Great Basin Naturalist

Volume 54 Number 4 Article 14

10-25-1994

Full Issue, Vol. 54 No. 4

Follow this and additional works at: https://scholarsarchive.byu.edu/gbn

Recommended Citation (1994) "Full Issue, Vol. 54 No. 4," Great Basin Naturalist: Vol. 54 : No. 4 , Article 14. Available at: https://scholarsarchive.byu.edu/gbn/vol54/iss4/14

This Full Issue is brought to you for free and open access by the Western North American Naturalist Publications at BYU ScholarsArchive. It has been accepted for inclusion in Great Basin Naturalist by an authorized editor of BYU ScholarsArchive. For more information, please contact [email protected], [email protected]. T H E REATGREATG BASIN MTURALISTNA

VOLUME 54 NQ 4 OCTOBER 1994

BRIGHAM YOUNG university GREAT basinnaturalistBASIN 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 84602020084602 0200 provo UT 84602687984602 6879 8013785053801 378 5053 8013786688801 378 6688 FAX 8013783733801 378 3733 emailE mail nmshbll1byuedu

associate editors MICHAEL A BOWERS PAUL C MARSH blandy experimental fannfarmearm 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 89154400489154 4004 JEFFREY J JOHANSEN PAUL T TUELLER department of biology john carroll university department of environmental resource sciences university heights OH 44118 university of nevada reno 1000 valley road reno NV 89512 BORIS C kondratieff department of entomology colorado state ROBERT C WHITMORE university fort collins CO 80523 division of forestry box 6125 west virginia university Morganmorgantowntown WV 26506612526506 6125 editorial board berranjerran I1 flinders chairmanChainnan botany and range science duke S rogers zoology william hess botany and range science H duane smith zoology all are at brigham young university ex officio editorial board members include steven L taylor college of biology and agriculture stanley L welsh 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great basin naturalist the great basin naturalist founded in 1939 is published quarterly by brigham young university unpublished manuscripts that further our biological understanding of the great basin and surrounding areas in western north america are accepted for publication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 1994 are 25 for individual sub- scriscribersbers 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 84602020084602 0200 scholarly exchanges braneslibrariesLi 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 84602688984602 6889

editorial production staff joanne abel technical editor jan spencer assistant to the editor

copyright 0 1994 by brigham young university ISSN 001736140017 3614 official publication date 25 october 1994 109410 94 750 11712 The Great Basin Naturalist PUBLISHED AT PROVO, UTAH, BY BRIGHAM YOUNG UNIVERSITY

VOLUME 54 31 OCTOBER 1994 NO.4

Great Basin Naturalist 54(4), © 1994, pp. 291-300

MYCORRHIZAL COLONIZATION, HYPHAL LENGTHS, AND SOIL MOISTURE ASSOCIATED WITH TWO ARTEMISIA TRIDENTATA SUBSPECIES

James D. Trentl, Tony J. Svejcar2, and Robert R. Blank!

ABsTRACT.-Mycorrhizal fungi are thought to benefit associated plant species via enhanced nutrient uptake and/or improved water relations. However, detailed descriptions of the components of mycorrhizal colonization and mycor­ rhizal hyphal growth are not available for Artemisia tridentata. This species occupies sites characterized by relatively low levels ofboth soil nutrients and moisture. We studied patterns ofvesicular, arbuscular, and hyphal mycorrhizal colo­ nization, mycorrhizal hyphal lengths, and soil moisture associated with two subspecies ofA. tridentata over a 2-year period. A. tridentata ssp. vaseyana (ATV) is generally associated with more mesic and slightly higher elevation sites compared to A. tridentata ssp. tridentata (ATI). Nearly twice as much precipitation was received the first year compared to the second. In general, there were higher levels of colonization and hyphal lengths associated with ATV than with ATT. The ATV site received slightly more precipitation and was lower in available nutrients than the ATI site. Hyphal lengths and arbuscular colonization appeared more responsive to precipitation than were either vesicular or hyphal col­ onization. Hyphal colonization did not necessarily follow the same temporal pattern as hyphal lengths. Thus, mycor­ rhizal activity was greater for the subspecies that received slightly more precipitation and occupied a site lower in avail­ able nutrients. Arbuscular colonization and hyphal lengths appeared to be most closely associated with soil moisture and thus plant activity.

Key words: vesicular-arbuscular mycorrhizae, hyphallength, Artemisia tridentata, soil moisture, soil temperature.

Understanding ecosystem processes requires 1988). Morphogenesis ofarbuscule and vesicle baseline data that describe spatial and tempo­ formation should be differentiated when ral variations in microbial mediated processes assessing functionality or dependency of the (Burke et al. 1989). Such information is also plant on VAM on a seasonal basis. The pres~ needed to assess the role of mycorrhizae and ence of arbuscules indicates plant"fungal other fungi in native plant communities. In interactions (Hirrel et al. 1978, Allen 1983) native plant communities, vesicular-arbuscu­ since arbuscules are the site for P and C trans­ lar mycorrhizal (VAM) colonization has been fer between syrnbionts (Cox and Tinker 1976, shown to vary both seasonally and among Wilcox 1993). Seasonal changes in extramatri­ .plant species (Read et al. 1976, Rabatin 1979, cal VAM fungal hyphae indicate that plant­ Daft et al. 1980, Gay et al. 1982, Allen 1983, fungal interactions are dynamic (Wilcox 1993). Giovannetti 1985, Brundrett and Kendrick Therefore, itis necessary to measure seasonality

1USDA-ARS, 920 Valley Road, Reno, Nevada 89512. 2USDA-ARS, HC 71, 4.51 Hwy 205, Burns, Oregon 97720. Address reprint requests and correspondence to this author.

291 292 GREAT BASIN NATURALIST [Volume 54 of arbuscular colonization and extramatrical Oryzopsis hymenoides, Elymus hystrix, and hyphae to adequately assess the changing rela­ Bromus tectorum. The soil is classified as a tionship between symbionts in the field. coarse~loamy, mixed, mesic Aridic Argixeroll. The Great Basin environment is character­ This is an alluvial fan soil that was mainly ized by winter precipitation, normally as snow, derived from granitic rocks. The soil consists followed by hot, dry summers (Comstock and of about 40 em of loamy sand to gravelly Ehleringer 1992). Root growth is most abtm­ loamy sand overlying a subsoil ofabout 50 em dant in upper soil horizons in the early spring. of sandy loam. The underlying material to over Rooting activity diminishes in upper soil hori­ 200 em is loamy coarse sand. zons as root growth follows the soil moisture The ATV community is at 1830 m elevation profile into deeper soil layers (Fernandez and and includes the following vegetation: Artemisia Caldwell 1975). In desert soils, N and Pare tridentata ssp. vaseyana, Purshia tridentata, most abundant in upper soil horizons (West Ribes sp., Chrysothamnus viscidijlorus, Stipa 1991), and their availability to plants diminishes columbiana, S. occidentalis, Elymus hystrix, and with decreasing soil moisture. Moisture move~ Bromus tectorum. the soil is classified as a ment from deep roots in moist soils to shallow coarse-loamy, mixed, frigid Ultic Argixeroll. roots in dry soils could make shallow soil nutri~ This type of upland soil formed in a residuum ents available through the process ofhydraulic from granodiorite and consists of about 60 em lift (Passioura 1988, Caldwell and Richards of gravelly coarse sand and loamy coarse sand 1989, Caldwell et al. 1991). Mycorrhizae could overlying a subsoil of about 30 em of loamy playa role in this process (Richards and coarse sand and sandy loam. The underlying Caldwell 1987, Caldwell et al. 1991); however, material consists of about 10 em of weathered little is known about seasonal dynamics of granodiorite. mycorrhizae in arid ecosystems. Fitter (1993) suggests that plant root systems METHODS evolve in a manner that optimizes the use of Sampling was conducted on seven dates in plant carbon. Mycorrhizal colonization and 1989 and five dates in 1990. Four replicate the formation of extramatrical hyphae should macroplots (20 X 20 m) were randomly selected also reflect an optimization of plant carbon at each site. One shrub within a macroplot was usage. But to date there is relatively little selected for sampling each year. We changed information on spatial or temporal variation in shrubs to ensure that the prior year's sampling mycorrhizal activity in the Great Basin. In this did not influence measured parameters. study we quantified VAM arbuscular toot col­ Within each macroplot a CRI0 micrologger onization, vesicular root colonization, hyphal (Campbell Scientific, Logan, Utah)3 was root colonization, and mycorrhizal hyphal equipped to measure soil temperature at 10 length through the plant growing season for 2 em and soil moisture at 10, 30, and 60 em, years inArtemisia tridentata ssp. tridentata and adjacent to the target shrub. Soil temperatures Artemisia tridentata ssp. vaseyana. We have were measured with thermocouples, and soil characterized the seasonality of the above moisture with gypsum blocks. Leader length parameters and show their relationship to was measured as a plant growth indicator changes in both soil moisture and temperature. (Barker and McKell 1986) to avoid destructive harvest ofArtemisia shrubs. Five marked lead­ STUDY SITE DESCRIPTION ers were measured on the target shrub within each macroplot at each sampling date. An Artemisia tridentata ssp. tridentata (ATT) Soil samples were collecteq with a spade to and an Artemisia tridentata ssp. vaseyana (ATV) a depth of 20 em from within the dripline of plant community were chosen for study. The the target shrub in each macroplot on each study sites are located approximately 30 miles date. Sample volumes were about 8000 cm3 northwest of Reno, Nevada, and are within 3 (20 X 20 X 20-cm cube). Roots were sieved miles of each other. The ATT community is at from soil, and both root and soil samples were 1555 m elevation and is composed of the fol­ placed in plastic bags and kept on ice prior to lowing vegetation: Artemisia tridentata ssp. tridentata, Chrysothamnus viscidiflorus, Ephedra viridis, Stipa comata, S. thurberiana, 3Mention oftrade names does not indicate endorsement by USDA. 1994] MYCORRHIZAE IN SUBSPECIES OF SAGEBRUSH 293

storage at -.2 0 C. During the sieving process and Sommers 1982); organic carbon was deter­ we discarded non-Artemisia roots, which are mined using the Walkley~Blackprocedure easily differentiated by color and morphology.. (Nelson and Sommers 1982); and partic1e=size In the laboratory roots were washed and cut analysis was determined using standard meth­ into l=cm segments; they were then cleared ods (Gee and Bauder 1982). with KOH and stained with trypan blue Data were analyzed by year. Artemisia sites (Phillips and Hayman 1910). We estimated per­ .and sampling date effects were assessed using cent total, arbuscular, hyphal, and vesicular analysis of variance with SAS (Statistical colonization using the gridline=intersect Analysis System). All VAM root colonization method with a compound microscope at 160X. data were transformed by taking the arcsine Hyphallengths were quantified using the and square root of VAM root intersects per follOwing modified Bethlenfalvay and Ames total root intersects prior to conducting analy­ (1987) procedure: (1) 20 g of soil was added to sis ofvariance. 100 ml of 0.05% trypan blue solution and boiled for 15 min; (2) samples were cooled and RESULTS 100 ml of sodium hexametaphosphate was added to each flask; (3) flask contents were The ATT site had significantly higher levels added to a Waring blender and blended for of bicarbonate extractable P, KCI extractable about 5 sec; and (4) an aliquot was added to a S04-2, and KCI extractable N03- than the microscope slide and scored for hyphal ATV site (Table 1). ATT soils had a Significant­ lengths at 400X. Hyphallengths were mea= ly lower proportion of sand and Significantly car~ sured on six different aliquots per soil slurry more silt and clay than ATV soils. Organic using an improved Neubauer ultra plane bon and NH4+ were not significantly different between the two sites. Neither soil chemical counting chamber with a depth of 0.1 mm. nor physical characteristics were significantly Four randomly selected transects from each different between years (Table 1). The signifi­ aliquot were scanned at 400X. Each scan was cant Site*Date interaction for extractable P is 1.9 mm long and 0.395 mm wide. From these evident since P increases in the second year at dimensions the liquid volume scanned was the ATT site, yet decreases at the ATV site. calculated (3.12 10-4 mll scan). Slides were x Leader lengths were shorter in both plant scanned and hyphallengths quantified using an communities in 1990 than in 1989 (Fig. 1). In image analysis system. Aliquots were averaged addition, leader length was slightly higher in prior to statistical analysis. We tried a 1:50 soil the ATV community in both years. Decreased to extractant ratio initially, but found we had to leader lengths in 1990 are attributed to lower use 1:10 because VAM hyphallengths were low 1:10 The extraction has been shown by Ingham TABLE 1. Means and probability values for soil physical and Klein (1984) to be adequate for measuring and chemical characteristics of the A. tridentata ssp. tri­ hyphallengths. Since roots were removed dentata (ATT) and A. tridentata ssp. vaseyana (ATV) sites before extraction, VAM hyphal lengths pre­ in 1989 and 1990. Soil samples used for mycorrhizal char­ sented do not include rhizoplane hyphae. acterization were used in this analysis, which accounts for the change in textural analysis between years. Inorganic P Inclusion of rhizoplane hyphae would no (bicarbonate extractable), N (KCl extractable), S (KGl doubt elevate hyphallengths; however, exclu~ extractable), and Walldey-Black soil organic carbon (O.C.) sion of roots allows a more accurate depiction are given in the table (n = 4a). ofthe hyphae that extend beyond the zone of ATT ATV phosphorus depletion around roots. Criteria Soil for determining VAM fungi were similar to variable 1989 1990 1989 1990 Site those established by Allen and Allen (1986). % sand 84.2 80.4 91.1 92.3 .006 Most mycorrhizal hyphae are branched, have % silt 11.5 14.8 6.2 5.8 .008 a knobby appearance, are aseptate, absorb try­ % clay 4.3 4.9 2.8 1.8 .046 pan blue, and are about 3-10 Mm in diametet P/-Lgg-l 12.1 14.3 9.8 7.9 .0005 Soil samples collected from each site were NH4/-Lgg-l 5.2 5.1 5.6 3.4 .39 N03-/-Lgg-l 6.6 8.3 4.6 4.0 .03 composited each year for soil chemical and S04-2/-Lgg-l 7.8 9.6 4.6 5.2 .001 physical analyses. Nitrate, NH4+, and S04-2 O.C.mgg-1 9.6 8.8 9.3 10.0 .62 were extracted using 0.1 M KCI; phosphorus "The effect ofyear was not significant (P > .05) for any variable, and only in was extracted using sodium bicarbonate (Olsen the case ofPwas the Site and Year interaction significant (P < .05). 294 GREAT BASIN NATURALIST [Volume 54

30 precipitation received during that year (Fig. . ATT89 2). Annual precipitation for the growing sea~ 0 ATV89 25 .. ATT90 son was measured beginning in November. In .. ATv90 1989 ATI and ATV communities received 384 20 and 436 mm of precipitation, respectively. However, during 1990 the ATI and ATV com~ 15 w munities received only 194 and 206 mm, 10 Sw respectively. The 4~year average precipitation Cl w > (1984-87) for ATI and ATV was 211 and 256 5 mm,respectivel~ The two sites were not different from each 0-l--+~~-+~~_~+--l-~+--l-~+---+--1 J F MAM JJ A SON D other in either maximum or minimum temper­ MONTH ature (Fig. 2). Maximum March temperatures at lO-cm soil depths for both sites and years Fig. 1. Seasonal change in leader length for Artemisia were 7=13°C; while minimum temperatures tridentata ssp. tridentata (ATI) and Artemisia tridentata were 24°C. Maximum soil temperatures in ssp. vaseyana (ATV) during the 1989 and 1990 growing seasons. Phenological stage appears at the top of the July=August at 10 cm for both sites and years graph. were 30-31°C, while corresponding minimum

35 -,--~~~~~~~~~~~~~~~~~-~~~-~ --ATT 30 --- ATV

o~ 25 W ~ 20 "'"~ W 15 C. r5 10 """ 5

o+-+-+-;--;....-;e______t-+_+____+__+_--i--'~'r=+=;__;...._;e______t__+_+__+____+__+__+_I 100 c:::::::J AIT 90 ~ATV E 80 .§. 70 z: o 60 i== i5 50 c:; 40 C3 ~ 30 C. 20

ND J F MAM JJ A 5 0 N D J F MAM JJ A 5 0 88 89 OAT6 90

Fig. 2. Daily maximum and minimum soil temperatures at the 10-cm soil depth and monthly precipitation for Artemisia tridentata ssp. tridentata (ATI) and Artemisia tridentata ssp. vaseyana (ATV) during the 1989 and 1990 grow­ ing seasons. 1994] MYCORRHIZAE IN SUBSPECIES OF SAGEBRUSH 295 temperatures were 19-21°C. By mid~October However, in 1990 both percent hyphal and maximum temperatures were 16-19°C and total colonization changed significantly through minimum temperatures were 8-9°C. the season (Fig. 4, statistics in Table 2). In both During early 1989 percent arbuscular colo­ years a Significant subspecies effect was ob­ nization was similar for the two Artemisia sub­ served for vesicular and total colonization. species but diverged to greater levels in ATV Subspecies vaseyana roots had greater levels roots by early June (Fig. 3). Arbuscular colo­ ofcolonization than tridentata roots. nization ofAU roots dropped to much lower Mycorrhizal hyphallengths changed signif­ levels than ATV roots during the midsummer icantly through the season for both years (Fig. dry period. This was followed by a minimal 5, st.atistics in Table 2). In 1989 and 1990 increase in colonization during fall for both hyphallength more than doubled from March subspecies. In 1990 percent arbuscular colo­ to May for all sites except the ATV site in 1990. nization was consistently lower for ATT than Hyphallengths decreased during summer and ATV roots (Fig. 4, Table 2). There was also a remained constant during fall for all sites with decrease in colonization from early spring the exception ofATV in 1990, which increased through summer for both subspecies. Overall slightly during the fall. Mycorrhizal hyphal~ analysis ofvariance indicates a significant sea­ lengths were significantly greater at the ATV son and subspecies effect for percent arbuscu­ site when compared to the ATT site for most lar colonization in both years (Table 2), sampling dates. In 1989 percent hyphal, vesicular, and total Soil moisture depletion data are presented in colonization did not significantly change Figure 6. In both years the ATV site had high­ through the season (Fig. 3, statistics in Table 2). er soil water potentials for a greater proportion

1989 7 1989 I 30 Art ATT Z • z • 0 0 ATV 06 0 ATV i= 25 1§ i= z 1§ 5 0 20 z ...J 0 0 54 u 15 u ...J ;i3 ~ a. 0 10 >- f- J:2 ~ ~ 0 5

0 0 z 14 0 z i= 20 Q12 ~ z ~ 0 Z 10 ...J 15 0 0 ...J U 0 a:: u 8 :) a:: =:> 10 :) 6 u =:> en U =:> ii5 4 [lJ w a:: 5 >

0 0 M A M JJ A S 0 M A M J .J A S 0 MONTH MONTH

Fig. 3. Monthly changes in percent total, arbuscular, vesicular, and hyphal root colonization for Artemisia tridentata ssp. tridentata (ATT) andArtemisia tridentata ssp. vaseyana (ATV) during the 1989 growing season. 296 GREAT BASIN NATURALIST [Volume 54

6 1990 35 • ATT Z 1 o ATV 0 30 6 - j::; F ~6 biz 25 Z 0 0 5 ....I d 0 20 u 84 d ....I <{ 15 « I- 0. 3 0 >- I- 10 :r: ~ ~2 5

0 0 ·~O z 0 20 516 ~ ~16 z 0 Z 14 d 15 0 0 u 6 12 0:: ~ 10 ::5 10 ::> - ::s 6 u ::> U) u ::> iii 6 III w 0:: 5 > 4 « ~ ;1!. 2 0 Ii M A M JJ A S 0 M A M J J A S 0 MONTH MONTH

Fig. 4. Monthly changes in percent total, arbuscular, vesicular, and hyphal root colonization for Artemisia tridentata ssp. tridentata (ATI) andArtemisia tridentata ssp. vaseyana (ATV) during the 1990 growing season.

TABLE 2. Site means and results of two-way ANOVAs showing the effect of sagebrush site (Artemisia tridentata ssp. tridentata vs. ssp. vaseyana) and time ofyear (Date) on mycorrhizal colonization (ARB = arbuscular colonization, YES = vesicular colonization, HYP = hyphal colonization, TOT,", total colonization) and mycorrhizal hyphallength (mg-I ). Site meansa ATI ATV Site Date Site*Date

------~------1989------% ARB 10.1 13.1 .009 .014 .316 %VES 5.4 8.8 .004 .517 .031 %HYP 3.9 4.8 .346 .205 .563 % TOT 19.4 26.'1 .004 .067 .149 Length 1.3 1.8 .018 .0004 .491

------~~-----_------1990---~~------% ARB 6.1 13.4 .0001 .005 .461 %VES 5.9 14.6 .0003 .116 .206 %HYP 3.1 4.6 .065 .003 .852 % TOT 15.1 32.6 .0001 .004 .511 Length 0.9 1.7 .005 .050 .373 aSite m~ans were avcrnged over sampling date for each year (Arr = Artemisia tridcnlala ssp. tridentata, ATV = Artemisia tridentata s;p. vaseyana). 1994] MYCORRHIZAE IN SUBSPECIES OF SAGEBRUSH 297

1989 3 • An o AIV

Q+---I---+--+--~--+---l---+--"'"

3 1990 • ATT o ATV en ..s iE 2 c:> Z LU -' -' - =!= 1

Q-'-__'---o.-_--'-__--'-__.L.-_--'-__--'-__.L-_-' M A M JJ A s o

Fig. 5. Seasonal changes in mycorrhizal hyphallengtbs (m ofhyphae/g of soil) in soil adjacent to Artemisia tridentata ssp. tridentata (ATI) andArtemisia tridentata ssp. vaseyana (ATV) during the 1989 and 1990 growing seasons. ofthe season than did the ATI site. Soil mois­ pied. Thus, differences must be attributed to ture declined more rapidly in 1990 compared the species/site combination. The ATV site to 1989. Soil moisture at 10 and 30 cm fell was slightly more mesic than tlle A'IT site, but below ~1.5 MPa at both sites by midsummer we did not detect any differences in tempera­ 1990. In 1989 late summer and early fall precip­ ture (Fig. 2). In general, ATV tends to occupy itation raised soil water potentials above =0.2 sites ofhigher elevation or with more summer MPa at both sites by October. precipitation than ATI (e.g., Winward 1980). In this study we also found that on a seasonal DISCUSSION average the ATV site was lower in available nutrients than the ATI site (Table 1). Nutrient The two Artemisia subspecies exhibited differences may be the result of either plant consistent differences in both colonization and community differences or the fact that the hyphal density of the associated mycorrhizae. ATV site had a higher sand content. Higher Because the study was observational in nature, levels of colonization and greater hyphal we could not separate genetics of two sub­ lengths at the ATV site compared to the ATI species from the differences in sites they occU- site were thus associated with more available 298 GREAT BASIN NATURALIST [Volume 54

< 1.5 .r.-.------.===4t--=---:~

1.5 .

1.0 . -a.Cll :2 0.5 -!,.. ...I 0.0 .I-e--r~+=~--+--t---l---+..!..--+I $ <1.5" .---. I- Z UJ 1.5 I- oa. 1.0 c::: W 0.5 ~>--t---¥c=V=:"""""I---+---t----+--t---+l i 0.0 ...I < 1.5 _. o 60 em 60 em en 1.5"

1.0 ..

0.5

0.0 .L-{'>=-,=(~<..F'=:>..L-_-'----'----'---'--_W MA M JJ A s o N MAM JJ A SON ·1989 1990

Fig. 6. Monthly change in soil water potential (-MPa) in Artemisia tridelltata ssp. tridentata (ATI) and Artemisia tri­ dentata ssp. vaseyana (ATV) plant communities. Water potentials were measured in 1989 and 1990 at 10-, 30- and 60-cm soil depths with gypsum blocks. soil moisture and lower nutrient availability. 1989. Previous research has demonstrated that These data tend to support the hypothesis that A. tridentata achieves highest photosynthetic mycorrhizae are more active when moisture is rates during late spring when moisture is available. and nutrients are limiting. available and temperatures are generally not Arbuscular colonization is thought to be limiting (DePuit and Caldwell 1973). The peak particularly important to carbon and phospho­ in root growth ofA. tridentata also occurs in rus exchange (Cox and Tinker 1976, Wilcox mid~April to late May (Fernandez and Caldwell 1993), and thus this feature plays a critical role 1915). Our measurements of leader growth in plant/fungus interaction (Hirrel et al. 1978, (Fig. 1) appear to confirm that peak above­ Chilvers and Daft 1982, Allen 1983). The two ground activity occurred in late spring or early study sites differed in arbuscular colonization summer during this study. However, during during both years. However, the difference was the dry year of 1990 there was no increase in more consistent during 1990 compared to 1989. mycorrhizal activity dUring the spring; rather, The difference in P availability between sites activity generally declined. The fact that pre­ was 23% in 1989 and 81% in 1990. Thus, the cipitation and leader growth in 1990 were greatest separation in mycorrhizal activity roughly halfof1989 values indicates that envi­ between sites corresponded to the greatest ronmental stress may have limited plant activ­ differenc~s in P availability. ity in 1990. Plant stress and limited carbon Timing of maximal mycorrhizal activity assimilation likely reduced root and/or mycor­ appeared to correspond to aboveground activ~ rhizal activity. Because mycorrhizal activity ity during the higher precipitation year of can be a rather substantial carbon cost to the 1994] MYCORRHIZAE IN SUBSPECIES OF SAGEBRUSH 299

associated plant (Chapin et al. 1987), it is not BURKE, r. C., W A. REINERS, AND D. S. SCHIMEL. 1989. surprising that mycorrhizal colonization and Organic matter turnover in a sagebrush steppe land­ scape. Biogeochemistry 7: 11-31. hyphallength would decline during a drought CALDWELL, M. M., AND J. H. RiCHARDS. 1989. Hydraulic year. lift: water efflux from upper roots improves effec­ Hyphal lengths appeared to be quite tiveness of water uptake by deep roots. Oecologia responsive to precipitation patterns within a 79: 1-5. CALDWELL, M. M., J. H. RICHARDS, AND W BEYSCHLAG. year. Increases in hyphallengths in the SPling 1991. Hydraulic lift: ecological implications ofwater of 1989 were associated with spring precipita­ effiux from roots. Pages 423-436 in D. Atkinson, ed., tion. I:Iyphallengths declined during the dry Plant root growth: an ecological perspective. summer period and were st~ple during fall as Blackwell SCientific Publications. CHAPIN, F. S., III, A. J. BLOOM, C. B. FIELD, AND R H. both sites began receiving precipitation again. WARING. 1987. Plant responses to multiple environ­ During 1990 hyphallengths ~ere stable dur­ mental stresses. Bioscience 37: 49-57. ing spring and declined during summer. In CHILVERS, M. T., AND J. F. DAFT. 1982. Effect oflow tem­ late summer and fall of 1~90; the ATV site peratures on development of the vesicular-arbuscu­ received precipitation an'd hyphallengths lar mycorrhizal association between Glomus caledo­ nium and Allium cepa. Transactions of the British - increased (Figs. 5, 6). However, the ATT site Mycological Society 79: 153-157. did not receive significant precipitation and COMSTOCK, J. P., AND J. R EHLERINGER. 1992. Plant adap­ hyphal lengths did not increase. It appears tation in the Great Basin and Colorado Plateau. that the pattern of arbuscular colonization Great Basin Naturalist 52: 195-215. Cox, G., AND P. B. TINKER. 1976. Translocation and trans­ more closely follows precipitation and hyphal fer of nutrients in vesicular-arbuscular mycorrhizas. lengths than does either vesicular or hyphal I. The arbuscule and phosphorus transfer: a quanti­ colonization. Arbuscular colonization and tative ultrastructural study. New Phytologist 77: mycorrhizal hyphallengths are more respon­ 371-378. DAFT, M. J., M. T. CHILVERS, AND T. H. NICOLSON. 1980. sive to seasonal and yearly variations in plant Mycorrhizas of the liliiflorae. r. Morphogenesis of growth than are either vesicular or hyphal col­ Endymion nonscriptus (L.) Garcke and its mycor­ onization levels. We suggest, from a functional rhizas in nature. New Phytologist 85: 181-189. standpoint, that measurement of mycorrhizal DEPUIT, E. J., AND M. M. CALDWELL. 1973. Seasonal pat­ tern of net photosynthesis of Artemisia tridentata. hyphal length and arbuscular colonization is American Journal ofBotany 60: 426-435. more relevant than is total colonization. FERNANDEZ, O. A., AND M. M. CALDWELL. 1975. Phenology and dynamics ofroot growth ofthree cool semi-desert shrubs under field conditions. Journal of ACKNOWLEDGMENTS Ecology 63: 703-714. FITTER, A. H. 1993. Characteristics and functions of root We thank Dr. M. Allen and Dr. C. Friese for systems. Pages 3-24 in Y. Waisel, A. Eshel, and U. reviewing the manuscript. We also thank Jill Kafkafi, eds., Plant roots: the hidden half. Marcel Holderman for teclinical assistance. Dekker, Inc., NewYork, Basel, Hong Kong. GAY, P. E., P. J. GRUBB, AND H. J. HUDSON. 1982. Seasonal changes in the concentrations ofnitrogen, phospho­ LITERATURE CITED rus and potassium, and in the density ofmycorrhiza, in biennial and matrix-forming perennial species of ALLEN, M. F. 1983. Formation of vesicular-arbuscular closed chalkland turf. Journal of Ecology 70: mycorrhizae in Atriplex gardneri (Chenopodiaceae): 571-593. -, seasonal response in a cold desert. Mycologia 75: GEE, G. W, AND J. W BAUDER. 1982. Particle size analy­ 773-776. sis. Pages 383-409 in A. Klute, ed., Methods of soil ALLEN, E. B., AND M. F. ALLEN. 1986. Water relations of analysis. Part 1. Physical and mineralogical methods. 2nd edition. Agronomy No.9, Part 2. American Soil xeric grasses in the field: interactions ofmycorrhizas Association and Soil Science Society of America and competition. New Phytologist 104: 559-571. publication. Madison, Wisconsin. BARKER, R, AND C. M. McKELL. 1986. Differences in J. GIOVANNETTI, M. 1985. Seasonal variations of vesicular­ big sagebrush (Artemisia tridentata) plant stature arbuscular mycorrhizas and endogonaceous spores along soil-water gradients: genetic components. in a maritime sand dune. Transactions of the British Journal ofRange Management 39: 147-151. Mycological Society 84: 679-684. BETHLENFALVAY, G. J., AND R N. AMES. 1987. Comparison HIRREL, M. C., H. MEHRAVARAN, AND J. W GERDEMANN. of two methods for quantifying extraradical myceli­ 1978. Vesicular-arbuscular mycorrhizae in the um of vesicular-arbuscular mycorrhizal fungi. Soil Chenopodiaceae and Cruciferae: Do they occur? Science Society ofAmerica Journal 51: 834-837. Canadian Journal ofBotany 56: 2813-2817. BRUNDRETT, M. C., AND B. KENDRICK. 1988. The mycor­ INGHAM, E. R, AND D. A. KLEIN. 1984. Soil fungi: rela­ rhizal status, root anatomy, and phenology of plants tionships between hyphal activity and staining with in a sugar maple forest. Canadian Journal of Botany fluorescein diacetate. Soil Biology Biochemistry 16: 66: 1153-1173. 273-278. 300 GREAT BASIN NATURALIST [Volume 54

NELSON, D. W, AND L. E. SOMMERS. 1982. Total carbon, READ, D. J., H. K. KOUCHEKI, AND J. HODGSON. 1976. organic carbon, and organic matter. Pages 539-577 Vesicular"arbuscular mycorrhiza in natural vegeta­ in A. L. Page, R. H. Miller, and D. R. Keeney; eds., tion systems.!. The occurrence of infection. New Methods ofsoil analysis. Part 2. Chemical and micro" Phytologist 77: 641-653. . biological properties. 2nd edition. Agronomy No.9, RICHARDS, J. H., AND M. M. CALDWELL. 1987. Hydraulic Part 2. American Soil Association and Soil Science lift: substantial nocturnal water transport between Society ofAmerica publication. Madison, Wisconsin. soil layers by Artemisia tridentata roots. Oecologia OLSEN, S. R., AND L. E. SOMMERS. 1982. Phosphorus. 73: 486"489. Pages 403-427 in A. L. Page, R. H. Miller, and D. R. WEST, N. E. 1991. Nutrient cycling in soils of semiarid Keeney, eds., Methods of soil analysis. Part 2. and arid regions. Pages 295-332 in J. Skujins, ed., Chemical and microbiological properties. 2nd edi­ Semiarid lands and deserts: soil resource and recla­ tion. Agronomy No.9, Part 2. American Soil Associ­ mation. Marcel Dekker Inc., NewYork. ation.and Soil Science Society ofAmerica publication. WILCOX, H. E. 1993. Mycorrhizae. Pages 731-767 in Y. Madison, Wisconsin. Waisel, A. Eshel, and U. Kafkafi, eds., Plant roots: PASSIOURA, J. B. 1988. Water transport in and to roots. the hidden half. Marcel Dekker, Inc., New York, Annual review. Plant Physiologist 39: 245-265. Basei, Hong Kong. PHILLIPS, J. M., AND D. S. HAYMAN. 1970. Improved pro" WINWARD, A. H. 1980. Taxonomy and ecology of sage­ cedures for clearing roots and staining parasitic and brush in Oregon. Oregon State University Station vesicular-arbusclllar mycorrhizal fungi for rapid Bulletin 642. assessment of infection. 'fransactions of the British Mycological Society 55: 158-161 Received 13July 1993 RABATIN, S. C. 1919. Seasonal and edaphic variation in Accepted 20 April 1994 vesicular-arbuscular mycorrhizal infection of grasses by Glomus tenuis. New Phytologist 83: 95-102. Great Basin Naturalist 54(4), © 1994, pp. 301--312

SOIL AND VEGETATION DEVELOPMENT IN AN ABANDONED SHEEP CORRAL ON DEGRADED SUBALPINE RANGELAND

James O. Klemmedson1 and Arthur R. Tiedemann2

ABSTRACT.-Vegetation and soils inside and outside an abandoned sheep corral on degraded subalpine range ofthe Wasatch Plateau were studied to determine the influence of approximately 37 years' use ofthe corral on soil and plant development. Vegetal and surface cover were estimated. Herbage, litter, and soils were sampled inside and outside the corral and analyzed for Corg> N, P, and S. Soil pH, bulk density, and C03-C also were measured. Storage (mass/unit area) ofCorg, N, P, and S was determined for each component. Yield and vegetal composition were significantly affected inside the corral boundary. Herbage yield was 2.2 times greater, litter mass 16 times greater, foliar cover of grasses 2 times greater, and forb cover 70% lower inside than outside the corral. Cover ofmeadow barley (Hordeum brachyantherum), a component ofthe predisturbance vegetation oftlle Wasatch Plateau, was nearly 12 times greater inside than outside the corral. These and ollier vegetal and cover differences reflect inside-outside differences in concentration, storage, and availability ofsoil Corg, N, P, and S. Concentrations ofCorg and total and available N, P, and S were greater in the surface 5 cm ofsoil inside the corral. Available P inside the corral was much higher in all soil layers. Because ofbulk density dif" ferences, storage was greater inside the corral only for Corg and N at 0-5 cm and for P at 5-15 cm. Lower soil pH inside the corral appears related to soil P distribution and C03-C storage. Results suggest a need to reexamine earlier conclu­ sions that tall forbs are the climax dominants ofthe Wasatch summer range.

Key words: summer range; soil Corg, N, P, S, COs-C, pH, and bulk density; plant composition and cover; biomass yield; litter.

