae Vertebrae Pelvic fin ravs Gillrakers Basibranchial teeth Scales in lateral series Scales above lateral line Anderson Brovvn SFWood Venus Wood Anderson Brown SF Wood Venus Wood Anderson Brown SF Wood Venus Wood Anderson Brown SFWood Venus Wood Anderson Bro\vn SFWood Venus Wood Anderson Brown SFWood Venus Wood Anderson Brovvn SF Wood Venus Wood 0.0S3 0.000 0.108 0.227 0.000 0.019 0.000 0.153 0.016 0.226 0.000 0.005 0.000 0.003 0.000 0.596 0.737 0.001 0.400 0.055 0.728 0.000 0.142 0.064 0.090 0.001 0.000 0,000 0.000 0.000 0.000 0.000 0,000 0.000 o.ooo I99(ij Variation in Yellowstone Cutthroat Trout 305 Table 5. Variation in mean meristic counts and standard deviations (in paren theses) of 3 readers. Means not signifi- cantly different indicated In hold (Tiike\'s /'< 0.05). Reader Structure 1 2 3 r Pyloric caecae 32.7 (6.3) 36,9 (9.5) 41.0 (11.7) < 0.0001 Vertehrae 60.5 (L6) 59.5 (2.0) 59.3 (1.2) < 0.0001 1\'!\ ic tin ra>s 9.0 (0.4) 8.8 (0.4) 9.4 (0.6) < 0.0001 Gillrakers 18.9 (L6) 18.8 (1.3) 19.3 (10.8) 0.83 Basihranchial teeth 13.7 (4.2) 15.3 (4.3) 14.2 (4.2) 0.003 Scales in later il series 178.0 (14) 187.5 (14) 187.4 (13) <0.0001 Scales ahoNC 1 iteral lini' 44 (4.2) 56.4 (5.2) 42.5 (3.6) <0.0001 Table 6. Mean meristic counts and standard deviations (in parentheses) for 175 fish 1)\ 1 reader with ranges among the 12 sample sites with >5 fish counted. A probability (P) of <0.05 indicates significant differences among sites. Structure Grand mean (s) Range in means among sites Pyloric caecae Vertebrae Pelvic fin rays Gillrakers Basibranchia! teeth Scales in lateral series Scales ab()\e lateral line 42.29 (10.89) 58.57(1.39) 9.23 (0.86) 18.80 (2.08) 13.96 (5.45) 182.70 (14.77) 40..39(3.51) 29.9-51.4 57.9-60.6 9.0-9.9 17.8-19.9 11.4-21.8 175.5-207.3 37.1-45.5 < 0.0001 0.()()()2 0.0001 0.0018 0.0025 < 0.0001 0.0001 among readers when counting meristic fea- tures of rainbow trout. Significant differences were observed in counts of meristic features among fish fi-om 12 streams (Tables 3, 6). Meristic features may be environmentally controlled within specific areas or drainages (Barlow 1961, Rinne 1985, Currens et al. 1989), but environmental vari- ables measured at each sampling site (eleva- tion, gradient, and stream size) were not corre- lated with meristic counts in the Greybull River drainage (Kruse 1995). Researchers have used meristic counts with varied success to identify subspecies of cut- throat trout (Loudenslager and Kitchen 1979, Loudenslager and Gall 1980, Marnell et al. 1987). Recent research has shown that meristic comparisons can provide potentially mislead- ing information (Busack and Gall 1981, Leaiy et al. 1984, 1985) because meristic characteris- tics are often specific to localized populations (Behnke 1992) and are strongly influenced by genetic variation (Leaiy et al. 1991). Behnke (1992) described typical meristic counts for Yellowstone cutthroat trout and rainbow trout (Table 7). Mean counts of meris- tic features of cutthroat trout from the Grey- bull River drainage (Tables 5, 6) were within ranges for Yellowstone cutthroat trout (Table 7); however, mean counts of pyloric caecae, vertebrae, and gillrakers were also within typi- cal ranges for rainbow trout. Variation and sim- ilarit)' in counts of meristic features of Yellow- stone cutthroat trout and rainbow trout make it difficult to determine species or hybrids using meristic counts alone. Only the presence of basihranchial teeth provided a distinction be- tween the 2 species. Variations among readers, and among sam- pling sites in a small geographic area, along with relatively wide ranges in counts for Yel- lowstone cutthroat trout and rainbow trout, make it difficult to differentiate these 2 species with certainty using commonly assessed meris- tic features (Tible 7). Furthermore, it is unlikely that Yellowstone cutthroat trout X rainbow trout hybrids can be identified due to the extensive variation in counts. 306 Great Basin Naturalist [Volume 56 Table 7. Ranges of meristie counts among species (YSC = Yellowstone cutthroat trout and RBT = rainbow trout), readers, and sampling sites. YSC /' RBT' Variation Vuiation among Variable Typical Overall Typical Overall among readers'^ sampling sites'^ Pyloric caecae 35-43 25-50 37-55 30-70 33-41 (36.9) 30-51 Vertebrae 61-62 60-63 62-64 61-66 59-61 (59.8) 58-61 Pelvic fin ravs 9 9-10 not reported 9 (9.0) 9-10 Gillrakers 19-20 17-23 19-21 17-24 18-21 (19.0) 18-20 Basibranchial teeth present present 14-16(14.4) 11-22 Scales in lateral series 165-180 150-200 125-150 120-160 179-188 (184) 176-207 Scales above lateral line 45-50 40-55 30-32 26-35 42-57 (47.6) 37-46 •'From Behnke (1992) ''Ranges are from the 9 readings taken for each strncture with means in parentheses (3 readings by 3 readers). '^Ranges are from means for the 12 sampHng sites that had > 5 cutthroat trout (> 150 mm total length) counted (Table 6). Acknowledgments We thank C. Ewers, C. Griffith, K. Harris, K. Krueger, and M. Wilhams for field assis- tance; E Rosebeny and R. Ziibik for technical support; K. Krueger and D. Simpkins for serv- ing as readers; and R. Behnke, B. Shepard, R. Wiley, and S. Yekel for critical review of the manuscript. This project was supported by the Wyoming Game and Fish Department and the U.S. Forest Sei"vice. Literature Cited Allendorf, K W., a.nd R. E Leary. 1988. Conservation and distribution of genetic variation in a polytypic species, the cutthroat trout. Consei"vation Biology 2: 170-184. Allendorf, E W, and S. R. Phelps. 1980. Loss of genetic variation in a hatcheiy stock of cutthroat trout. Trans- actions of the American Fisheiy Society 109; 537-545. AviSE, J. C. 1974. Systemic value of electrophoretic data. Systemic Zoology 23: 465—181. Ayala, E J., AND J. R. Powell. 1972. Allozymes as diagnos- tic characters of sibling species of Drosophila. Pro- ceedings of the National Academy of Sciences, USA 69: 1094-1096. Barlow, G. W. 1961. Causes and significance of moipho- logical variation in fishes. Systematic Zoology 10: 105-117. Behnke, R. J. 1992. Native trout of western North Amer- ica. American Fisheries Society Monograph 6, Betliesda, MD. Bu-SACK, C. A., AND G. A. E. Gall. 1981. Introgressive hybridization in a population of Paiute cutthroat trout {Salmo clarki .seleniris). Canadian Joimial of Fisheries and Aquatic Sciences 38: 939-951. Currens, K. R, C. S. Shahpe, R. Hjort, G. B. Schreck, AND H. W Ll. 1989. Effects of different feeding regimes on the morphometries of chinook salmon (Oncorhynchtts fsliaici/tsclia) and rainbow trout (O. iwikiss). Gopeia 1989: 689-695. Gresswell, R. E. 1988. Status and management of inte- rior stocks of cutthroat trout. American Fisheries Society Symposium 4. Hubert, W. A., and C. B. Ale.v\nder. 1995. Observer variation in counts of meristie traits affects fluctuat- ing asymmetry. North American Journal of Fisheries Management 15: 156-158. Kruse, C. G. 1995. Genetic purity, habitat, and population characteristics of Yellowstone cutthroat trout in the Greybull River drainage, Wyoming. Unpublished master's thesis, University' of Wyoming, Laramie. Leary, R. E, F W Allendorf, and K. L. Knudsen. 1985. Developmental instability and high meristie coimts in interspecific hybrids of salmonid fishes. Evolution .39: 1318-1326. . 1989. Genetic divergence among Yellowstone cut- tliroat trout populations in the Yellowstone River drain- age, Montana: update. Population Genetics Labora- ton Report 89/2, Unixersity of Montana, Missoula. . 1991. Effects of rearing density' on meristics and developmental stability of rainbow trout. Copeia 1991: 44-49. Leary, R. E, E W Allendorf, S. R. Phelps, and K. L. Knudsen. 1984. Introgression between westslope cutthroat trout and rainbow trout in the Clark Fork River drainage, Montana. Proceedings of the Mon- tana Academy of Sciences 43: 1-18. . 1987. Genetic divergence and identification of seven cutthroat trout subspecies and rainbow trout. Transactions of the .\merican Fisheries Society 116: 580-587. LouDENSL.\(,ER, E. J., AND G. A. E. Gall. 1980. Geogi-aphic patterns of protein variation and subspeciation in cutthroat trout. Sdliiw chirki. S\stemic Zoology 29: 27-12. LouDENSLAGER, E. J., AND R. M. KiTCHEN. 1979. Genetic similarity between two forms of cutthroat trout, Salmo clarki, in Wyoming. Copeia 4: 673—678. \L\HNELL, L. F, R. J. Behnke, and F W Allendorf. 1987. Genetic identification of cuttliroat trout, Salmo clarki, in Glacier National Park, Montana. Canadian Journal of Fisheries and A(}uatic Sciences 44: 1830-1839. Nielsen, J. D. 1995. Exolulion and the aciuatic ecosystem. American Fishi-ries Socii'tx Symposium 17. 1996] Variation in Yellowstone CuitiihoatTrolit 307 Perkins, D. L., C. C. Kiuecer, and li. May. 1993. Her- SPSS Inc. 199L The SPSS guide to data analysis for itage brook trout in nortlieastem USA: genetic vari- SPSS/PC-(-. 2nd edition. SPSS Incorporated, Cliicago, ability within and among popnhition.s. Transactions of IE. the American Fisheries Society 122: 515-531. RiNNE, J. N. 1985. Variation in Apaelie trout popuhitions in Received 10 April i.996 the White Mountains. Arizona. North American Accepted 2fi Aii-imt 1996 h)unial ol Fisheries Management 5: 146-158. Great Basin Naturalist 56(4), © 1996, pp. 308-318 STUDIES OX XEARCTIC XEGASTRIUS (COLEOPTEK\: EL.\TERIDAE) Samuel A. Wells^ Abstract. — New species descriptions of Xegastrius rupicola from California. Oregon, \\ashington, and British Columbia: .V. stibicki from California. Mont;ma, imd British Columbia: .V. solox from .\rizona and New Mexico: and .V. atrosus from Ontario and Quebec are gi\'en. Negastrius colon is returned to species status, and a neot>pe is designated for A^. choris. Fleutiaiixellus extricatits is a new combination. A ke\' is proNided to Nearctic species of Xegastrius. Key words: Xegastrius. Elateridae. holotijpe. parafype. neotype. History Xegastrius was established in the famiK Elateridae b\ Thomson (1859) to distinguish those species of Cn/ptohypmis Eschscholtz ha\ing arcuate prosteniiil sutures from species with straight or double sutin-es. Candeze (1860) did not use Thomson s assignments and placed all Negastrius species in Cryptohypnus. Honi s (1891) monograph of the species of Cryptolujp- nus of Boreal America rejected die name Xegas- trius and included all North American forms into 9 groups within the genus Cryptohypnus. Horns ehoris group included \. delumhis (Horn), xV. choris (Say), .V. exiguus (Randall), and A", ornatus (LeConte), which were equi\ a- lent to Thomsons Xegastrius. Schwarz (1906) included Xegastrius and Cryptohypnus with the genus Hypnoidus Stephens in the tribe H>pnoidini. Leng (1920) also placed iill species, e.\cept .V. e.xiguus. in the genus CryptoJnjpnus. Using mesostenial chai-acters, Xakane and Kisliii (1956) made die distinction bet\veen the subfamilies Xegastriinae and H>politliinae (which die\ s\"non\iiiized under the Cteniceri- nae). Aniett (1963) recognized onl\- the genera Xegastrius and Oedostethus LeConte in the Xegastiiinae from Xoitli .America. Stibick (1971) recognized or established Xeohypdonus Stibick, Migiwa Kisliii, OedostetJuis, FleutiauxeUus Mequignon, Xegastrius. Zorochrus Thomson, and Paradonus Stibick from Xorth America. He restricted Xegastrius to diose species with coarse pronota, single prosternal sutures, and species with the 2nd and 3rd antenual seg- ments equi\alent in length. Later, Kishii (1976) erected Microlu/pnus. to which Stibick (1991) assigned the single Xorth .American species of M. striatidus (LeConte). Discussion of Char.\cters With the exception of die Cardiophorinae, the subfamiK" Xegastriinae is distinguished from odier subfamilies of Elateridae b\' liaxing the nieso- and metastenia adjacent and sepa- rating the mesocoxal ca\it> from the niese- pimeron and mesepistenium. The Xegastri- inae is distinguished from the Cardiophorinae by possession of a pointed prosternal process, which is shortened and tianicated in the Car- diophorinae. Within the Xegastriinae, Xegas- trius is apparentK most closeh" related to the genus Microhypnus, bodi genera haxing a strigate and/or rugose pronotum. Following Stibick s (1971) presumed natural afrinities, die sister group of Xegastrius could be mi) of die Xorth .American genera, e.xcept Paradonus, w liicli is more closeK related to the Old World species of Thurana Stibick and Optitarynus Stibick, both of w liich are without e.xtemall>' \isible el\ tnil striae. Zivochrus is distinguished from Xegastrius b\ die double prosternal sutm-es and/or b\" the arcuateK" extended pronotum that projects o\er die head. In addition, the pronotum in Zorochrus is more coiU'seK gi-anu- late on the anterior half Fh'utiau.xeUus differs from the other genera of Xegastiiinae b\ lia\ - ing tlie 3rd antennal segment nearly twice as long as the 2nd. The genus Xeohypdonus is 'Biosvs Inc.. 10150 Old Columbia Rixid, Columbia. MD 2104(>-17O4. 308 1996] Studies ox Nearctic Negastrics 309 separated from Xi'fiastrin.s b\- a smooth to sliglitK punctate pronotum that is often shiny (^^'ells 1991). Oedostethiis is chstinguished from Xegastriiis h\' ha\in'pe. Fall's description of N. extricatus is clear on several points. The api- cal 3 segments of the antennae extend past the hind angles of the thorax whereas N. nadezha- dae has the antennae not attaining the hind angles. Fall also refers to an impressed vertex that is absent in all Negastrius species. Several specimens from Cornell (taken from the type locality of F. extricatus), the USNM, Chicago Field Museum, and the Canadian National Collection, all of which were taken from Alaska, fit Falls description on all points and have the 2nd segment of the antennae reduced; this is indicative of the genus Fleutiaiixellus. Acknowledgments I thank the following institutions and indi- viduals for providing material: G. E. Ball, Strick- land Museum, University of Alberta; E. C. Becker, Canadian National Collection; R. W. Brooks, Snow Entomological Museum, Univer- sity of Kansas; S. D. Cannings, University of British Columbia; J. A. Chemsack, University of California, Berkeley; D. A. Kavanaugh, Cali- fornia Academy of Science; J. K. Liebherr, Cornell University; K. C. McGiffin, Illinois Natural History Sui-vey; C. A. Olson, University of Arizona; E. Riley, Texas A & M University; C. Salvino, Chicago Field Museum; S. R. Shaw, Museum of Comparative Zoolog)', Hai-vai'd Uni- versity; J. N. L. Stibick, USDA,' APHIS; C. A. Triplehom, Ohio State University; and the late D. R. Whitehead, National Museum of Natural Histoiy, Smithsonian Institution. I also thank the Department of Zoology at Brigham Young University and the D Elden Beck family for financial support, and Richard W. Baumann, C. Riley Nelson, and Michael F Whiting for assistance in collecting specimens throughout the western United States and Canada. Special appreciation is extended to J. N. L. Stibick, E. C. Becker, Paul Johnson, and Stephen L. Wood for editorial suggestions. Literature Cited Arnett, R. H., Jr. 196.3. Fascicle 46, Elateridae in: The beetles of the United States. Catholic Uni\ersit>' of America Press, Washington, DC. Becker, E. C. 1977. New and noteworthy records of Coleoptera in Canada (1). Annales de la Societe Entomologique dn Quebec 22:14-17. Blatchley, W. S. 1910. Coleoptera or beetles known to occur in Indiana. Indianapolis. 1.386 pp. 318 Great Basin Naturalist [Volume 56 Brooks, A. R. 1960. Adult Elateridae of southern Alberta, Saskatchewan and Manitoba (Coleoptera). Canadian Entomologist, Supplement 20: 5-63. Candeze, E. 1860. Monographie des Elaterides III. Liege. 512 pp. Dietrich, H. 1945. The Elateridae of New York State. Cornell University Agricultural Experiment Station Memoir 269. 79 pp. DOLIN, V G. 1971. New species of click beetles (Cole- optera: Elateridae) from the Soviet Union. Entomo- logical Re\'iew 50: 362-370. Fall, H. C. 1926. A list of the Coleoptera taken in Alaska and adjacent parts of the Yukon Territoiy in the sum- mer of 1924. Pan-Pacific Entomologist 2: 127-154, 191-208. Fattig, P W. 1951. The Elateridae or click beetles of Geor- gia. Emeiy University Museum Bulletin 10. 25 pp. Horn, G. H. 1891. A monograph of the species of Crypto- hypnus of boreal America. Transactions of the Ameri- can Entomological Society 18: 1-29. KiSHll, T. 1976. New Nega.strius with some notes. Bulletin of Heian High School 20:17-46. Lane, M. C. 1971. Family Elateridae in: M. H. Hatch, The beetles of the Pacific Northwest, part V. University of Washington Press, Seatde. 662 pp. LeConte, J. L. 1853. Revision of the Elateridae of the United States. Transactions of the American Philo- sophical Societ>' 10: 405-508. Leng, G. W. 1920. Catalogue of the Coleoptera of America, north of Mexico. John D. Sherman, Jr, New York. 470 pp. Nakane and Ki.shii. 19.56. On the subfamilies of Elateri- dae from Japan. Kont\'u 24: 201-206. R\NDALL, J. W. 1838. Description of new species of cole- opterous insects inhabiting the State of Maine. Boston Journal of Natural Histoi-y 2: 1-52. Say, T 1839. Descriptions of new North American insects, and obseiAations on some alread\- described. Trans- actions of the American Philosophical Societv 20: 155-190. SCHWARZ, O. 1906. Elateridae, Genera Insectorum. Riscicle 46. SCHENKLING, S. 1925. Ill: Junk: Goleopteromm Catalogus 81:212-215. W. Junk,' Berlin. Stibick, J. N. L. 1971. The generic classification of the Negastriinae (Coleoptera: Elateridae). Pacific Insects 13: 371-390. . 1991. North American Negastriinae (Coleoptera, Elateridae): the Negastriinae of the northeastern United States and adjacent Canada (Coleoptera: Ela- teridae). Insecta Mundi 4 (1-2): 1-;31. Thomson, G. G. 1859. Coleoptera Scandinaviae 1. Wells, S. A. 1991. Two new species of Neohypdonus (Coleoptera: Elateridae) from North America widi a key to Nearctic species. Entomological News 102(2): 73-78. Received 15 November 1994 Accepted 4 October 1996 \ Gveal Basin Naturalist 56(4), © 1996, pp. 319-325 BIGHORN SHEEP RESPONSE TO EPHEMERAL HABITAT FIUGMENTATION BY CATTLE J. A. Bissonette' and M(>laiiic" J. Steinkampl'2 ABSThL\CT. — Wt' studied Sfasoiial tattk' mazing as an agent of eplieuierai habitat liagineiitation on a newly reintro- duced popidation of California bighorn sheep (Ovia canadensis californiana) in Big Cottonwood Canyon, Idaho, 1988-89. W'e evaluated the In pothesis that bighorn sheep avoid cattle. We dociunented sheep response to the pro.ximity to cattle by direct observation. The core areas used by bighorn and distances to escajje tenain generally decreased as cattle moved closer to sheep. Likewise, sheep moved from cattle as cattle approached them. Severity of response we obsened is in marked contrast with that reported for established bighorn populations, suggesting that newK' reintro- duced bighorn sheep are more highly sensitive to the presence of cattle. Krij words: hifjiorn slicep, cattle, disturbance, Idaho, Ovis canadensis. Prior to the 20th century, CaHfornia big- horn sheep were abundant in montane regions of the western United States (Van Dyke et al. 1986). However, since 1840 population num- bers of bighorn sheep and their area of distrib- ution have decreased (Cowan 1940, Buechner 1960). Disease, excessive hunting, activities associated with mining, human disturbance, and pressure from hvestock for resources and space reportedly contributed to the extiipation of the subspecies from most of its range (Smith 1954, Geist 1971, Graham 1971, Demarchi and Mitchell 1973, Demarchi 1975, Trefethan 1975, Van Dyke 1978, Smith et al. 1988). California bighorn sheep were once abun- dant in parts of southwestern Idaho; the last observations were recorded during the 1920s (Hanna 1978). The Idaho Department of Fish and Game (IFG) initiated reintroduction pro- grams of returning California bighorn to parts of their historic range in 1963. Thirty-eight sheep from the Chilcotin River herd in British Columbia were transplanted into the drainages of the East Fork of the Owyhee River between 1963 and 1966 and have provided a base for subsequent reintroductions. In 1967, 12 addi- tional bighorn were reintroduced into the nearby Little Jack's Creek drainage. Both pop- ulations were allowed to expand until 1980 (Toweill 1985). From 1980 to 1989, >100 sheep were relocated to 5 different regions in south- ern Idaho. Livestock pressures have been heavy on rangelands in the western United States that historically supported populations of bighorn sheep (Mackie 1978). Seventy percent of the public land area in the 11 westernmost states is grazed at least seasonally. Within Idaho range- land conditions varied. In 1986 surveys from the Owyhee range in Idaho reported 57% of the range in poor condition, 35% fair, and only 5% in good condition (Bureau of Land Man- agement, Owyhee rangeland program sum- mary, Burley District, ID, files, 16 pp., 1986); while in 1982, 30% of the range was in poor condition, 57% fair, and 18% in good condition (Bureau of Land Management, Twin Falls, land use decisions summaiy and rangeland program summary, Burley District, ID, files, 26 pp., 1982). Peiper (1988) reported that improve- ment in range condition has been slow since 1973. Bighorn sheep are more sensitive to land uses associated with development than most native ungulates (Andryk and Irby 1986). Addi- tionally, bighorn sheep are comparatively less abundant, react adversely to disturbance, and occupy habitats sensitive to change (Van Dyke et al. 1986). Livestock activities on these sites can negatively affect sheep through resource exploitation (i.e., forage, space, cover, water) or behaviorally (Geist 1971). On shared ranges social intolerance may impose greater limita- tions on distribution and habitat use of bighorn 'U.S. National Biological Service, Utah Cooperative Fish and Wildlife Research Unit, Department of iMslieries and Wilclliic. College of Natural Resources, Utah State University, Logan, UT 84321-,5290. ^Present address: U.S. Fish and Wildlife Service, Box 2676, Vero Beach, FL .32961. 319 320 Great Basin Naturalist [Volume 56 than competition for forage; however, biologists disagree whether livestock impact bighorn sheep spatial boundaries, limiting distribution. Wilson (1975) and Van Dyke et al. (1986) re- ported that bighorn show aversions to cattle and avoid them when unaccustomed to their presence on the range (Drewek 1970, Kornet 1978), while others did not detect reactions between sheep and cattle (King 1985, King and Workman 1985). Analyses that test the avoidance of livestock by bighorn sheep are limited. Habitat fragmentation theoiy has applica- tion to seasonal livestock grazing. Habitat frag- mentation may be permanent (e.g., subdivision constmction) or ephemeral, as in seasonal lixe- stock grazing. Effects of pemianent fragmenta- tion on habitat use have received increasing attention in recent > ears; ho\\ e\er, less is under- stood about effects of seasonal fragmentation. We postulated that areas used b\' bighorn sheep are fragmented during spring and summer by cattle on grazing allotments. An area ma\' appear large but, due to fragmentation, ha\'e a much smaller useable area. If bighorn sheep a\'oid livestock, the area available to them is reduced temporarily as lixestock graze seasonally in sheep habitat, resulting in sheep exclusion fiom areas of potential use. A population may be in- fluenced as sheep are restricted to smaller patclies of habitat and effects of densit>' depen- dence are felt. In our study we wanted to detennine whether avoidance occurs, assess its effect on habitat use by sheep, and consider how a\'oidance, if it occurs, might influence future decisions for reintroductions. Study Area We conducted the studv' in Big Cottonwood Canyon 16 km northwest of Oakle>' (Cassia Co.), Idalio. The canyon is approximately 18 km long, with Cottonwood Creek flowing to the northeast through the canyon bottom. Eleva- tion of the canyon floor increases gradually from 1400 to 2100 m. Average elexation gain from the canyon floor to the mesa top is 365 m. Canyon walls are steep and characterized by a combination of cliffs, boulder slopes, grass, and shrub slopes. Woody vegetation includes four- \ving salt brush {Atn})Icx cancsccns), spin\' hop- sage [Grayia spinusa), low sage [Aiieiuisia arhii- sciila), horse brush [Tetradymia canescens), rabbit brush (Chn/sothamniis nauseosiis). blue- bunch wheatgrass {Agropyron spicatuin), and juniper {Jiinipenis occidentaUs). Big Cottonwood Canyon lies within the Saw- tooth National Forest and contains a cattle graz- ing allotment that is leased fi-om late Ma\' until earh' October This grazing allotment consists of 5 pastures managed on a reverse-rotation basis and supports 400 cows with calves. Mesas south of the canyon contain another allotment of 3 pastures; this allotment is managed on a deferred-rotational system with 100 cows with calves. Permit dates for the Big Hollow allot- ment are late May to late October. Methods Thirt\'-seven California bighorn sheep (19 with radio-collars, 18 with patteni-coded col- lars) were released into Big Cottonw ood Canyon b\' the Idaho Department of Fish and Game during December 1986, December 1987, and Noxember 1988. Collars marked with different designs in pennanent ink allowed us to distin- guish between non-transmittered indi\'iduals. The population at the beginning of our 1st summer field season (1988) was 23, 13 from the 1st reintroduction in 1986 and 10 fi"om the 2nd in 1987. Fourteen additional sheep were released in November 1988. We recorded daiK- locations of bigliom sheep by visual obsenation fi-om May to September 1988 and June to September 1989. Telemetiy was used onl>' to aid in locating radio-collared bigliom sheep. We conducted weekly visual sur- veys to locate an)' uncollared sheep not close to collared indi\iduals. Sheep were viewed fi-om > 500 m using a spotting scope to reduce chance of detection and disturbance. If we were detected and sheep moxement followed, we disregarded subsecjuent obsenations of those indi\ iduals for die remainder of the day. Every effort was made to identify individuals within groups. \\'e determined indi\iduals by collar design or by telemetn' frequency. Loca- tions were recorded in Universal Transverse Mercator (UTM) coordinates. For each location we recorded group size and composition. We defined escape terrain as broken habitat on which mountain sheep max* safel)' outma- neuver or outdistance predators (Gionfriddo and Krausman 1985). SpecificalK; escape ter- rain ma\ be characterized b\- a ruggedness index as defined by Beasom et al. (1983), and terrain class and number of cliff faces > 120% 1996] Sheep Response to Fragmentation 321 following Krausnian and Leopold (1986). For every location we nieasnred distance to escape terrain using a range Finder once sheep left the area. We determined slope with a clinometer. We located cattle by hiking a systematic route on foot 3-4 times/wk. With the exception of group composition, data lecorded for each cat- tle location were identical to sheep locations. We recorded cattle and sheep locations simul- taneously iillowing sheep movements to he anal- yzed in response to cattle mo\'ement for that specific time. Data not taken during identical time periods were not used in paired analyses. Even though a controlled test was not possi- ble, we wanted to obsene tlie response of sheep when livestock were in proximity to sheep. On 14 August 1989, 5 cows were moved directly into an area of continuous sheep use and held continuousK' for 40 h. Cattle were kept within approximately a 0.8-km- area by 2 cowboys. Sheep response was observed and recorded. Cattle were watered every 5 h by removing them from the group one at a time and taking them to a trough in the bordering pasture. After 40 h all cattle were removed. We located sheep daily for the next 10 d. We combined individual bighorn sheep loca- tions for each group for analysis with Program Home Range (Samuel et al. 1985); thus, each location represented a group of bighorn sheep, not an individual. We used 95% haniionic mean measures of activity to estimate home ranges and core areas. We defined core areas as the maximum area where the obsei-ved utilization distribution as detemiined from the harmonic mean values was greater than a uniform uti- lization distribution (Samuel et al. 1985). Kol- mogorov's test was used to determine if ob- sei-ved use was significantly (P < 0.05) greater than expected. All comparisons were consid- ered significant at the 0.05 level. All data points were plotted at a scale of 1:12,000. We recognize that harmonic mean measures have been criticized. Naef-Daenzer (1993) tested the spatial resolution of the conventional har- monic mean measure and a bivariate normal kernel estimator with a new kernel estimator he developed. The harmonic mean estimator generalized the distributions of 2 parallel gra- dients and estimated density at higher than zero for areas containing no sample points. Worton (1989, 1995) and Boulanger and White (1990) have outlined some undesirable proper- ties of harmonic mean measures that were eliminated from kernel estimators using appro- priate smoothing techni(iues. Specifically with the harmonic measure, estimates of zero area can occur, and isopleths may include areas witli no sample points (Worton 1995). We had no estimates of home range or core areas that approached or even came close to zero. Addi- tionally, the isopleths we generated were based on tightly grouped locations of sheep, thus avoiding the problem ol' areas with no sample points. Finally, we did not employ interstudy comparisons, thus avoiding the onerous prob- lem of comparing between methods, thereby reducing the effect of inherent bias. We plotted mean monthly home ranges and core areas of sheep and cattle and then over- laid them to determine changes in size and location between consecutive months. We measured avoidance by quantifying changes in size and location of bighorn sheep range and core areas as cattle moved through bighorn sheep habitat. Changes in location were deter- mined from harmonic means. We compared data collected during the 1st and 2nd field sea- sons to determine whether range and core areas were related to seasonal changes. We calculated daily distances between big- horn sheep and cattle using UTM location co- ordinates. We defined consecutive locations as locations taken 1 d apart. Only cattle and big- horn sheep paired locations recorded at the same time were analyzed. Simple linear regres- sions were used to test for associations between 3 variables: distance (m) between cattle and big- horn, distance sheep moved in response, and distance fiom location of sheep to escape ter- rain. First, we tested sheep response to prox- imity of cattle; then we tested to determine whether distance between sheep and escape terrain was related to proximity of cattle. Results Response of Bighorn Sheep to Cattle Sheep range size did not change signifi- cantly in size or location (P < 0.05) from June to July in 1988 or 1989. Cattle were in adjacent pastures but because of topography were usu- ally not visible to sheep or the observers. Dur- ing August 1988, when cattle were moved to an allotment adjacent to areas receiving high sheep use, home range position shifted and range size decreased (Table 1). In September sheep expanded their range, coincident with 322 Great Basin Naturalist [Volume 56 Table 1 Sp atial responses of bighorn sheep in Little Cottonwood Canyon Idaho, to the pro.ximity of cattle. Range Core area Mean chstanee (in) size Size % use % area^ c-s'' e-t^- Sheep'' Date (kni2) (kni2) 6/88 13.4 4.3 61.4 32.1 4019 101 1616 7/88 13.7 4.7 53.9 27.3 4045 86 1246 8/88<^ 5.0 1.5 59.0 42.9 2251 55 1046 6/89 13.4 4.7 57.5 40.0 4820 112 1698 7/89 13.5 1.5 67.0 40.0 5148 63 1008 8/89^ 7.2 1.5 0.5' 55.6 40.0 3346 56 11 1276 ■'Percent >l total Ik nie range ai ■a that c ore an a encompasses ''Mean daiK distance sheep nii >\ ed dun ig the 1 lontli 'Wlean distance between cattle and big loni '"Cattle plac =d in allotments