After 35 years ofdestructive grazing by cattle 1964, Intermountain Research Station, and sheep in the late 1800s, the subalpine range Ogden, Utah, unpublished data). Our observa­ ofthe Wasatch Plateau east ofEphraim, Utah, tions suggest that soil and vegetal conditions was in extremely poor condition (Reynolds have essentially stabilized since Ellisons last 1911, Sampson and Weyl 1918, Sampson observations in the mid-1950s. We believe the 1919). Erosion and alteration of vegetal cover slow rate of succession and range improve­ reached such severe proportions that most of ment in the Wasatch subalpine since then is the soil A horizon was lost to erosion, and directly attributable to extreme amounts of mud-rock floods were a common occurrence soil loss and relatively low fertility ofsoils that in the canyons leading to valleys and settle­ remained after the period ofdegradation. Based ments at the base of the Wasatch Front on examination of numerous soil profiles on (Reynolds 1911, Croft 1961). In some places the plateau and those of similar soils else­ only subsoils remained when control of graz~ where, we believe at least 50%, and possibly ing was finally achieved with establishment of as much as 80 or 90%, of the A horizon was the Manti National Forest in 1903 (Reynolds lost from this summer range via accelerated 1911, Sampson and Weyl1918, Ellison 1949). erosion. Such a loss would certainly remove a Although condition of the range improved large portion of the soil's organic matter and steadily over the next several decades, most of nutrient capital and significantly alter produc~ the summer range was still unstable in 1950, tive potential. and accelerated erosion was continuing but at In recent years we have pursued this greatly reduced rates (Ellison 1954, Meetiwig hypothesis with several studies. This paper 1960). reports the results ofa fortuitous observational Under moderate grazing secondary sUcces­ study designed to demonstrate the effect of sion occurred from 1903 to about 1940 when it organic matter and nutrient additions over slowed perceptibly (Ellison 1954). Since then time on development ofsoils and vegetation of succession has been extremely slow (Johnson the Wasatch subalpine range. During field

lSehool of Renewahle Natural Resources. University ofArizona, Tucson, Arizona 8572l. 2Paeifie Northwest Research Station, La Grande, Oregon 97850.

301 302 GREAT BASIN NATURALIST [Volume 54

studies in 1988, we happened upon an aban­ ied Watersheds A and B of the Great Basin doned and dilapidated sheep corral that obvi­ Experimental Range established by Dr. Arthur ously had not been used in many years (Fig. 1). W Sampson in 1912 (Sampson and Wey11918, No remains ofmanure were present inside the Meeuwig 1960). Cherry Flat is typical of the COITal; vegetation and litter development were crest of the Wasatch Plateau, which is about advanced and perennial grasses were abun­ 3150 m elevation. The plateau is long, narrow, dant (Fig 2A). The contrast with vegetation and oriented approximately north and south and litter outside ofwhat remained ofthe cor­ with riblike ridges extending east and west. ral fence was striking (Fig. 2B). The top ofthe plateau is gently rolling to nearly The corral offered an opportunity to docu­ level. Average annual precipitation is about 840 ment effects ofuse ofthe COITal (1936-73) and mm; two-thirds of this falls as snow between inputs oforganic matter and nutrients via sheep November and April. Precipitation averages manure during its use to soil and vegetal 173 mm during the summer months (June development inside the corral boundary. through September) but varies considerably. Mean annual temperature is about O°C (Ellison STUDYA.REA 1954). In the vicinity ofthe COITal, Cherry Flat has The Buck Ridge COITal study site (39° 15'N, a gentle 2% slope to the east; microtopography 111°26'W) is located about 18 km east of is smooth. Soil parent materials are of the Manti, Utah, on Cherry Flat adjacent to Buck Flagstaff Formation (Stanley and Collinson Ridge Road and about 1.6 km east of Skyline 1979) that crop out over about 7200 km2 in Drive. This location is 5 km south of the central Utah (Schreiber 1988). Dominant lith­ Alpine Station and the well-known and stud- ology is freshwater lacustrine limestone and

'''\

'""'""'".\

Fig. 1. Remains ofBuck Ridge corral as it appeared in 1989. 1994] SOIL AND VEGETATION DEVELOPMENT ON SUBALPINE RANGE 303

Fig. 2. Close-up view ofvegetation and groundcover inside (A) and outside (B) Buck Ridge corral in 1989.

calcareous shales with minor interbeds of sand­ exposures of east-west ridges and dot the stone, oil shale, conglomerate, gypsum, and plateau landscape. Because there were no volcanic ash (Weber 1964, Schreiber 1988). remnants of the original pristine vegetation Soils in this region of the plateau are mostly (Ellison 1949, 1954), opinions differ regarding fine, mixed Argic Cryoborolls, but lithic, its exact character. Ellison (1954) describes the pachic, and vertic Cryoborolls also are present. original plant community as mixed-upland They are shallow to moderately deep; the sub­ herb dominated by tall forbs, while Sampson soils are silty clays or clay loams. Thickness of (1919) considered wheatgrasses to be the pri­ the A horizon averages about 4 cm; the B hori­ mary species of the herbaceous climax (i.e., zon averages approximately 52 cm thickness. what he referred to as summer range). Based on typical profile descriptions (H. K. Based on file records and discussion with Swenson, Soil Conservation Service, Boise, former permittees, we determined Buck Ridge Idaho, personal communication), these rela­ corral was built and first used in 1936. It was tive horizon thicknesses suggest that much of last used about 1973 (Ed Shoppe, Manti-LaSal the original A horizon was lost by wind and National Forest, Ephraim, Utah, personal water erosion following the period of unre­ communication). Hence, the corral was used stricted grazing prior to 1903. annually by sheep for about 37 years. During Vegetation ofthe Wasatch Plateau is chiefly this period undetermined and variable herbaceous, but small patches of Engelmann amounts of organic matter and nutrients were spruce (Picea engelmannii) and subalpine fir added annually via dung and mine, depending (Abies lasiocarpa) occupy steep northerly on the number and size of bands using the 304 GREAT BASIN NATURALIST [Volume 54 corral and frequency of use. For the past 10 196;2.). Total P was determined in plant materi­ years, the Bilck Ridge allotment, comprising al using the vanado~molybdo-phosphoricyel~ 4235 ac, has been grazed by sheep at the rate low color method after dry-ashing (Jackson of3.3 ac/AUM from 1 July to 30 September. 1958) and in soils using ascorbic acid color development (Olsen and Sommers 1982) fol­ METHODS lowing hydroflUOriC acid digestion (Bowman 1988). Available nutrients in soils were deter­ As a basis for conducting this observational mined as follows: P ilsing ascorbic acid color study, we first assured ourselves that areas development follOwing 0.5 M sodium bicar­ inside and outside the corral boundary were bonate extraction (Olsen and Sommers 1982), initially alike in every respect and that site N by steam distillation of 2 N KCI extracts characteristics had no bearing on the precise (Keeney and Nelson 1982), and S with 1:1 location at the corral at the outset or on the water extracts, followed by ion chromatogra­ findings. There were no differences in topog~ phy (Dick and Tabatabai 1979). raphy (or microtopography) and no evidence Tests of significance for difference between that other state factors (climate, biotic factor, inside and outside values for all variables stud­ parent material) differed within the small ied were carried out with the t test. We recog­ study site (about 0.75 ha) ofthe corral area. To nize the desirability of replicating the inside~ assess the differential effect of nutrient addi~ outside corral comparison. Unfortunately, that tions inside and outside the corral, we sampled was not a design feature we could control; vegetation, litter, and soil in midsummer. other abandoned corrals-even less than 37 Moisture conditions were dry at the time and years age-simply do not exist on this summer little grazing had occurred inside or outside the range. corral. Present condition of the corral fence (Fig. 1) indicates that sheep have had near RESULTS equal access to both sides ofthe corral bound~ ary for many years, but there is no record of Vegetation how long fence cross rails have been down. Obvious visual differences in vegetation, Cover by species (foliar projection), litter, litter; and soil surface conditions inside and soil, and rock were estimated in 10 randomly outside the old corral (Figs. 2A, 2B) were con­ located 0.5~m2 plots inside and outside the firmed in the data (Table 1). Herbage yield in­ corral (within 10 m of the corral boundary). side the corral was 2.2 times greater than out­ Herbage and litter were harvested in the same side. Although total herbage cover inside and plots, oven-dried (70°C), and weighed. Six outside the corral was the same (65%), grasses randomly located soil pits were sampled comprised a much greater percentage offoliar inside and outside the corral. At each pit, soil cover than did forbs inside than outside the cores (5.197 cm dia.) were collected from the corral. Three perennial grasses (Agropyron 0-5~, 5-15-, and 15-30~cm layers. Plants and trachycaulum, Hordeum brachyantherum, and litterWere ground to pass through a 0.425~:tnrn . Stipa lettermani) dominated vegetal cover sieve. Soils were air-dried, sieved to remove inside the corral (Table 2); outside the corral the >2-mm fraction; and then ground to pass A. trachycaulum and S. lettermani were equal­ through a 0.150~mm sieve. ly important, but H. brachyantherum was Plant and soil samples were analyzed for unimportant. total N by semi-micro-Kjeldahl (Bremner and Forbs were represented by 7 species inside Mulvaney 1982) and total S by dry combustion the corral and 11 species outside (Table 2). (Tiedemann and Anderson 1971) in a LECO Taraxacum officinale and Achillea millefolium high~frequencyinduction furnace (LECO were the dominant forbs inside and outside Corp., St. Joseph, Michigan). Plant and soil the corral, but their cover outside was much samples were analyzed for total C by dry com­ greater than inside. No other forb species con­ bustion (Nelson and Sommers 1982) in the stituted more than 4% of the herbage compo­ LECO high-frequency induction furnace. sition. Organic C (Cor) of soils wa~ determined by Soil surface protection by litter differed correcting totifC for carbonate-C as deter­ markedly inside and outside the corral. Mass mined by a gasometric method (Dreimanis oflitter inside the corral was 16 times greater 1994] SOIL AND VEGETATION DEVELOPMENT ON SUBALPINE RANGE 305

TABLE 1. Influence oflong"term corral effects on vegeta- TABLE 2. Percentage of species composition (by foliar tion, litter, and soil surface characteristics at Buck Ridge. cover) of vegetation inside and outside Buck Ridge corral. Component Difference Percentage and attribute Inside Outside sign. atP < Inside Outside Herbage yield (g m-2) 139 ± 21" 64± 9 .005 GRASSES Litter mass (g m-2) 312:t78 19 ± 2 .005 Agropyron trachycaulum 8.2 10.3 Alopecurus pratensis 1.6 Foliar cover (%) Bromus carinatus 4.0 Hordeum brachyantherum 35.6 3.0 Grasses 55:t 5 28 ±5 .001 Poa pratensis 0.7 Stipa columbiana 0.4 Forbs 11 ± 3 37± 8 .05 Stipa lettermani 30.4 26.5 Total grasses 78.6 42.1 Total 66 ± 3 65:t 7 NS FORBS Basal cover (%) Achillea millefolium 8.8 18.0 Androsace septentrionalis 0.3 0.3 Litter 71 ± 4 18 ± 4 .001 Artemisia ludoviciana v.imcompta 2.7 Bare ground Asterfoliaceus v. canbyi M + rock 3:t2 25 ±4 .005 Cymopteris lemmonii 0.2 aMean ± standard error; n = 10. Descurania richardsonii 0.9 0.1 Erigeron ursinus 2.6 Gilia aggregata 0.2 than outside, while cover oflitter was 4 times Lesquerella utahensis 2.1 greater (Table 1). This is consistent with the Polygonum glandulosa 0.2 12-fold difference in bare ground between Ranunculus inamoenus 0.3 inside and outside locations. Only 2% of the Rumex mexicanus 3.5 soil surface was bare inside the corral. Taraxacum officinale 7.4 29.2 Viola nuttallii v. nuttallii 0.1 NUTRIENTS Total forbs 21.4 57.9

Concentrations ofall nutrients studied were influenced by dung and urine accumulation DISCUSSION AND CONCLUSIONS inside the corral (Table 3). Nitrogen concentra~ tion was higher in the herbage, litter, and 0=5­ Development of vegetation and stabiliza­ em soil layer, but lower in the 15-30-cm soil tion of the soil surface at Buck Ridge corral layer inside than outside the corral. Concentra­ were indeed striking considering the slow pace of secondary succession of the Wasatch tion ofCorg was parallel to that ofN for the litter and soil layers, while P concentration was higher summer range from 1903 to 1940 (Ellison 1954, inside than outside only for litter and the 0~5~ Meeuwig 1960) and the apparent lack oftrend em soil layer (Table 3). Concentration of S was since 1940. Values fOr herbage production, litter higher inside than outside only for the upper mass, and cover data portray control ofthe soil soil layer. surface inside the corral and are in marked Storage (mass/unit area) ofall four nutrients contrast to conditions outside the corral and was significantly greater inside than outside the surrounding summer range, which is still the corral for herbage and litter components relatively unstable and subject to accelerated (Table 4). Amounts of Corg were greater inside erosion. We believe the vegetation trend ob­ the corral in the surface soil, but lower in the served inside the corral has occurred within a 15-30-cm soil layer. Storage of N was greater relatively short time~no more than .20 years, inside the corral in the O-$-cm soil layer, assuming vegetal development did not com­ while storage of P was greater inside the cor~ mence until after abandonment of the corral. ral in the 5=15~cm soil layer. This is not to say that ephemerals did not Availability of P was much higher inside occupy the corral annually between periods of than outside the corral in all soil layers (Table use, only to be eliminated during use. 5). Availability of Nand S was significantly The large changes in vegetation and soil higher (P < .10) inside the corr1l.l only in the surface conditions are consistent with changes 0--5-cm soil layer. in nutrient status ofthe soil~plant~littersystem 306 GREAT BASIN NATURALIST [Volume 54

TABLE 3. conceAtration of Corg, N, P, and Sin components ofthe soil-plant-litter system inside and outside Buck Ridge corral.. I .' Difference Nutrient I Component Inside Outside sign. at P < ______gkg-1 ___ ------Herbage 447.4 ± 7.1" 433.0 ± 4.2 NS Litter 420.0 ± 8.3 448.9 ± 7.2 .025 Soil, 0-5cm 134.2 ± 17.6 45.6 ± 2.2 .001 5-15cm 34.8 ± 2.7 39.5 ± 1.4 NS 15--30 cm 21.8 ± 2.5 36.4 ± 2.0 .005

N Herbage 21.75 ±·0.75 15.50 ± 0.82 .001 Litter 19.48 ± 0.83 11.77 ± 0.77 .001 Soil, 0-5cm 14.05 ± 1.95 3.39 ± 0.21 .001 5-15cm 2.99 ± 0.23 3.05 ± 0.10· NS 15--30 cm 1.89 ± 0.17 2.73 ± 0.17 .005 p Herbage 2.23 ± 0.10 2.22 ± 0.21 NS Litter 1.99 ± 0.12 1.47 ± 0.08 .001 Soil, 0-5cm 2.36 ± 0.22 1.61 ± 0.06 .01 5-15cm 1.70 ± 0.13 1.44 ± 0.08 NS 15--30 cm 1.36 ± 0.13 1.39 ± 0.09 NS

S Herbage 1.29 ± 0.04 1.22 ± 0.04 NS Litter 1.10 ± 0.07 0.96 ± 0.10 .05 Soil, 0-5cm 1.42 ± 0.18 0.88 ± 0.08 .025 5-15cm 0.69 ± 0.22 0.74 ± 0.11 NS 15--30 cm 0.68 ± 0.11 0.71 ± 0.05 NS aMean ± standard error; n = 6. over the life ofthe corral and subsequent to its should be expected in components ofthe Buck abandonment (Crocker and Major 1955, Olson Ridge soil-plant~litter system, especially be­ 1958, Blackmore et al. 1990). Unfortunately, tween Corg and N because soil N is almost mU:ch of the corral history (actual herd use, entirely organic (i.e., about 98%). On the other inputs of dung and urine, and character of hand, because of certain dissimilarities in the vegetation that occupied the corral between chemistry of these four nutrients and differ­ periods ofuse by sheep) was not documented. ences among them in physiological separation However, changes in concentration and accu~ into dung and urine pathways (C and P entirely mulation of nutrients inside the corral seem via dung, Nand S predominately via urine; reasonable, based on what might be expected Floate 1970, O'Connor 1981, Barrow 1987, from traditional use of a subalpine corral by Sagger, MacKay et al. 1990), certain differences sheep for 31 years, experience from other in nutrient accumul~tion patterns can be grazed systems (Blackmore et al. 1990, Scholes expected. 1990), and an understanding of the chemistry Close association of Corg and N was appar­ ofthe elements studied here. We can be confi­ ent even in the 15-30-cm soil layer where dent that concentrations of soil Corg and N in~ concentration of Corg and N and amount of side the corral have declined since abandon­ Corg were lower inside than outside the corral. ment and change in the biotic factor (Jenny Such an unexpected difference at this depth 1941). However, whether a new steady state has lacks explanation, certainly none related to been reached yet is conjectural (Jenny 1941, corral effects. It is not consistent with the Tiedemann and Klemmedson 1986), even large differences in Corg and N in the litter though observed concentrations of soil Corg and 0-5~cm soil layer where effects ofthe cor~ and N are within the range for comparable ral would be most expected. Parent material undisturbed soils (Retzer 1956, Youngberg seems the most likely cause of this difference, and Dyrness 1964). but samples from the study site were uniform Because C, N, P, and S are ubiquitous in soil in C03~C and varied randomly in N, P, and S. organic matter and its precursors (Stevenson However, limestone and shales are noted for 1986), close association among these elements spatial variation, even within very short 1994] SOIL AND VEGETATION DEVELOPMENT ON SUBALPINE RANGE 307

TABLE 4. Storage ofCorg, N, P, and S in components ofthe soil-pJant-litter system inside and outside Buck Ridge corral. Difference Nutrient Component Inside Outside sign. atP < ------kgm-2 ------______Herbage 0.064 ± 0.007a 0.028 ± 0.004 .001 Litter 0.129 ± 0.032 0.009 ± 0.001 .005 Soil, 0-5cm 3.46 ± 0.17 1.84 ± 0.13 .001 5-15cm 4.44 ± 0.61 3.94 ± 0.25 NS 15--30 cm 4.30 ± 0.62 6.44 ± 0.63 .05 Soil totalb 12.19 ± 1.29 12.22 ± 0.80 NS Total system 12.34 ± 1.27 12.25 ± 0.80 NS

-_------gm-2 ------______N Herbage 3.17 ± 0.27 1.03 ± 0.16 .001 Litter 6.23 ± 1.68 0.23 ± 0.04 .001 Soil, 0-5cm 348 ± 16 138 ± 10 .001 5-15cm 379 ± 50 304 ± 17 NS 15--30 cm 370 ± 46 485 ± 50 NS Soil total 10'79 ± 96 927 ± 62 NS Total system H15 ± 105 928 ± 152 NS p Herbage 0.33 ± 0.04 0.15 ± 0.03 .001 Litter 0.66 ± 0.18 0.03 ± <0.01 .005 Soil, 0--5 cm 63 ±4 64 ± 3 NS 5-15cm 216 ± 30 145 ± 14 .10 15--30 cm 231 ± 43 250 ± 32 NS Soil total 551 ± 70 460 ±44 NS Total system 552±70 461 ± 44 NS

S Herbage 0.19 ± 0.02 0.08 ± 0.01 .001 Litter 0.37 ± O.H 0.02 ± <0.01 .01 Soil, 0-5cm 37±2 35±3 NS 5-15cm 90 ± 16 77± 14 NS 15--30 cm 134 ± 25 128 ± 18 NS Soil total 261 ± 42 241 ± 33 NS Total system 262 ± 42 241 ± 33 NS aMean ± standard error; n = 10 for herbage and litter, 6 for soil components. b0-30cm distances (C. F. Lohrengel, Department of control levels (Doak 1952, During et al. 1973, Geology, Snow College, Ephraim, Utah, per­ Haynes and Williams 1992). Indeed the initial sonal communication). Of 23 rock samples impact ofdecomposition ofmost plant materials from the near vicinity (4-km radius) classified is an increase in bulk pH (Williams and Gray by Schreiber (1988), P concentration ranged 1974). However, products of organic decay are 31-fold, with a· C.Y. of L46. The sample high~ predominantly acid; hence, acidification even­ est in:P content, an organic-rich shale, burned tually dominates. Those horizons or soil layers under a match flame. that contain the products ofprimary decompo­ Phosphorus is relatively immobile but should sition, in this case the litter and 0-5-cm layer accumulate in soils over time where P inputs (Table 6), will show the greatest acidity (Swift exceed removal in grazed herbage (Sagger, et al. 19'/9) and a tendency for enhanced solu­ Hedley et al. 1990). This would characterize bility and mobility of P. Significantly lower the situation in Buck Ridge corral with the carbonate-C of soil inside than outside the large inputs of animal excreta from 1936 to corral (Table 6) manifests increased soil acidity 1973. Moreover, decomposition ofthis material inside the corral. James Clayton (personal should facilitate P mobility, especially after pul­ communication, Intermountain Research dif~ verization by hoofaction (Bromfield and Jones Station, Boise, Idaho) suggests the C03-C 1970). Although urine and dung hydrolyze ference inside and outside the corral is reason­ rapidly causing NH4 to accumulate and pH to able, based on estimated H+ supplied by rise, nitrification quickly takes over and, in the nitrification of urea and organic matter case of urine, within days pH will drop below decomposition over a period of37 years. 308 GREAT BASIN NATURALIST [Volume 54

TABLE 5. Concentration ofavailable soil N, :e and S inside and outside Buck Ridge corral. N P S Diff. Diff. Diff. sign. sign. sign. Soil Layer Inside Outside P< Inside Outside P< Inside Outside P< ----- mg kg-I -_------mg kg-I ------mg kg-I ---- 0-5cm 105 ± 26" 52 ± 11 .10 158 ± 19 57::t 6 .001 50 ± 13 22 ±3 .10 5--15 cm 29 ± 7 20± 3 NS 142 ± 16 22±4 .05 24±2 22± 1 NS 15--30 cm 13 ± 3 13 ± 2 NS 70 ± 11 10 ± 3 .001 15 ± <1 14 ± 1 NS 3Mean ± standard error; n = 6.

The P distribution pattern described here Heavy clay subsoils ofthis site should compact is similar to that found by Sagger, MacKay et readily. Sommerfeldt and Chang (1985) found al. (1990), who closely predicted observed P that 10ng4erm manure treatments reduced accumulation in soil of sheep pastures. In bulk density ofthe upper 15 cm ofa cultivated areas where sheep camped, 85~90% ofP accu~ soil by as much as 39%. mulated in the upper 15 cm of soil was Increased availability of nutrients in the accounted for by animal waste. Williams and upper 5 cm ofsoil ofthe corral soil is associat­ Haynes (1992) noted significant increases ofP ed with higher concentration of nutrients and in the top 20 cm of soil in pastures grazed by the large pool of organic matter in that layer. sheep for 38 years and treated with super­ Carbon/element ratios of all soil-plant compo" phosphate. nents (Table 7) indicate that conditions gener­ Nitrogen and S losses from the corral soil~ ally more favorable for net mineralization of N plant system could have been large. Nitrogen and P (Stevenson 1986) prevailed inside than may be lost by volatilization of NH3, leaching outside the corral at the time of sampling. and surface runoff of N03, or denitrification Greater availability ofP in all soil layers inside under appropriate conditions (Ball et al. 1979, the corral also can be associated with pH in a Floate 1981, O'Connor 1981), while S04 may range that one might expect maximum avail­ be lost by surface runoff and leaching, depend­ ability of the labile inorganic P fraction ing on S04 retention capacity of soils (Sagge~ (Stevenson 1986). Coupled with tllis is low Hedley et al. 1990). Williams and Haynes mobility of p, in contrast to Nand S, which (1992) assumed most ofthe S loss they observed allows P to be retained in place. (4~73%) was due to leaching. Thus, whether Since abandonment of the corral, organic Buck Ridge corral was devoid of vegetation matter would have continued to accumulate, during much of the year and hence subject to butfrom anew source, i.e., autotrophic produc. leaching and runoff, or whether its use was tion of vegetation that presumably developed intermittent so as to permit vigorous growth of soon after corral abandonment. Significant im" ephemerals and uptake of available nutrients, port of new nutrients since abandonment is we would not expect mineralized N and S to unlikely. Almost all nutrients in post-abandon­ accumulate in the soil profile. ment crops ofherbage would have been recy­ Similarity in P and S accumulation in the cled from the soil. Presumably, the level of O~-cm soil layer, in view ofhigher concentra­ herbage production inside the corral has tions of these nutrients, is attributed to lower exceeded that outside almost since abandon­ bulk density of the upper soil layer inside the ment owing to higher fertility status of soils con-al (Table 6)_ By contrast, bulk density of inside the corral. The present condition ofthe the 5~15-cm layer was significantly greater corral fence (Fig. 1) would indicate that it has inside than outside the corral. These opposite not been a barrier to sheep for many years. trends in adjacent soil layers appear to be due Hence, differential grazing probably has to compaction of the entire upper 15 cm dur­ played a minor role in vegetal differences ing 37 years of use ofthe corral by sheep, fol~ inside and outside the corral boundary. lowed by amelioration of this effect in the Although we have emphasized the role of absence of trampling after the corral was nutrients in the observed changes, we cannot abandoned, especially in the top 5 cm where dismiss the possibility that improvement in organic matter was concentrated (Tables 3, 4). moisture-holding capacity of surface soils 1994] SOIL AND VEGETATION DEVELOPMENT ON SUBALPINE RANGE 309

TABLE 6. Effects ofinside l!lld outside positions ofBuck Ridge corral on bulk density, pH, and carbonate-C ofsoil layers. - - Soil Difference Soil property layer (cm) Inside Outside sign. atP < - Bulk density (mg m-S) 0-5 0.59 ± 0.08" 0.84::!: 0.06 .05 5-15 1.26 ± 0.04 1.10 ± 0.02 .005 15-30 1.39 ± 0.06 1.26 ± 0.10 NS pH 0-5 6.58 ± 0.11 7.23 ± 0.13 .005 5-15 7.15 ± 0.06 7.28 ± 0.09 NS 15-30 7.17 ± 0.06 7.37 ± 0.09 .10

Carbonate-C concentration (g kg-I) 0-5 5.20 ± 0.82 8.92 ± 1.87 .10 5-15 5.35 ± 1.~ 9.23 ± 2.10 NS 15-30 4.13 ± 0.94 10.63 ± 2.19 .05

amount (kg m-2) 0-5 0.15 ± 0.04 0.38 ± 0.10 .10 5-15 0.65 ± 0.13 0.74 ± 0.17 NS 15-30 0.77 ± 0.15 1.80 ± 0.36 .025 Soil 1.57 ± 0.25 3.05 ± 0.54 .05 aMean ± standard error, n = 6.

TABLE 7. Carbon-element ratios of soil~plant~littercomponents for inside and outside positons ofBuck Ridge corral. Difference Ratio Component Inside Outside sign. atP < CIN Herbage 19.6 ± 0.8" 30.'/ ± 2.6 .05 Litter 22.4 ± 1.0 37.2 ± 3.6 .05 Soil, 0-5cm 10.0 ± 0.2 13.3 ± 0.6 .001 5-15 cm 11.7 ± 0.5 13.0 ±0.5 NS 15-30cm 11.5 ± 0.6 13.4 ± 0.6 .05

C/P Herbage 210 ± 15 241 ± 44 NS Litter 241 ± 23 308 ± 15 .05 Soil, 0-5cm 56 ±5 29 ±2 .001 5-15cm 21 ± 1 28:t 2 NS 15-30 cm 16 ± 1 27±2 .025

CIS Herbage 346 ± 16 368 ± 17 NS Litter 406 ± 43 439 ± 55 NS Soil, 0-5cm 95± 5 54±5 .001 5-15cm 53 ± 5 61 ± 11 NS 15-30cm 34::t 3 53 ±5 .01 aMean ± sbndard error; n = 10 for herbage and litter, 6 for soil components. inside the corral also may have influenced the tually static in the last 40-odd years. Although successional trend following abandonment. Elk Knoll had been previously grazed, it has Comparison of basal cover data for Buck been protected from grazing, except for wildlife, Ridge corral (inside and outside) with that from 1903 (:Ellison 1954) to the present. portraying conditions in 1946 for six "relic nat­ According to Ellison, the natural areas, which ural areas" (lIld four stands on Elk Knoll (Elk he claimed had never.been grazed by domes­ Knoll Research NaturaI Area), as described by tic livestock or had been grazed only lightly Ellison (1954), is revealing (Table·8). Similarity for many years, provided a partial description between cover oflitter, bare ground, and rock of pristine vegetation in the Wasatch sub­ at Elk Knoll (3.2 km west~northwest of Buck alpine at that time. Ridge corral) in 1946 and in 1989 and what we Vegetal covet data (Table 1) demonstrate a found outside Buck Ridge corral supports ob~ marked trend toward perennial grasses inside servations that successional trend has been vir- the corral. We interpret this as an upward 310 GREAT BASIN NATURALIST [Volume 54

TABLE 8. Comparison ofcover data for Buck Ridge cor­ this case where livestockmen controlled the ral with Ellison's data for "relic natural areas" and sites at input of manure and associated effects for 37 Elk: Knoll. years, ecosystem development was essentially Litter Bare ground one"sided; soil development advanced rapidly Location cover + rock for 37 years before sheep use of the corral ------% ~------ceased and development of vegetation was Buck Ridge corral allowed to proceed. The fact that development Inside corral 71 3 in vegetation, litter, and soil surface conditions Outside 18 25 has advanced so far in just 20 years, far out­ "Relic natural areas"· 11-20 11-50 pacing comparable development outside the corral, even in Ellison's "relic natural areas" Elk: Knoll leads us to conclude that soil fertility has be~n 1946' 26-31 29-37 im~ 1989b 27-49 17--39 a key factor controlling succession and provement ofthe Wasatch summer range.. ·See Ellison (1954). bKlemmedson and TIedemann, unpublished data ACKNOWLEDGMENTS trend in succession; vegetation changes are accompanied by greater herbage production, This research was supported by Grants 85" increased litter mass and covel; and stability of CRSR-2-2717 and 91-38300~6156, Range the soil surface. These characteristics have Research Grant Program, USDA-CSRS. The been commonly associated with improvement authors gratefully acknowledge the Shrub toward high range (ecological) condition Improvement and Revegetation Project, (Laurenroth and Laycock 1989). Intermountain Research Station Provo Utah Interestingly, herbage composition inside the for providing facilities for this ;esearch; field corral is in marked contrast to that described assistance of Gary Jorgensen, range techni­ by Ellison (1954) for his six relic natural areas. cian, Intermountain Research Station, and Dr. He said that "one of the most striking things Clyde Blauer, professor, Snow College, both of about the natural areas is the abundance of Ephraim, Utah; and laboratory analysis by perennial forbs"; they constituted 70-88% of Justine McNeil, University ofArizona. Thanks the vegetation in relic natural areas 'in 1946. are also due to Drs. James Clayton, Fred The trend toward perennial grasses (79% of Provenza, and E. Lamar Smith for reviews of total foliar cover) we have observed inside the the manuscript. corral is quite the opposite ofEllison's compo" sition data and corresponds more to vegetation LITERATURE CITED development of the summer range described by Sampson (1919). The increase in Hordeum BALL, R., D. R. KEENEY, P. W THEOBALD, AND P. NES. 1979. Nitrogen balance in urine-affected areas of a brachyantherum also suggests an upward New Zealand pasture. Agronomy Journal 71: trend. Both Ellison (1954) and Sampson (1919) 309-314. noted the presence ofH. nodosum, a synonym BARROW, N. J. 1987. Return ofnutrients by animals. Pages misapplied to H. brachyantherum (Hitchcock 181-186 in R. W Snaydon, ed., Managed grasslands. 1950, Holmgren and Reveal 1966), on the Elsevier, Oxford. BLACKMORE, A. C., M. T. MENTIS, AND R. J. SCHOLES. summer range. But only Sampson (1919) dis­ 1990. The origin and extent of nutrient-enriched cussed successional status; he described H. patches within a nutrient-poor savanna in South nodosum as a shallow-tooted species that Africa. Journal ofBiogeography 17: 463-470. occupied space between bunched wheat­ BOWMAN, R. A. 1988. A rapid method to determine total grasses, the primary species of the subclitnax phosphorus in soils. Soil Science Society ofAmerica Journal 52: 1301-1304. type. He did not list this plant with types of BREMNER, J. M., AND C. S. MULVANEY. 1982. Nitrogen­ lower developmental stage. total. Pages 595-624 in A. L. Page, ed., Methods of Normally where ecosystem degradation has soil analysis. Part 2. 2nd ed. Agronomy Monograph been as severe as experienced here, with almost 9. American Society of Agronomy and Soil Science complete loss of the A horizon, we would Society ofAmerica, Madison, Wisconsin. expect successional processes in soil and vege­ BROMFIELD, S. M., AND O. L. JONES. 1910. The effect of sheep on the recycling of phosphorus in hayed-off tation to occur simultaneously (Sampson 1919, pastures. Australian Journal ofAgricultural Research Crocker and Major 1955, Olson 1958). But, in 21: 699-111. 1994] SOIL AND VEGETATION DEVELOPMENT ON SUBALPINE RANGE 311

CROCKER, R L., AND J. MAJOR. 1955. Soil development in MEEUWIG, R O. 1960. Watersheds A and B-a study of relation to vegetation and smface age at Glacier Bay, surface runoff and erosion in the subalpine zone of Alaska. Journal ofEcology 43: 427-448. central Utah. Journal ofForestry 58: 556-560. CROFT, A. R. 1967. Rainstorm debris floods. Arizona NELSON, D. W, AND L. E. SOMMERS. 1982. Total carbon, Agricultural Experiment Station, Tucson. 36 pp. organic carbon and organic matter. Pages 539-580 in DICK, W A., AND M. A. TABATABAI. 1979. Ion chromato­ A. L. Page, ed., Methods ofsoil analysis. Part 2. 2nd graphic determination of sulfate and nitrate in soils. ed. Agronomy Monographs 9. American Society of Soil Science Society ofAmerica Journal 43: 899-904. Agronomy and Soil Science Society of America, DOAK, B. W 1952. Some chemical changes in the nitroge­ Madison, Wisconsin. nous constituents of urine when voided on pasture. O'CONNOR, K. E 1981. Comments on Dr. Floate's paper Journal ofAgricultural Science 42: 16Z-171. on grazing effect by large herbivores. Pages 707-714 DREIMANIS, A. 1962. Quantitative gasometric determina­ in E E. Clark and T. RosswaIl, eds., Terrestrial nitro­ gen cycles. Ecological Bulletin (Stockholm) 33. tion of calcite and dolomite by using Chittick appa­ OLSEN, S. R., AND L. E. SOMMERS. 1982. Phosphorus. ratus. Journal ofSedimentary Petrology 32: 520-529. Pages 404-430 in A. L. Page, ed., Methods of soil DURING, C., W C. WEEDA, AND E D. DOROFAEFF. 1973. analysis. Part 2. 2nd ed. Agronomy Monographs 9. Some effects of cattle dung on soil properties, pas­ American Society of Agronomy and Soil Science ture production, and nutrient uptake. II. Influence Society ofAmerica, Madison, Wisconsin. of dung and fertilizers on sulphate sorption, pH, OLSON, J. S. 1958. Rates ofsuccession and soil changes on cation-exchange capacity, and potassium, magne­ southern Lake Michigan sand dunes. Botanical sium, calcium and nitrogen economy. New Zealand Gazette 119: 125--170. Journal ofAgricultural Research 16: 431-438. RETZER, J. L. 1956. Alpine soils of the Rocky Mountains. ELLISON, L. 1949. Establishment ofvegetation on deplet­ Journal ofSoil Science 7: 22--32. ed subalpine range as influenced by microclimate. REYNOLDS, R V. R 1911. Grazing and floods: a study of Ecological Monographs 19: 97-121. conditions in the Manti National Forest, Utah. U.S. _.__. 1954. Subalpine vegetation ofthe Wasatch Plateau, Forest Service Bulletin 91. 16 pp. Utah. Ecological Monographs 24: 89-124. SAGGER, S., M. J. HEDLEY, A. G. GILLINGHAM, J. S. FLOATE, M. J. S. 1970. Mineralization of nitrogen and ROWARTH, S. RICHARDSON, N. S. BOLAN, AND P. E. H. phosphorus from organic materials ofplant and ani­ GREGG. 1990. Predicting the fate offertilizer sulphur mal origin and its significance in the nutrient cycle in grazed hill country pastures by modelling the in grazed upland and hill soils. Journal ofthe British transfer and accumulation of soil phosphorus. New Grasslands Society 25: 295--302. Zealand Journal of Agricultural Research 33: ___. 1981. Effects of grazing by large herbivores on 129-138. nitrogen cycling in agricultural ecosystems. Pages SAGGER, S., A. D. MACKAy, M. J. HEDLEY, M. G. LAMBERT, 585-601 in E E. Clark and T. R. Rosswell, eds., AND D. A. CLARK. 1990. A nutrient_transfer model to Terrestrial nitrogen cycles. Ecological Bulletin explain the fate of phosphorus and sulphur in a (Stockholm) 33. grazed hill"country pasture. Agriculture, Ecosystems HAYNES, R J., AND P. H. WILLIAMS. 1992. Changes in soil and Environment 30: 295--315. solution composition and pH in urine-affected areas SAMPSON, A. W 1919. Plant succession in relation to range management. U.S. Department of Agriculture ofpastures. Journal ofSoil Science 43: 323--334. Bulletin 791. 76 pp. HITCHCOCK, A. S. 1950. Manual of the grasses of the SAMPSON, A. W, AND L. H. WEYL. 1918. Range preserva­ United States. 2nd ed. Rev. by Agnes Chase. U.S. tion and its relation to erosion control on western Government Printing Office, Washington D.C. 1051 grazing lands. U.S. Department of Agriculture pp. Bulletin 675. 35 pp. HOLMGREN, A. M., AND J. L. REVEAL. 1966. Checklist of SCHOLES, R T. 1990. The influence of soil fertility on the the vascular plants of the Intermountain region. ecology of southern African dry savannas. Journal of USDA Forest Service Research Paper INT-32. Biogeography 17: 415-419. Intermountain Forest and Range Experiment SCHREIBER, J. E, JR. 1988. Final report on the Flagstaff Station, Ogden, Utah. 160 pp. Limestone (Paleocene-Early Eocene) in the Manti­ JACKSON, M. L. 1958. Soil chemical analysis. Prentice­ LaSal National Forest, east of Manti-Ephraim, HaIl, Englewood Cliffs, NewJersey. Sanpete County, Utah. Department of Geosciences, JENNY, H. 1941. Soil formation. McGraw"HilI Book Co., University ofArizona, Tucson. Unpublished report. New York. SOMMERFELDT, T. G., AND C. CHANG. 1985. Changes in JOHNSON, H. B. 1964. Changes in vegetation of two re­ soil properties under annual applications of feedlot stricted areas ofthe Wasatch Plateau, as related to re­ manure and different tillage practices. Soil Science duced grazing and complete protection. Unpublished Society ofAmerica Journal 49: 983-987. master's thesis, Brigham Young University, Provo, STANLEY, K. 0., AND J. W COLLINSON. 1979. Depositional Utah. history of Paleocene-Lower Eocene Flagstaff KEENEY, D. R, AND D. W NELSON. 1982. Nitrogen-inor­ Limestone and coeval rocks, central Utah. American ganic forms. Pages 643-698 in A. L. Page, ed., Association of Petroleum Geologists Bulletin 63: Methods of soil analysis. Part 2. 2nd ed. Agronomy 311--323. Monographs 9. American Society of Agronomy and STEVENSON, E J. 1986. Cycles of soil: carbon, nitrogen, Soil Science Society ofAmerica, Madison, Wisconsin. phosphorus, sulfur, micronutrients. John Wiley and LAURENROTH, W K., AND W A. LAYCOCK, EDS. 1989. Sons, NewYork. Secondary succession and the evaluation of range SWIFT, M. J., O. W HEAL, AND J. M. ANDERSON. 1979. condition. Westview Press, Boulder, Colorado. Decomposition in terrestrial ecosystems. Studies in 31.2 GREAT BASIN NATURALIST [Volume 54

Ecology. Vol. 5. University of California Press, WILLIfuVlS, P. H., AND R. J. HAYNES. 1992. Balance sheet of Berkeley. phosphorus, sulphur and potassium in a long-term TIEDEMANN, A. R., AND T. D. ANDERSON. 1971. Rapid grazed pasture supplied with superphosphate. analysis of total sulphur in soils and plant materials. Fertilizer Research 31: 51-60. Plant and Soil 35: 197-200. WILLIfuVlS, S. T., AND T. R. G. GRAY. 1974. Decomposition TIEDEMANN, A. R., AND J. O. KLEMMEDSON. 1986. Long­ oHitter on the soil surface. Pages 611-632 in C. H. term effects ofmesquite removal on soil characteris­ Dickinson and G. L. E Pugh, eds., Biology of plant tics: I. Nutrients and bulk density. Soil Science litter decomposition. Academic Press, NewYork. Society ofAmerica Joumal50: 472-475. YOUNGBERG, C. T., AND C. T. DYRNESS. 1964. Some physi­ WEBER, J. N. 1964, Carbon"oxygen isotopic composition cal and chemical properties of pumice soils in of Flagstaff carbonate rocks and its bearing on the Oregon. Soil Science 97: 391--399. history of Paleocene-Eocene Lake Flagstaff of cen­ tral Utah. Geochimica et Cosmochimica Acta 28: Received 4 January 1994 1219-1242. Accepted 29 March 1994 great basin naturalist 544 C 1994 appp 313 328

geostatisticalSTATISTICALGEO ANALYSIS OF RESOURCE ISLANDS UNDER ARTEMISIA tridentatatridentateDENTATATRI IN THE SHRUB STEPPE

jonathan J Halvorson 1 harveyharve bolton jrjrejr22 jeffrey L smithasmith3 and richard E rossiarossi2

ABSTFLNCT desert plants can influence the pattern ofresouicesof resources in soil resulting in smallsmail scaleseale enriched zones although conceptually simple the shape size and orientation of these resource islands are difficult to studystud in detail using conventional sampling regimes to demonstrate an alternatealternative e approachappiappl oach we sampled soil under and around indi- vidual arteartemisiaArle misiamisla tntritrldentata sagebrush a dominant shrub of cool desert environments and analyzed the samples with univariate statistics and geostatisticsgeostatistics univariateUnivanate statistics revealed that soil variables like total inorganic N soluble C and microbial biomass C were distributed with highest mean values within about 25 cm of the plant axisixisaxis and significant lv1 lower mean values at distances beyond 60 cm however such simple analyses restricted our view of resource islands to identically sized symmetrical accumulations of soil resources undertinderlinder each plant Geogeostatisticsstatistics provided additional information about spatial characterscharacterischaractensticscharaecharacteris ties of soil nahlesnablesnabiesvavariables varloVarioanographvvariographygraphy revealed that samples separated byb a distance of less than about 70 cm were correlated spatially over 75 of the sample variance was attributable to spatial variability we modeled these spatial relationships and used krilingkngmgkriging to predict values forfoiholbolborror undamunsam pled locations resulting maps indicated that magnitude size and spatial distribution of soil resource islands varyvaly between individual plants and for different soil properties maps together with cross variovarlovanographvvariographygraphy further indicate that resource islands under A tntritrldentata are not always distinguishable from the surrounding soil bby slarpshaipshalp transition bound ariesarlesanes and may be asymmetrically distributed around the plant axis

key words resource islands geostatisticsgeostatistics artemisia dentatatntritrltridentatatndentatatridentate nutrient availability kriginkrilingkriging spatial correlation

recognition that individual plants can sig- whitford 1986 resource islands reynolds nificantlynific antly affect the local soil environment et al 1990 or ecotessaraecotessara jenny 1980 are dates back to at least the middle of the nine-nine hypothesized to result from several mecha- teenth century see zinke 1962 and has been nisms garner and steinberger 1989 includ- documented for many plant forms including ing litter fall or steinstemsternstemflowsternflowflow zinke 1962 broadleaf and coniferous trees zinke 1962 decreased erosion or increased deposition everett et al 1986 doescher et al 1987 coppingerCoppgoppmger et al 1991 microclimatological belsky et al 1989 bunch grasses hook et al amelioration of the soil smith et al 1987 1991 herbaceous legumes halvorson et al pierson and wight 1991 or inputs of resources 1991 and in particular desert shrubs ege g viavla insects birds or animals davidson and fireman and hayward 1952 garcia moya and moltonmorton 1984 mckell 1970 nishita and haug 1973 barth and detailed knowledge of the size and internal klemmedson 1978 burke 1989 burke et al dynamics of resource islands is important foiforholbol 1989 virginia and jarell 1983 bolton et al understanding energy flux mass transport 1990 1993 soil associated with plants typi- and nutrient cycling processes at the scale of cally contains greater concentrations of limit- the individual plant resource islands mavmay also ing resources eg N P contains larger pop- connote a tier in a progressive hierarchical ulationsulations of soil microorganisms and exhibits mosaic of plant and animal habitats resource higher rates of nutrient cycling processes like distributions and geochemicalbiogeochernicalbioblo processes mineralization charley and west 1977 bolton iei e patches sensugensu kotliar and wiens 1990 et al 1990 and denitrification virginia et al estimates of the distribution and numbers of 1982 these small scale enriched zones vari- resource islands in the landscape maymav aid in ously termed fertile islands camercarnergamergarner and stein- understanding population level processes and berger 1989 isles of fertility west 1981 can be used to refine regional estimates of

ipajficpacific northwest laboratory richland washington 99352 dietdirect correspondencee to 215 johnjohnsonsorisorl hillhallhalihail washington state Lnieruniversitysiisitsli pullmanpu washingtondslnngton991b412199164642199164 6421 2pacllc2pqcific northwest lahoiatonlaboratonLab oraton richland washington 99352 alandand managementManaKementgement and wateatrr conservation research unit LSDUSDAARSUSDA absARSVRS washington state universihrUniverL nierniersihsihsibrsihrvib Pidpullmanlinan washington 99164

313 314 GREAT BASIN naturalist volume 54

energy flow and mass tiantransfertiansfersfer furthermore correlation of the samples evaluated lesourceresource inter relationshipsinterrelationships ofoflaigelaigelarge numbers of individ- islands of individual plants noinolnor quantified the ual resource islands mayma influence ecosystemecos stem scale of soil heterogeneityhetel ogeneitogeneitnelt beneath A ttltntrltridentata structurestruestructuietule function and stability realReTirevnoldsreynoldsoldsoids et al ceostatisticsCeoGeogeostatisticsstatistics has previously been used to 1990 schlesingerschiesSchlesmger et al 1990 halvorson et al describe environmental and soil parameters 1991 associated with A dentatatntritrltndentatatridentatatridentate for example although conceptualconceptually simple the size pierson and wight 1991 used one dimen shape and orientation ofieofleofiesourceof sourceresource islands are sionaldional geostatisticsgeostatistics to analyze spatial and tem- not easy to evaluate previous studies have poral variability of soil temperature under A typically been based on relativerelatively small num- tntritrldentata halvorson et al 1992 demonstrat- bers of samples collected using a binarybmaramar regime ed that geostatisticsgeostatistics was an appropriate means iei e samples collected beneath the plant versus of measuring resource islands at the scale of samples collected away from the plant or an individual A tntritrldentatatridentatatridentate plant jackson and along a transect passing from plant to bare soil caldwell 1993a attempted to quantify the such an approachappiappl oach cannot be used to provide a scale of nutrient heterogeneity around indi- detailed spatial analysis ofieofleofiesourceof sourceresource concentra- vidual A dentatatntritrltndentatatridentatatridentate and pseudoroegneriapseudot oegnena spi tions or processes in the soil that aiealeafeare likely to cata in a native sagebrush steppe using semi exhibit complex responses to landscape and vanogiamsvariogramsvariovarlo grams they demonstrated increasinginclinci easing micmiemicrolitemicrositemiciositemici ositerosite variations burke et al 1989 autoconautocorrelationelation of soil nutrients at spatial additionaladditionally data collected from different scales 1 m but did not determinedeteimme whether locations 01or depths have often been analyzed small scale effects were attributable to individ- using inferential statistics such as anovaANOVAANOXA or ual plants or an artifact of the nested sampling t tests that assume samples aiealeare spatialspatiallylv inde- design used more recently they constructed pendent and identically distributed however krigelkngedkriged maps that showed relatively high con these assumptions may be dubious if untested centrationscentiationscentrations of soil variables like soil organic since ecological phenomena are often spatially matter extractable phosphate and potassium or temporally coneoncorrelatedelated and thenthelitheir frequencfrequence near pseudoroegnenapseudoroegneria tustussockstussockysocks but not artemisia distributions aiealeare rarely normal rossi et al shrubs jackson and caldwell 1993b however 1992 recently a branch of applied statistics these krigelkngedkriged maps did not directly quantify known as geostatisticsgeostatistics has been demonstrated spatial covariation between locations of indi- to be useful for determining spatial correlations vidual plants and resource islands further among ecological data and foiforboibolhorbor estimating values jackson and caldwell did not observe autocorautocar at unsampled locations ege g robertson 1987 i elationrelation foiforooiool microbialmicio bialblai processes at any scale robertson et al 1988 rossi et al 1992 that was measuimeaskimeasureded objectives of this study were 1 to use geo to meet ouioulour objectives we applied giostageosta statistics to describe and model the spatial tisticsristics in thieethree steps first we characterized continuity of soil variables around individual and modeled the similarity between samples plants i2 to use this information to produce as a function of their sepaiationseparation distance and graphical representations or maps of specific direction second we used this relationship to resource islands and finally 3 to quantify interpolate values at unsampled locations directly and the spatial correlation between plants and soil under near individual plants finally we quantified spatial be- variables wevve examined artemisia tridentatatridentate covariation tntritrldentata tween soil properties and plants sagebrush a prominent shrub of cool desert environments west 1983 previously known stumSTUDY SITFSITE to affect the distribution of resources in the soil several workers have measuredmeasuiedsured highelhigherhighet the studastudv was conducted at the andaridarld land concentrations of resources such as total C ecology ALELE reserve located on the total N inorganic N and higher rates of N hanford site inm south central washington see cyclingc cling in soil beneath A tntritrldentata than in bolton et al 1990 foifolfor details there remnants nearby open soil using a binary sampling of the native artemisia tntritrldentata elneinElyelytngiaelytrigiatrigia regime ege g burke 1989 burke et al 1989 spicatespicata association occur on silt loamscoams of the bolton et al 1990 1993 however these stud- warden or ritzvilleRitzville sellesseriesserles this perennial iesles have not accounted foiforhorbor possible spatial auto shrub steppe is the largest grassland type in 199411994 geostatisticalGEO STATISTICAL ANALYSIS OF RESOURCE ISLANDS 315 north america and covers more than 640000 high all plots were located within approxi- km2 of the intermountain pacific northwest mately 20 in of each other within a flat area too dry to support forests daubenmire 1970 with randomlylandomlv spaced plants multiple plots rogers and rickard 1988 in an undisturbed were sampled for two reasons first to assess state the A tntritrlfridentata E spicatespicata association spatial characteristics of sourceleresource islands bvby would be composed typicatypicallylh of three layellayerss of basing oniouioulour calculations on several examples vegetation an overstoverstoryorv shrub artemisia tntritrl rather than a single instance and second to dentata tridentata a laigelarge caespitose perenni- provide replicates in the event that no spatial al grass elytrigiaElytrigia spicatespicata rolbolformerlyfoi merlamerlv agropyron dependence of soil properties was observed spicaspicatuinspicatumtuintum and a small catcaespitosecae spitosespinose perennial data from all plots were combined to consider grass poa secunda growing on soil with a thin spatial dependence of soil properties around cryptogamic crust daubenmire 1970 several A tntritrldentata plants simultaneously however following disturbance such as this approach was chosen because it provided tillage glazinggrazing or fire the alien annual grass a more generalized evaluation of resource bromus tectoriumtectortuntectorumtectectortorumtun becomes established islands under individual A tntritrldentata and greatly increased the number of data pairs at METHODS advanvany separation distance estimates of plant location were required soil collection and analysis weiewele for cross variovarlovanographvvariographygraphy see below thereforeTherefoieoole cores of surface soil 10510.510 5 cm dia X 5 cm vegetation maps weiewelewere produced from vertical deep were collected at 41 specific locations photographs each plot was divided into 5 X within 2 X 2 m plots centered on mature A 5 enicm squares each square was classified into tridentata individuals fig 1 samples weiewelewere one of three groupings bare grass species located so as to minimize the number of data or A tntritrlfridentata based on predominant cov- points needed for analysis of spatial character- erage for this work no attempt was made to istics and to avoid preferential clustering we distinguish among grass species sampled five identically oriented plots 205 each soil sample was sievedsievek 5 nimmm mixed points in maichmalchmalehmarch 1991 when levels of soil and analyzed for a variety of soil variables for moisture and microbial biomass activity were this workwordwore we present data only for water solu ble forms of C total inorganic N iei e nitrate ammonium and soil microbial biomass C soluble soil C HOCHO1120 C was extracted with N room temperature deionized water and ana- lyzed using an infrared gas analyzer ionicalonicsionics 2202.00 Ee lehe eeii e inc watertown massachusetts total inor-inorlnor 00oo 00oo0 ganic nitrogen TINTI N was extracted within 48 h 0 of collection from 10log g subsamples of soil 151.51 5 using 25 ml 2mam KCI and analyzed colorimetri- 0 0 0 0 cally alpkem corp clackamasClackamas oregon soil E 00000ooo0 microbial biomass C SIRCSIR C was estimated 101.0iolo 0 0 0 from the respiratory response of soil to glucose 00000ooo a source of C readily utilized bvby heterotrophic 10 ZI 000ooo soil microorganisms anderson and domsch 050.5 0 0 1978 tentengramgram samples of soil were placed 0 in 40 mml glass vials moistened with deionized hgo covered with Paraparafilmfilm and incubated in 000 000 the dark at 23.523235 5 0 ac foiforholbolhorbor 1 wk i 235 I each sam- 0.0oo00 D i E 1 E 05c5c 00 0 e iee e e ple was then amended with a glucose solution oo 0.000 0.505 1.0loio 1.515 2.020 1 00 05 10 15 20 at the rate of 600 mg glucose 240 inglugmg QC kg east m soil bringing the final HO content of the soil to 20 25 equivalent to 30 kpa fig 1 schematic of a typical sampling plot ww 50 each plot glass hiveelceleficfive total wasas centered on an Artenartentisiaartcmwatisia trifritridefridedentatawafamafawata plant vials were sealed with silicone septa and dashed circles show the location of 41 soil coleseolescores 10510510.510 5 cm incubated foiforbolror 3 h soil respiration was mea- X 5 five dia cm deep all plots were oriented as shown sured bvby gas chromatography and i elatedrelated to 316 GREAT BASIN naturalist volume 54 estimates of soil microbial biomass C with deviations of tail and head values of the lag equations developed by anderson and domsch respectively we chose the correlcorrelogramogram 1978 because it removes the effects of lag means and standardizes by the lag variances rossi et univariate statistics al 1992 for this work we express correlocarrelo univariateUnivanate statistics were calculated for grams in the form of a standardized varivarlvariogramvanvanogramvagogramogram each soil parameter for classical inferential bvby subtracting each ph from I1 isaaks and statistics data were also assigned to one of five srivastava 1989 rossi et al 1992 distance classes depending upon sample loca- correlograms were first calculated solely as tion within a plot these classes can be enviedvienvi- a function of lag distance iei e the omnidirec sioned as concentric rings located at increas- tionaldional case without considering any differ- ing distances from the center of the plot the ences in spatial continuity with direction iei e first distance class comprised samples collected anisotropy however since resource islands directly under A tntritrlfridentata distance 0 cm need not be symmetric ege g zinke 1962 we n I1 per plot followed by the second also calculated directional correlograms for approximate distance 25 cmem n 8 per each soil property a separate correlogramcorrelogram was plot third approximate distance 60 cm n calculated foiroirolfor samples oriented 0 45 90 8 per plot fourth approximate distance and 135 115l1515 tolerance from each other 110 cm n 12 per plot and fifth approximate since correlograms are symmetric about the distance 130 cm n 12 per plot average origin ie 0 180 45 225 etc 0 values of soil properties in each distance class 45 90 and 135 directions correspond to were plotted as a function of radial distance samples aligned along east west northeast from the plant axis following vanovariovarlovanographyvariographyvapographygraphy see southwest north south and northwest south below login10910 transformed samples deemed east axes spatially independent were compared with because the data of each plot were concate- ANOVA using plot as a blocking factor nated during this analysis local anisotropies i e anisotropies specific to each plot were in Geogeostatistics ie in statistics effect combined thus any directional effects varlovariograpmvariographyVARIOGRAPHY we evaluated spatial char- we observed were a composite of the five plots acteristicsacteristics of each soil parameter with the non and presumably indicative of overall direction- ergodic autocorrelation function srivastava al trends to identify directions of maximum and parker 1989 and summarized results and minimum continuity we estimated the lag graphically as correlograms like variovarlovariogramsvanovanogramsgrams distance corresponding to a common value for correlograms represent the average degree of each directional correlcorrelogramogram isaaks and similarity between samples as a function of srivastava 1989 the directional correlcorrelogramogram their separation distance lag and direction with the greatest lag associated with a carrelocorrelo unlike the variogramvanvarivarlvanogramvagogramogram the correlcorrelogramogram filters gram value of I1 was identified as the direction out the effects of changes in both lag means of greatest continuity the correlogramcorrelogram with the and lag variances each point in a correlogramcorrelogram smallest lag corresponding to I1 was deemed was calculated from this equation the direction of minimum continuity the empirically determined scatterseatter of data I1 ph points in each correlogramcorrel ogram was fit with models nh known to produce a positive definite krilingkngingkriging nh system iei e matrices that provide both a unique s 11 faxaxizxizx ni jzxh11zxjh mijmjjmajmh11 solution and a positive estimation variance ii 1 isaaks models s hb shh and srivastava 1989 such typi- cally contain several salient features known as where zxizxjzuj and zxiaxizxax h arearc two data points nugget range and sill the nugget is the separated by the distance lag h datum zxizxaxiax amount of variance not explained or modeled is the tail and zxizxaxiax h is the head of the vec- as spatial correlation it is the apparent ordi- tor nh is the total number of data pairs sepa- nate intercept and is due to 1 unsampled cor- rated by lag h m h and mhm are means of the relation below the smallest lag and 2 experi- points that correspond to tail and head of the mental error rossi et al 1992 A small nugget lag respectively and S h and shS are standard relative to the sill indicates that a large pro 199419941 geostatisticalGEO STATISTICAL ANALYSIS OF RESOURCE ISLANDS 317 portion of the sample variability is modeled as CROSSgross varloariocraphyvariographyVARIO GRAPHY in addition to spatial spatial dependence conversely a large nugget chaiactensticscharacteristics of single soil properties we indicates less sample variability can be mod- also determined how soil properties covaried eled as spatial dependence the sill is charac- with plants we modeled spatial covariation terized by a leveling off of the correlcorrelogramogram with p abaabhgh the estimated nonergodicnonergodic cross model if present it indicates that spatial cor- correlogramcorrel ogram like the correlcorrelogramogram it accounts relation is on average constant however if for both variables fluctuating lag means and spatial correlation continues to change at lags variances isaaks and srivastavasnvastava 1989 rossi et greater than those considered in the conelocarrelocorrelo al 1992 because comparisonscompansons between contin- gram then a sill will not be apparent the lag uous variables iei e TINTI N SIRCSIR C and HOCHOhgo C value when the conelcorrelcorreiconelogramcorrelogramogram model reaches data and discrete variables iei e plant data might be complicated by a the sill is known as the range it represents contact effect 1985 converted N C the maximum separation distance within luster we TINTI SIRCSIR and hgo C data to binary variables which samples are spatially correlated at lags HOCHO using an indi- cator transformation journel 1983 this the range the sill of the variogramvanvarivarlvanogramvagogramogram may for work continuous data values of TINTI N SIRCSIR C approach the sample variance barnes 1991 and HOCHO C were coded 1 if they were greater KRIGINGKRICING TO ESTIMATE DATA AT UNSAMPLED than the local within plotpiot median or 0 follow- LOCATIONS kngingkriging has been likened to kriling ing halvorson et al in review crossgross coirelocorrelocorbelo multiplemulti le linear regression with a few twists P grams were then calculated for grass species rossi 1989 in classical multiple linear regres- or A tntritrldentata and indicator transformed TINTI N an of the dependent variable Y sion estimate SIRCSIR C and HOCHO C data using the equation is calculated from a weighted linear combina- tion of independent variables where each is P abhaba measured at about the same location in time nh or space usually only a single value of Y is is nh ax nh estimated similarly in krigkrilingkngingkriginging the adiza zxzy 1 I1 axizaAxiZA MAA iilkzb111ibxkzb abhimbhi estimated value of the variable for an unsam- i1iaI k1ka 3 pled location xtp is calculated as a weighted SAs hsahhsbh linear combination of the surrounding sam- hh pled neighbors where nh is the total number of data pairs separated by vector hb iaxizaixziaz is the coded N plant data equal to 1 if the specified plant ax V ar type was present or 0 if absent at some loca- zxAQ L giff zxi7raxi falf2l2 7r x m SAS I1 1 tion xi ma h and b are the mean and standard deviation respectively for the plant variable at those data locations that h where the zxizxaxiax s are the sampled values at aiealeare away from a soil their respective locations and the gsgi s are the property data location weights associated with each sample value in similarly ibxkzbigfxjpzg is the coded soil variable data equal to 1 if the data value is ordinary krilingkngingkrigkriginging weights used to estimatee is greater estima than the local plot median or else 0 location zxpaxzap are chosen so that the resulting estimate at zx akx xk i an birSB i are thee mean anddn standard is unbiased and has a minimum estimation nibmb khandband bihB 1aaah deviatiodeviationdeviatioatlo n of the sohsouSoii011 variable indicator calcu- variance and sum to unity krilingkngingkriging incorpo- lated for those locations that are h away rates a model of spatial here the continuity cor from a plant variable data location note relogramrelmelogramogram model and accounts for the degree when h is 0 equation 3 is equivalent to the of clustering nearby of samples and their dis- pearson correlation coefficient isaaks and tance to the point being estimated isaaks and srivastava 1989 srivastava 1989 we used ordinary point krilingkngingkriging unlike the correlcorrelogramogram values calculated for to estimate values of soil properties at unsam- the crosseross correicorrelogramcorrelogram may not be symmetric pled locations for each plot we estimated val- about the origin because both the olderorder and ues for the nodes of a 5 X 5 cm2 grid each direction are switched when variables are re- predicted value was based on a minimum of 6 versed isaaks and srivastava 1989 conse- and a maximum of 12 neighbors located with- quentquentlyly we calculated individual cross correl in a 080 8 m circular search radius obramsograms for the 0 45 90 135 180 225 318 GREAT BASIN naturalist volume 54

270 and 315 directions 130l3030 tolerance A correspond to soil samples aligned to loo100 these number 205 the east northeast north northwest west mean 383.835 southwest south or southeast of a plant 80 standard deviation 393.9 skewnessSkew nesinuminues 272.7 21 Kurtoskurtosisls 11111.1 RESULTS 60 itt

summary statistics indicated that samples 40 of TINTI N and HOCHO C were positively skewed while samples of SIRCSIR C were more normally 20 distributed figs 2aaa C total inorganic N F ranged from 06og0 6 mg kg soil L 060.6 to a maximum of 0 H L 236 23623.623 6 mg kg soil fig 2aaa the mean value for 0 8 1166 24 TINTI N of 3.8383 8 mg kg soil compared reasonably 38 TI N mgkg soil to the values reported by bolton et al 1990 of 4144.11 and 4944.99 mg kg soil for open soil crust and B A dentata soil respectively values observed 150 tn number 205 for HOCHOhog C ranged widely from 989.89 8 mg kg soil mean 62662.6 to 6339633.9633 9 mg kg soil fig ab2b estimates of 120 standard deviation 73073.0 SIRCSIR C ranged from less than 200 to over 1800 skewness 393.9 kurtosiskuriKurtosl 24924.9 mg kg soil fig 2q2cacaq the average value for 0 90 SIRCSIR C 750 mg kg was within the range reported bvby burke et al 1989 and equivalent 60 to about 90980 kg C ha soil assuming a bulk density of 131.31 3 bolton et al 1990 comparative- 30 ly smith and paul 1990 reported average microbial biomass pool size for grassland sys- 0 tems of 1090 kg C ha 0 300 600 99000 also indicated soil univariate statistics how h20 C mgkg soil properties varied with distance from the A dentatatntritrltridentatatridentate axis iei e center of the plot figs C C C C 40 T 3aaa concentrations of HOCHOhog and SIRCSIR number 205 were greatest within 25 cm of the plant axis mean 748 and lowest at distances beyond 60 cm figs standard deviation 281 30 og ab3b QC A similar pattern was observed for TI N 1 skewness 090.9 TIN Kurtoskurtosisls 434.3 except that mean concentration was low in soil r collected from directdirectlylv beneath the A triden 20 tata plant fig 3aaa anand from distances beyond 60 cm this somewhat unexpected finding of a 10 resource hole in the center of the resource island may be indicative of differences in the cycling of N under sagebrush and grass plants 0n rrfrd n vanographyvariographyvapographyVanoVariomano graphy indicated that samples of TINTI N 0 400 800 1200 1600 2000 hgo C and SIR C spatially correlated HOCHO SIRC were SIR C mgkg soil fig 4 correlograms for SIRCSIR C and timTIATINTI N exhibited similar ranges of about 070.70 7 or 080.80os 8 in fig 2 frequency histograms and some basic summary the correlcorrelogramogram for hgo C was similar to the statistics for A total inorganic N TINTI N B water solu others at small lag distances and equaled the bleblegbieg C ilolloFIOCFIO C and C soil microbial biomass C sirosirgSIRQSIRCSIR QC sample variance at a range near 07m070.7 m how- ever at greater lags correlcorrelogramogram values for majority of sample values are collected from HOCHOhog C increased above the sample variance an area with dimensions equal to or less than and did not appear to reach a sill until lags the variogramvariogram range barnes 1991 or if dis- were greater than 1mIam1 in A correlcorrelogramogram sill crete regions of high and low concentration greater than 1 for HOCHO C can occur if the occur at lags greater than the maximum lag in 199411994 geostatisticalGEO STATISTICAL ANALYSIS OF RESOURCE ISLANDS 319

t 1501.50 T a A 7 0 1251.25 0 0 0V e 6 t 1001.00loo E f z 5 0750.75 Wj ft M 0500.50 4 S 0 total inorganic N 0 0 H 0 soluble C 0250.25 2 3 V SIR biomass C X W 3 000ooon nn i i i 0000.00 tto1.00100 j 0000.00 0250.25 00500.500 00750.7575 1001.00too 1251.25 1501.50 0 2 laglog m

fig 4 omnidirectionalOmni directional correlograms for TINTI N HOCHO C and SIRCSIR C each point shown was calculated giom a ininianinimini- 200 mum of 253 pairs range 253 644 0 175 the correlcorrelogramogram ie an incompletely mod- OT 150 5 eled hole effect apparent nuggetsnnggets for 40 the all three soil parameters suggested that more 125 E than 75 a correlogramcorrelogram value of 0250.25 or less 0 100 of total sample variability could be modeled as spatial dependence 75 M 40 ranges observed in correlograms of soil j properties were used to establish the separation 0 50 60 tn distance beyond which correlation between 60 0 25 samples could not be distinguished from sam- X ple variance other words samples i i in separat- 0 i i ed by distances greater than the range were candidates for analysis using more traditional

C statistical techniques that assume indepen- dence v20 1000 i- such as ANOVA for our data samples in the first two distance classes were thus 0 combined and compared to samples from the last two classes because they were separated by 003 900 E mmore0re than about 080.8os m samples in the third 0 distancedistance class lying at an intermediate dis- 1 tance between the center and outer boundary boo800soo of the plot were excluded from analysis z analysis of variance of logo transformed data 20 1090 m using the five plots as blocking factors showed ji 700 mean concentrations of TINTI N hgo C and m SIRCSIR C to be significantly greater P ooi001001.001 of0 within 29 cm of the plant axis than values col- u 2 600SCOgoo lected 1071.07 m away from the plant axis table 0 25 50 75 100 125 150ISO 1 omnidirectionalOmni directional correlograms character- DISTANCE FROM PLANT cm ized spatial correlation or continuity purely as a function of lag distance however by consid- fig 3 mean standard error for A TINTI N B HRCHOCHRHOhog C and C SIRCSIR C numbers in parentheses are the number ering the orientation of samples in addition to of data points from five plots combined that contribute their lag distance directional anisotropies were to each estimate suggested directional correlograms showed 320 GREAT BASIN naturalist volume 54

TABLETVBLE 1 summary statistics and randomized complete block ANOVAs for logio109log transformed TINTI N HOCHOhog C and SIR- C data are summarized into two sample location classes near all measurements collected from within 29 cmern of the plant axis it 45 and away samples collected at distances greater than 107 cm fromfroinhrombrom the plant axis n 120 aerageaverage separation distance between the two location classes was 99499 4 cm

meanM eanpan standard deviation standardS tandard error near anavawav near anavawavaw a near awa TINTI N 058 039 034 032 005 003 h20HOCHO C 201 149 031 026 005 002 SIRCSIR C 293 282 014 017 002 002

source of variation df MSXI S F P plot 4 059 614 001 TINTI N sample location 1 1141 14 1181 001 error 159 010

plot 4 021 285 026 HOCHOhgo C sample location 1 879 12249 001 error 159 007

plot 4 006 232 060 SIRCSIR C sample location 1 036 1363 001 error 159 003 differences in both nuggets and ranges fig 5 by the variation observed between individual generally the largest apparent nuggets were plants and specific soil properties for exam- observed in correlograms oriented in the 0 ple distinct islands of TINTI N were not always east west and 45 northeast southwest clearly associated with A tntrltrifridentatatridentatatridentate instead directions with the exception of HOCHOhog C in three of five plots highest concentrations of these correlograms hadbad estimated nuggets of TINTI N appeared to be associated with location 040.40 4 or more conversely correlograms calcu- of grasses fig 6 plots in plots C and goo ABE lated for the 90090 north south and 1351350 D concentrations of TINTI N were highest in the northwest southeast directions generalgenerallylv vicinity of the A tntritrldentata canopy however exhibited nuggets of 020.20 2 vegetation maps of these plots indicate that of directions maximum and minimum con- grass species were present near the A triden tinuity identified TI N were for TIN maximum tata plant in plot E smallest concentrations of continuity was observed in the 45 direction in TINTI N were predicted to lie under the A tntritrl while lower but similar ranges of continuity dentata plant were observed in the other three directional weiewele in conversely highest accumulations of HO- correlograms fig 5 left column little C were clearly associated with A tntritrltndentatatridentatatridentate in anisotropyamsotropy was observed in directional correl dentata krigelkngedkriged maps in each plot high concentrations ograms foiforholbol C suggesting that spatial cor- obrams HOCHO of C were localized near the plot center relation could be adequately modeled with a HOCHO under the plant canopy of grass single isotropic correicorrelcorrelogiamcorrelogramogram fig 5 center location did not always appear to coincide column like TINTI N the direction of maximum species strongly with high continuity observed for SIRCSIR C was 45 with a concentrations of HOCHO C see NW of plot A NE and SE of plot direction of minimum continuity in the 135 14 14 B and plot direction fig 5 right column the aniso- NE 14 of E howeverHoweveivel high con- tropies for TINTI N and SIRCSIR C were accounted centrationscent rations of both TINTI N and HOCHOhog C did foifotformot in krilingkrigingki igmeigmg by using a model that evaluated coincide with location of grasses in the NW both distance and direction table 2 14 of plot B and SW 14 of plot E maps of the estimates generatedgeneiagenerated with these maps of SIRCSIR C indicate that islands of soil models and ordinary krilingkngingkriging were constructed biomass C were plesentpresent under plants but not for each soil property in each plot taken apparently specific to any particular type of together they suggest that generalizations vegetation relatively high concentrations of about spatial distribution of resources in the SIRCSIR C were estimated under A tntritrldentata soil beneath A tntritrldentata can be complicated plants in all plots however SIRCSIR C was also 199411994 geostatisticalGEO STATISTICAL ANALYSIS OF RESOURCE ISLANDS 321