ck «e to she ep '-'Mean distance of sheep to escape terrain fpield experiment data the movement of cattle during late August into a pasture adjacent to a high use sheep area. Sheep tended to concentrate into smaller core areas in 1988 and 1989 as proximit)' to cattle decreased. No significant change (<3%) in core area of bighorn sheep occurred between July and August 1989 prior to moving cattle close to big- horn. When cattle were moved puiposefully to within 800 m, bighorn sheep responded by immediately vacating the area and creating a new distinct core area. Distances moved by bighorn sheep directly after movement of cat- tle into the sheep core area were 355% greater than daily sheep movements during early August (3000 vs. 845 m, respectively). Sheep remained together and stayed within 35 m of escape ter- rain for the following 9 d. This was the longest time period during the study that sheep re- mained within 35 m of escape tenain. Distances between cattle and bighorn sheep remained >4000 m for the following 5 d. Response of Bighoni Sheep Relative to Escape Terrain As mean daily distance between cattle and sheep decreased, the mean distance between sheep and escape terrain tended to decrease. Core-area size appeared to be directly related (adjusted r^ = 0.81) to distance to escape ter- rain (Fig. 1); the closer to escape terrain, the tighter sheep grouped together A correlation matrix, generated from these spatial data, adds further corroboration for the association (Table 2). The mean daily distance that bighorn moved dining the month was positively corre- lated (r^ = 0.88) with increasing distance of sheep to escape terrain. Discussion Hicks and Elder (1979) suggested that big- horn sheep were more likely to move greater distances when cattle were close, but were less likely to relocate when cattle were distant. Our data show increased movement by bighorn sheep as cattle mo\'ed closer When we moved cattle to within 800 m, bighorn left the area. Sheep response to cattle was much more ex- treme than at any other time or when com- pared to their behavior when confronted by humans at other times during the field season. We were unable to differentiate between the effect that cattle had and the potential effect of the personnel involved. We do not doubt that personnel moving the cattle had an effect. Fur- thermore, the presence of both cattle and per- sonnel close to sheep may well have augmented bighoni response nonlinearly However, at other times when we accidentally alerted sheep dur- ing the study [n = 10), bighorn responded by relocating much shorter distances (between 872 and 1190 m). Additionally, their response was t>q3ically short-lived and they left the prox- imity of escape cover by the next day or sooner. Although lx)th the proximity of cattle and per- sonnel influenced bighoni response, the impor- tant point is that extreme proximity evoked a higliK charged response. E\en without our intentional moxement of cattle toward sheep, their increasing affinity for escape cover as cat- tle moved closer suggests strongly that live- stock were perceived as a threat. Escape terrain is an important component of good sheep habitat (McQui\ e> 1978, Leslie and Douglas 1979, Weyhausen 1980, Krausman and Leopold 1986). We would lia\e predicted 1996] Sheep Response to Fragmentation 323 0 50 100 150 DISTANCE TO ESCAPE TERRAIN (m) Fig. 1. Relationship between size of core area of l)ishorn sheep, Cottonwood Canyon, Idaho, and distance to escape terrain, 1988-89. that tighter grouping should result as sheep moved farther from escape cover. However, our data show the direct opposite result, sug- gesting tliat when sheep move farther from escape terrain, they do so under less threaten- ing situations. Selective pressures under these conditions appear not to result in tighter groups. The response of bighorn sheep to cattle we obsened is in contrast with bighorn sheep in national parks. In some parks sheep approached humans closely and were photographed from car windows (Van Dyke et al. 1986). Smidi (1954) reported sheep eating from his hand, whereas others reported that sheep unaccustomed to people or cattle fled at the sight of humans or \ chicles >1600 m (Van Dyke et al. 1985). It appears that newly reintroduced sheep are more sensitive to disturbance, perhaps resulting from recent transplant activities, and react differ- endy dian do established, undisturbed j-jopula- tious. Sheep reintrockiced into Big (Jottonwood Canyon were net-gunned from helicopters, blindfolded, and flown to a base. They then had blood drawn, were given inoculations, weighed, measured, placed into the l)ack of a covered pickup with several conspecifics, and then transported approximately 160 kni and kept overnight in the vehicles. All were released the following moniing into an area foreign to them. As a result of exposure to such activities, any disturbance may more likely be viewed as a threat. In the Big Cottonwood Canyon popula- tion, alert-alarm behavior appears to be rein- forced yearly with each new group of reintro- duced animals. Age may also play a part; 55% of individuals released were <2 years of age. Heightened sensitivity and subsequent fre- quent reinforcement of alert behaviors appear to characterize the population and may be a general phenomenon for newly reintroduced populations placed into new areas. Sensitivity of these populations to disturbance may diminish over time as populations become estal^lished. Avoidance has implications for reintroduc- tions of bighorn sheep. The total area of poten- tial habitat may not be used by sheep if live- stock are present. If cattle allotments remain in use, it would appear wise to consider the possi- bility of ephemeral fragmentation by cattle when goals for desired bighorn population sizes are developed. Goals should be consistent with total useable habitat. Control of disturbance for recently reintroduced populations of bighorn sheep is certainly appropriate. Table 2. Correlation matrix for home rantie, core area, and mean distance variables for bighorn sheep in Bi wood Canyon, Idalio, 1988-89. Cotton- Range size Core area Mean distance (m) Size % use % area-* c-sl> e-f^- Sheep'' Range size 1.0 0.694 0.234 -0.601 0.887 0.721 0.440 Size 1.0 -0.380 -0.704 0.38.5 0.916 0.765 % use 1.0 0.335 0.410 -0.144 -0.272 %area 1.0 -0.220 -0.458 -0.266 c-s 1.0 0.520 0.308 e-t I.O 0.887 Sheep 1.0 ^Percent of total home range area that core area enconipassi "Mean distance between cattle and bighorn ■^Mean distance of sheep to escape terrain 324 Great Basin Natir.\i,ist [\'olunie 56 Acknowledgments We thank J. J. Beecham, \\; L. Bodie. T. C. Edwards, D. G. Oman, H. G. Hudak, E R. Krausman, R. B. Smith, D. E. Toweill, and E J. Urness (deceased) for tlieir ad\ice and help during the project. We also thank the Harold Cranneys who allowed us to use dieir land and utilities. We extend a special posthimius thanks to D. Balph for his insight and wisdom. We iilso thank the editor and associate editor of GBN and an anon>nious reviewer for constructive comments that helped us impro\'e the manu- script. This stud\ was fimded b> the Idaho De- partment of Fish imd Game dirough the Utah Cooperatixe Fisheries and Wildlife Research Unit (NBS) at Utah State Uni\ersit>-. We thank the United States Forest Senice for pro\iding aerial photographs and topographical maps. The research proposal was e\aliiated b\" the Animal Care Committee at Utah State for ad- herence to established animal cai-e guidelines. Data were collected follow ing acceptable field methodolog) established b> the American Society' of Mammalogists (1987). The United States National Biological Senice, tlie Utah Di\ision of \\ildlife Resources, the Wildlife Management Institute, and Utah State Uni\er- sit>- jointK support the Utiili Cooperati\ e Fish and Wildlife Research Unit and make our research possible. We tliank them. Literature Cited American Society of Mammalogists. 1987. Acceptable field inetliods in maniiniilog\ : preliniiniin guidelines approxed by tlie .\merican SocieU' of Mammalogists. Jouniid of Maniinalog)' 65 (4. supplement). 18 pp. Andryk, T. a., and L. R. Irby. 1986. Eviiluation of a moun- tain sheep transplant in noitli-cenbiJ Mont;uia. Jour- nal of Enxironmental Management 24: 337-346. Be.\som, S. L., E. P Wiggers, .\nd J. R. Giardino. 1983. A technique for assessing land surface ruggedness. Journal of ^^'ildlife Miuiagement 47: 1163-1166. BouL.\NCER, J. G., AND G. C. WHITE. 1990. A comparison of home nmge estimators using Monte Carlo sinuila- tion. Journal of Wildlife Management 54: 31t)-315. Blechner, H. K. 1960. The bighorn sheep in tlie United States, its past, present, and future. Wildlife Mono- graph 4. 174 pp. Cowan, I. M. 1940. Distribution and \ariation in the native sheep ot North .\merica. American Midland Natiuiilist 24: 505-580. Demarc:hi, D. .\. 1975. Report and recommendations of the workshop on California bighorn sheep. Pages 143-163 /'(I J. B. Trefctlien, editor. The wild sheep in modem Xordi .-Vmerica. Winchester Press, New York. Dem.vrcih, D. A., .\ND H. B. Mirhkll. 1973. The Chil- cotin Ri\er bighorn population. C'anadian Field-Nat- uralist 87: 433-454. Dkewek, J., Jr. 1970. Population characteristics and be- havior of inhoduced bighoni sheep in Owyhee Coimt>, Idalio. Unpublished master's thesis, Universit\ of Idalio, Moscow. 46 pp. Geist, \'. 1971. Mountain sheep: a study in behavior and evolution. Uni\ersit>- of Chicago Press, Chicago, IL. 383 pp. GlONFRlDDO, J. P, AND P R. Kralsman. 1985. Summer habitat use h\ mountain sheep. Journal of Wildlife Management 50: 331-336. GlullAM, H. 1971. Environmental analysis procedures for bighoni in die S;ui Gabriel Mountiiins. Desert Bigliom Council Transactions: 15: 38 — 15. Hanna, P 1978. OwAhee connection. Idaho Wildlife 1(2): 7-9. tllCKS, L. L., AND J. M. Elder. 1979. Human disturbance of Sierra Nevada biglioni sheep. Journal of Wildlife Management 43: 909-915. Klng, M. M. 1985. Behavioral response of desert bighoni sheep to himian harassment: a comparison of dis- turbed and undisturbed populations. Unpublished doctoral dissertation. Utah State Universitv. Logan. 137 pp. Klng, M. M.. and G. W. Worknlxn. 1985. Response of desert bighoni sheep to liimiim harassment: niiuiage- nient implications. Transactions of the 51st North American \\'ildlife and Natural Resources Confer- ence 51: 7-1-85. Kornet, C. A. 1978. Status and habitat use of C;ilifornia bighoni sheep on Hart Mountain, Oregon. Unpub- lished masters tliesis, Oregon State Universit\; Cor- V iillis. 49 pp. Kr\lsnl\n, P R.. and B. D. Leopold. 1986. Habitat com- ponents for desert bighoni in die Harcjualiala Moun- tains, Arizona. Journal of Wildlife Management 50: 504-508. Leslie, D. M.. Jr., .\nd C. L. Doiglxs. 1979. Desert big- honi sheep of the River Mountains. Nevada, ^^ildlife Monogiaphs 66: 1-56. McQuiVEV, R. P. 1978. The bighorn sheep of Nevada. Nevada Department of Fish and Game, Biological Bulletin 6. 81 pp. M.VCKIE, R. J. 1978. Impacts of livestock giiizing on wild un- gulates. Transactions of the Nordi .\nierican Wildlife and Natural Resources Conference 43: 462—176. Naef-Daenzer, B. 1993. .\ new transmitter for sniidl ani- mals ;md enhanced methods of home-riuige aiiidv sis. Jouniid of Wildlife Nhuiagement 57: 680-689. Peiper, R. D. 1988. Grazing svstems and management. Pages 9-24 in B. .\. Buchanan, editor, Rangelands. University' of New Mexico Press, ,\lbuquerque. Samuel, M. D., D. J. Pierce, .\nd E. O. Carton. 1985. Identifying areas of concentrated use within the home nmge. Jouniiil of AniuKd Ecologv 54: 711-719. Smith, D. R. 1954. The bighorn sheep in kkilio: its status, lite historv and management. Idaho Department of Fish and CJanie. Wildlife Bulletin 1. 154 pp. Smith, N., P R. Kiulsnlvn, and R. E. Kirby. 1988. Desert bighoni sheep: a guide tt) selected management prac- tices. U.S. Department oi Interior. Biological Report. 88. 35 pp. Tow EILL, D. 1985. The Calitbrnia bigiiorn sheep in Idaho. Pages 45-.56 in Proceedings of the CiJifoniia Bighorn Workshop, Reno. N\'. 1996] SiiKKp Kksi'onsk k) Fragmentation 325 I'RKFETHEN, J. B., !• DITOK. 1975. The wild sIk'C'p in North America. Wincliester Press, NY. 302 pp. \ \\ DvKi;, W. A. 1978. Population characteristics and habitat utihzation of hi^horn sheep, Steens Moun- tain, Oregon. Unpuhhslied master's thesis, Orej^on State University, Corxalh's. 87 pp. \ AN DvKE, W. A., A. Sands, J. Yoakum, A. Polenz, and J. Bl^MSDELL. 1986. Wiklhle habitats in manaj^ed range- land.s — the Great Basin oi southeast Oreuon — bij^honi sheep. 2nd edition. Cleneral 'leclinical ]^e]K)rt PNVV- 159. 37 pp. Weyhausen, J. D. 1980. Siena Nevada bitihoru slieep: his- tory and population ecology. Unpublislied doctoral dissertation. University of iMichigan, Ann Arbor 240 pp. WIL.SON, L. O. 1975. Report and recommendations of the Desert and Mexican bighorn sheep workshop. Pages 110-13! in J.B. Trefethan, edit(jr. The wild sheej) in modern North America. Winchester Press, N.Y. WoRTON, B. J. 1989. Kernel methods lor estimating the uti- lization distribution in home-range studies. Kcology 70: 164-168. • 1995. Using Monte Garlo simulation to evaluate kernel-based home range estimators. Journal ol Wild- lile Management 59: 794-800. Received 23 October 1995 Accepted 3 April 1996 Great Basin Naturalist 56(4), © 1996, pp. 326-332 A FIELD STUDY OF THE NESTING ECOLOGY OF THE THATCHING ANT FORMICA OBSCURIPES FOREL, AT HIGH ALTITUDE IN COLORADO John R. Conway^ Abstract. — ^A field study of the thatching ant, Fonnicu obsciiripes Forel, at 2560 m elevation in Colorado provided information on mound density, composition, dimensions, and temperatures; worker longevity; and mite parasitization. Density was 115 mounds/lia. Mounds had 1-52 entrances and Peromysciis fecal pellets in the thatch. Mounds conserved heat and exhibited thermal stratification. Excavations of 4 nests revealed depths of 0.3 m to almost 1 m, novel myrme- cophiles, and 0-198 wingless queens per nest. Marking experiments demonstrated that some workers overwinter and live more than a year Key words: Formica obscuripes, thatching ant, Colorado, ant mounds, mijrnwcophiles. Formica obscuripes Forel is in the Formica nifa-group (Weber 1935) and ranges from Indi- ana and Michigan westward across the United States and southern Canada. It is one of the most abundant ants in western North America, especially in semiarid sagebrush areas (Gregg 1963), and has been found at altitudes up to 3194 m (Wheeler and Wheeler 1986). The objective of this field study was to com- pare mound density, formation, composition, dimensions, and temperatures, worker longevit>' and parasitization, nest depths, mynnecophiles, and the number of wingless queens per colony of this species at high altitude in Colorado with findings from lower altitude studies in Colo- rado (Jones 1929, Gregg 1963, Windsor 1964), Idaho (Cole 1932), Iowa (King and Sallee 1953, 1956), Michigan (Talbot 1972), Nevada (Clark and Comanor 1972, Wheeler and Wheeler 1986), North Dakota (McCook 1884, Weber 1935), Oregon (Mclver and Loomis 1993, Mclver and Steen 1994), Washington (Hender- son and Akre 1986), and Canada (Bradley 1972, 1973a, 1973b). Although this species seems to be most common at altitudes of 1524-2743 m in tlie mountiiinous states (Gregg 1963, Wheeler and Wheeler 1986), the highest previous study site was at an elevation of 1550 m (Clark and Comanor 1972). It is hypothesized that cli- matic and vegetational changes associated with higher altitude may alter the nest ecology of this species. Materials and Methods The study site is in Gunnison County north of Blue Mesa Resei^voir and west of Soap Creek Co. Rd. in western Colorado at an alti- tude of 2560 m. Field obsei^vations were con- ducted 5-6 August 1990, 20 June -11 October 1992, 28 June-16 August 1993, 29 June-31 July and 14-16 August 1994, and 3, 29-31 July and 15-16 August 1995. The area, dominated by big sagebmsh {Arfemcsia tridcntata Nuttall) and to a lesser extent by rubber rabbitbrush {Clirysothammis miiiseosiis [Pallas] Britton), is adjacent to a grove of quaking aspens {Popiihis tremuloides Michaux). The locations of 85 mounds were mapped in a study area (64.6 m x 114 m) using a sur- veyor's transect and compass in JuK' 1993 to determine densit). The diameters and heights of 97 mounds in the stud\' area and sunounding area were mea- sured. The number of entrances per mound was determined by inserting sprinkler flags into the active openings on each mound. Mound temperatures were measured with a Model 100-A VWR digital thenuometer probe. Sixty-seven temperatiue measurements were made on 34 mounds in the evenings (1915-2045 h) 2-14 July 1993 by inserting the probe appro.ximately 15 cm into tlie top of each mound. The temperatures of 4 of these mounds were also recorded in the afternoon (1538-1600 h) Inipartiiii-iit of Hiolug), Liiivt-rsily ofScruiitoii, Scnintoii, P.\ 1S.t1(I. 326 1996] Nkstinc; Ecolcx;y of Thatching Ant 327 on 2 July 1993. In addition, hourly temptM-a- tnres were reeorded at 4 loeations (air, liround, mound top, and mound base) lor 3 diflerent- si/ed mounds in July 1994 between 0700 and ilOOO h tt) determine how moimd size aifeets thermal d) namies. Temperatures were taken at a mid-sized mound (height = 25.4 cm, average diametei- = 1 m) on 16-18 July, at a large mound (height = 49.5 cm, a\'erage diameter =1.21 m) on 18 JuK; and at a small mound (height = 27.9 cm, average diameter = 0.51 m) on 17 July. The small and mid-sized mounds were about 4.6 m apart and about 34-37 m from the large mound. The probe was inserted approxi- mately 15 cm into the top, base, and ground adjacent to each mound. Temperatures were also recorded in the shade about 15 cm above the ground near each mound. Hundreds of workers were marked on 8 nioimds and 5 plants in 1992-93 by applying model aiiplane paint with a fine-tipped brush and by spraying 5 mounds in 1994 with colored acnlic enamel. Although many workers were incapacitated or killed, especially by spraying, most sun'i\ ed. Spraying was the most efficient technique for marking large numbers of ants. Four nests were excavated, 1 each on 6 August 1990, 27-28 June 1992, 12 July 1993, and 11-25 July 1994. The 1993 nest was poisoned with IV2 cups Hi-Yield ant killer granules (Diazi- non) wetted down with about 7.6 L of water prior to excavation to investigate another tech- nique for collecting queens and mynnecophiles. Results Nest Density The extrapolated density for the 85 mounds mapped in the 7364-m2 area was 115 mounds/ ha. The closest mounds were 2.36 m apart. Mounds Formation and composition. — Mounds are composed of thatch and are usually dome sliaped. Some moimds are exposed while oth- crs are overgrown or shaded by low vegeta- tion. Dead sagebrush protruded from or was found on 63 of 98 mounds (64%). The largest mound was built aroimd the base of a fence post. No mounds were found inside the aspen grove, but 2 were built around small aspen trees on the forest edge. Mound thatch consisted mainly of twigs but also contained fecal pellets, probably from the deer mouse {Pcwmijscus maniculatus [Wagner]) or vole {Micwlm sp.). Thatch (n = 58) from 1 mound consisted mainly of small twigs 4-89 nnn (mean = 24.19 nnn) long and 1-5 mm (mean = 2.19 nmi) in diameter Workers were observed carrying fecal pellets into or out of mound entrances, but not on trails. Dimensions and entrances. — The diame- ters of 97 mounds ranged from 19 cm to 142 cm (mean = 65 cm). Mound heights ranged from 6 cm to 58 cm (mean = 26 cm). The number of entrances to 97 mounds ranged from 1 to 52 per mound (mean = 12), but their number, size, position, and activity changed over time. For example, 1 mound had 10 or more entrances in August but only 2 in October. Some entrances were larger than oth- ers, and some surrounded plant stalks growing out of mounds. Temperatures. — Measurements of mound- top and air temperatures in July 1993 demon- strated that moimds are warmer than air tem- peratures and that the differential is greater in the evening than in the afternoon. Evening temperatures {n = 67) for 34 mounds were 1.0°-15.5°C (mean = 8.6 °C) warmer than cor- responding air temperatures. Afternoon tem- peratures for 4 of these mounds were slightly warmer (0.5°-0.9°C; mean = 0.7 °C) than cor- responding air temperatures. Hourly mound-top and mound-base tem- peratures recorded in July 1994 were almost always higher than ground temperatures, and top temperatures were warmer than air tem- peratures (Figs. 1-3). Differences in top and air temperatures were greater in the evening (1900-2000 h) for a large nest (8.9°-irC) and mid-sized nest (6.8°-14.4°C) than their after- noon (1500-1600 h) differences, 2.6°-6.3°C and 0.6°-8°C, respectively. On die otlier hand, hour- ly top and air temperatures did not differ much for the small nest in die evening (1.1°-3.3°C) and in the afternoon (1.6°-2.2°C). Average hourly top and base temperatures were higher than average air temperatures for die mid-sized and large mounds (Figs. 1-3). For example, average top and base temperatures were 6.2 °C and 3.1 °C higher than average air temperatures for the large mound and 4.6 °C and 0.5 °C higher for the mid-sized mound. However, for the small mound the average top temperature was actually 0.8 °C lower, whereas the average base temperature was 2.7 °C higher than the average air temperature. 328 Great Basin Naturalist [Volume 56 MOUND #3 JULY 16-18, 1994 MOUND #14 JULY 18, 1994 35.0 35.0 ^ Time of Day -•- Mound Top "♦■■ Mound Base -•♦-• Ground -O- Air Fig. 1. Average mound-top, mound-base, ground, and air temperatures around a mid-sized Formica ohscuripes mound from 0700 to 2000 h on 16-18 Jul\ 1994 at 2560 m in Colorado. 10 11 12 13 14 15 16 17 18 19 20 Time of Day -•- Mound Top -*^' Mound Base -♦-■ Ground -O- Air Fig. 2. Mound-top, mound-base, ground, and air tem- peratures around a large Formica obsciiripes mound from 0800 to 2000 h on 18 Jul\ 1994 at 2560 m in Colorado. Hourly top and base temperatures showed thenual stratification. Average top temperatures were 3.2 °C and 4.1 °C higher than average base temperatures for the large and mid-sized mounds, respectively. However, for the small mound the stratification was reversed: average top temperature was 3.5 °C lower than the average base temperature. The poor thermal regulation of smaller mounds was also reflected by a greater fluctua- tion of hourly top and base temperatures. Daily ranges of top/base temperatures were 7.6/8.7 °C, 13.3/15.9 °C, and 13.8/26.3 °C for the large, mid- sized, and small nests, respectively. Thus, larger mounds exhibited less daily temperature fluc- tuation than smaller mounds. Worker Longevity Most marking experiments {n = 14) indicated that some workers live 19 to 44 d (mean = 31.6 d). However, 2 workers marked on a mound between 7-9 July and 15-27 JuK' 1994, respec- tively, were observed on 30 July 1995 on another mound and on the original mound. Thus, some workers overwinter and live more than 1 yr Mites Mite infestation was not common. Orange, spherical mites were noted on only 1 worker at 3 of the many mounds observed. The largest number of mites obsened was 4-5 on the tho- rax and gaster oi 1 worker. Excavated Nests Each oi the 4 nests excavated contained numerous workers, larvae, and pupae, but the nimiber of wingless queens per nest varied greatK : 0, 1, 32, and 198. No winged reproduc- ti\ es were found except a male in 1 nest. The 1996] Nesting Ecology ok Tiiai-giiing Ant 329 MOUND #98 JULY 17, 1994 ers); and Lepidoptera Table 1). 'E Noctuidae — lan'ae; 45.0 T 40.0 .. Q 35.0 e n 30.0 t i 25.0 g r 20.0 a 15.0 e 10.0 5.0 0.0 H 1 1 1 1 1 1 1 1 I ' ' 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Time of Day *-• Mound Base ■O- Air »- Mound Top ►-• Ground Fig. 3. Mound-top, mound-base, ground, and air tem- peratures around a small Formica obsciihpcs mound from 0800 to 2000 h on 17 Jul\ 1994 at 2560 m in Colorado. depth.s of the nests were 0.3 m (estimated), 0.3 m, 0.64 m, and 0.97 m. The nest excavated in 1993 contained the following arthropods: pseudoscorpions, coUem- bolans, beetles and beetle larvae (1 Ctenicera sp. [E Elateridae] and 4 Eleodes sp. [E Tene- brionidae]; Table 1). The following insects were identified in the 1994 nest: CoUenibola (E Entomobryidae); Hom- optera (E Cicadellidae — 1 inmiature, E Aphid- idae — 2 immatures); Hemiptera (E Anthocori- dae — 1 specimen); Coleoptera (E Curcnlion- idae — 5 adults, E Scarabaeidae — 1 adult and Cremastocheihis pupa and larval skin, probable E Carabidae — 1 adult, probable E Anthri- bidae — 2 larvae, E Tenebrionidae — unidenti- fied lai-vae, probable Eleodes sp. lai-vae, and Eleodes sp. pupae, E Cerambycidae — Lepturi- nae, probable Leptiira sp. larva); Diptera (probable E Asilidae — pupa); Hymenoptera (E Formicidae — few Tapinoma sessile [Say] work- DiscussioN .\ND Conclusions The extrapolated densit\' of 115 mounds/ha is about 1.8 times greater than the highest den- sity reported: 64Aia of Jack pine in Manitoba (Bradley 1973a). Colonies are known to be polydomous and to reproduce by budding (Herbers 1979). Some primarv' mounds and small secondaiy mound- lets along trails appeared and disappeared in our study area over the years as previously reported, and some may have moved. For ex- ample, a primary mound that was active in 1990 was largely deserted by 1994 and com- pletely abandoned in 1995. Colonies have been reported to move at least 3 times during their life and to move 18 m from their original location, or 1.3-;33 m after transplantation (Brad- ley 1972, 1973a). King and Sallee (1953, 1956) noted desertions of many old nests and the establishment of 1 or more new ones from each of them. All our mounds were in open sagebrush ex- cept for 2 built around aspens at the forest edge. Weber (1935) also noted that most mounds are in the open, but did find some mounds par- tiallv shaded and 1 enormous mound almost completely shaded in an aspen grove. In our study, 63 of 98 mounds (64%) showed evidence of being built around sagebrush as reported by Weber (1935), but a few were built around other structures such as trees and a fencepost. Weber noted that workers kill sage- bmsh by chewing bark at the base and spraying formic acid on the cambium. After 3 months, the stem is removed to form a longitudinal pas- sage in the center of the mound leading to the main entrance. Weber (1935) reported that mounds are composed of slighdy longer twigs (1-12 cm) than the ones we measured (0.4-8.9 cm), but these slight differences may simply reflect the availability of materials. A new discoveiy was the presence of fecal pellets of P. manicuhitits or Microtus (Clark personal communication) on the surface and in the thatch of Colorado mounds. Since workers were never observed canying pellets to mounds, their origin is unclear. Although Clark and Comanor (1972), Tilbot (1972), and Wheeler and Wheeler (1986) 330 Great Basin Naturalist [Volume 56 Table 1. Arthropods in Formica obsctiripcs Forel nests reported in die literature luid identified troni 2 excavated nests near Soap Creek, Colorado (*). *Collembola (unident.) Weber (1935) *E Entomohnidae *Honioptera *F. Aphididae — 2 immatin-es *E Cicadellidae — 1 innnature *Heiniptera *E Anthocoridae — 1 specimen Diptera *E Asilidae — pupa E Milichiidae PliyUonujza seciiriconiis Weber (1935) E Leptidae — Ian ae Weber (1935) E Anthoniyiidae — lanae \\'eber (1935) E There\ idae — lai-\'ae Weber (1935) E Phoridae — lana W'indsor (1964) Lepidoptera *E Noctuidae — lar\ae Epizcitxis sp. — lar\ae \\'eber (1935) H\nienoptera E Eormicidae Lasius hitipcs W;dsh \\eber (1935) Lcptoflwrax hiiiicornis Emen Weber (1935) *T(ipiiio)na sessile Say \\'eber (1935) Th\ sanura — sil\ ertisli ^^■indsor (1964) Coleoptera Unident. beeUe pupa Windsor (1964) E Elateridae *Cteiucera sp. lana Mehiiwtus sp. lanae Weber (1935) E Tenebrionidae I'nident. lanae and adults \\'indsor (1964) *Eleo(Ies sp. — lanae and pupae *Unident. lanae E Carabidae Ainara sp. — adult temale \\eber (1935) *Prob. adult Unident. adults Windsor (1964) *E Anthribidae *2 prob. lanae *E Ceranib\ cidae *Prob. Leptuni sp. — lana *E Cmculionidae *5 unident. adults Coleoptera (continued) E Scarabaeidae *Unident. adult *Crein' of mound entrances changed o\"er time as re- ported b\ Weber (1935). The number of en- trances per mound in our stud); 1—52, is close to the range of 3^50 per mound reported (Cole 1932. \\heeler and Wheeler 1986). In the earl>- morning ants use openings in die sunlight; later as the temperatine rises tiie\' use onK' shaded entrances as reported b>" Weber (1935). Hen- derson and Akre (1986) speculate entrances are opened dining the da\ and closed widi diatch at night to help control nest temperatures. Our mounds. especialK mid-sized and large ones, were geneialK warmer than groimd and 1996] Nestinc: Ecoixxn of Tiiatchinc; Ant 331 air temperatures and exhibited Hiennal stratifi- cation from top to base. Weber (1935) and Andrews (1927) also noted that mounds are w anner than the ground, and Andrews reported that the upper parts are warmer tlian the lower parts of" mounds. The differential between our mound-top and air temperatures was greater in the evening dimi in the iiftenioon. Sniiill mounds showed a re\ ersal of diermal stratification and greater liourK fluctuation of top and base tem- peratures, whicli is indicati\ e of poorer ther- mal regulation. Marking experiments suggest that worker longevit}' is often short but that some workers oxenvinter and lixe more than a year. Little information is available on the longevity of worker ants and none was found for diis species. Akhough maximum longevit>' is known to be 3 \ r for workers of some species, such as Aphae- )u) area. Weber (1935), on the other hand, noted that mites (Parasitidae, Tyroglyphidae, Vropoda sp.) were common on workers and sexuals, especially on the tibia-tarsal joint, and estimated over 200 on 1 queen. Excavated nests varied in depdi from 0.3 to 0.97 m, or less than the maximum depth of 1.37 m noted by McCook (1884) and 1.58 m reported by Weber (1935). Weber speculated that the water table (below 1.52 m) limits nest depth. Nests excavated from 27 June to 6 August did not contain winged reproductives except a male in 1 nest. This finding differs from Cole's (1932) obsei-\ations of large numbers of winged reproductives through June and July. Many species of Formica are polygynous (Kannowski 1963). The number of wingless ( [ueens per Colorado nest varied from 0 to 198 (Conway 1996). The latter number far exceeds the 2 or more queens per colony reported by Cole (1932). The following arthropod groups found in our excavated nests had not been reported associated with this species: pseudoscorpions, coUembolans (E Entomobiyidae), homopterans (E Aphididae, E Cicadellidae), hemipterans (E Anthocoridae), dipterans (E Asilidae), and cole- opterans (E Carabidae, E Anthribidae, E Cur- culionidae, E Elateridae — Ctenicera sp., E Tenebrionidae — Eleode.s sp., and E Ceramby- cidae — probable Leptura sp.; Table 1). On the other hand, Windsor (1964) and Henderson and Akre (1986) reported 3 major groups not found in our limited sampling: Arachnida (small spiders), Tlnsanura (silvcHlsh), and Orthoptera (E Cryllidae). In addition, Weber (1935), Windsor (1964), and MacKay and MacKay (1984) noted many dipteran and coleopteran families not in our nests and new genera and species in a few of the same fami- lies found in our nests (Table 1). The relationship of these myrmecophiles with the host colony is unclear. Larval and adult coleopterans and noctuid lawae may use the chambers for hibernation or development (W^eber 1935). Staphylinid beetles may prey upon brood or workers. Jones (1929) suggested that lepidopteran, coleopteran, and dipteran larvae are tolerated because they feed on decaying vegetable matter in the nest. Cremas- tochcihts is a well-known symbiont in the nests of a number of ant species (Holldobler and Wilson 1990). The scarab genus Euphoria may be a symbiont treated with indifference liy the host colony (Wheeler 1910). On the other hand, ants are aggressive to other guests, such as the m>rmecophilous cricket {Mijnnecophihi manni Schimmer; Henderson and Akre 1986). Weber (1935) reported 3 ant species in nests (Table 1) and noted that Leptothorax hirticornis Emery may prey upon brood or isolated work- ers. Tapinoma sessile, one of the species in our nests, often steals honeydew from thatching ants throughout its territoiy, but seems to elicit little defensive response (Mclver and Loomis 1993). The high altitude of our study site did not seem to significand)' alter nest dimensions and ecology, but the work did provide new findings on this species, such as the greatest mound density per hectare, first report of probable P. maniculatm fecal pellets associated widi mound thatch, new information on the thermal prop- erties of mounds, new^ information on worker longevity', greatest number of wingless queens reported in a nest, and possible new myrme- cophiles. Acknowledgments Support for this research was pro\ided by grants from die Howard Hughes Medical Insti- tute through the Undergraduate Biological 332 Great Basin Naturalist [Volume 56 Sciences Education Program in 1993-94 to the following University of Scran ton students: John Bridge, Tom Sabalaske, Anthony Musingo, and Jeanne Rohan. I would also like to thank the Systematic Entomolog) Laboraton' in Beltsxille, Maiyland, for identification of myrmecophiles; William Clark at the Onna J. Smith Museum of Natural Histoiy, Albertson College of Idaho, in Caldwell, Idaho, for identification of mam- malian fecal pellets; and the Forest Sei-vice in Gunnison, Colorado, for identification of plants. LiTER.