TI N h10 C SIR C 000 haoh2oIO 151.5 r 101.0iolo 0 0 0 w 0050.55 w

000.0oon n i i ji L 1 1 i

45450 151.5i1 5.5 r 101.0lo 050.5 000.0oo Q 900goo90 151.5 101.0loio 0 0050.55 000.0oo 1350 1 135 151.51 5.5 T 101.0iolo 050.5 000.0oo 000.0oo 050.5 101.0lo 151.5 000.0oo 050.5 101.0iolo 151.5 000.0oo 050.5 101.0loio 151.5 LAG m

fig 5 directional correlograms for TINTI N HOCHOhog C and SIRCSIR C in the 0 45450 90 and 135 directions calculated witwitt a tolerance of 15 each point shown was calculated from a minimum of 47 pairs range 47 212 verticalvetVei tical lines corre spond to a correlcorrelogramogram value of I1 and were used to identify directions of maximum and minimum spatial contimlibcontinuitacontincontinuityultauitA open symbols are estimates of the apparent nugget the apparent ordinateordmate fit bnbabn eveeye

accumulated elsewhere in relation to the loca- observed in the 45 northeast and 90 north tion of grass species high concentrations of directions fig 7aaa the range over which A SIRCSIR C were observed in several instances not dentatatntritrlfritridentatatridentate remained positively correlated ditlwitl associated with high concentrations of either TINTI N was longest in the 45 and 90 directions TINTI N or h20 C ege g NE and SW 14 of plot extending to about 1 in and 0750.750 75 in respectrespec B NW 14 of plot D tivelydively positive correlations with TINTI N wedwerwen cross variovarlovanographyvapographyvariographygraphy indicated how TINTI N observed for other directions too but only to HOCHOhog C and SIRCSIR C covaried spatially with lag of about 050.50 5 in in contrast grass speciespecies respect to A tntritrldentata and grass species were positively correlated to above fediaimediai indicator transformed TINTI N data were similarly TINTI N concentrations in the 315 southeast and positively correlated with A dentatatnedentatamdentatatndentata and 270 south and 225 southwest directiondirections grass species at a lag of 0 equivalent to the fig ab7b the range over which grass specie pearson correlation coefficient figs 7abaab were positively correlated to TINTI N in thes however correlation varied with increasing lag directions was less than that for A tntritrlfridentata distance iei e showed spatial dependence in a in other directions grass species were ancoruncor different manner for each for A tntritrlfridentata related or negatively correlated to TINTI N at lag highest positive correlations with TINTI N were above 0 322 GREAT BASIN naturalist Vovolumeluinelurne 54

TBLLTABLE 2 modellamodell2model12 parameters used for ordinary krilingkngingkriging observed out to a lag of about 050.50305min beyond ALE soil properties this correlations of grass species with SIRCSIR C soil parameter remained approximately constant TINTI N lplI p45p 5 154 0428 sph0spho2sphon 277 052403240 524324 sph180Sph 180 discussion lpl1 p135p 35 154 04280 428 sph071Sph 071 0524 sph120Sph 120 Geogeostatisticsstatistics can be applied to resource SIRCSIR C island data to provide several useful diagnostic 1 072 03810 381 sph023spho23Sph 023 0717 sphlsphj8787 p45 sahl features to actual of the land- lpl1 p135p 072 0381 spho74sph074Sph 074 07170 717 sph075Sph 075 prior mapping scape itself for example variovarlovariographyvanographvgraphy can IIOCIIOliollo1120 C define the presence and extent of spatial cor- 1 p 140 sahlsphlsph13 3 relation and alert the researcher to apply with 1 models shown for timTIATINTI N and SIRCSIR C are a coincorncombinationbination ofaot a nugget constant caution classical statistical comparisons that and two splicrical rnodcls the spherical model denoted sabsphspb is an autho nedriedmed modimodeodel 0commonhcommonso on used to mode varioanoanogramsanagramsgranisgrams the number that precedes assume samples are independent and from sph can be thought of as the local sill for that model while the number inin identically distributed populations for these parentheses isis the range at which the local sill is reached iseeiveesee isaaks and snastaasriastavasrivastava 1989 for a correlcorrelograinogrAin the standardized formforni is iphilphl1 phphi 15 methods to be more properly applied to spa- lagrange 050 5 lagrange 3 iflagil lag langerange else I1 if otberwotbergotherwiseise tial data should probably be lim- oifoteOTF Geoceogeostatlsticiansgeostatisticiansstatisticians oltenoften distinguish behenbehbetweenen the nugget used for diag- comparisons nostic purposes which is the apparent ordinateordmate intercept and thetlletile nugget ited to those samples separated byb distanceslistances alue which is used in rnornodelingdeling range of the correlcorrelogramogram table 1 webster 1985 robertson 1987 this is true for studies indicator transformed HOCHOhoghgo C data were that compare samples collected along a con- positively correlated with A tridentata but not tinuum such as distance depth or concentra- with grass species at a lag ofoof 0 mcd As figs 7cd tion iei e resource gradient or as a function of with TINTI N highest correlations with A triden time observed in the 45 and 90 tata were direc- another promising use of variovarlovanovanographyvapographyvariographygraphy is to tions fig 7q7cacaq similar patterns of spatial relate spatial continuity of two or more variables dependence were observed for all directions at the same site or the same variable at two or the distance to which HOCHOhog C remained posi- more sites by comparing variogramsvariovarlovanogramsvano grams covariocevariocovano tively correlated with A tridentata ranged grams or correlograms with one another this from a minimum of about 050.5 in to a maximum approach may be useful for comparing the of near 0750.75 in in the 45 and 90 directions scale of ecolodecoloecologicalcicalgical processes 01or ecosystem unlike A dentatatritridentatatridentate HOCHOhog C was not positive- boundariesboundanes but should be approached with ly correlated with grass species fig ad7d caution for several reasons first each point in instead h20HOCHO C was moderately negatively a traditional variogramvanvailvarivafivanogramvagogramogram represents the average correlated in the 0 45 90 anand 135 direc- value of the squared difference between many tions meaning grass species were more asso- pairs of data points while an average value ciated with below median concentrations of may be appropriate for modeling spatial conti- HOCHOhog C at lags greater than about 020.2 in little nuity as a summary statistic it does not indi- change in crosseross correlograms was observed cate the range of individual squared differ- indicating only a weak spatial dependence ences or provide an estimate of the goodness in contrast to HOCHO C erosscross correlograms of fit about each point in a varivarlvariogramvanvanogramvagogramogram the indicated that SIRCSIR C was slightly more corre- range of the deviation about the average value lated with grass species than with A tridentatridental may be large or small see webster and oliver ta at a lag of 0 figs 7efcef like other soil 1992 complicating comparisons of variovarlovariogramsvanovanogramsgrams properties strongest positive correlation thus for comparative purposespuipul poses other more between A tridentata and SIRCSIR C was robust representations of spatial dependence observed in the 45 direction which also such as journel s mad estimator cressie remained positively correlated to lags in hawkins robust estimator or the rodogram excess of I1 in fig 7eaeE lowest correlations see rossi et al 1992 may be more appropriate with A tritrldentata were to the 270 and 225 choices second even if variovarlovariogramsvanovanogramsgrams for two directions indicator transformed SIRCSIR C data properties are similar resultant estimates may were most correlated to grass species in the yield very different maps ege g TINTI N and SIRCSIR C 225 270 and 315 directions fig af7f in this study this is because estimation of un- spatial dependence of correlations was known data values by krilingkngingkriging depends not only 199419941 ceoGEOceostahsticalgeostatisticalSTATISTICAL ANALYSIS OF RESOURCE ISLANDS 323

plot A plot B plotpiot C piotplot D plotpiot E

9 i 11 1

4 2.2 if

IF

D 9

52 3

3 2 3

TO 50 30 so eb I CV 0 iopoptop 1110 60 0 0 014 lo100 0 tiolioeio 0

0 A 11001 10 0o o 000 6001googoop 800 1 00 0 600 1 0 7000 L alooaioo p.4 cac& rs 0 0 0 04 00 goa N 0 0 1 700 700100 120s0 boo800 V 0 0 0 boo100 00 700

fig 6 2 X 2 m maps ofofvegetationvegetation and krigedbriged estimates eftiofti N 110iio C and SIRCSIR C foroor five ALE plots each krigedbriged plotpiot is composed of 1681 points estimated by ordinary point krilingkriging vegetation maps indicate vertical projections ofa tritrlfridentata blackblaek and grass species crosshatch as determined from photographs soilsoli properties are depleteddepicted in nig kg teassssssssssoil dw upon a model of spatial continuity but ultimate- TINTI N on a seasonal or shorter time scale ly upon degree and configuration of the known other environmental variables such as soil sample values in the field texture total N or C might change more slow- during our analysis ofresouiceof resource island data ly thudthildthird we chose to analyze data for the five using geostatisticalgeostatistical methods we made several plots collectively rather than for each plot assumptions or decisions about the data that individually this choice reflects an interpreta- could have affected our interpretations first tion that correlograms for individual plots were we assumed that resource islands under A tntritrlfri leasonreasonablyably similar to each other and allowed dentata could be monitored using a particular our calculations to be based on a greater num- configuration of samples located within a 2 X ber of paired comparisons we reasoned that 2 m plot A different number of samples col- conelconelogramcorrelogramcorrelcorreiogram models of spatial continuity de- lected from a larger plot with a different shape rived from concatenated data would summarize or in a different pattern might have generated typical patterns of spatial continuity and direc- different correicorrelcorrelogramcorrelograrnogram models or krigelkngedkriged esti- tional anisotropies alternativelyAlternative lv although sin mates webster and oliver 1992 second we gle plot analyses would result inm plot specific collected data at a single time during the year models of spatial continuity with greater thereby making the krigelkngedkriged maps snapshots specificity they might make generalizations in time and space data values and spatial difficult finally we assumed that spatial con- continuity undoubtedly vary for some types of tinuity for HOCHOhoghgo C was reasonably described environmentalenanenvn mentalononmentalunmental variables ege g soil moisture or bvby a single isotropic model but wee concluded 324 GREAT BASIN naturalist volume 54

A with artirynisia with graasgrass 0 B species 060.6og 0 0 a 180 450 0 2252250 47 goo 040.4 1 7 90900 A 2700270 z v 1351350 A 3153150 020.2 0

A oo al414l 000.0 W 0 Av 02020.2

04040.4

080.8ob

080.8og U A 0 4

0 020.2 Q v 000.0oo 1.1 A 0 V 02020.2 z

V IV

060.6og

040.4

020.2

000.0oo V 0

02020.2

04040.4 0000.00ooo 0250.25 0500.50 0750.75 1001.00loo 1251.25 0000.00ooo 0250.25 0500.50 0750.75 1001.00loo 1251.25 m laglog laglog m

fig 7 directional crosseross cooconeoncorrelogramseloelogramsgrams shoeingshowing spatial covariancecoaeoa nancenanee of indicator transformed TINTI N HOCHO C sibSIRSI RCR C data and presenceabsencepresence absence coded as I1 or 0 of A tntritrldentata and grass species correlograms were calculated for 0 easteastcast 45 northeast 90 north 135 northwest 180 west 225 southwest 270 south and 315 southeast directions t3030 tolerance each point shown summarizes a minimum of54of 54 pairs iangelangerange 54 158 199419941 geostatisticalGEO STATISTICAL ANALYSIS OF RESOURCE ISIANDSISLANDS 325 that directional anisotropies observed for TINTI N rableTABLETABLF 3 frequency of occurrence of wind at the ALE and SIRCSIR C were important enough to be site source H bolton pacific northwest laboratorvlaboratory accounted for in the estimation process source quadrant 1990K 90 Kiggi199191 results of a geogeostatisticalstatistical analysis cannot hours percent hours percent completely replace sound ecological reason- ing or theory rossi et al 1992 thus the re- 0 90 NE 1161 1325 1208 1397 searcher must decide whetherwhethei directional ani- 90 180 SE 1557 1778 1613 1866 sotropiessotropiesples observed in descriptive variographyvanographvvariovarlography portray significant spatial patterns or are 180 270 SWSNNI 4090 4670 3974 4596 merely a coincidental resultresuit of the number and arrangement of data the decision to 270 360 NW 2227 2227 1851 2141 account for spatial anisotropy in the krilingkngmgkriging procedure is in part related to the desired end evidence for the occurrence of resource product of the geogeostatisticalstatistical analysis for ex- islands in the ALE landscape was provided by ample if the goal of an analysis is the most accu- comparing concentrations of soil resources rate representation possible of a particular collected near A tridentata vegetation to resource island under a specific A tntritrlfridentata those collected away from the plant however then a highly detailed model of spatial continu- the specific sampling regime employed to ity would bee appropriate regardless of the evaluate near vs away influenced the par- source of spatial variability in this case variagvariogvanog ticular conclusion reached for example raphy based on a concatenated data set might bolton et al 1990 were unable to conclude be less appropriate than analysis based only that concentrations of TINTI N in soil under A on the single plot however the goal of geo tridentata were significantly greater than con- statistical interpretation of ecological data may centrationscentrations measured in open soil crust based not be to produce detailed site maps instead on six samples drawn at random from each soil the ecologist may be more interested in pat- type see also doescher et al 1984 in con- terns that aiealeare broadly applicable anisotropies trast we found that evaluating TINTI N vs dis- can often be related to information about the tance away from the A tridentata axis resulted environment such as stratigraphic meteoro- in the naive conclusion that significantly high- logical or hydrogeological patterns isaaks and er concentrations eftioftiof TINTI N would always occur srivastava 1989 and mavmay suggest linkages be- under A dentatatritridentatatridentate plants fig 3aaa table 1 tween environmental variables our decision to such a conclusion for TINTI N and other soil account for anisotropies in the krilingkngmgkriging process properties would lead to a model of a land- was in part influenced by information about scape composed of identically sized symmet- another environmental parameter prevailing rical resource islands centered on lacfieacheacfi A tri wind direction for this reason we would expect dentata individual and would infer some sort the anisotropies observed for TINTI N and SIRCSIR C of causal relationship between concentration to be a consistent feature of the ALE landscape of TINTI N and A tridentata presence however directional coirelogramscorrelograms revealed greatest brigedkriged maps suggest that greatest concentra- spatial continuity for samples of TINTI N and tions of TINTI N were not always associated with SIRCSIR C in the 45 direction northeast south A tridentata west cross correlograms more specificalspecificallylv autocorrelation of soil properties was indicated that above median concentrations of described using variovariographygraphy the association soil properties were most correlated to A tntritrlfri of soil variables with A tridentata individuals dentata in the 45 direction or northeast for was supported jointly by krigedbriged maps and the ALE site cumulative records indicate that crosseross correlograms with the latter showing prevailing local wind direction is from the that soil properties especially hgo C were southwest quadrant table 3 corresponding positively correlated to A tridentata krilingkriging to the downwind direction of greatest spatial is a means for producing visually satisfying continuity prevailing wind direction might maps of soil properties and provided addition- influence spatial patterns of soil resources byb al insight into characteristics of resource dis- affecting distribution of litter deposition which tributiontribution under A dentatatritridentatatridentate however we for A dentatatntritrltndentatatridentatatridentate at the ALE site may exceed relied on these maps primarily as heuristic 60 kg ha annually mack 1971 tools because we recognized that krigedbriged maps 326 GREAT BASIN naturalist volume 54 are models that can be influenced by decisions these factors would become more important about the data set ege g concatenated vs single during the hot dry summer months and could plot the art of varivarlvariogramvanvanogramvagogramogram modeling the limit distribution of SIRCSIR C to locales closer to hpetype ofkrigidgofkiiging chosen ordinary kngingkrigidg is a A tntritrlfridentata data smoother the specific search strategy assessing the distribution of soil microbial used and the method of graphical representa- populations or micromicrobiallybially mediated nutrient tion finally krilingkngmgkriging by itself does not provide cycling processes such as mineralization or a measure of estimate confidence or reliability denitrification is complicated by multiple like parametricnonparametricnonparametncnon methods journel 1983 or resource requirements and compensatory stochastic conditional simulation rossi et al capabilities of living microorganisms smith et 1993 al 1985 for example microbial population krigelkngedkriged maps of HOCHOh20hog C appeared to be size or activity within a C substrate resource most similar to graphs of summary statistics vs island might be limited by the availability of distance from plant axis fig 3aaa in each krigelkngedkriged soil N conversely the same microbial popula- map fig 6 high concentrations coincided tion might be limited by the availability of C with A tntritrldentata in a classic sourceresourcele island substrate despite an N nchrich environment pattern these accumulations might be tied under such a scenario the greatest population closely to inputs from A tntritrlfridentata litter fall size or activity might occur in a location with presentinglerepresenting a source of C that could be low or intermediate quantities of both C and accessed byb heterotrophic soil microorgan- N and the resource island for soil microorgan- ismsiams alternatively high concentrations of ismsiams or mineralization potential would appear HOCHOhoghgo C under A dentatadentatotntritrlfritndentatatridentatatridentate might not indicate distinct spatially from other resources large C inputs instead they might indicate estimation of soil properties like SIRCSIR C that the accretion of soluble but recalcitrant depend on the distribution of one or more forms of C not readily useable by soil micromieromicro- other resources may need to be evaluated with organisms in this case the term resource respect to temporal and spatial distributions of island would be ambiguous to have ecologi- alternative resources cal significance a resource island must be kneedkngedkrigedbriged maps of various soil properties can evaluated for resource quantity sourceresourcele qual- be interpreted within the context of the rela- ity and presence of alternative lesourceresource sub- tiontionshipship between the particular soil parameter stitutesstitutes further the significance of resource and A tntritrlfridentata our data indicate that shape accumulation into islands might change with and orientation of resource islands under A time in relation to diurnal cycles growing sea- tntrifridentata vary with the specific soil property son or successional stage halvorson et al considered need not be centered on the axis 1991 of an A dentatatntritrlfritndentatatridentatatridentate plant and need not be sym- krigelkngedkriged maps of SIRCSIR C showed accumula- metrical the maps also provide evidence that tions of soil microbial biomass in close proxim- suggests a vertical projection of the plant ity to each A tntritrldentata individual however canopy is not well correlated to the distribu- high concentrations were also observed for tion of soil variables and thus should not be locations corresponding to other plant species used as a basis foifolfor sampling designs fig 6 demonstrating that resource islands of micro-mielomicro for some soil properties ege g hgo C the dif- bial populations or activity can be numerous ference between values characterizing the and are nonspecific to A tntritrlfridentata addition- resource island and those characterizing the ally a significant amount of SIRCSIR C was esti- surrounding matrix may be large and the mated for locations not associated with any resource island may appear to have sharp plant this suggests that while local inputs by boundaries conversely the range of data val- plants may stimulate microbial population ues for other soil properties ege g SIRCSIR C may growth or activity sufficient resources exist in be smaller and the transition from resource the environment to support moderate amounts island to the surrounding soil matrix more of SIRCSIR C during some times of the year complicated resource island boundaries mavmay however plant location may control the dis- also change with direction making sampling tributiontribution of SIRCSIR C indirectly through influ- designs based on only a few transects ques- ence on microclimatologicmicroclimatologicalaf factors such as tiotionablenable finally resource islands do not occur soil temperature and evapotranspiration under all A tntritrldentata or for all soil properties 199419941 geostatisticalGEO STATISTICAL ANLASISANALYSIS OF RESOURCRESOURCE E ISLANDS 327 other plants like annual and perennial grasses trees on their phphysicalsical chemical and biological envi- can be the focal points for resource islands of ronronmentsments in a semiserniaerni andaridarld savanna in kenyaken a journal of applied ecology 1024 some variables jackson and calcaldwellgalgaidwell 1993b 26 1005 BbolionBOLTON H JR J L SMITH ANDn SO LINK 1993 soil Geogeostatistics allows statistics estimation and map- microbial biomass and acactivitytiittilt of a disturbed and ping of resource islands in considerable detail undisturbed shrub steppe ecoecossystemstem soil biologybloh such maps can be used to further our under- and Biochemistbiochemistryrv 25 545 552 standing of the ecology of A tritrlfritridentdentataatalatay refine boltoboltonboitonBOITO H JR J L SMIIHSMITH ANDD R E XVILDUNGILDLG 1990 nutrient budgets for shrub steppe ecosystems nitrogen mineralization potentials of shrub steppe soils with different disturbance histories soil reveal the existence of resource and process science society ofamericaof america journal 54 887 891 dependent patterns and help provide a ratio- burkhbunkeBURKE I1 C 1989 conticontrolol01 ofnitrogenof nitrogen mineralization in a nale for sampling designs based on natural sagebrush steppe landscape ecology 70 1115 1126 boundaries besides two dimensional space burkhBLRKFBURKE 1 C W A RFIERSREINERS ANDD D S SCHIMELSCHIMLL 1989 geostatisticsgeostatistics can be used to consider differ- organic matter turnoturnoverer in a sagebrush steppe land scape biogeochemistry 7 11 31 ences in spatial with soil depth continuity ie CHARLEYCHRI E J L ANDD N E westNVEST 1977 mieromicro patterns of a third dimension or time via repeated mea- nitinitrogenogen mineralization activity in soils of some surementssurements however even with geostatisticsgeo statistics shrub dominated semidesertsemi desert ecosystemsecos stems of itahutah our definition of a resource island can be soil biology and biochemistry 9 357 365 improved whether a resource island is more COPPINGFR K D R A REIERSREINERS I1 C burkeaniBLRKEBURKE axiANi R K OLSON01 SON 1991 net on a sagebrush properly delineated by some differ- ol erosion steppe minimum landscape as determined bby cesium 137 distribu- ence in resource concentration or related to tion soil science society of amelicaamericaamenea journal 55 the ecological significance of small differences 254 258 in concentration remains to be answered DLBEMIRE R 1970 steppe vegetation of washington further the resource island effect may be technical bulletin 62 washington agricultural related to a single experiment station college of agriculture more than environmental washington state university pullman parameter consequently methods must be DVMDSONDAVIDSON D W annANDND S R morlonMORIONMORTON 1984 dispersal developed to simultaneously integrate infor- adaptations of some acacia species in the australian mation for several environmental variables and zone ecologyccokecok 65 1038 1051 and summarize them spatially DOESCHERDOESCIIER P S L E EDDLEMANEDDLENIAN AND M R VAITKUSVAITKLS 1987 evaluation of soil nutrients phpli and organic matter in rangelands dominated bby estemwestern juniper acknowledgments northwest science 61 97 102 dorsenorkenorseDOESCHERHER PSP S R E MIILERMILLER ANDND A fl WIARDWINWARD this research was supported in part by the 1984 soil chemical patterns undertinderlinder eastern oregon ecological research division office of plant communities dominated bvby big sagebrush soil and science society ofamericaof america journal 48 659 663 health environmental research US everetrevereteEVEREE ERFITTr R S SHARROSHARROW aniakiaklu D THRATHMN 1986 soil nutri- department of energy DOE pacific north- ent distribution undertinderlinder and adjacent to singleleafsingleleaf west laboratory is operated for the DOE by pirnpinyonon crowns soil science societysocietsochet of america battelle memorial institute under contract journal 50 788 792 DE ac06 1830 we thank joe FIRFMVSFIREMAN M ANDND II11 E HAYWARDI lvmbvm aiduid 1952 indicator signif- 76rlo icance of some shrubs in the escalante desert aufdermaur debbie bikfasy kirk stallcop and utah botanical gazette 114 143 155

GARCIA moya 1 jeff sullivan for analytical assistance M W carciaGARCIMOGARCI MO E AN DCC M bickellxickellme km 1970 contributions palmer and J price provided helpful editorial ofshrubsof shrubs to the nitrogen economeconomy of a desert avashnvashash comments plant community ecology 51 81 88 GRERgherGARNER W ANDD Y steinberger 1989 A proposed mechanism for the formation offertileof fertilebertlie islands in the literature CITED desert ecosystemecos stem journal ofariaofaridof and environments 16 257 262 ANDERSON J P E ANDND K H DOMSCH 1978 A physio- halvorsonHLORSO J J H boltonBOLTO JR ANDn J L SMITH 1992 logical method for the quantitative measurement of measurement of resource islands underneath microbial biomass in soils soil biology and arteinisiaartemistaartemishaArtemista tntritrlfridentata a geostatisticalgeostatistical approach pages biochemistry 10 215 221 117 120 in pacific northwest laboratory annual bricsbrfcsbreesBARNES R J 1991 the varivarlvariogramananagramanogramogram sill and the sample report USU S department of energy office of variance mathematical geology 23 673 678 scientific and technical information PNL 8000 LCUC BARTH R C VDAND J 0 klemmedsonkl l MMEDSON 1978 shrub 407de407 DE agog 76rlo 1830 oak ridge tennessee induced spatial patterns of dndry matter nitrogen and HLORSOHALVORSON J J J L SMITH H boltonBOLTOM IRJR ANDD R E organic01 game carbon soil science society of america rossi in reielreiewrelewreview defining resource islands using journal 42 804 809 multiple variablesv and geostatisticsgeostatistics BELSKYBELSKAA JRJ R G AIUNDSONAMLDSOJJ M DLBLISdubuisDUXBURYS J rihrimRIHA HAIAORSONHALVORSON J J J L SMITH ANDD E H FRNZFRANZpranz 1991 A R altallALI AND S M MONGmoncNINVONGA 1989 the effects of lupine influence on soil carboncalbon nitrogenniti ogen and 328 GREAT BASIN naturalist volumeVolurne 54

microbial activity in developing ecosystems at rossi R E P V BORTH AND0 o J J ToLiFFTOLLEFSONSO 1993 mount st helens oecologia berlin 87 162 70 stochastic simulation foifolfor characterizing ecological hook PPBB I1 C BLRKEBURKE ANDD W k LAUENROTHLALCNROTH 1991 spatial patterns and appraising risk ecological heterogeneity of soil and plant N indand C associated applications 3 719 735 with individual plants and openings in notnorneitlinoitlitiitit rossi R E D J MLLLAMULLA A G JOLRELJOURNEL MDAND E 11 amerleanamericanAmeiamer llanicaniLan shortshortgrassshoitgiassgrass steppe plant and soil 138 FRZFRANZpranz 1992 Geogeostatisticalstatistical tools for modeling and 247 256 interpreting ecological spatial dependence IsISAAKstaksYAKS E II11 ANDD R Mn1na srlstysrivastwa 1988 spatial conti ecological monographs 62 277 314 netynmty nuitv measures for probabilistic and deterministic schlesinger W ifII11 J F REYNOLDS G L CLICHMcunningham geostatisticsgeostatistics mathematical geology 20 313 341 L FE HLEANEKEHUENNEKE NV M JARRELLJVRRELL R A VIRGINIA 1989 an introductionatimti eduction to applied xeogeostatisticsgcostatisticsgeostatistics ANDID W G WHHFORDWHITFORD 1990 biological feedbacks press in ofordoxford university pi ess new yorkyolk global desertificationdecertification science 247 1043 1048

JACKSON R B M 1 JCRSN ANDD M CLDELLCALDWELL 1993a the scale of smiriljSMITI I J L ANDD E A PULPAUL 1990 the significance ofot soil nutrient heterogenheterogeneitycitelteit around individual plants and microbialmicromiero bialhialblai biomass estimations in soil pages 357 396 its quantification avithnvithith geostatisticsgeostatistiesxeogeo statistics ecology 74 inn G stotzkstotzkistotzkv and J bollag eds soil biocheblochebiochemistrymishimisny 612 614 vol 6 marcel dekker new york 19931 geostatistical Geo statistical patterns of soil heterogene SMIIHSMITH J L B L XIcNEALMCNEAL ANDND H H ghegghengCHENC 1985 ityit around individual perennial plants journal of estimationestitriation of soil microbial biomass an analysis of ecology 1993 81 683 692 the respiratory responselesiesponse of soils soil biology and jeijeljennijenny H 1980 the soil resource origin and behavior biochemistry 17 11 16 springer variagveriagei lag berlin SMITH S D W E SHTIIANDSMITH D D T PATTEN 1987 effects IOLRMLJOURNEL A G 1983 nonparainetncnonparametrie estimation ofspatialof spatial artificiallyofartificiallyof imposed shade on a sonoran desert distributions mathematical geology 15 445 468 ccosstemecosvstemecosystem arthropod and soil chemischemistschemistnchernichernistrystrytn responses kotliar N B AND0 o J A WIENSlesIES 1990 multiple scales of journal ofariaofaridof and environments 13 245 257 patchiness and patch stiucturestructure a hierarchical frame srivastavaSHIST R M ANDNO II11 NI PVRKI- R 1989 robust mea work foifolfor study the of heterogeneity oikosbikos 59 sure of spatial continuity pages 295 308 in M 253 260 armstrong ed geostatisticsGeo statistics 01volvoiol 1 kluverkluwer llsterLLSIFRLUSTER G R 1985 raw materials for portland cement academic dordrechtDordrecht the netherlandsnethel lands applications of conditional simulation of coregional VIRGINIA71 RGINIA R AANDA YND W M JARRELL 1983 soil properties ization unpublished doctodoctoraliallallai dissertation depart in a mesquite dominated sonoran desert ecosecosystemstem mcntofeaithment of earth sciences stanfoldstanfordStan foidfold university soil science society of americaofamerica journal 47 138 144

Stan V stanfoldstanfordloldfoidfold california vlrcimVIRGINIA R A nan7 M JRRELLJARRELL ANDD E FKXNFRANC VIZCAINOOVIZCAINO0O vlzcvizc UNO MACKM v K R N 1971 mineral cyclingc cling in artendsiaaitcmisia tridentata 1982 direct measurement of denitrification in a unpublished doctodoctoraliallallai dissertation ashingtonwashingtonNV state prosopis mesquite dominated sonoran desert university pullman ecosystemecos stem oecologia berlin 53 120 122 NISHITA II11111 AND HAUG isarismr ilslis annNDRMRM huehwe 1973 distribution ofdifferof differ NVEBSTERebsterEBSTLR R 1985 quantitative spatial analysis of soil in ent forms of nitrogen in some desert soils soil the field advances min soil science 3 1 70 science 116 51 58 WEBSTERNVEBSTER R ANDD M A OLIVEROEIVER 1992 sample adequate PIFRSONPIERSON F B AND J R lohilohlWIGHTIGHI 1991 variabilityanabil it ofnearolnearof neal iv to estimate variovarlovariogramsvanogranisvanogranisgrams of soil properties journal surface soil temperaturetemperatinc on sagebrush rangeland of soil science 43 177 192 journal of range management 44 491 497 rstEST N E 1981 nutrient cyclingcv cling in desert ecosystems REYNOLDS J FE R A arniirmiVIRGINIAirni annANDn J M GORNLHLSCORNELIUS pages 301 324 min D W goodall and R A pernperry 1990 resource island formationfoiroibol mationmatlon associated with the eds andaridarld land ecosystems structure function and desert shrubs creosote bush lanealarrea tritrldentata and management cambridge lnieisituniversitvuniversity press mesquite prosopis landulosalanduglandulosaglanduloseiosalosa and its role in the cambridge united kingdom stability of desert ecosystems a simulation analysis 1983 temperate deserts and semi deserts ecosys-s- supplement to bulletin of the ecological society of temsterns of the world elsevier scientific publishing americaAmeneaamenea7070 299 300 co ncnew york 522 appp ROBERISROBERTSON G P 1987 Geogeostatisticsstatistics in ecology interpo- WHITFORDWHITEORD W G 1986 decomposition and nutrient lating with known variance ecology 68 744 748 cyclingc cling in deserts pages 93 117 in WV G whitford RORLRISONROBERTSON G P NI A HUSTON FFGC EsEVANS ANDD J M ejed patterns and processes in desert ecosystemsecos stems TICDIETIEDJE 1988 spatial variabilityvalvai lability in a successional lmersituniversity of new mexico presspi ess albuquerque plant community patterns of nitrogen availability 1 zilZINKE P J 1962 the pattern of influence of individual ecology 69 1517 1524 forest trees on soil properties ecology 43 130 133 ROGERS L E ANDND W H rickard 1988 introduction shrub steppe lands pages 1 12 H in W rickard received 5 october 1993 EL E rogers B E vaughan and S FE liebetrau accepted 23 march 1994 eds shrub steppe balance and change in a seniigeniisemi andarid terrestrial ecosystem elsevier new york rossi R E 1989 thethegeostatisticilmteipretationofecoxeogeostatisticalgeostatistical interpretation of eco logical phenomena unpublished doctoral dissertadis serta tion washington state university pullman great basin naturalist 544 0 1994 appp 329 334

HABITAT preference AND DIURNAL USE AMONG GREATER SANDHILL CRANES

donald E mcivorlmclvor1mcivory and michael R conoverlconovei1Conoverl

ABSTRACT ABSTRAC r we examined patterns of habitat use hby greater sandhill cranes critsgruscrusgrits canadensiscanudensis tabidababida in the intermountain west april october 1991 92 to determine whether cranes exhibited a specific preferencepieferenceferenee foiformol crops fields and areas within a field this information will help farmershalmers and wildlife managers direct nonnonlethallethal control meth- ods to the where is likely sites sheiewheie crane damage is most to occur we conducted surveys along two 37 km transectstiansects weekly in cache valley utah and biweekly in bear river valleyvailey rich county utah and lincoln county wyoming we recorded 5814 cranes in 662 separate groups most were located in pastupasturehavpasturebaypasturerebayhav 34 small grain 39 alfalfa 9 plowed 9 fallow 4 or corn 1 fields an index of feeding activity for each field and habitat type suggested cranes fed at approximately the same rate in each field and habitat type crane diurnal activity patterns during summer and fall leveieverevealedaled that giamgragramgiamfieldsgrainfieldsfieldsinfields were used heavily throughout the day

key words greater sandhill cranes grus canadensis tabidababida habitat depredation diurnal activityactiuti utah wyomingWyonling

the most recent population estimate for awsows and grainfieldsgrainfields ranged from 69 to 100 the rocky mountain greater Sandhisandhillfl crane rowland etalet al 1992 is 17000 20000 drewien et al 198727 we examined habitat preferences and for- records of local summer populations are less aging habits of summer resident sandhill complete but the crane population in cache cranes because of increasing depredation valley utah has increased from 14 individuals complaints from farmers growing corn and in 1970 drewien and bizeau 1974 to approx- small grains ege g barley oats rye wheat in imately 200 in 1990 bndgerlandbridgerlandBridgerland audubon cache and rich counties utah As one means society 1990 between 1985 and 1987 of evaluating these problems and potential rowland et al 1992 reported 255 cranes sum- solutions we tested the hypothesis that crane mering in lower bear river valley wyoming use was concentrated in corn and smallsmail gram crop grain depredation complaints attributed to fields inm particular and inm agricultural fields in cranes are rising concomitantly with population general high use of a field may alarm a numbers lockman et al 1987 in response to farmer but little damage may occur if birds depredation complaints wyoming instituted a are not foraging hence we also tested the limited sandhill crane hunt in 1982 utah in-in hypothesis that cranes forage in habitats in stituted a hunt in 1989 but the decision gen- proportion to their availability in addition we erated enough public controversy that the assessed whether habitat use varied diurnally hunt was canceled in 1992 during summer and fall additional questions cranes are omnivorous mullins and bizeau relevant to selecting an appropriate scale for 1978 and readily feed in agricultural lands management include 1 whether cranes use although habitat use seems to vary widely all fields available to them or concentrate their agricultural fields comprised 91 of habitat activities min a few fields and 2 how cranes used by winteringwintenng cranes in western texas distribute their activities within fields iverson et al 1985 during spring staging in nebraska krapu et al 1984 reported that METHODS 70 of habitat use was in agricultural lands within agricultural fields 99 of use was in corn the studastudvstudy area is in cache valleyvailey utah and stubble approximately 80 of spring diurnal bear river valleyvailey utah and wyoming and habitat use in alaska was in barley iverson et includes three contiguous counties cache and al 1987 in wyoming crane use of wet mead rich counties in northern utah and lincoln

department offisheriesrisheriesoftiituriisrishfishpish erieserles and Wildwildlifelif uta statstate uiumversihltv logalopin uta 84322

329 330 GREAT BASIN naturalist volume 54

countecounucountv inm southwesternsouthwestein wyoming A compre- bvby the distance within which 90 of all cranes hensive description of the area is included in had been located during weekly surveys melvolmcivor and conover 1992 cranes normallynoinol mally an index of feeding activity was developed occupy the region from april untilmitil early to allow comparison among habitat types october when I1 crane was sighted an individual was to determinedete imine patteinspatternspatteins of field use we chosen at random from the flock and observed established a 37 km transect in cache valleyvailey foiforboibol 1 minmm to determine if the bird was feeding and another in bear river vallevvaclev the tran- the result was a logical variable feedno sects traversed a sample of habitat types avail- feed and these data were compiled and com- able to cranes including cultivated fields pas- pared across habitat types tures and natural habitats sampling was con- quantitative analyses were based on meth- ducted based on a visual survey method simi- ods devised by neu et al 1974 we used a lar to that used bvby iverson et al 1985 1987 goodness of fit test P 05 to examine the transectsTian seetssects were sunevedsurvesurvivedsurvevedved weekly in cache hypothesis that cranes used habitats in pro- vallevvaclev and biweekly in beaibealbearhear river valleymailey from portion to habitat availability and to deter- aprilaprifthroughthioughthrough mid october 1991 92 mine whether cranes fed preferentially in cer- surveys began 2 h attelaftelafterahneranner sunrise flomfrom a tain habitat types we used a BonbonferrombonferroniBonferferronirom Z sta vehicle moving at 40 kahkmh habitats on both tstctisticcistic to test foifolformoloor habitat and feeding prefer- sides oftheodtheof the transects were scanned systemati- ence the Z statistic and resulting family con- oor callyeallyeailycal As cranes were located a variety of para- fidence interval foifolfor testing each contingency table cell were generated a meters including type of habitat in use weiewelewere using monte carlo sampling simulation from a binomial distribu- recorded habitat was categorized byb clopelopcrop on a mainframefiameframe alfalfa corn small grain pasture hay or mixed tion mam computer using mmitabminitab 1989 use 01or groundgioundcovergroundcover ty cover typeI1 Pe riparian sage arte- in 1992 we mapped the distribution of misiamisla sppapp scrub to examine the distribution grain and cornfields along the we of cranes within fields we recorded the dis- survey transects then compared distribution available tance between field edge and the individual of fields with the frequency distribution of cranes eranecrane closest to the edge using a range finder observed to test the HQ of equal use among all these data produced a distance to edge esti- ho gram- and cornfields data were ana- mate which we used as a general indication of these weiewele lyzed using a goodness of fit test whether cranes preferentially used edges or patterns of diurnal habitat use were recorded interiors of fields using the range finder we over a 5 d period inm june 1991 using both tran- also recorded distance flomfrom the minimum sects and during september 1992 using only the to the flock tiantransectseetsect crane cache valley transect data collection methods each sighting of cranes was equal given were identical to those used in the habitat use weighting in tables 1I in constructing contingency survey described above except that transects to statistical maintain independence among weiewelewere sampled 5 timesday sunrise 2 h aftelafterahter field use observations A few observations of sunrise noon 4 h before sunset and 2 h before cranes were made in mixed use fields and on sunset rural roads these sightings were combined we used PCSASPC SAS SAS institute inc 1988 under the miscellaneous category early sea- and the PROC CATMOD routine to examine son hayfields weiewelewere difficult to distinguish from june 1991 diurnal use data and the PROC pastures and these observations were pooled FREQ routine to examine september 1992 habitat availability was quantified along data both SAS routines used a goodness of fit each transect in july 9911991 and 1992 A sample test P 05 to examine the null hypothesis of 125 landomrandom points on each transect was that cranes maintained the same pattern of selected a priori and each point was located field use throughout the day and its habitat type recorded to be selected as representative of habitat each point had to RESLLIS meet two criteria first advanvany sampling point not visible flomfrom the transecttianseetsect was not used fiftyfletyfiftfiat thieethree surveys were conducted in cache second the perpendicular distance from tran- vallevalleyvaileymailemalie and 29 in bear river valley during two sect to sampled habitat locations was bounded field seasons we recorded 5814 cranes in 662 199419941 SANDHILL CRANE HABITAT preference 331

groups most groups were observed in pas 1991 92 a crop not cultivated in bear river tureturehayhay 34 small grains 39 alfalfa 9 valley analysis indicated variation in habitat plowed fields 9 fallow 4 or cornfields availability between years along each transect 1 remaining cranes were located in ripari- although the change was not statistically sig- an 3 sagebrush 1 and miscellaneous nificantnificant P 2230 for these reasons col- 2 habitats lapsing the contingency tables across sample habitat availability differed between the sites or across vearsyears would have made the two survey transects table 1 although the results ambiguous cache valley transect contained no sagebrush cranes were not distributed randomrandomlvrandomlylv habitat the beaibealbear river valleyvailey transect con- among nine available habitats inm either 199iiggi1991 tained extensive sagebrush 61 in 1991 58 xax2 3740374.0374 0 df 13 P 005 or 1992 xax2 in 1992 conversely the cache valley tran- 4641464.1464 1 df 14 P 005 along the cache sect contained a small amount of corn 7 in valley transect cranes avoided alfalfa and

TBLETABLE 1 habitat availability use and selection among sandhill cranes in cache vailesvalleyvaileyvallevailevalie C utah and beaibeatbear river vallevailevalievalleymailey B utah and wyoming in 1991 and 1992

habitat studstudy crane expected pi oportionopproportionortionortlon proportion 95 farnilyfarnolyfa mi iv use area obser- crane ofofstudofstunstudastudv observed confidenceconeonfid enientse preferencepreference1preferenceapreferencesa1 vations obser area popio in each intervalmte n ilclieon allonslationsvationsv area piP iouiluyiiggi1991lyyilooi alfalfa C 6 25 0130 0031 0 pi 067 B 14 12 0158 0182 065 pi 325 0 corn C 6 13 0065 0031 0 piP 067 0 B 0 0 00 00 jtbntbnt iintbjtb fallow C 1 4 0023 0005 0 pi 021 B 0 0 00 00 ntilantilnt11 nntbjtb grain C 60 34 0174 0308 205 pi 405 B 18 7 0096 0234 104 pi 364 mise C 4 30 0152 0021 0 pi 051 B 1 3 0044 0013 050 pi 065 0 pasture C 87 70 0359 0446 346 pi 544 0 B 35 5 0070 0455 299 pi 610 plowed C 22 6 0033 0113 051 pi 174 B 1 1 0009 0013 0 pi 065 0 riparian C 10 13 0065 0051 010 Ppi 097 0 B 6 1 0009 0078 013 pi 169 sage C 0 0 0 0 ntbjtbnt11 jtbntbnt1 B 2 47 0614 0026 0 pi 091 1992 alfalfa C 29 61 0194 0092 051 pi 153 B 9 9 0121 0123 027 pi 246 0 corn C 2 20 00065 0006 0 pi 022 B 0 0 00 00 nth ntbjtb fallow C 20 24 0075 0064 025 pi 102 0 B 3 1 0010 0041 0 pi5pispiapi 123 0 craingraingram C 144 57 0183 0459 382 pi 535 B 32 4 0061 0438 274 pi 616 mise C 7 34 0108 0022 003 pi 051 B 0 1 0020 00 ntntbjtb nt11 pasture C 78 84 0269 0248 172 pi 322 0 B 25 11 0152 0342 192 pi 507 4 plowed C 28 14 0043 0089 051 pi 134 B 5 1 0010 0068 0 Ppi 164 0 riparian C 6 20 0065 0019 003 pi 048 B 1 4 0051 0014 0 pi 068 0 sage C 0 0 00 00 nth nt1ntanty B 2 42 0576 0027 0 pi 096 habitHibithibitdthabitatdt use isis expressed isas selection for use in proportion to availabilityi lilahilitv 0 and avoidance notot tested tinltinsthis habitat type was not recorded in the studstndsand area 332 GREAT BASIN naturalist volume 54

miscellaneous habitats in both years selected transect primarily with newly planted corn grain and plowed habitats in excess of their crops cranes pulled up corn plants and con- availability and used pasture in proportion to sumed the still attached seed farmers also its availability along the bear river transect reported minor damage from cranes trampling cranes avoided sagebrush habitat selected emergent alfalfa and small grains winter grain and pasture habitats and used alfalfa wheat barley oats the growing season along and plowed habitat types in proportion to the bear river transect in rich and lincoln their availability results from other habitat counties is too short for corn production and types along the two transects either varied crop damage occurred primarily in the fall between years or weiewelewere not tested due to pat- affecting small grain crops lockman et al terns of sampling or structural zeros in the 1987 mcivormelvor and conover 1994 some tram- contingency tables pling damage in spring was also reported in we examined distribution of cranes using this area graingram- and cornfields in 1992 and found that gianescranesgranes concentrated activities in smallsmail graingram certain gramfieldsgrainfieldsgragram fieldsinfields received preferential use in fields during our surveys fields planted in cache valley xax2 2724272272.44 df 72 P 001 corn constituted only 7 of available habitat and in bear river valley xax2 42642 6 df 10 and 3 of cranes sighted were in corn most P 001 insufficient lataiatadata were available for activity in cornfields occurred during germi- cornfields in 1992 cranes tended to exploit nation or while plants were young thereafter field interiors but weiewelewere broadly distributed cranes avoided cornfields until harvest within fields in 1991 92 mean distance to large expanses of sagebrush habitat were field edge for flocks in corn was 82282.282 2 in n little used although they constituted about 7 SE 21221.221 2 and 7217272.1791 1 in n 250 SE 60 of available habitat sagebrush habitat 7267 26 foiroirolfor flocks using gragramgramfieldsgrainfieldsfieldsinfields may have reduced crane foraging efficiency bvby cranes were recorded feeding in 75 of creating dense cover limiting movement anand our observations A goodness of fit test was offering few plant foods agricultural fields in used to examine the distribution of cranes bear river valley were surrounded by vast feeding in each habitat tytapetvpepe in compcomparisonarisonarlson to expanses of sagebrush a condition that may habitat availability table 2 feeding cranes have concentrated cranes into agricultural were not distributed randomly in 1991 xax2 fields 2428242.8242 8 df 13 P 0005 or 1992 xax2 feeding activity closely approximated pat- 3324332 4 df 14 P 0005 distribution of terns of habitat use suggesting cranes fed with feeding cranes approximated distribution of all the same intensity in each habitat tapetvpetype migrat- cranes observed except in the case of riparian ing cranes in nebraska relied on a diversity of habitat along the bear river transect while habitats to provide various components of cranes used this habitat type disproportionate- their diet meckemeekereineckeRe and krapu 1986 alfalfa ly to its availability in 1991 they appeared to fields walker and Schemmschemnitztz 1987 and grass- feed in this habitat type in proportion to its lands reineckeRemeekemecke and krapu 1986 provided a availability data for 1992 were insufficient for source of invertebrates for cranes although analysis invertebrates mavmay provide certain proteins crane diurnal use of field types varied with absent from plant foods reineckeRemeckemeeke and krapu time of day summer diurnal sampling x2xa 1986 they comprise only a small component 910491.0491 04 df 48 P 0002 fall diurnal sam- of the diet varying from 3 reineckeRemeekemecke and pling x2xa 726572.6572 65 df 24 P 01 crane krapu 1986 to 27 mullins and bizeau 1978 numbers peaked after sunrise decreased in this study cranes appeared to avoid feeding steadily throughout the day and then in cache valley alfalfa fields possibly obtain- increased again before sunset ing invertebrates from pastures or plowed fields in bear river valley cranes fed actively discussion in pasture corn reineckeRemeekemecke and krapu 1986 and cere- crop depredation attributed to cranes was al grains krapu and johnson 1990 provide reported by farmers in cache rich and important nutrient sources for fat synthesis in lincoln counties mcivormelvormcivor 1993 crane dam- cranes habitat use and feeding activity in age occurred in spring in the cache valley gramgragramfieldsgrainfieldsfieldsinfields along both transects and in both 199411994 SANDHILL CRANE HABITAT preference 333

TBLETVBLE 2 distribution of sandhill cranes observed feeding in various habitat types in cache valley C utah and beaibeatbear river valleyaileyalley B utah and wyoming in 1991 and 1992

habitat study crane expected proportion proportion 95 familyfami h use area obser crane of study observed eonconconfidenceid sibdib e preference a y ationslationsvations obser- area plopiopo in each inteintervallclielleileonn vations area pip iggi19911qq1 alfalfa C 5 17 0130 0037 050 pip 090 B 11 9 0158 0186 068 pi 356 0 corn C 4 9 0065 0030 005 pi 075 0 B 0 0 00 00 nthn nth fallow C 0 3 0023 0 nth ntnth B 0 0 00 00 nth nth grain C 44 23 0174 0328 209 pi 440 B 14 6 0096 0237 102 pi 407 mise C 2 20 0152 0015 050 pi 045 B 1 3 0044 0017 005 pi 068 0 pasture C 57 48 0359 0425 313 pi 560 0 B 27 4 0070 0458 271 pi 644 plowed C 18 4 0033 0134 052 pi 239 B 1 1 0009 0017 005 pi 082 0 riparian C 4 9 0065 0030 0 pi 5 075 0 B 4 1 0009 0068 0 pi 169 0 sage C 0 0 00 00 nth nth B 1 36 0614 0017 050 pi 085 1992QQ alfalfa C 20 43 0194 0091 041 pip 155 B 7 6 0121 0149 021 v 319 0 coincolncorn C 1 14 0065 0005 0 pi 023 B 0 0 00 00 ntnth nth fallow C 13 17 0075 0059 023 pip 100 0 B I1 1 0010 0021 0 pi 085 0 gram craingrain C 106 40 0183 0482 391 pi 577 B 19 3 0061 0404 213 pi 617 mise C 2 24 0108 0009 050 pi 032 B 0 1 0020 00 nt nf pasture C 54 59 0269 0245 1735173 pi 327 0 B 15 7 0152 0319 170 pi 532 plowed C 22 9 0043 0100 050 pi 164 B 4 1 0010 0085 005 pi 234 0 riparian C 2 14 0065 0009 050 pi 032 B 0 2 0051 00 ntn nfnt sage C 0 0 00 00 nth nth B 1 27 0576 0021 050 pi 085

habitat use isis expressed as selection fortor use in proportion to availabilityliability 0 and ivoidanceavoidance not tested this habitat tapetype vaswas not recorded in the studsstudysiudy area I1 kotnotot tested inofficiiniuflicientt observed irqinnciesfequenci to test iotheihpotlieilotheliotherIotheithel years were greater than expected although crane problems iverson et al 19874561987 456 midseason gragrainfieldsinfields are unlikely to provide reported that over 90 of the variation in dis- dietary components other than invertebrates tributiontribution of staging cranes min nebraska could cranes probably forage for waste grain in spring be explained by the composition and juxtapo- stubble and for ripening and waste grain sition of essential habitat types certain fields before and after fall harvest in our study area mavmay receive chronic use certain grainfieldsgrainfields and possibly certain because of thentheir proximity to othelother habitat cornfields are more attractive than others to types such as wetlands and roost sites or cranes any burden imposed on the agricul- because they possess characteristics that tural community by crane depredation is not enhance predator detection and escape shared evenly by producers determining why it is unlikely that crane presence has a certain fields are more attractive to cranes and significant negative effect on productivity of lessening these attractattractantsants mavmay help reduce pasture havhay and alfalfa fields however the 334 GREAT BASIN naturalist volume 54

concentration of cranes in small grain fields and distribution unpublished report logan utah particularly in the fall poses a potential eco- 88ppappappp nomic threat to farmers delayed harvest of DRFIFDRENVIEN R C ANDVND E G BIZEUBIZEAU 1974 status and dis- tributiontrib ution of greater sandhillSandbill cranes in the roekrock grains in fall due to wet weather is likely to mountains journal of wildlife management 383sas exacerbate the problem because standing 720 742 graingram remains available to an increasing num- DRFIFDRENVIEN R C W M bbownBROWN ANDD J D VVRLEIVARLEY 1987 of the greater sandhill crane in yellowstone national ber prestagingprestaging cranes lockman et al park 1987 a preliminary survey proceedings of the north american crane workshop 4 27 38 diurnal changes in habitat use may allow IVERSON1 ersonEPSON G C P A vonsVOHS ANDD T C TCHATACHA 1985 cranes to forage while minimizing heat stress habitat use bybv sandhillSandbill cranes wintering in western cranes using pasture and hayfields in midafter- texas journal of wildlife management 49 1074 1083 noon were probablprobably loafing before feeding 1987 habitat use bv mid continent sandhill cranes during spring migiamigrationtion journal of wildlife priorenoipnoi to sunset for reasons that are unclear management 51 448 458 activity patterns observed in cache valley JOHSGARDJOHNSGARD P A 1991 crane music a natural history of were less distinct in bear river vallevvaclev cranes american cranes smithsonian institution press mavmay have moved less visiting fewer habitat washington DC 136 appp KKPLkrapuKRAPL G L woannAND D H JOHNSON 1990 types as a result of the of habitat JOHSON conditioning of pattern distri- sandhill gianescranesgranes during fall migration journal of bution in bear river valleymalleymailey additionally the wildlife management 54 234 238 bear river valley survey may have included a krapuKRAPL G L D A fufkcein E K FRITZELLFBITZELL ADannAND D H greater proportion of paired individuals JOIINSONJOHNSON 1984 habitat use bvby migrant sandhill which remained on during early cranes in nebraska journal of wildlifeoflvvildlife management territories 4840748 407 417 summer johnsgard 1991 and subsequently LOCKMANLOCKMVDD CLC L SERDIUKSERDILK AND R DREUEDREWIEN 1987 an visited fewer habitat tytypesPes expelexperimentalimentalamental greater sandhill crane and canada crane depredation occurs under two dis- goose hunt in wyomingW oming proceedings of the north crane parate conditions in association with american workshop 4 47 57 spring x1civor D E 1993 planting of and fall mciormacior incidence and perceptions of corn just before harvest of sandhill crane crop depredations unpublished cereal grains encouraging rapid germination master s thesis utah state university logan 75 appp of corn and early harvest of grains would mini- MCIVOR D E ANDD M R CONOERCONOVER 1992 sandhisandhill mize availability of these resources to cranes crane habitat use in northeastern utah and south- during periods of susceptibility to depreda- western wyoming proceedings of the north crane american crane workshopnvorkshop 6 81 84 tion damage was concentrated in a few 1994 impact of greater Sandhisandhillfl cranes foraging fields rather than being evenly distributed in on corn and barleybarlebarie crops agriculture ecosystems all fields indicating that nonnonlethallethal techniques and environment 49 233 237 to alleviate these problems need to be focused 1MIMITBI 1989 minitab statistical software release 727 2 minitab inc state college pennsylvania in these same fields farmers who experience MJLLISMULLINS W H AMannAND E G BIZFLBIZEAU 1978 summer foods chronic depredation problems may wish to of sandhill cranes in idaho auk 95 175 178 consider the economic feasibility of producing NEU C W C R BERSBYERS ANDNDJJ M PEEK 1974 A tech- crops less prone to crane damage nique for analysis of utilization availability data journal of wildlifeofwildlife management 38 541 545 acknowledgments REINECKE K J AADAND G L KRAPU 1986 feeding ecology of sandhill cranes during spring migration in nebraska journal of wildlife management 50 project funding was provided by the utah 71 79 state university wildlife damage management ROLNDROWLAND M M L KINTER T BANKS ANDD D C program lockmanLOCKMA 1992 habitat use bv creatergreatelgreater sandhill and the utah division of wildlife cranes resources we in wyoming proceedings of the north thank D A brink K V dustin american crancranegrane workshop 5 82 85 and H low for field assistance R S krannich SAS INSTITLTEINSTITUTE INC 1988 SASSTAT user s guide J A bissonette and P M meyers provided release 6036 03 edition SAS institute inc cancarsgangaocary north helpful manuscript reviews and S L durham carolina 1028 appp WLKERWALKER D L ANDD S D schemnitzSCHEMNITY 1987 food habits gave invaluable assistance with matters statis- of sandhill cranes in relation to agriculture in cen- tical tral and southwestern new mexico proceedings of the north american crane workshop 4 201 212 literature CITED received 2 september 1993 BRIDbridgerlandbrlugfcrlandGERLAND ALDLBONAUDUBON SOCIETY 1990 resident sandhill accepted 20 april 1994 crane populations in cache county utah numbers great basin naturalist 544 D 1994 appp 335 341

SELENIUM geochemical relationships OF SOME NORTHERN NEVADA SOILS

stephen poolelpoole1pooled glenn crosslgrossl and robert pottslpottelpotts1

ABSTRACT soil samples one from each of 10 locations in northern nevada were evaluated for redox potential total and extractable selenium phosphate free iron oxide total and ferrous iron mole fractions for extractable selenium species were calculated from redox potentials data were used to extrapolate general geochemical relationships for soil selenium at the sample sites results obtained from one sample per location allowed only the most general conclusions to be drawn soil phosphate levels which affect the adsorption of selenite species on iron oxide bvby competing for adsorption sites were not correlated with levels of extractable selenium in this study this would suggest that selenium would exist in solution having been displaced from adsorption sites by phosphorus ferrous iron iron oxides and redox potential had a combined effect on the level of extractable selenium at all sites soils in this study support selemteseleniteselemie species that are not readily available to plants and therefore could not support vegetation adequate in se

key words selenium soil redox potential geochemistry plant availabilitybioavailabilitybio

selenium se is a significant micronutrient dominant mobile forms in a soil solution and in production agriculture because of this are available for plant uptake knowledge of the se status of rangelands is redox potentials are important in soils and important distribution of total and extractable theoretical relationships can be used to pre- se can vary widely over short geographic dis- dict and interpret metal solubilities lindsay tances fisher et alA 1990 because the geology and sadiq 1983 redox potentials have been of nevada is complex relationships between used in nevada to interpret observed critical plant se levels and geological forma- sequences of minerals in an alteration zone in tions are difficult to define recently a leviewreview ely nevada raymahashayraymabasbayRaymahashay and hollard 1969 of the se status of soils plants and animals in interpret hydrogeochemistryhydrogeochernistryhydrogeochemistry of the red rock nevada reported deficiency problems in west- nevada area fricke 1983 and evaluate trace ern nevada variable amounts in northern and element content of sediment and water in west central portions of the state and adequate lev- central nevada rowe et al 1991 soil redox els in the southern portion of the state potential data are lacking for the state selenium accumulator plants grow throughout the purpose of this study was to investigate nevada on limited seleniferousselemferousberous geological for- soil se geochemical relationships for 10 mations poole et al 1989 fig 1 the nannarrowow nevada sites using redox potential pe ph gap between essential and toxic concentra- and extractable and total se levels phosphate se tions of makes it imperative that processes P iron fe and iron oxide feofe203 levels controlling the distribution of this element be were also investigated to determine their understood mcneaimcnealmeneal and balistrieri 1989 effect on se availabilitybioavailabilitybio for plants growing uptake of se by plants is governed by many on the soils soil and plant factors including type of plant soil ph clay content and mineralogy most experimental PROCEDURE important factors determining uptake are form and concentration in the soil chemical form A soil sample was taken from each of 10 is controlled by redox potential parameters pe sites battle mountain and gund ranch in ph rashidielrashidiEl et al 1989 mikkelsen et al central nevada eureka and lander counties 1989 although se may exist in four oxidation minden douglas county reno red rock states selenate VI and selenite IV are pre area north of reno spanish springsspi ings washoe

sierrasierra environmental monitoring inc 1135 financialFin amial blvd reno nevada 89502

335 336 GREATGREVT baslbasinBASI naturalist volume 54

97

10 9998 3 92 94 066

5

las vegas

fig 1 selenium in nevada forage very low 81 of fig 2 soil sample locations in northern nevada samples with se concentration of ooi0010 01 0050 05 ppm se I1 battle mountain 2 gund ranch 3 spanish springs variable 74 of samples with se concentration of 4 reno 5 minden 6 fallon 7 salmon falls creek 0050 05 050 5 ppm se adequate 78 of samples with se 8 huntington valleyvailey 9 clociocloverer valleyvailey 10 red rock concentration ofosofo1ofo 1 101 0 ppm se county fallon churchill county salmon calculated from millivolt readings using the falls creek near contact and two locations relationship pe ehmillivolts59ehmillivolts592 2 near the ruby mountains elko county in to obtain total soil se levels we digested nevada fig 2 samples were taken approxi- samples in aliquoaliquotsts of 111 1 hydrochloric acid mately 12 15 cm below the surface so as to in- and 5 potassium perpersulfatesulfate for 15 minmm fol- clude the root zone air dried samples 2 minmm lowed bvby 353.5333 5 oxalic acid solution for 15 minmm no 10 were used for analysis the resulting solution was then treated with redox potentials were measured according concentrated hydrochloric acid foifolfor 42 minmm to the procedure of lindsay and sadiq 1983 prior to diluting to 100 ml volume with soil suspensions were prepared in conical ionizeddeionizedde water se concentrations of the flasksflacks to contain 50 g air dried soil and 100 digests were determined using hydride gener- ml deionized water each treatment was pre- ation atomic absorption spectroscopy AAS pared in duplicate degasserdegasseddegassed with argon ar varianvarlanvananmanan spectralspectraaSpectrAA 10 with VGA accessory stoppered and shaken millivolt readings soluble se was measuredmeasuied in a saturation were taken on soil suspensions with a plat- paste extract from each soil jump and sabesabey inum pt electrode and a glass agagclagagc1 refer- 1989 using hydride generation AAS the ence electrode using an altex selectionSelect lonion 5000 extract se concentration was used for calcula- the platinumreferenceplatinum reference electrode system was tion of se species the mole fraction of soluble standardized using a ferrousferricferrous ferric ion refer- se species was calculated employing methods ence solution zobellszobellaZoBells ASTM 1978 soil sus- ofehashidietalof Elrashidi et al 1987 pension ph was determined using a combina- bicarbonate extractable P was determined tion electrode that was calibrated with stan- using the method of olsen council on soil dard buffers ASTM 1978 suspension pe was testing and plant analysis 1980 to evaluate 199411994 SELENIUM IN NORTHERN NEVADA SOILS 337 fe03fe203 levels we extracted 4 g soil overnight TABLE 1 total selenium extractableextractabletahletabie selenium redox with 4 g sodium dithionite and 75 potential pe ph and predominant selenium species in n08304na2s204 soil samples ml ionizeddeionizedde water suspensions were fil- tered brought to volume kilmer 1960 and seleniumseltseitseleicumimum analyzed for fe by AAS total soil fe was sample total Extracextractableextractahletabletahletabie predominant determined bvby flame AAS on nitric acid location ingingkgmgkgingkekg lugjugjugkgggkg pe phpil species digests ofsoilof soil samples battle ferrous iron fe II11 in sample soils was mountainMoun taina 062 4 100 selenite determined colorimetrically samples were gund digested using concentrated sulfuric acid and ranchranch11 074 8 117 scieScleselenitenite spanish 4 30 hydrofluoric acid neutralized with Springspringssc 010 102 757 5 selenite boric acid and made to volume with deion reno oil011 2 iiiili11111 1 izediced water walker and sherman 1962 to an Minden 1 020 50 109 seleniteselescieSclenite aliquot of the digest we added 00010.0010 oolooi001 M batho fallonfailonfallone 12 3 ililii11111 1 selenite salmon phenanthroline in 50 ethanol and acetate creekfcreek1creeke 037 26 85 selenite buffer isoamyl alcohol extracted the ferrous clovers 013 2 76 is phenanthrolinebathbathophenanthrolinebatho ophenanthrol inelne complex from the solu- Huntinghuntmgtonshuntingtongtong 010 2 96 tion the alcohol layer was drained into a 25 red rockhbockhrock11 010 7 119 selenite ml volumetric flask made to volume with airir dried basis could not belielyeiye calculated dmdue to lack ofineasurableof me isiirabk selenium inin extract 95 ethanol and the absorption of the solu- asteartstew artantann and mckeevickee 1977 tion read on a spectrophotometer baush and tinaisnaILUIDASCSUIDAsiwSIM sosSCS 1978 lsdse1980USDASEAUSDA SEA 1980 LSDX SLSSCS 1983 lomb spectronicSpecspectromcspeetronictronie 710 at a wavelength of 538 1 USDASCSUSDA SCS 1984 ilibildennildenNIl den and speed 1974 nm standards and blanks were treated simi-simisiml fseliraderchrader 1934 luellertheller 1975 larly ferrous iron standards were derived gTheileroelleruellertueller 1975 huso SCS 1983 nekeineke1 1983 from a stock solution of ferrous ammonium sulfate for the interpretation of data we used 8 adsorption decreases to complete desorp- redox and adsorption relationships developed tion at ph 11 selenite shows a strong affin- by howard 1977 bahstnenbalistrieri and chao 1987 ity for fefeo203 surfaces balistrieri and chao 1990 and schwab and lindsay 1983 for the 1987 forming stable ferric oxide selenite behavior of se fe and POp04 and equilibria fe2oh4fe2oh4se03se03 complexes that causecanse immo- described by Elrashidi et al 1987 for se in bilization of se selenate on the other hand soils shows a weaker affinity for oxide surfaces regression and multiple regression analy- forming compounds that are soluble and ses were performed following methods of therefore mobile howard 1977 Elrashidi et damon and harvey 1987 regressions were al 1987 presser and swain 1990 and easily evaluated for significance at the 95 confi- transportable in groundwater and available for denceleveldence levellevei plant uptake lakin 1961 levels of extractable feofe203 are presented RESULTS in table 3 because adsorption of selenite increases with increasing concentration of total se extractable se redox parameters fefe03203 due to the greater number of available and general site descriptions are presented in binding sites bailbalistrieriBali strien and chao 1987 it table I1 for sample soils mole fractions of the follows that soils at gund ranch and spanish se species for each sample are presented in springs have the potential to adsorb the table 2 largest amounts of selenite the clover valley howard 1977 summarized se geochem- sample would be least likely to adsorb selen- istry on an eh ph diagram and found that fe ite ph would not have an effectekecteject on the ability with which se is closely associated in both oxi- of fefe03203 to adsorb selenite for all the soils dizing and reducing environments controls se except clover valley ph 8 levels of geochemistry in aerated soil suspensions the feofe203903 and fehfeilfell had a combined effect on the se IV oxyanionsoxyanions hsehsechaec03 and sec2secase032 are amount of se extracted from soils r 3196.3196 strongly adsorbed by hydrated surfaces of fer- iron oxide and fehfelifeilfen were not affected by ricrie oxides over the ph range 2 8 above ph redox potential r 0705.0705 338 GREAT BASIN naturalist volume 54

TABLE 2 log mole fraction of selenium species 1 species

1 seca se2sea SamsampieSampleplell sec4se04sectseca hsec4hse04 hse04h2se04 sec3se03 hsec3hse03 hasecchasec112seo 3 hseilseiiselise haseh2sehgse 1 92 149 292 015 055 56 229 156 196 2 59 123 226 002 132 70 333 266 311 3 122 179 244 018 048 73 139 63 104 4 74 123 212 055 015 45 299 219 249 5 61 107 191 086 006 39 351 266 299 6 75 129 222 030 030 50 285 208 253 7 199 198 288 059 039 49 839 05 379

bastbasebasc 10 logarithm ISasamplesampieniple identification I1 battlebatilebattie mountain 2 cundgund ranch 3 salinonsalanonsa i creek 4 mionfallonhailon 5 redreci rock 6 minden 7 spanish springs

activity of feiifelifellfebi is controlled by fec03 mgkg P to evaluate the effect of P on selenite dentesisiderite at pe ph 8 and by fc3oh8fe3ohg adsorption we calculated total anion concen- ferrosicferrosic avdihydroxidehvdi oxide at pe ph 8 in svsavssys tration ratios anionselemtefanionselenitel results aiealeare tems below ph 60 with stable ledoxredox less presented in table 4 A stronger affinity and soluble iron oxides such as grothitegeothite alpha larger concentration of one anion should result feooh can control fe solubility schwab in more sites being occupied by that anionanlon vs and lindsay 1983 levels of feiifelifellfebi are pre- another balistrieri and chao 1990 levels of sented in table 3 ferrous iron levels had a P and fe203feo and P did not have an effect on significant effect on the amount of se extract- the amount of se extracted from the study ed from soils r 5843 siderite would con- soils r 3030 and r 1019 respectively trol feiifellfeilfebi activity in the spanish springs and absence of a significant correlation between valley samples in clover ferrous iron activity in levels of fe203 and P and extractable se sug- the samples would be controlled by feo remaining ggestsests that se would exist in solution for these fer would control ferrosicrosic hydroxide hematite soils phosphorus would have displaced se ferric iron of sample soils except for iron activity from available binding sites selenium phos spanish springs and clover valley where feifelfer phate interactions are generally not of conse- roselite would control ferric iron fe III111 any for of se except for plants could be associated with quence plant uptake remaining feiiifeidi se inadequate to hydrous selenite complexes growing where levels of are total fe levels ranged from 8700 to 28000 meet animal nutritional needs mikkelsen et mgkg total se and total fe were not correlat- al 1989 mean total se soils ed for these soils r 1028 redox potential concentration of in and and total fe had an effect on soil felifell and surficial materials for the western united range fe203 content r 4565 and r 3998 states is 0230.230 23 mgkg with an observed 10 4.3434 3 se most soils from low- respectively A decrease in selenite adsorbed of olo0100.100 43 mgkg se 0.50 5 on iron oxide would depend on the adsorption areas in the united states contain 0505 density of selenite moles of ion adsorbedadsorbedkgkg mgkg se national research council 1983 of oxide balistrieri and chao 1987 other boon 1989 A limited survey of nevada soils ions in a soil solution including P can com- as part of a trace element survey of soils pete with selenite for adsorption sites on solid throughout the united states revealed a variety surfaces anion adsorption relies on several of se levels shacklette et al 1974 observa- factors including ph formation of solution tions in the fallenfallonfailon area demonstrate that complexes and competing adsoibatesadsorbatesadsorb ates seleniferous spots may be found in alluvial mikkelsen et al 1989 phosphate displaced pliocene deposits occurring over a large paitpart all the adsorbed selenite on allophane clays of nevada particularly in the carson and rajan and watkinson 1976 and has been humboldt sinks lakin and byers 1948 rowe shown to desoradesorb selenate singh et al 1981 et al 1991 total soil se levels in this study most P found in alkaline soil exists as calci- ranged from olo0100.100 10 to 0740.740 74 mgkg As with um phosphate cahoogahcahp04cahpoCaH PO lindsay and moreno many other elements total concentration of se 1960 boyle and lindsay 1986 phosphate in soils shows little relationship to se concen- levels of study soils ranged from 898.98 9 to 147 tration in plants grown in those soils 199411994 SELENIUM IN NORTHERN NEVADA SOILS 339