\TURE Cited Andrews, E. A. 1927. Ant-mounds as to temperature and sunshine. Journal of \Iorpholo^\ and Ph\ siologv' 44: 1-20. Br.\dley, G. a. 1972. Transplanting Formica obsciiripes and DoUchodenis taschenbcrgi (H\ nienoptera: Fonni- cidae) colonies in Jack pine stands of southeastern Manitoba. Canadian Entomologist 104: 245-249. . 1973a. Interference between nest populations of Formica obscuripes and Dolichoclcrus taschenbcrgi. Canadian Entomologist 10,5: 1525-1528. . 1973b. Effect ofFonnica obscuripes (H\inenoptera: Fonnicidae) on tlie predator-pre\' relationship between Hyperaspis congressis (Coleoptera: Coccinellidae) and ToiimeycUa uumismaticiim (Homoptera: Coccidae). Canadian Entomologist 105: 1113-1118. Clark, W. H., and P. L. Com.\.\'OR. 1972. Flights of the western thatching ant, Formica obscuripes Forel, in Nevada. Great Basin Naturalist 32: 202-207. Cole, A. C, Jr. 1932. The thatching ant, Formica obscur- ipes Forel. Ps\'che 39: 30-33. Conway, J. R. 1996. Nuptial, pre-, and postnuptial acti\ it\ of the thatching ant, Formica obscuripes Forel, in Colorado. Great Basin Naturalist 56: 54—58. Grecg, R. E. 1963. The ants of Colorado. University of Colorado Press, Boulder 792 pp. He\der,so\, G., and R. D. Akre. 1986. Biologv of die myr- mecophilous cricket, Mynnccophila monni (Oitlioptera: Gnllidae). Journal of the Kansas Entomological Soci- ety' ,59: 4,54^67. Herbers, J. M. 1979. The evolution of se.x-ratio strategies in hvnienopteran societies. American Natiualist 114: 818-834. Holldobler, B., and E. O. Wilson. 1990. The ants. Bel- knap Press of Han ard University' Press, Cambridge, MA. 732 pp. Jones, C. R. 1929. Ants and their relation to aphids. Bul- letin of the Colorado Agricultural E.\periment Station .341: 1-96. KiNt;, R. L., AND R. M. Sallee. 1953. On the duration of nests oi Formica obscuripes Forel. Proceedings of the Iowa Academ\' of Science 60: 656-659. . 1956. On the half-life of nests of Formica obscur- ipes Forel. Proceedings of the Iowa Academy of Sci- ence 63: 721-723. Kannowski, R J. 1963. The flight acti\ ities of formicine ants. Symposia Genetica et Biologica Ittilia 12: 74—102. MacKay, E. E., and W. P M.\(:K.\y. 1984. Biology of the thatching ant Formica Jiaemorrlioidalis Emeiy (Hyme- noptera: Formicidae). Pan-Pacific Entomologist 60: 79-87. Mann, \V. M. 1911. On some north western ants and their guests. Psyche 18: 102-109. McCoOK, H. C. 1884. The mfous or thatching ant of Dakota and Colorado. Proceedings of the Academy of Nat- ural Sciences of Philadelphia, part 1: 57-65. Mcl\ ER, J. D., AND C. LooMis. 1993. A size-distance rela- tion in Homoptera-tending thatch ants (Fonnica obscuripes, Formica planipilis). Insectes Sociau.x 40: 207-218. McIvER, J. D., .^ND T. Steen. 1994. Use of a secondary nest in Great Basin Desert thatch ants {Formica ob- scuripes Forel). Great Basin Naturalist 54: 359-365. Talbot, M. 1972. Flights and swamis of the ant Fonnica obscuripes Forel. Joinnal of the Kansas Entomologi- cal Society 45: 254-258. Weber, N. A. 1935. The biology of the thatching ant Formica obscuripes Forel in North Dakota. Ecological Monographs 5: 16.5-206. \\'heeler, W. M. 1910. Ants, their stioicture, development and beliavior Columbia University Press, New York. 663 pp. Wheeler, G. C., and J. N. Wheeler. 1986. The ants of Nevada. Natural Histoiy Museum of Los Angeles County, Los Angeles, CA. 138 pp. Windsor, J. K. 1964. Three scarabaeid genera found in nests of Formica obscuripes Forel in Colorado. Bul- letin of the Southern California Academy of Sciences 63: 205-209. Received 5 January 1996 Accepted 2S June 1996 Cri'at Basin Naturalist 56(4), © 199fi, pp. 333-3 10 GAS EXCHANGE, 8l^C, AND HETEROTKOPHY FOR CASTILLEJA LINARIIFOLIA AND ORrHOCARPUS TOLMlEl FACULTATIVE ROOT HEMIPARASITES ON ARTEMISIA TRIDENTATA Lori A. Duchainu'' and James H. Klilfiiii^cr' Abstract. — Gas e.xchange and carbon isotope ratios were measured on 2 Facultative heniiparasites, CasUllcja liiiari- ifolia Benth. (Indiau paintbrush; Scropliulariaceae) and Orthocarpu.s toliniei II. & A. Oblmie owl clover; Scropluilari- aci'ae), and their Ar/e//u'.s/V; trklentata L. (bit;; sauebrush; Asteraeeae) hosts. Photosynthetic rates differed j^reatly between \ cars; rates in 1995 were more than double those in 1994, likely due to more precipitation and less water stress during U)95. Despite this difference in precipitation, photos\nthetic rates for C liuanifoUa were not different from those of (heir hosts for either year. However, carbon isotope ratios of C. linariifolia and O. tolm'wi were up to 3%o more negative than those of their A. tridentata hosts. Using measmed 8'^C ratios in conjunction with S^^C ratios predicted from gas- ixchange measurements, we calculated that C. linariifolia derived, on average, 40% of its leaf carbon heterotrophically. (^outran' to current suggestions that high photosynthetic rates of heniiparasites are an indication of reduced heterotro- ph\, C. linariifolia e.xhibited photosynthetic rates similar to autotrophic plants and used a substantial amount of host- (iciixed carbon. Moreover, this evidence shows that manipulation of a heterotrophic carbon supply transcends obligate lii'miparasites to include those plants whose parasitism is facultative. Key words: hetcrotropluj, luiniparasitc, photofiijnilu'fiis, carbon isotope ratios, slirith ecolofiy. Heniiparasites, chlorophyllous parasitic plants, form an apoplastic continuum with host x\ lem (Ra\'en 1983). R has been assumed that these plants are largely autotrophic plants, being parasitic only for water and minerals (Smith et al. 1969). However, heniiparasites may also gain carbon through the passive uptake of dilute concentrations of organic car- bon contained within host xyleni sap (Raven 1983). Early studies using radiocarbon labeling demonstrated the transfer of solutes from host to parasite (Hull and Leonard 1964, Govier et al. 1967), although it was not possible to quantify tliis flux. Experiments of Govier et al. (1967), in which [^"^C]urea or ^"^COq was fed to hosts, showed the movement of ^'^C labeled coni- poimds to all parts of the hemiparasite Odon- tites venia (Scropliulariaceae). More recent stud- ies used a carbon budget model and/or a h^'^C method to (juantify the extent of heterotrophy (Press et al. 1987a, Graves et al. 1989, Marshall and Ehleringer 1990, Schulze et al. 1991, Mar- shall et al. 1994, Richter et al. 1995). Using the latter method, Press et al. (1987a) calculated tliat 28-35% of total carbon in Striga hennonthica and Striga asiatica (Scropliulariaceae) is host- derived carbon. There is also ample evidence that hemiparasitic mistletoes utilize host-derived carbon, although the values vaiy greatly, from 5% to over 60% (Marshall and Ehleringer 1990, Schulze et al. 1991, Marshall et al. 1994, Richter et al. 1995). Despite the potential importance of heterotrophy to carbon acquisi- tion in parasitic plants, relatively few studies have addressed this aspect of parasite-host interactions. Moreover, none have evaluated the exploitation of this carbon source by facul- tative root heniiparasites. Photosynthetic rates of heniiparasites fall within the lower range reported for C3 plants and are generally much lower than photo- synthetic rates of the host. S. hennonthica has a poorly developed palisade mesophyll, con- tributing, in part, to photosynthetic rates as low as 2.5 /zmol ni-2 s-1 (Shah et al. 1987). Moreover, these rates are half those reported for their Sorghum hosts (Press et al. 1987b). Striga species are the most extensively studied root heniiparasites because of their importance as agricultural weeds in semiarid Africa, and as obligate hemiparasites they require host attachment for sui-vival. Similarly, low photo- synthetic rates were found in facultative root hemiparasites. Press et al. (1988) measured iDqwrtiiK-iitofBioIogy, Universih' of Utah, Salt Lake Cih; UT84U2 333 334 Great Basin Naturalist [Volume 56 light-saturated photosynthetic rates of 2.1 to 7.5 fimol m~^ s~^ for 8 facultative species of Scrophulariaceae. However, 1 exception to this trend of low photosynthetic rates is the Medi- teiTanean facultative heniipai"asite Bartsia trixago (Scrophulariaceae), which has CO2 assimilation rates ranging from 12.4 to 18.8 jU-mol m"^ s"^ well within the range measured for potential hosts (Press et al. 1993). Castilleja and Orthocarpiis are facultative hemiparasites, those with the abilit\' to sui-\'ive in the absence of a host. It is this facultative parasitism that distinguishes them from Striga. The majority of Castilleja species are perennial, while Orthocarpiis are annuals. Both occur throughout the Intemiountain West most com- monly in the pinyon-juniper, mountain brush, and aspen-conifer zones (1140-3140 m eleva- tion), with Orthocarpiis tohniei occurring only at the higher elevations (2195-3265 m; Welsh et al. 1987). Castilleja hnariifolia and Oiihocar- piis tohniei parasitize a variety of host species (Heckard 1962, Atsatt and Strong 1970). Artem- isia tridentata is the common host for both hemiparasites at the sites studied in the paper Our overall objective was to investigate gas exchange and heterotrophy characteristics for facultative hemiparasites. We focused primarily on the facultative root hemiparasite Castilleja linariifolia infecting Artemisia tridentata hosts. A secondaiy focus of this study was Orthocar- piis tohniei, a closely related annual facultative root hemiparasite, also infecting A. tridentata hosts. We asked the following questions: Do C. hnariifoha and O. tohniei exhibit gas-exchange activities similar to those of their hosts? Does C. hnariifoha utilize heteroti-ophic carbon? Does hemiparasite infection impact water availability and gas-exchange rates of A. tridentata hosts? To evaluate these questions, we measured gas exchange and analyzed carbon isotope compo- sition for C hnariifoha, O tohniei, infected and uninfected A. tridentata. In addition, predawn water potentials (^pj) were measured for infected and uninfected A. tridentata to exam- ine the impact of hemiparasite infection on host water availability. Materials and Methods Study Sites This study was conducted at 2 sites in Utah where the hemiparasites have different grow- ing seasons. The first site, Tintic, is located just off Mclntyre Road, approximately 12 km south of Eureka, Utah (Juab Co.), at the Desert Range Experimental Station operated by Utah State University (latitude 39°51'N, longitude 112°12'W). The area is a sagebrush steppe habitat at about 1525 m elevation where sage- bi-ush is interspersed with herbaceous species such as Erigeron, Castilleja, Astragalus, and Phlox. The growing season for Castilleja at this site begins in late April and ends in late June to early July. The second site, Sti^awbeny Reser- voir (Wasatch Co.), is about 130 km southeast of Salt Lake City and approximately 800 m north of Highway 40 along Coop Creek Road (latitude 40°15'N, longitude 111°8'W). This site lies in the southern tip of the Uinta National Forest at about 2280 m elevation. Sagebnish is the dominant shrub mixed with a few herbaceous species such as Castilleja, Oiihocarpiis, and Malta. The growing season for C. hnariifoha at Strawbeny Reservoir begins in early June and extends through August; O. tohniei begins a few weeks later and extends into September. Twenty pairs of C. linariifolia and A. triden- tata hosts were selected at each site. At Straw- beny Reservoir an additional 20 pairs of O. tohniei and A. tridentata hosts were selected. In addition, 5 uninfected A. tridentata were selected at both sites as hemiparasite-free controls. Gas Exchange Photosynthesis and stomatal conductance were measured with a portable gas-exchange system (LI-6200, Licor Instruments, Lincoln, NE, USA) twice during the C. hnariifolia grow- ing season at the Tintic and Strawbeny Reser- voir sites. Specific dates were chosen to corre- spond with the early and late parts of the para- site growing season. At both sites data were collected during diurnal peak photosynthesis (1000-1300 h MST) on 20 pairs of C. hnariifo- lia and infected A. tridentata, and on an addi- tional 5 uninfected A. tridentata. Dining the late season at Strawbeiry Reservoir, measure- ments were taken on an additional 20 pairs of O. tohniei and infected A. tridentata. After gas- exchange measurements were completed, foli- age was removed for leaf-area measurements using a leaf-area meter (LI-3100, Licor Instru- ments, Lincoln, NE, USA). mm Heterotropiiv in Castilleja and ORTHOCARI'US 335 Water Potentials Stems of approximate!) ecjual leut^th and diameter were seleeted lor predawn water- |i()tential (^p^) measurements using a pressure homl) (PMS Instruments, Comillis, OR, USA) lor 20 inleeted and 5 iminfeeted A. tridentata at l)oth sites. These measurements were taken approximately every 2 wk from May through eaily July at the Tintic site and late June through the end of August at the Strawberry Hesenoir site. Carbon Isotope Composition Carbon isotope ratios (S^'^C) were analyzed for the same plants used to measure gas exchange. The foliage was dried for 24 h and then finely ground with a mortar and pestle to homogenize the tissue (Ehleringer and Osmond 1989). Subsamples of 1-2 mg were combusted to produce CO2, which was mea- sured using an isotope ratio mass spectrometer (delta-S, Finnigan MAT, Bremen, Germany). Results are expressed using the S^'^C notation (%o), which relates the isotopic composition of the sample to the PDB standard as follows: 813c = r5;^^i^^^-il*iooo%r, (1) L ^staiulaid J where R represents the ratio of ^■^C0.2/^'^C02 of the sample and standard, respectively (Ehleringer and Osmond 1989). All isotope ratio anah'ses were conducted at the Stable Isotope Ratio Facility for Environmental Research at the Universitv of Utah, Salt Lake Cit>', Utah, USA. Calculation of Heterotrophy Heteroti'ophy was calculated using measured and predicted b^-^C ratios. The predicted b^-^C ratio (8pp), the carbon isotope composition of a leai provided that all carbon is autotrophic, was estimated with intercellular COo concen- trations (cj) from gas-exchange measurements. E(iuation 2 relates cj to the leaf carbon isotope ratio as modeled by Farquhar et al. (1982): K = 8a-a-(l3-il)(Ci/Ca), (2) where 8p is the h^^C of the plant (= 8 in this study), 8^ is the approximate b^'^C of the air (— 7.8%c), a and b are discrimination factors due to diflhision (4.4%c) and carboxylation via RuBP carb()x>lase (27%c), respectively, c., is the con- centration of CO2 in air (ppm) and Cj was cal- culated from gas-exchange measurements described abovc>. Ileterotrophv (H) was calcu- lated for the 1994 data (9 C. linariifolia, 5 infected and 5 iminfected A. tridentata) using E(][uati()n 3: 5 2 jj _ "pp ~"in Spp -^1. (3) where 8pp is the predicted S^'^C for the para- site, 8^ is the h^-^C measured in the parasite tissue, and 8], is the S^^C measured in the host tissue (Press et al. 1987a). Statistical Analysis Analysis of variance was used to compare yearly, seasonal, and plant means within a site for all photosynthetic data, and yearly and sea- sonal means for carbon isotope ratios (JMR Version 3, SAS Institute Inc., Caiy, NC, USA). The Tukey-Kramer Honestly Significant Dif- ference test (HSD) was used to make specific comparisons. In addition, for each hemipara- site, carbon isotope ratios were compiled for all seasons and sites, and differences between hemiparasites and hosts were compared using a t test. A paired t test was used to determine differences between predicted and measured b^'^C for each C. linariifolia, uninfected and infected A. tridentata. Differences in ^p^j water potential between infected and unin- fected A. tridentata were determined b)- 1 tests within each date. Results Analysis of annual trends in photosynthetic rates for Strawbeny Resewoir (Fig. 1) revealed that plants had significantly higher rates in 1995 than in 1994 for both parasite and host (Tukey-Kramer, a = 0.05). For example, in 1995 photosynthetic rates for C. linariifolia and infected A. tridentata were 18.3 ± 2.1 and 16.0 ± 0.6 ^tmol m"2 s'^, respectively, more than double those during the 1994 season. We also found seasonal differences in photosynthetic rates at StrawbeiTy Resenoir Both C. linariifo- lia and infected A. tridentata at Strawberry Reservoir experienced a significant decline in photosynthetic rates late in the season, with rates falling —6.7 and 8.6 ^tmol m'^ s'^ re- spectively (Tukey-Kramer, a = 0.05). However, 336 Great Basin Naturalist [Volume 56 Strawberry Reservoir, Utah Tintic, Utah o E CO CO (D c >> CO o ■I— « o sz CL %0^ % Fig. 1. Mean pliotosynthetic rates for liosts and parasites. Sites and sample sizes are as follows; Uninfected A. triclen- tata (Tintic: » = 3 for early season, n — 4 for late season; Strawberry Resenoir: n = 3 for late season), infected A. triden- tata (Tintic: ii = 12 for early season, n = 7 for late season; Strawberiy Resei^voir: n = 7 for early season, n = 19 for late season), C. linuriifoliu (Tintic: n = A for early season, /! = 6 for late season; Strawberiy Resewoir: n — 3 for early season, n — 5 for late season), O. tohniei (Strawberiy Resenoir: n = 5 for late season). Data are shown for Strawberiy Resei^voir (left panel) and Tintic (right panel) during the 1994 early season (open bars), 1995 early season (hatched bars), and 1995 late season (solid bars). Letters denote significant differences within each site. Error bars represent ± 1 s^. photosvnthetic rates at Tintic showed no sea- sonal differences (ANOVA, F = 1.88, F = 0.134; Fig. 1). In spite of annual and seasonal differences in photosynthesis for parasite and host plants, we found no difference in photo- synthetic rates between C. linanifolia and infected A. tndcntata. In contrast, O. tohniei rates (14.0 ±1.1 jitmol m~- s"l) exceeded those for infected A. tridentata (9.3 ± 0.4 ^tniol m~- s~l; Tukey- Kramer, a = 0.05; Fig.l). At both sites we found no significant differ- ence in predawn water potentials (^p^) between infected and uninfected A. tridentata (F > 0.05 for all dates, t test), although there was a general decline throughout the season (Fig. 2). The range in ^ | was similar be- tween sites; however, the values were slightly more negative at Tintic. Carbon isotope ratios differed between years for infected and uninfected A. tridentafa, with more negative values in 1995. However, 6^'^C values for C. linariifolia did not differ between years (Tukey- Kramer, a = 0.05; Table 1). Our results showed a slight seasonal decline in S^'^C values for parasites and hosts at Straw- beriy Resen'oir, although only O. tohniei and infected A. tridentata were significantly different (Tukey-Kramer, a = 0.05; Table 1). This trend in seasonal reduction was not evident for plants at the Tintic site. Furthermore, we found that hemiparasite S^'^C ratios were significantly more negative than those of the hosts (C. hnariifoha, t = 12.57, F < 0.001; O. tohniei, t = 11.94, F < 0.001). In 1994 C. hnariifoha b^'^C values (-28.9 ± 0.34%<:i) were nearly 3%c more nega- tive than those of the hosts (-26.2 ± 0.13%c), while this difference naiTOwed in 1995 to ~2%c at Tintic and ~l.5%c at Strawbeny Resei-voir. Results from experiments in 1994 showed a significant mean difference of 1.34%c between predicted and measured 8^^C ratios for C. hnariifoha (paired t test, f = 2.745, F < 0.05; Table 2). Using this difference we calculated that, on average, 40% of C. hnariifoha leaf car- bon was host derived; individual plants ranged from 16 to 60% (Table 3). C. hnariijoha hetero- trophy is well within the range of xalues calcu- lated for obligate hemiparasites. There was no statistical difference lietween measured and predicted 8^'^C values for either infected or 1996] Heterothophy in Castilleja and Ortiiocarpus 337 -0.5 -0.5 1 1 1 Tintic, Utah 1995 o 1 infected A. tridenlala • uninfected - A. tridentata • ^^^^^ ^~~'^'^->^ - o ^^^^^ 1 1 1 1 1 1 Strawberry Reservoir, Utah 1995 1 \ • o \^ - ^\^ • "^~- __• 6 1 1 t 1 Month Fig. 2. Seasonal course of predawn water potentials tor infected A. tridentata (open circles; n = 11) and unin- fected A. tridentata (solid circles; n — 5). Data are pro- vided for Tintic (upper panel) and Strawberw Reser\'oir (lower panel) from May to late August of 1995. Error bars represent ± 1 % • uninfected A. tridentata, indicating no hetero- trophic carbon gain as expected. Discussion Our results suggest that, with the exception of photosynthesis, the hemiparasites in this study behaved similarly to other hemiparasites. Photosynthetic rates for hemiparasites in this study were higher than rates for most other hemiparasites and similar to those of their auto- trophic host plants. We also found large differ- ences between years, which likely reflected differences in precipitation. In agreement with other studies, hemiparasite S^'^C ratios were more negative than those of the host (Press et al. 1987a, Marshall and Ehleringer 1990, Schulze et al. 1991, Richter et al. 1995). Furthermore, large differences in S^'^C ratios between the parasite and host suggested that the hemipara- site utilized a substantial amount of host- derived carbon. Despite relatively high photo- synthetic rates, heterotrophy estimates for C. linariifolia range fi-om 16% to 69%. We found large interannual differences in photosyntlietic rates and carbon isotope ratios for C. linariifolia and A. tridentata, which most likely indicated a response to precipitation dif- ferences. Climate records showed that the grow- ing season at Strawbeny Resenoir in 1994 was consideral)ly drier than in 1995; the spring (March-May) of 1994 received only 96.3 mm of precipitation, whereas precipitation in the spring of 1995 totaled 216.4 mm (Utah Climate Center, Meber station). Differences in precipi- tation during the spring influence the amount of soil water available to the plants. This water supply can be indirectly assessed by measuring the plant's water potential before the sun rises and photosynthesis commences. Our "^Pp^j mea- surements corroborated that 1994 was a drier growing season; during 1994 the ^ | range for A. tridentata (-1.7 to -3.2 MPa) was much more negative than the ^ j range for A. tri- dentata in 1995 at either Tintic (-0.7 to -1.1 MPa) or Strawberry Reservoir (-0.3 to -0.9 MPa). Photosynthetic rates doubled during 1995, presumably in response to this increased precipitation. Interannual differences wei^e most pronounced for C. linariijolia, which showed photosynthetic rates 3-fold higher in 1995 rela- tive to rates in 1994. Carbon isotope ratios for autotrophic C3 plants represent an estimate of long-term water-use efficiency (mmol C/mol H2O; WUE), with more negative S^'^C ratios reflecting a lower WUE (Ehleringer and Osmond 1989). S^'^C ratios for infected and uninfected A. tridentata were significantly more negative during the wetter year, thus suggest- ing they were less conservative in their water use. Using S^'^C ratios as a measure of water- use efficiency is inappropriate for hemipara- sites because of the potentially confounding effects of assimilating heterotrophic carbon. Therefore, it follows that the 8^'^C ratio for C. linariifolia should also reflect influences from the import of host-derived carbon rather than simply the influences of increased precipita- tion. This prediction was supported by C. lin- ariifolia data, where, despite the large increase in precipitation, we saw no difference in S^^C ratios between years. Photosynthetic rates also responded to sea- sonal inffuences, although rates were not dif- ferent between parasites and hosts. Photosyn- thetic rates declined during the growing sea- son, which, in part, may be attributed to the drier conditions late in the season as indicated 338 Great Basin Naturalist [Volume 56 Table 1. Carbon isotope ratios (S^-^C) for hosts and parasites. Sites and sample sizes are as follows: Uninfected A. tri- dentata (Strawberry Reservoir 1994; n = 5; Strawberiy Resei-voir 1995: n = 3 for early season; Tintic 1995: »! = 4 for early season, n — 5 for late season), infected A. tridenfata (Strawbeny Reservoir 1994: n = 5; Stravvberr\' Reservoir 1995: n = 20 for early season, n — 30 for late season; Tintic 1995: n = 10 for early season, n = 11 for late season), C. linariifolia (Strawbeny Reservoir 1994: n = 9; Strawbeny Resei-voir 1995: n = 7 for early season, n = 8 for late season; Tintic 1995: 7i = 11 for early season, n — 8 for late season), and O. tobniei (Strawberrx* Resenoir 1995: 0=9 for early season, n = 19 for late season). Letters denote significant seasonal differences within a site and species (Tukey-Kramer HSD, a — 0.05). Values shown are means ± 1 .s^. NA denotes data not available. A. tridentata Site Year Lhiinfected Infected C. linariifolia O. tobniei 1994 -25.56 ± 0.32« -26.24 ± 0.13« -28.93 ± 0.34^' NA 1995 Early -27.S6±0.ll' -27.30 ± 0.11'' -28.91 ± 0.15^' -28.66 ± 0.09^' Late NA -27.80 ± 0.09^ -29.33 ± 0.25« -29.50 ± 0.10' 1995 Earlv -27.57 ± 0.19'' -27..32 ± 0.2ll'^' -29.19 ± 0.22^1 NA Late -27.33 ± 0.19'' -27.17 ± 0.16'' -29.23 ± 0.1.5-' NA Strawberry Resenoir Tintic by predawn water potentials. Perhaps, the de- cline in C. linariifolia photosynthesis was also related to the phenology' of the hemiparasite. It is possible that late in the season when these hemiparasites set fruit, they rely less on cur- rent photosynthesis and more on heterotrophic carbon gain. Most studies of hemiparasite-host gas-exchange dynamics found that hemipara- site photosynthesis was much lower than that of the host'(Hollinger 1983, Press et al. 1987b, Pate et al. 1990, Marshall, Dawson, and Ehle- ringer 1994). S. hennonthica and S. asiatica have photos>'nthetic rates that are half of those for Sorghum hosts (Press et al. 1987b). In con- trast, the photosynthetic activities of C. linari- ifolia in this stud}' were similar to rates of A. tridentata hosts. This pattern remained stable from year to year, despite large differences in precipitation. Hemiparasite gas-exchange rates have been used to make inferences about potential het- erotrophic carbon use. After calculating that 8.8-18.9 h of light-saturated photosyntliesis was necessaiy for 8 different species of facultative hemiparasites to reach zero net foliar carbon gain. Press et al. (1988) hypothesized that they must have had access to a heterotrophic car- bon supply. Converseh; in Baii.sia trixago and Parentucellia viscosa (Scrophulariaceae), where photosyndietic rates were veiy similar to auto- trophic plants, it was predicted that these fac- ultative root hemiparasites were less reliant on host-derived carbon (Press et al. 1993). Since C. linariifolia also has photosynthetic rates sim- ilar to those of its host, it follows that C. linari- ifolia might not contain significant amounts of heterotrophically derived carbon. However, in our study this was not the case. We found a relatively large difference in 'b^^C ratios between C. linariifolia and A. tridentata hosts, which likely indicates hemiparasite het- erotrophy. Indeed, we calculated that C. linari- ifolia in this study utilized an average of 40% host-derived carbon. As with other parasitic plants, unusually high transpiration rates rela- tive to the hosts represent the most likeK' dri- x'ing force for this assimilation of host-deri\ed carbon. While the estimates of heterotrophy found in this study are well within the range of those reported for other parasites, one must consider the inherent obstacles in using an instantaneous measure of photosynthesis as a basis for the predicted 8^^C ratios with an inte- grated measure of actual leaf 8^'^C ratio. For instance, differences in gas-exchange charac- teristics at the time leaf carbon was incorpo- rated may contribute to differences between predicted and measured S^'^C ratios. Although, we found no significant difference between pre- dicted and measured 8^'^C ratios for infected and uninfected Aiieniisia, a better contiol would have been autotrophic C. linariifolia plants if they had been a\ ailable. As mentioned earlier, parasites may also access different pools of car- bon at different times throughout the growing season; in turn, this ma\' influence the S^'^C ratios measiued in the leaf carbon. While these factors may appear troublesome at first, the\' represent a few of the many areas open to investigation in parasitic plant ecophysiolog)'. 