TABLE 3 total iron extractable iron oxides and ferrous TABLE 4 extractable phosphate phosphorus selenite iron in soil samples selenium and phosphatephosphateselemtephosphateseleniteselenite molar ratios

milligramskilogramamiilhgramskilo ramlramaram1 sample phosphate selenite phosphatefpbosphate sample location ioeloe10e 5mam5 M ioe10e 7 M selenite I1 location total fe oxide fe ferrous fe battle mountain 34 13 271 battle mountain 17000 4575 3290 gund ranch 35 25 136 gund ranch 28000 12175 6090 spanish springs 15215 2 28928 9 52 spanish springs 19000 8900 7920 reno 239 0 0 reno 27000 6550 5800 minden 14 29129 1 5 minden 19000 3950 10940 fallonfallen 34 13 269 fallon 12000 3700 3010 salmon creek 43 50 85 salmon creek 12000 2525 8580 clover 58 0 0 clover valleyvaileyvallevalievaile 8700 875 7060 huntington 46 0 0 huntmgtonhuntington valley 19000 4350 5300 red rock 41 25 160 red rock 17000 6350 3220 airir dried basis forage at the fallon site would not be borkmannworkmannWorkmann and soltanpourSoltan pour 1980 have expected to contain appreciable amounts of reported that water soluble se is usually 50 se soil redox potential does not allow for for- agkgjugaggkg in normal cultivated soils soils in this mation of plantpiant available selenate in areas study had soluble se levels of 1 102 gkgagkggagju adjoining carson valley including fallon existing primarily as selenite the significant white muscle disease in sheep has been a rec- correlation between pe ph and levels of ognized problem vawter and records 1947 extractable se r 4475.4475 suggests a relation- kuttler and marble 1958 for animals raised ship between the amount of se available for on native forage soil at the gund ranch site plant uptake and soil redox potential at the supports a small fraction of selenate allowing study sites for growth of forage marginally deficient in se certain native plants of the great basin grazing cattle have been found to be border- have tendencies to aggregate in relation to line deficient in plasma se at the gund ranch temperature gradients precipitation patterns site poole et al 1986 upper rangeland for- physiography and soils tueller 1975 reveal age in the salmon creek area would be defi- 1979 approximately 80 of all forage and cient in se because of lack of available soil grain sampled in western nevada has been selenate samples lacking measurable amounts shown to contain 0100100.10 ppm se less than the of extractable se would not support growth of dietary requirement of 0100.10olo ppm for grazing se bearing forage animals kubota et al 1967 mcdowell et al 1983 national research council 1983 soil conclusion se concentration can vary widely over a very short geographic distance fisher and total and extractable se redox potential munshower 1991 upper rangeland forage of ph and P feiifeilfellfebi and fe203feo levels were dif- extreme northeastern nevada growing on ferent for each of the sample sites redox idavada volcanivolcanicscs and silicic rocks of volcanic potential and fehfelifellfen and free eaf6203e0 levels origin was found to contain low levels of se would affect the quantity of se available for carteretcarter et al 1969 in contrast lower range- plant uptake in study soils anion concentra- lands surrounding a portion of these areas pro- tion ratios indicate that P would influence duce forage adequate in se carter et al adsorption of selenite on iron oxide soils in 1968 alfalfa samples taken from the carson this study support selenite species that are not valley area were found to be below 005oos0050.05 readily available to plants and therefore could ppm the dietary requirement of 0010.1oi 1 ppm se not support vegetation adequate in se allaway and hodgson 1964 forage at gund soil se concentration can vary widely over ranch has been shown to contain 0130.13 0170.17 a very short geographic distance nevada s ppm se poole et al 1989 selenium indicator complex geology therefore requires evaluation plants are limited to localized areas on selenif- of the se status of soils and vegetation on a erous geological formations in nevada poole site basis further studies are needed to devel- et al 1989 and are not reported to occur with- op a better understanding of the se status of in sample site areas the state 340 GREAT basenBASIN naturalist volume 54

literature CITED ment soil science societsocietysochet of america special publication no 23 madison wisconsin KILMER V 1960 ALLWallawayALLAWAINA hannHANDH djJ FE HODGSON 1964 smposiumsymposium on J the estimation officeof free iron oxides in nutritionnuti itionaition forage and pastures selenium in foragesfbi ages as soils soil science society proceedings 24 420420121421 related to the geological distribution of muscular KUBOTAKLBOTA J NV H ALLWAIALLANNAY D L carderCARTERCVRTER E E CRCARY dsdystrophytroph in livestocklicilcile stock journal of animalofanimal science 23 ANDn V A LAZAR 1967 selenium in crops in the 271 277 united states in relation to selenium responsive dis ASTM 1978 annual book of ASTM standards paitpart 31 eases of animals lournaljournal of agricultural food chemistry watelwater philadelphia Pennpennsylvaniaslamasiama ChemisChemistchemistnchemistrvtnrv 15 448 453 KLTTI FR K L ANDn NV MARBLE balistrieri L S ANDfd T T CIIAO 1987 selenium kuttler D irbieirbleIRBIL 1958 relationships adsorption bby grothitegeothite soil science society of of serum transaminase to naturally occurring and artificiallyfly america journal 51 1145 1151 artificia induced white muscle disease in calves and lambs journal of veterinary 1990 adsorption of selenium bby amorphous jrironon american research 19 632 hydroxideoxyhydroxideoxyoxyhydroxidesoxhoah dioxide and manganese dioxide geochimica LAKIN NV 1961 geochemistry ctet cosmochimica acta 54 739 751 LKI H of selenium in relation to agriculture selenium in agricultureagn agricultural booBOON D Y 1989 potential selenium problems in great in culture handbook 200 USDA 3 12 plains soils pages 107 121 in L Wnn7nna jacobs ed 312 LKILAKIN H W ANDD H G BVFRSBYERS 1948 selenium occur selenium in agriculture and the emironmentenvironment soil science society ofofamencaofAamericamenca special publication no rence in certain soils in the united states with a 23 madison wisconsin discussion of related topics seventh report USDA technical bulletin 950 washington DC BOYLEu FE W ANDNI WNV L LINDSAY 1986 manganese phos- bl linosayLINDSALINDSAY W L AND E C MORENO 1960 phosphate phate equilibrium relationships in soils soil science phase equilibria in soils society society ofamericaof america journal 50 588 593 in soil science of Aramerica pi oceedingsoceproceedingsedings 24 177 182 CARTERGYRTCR D L M J BROWNBRO 11 ALLAWAY AND nerica u W II anor E E aliw LINDSAYLINDS NV L ANDVND M SDIOSADIQ 1983 use ofofpepe to CRIcrycabyCARY 1968 selenium content ofoffoiageforage and hahay linus ofee ph predietpredictpi edictedlet and interpret metal solubility relationships crops in the pacific northwestsoith west agronomy journal in soils of the 605326033260 532 534 in science lotutotal environment 28 169 178 CARTERCAKTFR D L C NNT ROBINS ADAND M J BROWNBRO 1969 mcdowelllcDoELL L R J H CONRAD G L ELLIS AND J K seleniumSeleniuniluni concentrations in forage on some high loosu 1 1983 mineralsalsais fortoroor in northnortlrw estein ranges journal of range management loosiloolu minelMinei grazing rumilumirumlruminantsluminantsnants in tropical bulletin of the 2323423 234 238 regions university of florida gainesvilleGamescames ville COLVIICOUNCIL 0ON soilsollson TESTING anoANDn planrplaar ANIISISANALYSIS 1980 piifil J M AND L S BALISTRIERL 1989 geochemistryrN pages 47 51 in handbook on reference methods for mcneal blistrieribalistrierlbalistriere Geochemist and of selenium pagepagess soil testing athens georgia occurrence an overview 1 13 inm L NV jacobs ed selenium in agriculture and DAMODAMON R AJRADWRA JR AND NV R HAREHARVEY 1987 experimental the envbenv soil society design ANOVAAN OVA and regression harperharpelharpet and row environmentnormentnonment science of america publishers new york special publication 23 madison wisconsin MIKKELSEN R L A L PAGEpace ANDD F T BINGHAMBINCHAM 1989 ELRSHIDIELRASHIDI M A D C ADRIOADRIANO ANDD NV L LINDSALINDSAY 1989 solubility speciation and transformation of factors affecting selenium accumulation bybv agricu- ltural crops pages 65 94 in L W jacobs selenium in soils pages 51 63 in L W jacobs ed in ed selenium in agriculture and the selenium in agriculture and the environmentenvnonment soil in environment soil science sociersociety of americaamenea special publication no science society of america special publication no 23 madison wisconsin 23 madison wisconsin NATIONAL RESEARCH COUNCILCOLNCIL ELRASHIDI M A D C ADRIANOADRIVNO S Mn1na WORKMWORKMAN ANDD 1983 selenium in nutri- XV L LINDSAY 1987 chemical equilibria of seleni- tion revised edition national academy press um in soils a theoretical development soil science washington DC 144141144 141 152 POOLE S C V R boiibohlaBOHIAMAN L A RHODES ANDVND R FISHER S E ANDD FE FE MUNSIIOCRMUNSHOWER 1991 selenium TORMTORELL 1 1986 the selenium status ofofiangeoriangerange cattle in issues in drastically disturbed land reclamation plan- northeastern and central nevada proceedings of the ning in andaridarld and semiarid environments hiin R C western section of the american society of animal seversonSe erson S E fisher jijr and L P gough eds science 37 220 223 proceedings of the 1990 billings land reclamation POOIFPOOLE S C V R bolBoiBOHMAiMAN ADAND J A YOLNCYOUNG 1989 symposium on selenium in and and semiarid envi-enviedvi reviewrev lev of selenium in soils plants and animals in ronmentsronments western united states US geological nevada great basin naturalist 49 201 213 suiesule circular 1064 prbssirPRESSER T S ahnabnkowvowANDADWW C SAIstisalSWAIN 1990 geochemical evievl frioFRICKEFRIC KE R A 1983 the hydro1 geochemistrydrogeochemistnhydrogeochemistry and aqueous dence for se mobilization bvby the weathering of uraniumurani inniuntun distribution of petersen mountain and red pyritic shale san joaquin valley california USA rock vallevalievaile washoe countcounty eadanevada unpublished applications of geochemistry 5 703 717 master s thesis university of neadanevada reno RAJANRAJ S S S ANDADJJ II11 WVTKINSONWAXKINSON 1976 adsorption of HOWARDho RD J H 1977 geochemistry of selenium formation selenite and phosphate on an allophane clay soil of ferferroseliteferroselitcferrosherrosberrosroseliteelitc and selenium behavior in the vicinitymcimt of science society of america journal 40 51 54 oxidizing sulfide and uranium deposits geochimica riyiallsnvrymaiiasiiay B ADAND H D holiHOLLARDabdkbd 1969 redox reac- et cosmochimicagosmochimica acta 41 1665 1678 tions accompanying hvdiotheimalhydrotherinal wall lockrockloek alter jutejulejuipjutpJLMPRR K tobannANDVDBB R SBEISABEY 1989 soil test extractantsextrac tants for atlon economic geology 64 291 305 predicting selenium in soil pages 95 105 in L W REELREVEAL J L 1979 biogeography of the intermountain jacobs ed selenium in agriculture and the environ region A speculative appraisal mentzelia 4 1 87 199411994 SELENIUSELENIUMSELENIUNINI IN NORTHERN NEVADA SOILS 341 rohrowhroeROWE T G 1MI S lico R J HILLOCKHALLOCK A S iestlestehstXIAESTANDihst vn 1983 soilsodlsodd surveysubsun e of washoeofwashoe county nevada south R J HOFFMANHOITMILN 1991 physicalPh sical chemical and biolog- paipalpartt ical torhorbor study data for detailed of irrigation drainage in 1984 soil sunsurveyc of douglas county aleaarea nevada and near stiistillwaterStil laterlaten fernley and humboldt ildhfewildlife USDASEAUSDA SEA 1980 physical biological and culturalcultuialeuitcultuialulal management areas and carson lake west central sourcesresourcesle of the gund research and demonstration nevada 1987 89 US geological survey open file ranch nevada J A dolingyoung and R A evans eds report 9118591 185 VWTFKVANNTER L R ANDD E RECORDS 1947 muscular dystro- SCHRDERSCHRADER F C 1934 the contact mining district phphy white muscle disease in young calves journal united states department of the interior bulletin of the american veterinary association 110 847 A 152 157 SCHABSCHWAB A P ANDD W L LIDSlindsay 1983 effect of redox lahrWALKERlkhr J L AND G D SIIFRMA 1962 determination solubility on the solubilitsolnbilihv and availability of iron soil science of total ferrous iron in soils soil science 5 325 328 society ofamericaofamerica journal 47 201 205 WILLDENILLDE C AND R C SPEEDSPCTD 1974 geology and mineral shackletteSHACKLETSHCKLETTFTF H T J G boerngeBOERNCEBOERNGEN ANDD J R KHIHKEITH deposits Churchillofchurchillof chin chill county nevada nevada bureau 1974 selenium fluorine and aisenicarseniearsenicbisenic in surficial of mines and geology bulletin 83 university of materialmaterials of the conterminous united states USU S nevada reno geological survey circular 692 WORKMANORKMVN S M ANDD P N SOLTANSOLTPOLRPOUR 1980 sinchsighSICHSINGH M N sinchSINCIISINGH ANDnttantt P S beilbellrelanbeibelREIANRCI v 1981 adsorption importance of reducingpiepiereducingprereducingpre selenium VI and and desorption of selenite and selenate selenium on decomposing organic matteimattelmatter in soil extracts prior to different soils soil science 132 134 141 determination of selenium using hydride genelageneiagenerationtion STERTSTEWART J H anoANDD E H mckee 1977 geology and soil science society of america journal 44 mineral deposits ofoflanderhoflanderlander county nevada nevada 1331 1333 bureau of mines and geology bulletin 88 TLLLLERTUELLER PTP T 1975 the natural vegetation of nevada receivedri ceiied 6 aiwustaugust 1993 mentzeliaMentzeha 1 3 6 23 28 accepted 19 april 1994 USDASCSUSDA SCS 1978 advance soil surveysurevgurev data cundgundguild ranch eureka and lander counties nevada great basin naturalist 544 C 1994 appp 342350

STATUS AND distribution OF THE LARIDAE IN WYOMING THROUGH 1986

scott L Findholtfindholtlfindholt11

ABSTRACTABSTPLACT to date 17 species oflandaeof laridae have been reported in wyoming six of these species have known breed- ing populations in the state the ring billed gull laaslamslants delauarensisdelawarensis california gull larnslamslarus califorcahfornicuscalifornicusnicus herring gull larus argenargentatustatus caspian tern sterna caspiacampiacaspia forster s tern sterna forjorforstensteristert and black tern chlidoniaschhdonias niger of these species the california gull is the most abundant and widespread in 1984 approximately 7300 nests existed in wyoming at six breeding locations consisting of 10 different colonies in contrast only small breeding populations have been discovered for the remaining fiefivefle species the herring gull is the most recent addition among laridae known to nest in wyoming likewise two ring billed gull colonies were recently found after not having been documented as breeding in the state for over 50 years although some nesting colonies aiealeare threatened byb habitat loss and human disturbance most seem secure at present limited nesting and foraging habitat precludes establishment of large breeding populations of most laridae in the state

key words laridaelandue historical records inventory population status distribution breeding wyomingWyorning

considerable interest and concern exist of flight time to locate new nesting areas regarding conservation and management of reservoirs lakes marshes and other potential colcoieolcoloniallyomally nesting waterwaterbirdsbirds in the united breeding locations not observed during aerial states and elsewhere these species occupy searches were checked from the ground with high trophic levels on aquatic food chains and binoculars or a 20 45x spotting scope are sensitive to disturbance of aquatic ecosys- breeding colonies were usually censcensusedcensusesused by tems especially loss of wetland habitat and con- making total ground counts of nests where taminationtamination by chemical pollutants in addition ground counts were not feasible I1 estimated the because most of these species nest min colonies number of nests ground estimates colonies they are vulnerable to human intervention were censcensusedcensusesused when most birds were in late findholt 1984 and fmdholtfindholt and berner incubation or early hatching stages and cen- 1988 reported on the status and distribution suses were based on a single visit of the ciconnformsciconiiformsciconiiformes in wyoming the purpose As discussed by buckley and buckley 1979 of this paper is to provide information on the a waterwaterbirdbird colony is difficult to define theref- historical and present status and distribution ore I1 used kushlan s 1986 definition which is of the landaelaridaecandae in the state an assemblage of nesting birds nests were con- sidered active if adult birds were sitting or METHODS standing on nests incubation was observed or eggs or young were present mccrimmon data collection methods utilized were pre- 1982 viously reported fmdholtfindholt 1984 1986a find- additional sources of information included holt and berner 1988 from 1981 through a literature review an examination of the files 1986 but more intensively during the 1984 86 of the wyoming game and fish department period I1 conducted a comprehensive statewide and correspondence with biologists naturanaturalistslists inventory for colomallycoloniallycolcoieolomally nesting waterwaterbirdsbirds in birdbirdwatcherswatchers and others considered knowl- wyoming from 4 april to 31 may 1984 and edgeableedge ableabie of the laridae in wyoming this from 28 march to 5 june 1986 I1 made 15 aerial paper includes records through 31 december surveys in fixed wing aircraft totaling 6716767.11 h 1986

lwyorninglWyonangorningonnng cainecalnegainecamegame and fish department 260 baenabuenabulena virtivistivista dildriDII e landdlinderlinden Wywyomingorning 82520 present address oregon department of fish and wildlife forestry mdand range sciences laboratory 1401 gekeler lane la grindegrande oregon 9850983097850

342 199419941 LARIDAE OF WYOMING 343

RESULTS AND discussion also A B mickey observed a franklin s gull near lake hattie southwest of laramie albany pomarine jaeger county on 7 may 1933 mccreary 1939 there is one record of the pomarine jaeger oakleaf et al 1982 considered this species a stercorarius pomarinuspomarinus from wyoming J common summer resident and recorded it in and V herold observed an adult individual at 20 71 of 28 degree blocks and breeding in burlington lake goldeneye reservoir 24 one block the only nesting record is from km northwest of casper natrona county on beck lake near cody park county where U 15 may 1980 on the following day 0 K scott kepler found 10 20 nesting Frankprankfranklinlilss gulls in and B stratton saw the jaeger at the same 1977 kingery 1977 in my intensive statewide location and confirmed its identity survey for franklin s gull nesting sites I1 according to the AOU checklistcheck list of north found none at beck lake or elsewhere in american birds 1983 the pomarine jaeger wyoming based on the lack of suitable nest- breeding range occurs along northern coastal ing habitat at beck lake the validity of this areas in north america thus only accidental breeding record is questionable occurrence is expected in wyoming breeding records exist for this species in adjacent states of idaho lamsonlarrison et al 1967 parasitic jaeger C H trost personal communication montana 0 K scott discovered the first parasitic skaar 1980 south dakota johnsgard 1979 jaeger stercorarius parasiticusinparasiticusmparasiticusin wyoming at and utah behle and perry 1975 soda lake casper natrona county on 2 bonaparte s gull september 1962 two more parasitic jaegers were observed at jackson lake grand teton knight 1902 considered the bonaparte s national park teton county on 22 june 1975 gull larus philadelphia a rather rare migrant by M and B raynes on 24 october 1977 H inm wyoming and provided details of several downing and M collins reported one imma- records from the state in addition to knight s ture individual at lake desmet near buffalo records grave and walker 1913 reported one johnson county one year later another individual taken from near sheridan sheridan immature bird was observed at lake desmet county by metz mccreary 1939 indicated by M collins on 28 august an immature para- that the bonaparte s gull was a frequent sitic jaeger was seen at soda lake natrona migrant in eastern wyoming and sometimes county by J and V herold on 14 november common this species has been listed as occur- 1981 on 4 september 1983 G scott found ring in yellowstone national park skinner one individual at bates creek reservoir south 1925 more recently oakleaf etetalal 1982 con- of casper natrona county the most recent sidered the bonaparte s gull an uncommon record of this species is from healy reservoir migrant and reported it from 10 36 of 28 east of buffalo johnson county when H latilonglatilong blocks downing and M collins observed an adult because this species breeds north of the bird on 21 june 1985 conterminous united states AOU 1983 and the parasitic jaeger is mostly pelagic does not appear to be expanding its range breeds north of the conterminous united southward it is highly unlikely that bonaparte s states and generally winters offshore along gulls will be discovered nesting in wyoming ocean coasts AOU 1983 therefore only acci- Heerheermann s gull dental occurrence is anticipated in wyoming mams on 26 september 1984 K scott et al dis- franklinFranklirss gull 0 covered a heermannHeermamsmahs s gull lafuslamslarus beenheenheennanniheermanniheermanninanni the first record of the franifranklinfrankilnfrankiinFrankliisillss gull larus at soda lake approximately 3 km north of pipixcan in wyoming is a specimen collected casper natrona county kingery 1985 this near wheatland platte county on 6 may 1912 is the first record of this species in wyoming grave and walker 1913 mccreary 1939 the heermannHeermaniss gull breeds in the vicinity stated that this species occasionally occurred ofbajaof baja california and is a coastal species rang- in the state and on 5 may 1933 37 individuals ing from southern british columbia south to were seen near torrington goshen county guatemala AOU 1983 thus only accidental some birds remained in the area until 12 may occurrence is expected in wyoming 344 GREAT BASIN naturalist volume 54

mew gull california gull two historical records exist in wyoming for two california gull larus califorcalifornicuscaliformcusnicus the mew cullcuilgull laaslamslarus banuscanus one juvenile nesting colonies existed historically in budbird was collected bvby V ballexbaileyballey on lake foikfolkfork a wyoming one colony was discovered on the tributary of the green river in the wind molly islands yellowstone lake yellowstone rieirielriver mountains sublette county on 28 national park inm 1898 when skinner 1917 august 1893 oberholser 1919 this speci- estimated about 1000 gulls were present the men is located in the USU S national museum other colony which contained an unknown another mew cullcuilgullgukl was taken near laramie number of california gulls was located on an albany county bvby A E lockwood prior to island in bamforth lake about 15 km north- 1913 grave and walierwalkerwaiker 1913 no recent west of laramie albany county since 1934 records exist foiformol this species in the state mcdearvl939mccreary 1939 according to the AOU checklistcheck list of north in 1984 there were six breeding locations american birds 1983 the mew gull breeds consisting of 10 different colonies that included north of the contiguous united states based approximately 7300 nests findholt 1986a the yellowstone on the paucity of reports for this species in six sites included both lake and states that adjoin wyoming only accidental bamforth lake in addition to four recently occupied lif gull occurrenceocchioccui i enceenee is anticipated in the state nesting areas the new cacaliforniaornia colonies are located at pathfinder reservoir ring billed gull carbon county 4223n 10656w ocean in the 1920s the ring billed gull laruslartis lake fremont county 4307n 10835wioslos108 35w delailarensisdelawarensis nested on the laramie plains sand mesa fremont county 4319n albany county and on yellowstone lake 10820wios108 20w and soda lake natrona county yellowstone national park knight 1902 4254n 10618w although california gulls continued to nest at all six locations dur- skinner 1917 kemsiesKenisiessles 1930 it is difficult to six assess when the ring billed gull disappeared ing the 1985 86 period none of the colonies were censcensusedcensusesused also only 5 10 appeared as a breeding species in these two areas of the pairs to be present at sand mesa in 1985 and none state this no longer breeds on the in species 1986 decline in the sand mesa laramie plains raperpaper 1975 findholt personal in the in nesting population is a result of intentional destruc- observation also the ring billed gull no tion of nests by the wyoming game and fish longer nests in yellowstone national park in department to supposedly enhance canada schallerSchallei 1964 and condon 1967 K L diem goose branta canadensis production diem personal communication the overall increase in the callcaliCaticaliforniaforniabornia gull two active ring billed gull colonies were nesting population in wyoming since historical present at two locations in wyoming during times is most likely a result of human induced the 1984 86 period on 21 may 1984 1 I count- environmental changes these changes have ed 102 adults of this 70 species and nests with cleatedcreated additional breeding habitat and new eggs at soda lake 4254n 10618w about food sources findholt 1986a 3 kinkm of casper north natrona county this species breeds in adjacent states of findholt 1986b although ring billed gulls colorado rvderovder 1978 idaho larrisonLarnson et al continued to nest at soda lake in 1985 and 1967 C H trost personal communication 1986 the colomcolony was not censusedcenscensusesused one addi- montana skaar 1980 and utah behle and tional ring billed cullcuilgull nesting colony was perry 1975 found in wyomingwyoniing at ocean lake 4307n 10835wlosios108 35w approximately 24 kinkm northwest of herring gull riverton fremont county on 22 mavmay 1985 I1 knight 1902 considered the herring gull counted 10 adults and 6 nests containing 2 3 larus argentatusargentatus vervvery rare in wyoming and eggs each on peninsula island twenty three noted that there was only one record from the active nests were present on 31 mavmay 1986 state this species apparently increased in breeding records exist in adjoining states of numbers during the early 1900s and was idaho lafrisonlarrisonLarnson et al 1967 C H trost per- reported as being a common summer resident sonal communication montana skaar 1980 at yellowstone lake and in the big horn basin and south dakota johnsgard 1979 grave and walker 1913 later McCremccrearymcgrearyarv 199411994 LARIDAE OF WYOMING 345

1939 considered the herring gull a moder- 1898 two more birds weiewelewere observed at dubois ately common migrant seen around the lakes fremont county on 22 october 1974 by M of thetiietiletlle eastern paitpaltpart of the state and along the back kingeryKingerv 1975 this is the only recent north platte river recently oakleaf et al record of this species in wyoming 1982 reported this species to be an uncommon because the black legged kittiwake is pri- migrant that had been observed in 12 43 of marily a pelagic species and breeds north of 28 degree blocks the contiguous united states AOU 1983 in 1984 three herring gull nests were only accidental occurrence is anticipated in located at bamforth lake albany county B H wyoming pugesek personal communication this is the sabine s gull first record of this species breeding in wyoming one to three pairs of herring gulls mccreary 1939 indicated that the sabine s continued to nest at bamforth lake in 1985 gull xelnaxema sabini is rare in wyoming two and 1986 specimens weiewelewere taken by A E lockwood in although the herring gull has been re- the fall near lakes on the laramie plains ported from adjoining states of colorado baileyballey albany county grave and walker 1913 and neidrichneidrachNe idrach 1965 ryder 1978 idaho another sabine s gull was found dead near larrisonLarnson et al 1967 montana skaar 1980 douglas converse county bvby K cook and A nebraska and south dakota johnsgard 1979 hay on 24 october 1937 mccreary 1939 and utah bebiebehlebehie and pentpenyperry 1975 I1 am unaware since 1954 there have been approximately of breeding records from these states except 24 reports of sabine s gulls in wyoming con- for recent evidence of nesting at antero sisting of 28 individual birds all sightings reservoir park county colorado chase 1987 were made in september and october except glaucous gull for a subadult observed at lake desmet johnson county bvby dalydaiy on 7 1981 J daivW june the first record of the glaucous gull laaslamslarus the sabine s gull hasas been located in 7 25 hyperboreushyperboreus in wyoming is of a bird collected of 28 latilonglatilong blocks and is considered a rare by E isberg at lake hattie albany county migrant in the state oakleafoakleafetet al 1982 on 23 november 1933 mccreary and mickey according to the AOU checklistcheck list ofnoithof north 1935 mccreary 1939 another report of this american birds 1983 the sabine s gull is species by A B klots in mccreary 1930 was primarily pelagic and breeds north of the con- not mentioned later mccreary 1939 possibly tiguous united states thus this species is ex- because the validity of the report was ques- pected to be seen rarelyi arelaarelv in wyoming and then tiotionablenable mostly during migration there are three recent observations of the caspian ternn glaucous gull in wyoming on 23 september teitel 1969 K L diem observed one individual near skinner 1917 observed caspian terns laramie albany county A second glaucous sterna caspiacaspidcampia on the molly islands yellow- gull was seen south of laramie by W hep- stone lake yellowstone national park but worth on 20 june 1979 the most recent re- was unable to determine whether they were port of this species from wyoming is of a bird nesting on 4 june 1932 wright 1934 also seen by 0 K scott at soda lake natrona saw this species on the molly islands and pre- county on I1 may 1982 sumed it to be a breeder but failed to locate a few observations of glaucous gulls are nest kemsies 1930 fustfirst documented breed- expected in wyoming because this species ing caspian terns on the molly islands when prefers coastal areas and large inland bodies of he found eggs and downy young on 29 june water and its breeding range is north of the 1929 between 1932 and 1966 the number of contiguous united states AOU 1983 caspian tern nests vanedvaried fromhiomblom a low of 4 nests on 5 july 1959 to a high of 18 on 24 june 1966 black legged kittiwake diem and condon 1967 in 1955 walkleywarkley the black legged kittiwake rissa tndactilatridactyltridactylaa found evidence of caspian terns nesting at was first reported in wyoming by knight 1902 ocean lake WHMA scott 1955 also one bird was collected by M jeserumjesurum near mccreary 1939 indicated that A B mickey douglas converse county on 18 november located a pair on an island at bamforth lake 346 GREAT BASIN naturalist volume 54

albany county in the summer of 1936 one 32 of 28 latilatilonglong blocks this species may nest 4of this species was discovered at bamforth occur in the state more frequently than lake inm 1974 and two pairs appeared to nest reports indicate because of its similarity in there in 1975 E raper personal communica- appearance to the more common forster s tion the most recent evidence of caspian tern terns nesting at bamforth lake is from 1983 breeding records exist for the common tern when I1 counted four nests with eggs on 10 june in adjacent states of idaho C H trost person- all nests were later destioyeddestroyeddesti oyed by high water al communication south dakota johnsgard in recent years caspian terns have nested 1979 and montana skaar 1980 at five locations in wyoming table 1 for forstelforster s tern unknown reasons there has been a precipi- tous decline in the state s breeding population both knight 1902 and grave and walker during recent surveys the only active colony 1913 considered the forster s tern sterna min 1985 and 1986 was at pathfinder reservoirReservon forsteriforsteri a rare migrant to be found only in the breeding records exist in idaho larrisonLarnson southeastern part of the state mccreary et al 1967 C H trost personal communica- 1939 indicated that this species was a com- tion and utah behle and perry 1975 I1 am mon migrant in eastern wyoming and summer unaware of nesting records inm other states that resident in the southeastern portion of the adjoin wyoming state one nest with two eggs was found on 31 may by B common tern 1936 bv A mickey at bamforth lake albany county and a colony containing 12 bond 1885 was the first to list the common nests was found at the same location on 2 july tern sterna hirundo as occurring in wyoming 1933 mccreary 1939 another nest of this this species was considered rare by both species was discovered in albany county as knight 1902 and mccreary 1939 two speci- late as 21 july mccreary 1939 although mens were collected by mccarthy along the kemsies 1935 speculated that the forster s sweetwater river natrona county in 1859 tern occurred fairly frequently in yellowstone and another bird was taken at cheyenne national park and indicated that it may possi- laramie county by F bond prior to 1902 bly breed in the marshes bordering knight 1902 blackwelder may have seen a yellowstone lake thus far there has been only common tern in the teton region grave and one record for the park walker 1913 apparently woodbury 1937 oakleaf et al 1982 considered the forster s collected a specimen at yellowstone lake tern a common summer resident and reported probably in 1931 oakieOakleoakleafetaloakleafafetalet al 1982 report- it as a breeding species from one 36363.63 6 lati ed that the common tern was an uncommon long and occurring in 20 71 of 28 latilongslatilongs summer resident in wyoming occurring in 9 during the 1982 86 period forster s terns

TBLETOLE 1 location number of nests and habitathalhaintatneat of caspian tern ccolonies in wyomingW ming 1983 86 number of nests name location 1983 1984 1985 1986 habitat albany county bamforth lake lake bamforth island 4124ni0544nv4124 n105 44w 4 0 0 0 carbon countycount pathfinder besenreservoirresen airoir reserviorReservior bird island 4323n10656nv4323 N 106log 56w nga 15 20 23 29 natronaatronacountcounty soda lake reservoir west island 42425454n106106log 19w19nn7 13 0 0 0 rattlesnake island 425442344254n10618nn7N 106log 18 0 1 0 0 yellowstone national park yellowstone lake lake molly islands 4419n11016w4419n110 16w 12 3 0 0 NC not censcensusedcensusesused datdata ftonaftonfroni K L diem personal communication 199419941 LARIDAE OF WYOMING 347

TABLE 2 location numnumberber of nestsniests anda nd habitath of forsterfc rsterarster s tern 0olonieseloniescolonies in wyomingw1waromagrommg 1982 1984 86 number ofnestsof nests name locationlotloilatiratidati on 1982 1984 1985 1986 habitat albamalbany county caldwell lake 41 09n10548w09 N 105 48w NCMO NC NC 19 lake carroll lake 4125ni0544nv41 25 N 105 44w NC NC 15 20 3 lake hutton lake NWR 4111n10544w41 11 N 105 44w 8 15 0 3 2 3 marsh kay ranch 4115n10542w41 15 N 105 42 W 2 3 0 0 0 lake pilger lake 4123n10550w41 23 N 105 50w NC NC NC 9 lake fremont county ocean lake 4307n10835w43 07 N ios108 35w NC 10 36 12 reservoir lincoln county bear river 4201n11058nv42 01 N iioilo110 58w 0 2 3 0 0 marsh ancC noti ot ansicenscensusedcensusesusedd nested at seven locations in wyoming table 2 mccreary 1939 no recent records exist in however not all of these sites were active the state for this species each vearyearyean based on the 1986 colony censuses the least tern breeds locally and irregularly approximately 45 46 nests were present this in south dakota and nebraska johnsgard compares to 10 18 active nests in two colonies 1979 I1 am unaware of nesting records from during 1982 the increase in the breeding other states that are adjacent to wyoming population is primarily a result of locating four black tern new nesting areas during recent surveys I1 am uncertain why forster s terns failed to nest on bond 1885 was the first to list the black the kay ranch and bear river in 1985 and tern chlidoniaschhdomas niger as occurring in 1986 significant declines in nesting forster s wyoming this species was considered a rare terns were also noted at ocean lake and migrant in the state by knight 1902 grave carroll lake in 1986 fewer terns probably and walker 1913 indicated that there were nested at carroll lake because of very low records of black terns from cody park county water levels that reduced nesting habitat at sheridan sheridan county lake como ocean lake the decline may have been albany county cheyenne laramie county caused by the addition of more cobble to the and douglas converse county mccreary man made nesting islands which made them 1939 noted this species as being a common more dome shaped and less suitable as nest- migrant in eastern wyoming and a summer ing substrate flooding of nests may also be a resident in the southeastern portion of the serious problem at ocean lake state henningerHenmnger 1915 found a nest contain- breeding records exist for this species from ing one egg near bamforth lake on 12 june adjoining states of colorado bailey and 1914 this was the first documentation of neidrichneidrachNeidrach 1965 idaho larrison et al 1967 nesting by black terns in wyoming in yellow- C H trost personal communication montana stone national park kemsies 1930 reported skaar 1980 nebraska and south dakota that the black tern was a frequent migrant johnsgard 1979 and utah behle and perry and probable summer resident this species is 1975 considered a common summer resident by tern oakleaf et al 1982 and has been reported least from 20 71 of 28 latilonglatilong blocks with strong mccreary 1939 indicated that the least evidence of breeding from 2 latilongslatilongs on 3 tern sterna albifronsalbifrons was a summer resident june 1982 I1 discovered 2 4 nesting pairs of along the north platte river the first sighting black terns with eggs on the kay ranch of this species was at torrington goshen about 10 km southwest of laramie albany county on 11 june 1929 mccreary 1934 county from 1984 through 1986 this species mccreary and mickey 1935 J W scott noted nested at three locations in albany county 8 or 10 individuals near fort laramie goshen and in the marshes associated with the bear county on 25 june 1932 mccreary 1934 river south of cokevilleCokeville lincoln county one year later on 27 may the least tern was table 3 since 1984 new black tern colonies again reported fromrom torrington goshen county have been discovered at carrollcan oil and caldwell 348 GREAT BASIN naturalist volume 54

TABLE 3 location number of nests and habitat of black tern colonies in wyoming 1984 86 N umbertimberlimber ofofnestbnests name location 1984 1985 1986 habitat abanymbanyalban count caldwell lake 414109n10509n105 48nn48 nc1acanca NC 2 3 lake carroll lake 41412525n10544nn105 44 NC 10 15 2 3 lake hutton lakenrlakenaLaklake eNRNWR 411in10544nv4111n105 44w 7 10 8 10 1 2 marsh kay ranch 41i5ni0542nv4115n105 42w 0 0 0 lake lincoln count bear ririverer 420inii058aj74201n110 58w 100 150 NC NC marsh

anc ot censuscensuscdt cds

lakes for unknown reasons this species failed increased or whether new colonies are the to nest at the kay ranch during the 1984 86 result of intensive surveys I1 believe that evievlevi- period and has not been documented as dence exists for recent population increases of breeding there since 1982 population trends the ring billed gull california gull and her- of black terns are unknown in wyoming ring gull in the state reasons for proliferation because most colonies have been monitored of california gull and possibly ring billed an insufficient number of years also numbers gull and heangherring gull populations in wyoming of black terns nesting in the marshes adjoin- include construction of large reservoirs with ing the bear river have not been censcensusedcensusesused isolated islands for nesting as well as creation since 1984 of new food sources such as garbage dumps this species has been found nesting in the other human refuse and agricultural land following states that adjoin wyoming findholt 1986a breeding populations of colorado bailey and neidrichneidrachNeidrach 1965 idaho these species apparently are expanding lamsonlarrison et al 19671667 C H trost personal com- throughout the western united states conover munimunicationcation montana skaar 1980 nebraska 1983 chase 1987 in contrast I1 believe most and south dakota johnsgard 1979 and utah new colonies of caspian terns forster s terns behle and perfyperry 1975 and black terns are a result of current sur- veys and not the result of recent breeding conclusionsCOCLLSIOS range expansions into wyoming however the addition of at least a few new colonies of in recent years observations of nonbreeding caspian terns and forester s terns in wyoming species of gulls terns and jaegers have inin- since historical times appears to be the result creased in wyoming I1 believe these increases of human caused environmental changes are primarily a result of more surveys being especially the construction of reservoirs which conducted by professional biologists and more have created nesting and foraging habitat time spent in the field by greater numbers of with the exception of the california gull amateur birdbirdwatcherswatchers of the 11 nonbreeding which is a relatively abundant and widespread laridae documented in the state the pomarine nesting species in wyoming breeding popula- jaeger and Heermanheermannrss gull were reported for tions of the other five species of laridae that the first time since 1980 also the majority of nest in the state are small it appears that lim- sightings of other nonbreeding species have ited nesting and foraging habitat restricts pop- occurred during the last 10 15 years ulation sizes of most gulls and terns also observations of these species will most likely because wyoming is at the edge of the breed- continue to increase as more individuals take ing range of most species currently nesting in up birdcatchingbirdwatchingbirdwatching as a hobby in wyoming the state populations may remain small alternative explanations for increased reports it seems unlikely that nesting populations of nonbreeding species of laridae are range of the 11 nonbreeding laridae will be docu- expansions or changes in migration routes I1 am mented inm wyoming except for the franklin s unaware of evidence from wyoming or else- gull and common tern because the breeding where for either explanation range of most species occurs along coastal it is unknown whether breeding populations areas or north of the contiguous united states of some laridae in wyoming have recently the only other species in addition to the 199419941 LARIDAE OF WYOMING 349

franklin s gull and common tern that nests congress ist session house executive document in states that adjoin wyoming is the least no 1 part 5 vol 2 tern since the least tern nests locally and BUCKLEYBLCKI PAP A decdegANDDFGE G BLCKUBLCKUABUCKLEYA 1979 what constitutes a waterbirdwaterwaterbirclbird colom r reflections from the irregularly in south and Ne northeast- in dakota nebraskabrasla and ern USU S proceedings of the 1979 colonial Waterwaterbirdbird does not appear to be expanding its range it group3croupgroupgroupe3 1 15 seems unlikeunlikelvunlikelylv that it will be found nesting in CHASE C A III111 1987 hvbridizationhybridization of herrinherring and wyoming uansgardwansgardjohnsgard 1979 california gulls pages 16 17 in proceedings oftheodtheof the 14th14tb annual pacific group pacific it is difficult to assess longtermlong term population meeting seabird clovegiovecrovegrove california trends of most laridae that currently breed in CONOVER NI R 1983 recent changes in ring billed and wyoming because of the limited number of california gull populations in the western united years that population data are available states wilson bulletin 95 362 383 however results presented in this paper will dieniDIEM K L AND4nddD D gondoCONDOgondonCONDON 1967 banding studies of serve as baseline data that be used eval- water birds on the mollymoll islands yellowstoneyellow stone lake can to yellowstoneyellow future population wyoming stone librarlibraelibrary and museum uate changes in the state Asassociationsociation yellowstone national park because most breeding colonies are cur- FINDHOLTFIVDHOLT S L 1984 status and distribution of herons rently protected in wyoming prospects for egrets ibises and related species in wyoming maintaining viable nesting populations appear colonial waterbirdsWater birds 7 55 62 good it is my hope that natural resource man- 1986a status and distribution of california gull nesting colonies in wyoming great basin naturalist agement agencies will continue efforts to 4612846 128 133 monitor longtermlong term population changes and 19861 the ring billed gull a rediscovered nest will implement appropriate management ing species in wyoming western birds 1718917 189 190 strategies to ensure that currently unprotected FNDHOLTFINDHOLT S L annANDD K L BERNER 1988 current status breeding populations of are and distribution of the ciconnfoimsciconiiformsciconiiformes nesting in laridae main- wyoming great tained in the state basin naturalist 48 290 297 crecravegreGRAVE B H ANDD E P WLKERWALKER 1913 the birds of wyoming universityUmer sit of wyoming laramie acknonvledgmentsACKNO LEDGMENTS heiterhelterHEICERHENNINGER W FE 1915 june birds of laramie wyoming wilson bulletin 27 221 242 JOHNSGARDjohnsoJOHNSC P A 1979 of great plains I1 thank K L diem B H pugesek B and W birds the breeding and M raynes species their distribution universitylniersitsit of 0 K scott H downing and oth- nebraska press lincoln Lnier ers for allowing me to use their observations of kemsiesKEMSIFSKENISI Fs E 1930 birds of the yellowstone national park rarely sighted laridae in this paper much with some lecentrecent additions wilsonlisonilson bulletin 42 information comes from sheridan area birds 198 210 an unpublished collection of sightings edited 1935 changes in the list of birds of yellowstone edited naionalnacional park wilson bulletin 47 68 70 by H downing K L berner provided invalu- KINGERY fl E 1975 mountain west american birds 29 able field assistance in 1986 I1 thank N E 93 98 findholt S H anderson J burger L spear 1977 mountain west american birds 31 1168 W H and 1985 mountain west american birds 39 82 85 behle an anonymous reviewer for KNIGHT kmchrwcW C 1902 the budsbirds of wyoming universityv of helpful comments on the manuscript this wyoming agricultural experiment station bulletin study was funded by the wyoming game and 55 fish department nongame program kushlanKLSHL J A 1986 Comcommentarymentan colonies sites and suressurvevs the terminologyterminolog of colonial waterbwaterbiidwaterbirdwaterbirdiidildlid stud iesles colonial waterbirdswaterhirdsWaterbirds 9 119 120 CITED llteratlreliterature LARRISON E J J L TUCKER ANDID M T JOBLIE 1967 guide to idaho birds journal oftheodtheof the idaho academy amennAMERKAIERICAN ornithologists uloklokioUNION 1983 checklistcheck list of of science 5 1 220 north americanamerleanamelAmei birds ath6th edition alienallenailen press icanlean inc MCCREARY 0 1930 some recent observations onOD wyoming Lay lawrencelavvience kansas birds journal of the colorado wyoming academy of bilmaballeyBAILEYA VI ANDND R J NEIDRACH 1965 birds of coloradocoloichloi ado science 12 26 museum denver of natural history denver 1934 the platte river as a migration louteroute foiforholbol coloradoado 2 volumes chloicoloi birds nebraska bird reviewre lew 22 38 39 behlBEHIBEHLE 1 XV H AND M L PERRY 1975 utah birds check- 1939 wyomingomegommg bird life burgess publishing co list seasonal and ecological occurrence charts and minneapolis minnesota guides to bird finding utah museum of natural MCRERNICCREARY 0 ANDD A B MICKEY 1935 bird migrationmigia tion history university of utah salt lake city records from southeastern wyoming wilson BOND F 1885 A list of the birds of wyoming pages bulletin 47 129 157 1138 1140 in annual report of the governor of MCCRIMMON D A JR 1982 populations of the great wyoming to the secretary of the interior inliiiiilit annual blue heron ardea herodiashero clias min new york state from report of the secretary of the interior 49th 1964 to 1981 colonial waterbirdsWaternvaterbirdsbirds 5 87 95 350 GREAT BASIN naturalist volume 54

OAKLEAFCIAKLEAF B 11 DOWNING B RAYNES M RAYNES AND 0 SCOTT 0 K 1955 great basin central rocky mountain K SCOTT 1982 wyoming avian atlas wyoming region audubon field notes 9 392393392 393 game and fish department cheyenne SKAAR P D 1980 montana bird distribution published oberholser H C 1919 larus banuscanus brachyrhynchus inm bvby P D skaar bozeman montana wyoming auk 36 276 277 SKINNER M P 1917 the birds of molly islands yellow- RAPER E L 1975 influence of the nesting habitat on the stone national park condor 19 177 182 breeding success of california gulls larus californi 1925 the birds of the yellowstone park roosevelt cus bamforth lake albany county wyoming un- wild life bulletin 3 1 192 published master s thesis university of wyoming woodberywoodburyWOODBLRY L A 1937 parasites of fish eating birds at laramie yellowstone lake wyoming condor 39 125 126 RDERRYDER R A 1978 gulls in colorado their distribution WRIGHT G M 1934 the primitive persists in bird life of status and movements proceedings of the 1977 yellowstone park condor 36 145 153 conference of the colonial waterbirdWaterbird group 1 3 9 SCHALLER G B 1964 breeding behavior of the white received 2 september 1993 pelican at yellowstone lake wyoming condor 66 accepted 7 february 1994 3 23 great basin naturalist 544 C 1994 appp 351358351 358

SEED production IN GENTIANA NEWBERRYI gentianaceae

myra E barnesibames1bargnesi and richard W rusti

ABSTRACT experimental manipulations and observations in one population ofcentianaof gentiana newberrnewberryiyi graygrav flowers over 2 years showed significant variation in seed production relative to pollinator and soil water availability when pollinatorspollinators were rare there was a significant relationship between number of bees present and number of mature seeds produced and supplemental hand cross pollination xenogamy did improve seed set in gentiana newberryinewberryi gray when pollina tors were abundant supplemental hand cross pollination did not increase seed set self fertilized seeds autogamy ger- minated at the same rate as cross pollinated seeds seed production in unvisited flowers is probably limited anatomical- ly and is not influenced by the type of fertilization there was a significant relationship between soil moisture and flower size in G newberrnewberryiyi with larger flowers found in wetter areas

key words gentiana seed production pollination bumblebees soil water potential

seed set can be limited by insufficient po- nator visitation rate zimmerman 1980 llinator visits levin and anderson 1970 mcdade and davidardavidaz 1984 zimmerman and thomson 1980 1981 bierzychudek 1981 pyke 1988 supplemental hand pollination gross and wemerwerner 1983 pleasants 1983 waser can be used to determine whether pollinatorspollinators 1983a motten 1986 galen and newport 1988 or climatic factors are limiting seed set calvo and horvitz 1990 harder 1990 ashman motten 1983 and stanton 1991 or by other resources such here we present both observational and as water or nutrients in populations with suf- experimental data on seed production in ficient pollinatorspollinators stephenson 1981 evenson Cengentianagentianafianaflana newberrnewberryiyi gray these perennial 1983 mcdade and davidardavidaz 1984 primack plants are restricted to high elevation wet and kang 1989 resource limitation may meadows in eastern california western result aborting the whole fruit or only some nevada and southern oregon munz 1973 in gen seeds in a fruit lee 1988 multiple reproduc- gentianaCentiana newberrnewberryiyi has protandrous funnel tive strategies in perennials including cross shaped flowers that are usually white with pollination self compatibility levin 1971 jain greenish spots munz 1973 each ramet has 1976 barrett 1988 karoly 1992 and vegeta- one or two flowering shoots with one or two tive reproduction evenson 1983 waller terminal flowers personal observation they 1988 are advantageous in populations where can reproduce sexually and vegetatively spira pollak pollinatorspollinators and other resources are un pre 1983 spira and 1986 dictabledictable motten 1982 sutherland 1986 initial observational data included pollina- soil ehrlen 1992 while self fertilized and vegeta- tion modemodelmodeo pollinator activity and mois- tively produced plants increase the risk of ture effects on seed production based on measured inbreeding depression those that are success- observational information we then soil potential and pollinator visitation ful may have coadapted genes that are advan- water the study and throughout the sea- tageous for current environmental stresses across area to determine relationships between polli- lloyd 1979 waser and price 1982 barrett son availability or soil moisture and seed 1988 nator production optimal flowering time is when a plant can attract the most visitors and still be able to set METHODS MATERIALS seed during the growing season pleasants AND waser primack 1987 1983 1983b when study site bees are abundant pollination does not limit seed set when bees are infrequent there is in august 1991 we selected a 2700 mm2ma2 often a correlation between seed set and polli study site at little valley 27327.3 km southeast of

I1blobiolobialogBiobiologlog109 department university of nevada reno nevada 89557

351 352 GREAT BASIN naturalist volume 54

reno nevada 11952w 3915n the site length by corolla tube width ultraviolet re- located at an elevation of 2000 in along the flectiveness was determined by photographing eastern edge of the sierra nevada escarpment fumeinumeinumerousous buds and open flowers on live plants is part of the 1200 ha whittellwbittell forest and in the field with a wratten isa UV filter wildlife area owned bvby the university of on 21 august 1991 10 mature bud flowers nevada reno rust 1971987 the area is cov- ready to open were marked and covered with ered by snow each winter but has an average waxed paper bags J1 in the wet area and 9 in of 120 days with minimum temperatures the dry another 13 buds were marked and above OC houghton et al 1975 Cengentianagentiana left uncovered 7 in the wet and 6 in the dry newberrnewberryiyi populations are found in meadows area at 0900 the following morning nectar that collect and retain snowbeltsnowmeltsnowmelt water longer volume in each flower was measured using a than surrounding meadow areas A small loi1011 jaljaitajult1 capillary tube on 13 september 1991 10 creek running through the southern portion of mature bud flowers in the dry area and 13 in the site keeps the areas near the creek wet the wet area wereweiwel e bagged the following after- throughout the growing season central and noon 1300 nectar was measured from each northern portions of the site dry out toward flower each week during 1992 nectar was the end of the growing season the study was measured at approximately 0900 from one ran- conducted during the fifth and sixth years of a domly selected uncovered dehiscingdebiscingdehlsdebiscingeing flower drought when snowfall was only 50 of nor- in each of the five soil moismolsmoisturetuietule areas n 60 mal james 1992 most flower species began pollinators flowering 4 weeks earlier than usual in 1992 Pollinators personal observation aftelafterabterabher a warm spring and bumblebees were common throughout early snowbeltsnowmeltsnowmelt james 1992 august and september 1991 but the number population characteristics of bees plesentpresent was not regularly recorded individual foraging bees were followed and in mid september 1991 and 1992 num- the flowers visited distance between flowers bers of plants cametsrametsramets and flowers on the and times were recorded in 1992 a 100loom m study site were estimated using 11 transects tiantransectseetsect was established across the drier part 38 54 in long placed 5 in apart and extend- of the site and a 50 m shorter due to limits of ing across the population to include all G the wet area tiantransectsectseet placed across the wet newberrnewberryiyi plants at each meter along each area these transects were walked hourly at transect all plants and flowers were counted least 2 davsdays each week and any bees observed and percentage of CG newberryinewberryi coverage was within 5 in of the transect were recorded A estimated within 5 m2ma 050 circular quadratsquadquadransrats sample of all insect visitors to CG newberrnewberryiyi floral characteristics was collected for identification by R rust university of nevada reno and R brooks in 1992 after observing the differential university of kansas lawrence drying of the study site in 1991 we divided the site into five areas area 1 was adjacent to seed production the southern creek always wet during 1991 in 1991 flower buds of CG newberrnewbernyiji were aleasareas 2 to 5 were equaliaequalivequally spaced away from randomly selected with not more than one the creekci eek with area 2 the closest to area I1 and flower per plant and mariedmarked with numbered area 5 the farthest removed soil water poten- paper tags n 113 three times each week tial was measured in each area weekly with a the phenology bud flower opening dehiscing quickdrawQuickdraw series 2900 soil moisture probe anthers receptive stigma and seed capsule moisture was measured at approximately 30 formation of each marked flower and soil cm three measurements were taken per area moismolsmoisturetuietule conditions adjacent to the plant per week petal length and maximum corolla visuallyvisuallv were recorded during 1992 two tube width were measured for 10 random G newly opened flowers on different plants with newberryinewberryi flowers each from a different plant dehisced anthers in each area were randomly and with dehlsdebisdehiscingdebiscingcingeing anthers in each area in selected and marked each week from early july 1992 n 50 an index was developed to through early october n 84 mature seed compare flower size by multiplying petal capsules for all marked flowers were collected 199419941 GENTIANACENTIANA NEWBERRYI SEED production 353 and placed in individual waxed paper bags RESULTS mature seeds and undeveloped ovules were counted using a dissecting microscope seeds population characteristics for each flower with any mature seeds were there was no significant difference between placed in individual waxed paper bags and 1991 and 1992 in number of plants F 0430.430 43 stored outdoors inin reno nevada df 1984 P 51 or number of flowers F during august 1991 53 CG newberrnewberryiyi buds 1161.16ilg1 16 df 19841 984 FP 28 peipelper quadrat were randomly selected and marked and the table 1 percentage cover was significantsignificantly plants covered with white nylon organdy 100 different between earsvears F 3973.973 97 df 1984 mesh bags over wire frames when the flow- FP 04 table 1 distribution of plants is ers opened 9 were handband pollinated using a clumped throughout the study site as indicat- paint brush bearing pollen from a flower on bvby 1 I 0.510 51 ed the Mormonsitaisita index P 051031 the same plant geitonogamy 12 with pollen from a different plant xenogamy and 34 floral characteristics were left to self pollinate autogamy during A significant difference in CG newberrnewberryiyi the week of 18 august 1992 10 plants 2 in flower size petal length X maximum corolla each soil moisture area with bud flowers were tube width was found among the five soil selected at random marked covered with moisture areas of the study site in 1992 F nylon organdy bags and allowed to self polli 370437 04 df 445 FP 00010061.0061 table 2 there nate ten other newly opened flowers in the was a significant regression y 6934693.4693 4 different areas were marked and cross polli 125x12 5xax between soil water potential and flower natedbated by hand after the stigma became recep- size F 11779117.79117 79 df 148 FP 0001 r2ra tive another 10 flowers in the different areas 71 with larger flowers found in wetter areas were marked and left alone for natural pollina- ultraviolet images of G newberrneitberryinewberryiyi flowers tion mature seed capsules for treatments show a dark central UV absorbing bullbullsevebullseyeseve used in 1991 and 1992 were collected and pattern in the corolla tube and a dark longitu- seeds counted and stored using the same dinal stripe on the outside of each petal from method the base to the apex outer petal stripes are each yeaiyear five randomly chosen seed cap- also visible on flower buds sules from each treatment and the open flow- in 1991 in a sample of newly opened flow- ers were germinated seeds for each flower ers at 0900 h there was no difference in the were placed in a petri dish on brown paper amount of nectar available between flowers over kilpackkimpack moistened with a 400 ppm covered with a bag overnight 0 1 0 1 mean giberellicgiberellic acid solution seeds were kept at and standard deviation range 0 030.30 3 alzlA and 15c15 C for 7 d and then alternated between those left open oi01010.10 1 010.10oi 1 range 0 030.30 3 A F 15c foiroiforbor 12 h and 25c for 12 h for 7 d in the 0030.030 03 df 1221.22 FP 97 or in the amount of daikdalkdark nectar in flowers between wet area 04040.40 4 040.40 4 range 0 141.41 4 alA and dry area 02020.20 2 020.20og 2 i angerange analyses statistical 0 070.70 7 1aaa1 F 2029022.022 02 4df 121 FP 17 in analysis of variance GLM in SAS 1990 1992 there was no difference in the amount of was used for all comparative analyses between nectar available in open flowers between years and between areas or weeks within each year bonferrombonferroniBonBonferferronitomrom t tests were used for multiple TBLETABLE 1 number of G neunewnewberrnewberryiherrliherryiyi plants flowers and comparisons when analysis of variance indi- percentagepeiper centage otof cover per 050511120 5 1112 quadrat at little vallenmailenmaileyailealiealle cated a significant difference arcsine transfor- neadanevada values are means standard deviationdeiation ii 493 total numbers ofplantsof plants and flowers in 2700 m2ma studystud site mations were used for analysis of percentage in aiealeare in parentheses data zar 1974 linear regression SAS 1990 plants percentpeicentcovei was used to determinedetel mine if there was a relation- no flowers no cover ship between soil moisture and flower size 1991 soil moisture and seed production or seed 12 373 7 3115 232 3 31 773 17 555 5 production and bee visits each week plant distribution was determined using the stan- 1992 111 1 414 1 2710 020 2 111 1 476 111 1 545 4 dardizeddardized Mormonsitamorisitamorishitaisita index 1 Ip krebs 1989 354 GREAT BASIN naturalist volume 54

TBLETABLE 2 relationship between soil water potential and flower size in G newberrnewberryiyi at little valleyvailey nevada in the second week ofjulyof july 1992 soil water potential petal length corolla width and a flower size index petal length X corol- la width are indicated for five areas of decreasing soil water potential valuesmahles are means standard deviation nio11 10 area soil water petalP italitai coicolcorollarolla flowerfic V er size potentialpotestialential lengthlerieniehirthigth w1widthwaathdth inindexdex mpaI1 pa atmm1tim mmITim

1 00 00 465 24 164 10 7641764 1 790 aaA 1

2 35 05 440 35 140 18 6176 1068 B

3 60 16 417 36 133 19 557753773377557357 7 1055 B

4 120 49 383 31 110 19 4251425 1 1023 C

5 350 00 331 29 86 18 2875 744 D

bonferromBonferrom t tests comparing differences in flflerflowerer size index among areas means withith the s tinelinesame letter anare not significantsignificantly different P 05 weeks week 1 010.10oi 1 010.1oi0 1 range 0 030.30 3 alzl bombus edwardedwardsiiedwardsiiisii were seen bombus cosnevosne og week 2 060 6 090.90 9 range 0 212.12 1 JLd week 3 senskii were observed visiting other flower 040.40 4 030.30 3 range 0 070.70 7 A week 4 0 week species and occasionally robbed nectar from 5 070.70 7 040.40 4 range 040.40 4 131.31 3 alA week 6 010.10oi 1 CG newberrnewberryiyi from outside the flower 020.20 2 range 0 040.40 4 alA and week 7 020.20 2 020.20 2 anthophora bomboidesbomboides kirby A urbana range 0 040.40 4 alfilaalaai F 1991.991 99 df 628 P 10 cresson and A teniteninalistennitenninalisnalisnallsnails cresson were occa- no difference in amount of nectar was found siosionallynally observed visiting G newberryinewberryi apis between areas area 1 driest o1001oa 1 020.2og0 2 mellifera L were common visitors to adjacent og range 0 0500.55 alJL area 2 030.30 3 020.20 2 range 0 060.60 6 lupinus sellulussellulus and one was seen visiting a allalibtl1 area 3 020.20og 2 030.30 3 range 0 0 7 A area 4 G newberrnewberryiyi flower anthophora species and 04 04 range 0 131.31 3 tlA and area 5 A mellifera were not seen in 1991 wettest 060.60og 6 080.80 8 range 0 222.22 2 alxl F in 1992 there were always many flowers loi1011.011 01 df 430 FP 41 most nectar was open with pollen available less than one bee usually found in one or two of the five nectar per 750 m2mam was observed when walking tran- tubes sects there was a significant positive correla- tion between number of bees observed each Pollinapollinatorstors week along the combined transects and num- in 1991 four species of bumblebees were ber of seeds produced per flower marked that observed visiting CG newberrnewberryiyi flowers table week fig 1 3 most visits were for nectar nectar forforagersagers seed production would pick up pollen on their ventral surface from the centrally located anthers bombus there was no significant difference in the appositusappo situs cresson and bombus eduardedwardhdwardedwardsnedwardsiiedwardsensiisilsllsn mean number of mature G newberrnewberryiyi seeds cresson were frequent visitors to CG newberrnewberryiyi produced per marked flower between 1991 from early august until the end of september 1162116.2 1436143.6 n 58 and 1992 1354 bombus vosnesenskiivosnesenskn oszkowskiradoszkowskiRad was fre- 1144114.4 n 76 F 0750.75 df 1132 P 3838.38 quently observed visiting lupinus sellulussellulus when we eliminated flowers with aborted kell adjacent to CG newberrnewberryiyi and occasional- seed capsules from the analysis 50 aborted ly visited a few G newberrnewberryiyi flowers bombus 1991 23 1992 the number of mature seeds ferfervidusvidus fabricius was seen visiting G new per capsule was higher in 1991 23242324232.4 1183118.3 berrtiberryi flowers during only one week in august n 29 than 1992 17461746174.6 1000100.0 n 59 F usually between one and six bumblebees 5765.76 df 186 FP oi0101.01 could be found visiting CG newberrnewberryiyi flowers in 1991 significantly more seeds were pro- on the study site anytime during august and duced by CG newberrnewberryiyi flowers in the areas september when the weather was warm and with a wet soil surface 21032103210.3 1751175.1 n 20 calm than in the dry areas 66766766.7 93493.4 n 38 F during 1992 bumblebee visits to G new 167016.70 df 156 FP 0001.0001 more seed berrtiberryi were rarely observed A few visits by capsules aborted in the dry than the wet area 199419941 GENTIAIVA NEWBERRYIXEWBERKYI SEED production 355

TTABLETVBLEVBLE 3 total number of visits mean length ofisofisitsof visitsits 400 and mean distance traveled between flowers for varying numbers of individualsmdi idealsiduals of four foraging bombus species at rr079079 p oot001000.01oo little valleyvailey neadanevada inm august 1991 values for timetune and 30000 000.0 distance are mean standalstandardd dedeviationlation species indiindividualsidealsiduals visits time distance 200 a G secsee cm 19

B 1 13 5 appoappositusappo&itussitus 6 103 07 135133 83 a 100 a G B eduardedwardeduardsiiedwardsiiedwardsiiisiisilsllsti 1 20 128 45 919 1 247 La 0 B ferferudiiffervidusoldusvidus 5 45 147 92 644 982 00 020.2 040.4 060.605 080.8ob 101.0lo B vosnesenskiibowesenfikii 3 26 82 56 83523583323 5 232 lleanmean beestransectweek

fig 1 correlation between mean number of bees observed along combinecombined transects anand mean number of 58 vs 35 A higher percentage of ovules seeds produced by G newberrnewberryiyi flowers that were open per capsule matured to seed in the wet area the week bees were observed iin 7 84784784.7 108ioslos10.8 n 13 than in the dry area 60260260.2 24024.0 n 16 F 5885.88 df 127 P between treatments in 1991 F 2009002.00goo df oi0101.01 315 P 1515.15 or 1992 F 2932.93 df 29 P in 1992 there was no significant difference iolo10.10 seed germination ranged from 91591.5 to in the number of mature seeds produced per 98 in 1991 and 84884.8 to 98898.8 in 1992 for all flower between the areas with varying soil treatments water area I1 driest 1337133.7 1125112.5 area 2 1531153.1 98498.4 area 3 57057.0 79379.3 area 4 discussion 1619161.9 1358135.8 and area 5 wettest 14211491142.1 12281928122.8 F 2262.26 df 479 P 07 or in the seed production per marked gentiana percentage of ovules that matured to seed per newberryinewherryinewberr yi flower was not significantly differ- capsule F 0560.56 df 460 P 69 there ent between 1991 and 1992 in little valley was however a significant positive relation- even though pollinator numbers and soil mois- ship between soil water at the time a flower ture were different between the 2 years the opened and number of mature seeds pro- large variation observed in all measurements duced y 1589158.9 23x and F 5965.96 df of CG newberrnewberryiyi population characters and 169 P 0202.02 flowerseed characters in the little valley there was a significant difference in the population both among and between years average number of seeds produced F 8448 44 suggests much individual variability within df 383 P 0001.0001 and the percentage of the CG newberrnewberryiyi habitat area individuals ovules that matured to seeds F 9069.06gog df respond to a variety of localized microenviron 342 P 0001.0001 between the open xenoga- mental parameters with the resulting variation mous and geitonogamous versus autogamous in sexual reproductive output pollination treatments in 1991 table 4 in bumblebees visiting CG newberrnewberryinetcberryiyi were 1992 there was also a significant difference abundant during 1991 it is unlikely that polli between pollination treatments of open and natorsgators limited seed set in 1991 since flowers xenogamous versus autogamous in the average with hand cross pollination did not set signifi- number of seeds produced F 113711.37 df cantly more seeds there was a higher num- 220 P 0005.0005 and the percentage of mature ber of aborted fruits in the dry area in 1991 seeds per capsule F 8158.15 df 225 P during 1992 when soil was wetter G newbernewbor 002002.002 table 4 hand cross pollinated ni plants appeared more vigorous throughout xenogamy seed production was highest but the study area low pollinator availability in was not statistically different fromfrontfroni open bee 1992 did appear to limit seed set as the nnum-um pollination in botiboth years autogamous seed ber of flowers that matured seeds each week production was lowest but not statistically dif- was correlated to the number of bees observed ferent from open pollination zimmerman 1980 and mcdade and davidardavidaz the percentage of germinated G newber 1984 both found that seed set was correlated ryl seeds was not significantly different with visitation rates when pollinator numbers 356 creatGRFVTGRFAT BASIN naturalist volumeVolinne 54

tabitI1 AR r 4 pciccntaucpcrecritagc of matuiematule seeds peipelper capsule and number ofmatureof matluematthemature seeds peipelper flower in foulfour pollination treat- ments treatment I1 open 2 hand cross pollpollinatedinated 3 self pollinated 4 hand pollinatedpollinatcd from floxcrfloweiflowed on same plant aborted seed capsules etewereerecic not included in the percentage ofofinatuicofinamaturetuicturcture seeds ptpeiperp t capsule analysisan alsis values aienieare mean standardsianvianstan daiddald deviation n is in parentheses matinmahnematinemannee seeds mature seeds treatments per capsule c peiper flower ilitiiiiolio110no 1991 1 72579572 5 25223225 2 16 2vavV 124312431155815581358 32 A B 2 699 27827 S 12 A 1896lb96916916 12 A 33 225 236 11 B 20420 4 49549349 5 34 B 4 49349 3 340 7 A B 1056105 6 1300 9 A B

1992 1 5835855s5275275273 6 A B 93893s947947 8 A B 2 908 80S 10 A 1565156 5 35333335 3 10 A 3 459248459439 248 7 B 460 41542542342 5 10 B

bofcroniBofbdiifficronlcronitoliloin t tctacttstsoiiipiiintuparlpari g teahtcahicahitiikiihmitlnikt ithiathi cabcaheabKIHI arii Nmtimuiimtlix ith linesainesaihe letterietter irelreare not sftnificantksmnilksmniak mth dillerditlerdiffditlereilttnteilteritant ip 05

were low when aborted fruits were eliminat- occasional visits to several CG newberrnewberryiyi in ed from the analysis in 1991 50 more 1992 the few bees observed visited a variety of mature seeds were produced per flower in flowers and none showed a preference for G 1991 than in 1992 this also suggests that the newberryincwbcrryinewberryi we observed a few nectar robbing larger number of bees present in 1991 visits from outside the flower by malemaie B vos increased the numbers of seeds produced in ne senski in 1992 plants with sufficient moisture to produce ultraviolet markings in the center of the mature fruits fruit initiation inaymay belielyeiye pollina corolla and on the outside ofpetalsof petals attract and tor limited but mature fruit and seed produc- guide bees to nectar sources silberglied 1979 tion are primarily limited by resources such as kevan 1983 waddington 1983 waser 1983b water galen and newport 1988 horvitz and bumblebees verewere often observed flying quick- schemske 1988 ashman and stanton 1991 ly and directly in a straight line fromfroinhrombrom G new it is not known whether the rarity of bees berryl flower to flower when entering a observed in 1992 could be related to drought flower a nectar foragiboragiforagingng bee positions its ven- conditions or snow and cold weather in late tral surface over thetiietile anthers in the staminate june many species ofbeesof bees were seen foraging phase and over the stigma in the pistillate on other flower species in little valley during phase bees were rarely observed collecting klaymay 1992 personal observation numerous CG newberryinewherryinewberryi pollen malemaie Boinbomhiisboinbitsbilsbitsblis vosnesenskiivosncsciiskii were observed for- there was no difference in the rate of seed aging on several flower species on the study germination between self and cross pellinatpollinat site during july but few females of any ed flowers seed production is not affected by Botbomhiisbottibitstibits species were seen during the CG leivnew type of pollination but it may be limited ana- berryl flowering period in contrast an open totomically in unvisited self pollinated flowers flower in 1991 could rarely be found with As the flower closes the stigma bends over pollen remaining on the anthersandiers throughout and touches only one or two of the anthers 1992 most open G newberrnewberryiizewbcrryiyi flowers had caged plants that were hand pollinated with nectar and abundant pollen available as poten- pollen fifromoni a flower on the same plant pro- tial pollinator rewards duced as many seeds as open plants or plants there were large populations of lupinus with supplemental hand cross pollination seiliselliseuuhissellithisthis adjacent to patches of G neubelnewberrnewberryi1 yi and plants in wetter soil hadbad larger showier aster species and rideridiaRperideridiafiderideridiafidia bolanderibolanderi craygravgray flowers than those in dry soil Pollinapollinatorstors are nels & macbrmacar were common from lulyjuly usually attracted to larger flowers galen and through september in 1991 bombus edward newport 1988 ashman and stanton 1991 in sii B fervidnervidferuidusferuferviditsjerufervidityidusits and B appositlisapposiffis were seen vis- 1991 the surface soil was very dry in a large iting only G newberrnewberryitiewberryiyi bombus vosnesenskiivostiesetiskii part 1800 m2n12ma of the study site fewer seeds vaswas usually observed visiting L seffsemfsemmidussellulusselseiseffiduslulusiulus withvith matured per flower and more seed capsules 199419941 CLMIAAGENTIANA lbhrmiNEWBERRYI SEED productionPRODLC no 357 aborted in the dry area suggesting that inade- literature gitenCITED quate soil water can limit the number of seeds produced per flower surface soil throughout ASHMASHMAN TTLL ANDMD M srtsitartSTANTON 1991 seasonal variation in pollination dynamics of sexuallyscuallsquall dimorphic the study site appeared wetter in 1992 the sidalcea oregaoregaiiaoicgunuilaiiatia sspasp spicatespicata malvaceae ecology positive relationship between soil water at 7299372 993 1003 flower opening and number of mature seeds brrhttBARRErr S C H 1988 the e olntionevolution maintainmaintainancemaintainanccanceanee and produced in 1992 indicates the importance of loss of self incompatibility ssystemsstems pages 98 124 mhi plant sufficient water resources in determining seed L lovett doustdonst and J loettlovett doust eds reproductive ecology patterns indand strategies set oxford amerstumerstuniversitvUm erst pipressess new tolktoikyoriyork facultative selfseif compatibility allows G bitbieblebierzychudekRU HLPFK P 19811981 pollinatorPollin atoiatol limitation of plant newberryinewberryi to produce seeds even when polli reproductive effort amerleanamerican naturalist 117 natorsgators are rare vegetative reproduction re- 838 840 clo R nannNANDN ncC C HORVITZ 1990 pollinatorPollinatoiatol limita- less 1988 and be cio quires energy waller may tion cost reproductionofreproductionofot and fitness in plants a used in addition to or in place of flower pro- transition matiixniatrixmataix demographic approach amelamerleanamericanAmei leanican duction or sexual reproduction of 10 plants naturalistNatmalist 136 499 516 that were dug tiplipup only one did not have an EHRLEEHRLEN J 1992 Piproximateoximate limits to seed piproductioneduction infilhil a herbaceous perennial legume lathyrtislathiillLathi attached lateral rhizome since data will con- ill vermis decoloscologecologecolo 73 1820 1831 tinue to be collected from this site we did not lensonEVENSON E 1983 rpenmcntalexperimental studies oftheodtheot the repro- dig up a sufficient number of plants to be able ductive energy allocation in plants pages 249 274 inhi to determine size of clones or average number C E jones and R J littlelittieeittle eds handbook ofot of rhirhizomeszomes per ramet experimental pollination biology reinholdreinhoid newe york101 k was no significant difference in pop- loi there calenGALENgi r C ANDn M A NEWPORTNE neiportnniport 1988 pollination qualityanqn uit ulation size number of flowers or mean num- seed set and flower traits in poleinonnnnpolemomilill tiscoviscosumviscoscos11111v1va sum ber of seeds produced per flower between complementary effects ofofaofanationvariationnation in flower scent 1991 and 1992 some plants marked in 1991 and size amciicanjonrnalofbotanamericanamerlean journal of botany 75 900 905 GROSS R S annAND0 o P A WERNER 1983 relationships that remained covered with 1 cm of water or hrfcr among floweringflofio ei ing phenolphenoiphenologyog insect visitors and seed more from the natural creek diversion did not set of individuals experimentalcxpcinncntal studies on fourhombom co survive occuringeuring species of goldenrod solidago compositae larger G newberrnewberryiyi flowers are found in ecological monographs 53 95 117 wetter areas and produce more mature seeds HARDER EL D 1990 behavioral responses byb bumblebuiriblebuibultibletibierible bees to variationvanation in pollen availability oecologia0ecologia 85 than flowers drier soil areas was a in there 41414747 significant relationship between number of horvitzhorwtzHORMTZ C C annANDD D NN schlmskhSCHEMSKE 1988 A test foiforboibol the pollinatorspollina tors present and number of seeds pro- pollinatorpollin atoiatol limitation hvpothesishypothesis foifor a neotropicaleo tropical duced only when pollinatorspollinators are rare there herb ecology 6920069 200 206 HOLCHIONHOUGHTON G C M SAKAMOTOSVKVMOTO ANDD R GIFFORDgitgil was no dididerencederencedifference in seed production between J 0 lordiohd 1975 badaseadasnevadas cathrweather and climate nevadane tdaida buleanbureau flowers with xenogamous and geitonogamous ofiofafi minesines and geology special publication 2 1 78 pollination facultative selfseif compatibility and JAIN1 l S K 1976 the c olntionevolution of inbreeding in plantspi mts vegetative reproduction allow plants to pro- annual reicheiehelereview of ecology and systematics 7 duce seeds or rametscamets when pollinatorspollina tors are 469 495 JAMESJYMES J NN 1992 nevada climate summary 01volvoiol 10 no0 o limiting CenGengentianagentiatiaflatla newberryinewberr yi appears well snminansummars centianatianatiatiafianaflana 5 unicrsitL m ei sit of nevadane ad i reno adapted to survive during unpredictable peri- krolyISYROU K 1992 pollinatorPollinatoiatol limitation in the facultativelyfacilitate cl ods of pollinator availability and7andband soil moisture autogamous annual lupinalupinuluphnis nansnany leguniinosaeeegnmmosae american journal of botany 79 49 56 kelanKENANkn PPGG 1972 floral colors in the high arctic othmthwith ref- acknowledgments erence to insect floweiflowedflo nteranter lationsrelationsle and pollination canadian journal of botany 50 2289 2316 we thank james cane auburn universitvuniversity 1983 floral colors through the insect eseeyeee what thethen what thethevchev pages 3 hi C PE and carvcarygarygarv vinvardminvardvinyard and robert nowak ununi-i are and mean 30 n jones and R J littlelittieeittle eds handbook of of reno for valuable com- experimencxpeiimcnexperiment versity nevada tal pollination biologic reinhold newie yorkork ments on the original manuscript comments KREBS C J 1989 ecological methodology halperharper and from journal reviewers and associate editor row new yorkork jeanne chambers were also constructive and eeeLEEECE T EL 1988 Pattpatteinspatternseins offruitof flintflunt and seed pioploproductionduction pages 179 202 in J lovettloveit doustdonst and EL loettlovett helpful funding assistance was provided by doustdonst eds plant leprodnctnereproductive ecology patterns the george whittell forest board and strategies oxford university pipressess new boikioikyork 358 GREAT BASIN naturalist volume 54