1996] Hetekotkoimiv in Castilleja and Ortiiocarpus 339 Table 2. Measvired and predicted S^-^C values for iiniii- Iccted and infected A. thdentutd {n = 5) and C'. linariifolia ui = 9) at Strawhem Reserxoir in 1994. Means ± 1 s^ are presented. Also shown is the difference between the pre- dicted and measured values. * denotes significant difler- ence at P < 0.05 (paired t test). A. tridcntdta Uninfected Infected C. liiiuriijolia ■^predicte Difference -25.56 ± 0.32 -0.98+1.19 -26.24 ±0.13 1.13 ±0.94 -28.93 ± 0.34 -1.34 ±0.48* Though no other study quantifies hetero- trophic carbon gain by a facultative hemipara- site, a study by Hansen (1979) impHed potential heterotroph)' in Castilleja chromosa. Experi- ments measuring the difference of ^"'C labeled sugar content in uninfected and infected Artemisia tridentata indivdduals showed less ^"^C in the infected host tissues. Hansen (1979) hypothesized that this difference represented sugar lost to the C. chromosa parasite. With this indirect method, C. chromosa utilized, on average, 10% host-derived carbon. Using a more precise method, we would suggest from our study that 10% heterotrophy may be an underestimate. Significant heterotrophic carbon gain by the hemiparasite can be associated with a decrease in host production. Graves et al. (1989) found that dry weight of Sorghum infected with S. hermonthica was 40% less than that of unin- fected Sorghum, and hypothesized that the effects of S. hermonthica were due to (1) the direct reduction in host carbon by parasite het- erotroph}' and (2) the indirect reduction of host photosynthetic potential. Press and Stewart (1987) showed that photosynthetic rates for Sorghum infected by S. hermonthica were re- duced by nearly half relative to those for unin- fected Sorghum; stomatal conductance rates were also significantly decreased. In contrast, we saw no decrease in photosynthetic rates nor stomatal conductance rates for infected A. tri- dentata. Interestingly, there was an increase in host photosynthesis relative to uninfected A. tridentata late in the season at Tintic. Our study also showed no difference in '^ | between infected and uninfected A. tridentata, suggest- ing that hosts in this study were not experienc- ing detectable water stiess. Taken togedier these Tablk 3. Calculated heterotropliy of C. liimhifolia in this study compared to heterotropiu' calculated for other hemiparasitcs. .Species Calculated heterotrophy in % (range) CastiUcja linariijolia Striiid Iwnnontlticd, Striga asiatica Phoredendron jiinipcrinum Mistletoe species Australian mistletoe 40(16-69) 28-35 61 60 (49-67) 15 (.5-21) This study Press et al. 1987, Graves et al. 1989 Marshall and Ehleringer 1990 Schulze et al. 1991 Marshall etal. 1994b data seem to suggest that C Unariifoha do not negatively impact A. tridentata hosts. How- ever, this conclusion may be relevant only dur- ing unusually wet years; A. tridentata may respond differently to hemiparasite infection when drought conditions prevail. One well-supported aspect of the host-para- site relationship is the unusually high transpi- ration rates of tlie parasite, often 10 times greater than those of the host. It is generally believed that this high water flux results in a water potential gradient from the host to the parasite. Therefore, through this mechanism higher transpiration rates are thought to represent the driving force for the transfer of solutes from die host to parasite. Schulze et al. (1984) sug- gested that high transpiration rates may be necessary for mistletoe to acquire adequate nitrogen for growth. The nitrogen-gathering hypothesis has been the focus of several stud- ies (Schulze et al. 1984, Ehleringer et al. 1985, Marshall, Dawson, and Ehleringer 1994). However, as Raven (1983) points out, these plants are also inextricably acquiring signifi- cant amounts of host carbon. Recent studies indicated that heterotroph)' may be a wide- spread phenomenon occurring in a variety of obligate hemiparasitcs (Press et al. 1987a, Graves et al. 1989, Marshall and Ehleringer 1990, Marshall et al. 1994, Richter et al. 1995). Evidence from this study indicates that the facultative root parasite C. linariifolia obtains a substantial contribution of host-deri\'ed carbon, thus extending further emphasis to the impor- tance of this carbon supply for hemiparasitcs. 340 Great Basin Naturalist [Volume 56 Acknowledgments The authors thank Nina Buchmann and J. Renee Brooks for critical review of the man- uscript and assistance with statistical analysis; Sue Phillips, Nathan Richer, and Sylvia Torti for field assistance; and Craig Cook for assist- ing with stable isotope measurements. All pre- cipitation data were generously provided by Utah Climate Center in Logan, Utah. Literature Cited Atsatt, E R., and D. R. Strong. 1970. The population biology of annual grassland hemiparasites. I. The host environment. Evolution 24: 278-281. Ehleringer, J. R., and C. B. Osmond. 1989. Stable iso- topes. Pages 218-290 in R. W. Pearcy, J. R. Ehle- ringer, H. A. Mooney, and P W. Rundel, editors. Plant physiological ecolog>'. Chapman and Hall, New York. Ehleringer, J. R., E. D. Schulze, H. Ziegler, O. L. Lange, G. D. Farquhar, and I. R. Covvar. 1985. Xylem-tapping misdetoes; water or nutrient parasites? Science 227: 1479-1481. Farquhar, G. D., M. H. O'Leary, and J. A. Berry. 1982. On the relationship between carbon isotope discrimi- nation and the intercellular carbon dioxide concen- tration in leaves. Australian Journal of Plant Physiol- ogy 9: 121-137. GoviER, R. N., M. D. Nelson, and J. S. Pate. 1967. Hemi- parasitic nutrition in angiosperms I. The transfer of organic compounds from host to Odontites verna (Bell.) Dum. (Scrophulariaceae). New Phvtologist 66: 285-297 Graves, J. D., M. C. Press, and G. R. Stewart 1989. A car- bon b;Jance model of the Sorghiiin-Striata. Oecologia 60: 396-400. Hull, R. J., and O. A. Leonard. 1964. Physiological aspects of parasitism in mistletoes {Arceiithobiuin and Phora- dendron) I. The carbohydrate nutrition of mistletoe. Plant Physiology 39: 996-1007. Marshall, J. D., and J. R. Ehleringer. 1990. Ai-e .xylem- tapping mistletoes partially heterotrophic? Oecologia 84: 244-248. Marshall, J. D., T. E. Dawson, and J. R. Ehleringer. 1994. Integrated nitrogen, carbon and water relations of a xylem-tapping mistletoe following nitrogen fertil- ization of the host. Oecologia 100: 430^38. Marshall, J. D., J. R. Ehleringer, E. D. Schulze, and G. Farquhar. 1994. Carbon isotope composition, gas exchange and heterotrophy in Australian mistletoes. Functional Ecology 8: 237-241. Pate, J. S., N. J. Davidson, J. Kuo, and J. A. Milburn. 1990. Water relations of the root hemiparasite Olax pliyllanthi (Labill) R.Br (Olacaceae) and its multiple hosts. Oecologia 84: 186-193. Press, M. C, and G. R. Stewart. 1987. Growth and pho- tosynthesis in SorgJiinn bicolor infected with Striga hennonthica. Annals of Botany 60: 657-662. Press, M. C., J. D. Graves, and G. R. Stewart. 1988. Transpiration and carbon acquisition in root hemipar- asitic angiosperms. Joiunal of Experimental Botanv 39(205): 1009-1014. Press, M. C, A. N. Parsons, A. W. Macxay, C. A. Vincent, V Cochr.'VNE, and W. E. Seel. 1993. Gas exchange characteristics and nitrogen relations of two Mediter- ranean root hemiparasites: Bartsia trixago and Paren- tucellia viscosa. Oecologia 95: 145-151. Press, M. C, N. Shah, J. M. Tuohy, and G. R. Stewart. 1987a. Carbon isotope ratios demonstrate carbon flux from C_^ host to C3 parasite. Plant Phvsiology 85: 1143_1145. . 1987b. Gas exchange characteristics of the Sorghitm-Striga host-parasite association. Plant Phys- iologx- 84: 814-819. Raven, J. A. 1983. Phytophages of .xylem and phloem: a comparison of animal and plant sap-feeders. Advances in Ecological Research 13: 136-239. RiCHTER, A., M. Popp, R. Mensen, G. R. Stewart, .vnd D. J. VON Willert 1995. Heterotrophic carbon gain of the parasitic angiosperm Tapinanthiis oleifolius. Australian Journal of Plant Ph\siolog>' 22; 537-.544. Schulze, E. D., O. L. Lange, H. Zeigler, .vnd G. Gebauer. 1991. Carbon and nitrogen isotope ratios of misdetoes growing on nitrogen and non-nitrogen fixing hosts and on CAM plants in the Namib desert confimi par- tial heterotrophy. Oecologia 88: 457^62. Schulze, E. D., N. C. Turner, and G. Glatzel. 1984. Carbon, water and nutrient relations of two mistle- toes and their hosts: a hypothesis. Plant, Cell and Environment 7: 293-299. Shah, N., N. Smirnoff, and G. R. Stewart. 1987. Photo- synthetic and stomatal characteristics of Sfrigo her- monthica in relation to its parasitic habit. Physiologi- cal Plantarum 69: 699-703. Smith, D., L. Muscatine, D. Lewis. 1969. Carbohydrate movement from autotrophs to heterotrophs in para- sitic and mutualistic s\mbiosis. Biological Re\ iew 44; 17-90. Welsh, S. L., N. D. Atwood, L. C. Hic;gins, .and S. Good- rich. 1987. A Utali flora. Brighani Young University, Prove, UT. 577 pp. Received 22 January 1996 Accepted 10 June 1996 Great Basin Naturalist 56(4), © 1996, pp. 341-347 HABITAT AND SPATIAL RELATIONSHIPS OF NESTING SWAINSON'S HAWKS {BUTEO SWAINSONI) AND RED-TAILED HAWKS (B. JAMAICENSIS) IN NORTHERN UTAH Thomas Bosakowski', R. Douglas Ramsey^, and Dwight C. Smitlv^ Abstract. — A total of 28 Swainson's I lawk {liiiteo swainsDiii) and 30 Red-tailed Hawk {B. jainaiccnsis) nests were found in Cache Valley, Utah, during the summers of 1992 and 1993. All nests were in trees, hut only Hed-tailed Hawks nested in dead trees (30^). In the intensive study area, ncstinu; densities were 0.10 nests/km- for Swainson's Hawk and 0.08 nests/km- for Red-tailed Hawk. Nearest-neighhor nest distances were significantly shorter among Swainson's Hawks (1.74 km) than among Red-tailed Hawks (2.83 km). Congeneric nearest-neighhor distances were significantly shorter tlian conspecific distances for Red-tailed Hawks (1.59 vs. 2.83 km) hut not for Swainson's Hawks (1.52 vs. 1.74 km). CIS analysis of hahitat types was made for 2-km radii around nest sites. Cropland was the dominant land cover type at nest sites of both species and no significant difference was found between species. Swainson's Hawk nest sites contained significantly more pasture, whereas Red-tailed Hawk nest sites contained significantly more juniper, maple, and sagebrush. Only Red-tailed Hawk nests {n — 8; 27%) were found on the periphery of the valley at the base of foothills of the Cache Mountains. This preference resulted in a significantly higher elevation for Red-tailed Hawk nest sites. Swainsons Hawk nests occurred only on the valley floor on level teiTain. Distance to the nearest paved road and building was veiy similar for both species, implying that little difference exists in tolerance levels for human activities. Overall, multivariate niche overlap for habitat was high (0.89), indicating a lack of habitat partitioning between these 2 Buteos in Cache Valley. Key words: Swainsons Hawk, Red-tailed Hawk. Buteo, nest sites, liabitat, CIS. Relatively few studies have included a com- parison of nest sites, habitat, or densities of Swainson s Hawks {Buteo swoinsoni) and Red- tailed Hawks {B. jamaicensis). Rothfels and Lein (1983) and Bechard et al. (1990) examined nearest-neighbor distances of these 2 species in Alberta and Washington, respectively, and Bechard et al. (1990) also compared habitats. Janes (1985) examined habitats associated with sightings of these 2 Buteos in Oregon. Consid- ering that these species are sympatric through- out much of their range in western North America, further information on their habitat use and nesting density in overlapping regions would be useful for understanding patterns of coexistence. The present study is also important because the Swainson's Hawk is considered to be de- clining in Utah, Nevada, and Oregon, and its status is listed as a "species of special concern " in Utah, Nevada, Oregon, and Washington, and "threatened " in California (Harlow and Bloom 1989). Conversely Red-tailed Hawks are con- sidered to be increasing (Harlow and Bloom 1989) due to an increase in perching habitat, at the expense of Swainson's Hawks (Janes 1985, 1987). Therefore, a comparative approach to the nesting ecology of these 2 species is not only of ecological importance but has implica- tions for the ftiture conservation of Swainson's Hawks. Study Area and Methods The study was conducted in the Cache Val- ley portion of Cache County in northern Utah (Fig. 1). The valley comprises cropland (alfalfa, hay, winter wheat, corn), pasture, grassland, marsh, sagebrush-grassland, barnyards/feed- lots, residential areas, and commercial com- plexes. During the summers of 1992 and 1993, we conducted a vehicle survey of the entire valley by driving on primary and secondary (dirt) roads along the valley floor and lower benches. Searches did not extend into moun- tainous terrain. lUtah Divisidii of Wildlife Resources, 146.5 W. 200 N., Logan, UT 84321. Present address: Beak Consultants, Inc., 12931 N.E. 12fitli Place, Kirkland. WA 98034-771.5. ^Department of Geography and Earth Resources, Utah State University; Logan, UT 84322. -^Biology Department. Southern Connecticut State University, New Haven, CT 06515. 341 342 Great Basin Naturalist [Volume 56 Fig. 1. GIS shaded relief map of the Cache Valley study area in northern Utah showing the distribution of Swainson's Hawk (circles) and Red-tailed Hawk (triangles) nests for 1992 and 1993 nesting seasons combined. Light lines indicate primary and secondaiy roads. Occupied nests were found by scanning likely trees, especially if adults were seen near- by or protested our approach. A nest was con- sidered occupied only if an incubating or brooding female and/or young were present. Nests were not examined for the presence of eggs. Because of the low density of ti^ees (200 m). Habitats were classified by the Utah Division of Water Resources (1991) into 11 major habitat types: cropland, fallow field, grassland, pasture, sage- brush, juniper {Jiiniperus spp.), maple {Acer granclidentatinn), riparian (wetlands, temporaiy marshes, mud flats), open water, residential, and commercial (non-residential buildings, indus- trial stiiictures, junkyards, and parking lots). Statistical analysis was performed on NCSS software (Number Cruncher Statistical Soft- ware, Kaysville, UT). Prior to analysis, habitat variables were tested for normality (D'Agostino 1990). A number of data transformations were attempted (Zar 1984), but none were able to normalize all variables. Therefore, a non-para- metric rank test (Mann-Whitney C/-test, 2- tailed) was selected for all habitat comparisons. To calculate habitat overlap from the GIS data variables, a full-model (all variables included) discriminant function analysis (DFA) was run to detemiine the extent of habitat partitioning between the 2 species. Multivariate niche overlap for habitat was calculated with log- transformed variables with the following for- mula presented by Maurer (1982): overlap = exp (-d^ /SI + S2); where d = the difference between mean dis- criminant scores for species 1 and 2, and S = the standard deviation of the discriminant scores. Maurer (1982) and Klopfer and Ganzhorn (1985) suggested that stepwise pro- cedures that eliminate variables always result in a biased underestimation of niclie overlap. Therefore, we used a full-model DFA instead of stepwise DFA because it considers the whole spectrum of habitat variables available for partitioning. Results and Discussion We located 58 occupied nest sites during 1992 and 1993 field seasons: 28 Swainson's Hawk nests and 30 Red-tailed Hawk nests. In a single breeding season (1992), a maximum of 22 occupied nests were found for Swainson's Hawks and 23 for Red-tailed Hawks. M\ nests were in trees, although a few cliff sites were available in the study area but not occupied. Only Red-tailed Hawks nested in dead trees (9 of 30 trees), which was statistically significant because Swainson's Hawks nested only in live trees (Fisher Exact Test, 2-tailed, P = 0.002). Red-tailed Hawks nested higher aboveground and in taller trees, but tree diameter was not significantly larger (Table 1). The intensive study area was completely searched for occupied nests in 1992 and con- tained 10 Swainson's Hawk nests and 8 Red- tailed Hawk nests (Fig. 2). Absolute nesting density in this area was 0.10 nests/km^ for Swainson s Hawks and 0.08 nests/km- for Red- tailed Hawks. Gilmer and Stewart (1984) reported a nesting density of 0.055 nest/km^ for Swainson's Hawk, which was almost half the density found in Cache Valley. Luttich et al. (1971) reported a nesting density of 0.145 red-tailed nestsAni^, which is higher than our study area. Rothfels and Lein (1983) reported nesting densities of 0.238 nests/km^ for Swain- son's and 0.508 nests/km^ for Red-tailed Hawks, which were much greater than the density for Swainson's and Red-tailed Hawks in Cache Val- ley. This difference probably reflects the fact that hawk nests can be dispersed because of areas of unsuitable and marginal habitat (e.g., note that nests were not located within areas of dense suburban road networks in Fig. 2). Newton (1979) stated that in continuously suitable habitat the nests of the same species are often separated by roughly equal distances. Mean nearest-neighbor distance (Clark and Evans 1954) is the measure that can be used to 344 Great Basin Naturalist [Volume 56 Table 1. Nest tree and topographic variables of Svvainson s and Red-tailed Hawk nest sites in northern Utali. Data represent means ± s with sample size in parentheses. Red-tailed Hawk Swainson's Hawk p. Nest tree height (m) 17.3 ±4.1 (25) 13.9 ± 2.9 (23) 0.001 Nest height (m) 14.8 ± 3.4 (23) 11.3 ±3.3 (21) 0.002 Nest tree DBH (cm) 87.2 ± 39.4 (22) 75.5 ± 44.6 (23) 0.226 Distance to paved road (m) 393.6 ± 580.9 (30) 311.6 ±484.2(23) 0.133 Distance to building (m) 246.1 ±174.3 (30) 250.4 ± 174.2(24) 0.649 Elevation (m) 1401 ± 193.6 (29) 1373 ±31.4 (27) 0.001 ^Mann-Wliitnev I'-tt-st, 2-tailfcl quantify these spacing patterns. In Cache Val- ley we found a significant difference (Student's t test, t = 2.61, P < 0.02) for this distance, which was 1.74 km for Swainson's Hawks {n = 10) and 2.83 km for Red-tailed Hawks {n = 8). In Alberta, Rothfels and Lein (1983) reported mean nearest-neighbor distances of 1.46 km for Swainson's Hawks and 0.88 km for Red- tailed Hawks. These results are similar to Swciin- son's Hawks in Cache Valley but are much shorter than our estimate for Red-tailed Hawks. Rothfels and Lein (1983) noted that their data on Red-tailed Hawks showed a much denser population than nomial. The mean for 7 other Red-tailed Hawk studies was 1.95 km (data from Rothfels and Lein 1983), which is closer to the distance found for Cache Valley. The mean for 5 other Svvainson s Hawk studies is 1.78 km (data from Rothfels and Lein 1983), veiy close to the mean for Cache Valley. Over- all, nearest-neighbor distances from our study area were consistent with the majorit>' of liter- ature values, demonstrating the regular disper- sion of nest sites that results from territorial behavior (Newton 1979). Congeneric nearest-neighbor distances were significantly shorter than conspecific distances for Red-tailed Hawks (1.59 km vs. 2.83 km) but not for Swainson's Hawks (1.52 km vs. 1.74 km; Student's t test, ^ = 2.18, P = 0.047 and t = 0.78, P = 0.44, respectively). These results sug- gest that Red-tailed Hawks are more tolerant of close nesting by Swainson's Hawks than their own species, but Swainson's Hawks are equally intolerant to congeners and conspecifics. In Alberta, Schmutz (1977) and Rothfels and Lein (1983) found that congeneric Buteos nested closer together than conspecifics probably be- cause competition among congenerics was less than among conspecifics. With regard to distribution in the study area, only Red-tailed Hawks (27%; n = 8) nested above the vallev floor at the base of foothills of tlie Cache Mountains (Fig. 1), and this difference resulted in a statistically significant increase in elevation (Table 1). Swainson's Hawks did not nest in this zone at all, possibly because many of these sites were already occupied by earlier- nesting Red-tailed Hawks or because of habitat preferences discussed below. Rothfels and Lein (1983) noted qualitatively that Swainson's Hawks usually nested on flatter tenain than Red-tailed Hawks. In this study, Swainson's and Red-tailed Hawk nests lacked a significant difference for the distance to nearest buildings or paved roads (Table 1). No previous studies of these 2 species have been conducted in areas with this much urbanization. Our data suggest that no signifi- cant differences exist in regard to tolerance of human activities and structures. Overall, the CIS indicated that habitat around nest areas was dominated by cropland and pasture for both Buteos (Fig. 3). Swainson's Hawk nest sites had significantly more pasture (22.4% vs. 12.3%) but not cropland, fallow field, or grassland. In eastern Washington, Bechard et al. (1990) noted that Swainson's Hawks uti- lized wheatland and grassland more than Red- tailed Hawks. In this study Red-tailed Hawks nested in areas with significantly more tree cover (maple and juniper) and sagebiTish, which predominated uplands along the edge of the valley floor. The importance of trees to Red- tailed Hawks was noted by Houston and Bechard (1983), who documented the increase in nesting by this species after the expansion of trees into the prairie regions of Saskatchewan. Similarly Knight et al. (1982) found that Red- tailed Hawks nested exclusively in riverine for- est land along the Columbia River, even though suitable cliff nesting areas were available. Janes (1985) noted that Swainsons Hawks depended more on aerial foraging and occurred in habi- tats containing few or no perches. In this study Red-tailed Hawks probably nested more in tree habitats because of greater perch availability/ 1996J Svvainson's and Red-tailku Hawk Nesting 345 [B Red-tailed Hawk m Swainson's Hawk N Fig. 2. CIS road map of 100-kni- intensive study area centered at Logan Municipal Airport sliowing nest site locations for Swainson's and Red-tailed Hawks during the 1992 breeding season. use, although other factors such as larger prey species may also be a factor. Connell (1980) explained that resource par- titioning (or low overlap) is due to the "ghost of competition past" (past competition), which has created evolutionary changes in morphology and behavior to avoid cuirent competition. Since raptors are at the top of the food pyramid and occur at extremely low breeding densities (Newton 1979, Scheoner 1984), resources are likely to be limiting, and high overlap in re- sources between species is likely to result in current competition (except in rare cases such as vole plagues). Despite some significant dif- ferences for 4 of the 12 habitat types, we calcu- lated a multivariate (DFA) niche overlap of 0.89 for habitat. Niche theory suggests that overlap values higher than 0.6 are needed to cause interspecific competition, while lower values indicate undemtilization of the resource continuum resulting in intense intraspecific competition (see reviews by Bosakowski et al. 1992, Bosakowski and Smith 1992). Prey overlap data were not collected in our study area, but Smith and Murphy's (1973) data from northern Utah showed a high prey overlap value of 0.80 for Red-tailed and Swain- son's Hawks (as calculated by Jaksic 1983). In Montana, Restani (1991) found an even higher food overlap (0.93) for tliese 2 Biiteos. Consid- ering the findings of high overlap for food (Jak- sic 1983, Restani 1991) and habitat (this study), competition between these Butcos sliould be expected whenever the species occur in close proximity. As further evidence, Schnuitz et al. (1980) found that reproductive performance was significantly reduced in cases where these Biiteos nested at close range. Due to man-made alterations, few of the na- tive plant communities presently exist in Cache Valley. Not surprisingly, we did not observe significant habitat partitioning between these 2 Biiteos for the existing habitat t\pes. Elsewhere, investigators have claimed that significant habi- tat partitioning (non-overlap) occuned between these Buteos in Oregon (Janes 1985) and Wash- ington (Bechard et al. 1990), but the extent of habitat overlap was not previously quantified. Our results indicate that statistical tests can show differences among several habitat vari- ables, while the overlap value can still remain critically high. Competition for habitat has also been demonstrated by behavioral observations of Swainson's Hawks frequently usurping por- tions of Red-tailed Hawk temtories with lower perch densities (Janes 1994). Alternately, Janes (1985, 1987) noted that the increase in perch- ing habitat, caused by the spread of junipers, homesteads, and utility poles, "favors the Red- tailed Hawk at the expense of the Swainson's Hawk. " In addition, Janes (1994) also reported that territorial Swainson's Hawks are occasion- ally displaced by Red-tailed Hawks. Bednarz (1988) noted that availabilit)' of nest trees could be a limiting factor for Swainson's Hawks because of their affinity for open grass- land and desert habitats that are often devoid of trees. Similarly, Houston and Bechard (1983) reported the expansion of Red-tailed Hawks in Saskatchewan following the spread of trees into prairie regions. In our study area the west- ern portions of the valley floor were often tree- less and usually supported little nesting for either species (note lower density of nests in Fig. 1). For such situations Schmutz et al. (1984) recommended installation of artificial nest plat- forms for Swainson's Hawks, which signifi- cantly increased nesting densit>' in his experi- ments. However, if artificial nest platforms are used, we recommend caution and close moni- toring so as not to give advantage to the more 346 Great Basin Naturalist [Volume 56 Hectares 300 400 700 CROPLAND Fig. 3. Habitat areas around nest sites (2 km radius) of Svvainsons {n — 26) and Red-tailed Hawks [n = 2S) from Cache Valley, Utah, as determined from CIS analysis. Bars represent the mean and stars indicate that a significant difference was observed between species (Mann-Whitney t/-test, 2-tailed, P < 0.05). common Red-tailed Hawks. In our study area only Red-tailed Hawks nested in snags (30% of occupied nests) and may be more likely to use an open-topped artificial platform than Swain- son's Hawks, which always nested in green trees. Many of the snags used by Red-tailed Hawks in Cache Valley were caused by failure to irrigate croplands during recent drought conditions, thus changing the suitability of nest sites in favor of Red-tailed Hawks. In the future close attention to inigation and surveillance of land-use changes are likely to be the most important factors in conserving Swainson's Hawks in Cache Valley. Economic conversion of agrarian land use to commercial and residential real estate is currently in pro- gress, and impacts to future Swainson s Hawk populations need to be carefully monitored. Due to the rapid human population growth in Cache Valley we recommend annual monitor- ing for Swainson's Hawk territories and nests, which may be impacted by future development or land-use changes. This monitoring manage- ment will recjuire frequent updating of the CIS database to track habitat changes in the futuie so that necessary mitigation steps can be evalu- ated. Acknowledgments This study was entirely a volunteer effort completely funded by the authors except for the use of the CIS system at Utah State Uni- versity. We thank J. C. Bednarz, M. J. Bechard, and R. C. Whitmore for rexiewang the manu- script and providing helpful comments and criticisms. Literature Cited Bechard, M. J., R. L. Knight, D. G. Snhth, and R. E. FiTZNER. 1990. Nest sites and habitats of sympatric hawks {Bitten spp.) in Washington. Journal of Field Ornitholog) 61: 159-170. Bednarz, J. CI 1988. Swainson's Hawk. Pages 87-96 in Proceedings of the southwest raptor management SNinposiiun. National Wildlife Federation, Scientific and Technical Series No. 11. Washington, DC. BosAKOwsKi, T, and D. G. Snhth. 1992. Comparative diets of SNinpatric nesting raptors in the eastern 1996] Swainson's and Rkd-tailkd Hawk Nesting 347 (lec'icluoiis forest l)ioini\ Canadian [oninal of ZooIogN' 70: 9S4-992. BosAKOWSKi, T, D. G. Smuii, and K. Spkiskh. 1992. Niche o\erlap of tvvo sympatiic-nt'stinK liavvks Accipiter spp. in tlic New Jersey-New York llitililaiuls. I-lcoi^rapln 15: 358-372. CUAHK, P [., AND E C. Evans. 1954. Distance to nearest neighbors as a measure of spatial relationships in populations. Ecology' 35: 445-453. CONNELL, J. H. 1980. Diversity and tlie cocNolntion of competition, or the ghost of com|ietition past. Oikos 35; 131-138. Cr.'\ighe.\d, J. J., .\NU E C. Ckak:iieai), Jk. 1956. Hawks, owls and wildlife. Stackpole Publishing Co., Harris- burg, PA. D'Agostino, R. B. 1990. A suggestion for using powerful and informative tests of normality. American Statisti- cian 44: 316-322. FiTZNER, R. E. 1980. Behavioral ecolog> of the Swainson's Hawk in southeastern Washington. Unpublished doc- toral dissertation, Washington State University; Pull- man. 194 pp. GiLjMER, D. S., and R. E. Stewart. 1984. Swainson's Hawk nesting ecology in North Dakota. Condor 86: 12-18. Harlow, D. L., and R H. Bloom. 1989. Biiteos and Golden Eagle. Pages 102-110 in Proceedings of the western raptor management symposium and workshop. National Wildlife Federation, Washington, DC. Houston, C. S., and M. J. Bechard. 1983. Trees and the Red-tailed Hawk in southern Saskatchewan. Blue Jay 41: 99-109. Jaksic, E M. 1983. The trophic structure of sympatric assemblages of diurnal and nocturnal birds of prey. American Midland Naturalist 109: 152-162. Janes, S. W 1985. Habitat selection in raptorial birds. Pages 159-190 in M. L. Cody, editor. Habitat selec- tion in birds. Academic Press, Inc., Orlando, EL. . 1987. Status and decline of of Swainsons Hawks in Oregon: the role of habitat and interspecific competi- tion. Oregon Birds 13: 165-179. . 1994. Partial loss of Red-tailed Hawk territories to Swainson's Hawks; relations to habitat. Condor 96: 52-57. Klopfer, P H., and J. H. Ganzhorn. 1985. Habitat selec- tion; behavioral aspects. Pages 436-454 in M. L. Cody, editor. Habitat selection in birds. Academic Press, Inc., Orlando, EL. Knight, R. L., D. G. S.\irni, and A. Ehickson. 1982. Nest- ing raptors along the Columbia River in ncjrth-cen- tral Washington. Murrelct 63: 2-8. LurnGii, S. N., L. B. Keith, and J. D. Stephenson. 1971. Population dynamics of the Red-tailed Hawk (Btileo janiaiccnsis) at Rochester, Albc-rta. Auk 88: 75-87. Maurek, B. a. 1982. Statistical inference for MacArthur- Levins niche overlap. Ecology 63: 1712-1719. Newton, I. 1979. i\)pulation ecology of raptors. Buteo Books, Vermillion, SD. 399 pp. Restani, M. 1991. Resource partitioning among three Bitlco species in the Centennial Vallev, Montana. Condor 93: 1007-1010. ROTHFELS, M., AND M. R. Lein. 1983. Territoriality in sym- patiic populations of Red-tailc-d I lawks and Swainson's Hawks. Canadian Journal of Zoolog\' 61: 60-64. Schoener, T. W 1984. Size differences among sympatric, bird-eating hawks: a worldwide siuA'cy. Pages 254—281 in D. R. Strong et al., editors. Ecological communi- ties: conceptual issues and evidence. Princeton Uni- versity Press, Princeton, NJ 613 pp. SCHMUTZ, J. K. 1977. Relationships between three species of the genus Buteo (Aves) coexisting in the prairie- parkland ecotone of southeastern Alberta. Unpublished master's thesis, University of Alberta, Edmonton. ScHMUTZ, J. K., D. A. Moore, and A. R. Smith. 1984. Arti- ficial nests for Ferruginous and Swainson's Hawks. Journal of Wildlife Management 48: 1009-1013. ScHMUTZ, J. K., S. M. Sghmutz, and D. a. Boag. 1980. Coexistence of three species of hawks {Buteo spp.) in the prairie-parkland ecotone. Canadian Journal of Zoology 58; 1075-1089. U.S. Department of Interior. 1979. Snake River Birds of Prey special research report. USDI Bureau of Land Management, Boise, ID. 142 pp. Utah Division of Water Resodrces. 1991. Water-related land use inventoiy report of the Bear River river basin (Utah portion). Utah Department of Natural Resources, Salt Lake City, UT. Zar, J. H. 1984. Biostatistical analysis. Prentice-Hall, Engle- wood Cliffs, NJ. Received 12 July 1995 Accepted 30 May 1996 Great Basin Xatxinilist 56(4). © 1996, pp. 348-359 WESTERN BALSAM BARK BEETLE, DRYOCOETES COXFLSUS SWAINE, FLIGHT PERIODICITY IX NORTHERN UTAH E. Matthew Hansen^ Abstract. — The flight periodicit>' of western balsam bark beetle {Dryocoetes confusits Swaine) in Big Cottonwood Can\'on, Utali, was studied during tlie summer months of 1992, 1993, and 1994. Contents of baited funnel traps uere tallied b\' species up to 3 times weekh'. Two main periods of flight acti\it>- were obsened each >'ear. The first and, gener- ally; largest occun^ed in eaih' summer soon after flight was initiated for the season. A 2nd period was obsened in late summer, generalK .\ugust. Timing of die 2 periods was influenced b\' imusualK wann or cool weather in each stud\' year. The 1st period had more niiiles than females while the 2nd period had a majority of females. E.vcept during periods of cool or wet weaUier, western balsam bark beedes were found to be active at least at mininiiil levels fioni June through September Key words: Dnocoetes confusus, flidit periodicity. Scolyfidae. insect control, insect plienology. .\bies lasiocaipa, Utah forests. The western balsam bark beetle. Dri/ococtes confusits S\\aine (Coleoptera: Scol\ tidae), is a serious insect pest of tiTie firs. This insects life cycle is not full\ understood (Johnson 1985), howe\er, possibK due to the traditionalK low commercial \ alue of its host. In British Colmn- bia, for example, timber losses from western biilsiuii bark beetles ha\e onb' relati\eK' recently been calculated (Doidge 1981). The need to understand the life c\ cle and beha\ ior of this bai^k beede has increased in relation to die in- creased commercial and aestlietic \iilue of titie firs. Drought-subjected subalpine tin Abies hisio- ctirpa (Hook.) Nutt.. in northern Utah has been experiencing a western balsam bark beetle out- break that began in 1989. Most of the iiffected trees cu-e on the hea\ il> \ isited \\'asatch-Cache Nationid Forest, including die can\oiis east of Salt Liike Cit\- where picnic ai-eas, campgiounds, and ski resorts are common. This caused local forest managers to seek beede abatement mea- sures. Baik beede conti'ol sti-ategies, such as de- plo\ nient of semiochemicals or cultural treat- ments, rec^uire know ledge of the time frame in which beetles emerge fi-oni infested host mate- rial to attack new hosts. This is a report of 3 \r of western biilsam bark beetle flight periodicitx data from Big Cottonwood Can\on. Utah. Sex ratios, weather influences, and associated scoly- tids are also presented. Materials and Methods Fi\ e plots w ere established on .5—6 June 1992 in Big Cottonwood Canyon, Utah, ranging fiom 2000 to 2840 m ele\ation. Plots were selected from areas of recent beetle actixit)' indicated b\ lading or red subalpine fir crowns. The plot at 2000 m has a white fir [Abies concolor [Cord and Glend.] Lindl. ex Hildebr.)/ Douglas-fir (Pseiidoisuga )nciizicsii [Mirb.] Franco) mix and also had the least amount of fading host material of an\ plot. This is gener- alK the low er ele\ ational band for subalpine fir in Big Cottonwood Can>on. Small amoimts of subalpine fir Ciui be found immediateK" uphill fi"om the plot. The higher plots are dominated b> subalpine fir sometimes in association with Douglas-fir tjuiiking aspen (Popuhis trcnwloides Michx.), or Engelmann spruce [Picca cngcl- )nannii Pam" ex Engelm.). Each plot contained three 16-imit Lindgien fimnel traps- spaced at about 50-m inten ids in a triangular pattern. Traps were baited with a semiochemical mixture containing <^vo-bre^ i- coniin (racemic) released at 1 mg/24 h at 24 °C'^ (Borden et al. 1987). Traps were hung as high 'I SDA Forest Senici-. Iiitt-nuoiintain Rese;irch Station. S60 Nortli 12(K) East. U>,aan. ITS4321. -Phero Tech Inc.. Delta. B.C.. Canada, 3phero Tech Inc.. Delta. B.C. Ciuiada, 348 1996] Dryocoetes coxfusus Flight Periodicity 349 as possible on branches, lea\ inu tlie trap eup about 1.5 ni abcneground. Trap eups were emptied up to 3 times weekly to reduce losses to predation. Cups w^ere emptied less fiequentK- late in each stud\- \ear as captures diminished. Western balsam bark beetles were tallied aloni:; with associated scoK tids and important preda- tors, nameh' checkered beedes {Enoclerius spp.). Identification of associated scoK tids was pro- \ided b>" Stephen L. Wbod^. D. coufusus cap- tures for the entire season were tlien totaled for each plot. The percentage of the annual total caught at each obsen ation w as then plot- ted against date for each location. The stud\' was repeated starting in mid-Ma\ 1993 with 5 plots installed from 1750 to 2840 m ele\'ation. Two sites from 1992 were reused, 2 were mo\ ed a short distance (about 100 m horizontal), and 1 was new. The lowest ele\a- tion plot was deliberately established in the white fir zone. Low^-elevation plots were in- stalled earlier in the \ear than in 1992 to ensure placement before beetle flight commenced. Plots at 1750 and 2350 m were installed 25 Ma\- 1993, and the plot at 2560 m was estab- lished 27 Max 1993. The plot at 2660 m was instiilled on 8 June 1993 and the plot at 2840 m on 21 June 1993. The 1994 flight periodicit) stud\ utilized the 4 highest sites from 1992, ranging in ele\a- tion from 2310 to 2840 m. These areas contin- ued to contain fading host material throughout each stud\ \ ear The low-elex ation sites, lacking a substantial sulialpine fir component, were dropped from the stud\' due to the small popu- lations of D. confiisus in those areas. The plot at 2310 m was established on 10 May 1994, while the remaining plots were instiilled on 25 Ma\ 1994. This gi\es 3 consecuti\e >ears of flight period data for 4 locations. The first 10 D. confiisus from each trap cup obsenation, totaling 30 per plot, were tiillied for sex in 1993 and 1994. Females were identified b\' a prominent setiU biiish on the frons (Borden et al. 1987). For 1993 and 1994 the sex ratio of each distinct flight surge was compared. The di\'ision between surges was determined from each significant flight activity' lapse not associ- ated with cool or wet weather. Weather data from Brighton-SiKer Lake'^ was compared with flight acti\ it> for the plots at 2600/2660 m and 2840 m (this station is geo- graphicalK' and elevationalK' between the 2 plots). Daily maximum/minimum temperatures and dail\- precipitation were plotted from 20 Ma\- dirough 31 October for each stud\' year. Results Flight Periodicity 1750 Meters. — This site was used onl\ in 1993 with a total of 42 D. confiisus captures. ConsequentK; I deleted it from consideration for the puipose of this study. Nearb\ white fir mortalit) that was examined contained evi- dence onh' of fir engraver beetle, Scolytus ven- tralis LeConte. 2000 Meters. — Because this site had rela- ti\el> few captures, I used it only in 1992. Flight acti\ it>' for that year sharpK peaked in mid- to late June (Fig. 1). A substmitiiilK- smaller surge occurred in early August. D. confusus captures totaled 1469. The 1st wa\e of acti\it\^ accounted for 84% of total captures. 2310/2350 Meters.— The substantial num- ber of captures at the first obsenation of 8 June 1992 indicates diat flight was likeK- initi- ated before plot establishment (Fig. 2). Cap- tures peaked in mid- June with activity contin- uing throughout the month. A 2nd surge began in mid-July and tapered off in mid-August. D. confusus captures totaled 19,071. Fort)-one percent of the total occurred in the 1st surge. In 1993 traps at this plot began to capture beedes in mid-June widi a sniiill peak occuning in late June. A 2nd wave of captures began in late Jub, peaking in mid- to late August. D. confusus captures were about 9% of those in 1992, totaling 1800. The 1st wa\e of captures accounted for 10% of the totiil. Capture patteiTis of 1994 were very similar to those of 1992. The first positi\e obsen'ation was on 3 June 1994. An early sunuuer peak occuned on 13 June 1994 with acti\it\^ taper- ing off in late June through earl\- JuK'. A late- summer surge occuned in mid- to late July with captmes gradualb' diiuinishing through early October. D. confusus captures totaled 2574, with 30% caught in the 1st surge. 2560 Meters.— The first 1992 obsenation was positi\e, indicating that flight w as possibb' initiated before plot establishment. Beetles ■'Prottssor i-meritus. Life Science Museum and Department of Zoolog\, Brighani Young L'niversit)', Provo, UT. -^Salt Lake Cit\ Watershed .Management. 350 Great Basin Naturalist [Volume 56 50 I 30 S20 + §10 0) ■ ■I 20 6/1 15 7/1 15 8/1 15 1992 9/1 15 10/1 15 31 Fig. 1. Percentage of total (;; = 1469) seasonal Dnjococtes conjmm captures per obsei-vation at 2000 m, 20 May-31 October Arrow indicates the trough between surges not associated with cool or wet weather x,20 10 20 6/1 20 Q. iK s V) (U ^ 10 20 15 7/1 15 8/1 15 1992 9/1 15 10/1 15 31 ■ ■ ll illl .1 111 20 6/1 15 7/1 15 8/1 15 1993 9/1 15 10/1 15 31 X3 0> gist V) (U Si 5 - ^^ I I I 20 6/1 15 7/1 15 8/1 15 1994 9/1 15 10/1 15 31 Fig. 2. Percentage of total (nigc), = 19,071, n]yy3 = 1800, /i,c)94 = 2574) seasonal Dnjococtes confii,siis captures per observation at 2310/2350 n», 20 May-31 October. Arrow indicates die trough between surges not associated with cool or wet weather 1996J DmOCOETES CONFUSUS FL1(;11T PERIOIDlCITi' 351 were eau^ht in large numbers throughout June with a 2ncl surge of aetivity in early to mid- August (Fig. 3). D. confiisiis trap eaptures totaled 9164, with 669r captured in Hie 1st surge. In 1993 eaptures began in late June, peak- ing in early to mid- July. A 2nd, substantially larger wa\e started in late July and peaked fi-om mid-August through early September. Captures were about 41% of those in 1992, totaling 3763. Twelve percent of that total were caught in the 1st surge. The pattern of 1994 captures was similar to that of 1992 with a sharp peak occurring in mid- to late June. A 2nd wave began in late July, peaking in early August. D. confusus cap- tures totaled 4476, half of which were caught in the 1st surge. 2600/2660 Meters.— 1992 captures began in mid- June with a sharp peak occurring in late June (Fig. 4). A 2nd wave began in late July with a mid-August peak. D. confusm cap- tures totaled 7548, with 68% caught in the 1st surge. In 1993 activity began in late June with an early summer peak in mid-July A 2nd, substan- tially larger wave started in late July and peaked from late August through early September D. confusus captures totaled 5882, 16% coming in the 1st surge. In 1994 activity began in early June with a shaip peak in late June. A 2nd wave began in late July, peaking in early August. Captures were the fewest for any study year, totaling 1331. ^30 3 25 f S2O CO 0) ii993 = 5882, /iiyg^ = 1331) seasonal Dnjocoetes confiisus captures per obsei-vation at 2600/2660 in, 20 May-31 October. Airow indicates the trough between surges not associated with cool or wet weather. Si.xty percent of these were caught in the 1st surge. 2840 Meters.— In 1992 the 1st capture was obsei-ved on 22 June. Captures peaked in late June, and considerable activity continued tlirough early July (Fig. 5). A 2nd surge occuned in mid- to late August. Captures totaled 17,542 with 72% caught in the 1st surge. In 1993 activity began in early July with a very large peak occurring in mid- to late July. A 2nd surge occurred in mid- to late August. Captures were down fiom 1992 levels but were still substantial, totaling 10,344. Seventy-six percent of these were caught in the 1st surge. In 1994 flight initiated in mid-June with a distinct spike in late June. A late-summer surge began in late July and continued through mid- August. Captures were the greatest of any plot in an\' year, totaling 20,600. Sixty-seven per- cent were caught in the 1st surge. Surge Activity Considering onb' the 4 plots common to each study year, there is a trend for the 1st surge to be larger than the 2nd w ith increasing elevation (Table 1). The lowest ele\ation plot consistentK' captured more beetles in the 2nd wave. The highest plot, however, consistentK captured more beetles in the 1st surge. Weather Influences Periods of cold and/or wet weather coin- cided with a reduction or pause in beetle 1996] Dryocoetes coneusvs Flicht PERionicnv 353 ■o30 0) = 25 - 820 CO 0) 0) £10 i 5 + 0) Q. 30 = 25 I 820 CO 0) ^15 Z 10 - £? 5 30 B 25 8 20 ^15 y 5 CL „ ■ 1 r L , 1 n. . 1. 1 1 . 20 6/1 15 7/1 15 8/1 15 1992 9/1 15 10/1 15 31 ■ • ■ 1 f • II .1 20 6/1 15 7/1 15 8/1 15 1993 9/1 15 10/1 15 31 20 6/1 15 7/1 15 8/1 15 1994 9/1 15 10/1 15 31 Fig. 5. Percentage of total (/!i9g2 = 17,542, 711993 = 10,344, nig94 = 20,600) seasonal Dryocoetes confusu.s captnres per observation at 2840 m, 20 May-31 October. Arrow indicates the trough between surges not associated with cool or wet weather. captures (Figs. 6-8). Ver>' little flight occurred when daily maximum temperatures were less than 15 °C, confimiing Stock's (1991) findings. Lapses in flight activity between the main surges, however, are not necessarily associated with cool or wet weather. With 4-7 wk between peaks, warm, diy days were available during these spans of reduced flight. One would e.xpect delayed emergence and flight timing with increasing elevation. Initial captures at 2840 m were about 2-3 wk later than at 2310/2350 m each year. Timing of peak flight activity' was similarly delayed with increas- ing elevation (Figs. 2-5). Febmary through May 1992 was the wannest on record for that period in northern Utah. June through August 1993 was the coolest on record while June through August 1994 was the warmest. The warm spring of 1992 coin- cided with an earlier than expected flight com- mencement. D. cunfustis were likely flying be- fore ti-ap placement, possibly as eaily as late May at lower elevations. In contrast, the snowy win- ter and spring of 1993 followed by a record cool summer resulted in a delayed beetle flight. In 1992 D. confimis were first captured at 2840 ni on 10 June compared with 6 July in 1993. In each year, regardless of the overall weather regime, I obsei-ved that flight did not initiate until the local snowpack was mostK melted and that the early summer peak occurred after all snow patches were gone. 354 Great Basin Naturalist [Volume 56 Sex Ratio The early summer surge typically had a higher portion of male beetles (Table 2). Males were especially doininant during the first 5—10 d of emergence, comprising nearly all of those sampled. The sex ratio tlien became more evenly mixed for the remainder of the early summer, including during peak activity. The late-sum- mer surge was dominated by females in each year, the ratio being more stable throughout the period. Secondaiy Scolytids and Predators Other scolytids captured include GiiatJiotri- chiis sulcatus LeConte (ambrosia beetle), Pityok- teines minutus Swaine, Xylechiniis montanus Blackman, Hylastes suhopacus Blackman, Scoly- Table 1. Percentage of total seasonal beetle captures per plot occuning in the 1st surge. Year 1992 199.3 1994 Plot 1st surge (elev. [m]) (%) 2310/2.350 41 2.560 66 2600/2660 68 2840 72 2310/2350 10 2560 12 2600/2660 16 2840 76 2310/2350 30 2560 50 2600/2660 60 2840 67 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 ^^1 mm prec Lotemp Hi temp ■o30 3 25 Q. M 0) e^o c <0 2 5 0) ^ 0 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 2840m Fig. 6. 1992 daily weather conditions at Brighton-Silver Lake (2700 ni) with Dnjocoetc.s confiisiis flight acti\it\- at nearby plots. Arrow indicates the trough bcKveeu surges not associated with cool or wet weather. Note the lag effect, resulting from 2- to 5-d observation intcnals, which can gi\c the illusion of (hght acti\it\ during acKerse weather • f 1 1 ,,. . 1. 1 1 , 1996] DRYOCOETES CONEUSUti FL1C;HT Pl^l^lODlCITi' 355 30 5 25 - 820 15 r 10 - i 5 (U 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 2660m 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 ^^H mm prec Lo temp Hi temp 0) 525 s 20 - ^ 15 i:io c Ji 5 0 °- 0 • I f ■ II 1 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 2840m Fig. 7. 1993 daih' weather conditions at Brigliton-Silver Lake (2700 m) with Dnjocoetes confusiis fliglit acti\it\- at nearby plots. Arrow indicates the trough between surges not associated with cool or wet weather Note the lag effect, resulting from 2- to 5-d obsei-vation intenals, which can give the illusion of flight activit)' during adverse weather. tus ventralis LeConte (fir engraver beetle), Ips spp., Cryphahis ritficollis Hopkins, Hyhirgops porosus (LeConte), Scohjtiis opacus Blaekman, Dryocoetes affaher (Mannerheim), and Dry- ocoetes sechelti Swaine. Checkered beetles, Enoclerus spp., were the most common and important predaceous insect trapped. Other predators captured include snakeflies (Raphidi- idae) and rhizophagids. Gnathotrichus siilcatus, Xylechinus montanus, Hylastcs siihopaciis , Dryo- coetes ajfaber, Enoclerus spp., and rhizo- phagids were caught in sufficient numbers to suggest that they possibly cue on exo-brevi- comin. Clerids, or checkered beetles {Enoclerus spp.), were captured before western balsam bark beetle flight commencement in each year. Captures generally peaked in mid- to late June, then tapered off through August. Few clerids were captured later than the end of August in any year though D. confusus continued to be active. At the 4 plot locations common to each study year, a total of 242 clerids were caught in 1992, 357 in 1993, and 307 in 1994. Discussion Flight Periodicity Western balsam bark beetles were caught throughout the summer months for all study years. Beede flight typically started in June and continued well into September with a few captures as late as early October. Once flight was initiated, only cool or wet weather could 356 Great Basin Naturalist [Volume 56 30 B25 - 8 20-- M ^ 15 ifiof c S^ 5 I. I ■ I I I I I ll I I I I , I ■ I 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 2600m 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 ^^1 mm prec Lo temp Hi temp ■o30 3 25 S20 (/> (U ^15 f 10 c 0) ^ 5 0) °- 0 20 6/1 15 7/1 15 8/1 15 9/1 15 10/1 15 31 2840m Fig. 8. 1994 daily weather conditions at Brighton-Silver Lake (2700 m) with Dryococtes confusiis flight activity' at nearby plots. Arrow indicates the trough between surges not associated with cool or wet weather. Note the lag effect, resulting from 2- to 5-d obsei-vation intei-vals, which can give the illusion of flight activity during adverse weather. completely curtail it. Some warm, diy periods, however, had dramatically reduced activity rel- ative to the peaks. At each plot 2 distinct peaks of flight activity were seen every year. Some plots appeared to have a 3rd peak in Septem- ber of 1992 and 1994, but this can be attrib- uted to a reduction in trap checking fi-equency at those times. Generally, the 1st peak was sharp and occurred within 2-3 wk of initial emergence. A 2nd peak was obsei'ved 4-7 wk after the 1st. This is similar to Stock's (1991) findings in British Columbia where there "were two major flight periods per year, the first commencing in mid- to late June, and the second in mid- to late August." Stock found this flight period to correspond well with the life cycle described by Mathers (1931). Mathers (1931) studied the western balsam bark beetle life cycle at Stanley, B.C., using caged subalpine fir bolts. Young adults were found to emerge and attack fresh trees in June and July. Eggs were laid through August before parents commenced feeding in tunnels before overwintering. Galleries were advanced the following June and July with continued egg lay- ing. Parents then reemerged in July to attack a fresh host with a 3rd set of brood tunnels exca- vated, eggs being laid through August. Math- ers concluded this to be the end of the life cycle. Eggs from the 1st brood ovenvintered as lai^vae before pupating the following sunnner. The 2nd winter was passed as >()ung adults that emerged to attack fresh trees the following June and July. Eggs from the 2nd and 3rd broods overwintered as lanae that pupated the 1996] Dryocoetes confusus Flight Periodicity 357 Table 2. Percentage of females amonc; sampled beetles in early and late-smnmer surges. 1993 1994 Plot 1st sur Se 2nd sui Ke 1st sm- Se 2nd singe (elev. [m]) {%) {7c) {%) (%) 2310/2350 47 74 38 51 2560 46 71 29 (K) 2600/2660 34 75 28 (i6 2840 52 64 44 60 following year and emerged as adults the next season. Gixen this life cycle, one would expect to find all life stages represented in any given year. Stock (1991) suggests that the August peak obsei"\'ed in his study corresponds to the July reemergence described b>' Mathers, the timing difference due to warmer conditions at Mathers site. Using Mathers life cycle, the 2 flight peaks must be of different generations, the June peak being of newly emerged young adults and the August peak being of reemerged 2nd-yr adults. This gives a 3-yr life cycle when the reemergence year is included. Beetles fi-om eggs laid in 1991, for example, presumably might not complete their life cycle until 1994. Bright (1963) suggests that parent beetles may die in their 1st or 2nd brood tunnels be- fore reemerging to attack another host. This would account for the late-summer surge often having less activit)' tlian tlie early summer surge. Bright also believes D. confusus to have a 1-yr life cycle in the western United States, the life cycle proposed by Mathers (1931) being a phe- nomenon restricted to the insect's northern range. No data or references, however, are cited for this assertion. The flight period data presented here pro- duced noticeably different results each year. This is almost certainly associated with the record-setting weather regimes seen each study year. The double peak pattern was not as evi- dent during the cool, wet summer of 1993. Only die highest elevation plot exliibited a large, early summer peak, July in this case. The lower elevation plots had a noticeably reduced early summer peak. The greatest activity at these plots occurred in late August. The record cool summer weather likely caused the delayed emergence seen in 1993, but this does not ex- plain the diminishment of the early summer peak. Even with delayed development, I ex- pected considerable activity once flight was initiated. Cold weather during emergence was explored as a possible cause for the reduction in the early summer peak. Night temperatures at 2700 m dropped to -4°C on 24 June 1993, which could have killed some new adults and lurthei- delayed the early summer flight (Bar- bara Bentz^ personal communication). Perhaps development of some new adults was delayed such that their emergence overlapped with that of reemerged adults in the late summer At the highest plot, Mafliers' (1931) hypoth- esized life cycle corresponds well with the data. The 2nd and 3rd broods described b\' Vlathers would assure an early summer surge each year even though 2 yr is required for sexual matu- rity. In other words, there is no "off year " such as with the 2-yr life cycle of the spmce beetle, Dcndroctomis nifipennis (Kirb\'). Some results from this study did not match Mathers' (1931) life cycle as well as did Stock's (1991). For example, assuming that the same local population was sampled each year, adults in the early summer flight of 1993 should be represented again in the late-summer flight of 1994. Allowing for some degree of mortality, I expected the late-summer surge to have fewer beetles than the early summer surge of the previous year for a given location. This study produced 2 examples where the late- summer surge contained several times more beetles tiian the early summer surge of the previous year. If, in fact, these plots did sample the same popu- lations each year, then Mathers' (1931) life cycle may not be accurate for northern Utah. Though funnel trap captures are not appro- priate for adequately describing a life cycle, there are several possible explanations for the unexpected results at the 3 lower plots. The record-setting weather regimes in each year would have certainly affected beetle phenol- ogy. Perhaps some critical thresholds were not achieved in 1993, resulting in retarded de\'el- opment or mortality (Barbara Bentz personal communication). This may have affected young adults more so than reemerging adults. Con- versely, record warm weather in 1992 and 1994 could have advanced development. Perhaps the 2nd and 3rd broods are not important in the overall life cycle. Amman and Bartos (1991) found reemerged mountain pine beetle, Dendroctonus ponderosae Hopkins, females to ^Entomologist, USDA Forest Sen'ice, Intennountain Research Station, Logan, UT. 358 C'UU- AT IvVSlN \ ATI KAl 1ST [NoUinie 56 produce siiiiiiticaulK tow or oti^^priui:; than now temiUes with nialos. Forhaps lower elevation beetles tend to ha\ e a l-\ r lite e\ ele while 2 \ r is required for beetles at hiiiher oloxations. Pivlxiblx some ciMubination ot tliese tiictors plus soiue not explored, sueh as disease and preda- tion, e».>ntributed to the results. Sex Ratio Stock (^IQQl'i found that the late-sunnner sui"ge is comprised larvioK of females. I sinu: AuiTust 1 to distiuiTuish 1st and 2nd tlight. Stock found femiiles to comprise 48*^. 29^, and 50^ of the 1st tlidit dining 3 consecutive \-eai-s. The 2ud flight had S0'>. 4S^f . imd GO-^ leniiiles. I found simiUu" tivnds in noitheni I tali. This predominance b\ feniiiles during 2nd tlight is t\pic;il tor other scol\ tids tliat exhibit ivemergence vAnderbrandt et al. 19S5. Flamm et al. 19S71 This suggests diat tiie late-summer sui-ge, in fact, comprises reemerged adults. The dominimce of males dining the initial da\s of eaiK' summer emergence suggests tliat die\ iU-e likely i-esponsible for host selection and mate atti-action. \\Vatlier Influences Stock 1991^ found tlie majorit>" of western balsam biuk beetle tlight to occur when tlie d;ul> maximum temperatme was greater dimi 15 *C. The Siune tivnd is seen hei^e. Periixls ot cool weather especi;illy when coupled widi precipitation. essentiiilK stopped beetle cap- tures. Given w^mn, di"\ da\ s. D. confiisus was found to be active as late as earh October. all>eit in greatb reducx^d numbei-s. An\ ctmbol strategx will need to cxinsider diis extended flight period. SinprisingK, tlie cooL wet summer of 1993 feiiled to ha\ e ;m\ obxions etlect on die 1994 beetle population other th;m to dela\ emer- gence. I h\podiesized diat diis w eadier pattern \\X)idd ha\ e inciviused lieede morhilit). resulting in fewer trap captures in 1994. Elsewhere in the region, mountain pine beede Ian ae were observ ed widi ivhuxied de\ elopment. possibK leaving o\ ervvintering life stages more xailnera- ble to cold weather mortUit\. Assuming a 2-\t life cycle for western IxUsiun Ixu-k beede. per- haps the more cxild-susceptible life stages would not have emerged until June 1995. It is also possible that winter temperatures in 1993-94 did not reach lethal le\ els. \\ hilo i\u"K (.Mnergeneo was associated with a record warm spring in 1992, die earl\ sum- mer peak was no earlier than after the more topical spring of 1994. Timing of the earh" sum- mer peak for each plot occurred on essentiidh" the same date in 1992 and 1994. Timing of die late-summer peak, how o\ er. was about 2 wk earlier dining the reeord warm summer of 1994 than the more h pieal summer ot 1992. CONCI.ISION Once flight begins, actixih t\pie;ill\ builds to a sliaip peak within 2 wk. This genenilK occurs fixim mid-June dirough emK" JuK. Acti\it> dien subsides before building to a 2nd peak 4-7 wk later. usuiilK in August. Significant acti\it\ CiUi cxmtinue into earl\ September with some bee- des fixing as late iis eaiK October Cultural or semiochemicid iniuiagement of western balsam bark beetle will need to con- sider the double peak flight pattern of piu^ent beetles luid die tact that adult beedes can be found in some quimtit> diroughout die wiuiiier months. Remoxid of infested host materiid. for example, should be done in the tdl w hen flight is complete. Anti-aggiegation pheromones will need to be foniiulated to efliuse up to 4 nion or, possibK. applied twice per seiison. Further resernxdi is needed to confinii or re- vise die life c>des described b\ Madiei^s il931) and suggested by Bright il963i since control sti-ategies for a 1-\t life cycle can be different from a 2- or 3-\t life c\ ele. Considering die 1st versus 2nd surge difterences. diis should be done for a nmge of eleN"ations. The role of re- emerged adults in brood production must be detennined for a more complete understimd- ing of die oxenill life c> ele. The conditions lead- ing to iUi outbre;dc also need to be quantified such diat cidtural guidelines can be estabUshed. ACKNO\\T_EDGM ENTS This pivject w:is initiated by Ste\e Munson • Forest Hciddi Protection. Ogden. VT). Criti- Cid re\iew of die manuscript w iis proxided b\' BiU-bara Bentz. -\rt Stock British Colimibia Forest Service. Nelson). Da\\-n Hansen, and John .\nhold both Forest Hcidth Protection. Ogden. UT . Jim Viuidygriff Inteniiountiiin Research Station. Logiui. UT produced die graphs. .\lan Dxinersld (Forest HciUth Protec- tion. Ogden, UT) proxided significant assis- tance with the annual installation of plots. 1996] Dryocoetes confusus Flight PpzRioDicrn 359 Thanks to the main who helped collect and count the trap captures o\er 3 seasons: Alan Dymersld, Dan Johnson (USDA Forest Ser- vice, Ogdcn, UT), Jill .\nsted (USDA Forest Ser\ice, Heber City, UT), Heather Schmidt, Keith Pflei^er. Rhonda Bishop, Josh Vierej^ge, Steve Deakins, and Chris Peterson (all Utah Department of Agriculture, Salt Lake City, UT). Irene \bit (Intennountain Research Sta- tion, Ogden, UT) conducted the literature search. Special thanks to Art Stock and Bar- bara Bentz for encouragement and insights. My thanks also to Jesse Logan and L\nn Ras- mussen O^oth Intennountain Research Station, Logan, UT) for allowing me to complete this manuscript. This project was partially funded by Intennountain Research Station, Mountain Pine Beetle Project, Logan, UT. LlTER\TLRE CiTED A,MMAN, G. D.. AND D. L. Bartos. 1991. Mountain pine beede offspring characteristics associated with females producing first and second broods, male presence, and egg galler\ length. Environmental Entomology 20: 1562-1.567.' Anderbrwdt, O., E Schlyter, asd G. Birgersson. 1985. Intraspecific competition affecting parents and off- spring in the bark beetle /;«• tyjinfiraphus. (Jikos 45: 89-98. Borden, J. H., A. .VI. Pierce, M. D. Pierce, Jr., L. J. Cik ag, A. J. STfx.K, AND A. C. Oemi.schi^(;er. 1987. .Semio- chemicals produced by western balsam bark beetle, Dryocoetes amfwms Swaine (Coleoptera: Sco^■tidae^ Journal of Chemical Ecology 1.3: 82.3-836. Bright, D. R., Jr. 1963. Bark beetles of the genus Dry- ocoetes (Coleoptera: Scolytidae) in .North Amerit-a. Annals of the Entomological Society of .America 56: 10.3-115. DoiDGE, D. 1981. Western balsam bark beetle in British Columbia. Canadian Forest .Service, Pacific Forest Re- search Centre, Victoria, B.C. Forest Pest Ix-aflet 64. Fl\mm, R. O., S. R Cook, T. L. Wagner, P E. Plleey, and R. .\. CoLLSON. 1987. Reemergence and emergence of Ips acuhwi and Ips calligraphm (Coleoptera: ScoKti- daej. Environmental Entomologv 16: 869-876. Johnson, D. W. 198.5. Forest pest management training man- ual. USD.\ Forest Service, Rock\ .Mountain Region. Lakevvood, CO. 138 pp. -Mathers, W. G. 1931. The biology of Canadian baH< beetles. The seasonal history of Dryocoetes confusus Sw. Canadian Entomologist 68: 247-248. Stock, A. J. 1991. The western balsam bark beetle, Dry- ocoetes confusus Swaine: impact and semiochemical- based management. Unpublished doctoral disserta- tion, Simon Eraser University, Bumaby, B.C. 13.3 pp. Received 11 December 1995 Accepted 12 August 1996 Great Basin Naturalist 56(4), © 1996, pp. 360-368 DISTRIBUTION OF A THERMAL ENDEMIC MINNOW, THE DESERT DACE {EREMICHTHYS ACROS), AND OBSERVATIONS OF IMPACTS OF WATER DIVERSION ON ITS POPULATION Gaiy L. Vinyard^ Abstract. — Population status sui-veys were performed from 1987 to 1996 for desert dace {Eremichthys acros), a cyprinid endemic to several small thermal springs in Soldier Meadow, Hinnboldt County, Nevada, where the species occupies 7 spring areas in a single valley. Because spring distributions are patchy and all areas are not linked by surface flow, each area comprises a more-or-less isolated population, although iiTigation practices or high runoff may occasionally link several of them. Although limited to thermal springpools and outflows, desert dace were foimd in temperatmes rang- ing fi'om 37°C near spring sovux'es to 13°C in downstream areas. Between May 1988 and October 1989, most of the dis- charge fiom a major spring outflow was diverted from its natmal channel into an inigation ditch. Trap catches in the orig- inal channel were reduced after the diversion, and fish densities were lower in die ditch dian in the channel. Reduced fish numbers still persist (1996), even though the affected site has been relatively undisturbed since 1989. To improve desert dace habitat and increase populations, inigation diversion should be discontinued and water returned to the original chan- nel. Continued protection and increased habitat presewation for desert dace are recommended because of their limited distiibution, apparently restincted habitat requirements, and the potential for environmental disniption in the area. Key words: conservation, endemisni, Eremichthys acros. Great Basin, habitat, irrigation, Nevada, springs. The desert dace {Eremichthys acros, Cyprini- dae) is a federally listed threatened species restricted to outflows of thermal springs in Soldier Meadow, Nevada (Hubbs and Miller 1948, La Rivers 1962). Desert dace have a unique homy sheath on both jaws and a greatly elongated intestine; they occupy exceptionally high temperatures relative to other cyprinids (Hubbs and Miller 1948, Nyquist 1963). The distinctive moiphology of desert dace suggests a long period of isolation extending beyond the most recent pluvial period. The desert dace differs significantly from other Great Basin minnows, and its original description and assignment to a monoty[Dic genus (Hubbs and Miller 1948) have been confirmed (La Rivers 1962, Nyquist 1963, Cavender and Coburn 1992). Its taxonomic relationship within the western cyprinid fauna remains unclear, and recent workers have judged desert dace most similar to either relict dace, Relictiis solitarius (Cavender and Coburn 1992), or to tui chub, Gila {Siphateles) hicohr (Lugaski 1980). Few investigations of desert dace have been undertaken since Nyquist (1963), and little is known of their behavior, ecology, or physiology'. I present results of studies of desert dace dis- tribution and document adverse consequences to the species fiom water diversion in die valley. Study Area Soldier Meadow and Mud Meadow to the south occupy a gently sloping valley in south- western Humboldt County, Nevada. Elevation ranges fi-om about 1400 m MSL at the noitli end of Soldier Meadow to about 1317 m MSL south of Mud Meadow Resei-voir (Fig. 1). Although the lower elevations are near maximum shore- line level of pluvial Lake Lahontan, the area was prol^ably not inundated during the Pleis- tocene (Benson 1978). Soldier Meadow is also home to an endemic plant, the basalt cinque- foil {Potcntilla hasaltica), and at least 4 un- described species of h\'drobiid springsnails (R. Hershler, Smithsonian Institution, personal communication). Ranching operations in Soldier Meadow be- gan in the late 1860s but ha\ e ne\'er been par- ticularly successful. After an active period in the 1960s, ranching was largely dormant in the val- ley through most of the 1970s and early 1980s. In 1994, \\ ith the help of the Nature Conser- vancy, much of the property' was transferred to I Dt'partinciit of Biolo©' /315. University of .Nevada, Reno, NV S95.57. 360 1996] DisTKiiJU rioN oi' Desert Dace 361 4586 4584 4582 - 4580 - TO SUMMIT LAKE / I 4578 4576 4574 GENERAL LOCATION SOLDIER MEADOWS Enlargement of Area 4 TO HIGH ROCK t- LAKE 1 mi TO GERLACH 1 km 310 315 320 325 Eig. L Map of Soldier Meadow and Mud Meadow showing sampling locations. Solid lines indicate water conrses; dotted lines are roads. Spring areas discussed in the te.xt are enclosed in ovals. Not all springs in the valley are presented, and in some areas numerons small springs are represented by single symbols. Springs and streams outside the ovals lack desert dace {Eremichthtjs acros), but not all sites within ovals contain fish. Values on axes indicate coordinates in relation to 1000 meter Universal Transverse Mercator Grid, digitized from USGS maps. Insets show location of Soldier Meadow in relation to Nevada state boundaiy (left) and an enlargement of the spring and ditch system associated with area 4 (right). The area 4 inset depicts the relationships between the spring sources, upstream zone, ditched zone, and old chaimel as discussed in the te.xt. The ditch was most recently dredged between May 1988 and October 1989. The approximate upstream limit of desert dace was near the start of die traps. the U.S. Bureau of Land Management. Most desert dace habitat now occurs on pubHc land, and the remainder of the privately held habitat is protected by a consei"vation easement (Nature Consei^vancy 1994). Springs and outflow streams in the valley are all subject to grazing by cattle, feral horses and burros, and pronghoni. There is also frequent recreational use of the area by hunters, campers, bathers, and others. Fish Distribution Desert dace distribution is strongly corre- lated with spring discharge. Fish are absent from small springs or seeps with little surface water and fiom larger pools of very hot water lacking organized discharge. All springs with perennial surface flow are occupied by desert dace, the most abundant fish in the valley. Although desert dace are most often found in habitats lacking other fish species, they are not confined to them and have been observed coexisting with tui chub, speckled dace {R. osciilus), and Tcilioe suckers {Catostomus tahoensis; La Rivers 1962, Nyquist 1963, Sigler and Sigler 1987). Desert dace habitat occurs in 7 distinct areas located within an 8-km (5-mi) radius (Fig. 1). Because spring distributions are patchy and all areas are not linked by surface flow, each area comprises a more-or-less isolated population, although irrigation practices or exceptionally high runoff may occasionally allow fish passage among several of them. Most areas described below contain many springs varying greatly in size, and it is often difficult to identif\' the exact number and location of discharge sources because of the dense vegetative cover. 362 Great Basin Naturalist [Volume 56 Area 1 Area 1 (Fig. 1) includes the type locality (Hubbs and Miller 1948) for desert dace (Table 1, site 19), a spring that issues from the base of a small cliff It is modified by a valved struc- ture diverting flow into a pipe for household use at Soldier Meadow Ranch. Undiverted dis- charge flows about 100 m east, where it enters a small impoundment or a series of irrigation ditches. Desert dace coexist with tui chub in the spring and impoundment. Desert dace habitat here is limited by the impoundment and by the shifting diversion into ditches. Area 2 In area 2 (Fig. 1) several small springs con- taining desert dace (Table 1, sites 9, 23-27) are located around the base of a small hill and flow generally southward or southwestward into a large meadow. A large springpool containing desert dace (Table 1, site 10) is the largest nat- urally occurring body of open water in the val- ley (approximately 15 m in diameter and 1.5 m deep). It has a minor surface outflow south into a marshy meadow. Nonnally, most of the springs in this area are unconnected, although proxim- it)' and common drainage suggest connections are likely during high lamoff. Natural drainage from this site is toward area 7, and the outflows of several of the more southerly springs are ditched southward through the meadow. The larger springs in this area receive frequent recreational use by bathers and campers. Area 3 Area 3 (Fig. 1) includes several small springs flowing south approximately 1 km north of Mud Meadow Resei-voir (Table 1, site 20). Although at least 3 springs in this area contain desert dace, and some populations are quite dense, all springs in the area have been heavily affected by livestock grazing and irrigation diversion. The 2 largest springs have long been diverted into irrigation ditches at a point within 20 m of the sources. Grazing by cattle, buiTOS, and feral horses has altered the vegetation and disrupted soils near the springs. This distin-bance has wid- ened tlie outflow channel, reduced water depth, and generally eliminated riparian vegetation. Area 4 Several large springs issuing from the side of a small hill are the source for area 4 (Fig. 1). Widi more than 2600 m of stream, this is the largest contiguous potential habitat for desert dace. Water issues from the highest springs at approximately 50°C (Table 1, sites 1-6, 11, 21) and cools gradually while flowing downstream with occasional augmentation by both warm and cool inflows. These springs produce an aggregate discharge of approximate!) 60 1/s. The upper reaches are fishless, presumabh' be- cause of high water temperatures. Headwater springs in area 4 probably receive the highest level of recreational use in the valley, primarily from bathers and campers. Several small cobble dams erected across the outflow stream in this area are mostly upstream from the dace habitat and pose little impediment to fish passage. Deposition of soaps and other water pollutants from bathers may constitute a risk of unknown magnitude. This area has also been heavily grazed by cattle and feral horses and buiTos. Several inigation diveisions ha\'e long existed in area 4; however, they were poorly maintained and little used for at least a decade. Between May 1988 and October 1989, the rancher in the valley dredged out an old ditch, moving water away from the original channel in area 4 (Fig. 1, inset). Before the dredging most of the discharge continued southeastward in the natural channel and spread into a large, wet meadow. Speckled dace historically coexisted with desert dace in the lower sections of this system near the wet meadow. By October 1989 most discharge in area 4 was diverted east to the irrigation ditch, and no water reached the meadow by the original channel. CurrentK, approximateh' 80% of the total combined discharge from the source springs in area 4 is diverted. Speckled dace are now absent from the system, and the amount of desert dace habitat was significantK reduced by this diversion. The loss of discharge into the lower portions of the wet meado\\' on the down- stream end of area 4 had additional adverse impacts on desert dace in area 5 (see below). Area 5 Area 5 includes a group of \ en- hot springs that enter a series of old irrigation ditches approximately 200 m from the source and then flow southeasterly toward Mud Meadow Reser- voir (Fig. 1). In 1988 a series of cool springs fed 1)\ discharge from the wet meadow below area 4 entered the outflow stream at area 5 approxi- mateh' 50 m downstream from the primary spring sources. Mixing of these waters produced 1996] DlS THIBUTION OF DESERT DaCE 363 Taiu.K 1. Characteristics of spring Iiahitats in Soldier Meadow. Data were collected at various times between 1987 and 1995. Column desiunations are as follows: AHKA = distribution area (numbers indicate areas indicated on Figure 1; sites without numbers lack tlesert dace); SITK = site identifier Irom field notes, refers to specific locations within areas; EAST and NORTH indicate site locations in relation to lOOO meter Universal Transverse Mercator Crid, di^jitized from uses maps — Mud Meadow, 1972; Soldier Meadow, 1972; FISH = fish six'cies present, E = Eremichlhtis arms, H = Rhiniclithij.s osnilu.s, C, = Gila bicohn C = Catostonuis taliooisis, a = fish absent; °C = water temperature; DO = dis- soKed o.xygen concentration (mg O2/I); CX)ND = electrical conductivity (;UMho/cm). AREA SITE EAST NORTH FISH "C DO COND 1 19 31SSS5 4584932 E,C 28-29 5.6-6.2 190 1 19A 319038 4584990 E,C 22-26 8.0-8.6 190 2 9 317625 4583030 E 34-38 3.4-4.5 432 2 10 316826 4582540 E 21-34 4.8 410 2 23 317547 4582650 E 21 3 20 318607 4578372 E,R 25 4.5-6.1 270 4 21 315827 4580116 E 17 4 1 314016 4580934 a 36-40 3.6-5.8 370-420 4 2 314068 4580654 a 36-38 5.3-5.8 380-420 4 3 314446 4579892 E 30-35 6.2-6.4 310 4 4 314869 4579580 E 27-28 6.4-7.0 4 5 314005 4580202 a 37-42 1.8-3.7 400-430 4 6 315961 4578982 E,R 19-25 5.8-7.6 280-285 4 11 315331 4579392 E 20-29 6.3-8.0 305 5 12 316691 4578454 E 13-50 1.9-6.3 480 5 13 316888 4578154 E,R 23 6.1 650 5 14 317089 4577836 E,R 23 5.6 650 6 16 314599 4581304 E 23 6.5 295 7 8 316929 4580550 E 34-57 0.9-6.9 470-750 — 1 316682 4580660 R,C 6 8.3 325 — 15 315512 4576460 a 25 5.6 280 — 17 314198 4581512 a 35 3.6 370 — 18 316939 4580684 a 50 3.8 280 — 22 317115 4581286 R 13 — 24 317536 4582660 a 35 — 25 317519 4582660 a 37 — 26 317498 4582640 a 40 — 27 317516 4582620 a 40 — 2.S 316383 4580864 a 10 8.9 — 29 316381 4580940 a 9 5.5 steep temperature gradients, as waters of ±45 °C and <20°C gradually mixed over about 100 m. In May 1988 water in the main spring outflow was 43 °C at the point where water at 13 °C entered fiom the meadow to the north. Desert dace were observed actively feeding in the 13 °C water mass at the point where cold \\ ater entered the primary channel. Fish also darted into the turbulent zone between the hot and cold water masses in pursuit of small drift- ing food. All observations since October 1989, after the diversion of water in area 4, have found the amount of water reaching the lower sections of the wet meadow above area 5 to be greatly reduced, and inflow from the cool springs flowing into area 5 has ceased. Conse- quently, several hundred meters of the approx- imately 1 100 m of (ditched) desert dace habitat have been lost from this area. This area is grazed and the outflow ditched, but it receives relatively little recreational or other use. Ai-ea 6 The desert dace population in area 6 (Table 1, site 16, Fig. 1) occupies a single spring stocked by U.S. Bureau of Land Management personnel in the early 1980s at a time of con- cern over the ftiture of desert dace. This spring has the smallest discharge of any containing desert dace (estimated at less than 5 liters per minute). A gauging box and wooden notch weir produce a small impoimdment (appro.ximatcK 3 m X 3 m) about 30 m from the source, \\'liicli contains most of the fish population at this site. Recently this impoundment was nearly lost by deterioration of the weir. This system also in- cludes a somewhat larger eailJien impoundment (±10 m diameter) approximately 50 m from the 364 Great Basin Naturalist [Volume 56 source, after which flow disappears into a small meadow. Desert dace are mostly restricted to the area above the larger impoundment and are most abundant near the gauging box and notch weir. This site is grazed but too small for recreational use. Ai-ea7 Area 7 includes several hundred meters of suitable habitat fed by several springs on the eastern side of the valley (Fig. 1). Because sev- eral springs issue at temperatures exceeding 50° C, the extent of suitable habitat varies with ambient air temperature. This area may be connected with outflows from area 2 during periods of high iimoff. Most of the outflow in this area has been modified to some extent for inigation, and it is subject to grazing and some recreational use. Mud Meadow Resei-voir Mud Meadow Reservoir contains large- mouth bass {Micropterus salmoides), goldfish [Carrasius auratus), and perhaps other species planted by unknown individuals (Ono et al. 1983). It is unclear whether it is a barrier to desert dace passage, but it is unlikeK' to provide any permanently suitable habitat. Although no nonnative fishes have been observed in any sites containing desert dace, the potential threat posed by nonnative fishes spreading into dace habitat is certainly enhanced by their establishment in the resei'voir. Materials and Methods All desert dace habitats recorded by Nyquist (1963) and most other springs in the valley were visited in 1987 to update distribution information (Vinyard 1988). The dredging of the irrigation diversion in area 4 significantly reduced the amount and quality of desert dace habitat in that area. Investigations thus were concentrated in the affected locality (area 4) beginning in 1989. Fishes in area 4 were sampled widi standard unbaited minnow traps on 5 occasions (14 May 1988, 20-22 October 1989, 3 November 1993, 20-21 October 1995, and 27 April 1996). Dur- ing May 1988 sampling included the entire original stream channel (>2.6 km) from spring sources to disappearance of the stream in a wet meadow. The section upstream from the diver- sion and the irrigation ditch were sampled in October 1989, November 1993, October 1995, and April 1996. The remnant natural channel downstream from the diversion was also sam- pled in October 1995 and April 1996 (Fig. 1, inset). Fish traps were 40 cm long by 20 cm diame- ter, constructed of 0.64-cm-mesh galvanized hardware cloth, with 2.5-cm entrance holes at the peak of each concave conical end section. Traps were placed at 20-m intei-vals along spring outflows and fished 2 h during dayliglit. Altliougli the traps were sometimes not completely sub- merged, they were always placed witli tire open- ings under water. Captin^ed fishes were identi- fied, enumerated, and released near the point of capture. Standard length (SL, in mm) and weight (gm) were recorded on some sample dates. Water temperature, dissolved oxygen, and electrical conductivity were measured using portable meters at regular intei-vals along the trap set. In October 1995 stream velocity' was mea- sured along cross-section transects with a Marsh-McBurney model 201D flow meter at 6 or 9 sites each in the upstream, ditch, and old channel zones of area 4. Measurements were at 5-cm vertical and either 10- or 20-cm hori- zontal increments, depending on channel widtli. In June 1995 electrofishing was performed using a 3-pass depletion methodology (Van- Deventer and Platts 1989). Stream sections 10 m long were isolated with blocking nets and depletion rates on successive passes used to estimate population size. Three groups of 6 sections were fished: in area 3, and in the old channel and ditched zones of area 4. Results Although resident in thermal springs and outflows, desert dace have wide temperature tolerances and were obsened in waters rang- ing from 13 °C to 38 °C. Occupied waters had conductivity ranging from 190 to 650 fuS and dissoKed oxygen concentrations generalK near saturation, ranging from 4.5 mg/1 to 8.0 mg/1 (Table 1). Although there was considerable overlap between species, desert dace were ioimd at higher temperatures and lower dis- solved oxygen concentrations than speckled dace, which were not observed at tempera- tures above 26 °C or in dissolved oxygen con- centrations below 5.2 mg/1. 1996] Distribution of Desert Dace 365 In area 4 in 1988 desert daee eateh in min- now traps was signifieantly negative!) corre- lated with temperature (linear regression; F = 19.98, /?2 = 0.122, n = 131, P < 0.001), al- though relati\ely little variance in catch was explained by temperatiue, and no such cone- lation was obsened in later years. Desert dace catch was also not significantly correlated with speckled dace catch. Catch rates generally reflect fish abundance but may also be affected by activity. Temperature, dissolved oxygen con- centration, and combinations of these and other factors may affect activity levels. In 1988, when desert dace and speckled dace were sympatric in the natural channel above the meadow in area 4, mean catch per trap-horn- was significantly greater {t = 2.83, P = 0.009) for desert dace (4.56 fish per trap- hour) than for speckled dace (1.04 fish per trap-hour). Where both species occurred, desert dace was more abundant, and maximum densities of both species were observed at temperatures of about 23 °C. Cross-section grid transect measurements of water velocit)' were used to assess mean val- ues at each transect and to compute discharge. In 1995 in area 4, velocities were significantly higher in the ditch (6 transects, z = 24.9 cm/s, n = 121) than in the upstream zone (9 tran- sects, z = 17.3 cm/s, n = 318; t test, df = 159, T = -3.663, P < 0.001), or in the old channel (6 transects, z = 16.7 cm/s, n = 111; t test, df = 188, T = 3.733, P < 0.001). However, veloc- ity measurements did not differ significantly between the upstream and old channel seg- ments {t test, df = 226, T = 0.504, P = 0.616). Volumetric computations indicated that dis- charge in the ditch was 46.5 1/s while discharge in the channel was 10.8 1/s, or 18.8% of the total. In October 1995 desert dace trapped in the upstream zone were significantly smaller (mean SL = 35.9 mm, n = 172) than those from either the ditch (mean SL = 38.7 mm, n = 82; t test, df = 105, T = -3.33, P = 0.001) or the old channel (mean SL = 38.4 mm, n = 135; t test, df = 284, T = -5.50, P < 0.001). Standard length of the fish in the ditch and in the old channel did not differ significantly {t test, df = 113, T = 0.328, P = 0.744). Electrofishing transects in area 4 June 1995 yielded mean values of 21.8 fish per 10 m {n = 6, ,s' = 27.2) for the old channel and 12.5 fish per 10 m (n = 6, s = 8.5) for the ditch zones. Densit>' estimates of 110 fish per 10-m section {it = 6, s = 51.53) were obtained in area 3 at that time. These values did not differ signifi- cantly among the 2 zones of area 4; however, densities in area 3 were significantly higher than in either zone of area 4 {t tests, P < 0.01 in both eases). The fish eleetrofished from the ditch were significantly larger (avg. SL = 37.2 mm, n = 67, s = 7.9) than those from the old channel (avg. SL = 32.7 mm, n = 122, .s- = 11.3;/test, f = 3.2, P = 0.002). Discussion Trap data from area 4 offer an opportimit>' to assess impacts of habitat alteration by com- paring catch rates in the zone upstream from the diversion, in the original channel down- stream, and in the ditched zone (Fig. 2). Because traps were set on the same spacing intei-vals in each sampling period, it is possible to examine cumulative catch per trap hour to compare fish densities. These values are com- puted by summing catch per trap hour for each trap along the trap set from the upper to the lower end (Fig. 2). The total cumulative catch in area 4 was much larger in May 1988 than at any other time (Fig. 2). In contrast, the lowest cumulative catch observed was in October 1989, the first sample after the ditch dredging. Although direct comparisons of these 2 samples may be confounded by seasonal differences, the con- trast between the largest catch obsen^ed (in May 1988) and the smallest catch observed (in October 1989) coincides with the dredging. Comparison of the autumn sample in October 1995 with the spring sample in April 1996 sug- gests that populations are larger in die spring than fall, but that this difference is probably insufficient to explain the difference between the 1988 and 1989 data. By November 1993 the cumulative catch had recovered somewhat from 1989 (Fig. 2). A not- able difference in 1993 relative to both earlier observations was the sharp increase in catch apparent at about 1000 m, slightly upstream from the diversion. However, with the excep- tion of the accumulation of fish at this point, the general slope of the cumulative catch curve was little changed from October 1989. In October 1995 and April 1996 shaip increases in catch immediately above the diversion were still apparent. The general slope of the catch 366 Great Basin Naturalist [Volume 56 cr D o I cc LU Q. I O o LU > 250 200 150 100 50 I 200 O 150 100 50 CHANNEL -APR 96 CHANNEL -OCT 95 DITCH - APR 96 UPPER ZONE APR 96 500 1000 1500 DISTANCE (m) 2000 2500 Fig. 2. Cumulative catch of desert dace in unbaited minnow traps in area 4 from Ma\' 1988 through April 1996: A, Data collected in May 1988, November 1993, and October 1989. B, Results from 1995 and 1996. In all cases single, un- bailed minnow traps were fished at 20-m intervals for 2-h sets during daylight. Cumulative catch per hour is computed by summing catch per hour for each trap beginning at the upstream end of the trap set. Distance on the ordinate is the distance downstream from the first trap. In May 1988 cumulative catch reached 448 at 2620 m downstream (off scale). Dredging of the irrigation diversion occurred between May 1988 and October 1989. Samples in October 1988 and 1993 included only the zone upstream above the diversion and the ditched zone downstream. In October 1995 and .\pril 1996, the old channel remaining below the diversion was also sampled. cui-ve for the ditclied segment changed little between 1993 and 1996 (Fig. 2). The slope from the renniant old channel in 1995 and 1996 (Fig. 2) was much steeper than that observed in the ditch, in spite of the roughly 4 times greater discharge measured in the ditch in 1995. Catch rates in the ditch or old channel have never reached levels observed in the channel in 1988, and even summing the cumu- lative catch from both the ditch and the old channel still does not \ ield results comparable with the catch rates ol^sened in 1988. Different responses b\- desert dace to the various habitats in area 4 are also evident in the percentage of traps with non-zero catch (Fig. 3). This measure can be used as an indi- rect indication of the amount of habitat occu- pied. In 1988 the channel zone had the highest percentage of traps catching fish, nearly 90%. In the 4 samples from the ditched segment. lfJ96] DisTKiiui ION OF Desp:rt Dace 367 12 ■?10 r .6 . 4 UJ 3 &2 H UPPER ZONE ■ DITCH □ CHANNEL J3_Jl 100 80 - MAY88 OCT89 NOV93 DATE OCT95 APR96 Fig. 3. Axerage catch per trap hour for all traps, area 4, from Ma>' 1988 tliroiigh April 1996. Bars indicate different stream segments. Dredging of the inigation ditch occuiTed between May 1988 and October 1989. The upper zone sampled was abo\e the point of the irrigation diversion. The ditched segment existed prior to dredging in 1988 but had become overgrown nearly to the point of obstructing any flow. After dredging, it received most of the discharge from the system. The channel received nearly all the flow from the springs in area 4 during the 1988 sample, but only 20% or less of the total flow in subsequent samples. never more than 75% of the traps eaught fish. The old channel zone continues (in both 1995 and 1996) to have a larger percentage of traps catching fish than either of the other 2 zones (Fig. 3). Catch per trap hour may also be used to estimate relative fish populations. The 3 high- est average catch rates were observed in the channel below the present point of diversion, and the highest value of any was obsei'ved in 1988, prior to the dredging (Fig. 4). Although catch values were still highest in 1995 and 1996 in the old channel zone, they have not returned to levels observed in 1988. Catches fi-om the upstream and ditched zones have var- ied much less during the sample period. These data indicate that the natural channel was the most productive site for desert dace prior to the ditch dredging, and that it still pro- vides habitat which is superior to the ditch, c\ en 8 yr after the dredging and with <20% of the total discharge. The obsei-ved aggregation of fish above the diversion (evident in the cumulative catch data since 1993) bears examination. If habitat in the ditch is unsuitable, desert dace may avoid the ditched zone and accumulate in the upstream zone. Because no aggregation of fish in this zone was obsei-ved in 1988 or 1989, it seems likely to be the result of a behavioral response to the changed conditions. i 60 40 20 ■ UPPER ZONE ■ DITCH □ CHANNEL MAY88 OCT89 NOV93 OCT95 APR96 DATE Fig. 4. IVrcent traps with non-zero catch, area 4, from May 1988 through April 1996, Bars indicate dilTerent stream segments. Dredging of the irrigation ditch occuned between May 1988 and October 1989. The channel received nearly all the flow from the springs in area 4 during the 1988 sample, but only 20% or less of the total flow in subse- quent samples. The higher mean water velocities observed in the ditch (24.9 cm/s) relative to the up- stream zone (17.3 cm/s) suggest that desert dace may avoid higher velocity flows. It is likely that smaller fish avoid higher velocity flows in the ditched section and accumulate in the region immediately upstream from it. Although this explanation does not account for the relatively low abundance of fish in the old channel, other factors, including reproductive success and differences in habitat qualit)-', may be important. The absence of the aggregation upstream from the ditch in 1988 may reflect a general population reduction resulting from the ditching. Distribution of desert dace reflects poteu- tially interacting factors including tempera- ture, dissolved oxygen concentration, and cur- rent velocity. Distribution may also be affected by interactions with other species, particularK speckled dace. Studies are necessaiy to iden- tify and assess the mechanisms of such interac- tions. An additional area of interest would be to assess the relative degree of isolation of the 7 population units identified in this study to determine whether there are any behavioral, ecological, or genetic differences among these groups. Conclusions In recent years desert dace have been sub- jected to relatively minor disturbance com- pared with many other native fish species in the Great Basin. Most of the sites historically 368 Great Basin Naturalist [Volume 56 occupied by desert dace retain suitable habitat, though it has generally been modified to some extent. Disturbance levels may have been higher at times in the 1960s (Nyquist 1963). Desert dace populations in Soldier Meadow have been relatively stable since 1989, but most desert dace habitats have been substantially altered over the years, and we cannot now directly assess the magnitude of any persistent popula- tion reductions that may have occurred before that time. Desert dace populations persist in the modified thermal waters that now charac- terize Soldier Meadow; however, the data from area 4 demonstrate that adverse effects of habi- tat modifications linger for many years. It is appropriate to consider management options for this unique fish. Their presei^vation requires continued physical protection of springs and flowing waters in Soldier Meadow from excessive grazing and prohibition of the introduction of nonnative organisms. Restoring the water to natural stream channels should also be incoiporated into any management plan because of the potential positive impacts from improving habitat quality. Consequences of such water management for the endemic spring- snails should either be neutral or positive. They are generally abundant in the springs where they occur, and losses fi^om restoration of flows should be offset by increased habitat stability. Desert dace seem relatively secure under current conditions. However, the small num- ber of occupied sites, restricted geographical distribution, and generally unknown but possi- bly specialized habitat requirements of the fish argue strongly for continued monitoring and increased investigation into factors regulating populations. Growing demands on aquatic resources of the Great Basin make it clear that increased awareness of and protection for this unique fish will be necessary for their long- term sui'vival. Acknowledgments I am grateful to A. Berglund, J. Dunham, T Kennedy, K. Obermeyer, R. McNatt, D. Sada, M. Sevon, and C. Stock-well for assistance with various aspects of this work, and to Dave Liver- more of the Nature Consen'ancy for his efforts on behalf of desert dace. Members of various Desert Ecosystems, Aquatic Ecology, and Ichthyology classes from the University of Nevada, Reno, assisted in fieldwork. Literature Cited Benson, L. V 1978. Fluctuation in the \e\e\ of pluvial Lake Lahontan during the last 40,000 years. Quaternaiy Research 9: 300-.318. Cavender, M. M., and T. M. Coburn. 1992. Interrelation- ships of North American cyprinid fishes. Pages 328-373 in R. L. Mayden, editor, S\ stematics, histor- ical ecology and North American freshwater fishes. Stanford Universit)' Press, Stanford, CA. HuBBS, C. L., AND R. R. Miller. 1948. Two new, relict genera of cyprinid fishes froiii Nevada. University of Michigan Museum of Zoology Occasional Papers 507:1-30. La Rivers, I. 1962. Fishes and fisheries of Nevada. Nevada State Fish and Game Commission. 782 pp. Lugaski, T. p. 1980. Comparative chemota.xonomy of selected Great Basin native cyprinid fishes. Unpub- lished doctoral dissertation. UniversitA' of Nevada, Reno. 2.54 pp. Nature Conservancy. 1994. Soldier Meadow conserva- tion project. Great Basin Field Office, Salt Lake City, UT Nyquist, D. 1963. The ecolog>' of Eremichthys acws, an endemic thermal species of cyprinid fish fi'om north- western Nevada. Unpublished master s thesis. Uni- versity of Nevada. 247 pp. Ono, R. D., J. D. VViLLUMS, and a. Wagner. 1983. Vanish- ing fishes of North America. Stonewall Press, Wash- ington, DC. 257 pp. SiGLER, W, AND J. SiGLER. 1987. Fishes of the Great Basin. University of Nevada Press, Reno. 443 pp. VanDeventer, J. S., AND W. S. Pu\TTS. 1989. Micro- computer software system for generating population statistics from electrofishing data — user's guide for MICROFISH 3.0. U.S. Forest Service General Tech- nical Report INT-254. Vinyard, G. L. 1988. Population status sune\ of tlie Soldier Meadows desert dace {Ercinichfliys acros). Submit- ted to the U.S. Fish and Wildlife Sei-vice. Contract 14320-87-00178. Received 19 Fchruanj 1996 Accepted 26 June 1996 Great Basin Naturalist 56(4). © 1996. pp. 369-374 HELMINTHS OF THE SOUTHWESTERN TOAD, BUFO MICROSCAPHUS, WOODHOUSE'S TOAD, BUFO WOODHOUSII (BUFONIDAE), AND THEIR HYBRIDS FROM CENTRAL ARIZONA Stephen R. CoUlluM-ui. Cliarles H. Bursey-, Keith 13. Malmo.s'^ Brian K. Snllivan"\ and Ha\ Cheain' AB-STRACT. — The gastrointestinal tracts, lungs, and iirinar\' hladders from 77 Btifo inicroscaijIiiLs, 61 Bufo woodhousii, and (S of their hyhrids were e.xamined for helminths. One species of trematode {Glyptlwlinins quiela), 1 species of ces- tode {Distoichotnetra bttfimis), and 5 species of nematodes {Aph'cianu incerta, A. itzocancims, Rhabdias americanus, Physaloptcra sp., and Physocephalus sp.) were found. The greatest prevalence (41%) and mean intensity (231.7) were recorded ihv Aplcctiina iiiccrta in Bufo icoodhoii.sii. !t appears h\hrids Iiarhoi- fewei- i:iarasites than either parent species. Key icords: Iwhniutlis. Bulo microscaphus, Huto woodhousii, hyhrids. Arizniia. The southwestern toad {Bnjo uiicroscapluts Cope, 1866) is presently recognized as 3 allo- patric subspecies: B. )n. californicus Camp, 1915, which occurs in coastal southern California and northwest Baja California; B. m. microsca- phus Cope, 1866, found in southern Nevada and Utah, Arizona, and New Mexico; and B. m. mexicanus Brocchi, 1879, which occurs in the Sierra Madre Occidental of central Me.xico south to Durango (Price and Sullivan 1988). Woodhouse's toad {Bufo woodhousii Cirard, 1854) is recognized as 4 subspecies: B. w. wood- housii Cirard, 1854 occurs in eastern Montana and North Dakota, south through the plains states to central Texas and west of the Rocky Mountains from Idaho south to Colorado and Arizona with isolated populations in west Texas, southeastern California, and along the Oregon- Washington border; Bufo w. austrolis Shannon and Lowe, 1955 is found from central Colorado through New Mexico and Arizona to Sonora, Mexico, and iilong the Rio Crande drainage into southwest Texas and adjacent Mexico; Bufo w. velatus Bragg and Sanders, 1951 is restricted to northeast Texas; and B. w. fowled Hinckley, 1882 is widespread throughout much of the eastern United States south to the Culf Coast and west to eastern Te.xas (Beliler and King 1979). The toads examined during this study, B. in. microscaphus and B. w. austraUs, are known to li> bridize in Arizona (Sullivan 1986, Sullivan and Lamb 1988). Altliough diere are reports of helminths from B. microscaphus (Pany and Cnmdmann 1965) and B. woodhousii (Trowbridge and Heflev 1933, Brandt 1936, Walton 1938, Reiber et af. 1940, Kuntz 1941, Kuntz and Self 1944, Rankin 1945, Fantliam and Porter 1948, Frandsen and Crund- mann 1960, Pany and Crundmann 1965, Camp- bell 1968, Brooks 1976, Jilek and Wolff 1978, Baker 1985, Hardin and Janovy 1988, McAUister et al. 1989), populations of tliese toads from Ari- zona have not been examined. Concern over declining amphibian populations (Heyer et al. 1994) has increased interest in die possible nega- tive effects of parasites on toads. The puqDOse of tliis paper is to report on helminths of tliese toads and dieir hybrids from Aiizona. This investigation of parasitism in these toads addresses a hypothesis of hybrid zone theory and species boundaries. The hypothesis that populations of hybrid individuals with reduced fitness act as barriers to gene flow between 2 species separated by a hybrid zone (Biuton 1979, 1980) could have several mechanisms. One mechanism, increased parasitism of hybrids, is evaluated in this study. Two previous studies of parasitism in vertebrates are split. Hybrid mice {Mus muscuhis X Mus domesticus), specifically backcrossed hybrids, had greater numbers of cestode and nematode parasites than either parental species (Sage et al. 1986). Prevalence of monogenean parasites for hybrid minnows {Barhus barhus X Barhus meridiomiUs) was ^Department of Biology, Whittier College, Whittier, CA 90608. ^Department of Biology; Pennsylvania State University, Shenango Valley Campus, 147 Shenango Avenue, Sharon, PA 16146. ■^Department of Life Sciences, Bo.x .37100, Arizona State I'niversity West, Phoeni.x, .AZ 8.5069. 369 370 Great Basin Naturalist [Volume 56 positi^'el^• associated with the percentage of B. meridionoUs genes (Le Brun et al. 1992). If we find that hybrid toads have greater parasitism than each toad species, then parasitism may be a mechanism tliat reduces hybrid fitness and contributes to the hairier between these 2 toad species. Materials and Methods One hundred forty-six toads were collected in Aiizona during 1991-1995; snout-vent length (SVL) was measured to the nearest mm after a minimum of 6 mon in 70% ethanol storage. Toads were identified using a hybrid index (HI) and advertisement call structure, if available. Following Blair (1955), Sullivan (1986), and Sul- livan and Lamb (1988), we evaluated die degree of expression of 4 characters to generate the HI score for each toad: dark ventral pigmenta- tion, cranial crest, dorsal stripe, and pale colora- tion across the eyelids. A numerical score (0, 1, 2, 3) was assigned for each of the following 4 character states: present, weakly present, very weakly present, or absent. A score of 3 was assigned for the presence of dark xentral pig- mentation, cranial crests, a dorsal stripe, and absence of a pale bar across the eyelids. This yields scores near 12 (4 X 3) for B. woodhoiisii and 0 (4 X 0) for B. microscaphus. Numerous other studies of hybridization between toad species have used a moiphological hybrid index such as this (Volpe 1959, Meacham 1962, Hen- rich 1968, Zweifel 1968). All toads from sites of sympatiy with scores of 4 through 8 were con- sidered hybrids, as were all toads with interme- diate advertisement calls. Inteniiediate calls are typical of hybrid toads between these species (Sullivan 1995), and calls have long been used to delimit hybrid toads of other species pairs (Blair 1956, Zweifel 1968, Green 1982). Sev- enty-seven Bufo microscaphus (mean SVL = 61.4 mm ± 8.7 s, range 34-86 nmi, 67 males, 10 females); 61 Bufo wooclhousii (mean SVL = 74.5 mm ± 8.8 s, range 49-91 mm, 53 males, 8 females), and 8 hybrids (mean SVL = 60.5 mm ± 8.4 s, range 45-72 mm, 7 males, 1 female) were examined. Kruskal-Wallis test statistic (45.92, 2 df, F < 0.001) indicates significant difference in SVLs for the samples examined. After examination all specimens were deposited in the heipetology collection of Arizona State University (ASU), Tempe. Collection localities and ASU accession numbers are given in Appendix 1. Toads were anesthetized by immersion in 1 g/1 solution of tricaine methane sulfonate (MS- 222, Sigma, St. Louis, MO). Heart, liver, thigh muscle, and kidne\' were remo\'ed and fiozen for future genetic analyses. Toads were then fixed in neutral-buffered 10% formalin and moved to ethanol for storage following proce- dure outlined by Simmons (1987). The body cavity was opened by a longitudinal incision from vent to throat, and the gastrointestinal tract was removed by cutting across the esoph- agus and rectum. The esophagus, stomach, small intestine, large intestine, lungs, bladder, and coelom were examined under a dissecting microscope. No helminths were found in the esophagus or urinar>' bladder. All helminths were removed and identified using a glycerol wet mount. Specimens were placed in vials of alcohol and deposited in the U.S. National Par- asite Collection, Beltsville, Maryland 20705: (accession numbers, Appendix 1). Results and Discussion Prevalence, site, and mean intensity for each parasite are given in Table 1. Temiinology is in accordance with Margolis et al. (1982). One species of trematode {Glijpthehnins quieta [Stafford, 1900]), 1 species of cestode {Distoi- chometra bufonis Dickey, 1921), and 5 species of nematodes {Aplcctana inccrta Caballero, 1949, Aplectana itzocanensis Bravo Hollis, 1943, Rlxahdicis americanus Baker, 1978, Fhijsaloptera sp. [lai-vae only], and PhysocepJuiIus sp. [lanae only]) were found. It would appear fiom Table 1 that both species and their hybrids are sus- ceptible to infection by the same parasites. The greatest prevalence (41%) and mean intensit)' (231.7) in our study were recorded kr Aplectana incerta in Bufo woodhousii. Thirty-four of 77 (44%) Bufo microscaphus (30/67, 45% males; 4/10, 40% females), 51 of 61 (84%) B. wood- housii (45/53, 85% males; 6/8, 75% females), and 4 of 8 (50%) hybrids (3/7 males, 1/1 female) were infected. Males and females of both Bufo microscaphus {x^ = 1.17, 1 df P > 0.05) and B. woodhousii {x^ = 2.79, 1 df P > 0.05) did not differ significantly in helminth prevalence. There were too few female hybrid toads for chi-square analysis. There was statistical differ- ence in abundance of nematodes betw een B. microscaphus and B. woodhousii (x^ = 23.72, 1996] Arizona Toad Helminths 371 Table 1. Prevalence, mean intensih' (range), and location of lielmintlis from Btifo microscaphus, B. tvoodhousii. and tlitir Iivbrids from Arizona. Biifo microscaphus Bitfo woodhousii llv l)rids (.V = 77) (\ = 6l) i^' = 8) Fr e\ale lice Intensih Location Pie\iilencc Intensih' L ocation Prevalence Intensih IjOcation r.u, isite species C^^) (range) (%) (range) (Vo (i range) 1 i;i \l VIODA ('.liiptlichiiiii.s quk'ta 1 1.0 h 2 2.0 1. 13 1.0 h (JsiODA Distoichoinctni hufoni.'i 14 2.9 (1-6) h 38 2.0 (1-8) b 13 1.0 b Nl MVrODA Aplectana incerta 1 156.0 b,c 41 231.7(23-564) b,c — Aplectana itzocanensis 19 75.0 (1-.373) b,c 26 43.2 (1-204) b,c 25 1.0 c Phtjsaloptera sp. (lar\;i) 16 5.5(1-31) a 5 6.0(2-11) a 13 1.0 a Pinjsocephalm sp. (lar\ae) 1 104.0 d — — RJialxlias americanus 5 2.0(1-31 V 38 21.7(1-11]) e — large intestine.s, d = cysts on stomach wall, e = lungs. 1 df, P < O.OOlj. When the intermediate prexalence (50%) of the small hxhrid sample (N = (S) was included in the chi-square calcula- tion, statistical significant difference remained iX- = 23.97, 2 df, P < 0.001). To test for difference in infection rate, we used a Kruskal-Wallis rank-order statistic be- cause of the great variation in mean intensit}' of parasites harbored by B. microscaphus, B. woodhousii, and their hybrids (116.3, 19.4, 1.3, respecti\'ely) and the relative!)' small sample of In biids {N = 8). This test revealed that hybrid indixiduals had fewer parasites than do indi- viduals of either species. E.xamination of more h\ brids could strengthen this result. Subsequent work to determine the importance of age, Uenetic factors, nutrition, and ecology would also help to establish the significance of hybrid ancestiy on parasite levels. Infected frogs appeared healthy; i.e., none \\ ere emaciated and there were no gross abnor- malities. Thus, tlie presence of helminths did not appear to adversely affect the populations oi B. microscaphus, B. woodhousii, or their hy- brids. In a stud\' on Couch's spadefoot (Scaphio- pus couchii) fi-om Arizona, Tinsley (1990) found no correlation between presence of the trema- tode Pseudodiplorchis americanus and mating success, although the presence of P. americanus reduced fat reserves during hibernation. Bufo microscaphus is a new host record for Distoichometra bufonis, Aplectana incerta, A. itzocanensis, Physocephahis sp., and Wwibdias americanus. Bufo woodhousii is a new host record for Aplectana incerta and Physaloptera sp. Bufo w. woodhousii has been reported by Baker (19'th 1994). Because these helminths are not species specific and occur in a variet\- of amphibians, the distribution of inteiTnediate hosts may play an important role in determining the distribu- tion of those parasites with indirect life cycles. The conditions responsible for determining distribution of the parasites with direct life cvcles have vet to be defined. 372 Great Basin Naturalist [Volume 56 Table 2. Helminth conimuniW of desert toads from Arizona. Helminth Host Reference Trematoda Glypthelmins quieta Pseudodiplorchis americamts Cestoda Distoichoinctra bitfonis Nematotaenia dispar Nematoda Aplectana incei'ta Aplectana itzocanensis Oswaldocriizia pipiens Physcdopfera sp. (larva) Physocephaliis sp. (lai^va) Wiahdias amcricaniis Biijo microscaphiis B. woodhoiisii Scaphiopu.s couchii Bufo cognatus B. microscaphiis B. punctatiis B. retifonnis B. woodhoiisii Scaphiopits couchii Bufo alvarius Bufo uiicroscaphus B. retifonnis B. woodhousii Scapliiopus couchii Bufo alvarius B. cognatus B. microscaphiis B. punctatiis B. rctijonnis B. woodhousii Bufo alvarius B. cognatus B. punctatiis B. retifonnis Scaphiopus couchii Bujo alvarius B. cognatus B. microscaphiis B. retifonnis B. woodhousii Bujo alvarius B. microscaphus B. retifonnis Bufo alvarius B. cognatus B. microscaphus B. retifonnis B. woodhousii This study This study Tinsley 1990 Goldberg and Bursey 1991a This study Goldberg and Burse\' 19911) Goldberg et al. 1996 This study Goldberg and Bursey 1991a Goldberg and Bursey 1991a This study Goldberg et al This study Goldberg and Goldberg and Goldberg and This study Goldberg and Goldberg et al This study Goldberg and Goldberg and Goldberg and Goldberg et al Goldberg and Goldberg and Goldberg and This study Goldberg et al This study Goldberg and This study Goldberg et al Goldberg and Goldberg and This study Goldberg et al This stud\ . 1996 Bursey Bursey Bursey Bursey . 1996 Burse\' Bursey Bursev . 1996 Burse\ Bursey Bursey . 1996 1991a 1991a 1991a 1991b 1991a 1991a 1991b 1991a 1991a 1991a Bursev 1991a . 1996 Burse\ Burse\' . 1996 1991a 1991a Acknowledgments Field assistance was provided by Mike Demlong, Erik Gergus, and Matt Conley. The Department of Life Sciences at Arizona State University West provided a vehicle for trans- portation to collection sites and funds for some costs of this project. We thank Steven M. Nonis and Michael E. Douglas for access to the Ari- zona State University Vertebrate Collection and the constructive comments of 2 anonymous reviewers. Rachael Schuessler and Elizabeth Stikkers assisted with collection of parasites. Literature Cited Anderson, R. C. 1992. Nematode parasites of vertebrates. Their development and transmission. CAB Interna- tional, Wiillingford, O.xon, U.K. 578 pp. Baker, M. R. 1985. Redescription of Aplectana itzocanen- sis and A. inceiia (Nematoda: Cosmocercidae) fi^om amphibians. Transactions of the American Micro- scopical Society 104: 272-277. Barton, N. H. 1979. The dynamics of Inbrid zones. HereditA 43: 341-359. . 1980. The h>'brid sink effect. Heredit>- 44: 277-278. Behler, J. L., AND E VV. KiN(.. 1979. The Audubon Societ>- field guide to North American reptiles and amphib- ians. Alfred A. Knopf New York. 743 pp. 1996] Arizona Toad Hki.mimiis 373 lii.AiR, A. P. 1955. Distribution, variation, and hybridization in a relict toad (Biifo iiiicroscaphiis) in sontliwi'stern U tall. American Museum Novitates 1722: 1-38. BiAiR, W. F. 1956. The mating calls of'lnbrid toads. Texas Journal of Science 8: 350- 355. Hii\\[n, B. B. 1936. Parasites of certain North C^arolina Salientia. Ecolosjical Monographs 6: 491-532. BiiooKS, D. R. 1976. Parasites ol' amphibians of the C.reat Plains. Part 2. Plat\ helminths of amphibians in Nebraska. Bulletin of the University of Nebraska State Museum 10: 65-92. (IwiPBELL, R. A. 1968. A comparative study of the para- sites of certain Salientia from Pocahontas State Park, Virginia. Virginia Journal of Sciencel9: 13-20. I'^NTHAM, H. B., .AND A. PoRTKR. 1948. The parasitic fauna of vertebrates in certain Canadian fresh waters, with some remarks on their ecolog>', structure and impor- tance. Prt)ceedings of the Zoological Society of Lon- don 117: 609-649. f'RANDSEN, J. C, AND A. W. Grundmann. 1960. The para- sites of some amphibians of Utah. Journal of Para- sitology 46: 678. Goldberg, S. R., and C. R. Bursev. 1991a. Helminths of tliree toads, Biifo (ili(iriii.s, Btifo cognatus (Bufonidae), and Scaphiopus coiichii (Pelobatidae), fi'om southern Arizona. Journal of the Helminthological Society of Washington 58: 142-146. . 1991b. Helminths of the red-spotted toad, Biifo piiiwtafiis (Anura: Bufonidae), from southern Aiizona. Joimial of the Helminthological Society' of Washing- ton 58: 267-269. (ioLDBERG, S. R., C. R. Bursey, B. K. Sullivan, and Q. a. Truong. 1996. Helminths of the Sonoran green toad, Bufo retifonnis (Bufonidae), from southern Arizona. Journal of the Helminthological Societ\' of Washing- ton 63: 120-122. (iKEEN, D. M. 1982. Mating call characteristics of hybrid toads (Bufo cnnericanus X B. fnwieri) at Long Point, Ontario. Canadian Journal of Zoolog\ 60: 329.3-3297. Hardin, E. L., and J. Janovt, Jr. 1988. Population dynam- ics of Disfoicho)netra hufonis (Cestoda: Nematotaeni- idae) in Bujo woodJwimi. Journal of Parasitology 74: 360- 365. H ENRICH, T. W 1968. Moiphological evidence of secondaiy intergradation between Bufo hemiophnjs Cope and Bufo americanus Holbrook in eastern South Dakota. Herpetologica 24: 1-13. Heyer, W R., M. a. Donnelly, R. W McDiarmid, L. C. Hayek, and M. S. Foster., editors. 1994. Measur- ing and monitoring biological diversity. Standard methods for amphibians. Smithsonian Institution Press, Washington DC. 364 pp. Jilek, R., and R. Wolfe 1978. Occurrence of Spinitectus gracilis Ward and Magath 1916 (Nematoda: Spiiin-oi- dea) in the toad (Bufo woodhousii fowleri) in Illinois. Journal of Parasitology 64: 619. KiNTZ, R. W. 1941. The metazoan parasites of some Okla- homa Anura. Proceedings of the Oklahoma Academy ofScience 21: 33-34. KUNTZ, R. E., .\ND J. T. Sele 1944. An ecological study of the metazoan parasites of the Salientia of Comanche County, Oklahoma. Proceedings of the Oklahoma Academy ofScience 24: 35-38. Le Brun, N., F Renaud, P Berrebi, and A. Lambert 1992. Hybrid zones and host-parasite relationships: effect on die evolution of parasitic specificits'. Evolu- tion 46: 56-61. Margolis, L., G. W. Esch, J. C. Holmes, A. M. Kuris, and G. a. Schad. 1982. The use of ecological terms in parasitology' (report of an ad hoc committee of the American Society of Parasitologists). Journal of I^ara- sitolog\' 68: 131-133. McAllister, C. T, S. J. Uiton, and D. B. Cow. 1989. A comparati\'e study of endoparasites in three species of sympatric Bufo (Anura: Bufonidae), from Texas. Proceedings of the Helminthological Society of Wash- ington 56: 162-167. Meacham, W R. 1962. Factors affecting secondaiy inter- gradation between two allopatric populations in the Bufo woodhou-sei complex. American .Midland Natu- ralist 67: 282-304. PARin; J. E., and a. W CiRUNDMANN. 1965. Species compo- sition and distribution of the parasites of some com- mon amphibians of Iron and Washington counties, Utah. Proceedings of the Utah Academy of Science, Arts and Letters 42: 271-279. Price, A. H., and B. K. Sullivan. 1988. Bufo inicrosca- phus Cope, southwestern toad. Catalogue of Ameri- can Ainphibians and Reptiles 415: 1-3. Prudhoe, S., AND R. A. BR.4Y. 1982. Platyhelminth para- sites of the Amphibia. British Museum (Natural His- toiy), Oxford University Press, London. 217 pp 4- 4 microfiche. Rankin, J. S., Jr. 1945. An ecological study of the helminth parasites of amphibians and reptiles of western Mass- achusetts and vicinity. Journal of Parasitolog)' 31: 142-150. Reiber, R. J., E. E. Byrd, and M. V. Parker. 1940. Certain new and alread\' known nematodes from Amphibia and Reptilia. Lloydia 3: 125-144. Sage, R. D., D. Heyneman, K. Lim, and A. C. Wilson. 1986. Wormy mice in a hybrid zone. Nature 324: 60-63. Simmons, J. E. 1987. Heipetologieal collecting and collec- tions management. Socieb,' for the Stud\' of Amphib- ians and Reptiles, HerjDetological Circular 16. 70 pp. Smyth, J. D. 1994. Introduction to animal parasitolo,g\'. 3rd edition. Cambridge University Press, New York. 549 pp. Sullivan, B. K. 1986. Hybridization bet\\een tlie toads Bufo rnicroscaphus and Bufo woodhousei in Arizona: mor- phological variation. Journal of Heq^etolog)' 20: 1 1-21. , 1995. Temporal stabilib.' in hybridization between Bufo rnicroscaphus and Bufi woodhousii (Anura: Bufonidae): behavior and moiphology. Journal of Evolutionary Biolog)' 8: 233-247. Sullivan, B. K., and T. Lamb. 1988. Hybridization between the toads Bufo microscaphus and Bufo woodhousii in Arizona: variation in release calls and alloz\nies. Hei-petologica 44: 325-333. TiNSLEY, R. C. 1990. The influence of parasite infection on mating success in spadefoot toads, Scaphiopus couchii. American Zoologist 30: 313-324. Trowbridge, A. H., and H. M. Hefley. 1933. Preliminaiy studies on the parasite fauna of Oklahoma anurans. Proceedings of the Oklahoma Academy of Science 14: 16-19. Volpe, E. P 1959. Experimental and natural Inliridization between Bufo terrestris and Bufo fowh'ri. American Midland Naturalist 61: 295-312. Walton, A. C. 1938. The Nematoda as parasites of Amphibia. IV. Transactions of the American Micro- scopical Societ)' 57: 38-53. 374 Great Basin Naturalist [Volume 56 || ZvVElFEL, R. G. 1968. Effects of temperature, bocK size, and h\'bridization on mating calls of toads, Bufo a. ainericaniis and Bufo woodhoiisii fowleri. Copeia 1968: 269-285. Received 27 March 1996 Accepted 29 July 1996 Appendix 1 Localities and museum (ASU) numbers for specimens examined: Bufo microscaphus. Maricopa County (xV = 6) (34°00'N, 112°45'W, elev 603 m) ASU 30360-61, 30369-72; Yavapai County (iV = 61); 7 fi-om (34°24'N, 112°13'W, elev. 1323 m) ASU 30328-31, 30347-49; 6 from (34°06'N, 112°09'W, elev 603 m) (ASU 29166-67, 29170-71, 30351, 30375); 4 from (34°04'N, 112°09'W, elev 488 m) (ASU 30377, 30379-81); 34 from (34°05'N. 112°07"W, elev 616 m) ASU 28845-50; 28852-57, 29172-83, 303.34-40; 30386-88; 10 fi-om (34°24'N, 112°08'W, elev. 1140 m) ASU 30487-96; Coconino County (N = 10) (34°24'N, 112°08'W, elev 2094 m) ASU 30477-86. Bufo woodhousii: Maricopa County (N = 53); 14 from (33°38'N, 112°28'W, elev 410 m) ASU 28821-27, 28829- 31, .30356-59; 19 from (.33°56'N, 112°08'W, elev 628 m) ASU 28818-19, 28828, 28835, 30.362-64, 30366-68, 29151- 59; 2 fi-om (33°36'N, 112°15'W, elev 365 m) ASU 28834, 28836; 7 from (33°39'N, 112°14'W, elev 389 m) ASU 30497-503; 11 from (33°36'N, 112°11'W, elev. 372 m) ASU 30504-14; Yavapai County {N = 8); 7 fiom (34°06'N, 112°09'W, elev 488 m) (ASU 29165, 29167-69; 30345, 30350, 30355, 30376); 1 from (.34°04'N, 112°09'W, elev 488 m) (ASU 30385). Hybrids; 'i'avapai County (N - 8); 7 from (34°06'N, ll'2°09'\V, elev 603 m) ASU 30346, 30352-54, 30373-74, 30382; 1 from (34°04'N, 112°09'W, elev 488 m) ASU 30378. Accession numbers for helminths in the U.S. National Parasite Collection (USNPC): Bufo microscaphus: Distoichometra bufonis (85910); Ghjp- thcbnins quieta (85921); Aplectana incerta (85911); Aplec- tana itzocanensis (85912); Physalopteridae (85915); Phij.so- cephalus sp. (85914); RJjabdias americamis (85913). Bufo woodhousii: Distoichometra bufonis (85916); Ghjpthehnins quieta (85921); Aph'ctana incerta (85917); Aplectana itzo- canensis (85918); Ph\ salopteridae (85920); Rliabdias ameri- camis (85919). Hybrids: Distoichometra bufonis (85922); Ghjpthehnins quieta (85921); Aplectana itzocanensis (85923); Physalopteridae (85924). C;reat Basin Naturalist 5fi(4), © 199fi, pp. 375-376 JUVENILE RAZORBACK SUCKER {XYRAUCHEN TEXANUS) IN A MANAGED WETLAND ADJACENT TO THE GREEN RIVER Tiiiiotlw Moclck'l Key words: razorhack sucker, floodplain, wctlaiKl. jiircnilc. The razorback sucker {Xyrauchen fexanus) is a kirge, endemic catostomid of the Cokirado Ri\'er drainage. It was once widespread and abundant throughout the basin (Minckley et ah 1991). Species abundance and distribution de- cUned following construction of mainstem dams and the introduction of many nonnative fishes (Behnke and Benson 1983, Carlson and Muth 19(S9). The razorback sucker was federally hsted as endangered in 1991 (USFWS 1991). ' The largest riverine population of razorback sucker is in the middle Green River (Lanigan and Tyus 1989). These fish spawn successfully (Tais and Karp 1990), but Lanigan and Tyus (1989) reported little or no recruitment. Razor- back sucker larvae in the Green River drift downstream from spawning sites (Robert Muth, Lanal Fish Laboratory', Colorado State Univer- sity, Fort Collins, CO), but few juvenile have been found and little is known of their habitat needs. Taba et al. (1965) captured 8 juveniles (90-115 mm total length [TL]) from Colorado Rixer back-water habitat in surveys from 1962 to 1964 between Moab and Dead Horse Point, Utah. More recendy, Gutermuth et al. (1994) collected 2 juveniles (37 mm and 39 mm) from a lower Green River backwater in 1991 and 2 others (59 mm and 29 mm) in a backwater on tlie Ouray National Wildlife Refuge in 1993 (Robert Muth, Larval Fish Laboratory, Colo- rado State University, personal communication). This note reports occurrence of juvenile and adult razorback suckers in a wetland adjacent to the Green River in Utah. Old Charley Wash is a 60-ha wetland on the Ouray National Wildlife Refuge in Uintah County, northwest Utah, adjacent to river kilo- meter (RK) 402 on the Green River. The wash is a historical type IV wetland (Cowardin et al. 1979) with smartweed {Polygonum sp.) and sago pond weed {Potarnogeton pectinatus) being the primary aquatic plants. The natural levees of the wetland have been reinforced with dikes to retain water through the siunmer and fall periods. Water in- and outflow is con- trolled at flows <481 ni'^/s. Water enters the inlet at river flows of approximately 240 m-^/s. Typical management is to fill in spring and then maintain water through the summer and autumn. The outlet structure at Old Charley Wash was modified in April 1995 to facilitate fish capture by creating a drainable, 12-m concrete- lined channel in which fish could be concen- trated and captured with seines. Spring flow of the Green River peaked at about 595 mVs in 1995 and inundated Old Charley Wash between 23 May and 1 July. Inundation was at flows >481 rn^/s. The wash was dry prior to inundation. Maximum depth of the wetland was >2 m. Fish in the wetland were isolated from the river; when runoff sub- sided, no additional water was added. Fishes were sampled by fyke and trammel nets, min- now and light traps, and seines. Collections were weekly from 23 May to 1 July and ever>' 2 wk from 2 July to 31 August. The wetland was drained from 25 September to 12 October, and fishes were collected from the outlet every other day during the first 2 wk and daily (except 9 October) during the 3rd week. Twenty-eight juvenile razorback sucker were collected when Old Charley Wash was drained in the fall of 1995 [x = 94 mm TL [range = 74-125 mm] and 9.5 g [range = 3-18 g]; voucher speci- mens, catalog number LFL 24874, Larval Fish Laboratory, Colorado State University). Eight (461-525 mm TL; 1034-1650 g) adults also were captured, 6 prior to and 2 during the draining process. A total of 10.1 metric tons of 'Colorado River Fish Project, U.S. Fish and VVildhfe Service, 266 West 100 North, Suite 2, Vernal, UT 84078. 375 376 Great Basin Naturalist [Volume 56 fish were collected during draining. The iol- lowing species were represented in order of contribution by weight: Cyprinus carpio, Pime- phales promelas, Lepoinis cijaneUus, Ictahinis pimctatus, Ameiurus melas, Cyprinella hitren- sis, Pomoxis nigromaculatus, Xijrauchen tex- anus, Esox liicius, Gila atraria, Catostotmis latipinnis, Catostomiis commersoni, Ptyclioche- iliis liicins (7 individuals ranging in TL be- tween 175 and 207 mm, and weight from 33 to 62 g), Gila rohiista, and Ciilaea inconstans. Tyus and Karp (1990) reported that razor- back sucker spawn on the ascending limb of the hydrograph, allowing drifting larvae to disperse during peak runoff and thus maximiz- ing access to wetland habitats. It is unknown whether the juveniles collected during drain- ing originated from riverine spawning sites or were produced in Old Charley Wash. How- ever, their occurrence in Old Charley Wash in 1995 supports speculation (Tyus and Kaip 1990, Modde et al. 1966) that floodplains may be important razorback sucker nursery areas. Support for this study was provided by the Recovery Implementation Program for the Endangered Fishes of the Upper Colorado River Basin. Thanks to T Hatch, C. Flann, N. Hoskin, D. Irving, B. Haines, R. Nicoles, K. Day, and K. Kaczmarek for assisting in fish collections. Literature Cited Behnke, R. J., and D. E. Benson. 19S3. Endangered and threatened fishes of the Upper Colorado River basin. Colorado State University Cooperative Extension Service, Bulletin 503A. Carlson, C. A., and R. T. Muth. 1989. The Colorado River; lifeline of the American Southwest. Pages 220-239 ((1 D. R Dodge, editor. Proceedings of the International Large River Symposium. Canadian Spe- cial Publication of Fisheries and Aquatic Sciences 106, Ottawa. CowARDiN, L. M., V. Carter, E C. Golet, and E. T. La Roe. 1979. Classification of wetlands and deepwater habitats of the United States. U.S. Department of Interior, U.S. Fish and Wildlife Service, FWS/OBS- 79/31. 131 pp. Gutermuth, F B., L. D. Lentsch, and K. R. Bestgen. 1994. Collection of age-0 razorback suckers {Xyrau- chen texamis) in the lower Green River, Ut;ih. South- western Naturalist 39: 389-391. Lanigan, S. H., and Tyus, H. M. 1989. Population size and status of the razorback sucker in tlie Green River basin, Utah and Colorado. North American Journal of Fisheries Management 9: 68-73. MiNCKLEY, W. L., R C. Marsh, J. E. Brooks, J. E. John- son, and B. L. Jensen. 1991. Management toward the recovery of the razorback sucker Pages 303-357 in W. L. Minckley and J. E. Deacon, editors. Battle against extinction: native fish management in the American West. University of Arizona Press, Tucson. MoDDE, T, K. P Burnham, and E. E Wigk. 1996. Popu- lation status of the endangered razorback sucker in the Middle Green River. Conservation Biologv 10: 110-119. Tara, S. S., J. R. Murphy, and H. H. Frost. 1965. Notes on die fishes of the Colorado River near Moab, Utiih. Proceedings of the Utah Academy of Sciences, Aj-ts, and Letters 42: 280-283. Tyus, H. M., and C. A. Karr 1990. Spawning and move- ments of razorback sucker, Xtjrauchen texamis, in the Green River basin of Colorado and Utah. Southwest- ern Naturalist 35: 427-433. U.S. Fish and Wildlife Service (USFWS). 1991. Endan- gered and threatened wildlife and plants: the razor- back sucker {Xtjrauchen fexaniis). Determined to be an endangered species. Federal Register 56(205): 54957-54967. Received 6 November 1995 Accepted 21 June 1996 Creat Basin Naturalist 56(4). © 1996, pj:.. 377-378 CONFIRMATION OF COSEXUALITY IN PACIFIC YEW {TAXUS BREVIFOLIA NUTT.) K. E. Hoggi, A. K. Mitchell'-, and M. R. Clayton' Key words: Pacific yew, Taxiis 1)il'\ iiolia, dioecious, cusexuality, British C'oliiinhia, pollen, seed. Unlike most evergreen conifers in our forests, which have both pollen and seed on a single tree. Pacific yew {Taxiis brevifolia Nutt.) is dioecious, the 2 sexes being segregated on dif- ferent trees (Rudolf 1974, Taylor and Taylor 1981, Bolsinger and Jaramillo 1990, Hils 19'93). In Jul\ 1993 branch samples of T. brevifolia were taken from an undisturbed stand of coastal Douglas-fir {Psciidotsuga menziesii) on southern Vancouver Island (48°26'N. lat.; 123°28'W. long.) near Victoria, British Colum- bia. One of the samples was obsei^ved to have both male and female reproductive structures (bud scales partially removed) on a single twig (Fig. 1). Occasionally, male and female structures can occur on the same tree (Taylor and Taylor 1981). In the instances reported (Owens and Simpson 1986, DiFazio 1995), female and male structures occurred together only on branches of predominantly male trees. We obsen'ed this phenomenon, termed cosexuality (Lloyd 1980), on a single yew tiee. On one branch, female and male reproductive structures were observed within a few mm of each other (Fig. 1) on an otherwise male tree. The structures were visu- ally identical to respective buds from other dioecious trees. In a study by DiFazio (1995), cosexuality was found in 17 of 58 male trees (29.3%). It is not known whether these female buds found on male trees produce viable seed. Reproductive buds of the Pacific yew can be visually differentiated throughout the year (Taylor and Taylor 1981) and are usually located on the underside of the shoot on noncurrent growth. Male buds are small (2-3 mm), round, and green, and they generally occur in clusters (Fig. 2). They consist of a number of distinct segments made up of pillowlike structures (microsporangia) in which the pollen mature. In spring microsporangia burst the bud scales (Fig. 3) and pollen is released. Female buds generally occur singly (Fig. 4) and are erect, oval (2-3 mm), and green. The female bud matures slowly through spring and summer with the ovule (Fig. 5) growing through the bud scales and revealing the micropyle (open- ing for pollen). Beginning in late July or early August, depending on location, a fleshy red aril (berry) around the hard-coated seed becomes visible. Acknowledgments The authors thank L. Kaupp (University' of Victoria) for field sampling and L. Manning (Pacific Forestiy Centre) for preparation of the photographic plate. Literature Cited Bolsinger, C. L., and K. E. Jaramillo. 1990. Taxtis brevi- folia Nutt.— Pacific yew. Pages 573-579 in R. M. Bums and B. H. Honkala, editors, Silvics of North America: 1. Conifers. Agriculture Handbook 654. United States Department of Agriculture, Forest Service, Washing- ton, DC. DiFazio, S. P. 1995. The reproductive ecology of Pacific yew {Taxtis brevifolia Nutt.) under a range of o\'erstory conditions in western Oregon. Unpubfished disserta- tion, Oregon State University, CorvaUis. 178 pp. HiLS, M. H. 1993. Taxus. Pages 424-426 in Flora of North America Committee, editors, Flora of North .\merica. Volume 2. Pteridophytes and Gymnosperms. O.xford University Press, New York. Lloyd, D. G. 1980. Sexual strategies in plants III. A (]uan- titative method for describing the gender of plants. New Zealand Journal of Botany 18: 103-108. Owens, J. N., and S. Simpson. 1986. Pollen from conifers native to British Columbia. Canadian Journal of For- est Research 16: 955-967. Rudolf, R O. 1974. Taxus. Pages 799-802 in Seeds of woody plants in the United States. United States Depart- ment of Agriculture, Forest Service, Washington, DC. 'Canadian Fcirest Service, Pacific Forestry Centre. 506 West Bumside Rd., Victoria, BC VSZ 1M5, Canada. -.\uthor to whom all correspondence should be addressed. 377 378 Great Basin Naturalist [Volume 56 Taylor, R. L., and S. Taylor. 1981. Taxus brevifolia in British Columbia. Davidsonia 12(4): 89-94. Received 28 March 1996 Accepted 5 June 1996 Figs 1-5 Scanniiis electron inicrotirapii confirming cose.xualitA in Pacific yew {Taxus brevifolia) fi-om soutliem yancouver Island British Columbia. Scale bar = 1 mm in each Rgme. 1, Male bud (left) aud fennJe bud (right), both vvidi bud scales partial'h' removed, on the same twig. 2, Young male bud (March) prior to shedding of pollen; bud scales intact. .3, \oung male bud (March) showing the emerging microsporangia (M); bud seniles intact. 4, Young female bud (Mmdi), bud scdes intact. 5, Mature female bud (August) showing die ovule tip (Ov) and micropyle emerging dirough die center of the intact bud scales. ( ;ifut Basin Naturalist 5(i(4), © 199(i, pp. 379-3S() DIURNAL ABOVEGROUND ACTIVITY BY THE FOSSORIAL SILVERY LEGLESS LIZARD, ANNIELLA PULCHRA l^a\icl J. (icMiiiaiu)' and l)a\i(l J. Moralka- Kcy words: uctivilij, lizardu, Anniella, Calijornki, reptiles, behavior. Anniella pulchra is a limbless, fossorial lizard. 11iis species occurs from Antioch, California, to northern Baja California, and is often found on dune fonnations and in sandy habitats where it t\picall> can be captured by raking the soil luider bushes (Miller 1944). It can also be found in se\eral low, coastal mountain ranges (Steb- bins 1985), and its range extends into the San Joaquin Valley and to the edge of the Sonoran Desert in eastern San Diego County (Klauber 1932, Jennings and Hayes 1994). It seems to prefer moist soils (Miller 1944) where it is able to drink (Fusari 1985). Because of its fossorial habit, A. pulchra is rarely found moving above- ground, but it sometimes can be found on the surface at dusk or in the evening (Stebbins 1 985). Here we report tlie previously unrecorded finding of a single A. pulchra moving above- ground during the middle of the day. On 27 April 1995 we were driving on Crocker Springs Road heading northeast over the southern end of the Temblor Mountains. This road is unpaved over the Temblors, and we found 1 A. pulchra on a hard-packed sec- tion of the road. The location was at 769 m (2500 ft) in San Luis Obispo County, approxi- mately 1.5 km west of the county boundary \\ ith Keni County. The lizard, an adult male 140 nmi snout-vent length (217 mm total length), w as found at approximately 1425 h. The day was partly cloudy and the air tem- perature when the lizard was found was about 24 °C. Although the road is not steeply inclined at the location, the surrounding topography traversed by this section of the road is a steep hillside of about a 45-degree slope. Dominant \ egetation on the hillside is alkali goldenbush {Haplopappus acradenius), and no sandy soil occurs near the location where we found the lizard. The lizard was stretched out on the road. which it probably was crossing when we saw it. Unfortunately, we ran over the lizard with our vehicle and were not able to watch its move- ment after we found it. We salvaged the body and deposited it in die museum of die Ctilifoniia Academy of Science (specimen #CAS201173, taken under California Department of Fish and Game peniiit #1111). Besides the injuries we inflicted on the specimen, there were no other signs of injuiy or obvious infestations by para- sites. This is the first obsei-vation we know of showing that A. pulchra sometimes makes above- ground movements during the day. Midday aboveground activity of A. pulchra appears to be a rare behavior. It is possible that this lizard has narrow physiological tolerances that often prevent surface activity, particularly in full sun. A. pulchra has a lower prefeired body temperature than most other lizards (Bury and Balgooyen 1976), and its requirement for moist soil and free water has been known for almost a century (Coe and Kunkel 1907). We found this lizard active at an air temperature of about 24 °C, which is consistent with its preferred thermal range of 24-25 °C (Buiy and Balgooyen 1976). Limited surface acti\4t>', especially away from plant cover, may also be due to predator avoidance. Because limbless lizards are adapted for liurrowing, their ability to move quickly aboveground is limited (Cans 1975). These phys- iological and behavioral constraints likely limit the aboveground activity of A. pulchra to short durations and distances. Literature Cited Bl'RY, R. B., and T. G. BALc;<)()Vii.\. 197tt. Tt'inperatuic selectivity in the legless lizard, Anniella pitlehra. Copeia 1976: 152-155. Coe, W. R., and B. W. Kunkel. 1907. Studies of the Ciili- fornia limbless lizard, Anniella. Transactions of the 'Departmrnt of Biology; California State University', Bakersfield, CA 93311. ^Department of Biolog>', California State University-, Dominguez Hills, Carson, CA 90747 379 380 Great Basin Naturalist [Volume 56 Connecticut Academy of Arts and Sciences 12: 349-403 + plates. FUSARI, M. H. 1985. Drinking of soil water by the Califor- nia legless lizard, Anniella pulchra. Copeia 1985; 981-986. Cans, C. 1975. Tetrapod limblessness; evolution and func- tional con-elates. American Zoologist 15: 455-467. Jennings, M. R., and M. P Hayes. 1994. Amphibian and reptile species of special concern in California. Cali- fornia Department of Fish and Game, Sacramento. 255 pp. KijVUBER, L. M. 1932. Notes on the silver}' foodess lizard, A7iniella pulchra. Copeia 1932; 4-6. Miller, C. M. 1944. Ecologic relations and adaptations of the limbless lizards of the genus Anniella. Ecological Monographs 14: 271-289. Stebbins, R. C. 1985. A field guide to western reptiles and amphibians. Houghton Mifflin Company, Boston, MA. 336 pp. Received 11 March 1996 Accepted 14 June 1996 THE GREAT BASIN NATURALIST N D E X VOLUME 56 — 1996 BRIGHAM YOUNG UNIVERSITY Great Basin Naturalist 56(4), © 1996, pp. 382-389 INDEX Volume 56—1996 Author Index Austin, Dennis D., 167 Baird, Craig R., 237 Barnard, David, 12 Barneby, R. C, 85 Barnum, Andrew H., 283 Bissonette, John A., 1, 319 Bleich, Vernon C, 276 Bloszyk, Jerzy, 59 Bosakowski, Thomas, 341 Bowker, Robert W, 38 Bradley Peter V, 48 Braker, H. Elizabeth, 172 Brusven, M. A., 22 Bursey, Charles R., 180, 369 Buskirk, Steven W, 247 Cashore, Brian L., 183 Cheam, Ha\', 369 Clancy, Karen M.. 135 Clary, W.E, 119 Clayton, M. R., 377 Conway, John R., 54, 326 Cranney, J. Stephen, 142 Crawford, John A., 177 Davidson, Diane W, 95 Davis, Jeffrey L., 276 Davis, Vicki L., 276 Dodds, Kimberly A., 135 Ducharme, Lori A., 333 Eckel, E M., 197 Ehleringer, James R., 333 Fielding, Dennis J., 22 Fraas, W. Wyatt, 205 Frisina, Michael R., 205 Gergus, Erik W. A., 38 Germano, David J., 379 Goldberg, Stephen R., 180, 369 Grayson, Donald K., 191 Greenberg, David, 135 Crier, Charles C, 211 Gutierrez, R. J., 87 Guyon, John C, 129 Haigh, Sandra L., 186 Haines, G. Bruce, 281 Halford, Anne S., 225 Hansen, E. Matthew, 348 Haiper, Kimball T, 95, 294 Hart, KimberK Hamblin, 188 Heckniann, Richard A., 142 Hoffman, James T, 129 Hogg, K. E., 377 Hubert, Wayne A., 300 Hysell, Molly Thomas, 211 Jenkins, Michael J., 28 Keiter, Robert B., 95 Kish, L. E, 22 Knapp, Eaul A., 162 KiTise, Carter G., 300 Ley\'a, Kathiyn J., 135 Lindquist, John L., 267 Livingston, Stephanie D., 191 Looman, Sandra J., 73 Malmos, Keith B., 38, 369 Manning, Sara J., 183 Mathiasen, Robert L., 129 Mattson, David J., 272 Mattson, Todd A., 247 Maxwell, Bruce D., 267 McAllister, Chris T, 180 McCutchen, Hemy E., 90 Mead, Leroy L., 294 Messmer, Terry A., 254 Mihuc, Janet R., 287 Mihuc, Timothy B., 287 Minshall, G. Wayne, 287 Mitchell, A. K., 377 382 1996] Index 383 Modde, Timothy, 281, 375 Moiison, Clark S.. 150 Moratka, Da\i(lJ..379 Ncwniark, W'iliam I)., 95 Nowak. Rohcrt S., 225 IV'cn, M. Zachariah, 87 IVlren, Eric C, 177 IVttengill Thomas, 12 rierce! Beck>- M., 276 Popper, Kenneth J., 177 Porter, Eric E., 172 Ports, Mark A., 48 i'rice, Peter W, 135 |{ahel, Frank J., 300 Hamse\', R. Douglas, 341 Ik-dak,' Richard A., 172 Reinhart, Daniel P, 272 Rickart, Eric A., 95, 191 Romin, Laura A., 1 Schroeder, Sue, 254 Schultz, Brad W, 261 Seamans, Mark E., 87 Shaver, Monson W, III, 191 Sherman, Paul W, 237 Shiozawa, Dennis K., 95 Shirley, Dennis L., 73 Sites, Jack W., Jr, 95 Smith, Dvvight (;., 341 Smith, llohart M., 180 Soule, Peter T., 162 Stanton, Nancy L., 247 Steinkamp, Melanie J., 319 Sullivan, Brian K., 38, 369 Szewczak, Joseph M., 183 Szyinkowiak, Pavvcl, 59 Tanner, Wilmer W., 279 Tausch, Robin J., 261 Taylor, Vicki L., 294 Tiedemann, A. R., 119 Truong, Quynh A., 180 Tueller, Paul T, 261 Van Buren, Renee, 188 Vinyard, Gar\' L., 157, 360 Wcxdleigh, Linda L., 28, 129 Wambolt, Carl L., 205 Weaver, T, 267 Wells, Samuel A., 308 Welsh, Stanley L., 85, 93 White, Clayton M., 73 Wurtshaugh, Wayne A., 12 Yensen, Eric, 237 Yuan, Andv C, 157 Key Word Index Taxa described as new to science in this volume appear in boldface type in this index. Ahies lasiocarpa, 348 acid precipitation, 167 Acrididae, 172 activity, 379 A^ropijron spicatiiin, 267 alfalfa growers, 254 alkalinity, 167 alpine, 225 lakes, 167 amphibian decline, 38 Aiiniclhh 379 ant mounds, 326 thatching, 54, 326 western hai-vester, 162 Arizona, 38, 180, 369 willow, 294 Astragalus laxmannii, 85 Barn Owl food, 73 reproduction, 73 bat(s), 48 silver-haired, 247 bees, 95 behavior, 276, 379 big game nutrition, 205 bighorn sheep, 319 biodiversity, 95 biogeography, 191 bitterbrush, 205 black bears, 90 blood parasites, 142 Blue Grouse, 177 British Columbia, 377 Bromus inennis, 267 brook stickleback, 281 bnophvtes, 197 Biifo microscaphus, 369 rctiformis, 38 bulk density, 211 bunch grass lizard, 180 Bureau of Land Management, 95 Buteo. 341 cactus prickly pear, 211 California, 276, 379 native grassland, 172 Owens Valley, 183 carbon isotope ratios, 333 cations, 119 cattle, 319 Centaiirea dijfiisa Lam., 22 maculosa, 267 Cercocarpus, 261 Cestoda, 180 characteristics, site, 225 check dams, 211 checklist, 197 384 Great Basin Naturalist [Volume 56 Chiroptera, 48 Choristoneura occidentalism 135 class maturity, 261 Colorado, 54, 326 competition, 22, 267 consei"vation, 95, 360 biology, 300 cosexuality, 377 creel, 12 crude protein, 205 ciyptobiotic soils, 95 dams check, 211 Daphnia. 157 decline amphibian, 38 deer, 1 den, 276 Dendra^apiis ol)sciints. 177 dendrochronologv, 294 density, 172 bulk, 211 detritus, 211 diffuse knapweed, 22 dioecious, 377 dispersal, 87, 150 distribution historic, 38 present, 38 disturbance, 319 diversity, 172 Douglas-fir, 129 Drijocoetes confitsus, 348 dwarf mistletoes, 129 ecology seed, 333 shrub, 333 ectoparasites, 237 Elateridae, 308 endemic, 225 species, 95 endemism, 360 Ereiuichthys acros, 360 eruptive species, 135 exotic plant(s), 183, 267 species, 95 Felis concolon 276 Festitca idahoensis, 267 filter feeders, 287 fire freciuency, 28 scar, 28 fish management, 12 parasites, 142 Flaming Gorge Resenoir, 150 flight periodicity, 348 floodplain, 375 flora, 197 foliar qualit}', 135 food habits, 90 forest subalpine spruce-tir, 28 Formica obscuripes, 54, 326 Fourier series, 272 Gastrophrijne olivacea, 38 genetics, 300 Gila atraria, 142 GIS, 341 grasshopper herbivoiy, 172 Great Basin, 59, 191, 360 ground squirrels, 237 growth, 12 habit, 211 habitat. 1, 48, 341, 360 hai"vest, 12 Hawk Red-tailed, 341 Swainson's, 341 helminths, 369 hemiparasite, 333 herbivoiy, 22 grasshopper, 172 heterotroph}, 333 highw a\' mortalit); 1 historic distribution, 38 holotype, 308 host, 186 hybrids, 369 Idaho, 237, 319 insect(s), 22 control, 348 phenology, 348 invasive plant, 183 irrigation, 360 island(s) biogeography, 191 of fertility, 2il juvenile, 375 Lahontan cutthroat trout, 157 Lahontan redside shiner, 157 Lasionycteris iwctivagans, 247 Lewisia longipetala, 225 pygtuaca. 225 limber pine, 90 litter, 261 lizard(s), 379 bunch grass, 180 lodgepole pine, 129 mammals, 191 maturity class, 261 Melanoplus sanguinipes, 172 meristic counts, 300 variation, 300 midden, 272 mineral nitrogen, 211 mistletoe(s), 186 dwarf, 129 mites, 59 Montana, 205 mortality highway, 1 mountain lion, 276 mahogany, 261 myrmecophiles, 326 Nassella pulchra, 172 needle age, 135 Negastrius, 308 atrosus, 308 rupicola, 308 solox, 308 stibicki, 308 Neotoma cinerea, 191 neot>'pe, 308 nest boxes, 73 density, 162 sites, 341 size, 162 Nevada, 48, 360 nitrogen, 119 enrichment, 211 mineral, 211 total, 211 noctin^nal, 177 nomenclature, 85 nonnative, 281 North America, 85 nuptial flight, 54 nutrient cycling, 119 nutrition big game, 205 OdocoiJetis hciuioniis. 1 Oncorhynchus clarki henshawi, 157 tnykiss, 12 Oochoristica scclopori. 180 Oregon, 177 organic carbon, 211 Osprey 150 oviposition, 135 Ovis canadensis, 319 Owens Valle\; [California], 183 Owl, Spotted, 87 1996J Index 385 I'ac'ific yew, 377 randion hdliai'tus, 150 panisite(s), 1(S6 blood, 142 paiatype, 308 rlioradendron calijoniicwn, 186 phosphorus, 119 photosxnthesis, 333 llinnosoiiiatidae, 180 pine lodgepole, 129 pinon-juniper woodlands, 21 1 planktix'on; 157 l)lant(s), 95 density, 225 invasive, 183 size, 225 pocket gopher, 183 Pogonoinynncx owyheei, 162 pollen, 377 Porter, 197 precipitation acid, 167 predation, 157 preference, 135 present distribution, 38 prickly pear cactus, 211 productivity, 12 protein crude, 205 Pternohylo fodiens, 38 Purshio tridentata, 205 Qiierciis gombclii, 119 rainbow trout, 300 range extension, 281 raptor, 150 razorback sucker, 375 recruitment, 261 Red-tailed Hawk, 341 regulation, 12 relationships spatial, 261 reptiles, 379 reservoir, 12 Richardsonius egregiiis, 157 roadkill, 1 llocky M()untain(s), 90, 197 streams, 287 roost(s), 177, 247 runofif, 211 SV///.V, 294 saltccdar, 183, 186 Sccloponis scalaris, 180 ScoKtidae, 348 seed, 377 seedling, 261 shrub ecology, 333 sighting, 272 silver-haired bat, 247 site characteristics, 225 size selectivity, 157 snags, 247 snow-bed vegetation, 225 soil(s) characteristics, 211 ciyptobiotic, 95 nutrients, 119 Sonoran Desert, 38 southern Utah, 294 spatial relationships, 261 species endemic, 95 -environment relationships, 287 eruptive, 135 exotic, 95 Spennophihis hrunneus, 237 Spotted Owl, 87 springs, 360 squirrels ground, 237 stem diameter, 294 strains, 12 Strix occidentalism 87 subalpine spruce-fir forest, 28 sucker razorback, 375 sulfur, 119 sui^veys, 129 Swainson's Hawk, 341 tamarisk, 183 Tainahx rainosi.s.siinu, 183, 186 Taxiis hrevifolia, 377 temperature, 276 thatching ant, 54, 326 Thomomys bottac, 1 83 total nitrogen, 211 Trachytes kaliszewskii, 59 transect, 272 trout, 12 rainbow, 300 Yellowstone cutthroat, 300 Trypanoplasma atraria, 142 turbidity, 157 Tyto (dha, 73 Uropodina, 59 Utah, 59, 73,95, 119, 167 forests, 348 southern, 294 vegetation clearing, 162 snow-bed, 225 vertebrates, 95 vocalization, 272 water quality, 167 weed control, 267 western harvester ants, 162 western spruce budworm, 135 wetland, 375 wilderness, 95 wildfire, 272 wildlife damage management, 254 damage perceptions, 254 management, 254 winter range, 205 Wyoming, 197 Yellowstone cutthroat trout, 300 yield, 12 386 Great Basin Naturalist [Volume 56 TABLE OF CONTENTS Volume 56 No. 1— January 1996 Articles Temporal and spatial distribution of highway mortality of mule deer on newly constructed roads at Jordanelle Resei-voir, Utah Laura A. Romin and John A. Bissonette 1 Exceptional fish yield in a mid-elevation Utah trout reservoir: effects of angling regulations Wayne A. Wurtsbaugh, David Barnard, and Thomas Pettengill 1 2 Consumption of diffuse knapweed by two species of polyphagous grasshoppers (Orthoptera: Acrididae) in southern Idaho Dennis J. Fielding, M. A. Brusven, and L. R Kish 22 Fire frequency and the vegetative mosaic of a spruce-fir forest in northern Utah Linda Wiidleigh and Michael J. Jenkins 28 Arizona distribution of three Sonoran Desert anurans: Bitfo retifonuis, Gastrophrijne oliiacea, and Pteniohyla fodiens Brian K. Sullivan, Robert W. Bowker, Keith B. Malmos, and Erik W. A. Gergus 38 Habitat affinities of bats from northeastern Nevada Mark A. Forts and Peter V Bradley 48 Nuptial, pre, and postnuptial activity of the thatching ant, Formica obsciiripes Forel, in Colorado John R. Conway 54 Trachytes kaliszewskii n. sp. (Acari: Uropodina) from the Great Basin (Utah, USA), with remarks on the habitats and distribution of the members of the genus Trachytes Jerzy Bloszyk and Pawel Szymkowiak 59 Productivity, food habits, and associated variables of Bam Owls utilizing nest boxes in north central Utah Sandra J. Looman, Dennis L. Shirley, and Clayton M. White 73 Notes Astragalus laxmannii Jacquin (Leguminosae) in North America . . . . R. C. Barneby and S. L. Welsh 85 Intermountain movement by Mexican Spotted Owls [Strix occidentalis lucida) R. J. Gutierrez, Mark E. Seamans, and M. Zachariah Peery 87 Limber pine and bears Heniy E. McCutchen 90 Book Review Utah wildflowers: a field guide to northern and central mountains and valleys. Richard J. SJiaic . . . Stanle\' L. Welsh 93 No. 2— April 1996 Articles Selecting wilderness areas to conserve Utah's biological dixersitv' Diane W Daxidson, William D. Newniark, Jack W Sites, Jr, Dennis K. Shiozawa, Eric A. Rickart, Kimbcill T. Haiper, and Robert B. Keiter 95 Nutrient distribution in (^)iierciis gamhehi stands in central Utah A. R. Tiedemann and W P Claiy 1 1 9 Comparsion of two roadside sui-vey procedures for dwaif mistletoes on the Sawtooth National Forest, Idaho Robert L. Mathiasen, James T. Hoffman, John C. Gu\on, and Uinda U. Wadleigh 129 1996] Index 387 Effects of Doiitilas-fir foliage age class on western spruce hudwonn oviposition choice and lan'al performance Kinil)erly A. Dodds, Karen M. Clancy, Kathryn J. Leyva, David Greenberg, and Peter W. Price 1 35 Tnjpanophmiui atrarid sp. n. (Kinetoplastida: Bodonidae) in fishes from tlu^ Sevier River drainage, Utah j. Stephen Craimey and Richard A. Heckniann 142 Geographical re\ic\\ of the liistorical and cnrrent stains of Ospreys {Pandion huliaetus) in Utah Clark S. Monson 150 Effects of turbidity on feeding rates of Lahontan cutthroat trout {Orworhi/ncluis clarki henslmwi) and Lahontan redside shiner {Richardsonius egrc nopsis of the mosses of Wyoming P M. Eckel 1 97 Variation in bitterbrush {Purshia tridentata Pursh) crude protein in southwestern Montana Carl L. Wambolt, W. Wyatt Fraas, and Michael R. Frisina 205 Dam-forming cacti and nitrogen enrichment in a piiion-juniper woodland in northwestern Arizona Molly Thomas Hysell and Charles C. Grier 211 Distribution and ecological characteristics of Leivisia longipetala (Piper) Clay, a high-altitude endemic plant Anne S. Halford and Robert S. Nowak 225 Larger ectoparasites of the Idaho ground squiirel {Spennophilus brunneus) Eric Yensen, Craig R. Baird, and Paul W. Sherman 237 388 Great Basin Naturalist [Volume 56 Roost sites of the silver-haired bat {Lasionycteris noctivagans) in the Black Hills, South Dakota Todd A. Mattson, Steven W. Buskirk, and Nancy L. Stanton 247 Perceptions of Utah alfalfa growers about wildlife damage to their hay crops: implications for managing wildlife on private land Teny A. Messmer and Sue Schroeder 254 Spatial relationships among young Cercocarpus ledifoliiis (curlleaf mountain mahogany) Brad W. Schultz, Robin J. Tausch, and Paul T. Tueller 261 Potential for controlling die spread of Ccntaitrca macuhsa with grass competition John L. Lindquist, Bruce D. Maxwell, and T. Weaver 267 Indicators of red squirrel {Tamiascinnis hiidsonicus) abundance in the whitebark pine zone David J. Mattson and Daniel P Reinhart 272 Thermal characteristics of mountain lion dens Vernon C. Bleich, Becky M. Pierce, Jeffrey L. Davis, and Vicki L. Davis 276 James William Bee, 1913-1996 Wilmer W. Tanner 279 Note Brook stickleback {Ciilaea inconstans [Kirtland 1841]), a new addition to the Upper Colorado River Basin fish fauna Timothy Modde and G. Bruce Haines 281 Errata 282 Book Review Snakes of Utah. Doiidas C. Cox and Wihner W. Tanner Andrew H. Bamum 283 No. 4— October 1996 Articles Species-environment relationships among filter-feeding caddisflies (Trichoptera: Hydropsychidae) in Rocky Mountain streams Timothy B. Mihuc, G. Wayne Minshall, and Janet R. Mihuc 287 Stem growth and longevity dynamics for Salix arizonica Dorn Vicky L. Ta\'lor Kimball T. Harper and Leroy L. Mead 294 Sources of variation in counts of meristic features of Yellowstone cutthroat trout {Oncorhijnchiis clarki boiivicri) Carter G. Kruse, Wayne A. Hubert, and Frank J. Rahel 300 Studies on Nearctic Negastriiis (Coleoptera: Elateridae) Samuel A. Wells 308 Bighorn sheep response to ephemeral habitat fragmentation b\' cattle J. A. Bissonette and Melanie J. Steinkamp 319 A field study of the nesting ecology of the thatching ant, Fonnica ohsciiripes Forel, at high altitude in Colorado J"hn R. Conwax 326 Gas exchange, 5^'^C, and heterotrophy for CastiUeja linariifolia and Orflnx^arpus tohnici, faculta- tive root hemiparasites on Artemisia tridentata Lori A. Ducharme and James R. Ehleringer 333 Habitat and spatial relationships of nesting Swainson's Hawks [Buteo swainsoni) and Red-tailed Hawks {B. jamaicensis) in northern Utali Thomas Bosakowski, R. Douglas Ramsey, and Dwight G. Smith 341 Western balsam bark beetle, Drijocoetes confusns Swaine, flight periodicity in northern Utah E. Matthew Hansen 348 Distribution ol a thermal endemic iniimow, the desert dace (Ereinichtluis aeros), and obsen'ations of impacts of water diversion on its population Gary L. Vinyard 360 1996] Index 389 Helminths of the southwestern toad, Bufo iiiicroscaphus, Woodhoiise's toad, Btifo ivoocUwtisii (Butonidae), and their h> hrids troni eentral Arizona Stephen R. Goldberg, Charles K. Bursey, Keith B. Malnios, Brian K. Snlli\an, and Hay C^heani 369 Notes Juvenile razorhaek sneker (Xi/rdiichoi fe.xaiiKs) in a inana^ed wetland adjaeenl to the Green River Timothy Modde 375 Cont'irmation ol'eosexualit) in Faeiiie yew [Taxiis brevifolia Nutt.) . . . . K. E. Hogg, A. K. Mitchell, and M. R. Clayton 377 Diurnal abo\ egiound aeti\it>' b\' the fossorial silvery legless lizard, Anniella pulchra David J. Germano and David J. Morafka 379 p INFORMATION FOR AUTHORS The Great Basin Naturalist welcomes previously unpublished mauuseripts pertaining to the biologi- cal natural history of western North America. Preference will be given to concise manuscripts of up to 12,000 words. Simple species lists are dis- coin-aged. SUBMIT MANUSCRIPTS to Richard W. Baumann, Editor, Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young University, Provo, UT 84602-0200. An accompanying cover letter must include phone number(s) of the author submitting the manuscript, and FAX number and E-mail address when applicable; the letter must also pro- \ ide information describing the extent to which data, text, or illustrations have been used in other papers or books that are published, in press, submitted, or soon to be submitted elsewhere. Authors should adhere to the following guidelines; manuscripts not so prepared ma\' be returned for revision. MANUSCRIPT PREPARATION. In general, the Great Basin Naturalist follows recommendations in Scientific Style and Format: The CBE Manual for Authors, Editors, and Publishers, 6th edition (Council of Biology Editors, Inc., 11 South LaSalle Street, Suite 1400,' Chicago, IL 60603, USA; phone 312-201-0101; FAX 312-201-0214). We do, however, differ in our treatment of entries in Literature Cited. \ Authors may consult Vol. 51, No. 2 of this journal for specific instructions on format; these instruc- tions. Guidelines for Manuscripts Submitted TO the Great Basin Naturalist, are printed at the back of the issue. Also, check the most recent issue of the Great Basin Naturalist for changes. TYPE AND DOUBLE SPACE all materials, including literature cited, table headings, and figure legends. Avoid hyphenated words at the right-hand margins. Underline words to be printed in italics. Use stan- dard bond (22x28 cm), leaving 2.5-cm margins on all sides. SUBMIT 3 COPIES of the manuscript (5 copies of fish manuscripts) and the original on a 5.25- or 3.5- inch disk utilizing WordPerfect 5.1 or above. j Number all pages and assemble each copy sepa- ! rately: title page, abstract and key words, text, i acknowledgments, literature cited, appendices, tables, figure legends, figures. TITLE PAGE includes an informative title no longer than 15 words, names and addresses of authors, a running head of fewer than 40 letters and spaces, footnotes to indicate change of address and author to whom correspondence should be addressed if other than the first author. ABSTRACT states the purpose, methods, results, and conclusions of the research. It is followed by , 6-12 key words, listed in order of decreasing importance, to be used for indexing. TEXT has centered main headings printed in all capital letters; second-level headings are centered in upper- and lowercase letters; third-level head- ings begin paragraphs. VOUCHER SPECIMENS. Authors are encouraged to designate, properly prepare, label, and deposit high-(juality voucher specimens and cultures docu- menting their research in an established permanent collection, and to cite the repository in publication. REFERENCES IN THE TEXT are cited by author and date: e.g., Martin (1989) or (Martin 1989). Multiple citations should be separated by commas and listed in chronological order. Use "et al." after name of first author tor citations having more than two authors. ACKNOWLEDGMENTS, under a centered main heading, include special publication numbers when appropriate. LITERATURE CITED, also under a centered main heading, lists references alphabetically in the fol- lowing formats; Mack, G. D., and L. D. Flake. 1980. Habitat rela- tionships of waterfowl broods on South Dakota stock ponds. Journal of Wildlife Management 44: 695-700. Sousa, W. E 1985. Disturbance and patch dynamics on rocky intertidal shores. Pages 101-124 in S. T. A. Pickett and E S. White, eds., The ecolo- gy of natural disturbance and patch dynamics. Academic Press, New York. Coulson, R. N., and J. A. Witter. 1984. Forest ento- mology: ecology and management. John Wiley and Sons, Inc., New York. 669 pp. TABLES are double spaced on separate sheets and designed to fit the width of either a single column or a page. Use lowercase letters to indicate foot- notes. PHOTOCOPIES OF FIGURES are submitted initially with the manuscript; editors may suggest changes. Lettering on figures should be large enough to withstand reduction to one- or two-column width. Originals must be no larger than 22x28 cm. NOTES. If the manuscript would be more appro- priate as a short communication or note, follow the above instructions but do not include an abstract. A CHARGE of $50 per page is made for articles published; the rate for individual subscribers will be $35 per page. However, manuscripts with com- plex tables and/or numerous half-tones will be assessed an additional charge. Reprints may be pur- chased at the time of publication (an order form is sent with the proofs). FINAL CHECK; • Cover letter explaining any duplication of information and providing phone number(s), FAX number, and E-mail address • 3 copies of the manuscript and WordPerfect diskette • Conformity with instructions • Photocopies of illustrations (ISSN 001 7-361 4) GREAT BASIN NATURALIST Vol 56 no 4 October 1 996 CONTENTS Articles Species -environment relationships among filter- feeding caddisflies (Trichoptera: Hydropsychidae) in Rocky Mountain streams Timothy B. Mihuc, G. Wayne Minshall, and Janet R. Mihuc 287 Stem growth and longevity dynamics for Salix arizonica Dorn Vicky L. Taylor, Kimball T Harper, and Leroy L. Mead 294 Sources of variation in counts of meristic features of Yellowstone cutthroat trout {Oncorhijnchiis clarki bouvieri) Carter G. Kruse, Wayne A. Hubert, and Frank J. Rahel 300 Studies on Nearctic Negastrius (Coleoptera: Elateridae) Samuel A. Wells 308 Bighorn sheep response to ephemeral habitat fragmentation by cattle J. A. Bissonette and Melanie J. Steinkamp 319 A field study of the nesting ecology of the thatching ant, Formica obscuripcs Forel, at high altitude in Colorado John R. Conway 326 Gas exchange, S^-^C, and heterotrophy for Castilleja linariifolia and Orthocarpus tolmiei, facultative root hemiparasites on Artemisia trideniata Lori A. Duchamie and James R. Ehleringer 333 Habitat and spatial relationships of nesting Swainson's Hawks {Buteo swainsoni) and Red-tailed Hawks {B. jamaicensis) in northern Utah Thomas Bosakowski, R. Douglas Ramsey, and Dwight G. Smith 341 Western balsam bark beetle, Drijocoetes confusus Swaine, flight periodicity in northern Utah E. Matthew Hansen 348 Distribution of a thermal endemic minnow, the desert dace {Eremichthijs acros), and observations of impacts of water diversion on its population Gaiy L. Vinyard 360 Helminths of the southwestern toad, Bufo microscaphus, Woodhouse's toad, Bufo woodhousii (Bufonidae), and their hybrids from central Arizona Stephen R. Goldberg, Charles R. Bursey, Keith B. Malmos, Brian K. Sullivan, and Hay Cheam 369 Notes Juvenile razorback sucker {Xyrauchen texanus) in a managed wedand adjacent to the Green River Timothy Modde 375 Confirmation of cosexuality in Pacific yew {Taxus brevifolia Nutt.) K. E. Hogg, A. K. Mitchell, and M. R. Clayton 377 Diurnal aboveground activity by the fossorial silvery legless lizard, Anniella pulchra David J. Germano and David J. Morafka 379 Index to Volume 56 381 !ii:isiiii>!ir'!lli!' 3 2044 072 231 145