LEVIN D A 1971 competition foifolformol pollinator service a SPIRA T P AMannAND 0 D POLLAK 1986 comparative repro- stimulus foifor the evolution of autoautogamygainy evolution ductive biology of alpine biennial and perennial gen- 2666826 668ggs 669 tians gentiana gentianaceae in california LEVINlefileei D A ANDND W W ADERSOANDERSON 1970 competition for american journal of botany 7373 39 47 pollinatorspollina tors between simultaneously flowering species STEPHFSOstephenson A G 1981 flower and fruit abortion proxi- american naturalist 104 455167435167433167455 467 mate causes and ultimate functions annual review LLOIDLLOYD D G 1979 some reproductive factors affecting of ecology and systematics 12 253 279 the selection of selfseifneif fertilization in plants american sllsliherlandsulSLISUTHERIHERLANDAND S 1986 patterns of fruit set what controls naturalist 113 67 79 fruit flower ratios in plantsplants3 evolution 40 117 128 mcdade L A ANDfd P olardavilarolah 1984 determinants of THOMSON J D 1980 skewed flowering distributions and fluitfruit and seed set in pavoniapatoma dasydasypetaladasijpttalapetalapetaia malvaceae pollinator attraction ecology 61 572 579 oecologia 64 61 67 1981 spatial and temporal components of resource MOTTEMOTTEN A F 1982 Autoautogamygainy and competition for polli assessment b1 flower feeding insects journal of natorsgators in hepaticaHc patica aineainericanaamenmnuricana ranunculaceae animal ecology 50 49 59 american journal of botany 69 129613051296 1305 waddingtonWWDINGTON K D 1983 foraging behavior of pollina 1983 reproduction of Erythronerythronimnerythroniuinimn unibilicatuntinnbihcatum tors in L real ed pollination biology academic liliaceae pollination success and pollinatorpollinatoiatol effec- press orlando florida tivenesstiveness oecologia 59 351 359 WLLERWALLER D M 1988 plant morphology and reproduction 1986 pollination ecology of the spring wildflower pages 203 227 in J lovett doust and L lovett communitcommunistcommunitvcomm unit of a temperate deciduous forest doust eds plant reproductive ecology patterns ecological monographs 56 21 1242 and strategies oxford university press new york muzMUNZ PAP A 1973 A california flora and supplement WASERNVASER N M 1983a competition for pollination and floral university of california press berkeley character differences among sympatric plant species PLEASPLEASANTSrs J M 1983 structure of plant and pollinator a review of evidence pages 277 293 min C E jones communities pages 375 393 in C E jones and R J and R J little eds handbook of experimental pol- little eds handbook of experimental pollination linationlination biology Remreinholdreinhoidholdhoid new york biologybiologv reinhold new york 1983b the adaptive nature of floral traits ideas PRIMACKPRIMYCK R B 1987 relationships among flowers fruits and evidence pages 277 293 in L real edcd and seeds annual review of ecology and pollination biology academic press orlando systematics 18 409 430 florida PRIMACKPRIMACY R B WDAND H kantKANC 1989 measuring fitness WASER N M AND M V PRICE 1982 optimal and actual and natural selection in wild plant populations outcrossingoutcrossmg in plants and the nature of plantpiant polhpolli annual relewreview of ecology and systematicsS stematicsste matics 20 natornatormteractionnatorinteractioninteraction pages 341 359 in C E jones and 367 396 R J little eds handbook of experimental pollina- ristrestrustRLST R W 1987 collecting of finufpinus pinaceaepoaceaepinaceacpmaceae pollen byb tion biology reinholdremholdreinhoidramhold new york osinia bees hymenoptera megachilidae environen iron zabZAPZAR J H 1974 statisticalbiostatisticalBio analanaianalysissis prentice hall new mental entomologyv 16 668 671 jersey SAS 1990 sasstatSAS stat users guide vols I1 and 2 SAS zilzitZIMMERMANMERMAN M 1980 reproduction min polemonium com- institute can north carolina petition for pollinatorspollinators ecology 6149761 497 501 silberglied R E 1979 communication in the ultraultraviovio ZIMMERMAN VI AND G PIKE 1988 reproduction in let annual review of ecology and systematics 10 bolemonPolemonpolemonimnpolemomumimn assessing the factors limiting seed set 373 398 american naturalist 131 723 738 SPIRA T P 1983 reproductive and demographic charac- teristicste of alpine biennial and perennial gentians received 2 july 1993 gentiana sppapp in the white mountains unpublished accepted 25january25 january 1994 doctoral dissertation university of california berkeley Great Basin Naturalist 54(4), © 1994, pp. 359--365

USE OF A SECONDARY NEST IN GREAT BASIN DESERT THATCH ANTS (FORMICA OBSCURIPES FOREL)

James D. McIverl and Trygve Steen2

ABsTRACI:-Workers ofGreat Basin Desert thatch ants (Formica obscuripes Forel) dig simple secondary nests at the base ofplants upon which they tend aphids and scales. These secondary nests house only foragers, with the number of foragers occupying each nest positively correlated with the number of worker-tended Homoptera feeding on plant foliage above. Thatch ant secondary nests are cooler than 25 em below the dome top ofthe primary nest and maintain a sigoificantly more constant temperature than is observed on the ground surface or in the plant canopy. Thatch ant foragers use secondary nests for at least two purposes: as a cool refuge for Homoptera tenders when midday plant canopy tem­ peratures rise dUring the summer months, and as the primary place within which Homoptera tenders transfer honey" dew to larger "honeydew transporters" for ultimate transport back to the primary nest.

Key words: honeydew harvest, thermal refugia, behavioral thermoregulation, red wood ants, desert adaptation, satel­ lite nests.

Although most ant specIes use a nest struc­ STUDY AREA AND SPECIES ture consisting of a single central location (a primary nest), many species also employ "sec­ Thatch ants were studied between June ondary" nests in which a portion ofthe colony 1987 and September 1991 at Pike Creek, 160 population is dispersed among several alter­ Ian southeast ofBurns, Oregon. The Pike Creek nate sites (Wheeler 1910). study site is at 1300 m elevation at the base of Several species ofCamponotus, for example, Steen's Mountain in the northern Great Basin use a secondary nest to which workers transport Desert. Sagebrush (Artemisia tridentata), rabbit­ late-instar larvae and pupae from a centralloca­ brush (Chrysothamnus nauseosus), horsebrush tion occupied by the queen and brood (Hansen (Tetradymia sp.), lupine (Lupinus caudatus and Akre 1987). Similarly, the dolichoderine Kellogg), and cheatgrass (Bromus spp.) are lridomyrmex sanguineus maintains secondary dominant plants at the site, which was grazed nests containing older larvae and ptipae, but moderately by cattle throughout the study workers bring young from several locations period. within oligogynous colonies (McIver 1991). A total offour colonies ofFormica obscuripes Many other ant species (Polyrachis simplex, Forel were observed for various parts of the Lasius niger; L. emarginatus, Formica pratensis, study. E obscuripes is a widespread and abun~ E exsectoides, Crematogaster pilosus) are known dant North American rufa-group species to use secondary nests in which only foraging (Wheeler and Wheeler 1983). Like E rufa­ workers reside (Forel 1921, Andrews 1929, group species elsewhere, E obscuripes builds Ofer 1970). These secondary nests are thought symmetrical, dome-shaped primary nests of to serve as refuges for the workers from the thatch, from which radiate trunk trails that physical environment, as a defense against access foraging territory. In all four study enemies, or as a protected site within which to colonies workers foraged for honeydew on tend Homoptera for honeydew (Wheeler 1910). sagebrush, rabbitbrush, horsebrush, and/or This paper characterizes the secondary nest lupine, and scavenged for arthropods in the area used by the thatch ant Formica obscuripes surrounding each nest. Although broodless Forelliving in the Great Basin Desert and dis­ satellite nests were occasionally observed, cusses its possible function within the context there was no evidence of primary nest poly­ ofthe desert environment. domy in any study colonies.

1Deparnnents ofBiology and Edncation, Eastern Oregon State College, La Grande, Oregon 97850, and Blue Mountains Natural Resources Institute, 1401 Gekeler Lane, La Grande, Oregon 97850. 2Deparnnent ofBiolog;; Portland State University, Box 751, Portland, Oregon 97207.

359 360 GREAT BASIN NATURALIST [Volume 54

METHODS AND MATERIALS tended Homoptera at Pike Creek during the study period. Viewed from above, secondary Secondary Nest Charactelistics nests were simple openings in the ground The aboveground structure of the sec­ adjacent to plant trunks (Fig. 1). Ground ondary nest is portrayed by a photograph taken around an opening was typically littered with from colony 5 at Pike Creek, August 1988. The thatch material, fallen from the plant canopy, belowground structure was investigated by blown in, or excavated from the gallery pouring a measured quantity ofdentallabstone beneath. down 10 different secondary nest entrances of Volume of10 secondary nests beneath active two colonies (colonies 4 and 26) during August tending groups of workers ranged from 35 to 1991. Quantity oflabstone required to fill each ' 125 cc. Secondary nest volume was not signifi­ secondmy nest was then correlated with basal cantly correlated with basal plant dimneter (R2 plant dimneter and number ofworkers tending = .02, P > .05, N == 10) but was significantly Homoptera in the plant canopy. Actual struc­ cOlTelated with number of tenders (R2 = .33, ture of the secondary nest interior was deter­ P < .05, Y = .54X + 43.3, N = 10). mined by excavating two nests, photographing Excavations of secondary nests into which the labstone "plug" in place, and drawing one labstone had been poured revealed that cavities ofthese to scale using the photograph as refer­ essentially conformed to morphology of the ence. plant trunk itself (Fig. 2). Thatch ant workers Temperature at 6 cm depth in a typical sec­ typically removed dirt, small stones, and other ondary nest (plant A, colony 2) was measured debris from within 5-20 mm of the plant during summer 1987 and compared to mea­ trunk, leaving a cavity punctuated with large surements for tending localities in the plant stones and roots. The nest represented in canopy, ground surface, and 25 cm below the Figure 2 was 10.8 cm deep and consisted of top ofthe primary nest dome. three separate chmnbers totaling 175 cc in vol­ Secondary Nest Use ume. Temperature within the secondary nest dif­ Use of the secondary nest by thatch ant fered considerably from temperatures record­ workers was explored by conducting intensive ed Simultaneously on the ground swface, in the observations on a selected sagebrush plant plant canopy, or deep within the primary nest (plant 13) at the Pike Creek study site during (Fig. 3). Over the l-wkperiod 13-19 June 1987, July 1987. Beginning 1 July 1987, thatch ants for exmnple, the secondary nest we measured working in the vicinity of plant 13 were indi­ vidually marked with "beenumbers" (Charles was an average ofabout 1°C cooler than 25 cm Graz Co., Frankfurt, Germany) so that the from the dome top ofthe primary nest (18.8 ° vs. activity pattern of each could be determined. 26.1°), with a little over twice the variance over By 23 July, a total of 66 workers had visited time (12.6 vs. 5.7). Compared to ground surface, plant 13 and been marked, 30 of which were the secondary nest was slightly cooler (18.8 ° still using the plant daily. At noon on 23 July, vs. 19.2°) but much less variable, exhibiting a we began a 24-h continuous period ofobserva­ variance of about one-ninth the ground sur~ tion ofworker behavior on plant 13. We record­ face (12.6 vs. 112.4). Compared to the canopy of ed the location and task of each worker at 15­ the smne plant, the secondary nest was slightly min intervals throughout the 24-h period and warmer on average (18.8° vs. 18.0°) but about noted its interaction with other workers. The one-fifth as variable (12.6 vs. 67.1). Temperature result was a time budget for 30 different work~ trends over the entire summer were similar to ers that frequented plant 13 during the 24-h those measured in this l-wk smnple period in period, from which we could infer how work­ lnid-June. ers ofvarious task specializations used the sec­ Secondary Nest Use ondary nest at the plant base. Observations of individually marked work­ RESULTS ers on plant 13 of colony 2 clearly show that the secondary nest is used throughout the day Secondary Nest Characteristics (Fig. 4). The greatest percentage of workers Thatch ant secondary nests were found at was found in the secondary nest during lnid­ the base of each plant upon which workers afternoon, cOlTesponding to highest daily tem- 1994] SECONDARY NESTS IN THATCH ANTS 361

Fig. 1. Aboveground appearance ofsecondary nest at base ofsagebrush plant, Pike Creek, southeastern Oregon, June 1994 (photograph by Trygve Steen). peratures. Secondary nest population was low­ mary nest, worker 13 often had a distended est between 1700 and 2000, and between 0600 gaster, indicating a crop swollen with honey­ and 0900, during principal times when work­ dew. Typically, workers like #13 were scaven­ ers deliver honeydew to the primary nest. gers, secondary nest excavators, and/or honey­ Two typical patterns of activity were dew transporters, receiving the majority of observed for plant-associated workers (Fig. 5). their honeydew from workers that concentrat­ Tenders spent the majority oftheir time tending ed on tending Homoptera in the plant canopy. Homoptera for honeydew in the plant canopy. Of the 30 workers associated with plant 13 Worker 84, for example, spent 54% ofher time during the intensive observation period, 19 tending aphids, with each visit to the plant were classified as tenders, 6 as honeydew transporters/scavengers, 2 had behavior inter­ canopy lasting between 2 min and about 3 h. mediate between tender and b'ansporter/scav­ Her visits to the plant canopy were inter­ enger, and 3 were not observed often enough spersed with frequent visits to the secondary to classify. nest at the plant base, where it is likely she transferred honeydew to larger nontending DISCUSSION individuals like worker 13 (chain transport). Twice per day she returned to the primary Great Basin Desert thatch ants use sec­ nest: once in the early evening and once in the ondary nests as a refuge from high midday morning. temperatures and as a site within which honey­ Honeydew transporters spend the majority dew is transferred from workers who collect it of their time in the secondary nest itself. in the plant canopy to those who help trans­ Worker 13, for example, spent 66% ofher time port it back to the primary nest. Ground tem­ in the secondary nest, 23% scavenging on the peratures above 50 0 C have been reported as ground surface, and 9% on twice-daily returns lethal to F. obsGuripes (O'Neill and Kemp to the primary nest. On her returns to the pri- 1990), and Mackay and Mackay (1984) 362 GREAT BASIN NATURALIST [Volume 54

.~

Q o

o Fig. 2. Scale drawing ofsecondary nest, taken from photograph of labstone plug, Pike Creek, southeast­ ern Oregon, August 1991.

observed that E haemorrhoidalis workers hide Thatch ants living at other sites in the Great under pine cones or retreat to shady places Basin also use secondary nests of this kind during midday heat. Chain transport appears (McIver personal observation); Weber (1935) to be an effective way to increase delivery of described secondary nests in his study of honeydew to the primary nest (McIver and South Dakota thatch ants. However, Weber Yandell 1994); thus, it is not surprising that reported that the function of these nests was honeydew transfer occurs at a site offering to serve as (1) an arborescent chamber within refuge from midday heat. which to tend Homoptera and (2) a potential The use of cool midday refugia by workers site for development into primary nests. may also reduce metabolic costs and increase Certainly, colonies of Formica rufa-group worker longevity. In a study on fire ant thermal species often reproduce by budding (Mabelis preferences, Porter and Tschinkel (1993) report~ 1979; E polyctena), and the site of a new pri­ ed that fire ant workers consistently choose mary nest is very often a secondary nest cooler temperatures than those selected for (Scherba 1959, McIver personal observation). the brood. They postulate that this tendency It is not known whether E rofa-group species increases longevity of workers not directly living in other habitats employ secondary associated with brood care. This idea is sup­ nests for these or other reasons. ported by Calabi and Porter (1989), who Other Formica species are also known to demonstrated that because temperature and employ secondary nests. The mound-building metabolic rate are highly corrrelated, fire ants ant E exsectoides (exsectoides-group) uses sec­ reared and maintained under high tempera­ ondary nests as shelters for b:eehoppers and as ture regimes have lower longevity. sites for food exchange (Andrews 1929). 1994] SECONDARY NESTS IN THATCH ANTS 363

GROUND Pike Creek Colony 2 -- 1987 50 / o 40 o w a: PRIMARY, 25cm below dometop ::> 30 ,.... ~ , w ,. a.. , :lE 20 .: •...... \ w ...... -'. I-

10 SECONDARY

o 8 16 0 8 16 0 8 16 0 8 1$ 0 8 1$ 0 8 16 0 a 1$ 13 June 14 June 15Jun~ 16June 17 June 18 June t9June

TIME

Fig. 3. Temperature (OC) during week of 13-19 June 1987, on ground surface, in sagebrush canopy, 25 cm below dome top ofprimary nest, and in secondary nest ofcolony 2, Pike Creek, Oregon.

Formica integra of North America and F. ported by the National Geographic Society pratensis of Europe construct secondary nests and the Systematic Entomology Laboratory of along covered paths (Wheeler 1910, Forel Oregon State University (Dr. John Lattin), 1921). where voucher specimens are held. Other Homoptera~tendingants, including the formicines Lasius niger (Forel 1921), L. LITERATURE CITED emarginatus (ForeI1921), L. flavus (Soulie ANDREWS, E. A. 1929. The mound_building ant, Fonnica 1961), and Polyrachis simplex (Ofer 1970), and exsectoides E, associated with treehoppers. Annals of the myrmicines Crematogaster pilosus (Fore! the Entomological Society ofAmerica 22: 369-391. 1921) and C. auberti (Soulie 1961), use sec~ CALABI, P., AND S. D. PORTER. 1989. Worker longevity in ondary nests as shelters for their homopteran the fire ant Solenopsis invicta: ergonomic considera­ tions ofcorrelations between body size and metabol­ symbiotes. ic rates, Journal of Insect Physiology 35: 643-649. FOREL, A. 1921. Le monde social des fourmis du globe. ACKNOWLEDGMENTS Vols. 1-3. Geneve, Libraire Kundig. 1921-1923. HANSEN, L. D., AND R. D. AlmE. 1987. Biology of carpen­ ter ants. Pages 274-280 in R. K. Vander Meer, K. Bryce Kimberling drew the secondary nest Jaffe, and A. Cedeno, eds., Applied myrmecology-a from photographs. We thank Courtney Loomis, world perspective. Westview Press, Boulder, Deborah Coffey, Joseph Furnish, and Bill Colorado. Clark for assistance in the field. Jeffrey C. MABELIS, A. A. 1979. Nest splitting by the red wood ant (E polyctena Foerster). Netherlands Journal of Zool­ Miller provided the datapod for temperature ogy 29: 109-125. recordings. Andre Francour kindly identified MACKAY, E. E., AND W P. MACKAY. 1984. Biology of the Formica obscuripes Fore!. Research was sup- thatching ant Formica haemorrhoidalis Emery 364 GREAT BASIN NATURALIST [Volume 54

• SI;CONDARY NEST D PLANT CANOPY !ill] GROUND 1]1] PRIMARY NEST

...... I- 30 c::i fi) 15 ~ 6 Cf) Cf) ....J W r:r:w o ~ o a: -..- o 1-25 ~ ~ 10 ::> o ..... w a:

o I 1200 1400 1600 1800 2000 2200 2400 200 400 600 800 1000 1200 TIME OF DAY

Fig. 4. Activity ofmarked workers ofplant 13, Pike Creek colony 2, 23-24 July 1981. Number ofworkers observed in secondary nest, in plant canopy, on ground, at primary nest, and temperature in degrees Celsius over 24-h period.

% TlMEi LOCATION Worker 84: tender, Honeydew Transporter 54% PLANT CANOPY -

2% GROI,IND

36% SECONDARY NEST '- '---- '-- '-

~ ~ ~ ~ ~ 12 13 14 15 18 17 18 18 1 2 3 4 5 e 7 8 10 11 12 mJdnlQI't

8% PRIMARY NEST

Worker 13: Scavenger, Honeydew Transporter 2% PLANT CANOPY

23% GROUND

66% SECONDARY NEST

12 13 14 15 18 17 18 1; ~ ~ ~ ~ ~ 1 2 3 4 5 e 7 8 a 10 11 12 mIdnIott

9% PRIMARY NEST

Fig. 5. Activity over 24-h period ofworkers 84 and 13 onplant 13, colony 2, Pike Creek, Oregon, 23-24 July 1987. 1994] SECONDARY NESTS IN THATCH ANTS 365

(Hymenoptera: Fonnicidae). Pan-Pacific Entomology SOULIE, J. 1961. Les nids et Ie comportem~nt nidificateur 60: 79-83. des Fourmis du genre Crematogaster d'Europe, McIvER, J. 1991. Dispersed central-place foraging in d'Afrique du nord et d'Asie du sud-est. Insectes Australian meat ants. Insectes Sociaux 38: 129-137. Sociaux 8: 213-297. ' McIVER, J., AND K. YANDELL. 1995. Honeydew harvest in WEBER, N. A. 1935. The biology of the thatching ant, Great Basin Desert thatch ants. National Geographic Formica rufa Obscllripes Fore!, in North Dakota. Research 11: In press. Ecology 5: 166-206. l OFER, J. 1970. Polyrachis simplex, the weaver ant ofIsrael. WHEELER, w: M. 1910. Ants. Their structure, develop­ Insectes Sociaux 17: 49-82. ment and behavior. Columbia University Press, New O'NEILL, K. M., AND w: P. KEMP. 1990. Worker response York. 663 pp. to thermal contraints in the ant Formica obscllripes WHEELER, G. C., AND J. N. WHEELER. 1983. The ants of (H:F). Journal ofThermal Biology 15: 133-140. Nevada. Natural History Museum of Los Angeles PORTER, S. D., AND w: R. TSCHINKEL. 1993. Fire ant ther­ County, Los Angeles, California. 138 pp. mal preferences: behavioral control of growth and metabolism. Behavioral Ecology and Sociobiology Received 24 August 1993 32: 321-329. Accepted26January 1994 SCHERBA, G. 1965. Analysis of inter-nest movements by workers of the ant Formica opaciventris Emery. Animal Behavior 12: 508--512. great basin N aturahstnaturalist 544 C 1994 appp 366 370

SPAWNING chronology AND LARVAL EMERGENCE OF JUNE SUCKER chasmistes LIORUS

timothy modde 12 and neal muirheadlM mi head1headd

ABSTRACTABSTILACT june sucker chasinistescha&imstes horus spawned in the prooprovo rivelriver utah over a 2 wk period in earl june dur- ing both 1987 and 1988 Emeemergentigent larvae emigrated from the nverriver to utah lake over a 2 to 3 wk period drift into the lake peaked between 1200 and 0400 during daylightda light hours emergent larvae tended to occur in pools peak emergence of larval drift was approximately 121 2 larvaem3 during late june in 1987 and 1988 recruitment failure of june suckerstickersueker is not due to reproductive failure

key worthwords june sucker chasmistes liorusdiorus spawning larvae habitat drift emergence ahnwermerriverriven

the june sucker chasinisteschasmistes horushornshofus is one of june sucker have been described as spawn- three contemporary species of the genus ing on gravel cobble substrate in relatively chasinisteschasmistes miller and smith 1981 and is high velocity habitats radant and hickman endemic to utah lake a 38000 ha remnant of 1984 sex products are broadcast over the prehistoric lake bonneville once june sucker substrate and eggs are adhesive to the sub- numbered in the millions jordan 1891 and strate shirley 1983 radant and hickman were one of the most abundant fishes in utah 1984 although information on spawning lake during the last century population size behavior and larval morphology shirley 1983 ofjuneof june sucker declined drastically in a survey snyder and muth 1988 exists no information of utah lake fishes less than 040.40 4 offishof fish col- is available on spawning success of the june lected were june sucker radant and sakaguchi sucker because natality is a vital element of 1981 of the population ofjunejune sucker has been recruitment information on spawning success estimated to 1000 and is be adults is listed on is important in understanding declining abun- the federal register as an endangered species dance of this species the objectives of our USU S fish and wildlife 1986 suspected service study weiewelewere to 1 estimate timing and magni- factors contributing to the decline of this tude of downstream drift of emergent june loss species include water to irrigation and sucker larvae and 2 describe habitats occu- drought degradation of water quality and pied by larval june sucker in the provo river negative interactions with nonnative fishes and 1984 radant hickman reduction of METHODS water quantity and quality impacted both the lake and spawning tributariestributaries drift sampling the direct cause of decline in the june sucker population has been lack of recruit- drift netting was conducted in the lower ment sigler et al 1985 in a survey of utah provo river to capture emergent larvae during lake radant and sakaguchi 1981 did not the 1987 and 1988 spawning periods netting capture any june sucker 400 mm total began 1 june 1987 and terminated when lar- length scoppettone 1988 reported that june vae ceased to appear in collections five drift sucker mavmay live to be 42 years of age thus in nets each with a mouth size of 30 X 45 cm and the absence of recruitment senescent individ- a mesh size of 560 microns were placed at a uals would dominate the population none of single site about 3 km upstream of utah lake the 18 fish he examined was younger than 20 immediately downstream of the lowermost vearsyears of age observed june sucker spawning activity nets

rational biological survey utah cooperative fishlish1 ish and ilcllifewildlife research unit department of fisheries and wildlife utah state Lnierlniersituniversitysit logan utah 84322521084322 5210 theth utah Ccooperativep ativeactive fish and itdltdildliteildwildlifeliteilfe research unitL nit isis a program jointly sponsored byh the USL S department of interior liahutah Didiviidiviadivviivilision ofnvildlifeof ildhfe resource and utah state lnnersirv 2centscentfresentaddrtaddress colorado river fish project LSUS fish and wildlifeidhfe seniceaenicesebsen ittlctlet 266 westaiteittytiit 100 north suite 2 vcvenial utah 84078

366 199411994 JUNE SUCKER SPAWNING 367

were anchored with 0 64 cm diameter rebar analysis along a single transect perpendicular to the means for egg and larval density in the channel when depth permitted nets were in drift were determined for daily and 4 h placed alternately at the surface and bottom peri- ods standard deviations were calculated from in 1988 drift nsmg the same sampling netting nimgusing daily means among periods and for periods scheme began 6 was june the netting site with days as replicates drift densities were moved about 50 of 1987 in downstream the estimated by dividing eggs and larvae collected site because of physical changes chan- in the during each sampling period by watelwater volume nel only four nets were used during 1988 passing through drift nets daily estimates nets were set on alternate days MWF were determined by computing the means of each hb week each 24 day was divided into six all six time periods estimates of total larlarvaeae- 4 h periods and drift was sampled continu- on the peak drift date were determined bvby ously during the middle 151.51 5 h of each starting averaging discharge recorded at the provo times were 1315 1715 2115 0115 0515 and city gauge station USGS on both davsdays sam- 0915 h drift from each net was rinsed placed ples were made and multiplying the volume in watertight plastic bags and preserved in estimate by daily mean larval density 5 buffered formalin 420 samples were because of the few sites in which june taken sucker larvae were present habitats were velocity losee10 sec average through each net grouped into pool and nonpool categories and water depth were measured before and chi square analysis was used to test the signif- after each set volume sampled was estimated icanceacance of differences in the incidence of larval by multiplying the average of the two velocity june sucker in pool and nonpool habitats and measurements by time sampled and area of odds ratio analysis fienberg 1980 was used the net opening water temperature was re- to quantify the magnitude of differences corded during each 4 h interval all samples observed were sorted for eggs and larvae which were identified to species snyder and muth 1988 RESULTS counted and measured to the nearest 0100.1oi 1 mm total length drift spawning as defined by egg drift was high- habitat sampling er on 3 4 june 1987 and peaked on 6 7 june fishes in a 2259252 25 km section of the lower 1988 figs 1 2 A malfunction of the velocity provo river were sampled during the 1988 meter on 6 and 8 june 1988 prevented accu- spawning season to determine larval habitat rate estimation of egg and larval concentra- use eighty four transects about 27 m apart tions however absolute numbers of eggs cap- and perpendicular to the thalthalwegweg were estab- tured on 7 june 000700070.0070 007 eggsseceggs sec and 8 june lished from aerial photographs of the river 0005000300050.0050 005 eggsseceggs sec exceeded those caught on 11 three samples were taken along each transect june 0 0007000070.0007 eggsseceggs sec average river tempera- tures during the one dearnear each shoreshofe and one in the middle of spawning period were 13 in 1987 and 12 in 1988 spawn- the river samples were collected with a im2ima1 m2ma 14c 17c in ing occurred over a relatively short time eggs bag seineseme with a 560 micron mesh substrate were collected foifolforoor I1 wk in 1987 and I111 I1 d in in a I1 m2ma area immediately in front of the in in 1988 spawning duration was probably longer seineseme was mechanically stirred at each sam- in both years than shown in figures I1 and 2 pling site and the seme was quickly pulled seine because eggs were already present in the river through in river samples were taken only during day- when sampling began however collections light hours from both years suggested that june sucker habitat types were described and widths spawning activity does not last more than 2 measured along each transect using a modifi- wk with the greatest number of eggs spawned cation of bisson et al 1982 all fish collected within a 3- 1987 to 5 d 1988 period were placed in plastic containers and pre- density of egg drift was variable and showed served in 5 buffered formalin larvae were no diel pattern fig 3 thus either fish were identified to species measured to the nearest spawning in both light and dark hours or eggs 010.1oi0 1 mm total length and counted were being randomly dislodged from the 368 GREAT BASIN naturalist volume 54

14 1987 larval density 14 1988 larval densitydensify 12 10 12

E 1 E L L 080.8 ID 06 E 06 z 04 04 02 02 t 0 0 1 1 3 5175 1 11 13 1111115 171911 211 1 231 11110511051725272911 1 VY 517 191113151711912 1z3 52729 008 0080.08 1987 egg density 1988 egg density 10 006 E 006 004 0040.04 E E 3 002 0020.02

n 1 0 1 1 1 1 1 Y 3 7 57117 1 1 1 1 1 5795 9 1113faf3 f 7192119 2325223 25 27729if 31 5 7 9111111133 15 1 2 12325127291 1

24 1 1988 temperature 20 1987 temperature 20 16 j T 16 12 12 cl 8 E af8f E 4rar aj4j 0 1 1 1 1 5 7 13 15 17 192 12 9 1 1 1 1 1 y3tst7w2v2s252729v32 52 72 01 53 5 7 9 11 1315171921232527293 15 17 19 2 23 25 27 29 1 579 june june

fig 1 drift rates of june suekersuckerstickerslickerslieker lanaelarvae and eggs and fig 2 drift rates of june suckelsticker larvae and eggs and daikdalkdalhdailydallydaliy average temperature collected from the prooprovo river daildalldalidailydaliy average temperature collected from the prooprovo river utah in june 1987 etabutahutab in june 1988 substrate throughout the 24 h period during larvae in the provo river were approximately drift netting operations june sucker were 60200 in 1987 and 73000 in 1988 observed during both day and night spawning habitat use larval june sucker fustfirsthirstbirst appeared in the drift on 3 june 1987 and 6 june 1988 fig 1 A total of 57 june sucker larvae were col- although velocity error precluded absolute lected inm 7 of 115 collections incidence of lar- measurement until after 10 june few larvae vae in poolpooi type habitats was different from were collected until 20 21 june peak densi- nonpool habitats xax2 70477.0404 05 5995.995 99 june ties of larvae in the drift occurred on 22 23 sucker larvae were 757.57 5 times more likely to be june 1987 and 22 23 june 1988 minimum found in pool than nonpool habitats during estimates of the time between egg deposition daylight hours and swim up measuimeaskimeasureded as the period between peak egg drift and peak larval drift were 19 d discussion in 1987 and 16 d in 1988 the difference in incubation time between years is probably shirlevshirley 1983 reported june suckersuekerstickernucker spawn- due to warmer rivernver temperature in 1988 ing in mid june when mean water tempera- 15 19c than in 1987 12 16c drift of ture was between 11 and 13c13 C similar obser- june suckersuekerhuekelbuekel larvae continued for about 3 wk vations were made by radant and hickman during both study years all june sucker lar- 1984 and radant and sakaguchi 1981 vae collected weiewelewere identified as either proto radant and hickman 1984 also observed a or mesomesolarvaelarvae short spawning period that lasted only 5 8 d A distinct daily pattern of larval drift densi- the cuieuieul ui chasinisteschasrmstes caiuscuiuscujus also spawns dur- ty was observed with most larvae captured ing a brief period males occupying the truckee between 2000 and 0400 h fig 4 few larvae river nevada 656.56 5 16516.516 5 d and females 4044.00 10510.510ioslos 5 were collected in driftdrinn nets during daylight d scoppettone et al 1986 temperatures of peak dailydaliy estimates of drifting june sucker the truckee river during cuieuieul ui spawning 19941 JUNE SUCKER SPANNINGSPAWNING 369

007 35 1987 1987 006 3

005 25 E 004 2

003 E 15 z 0 002

001 05

0 0 0200 0600 1000 1400 1800 2200 0200 0600 1000 1400 1800 2200 007 35 1988 1988 006 3

005 25 E E 004- 2

E 003 E 15

002

001- 05

0 0 0200 06d0 000ooo1000 1400 1800 2200 0200 0600 1000 1400 1800 2200 hour hour

fig 3 diel drift rate of june sucker eggs collected in fig 4 diedlediel drift rater ite of june sucker larvaeianlan ac collected in the provo rieirielriver etahutah in june 1987 and 1988 verticalei tical the proveprovo river utah in june 1987 and 1988 vertical baibalbars s lepresentrepresent one st edardmdardstandard dedeviationlation barshars represent one standardstand ud deviation ranged from 12 to 15c15 C cui ui spawned provo rivelriver foiroirolfor an extended time most larvae between 2000 and 0600 it during a 3 d period drifted out of the provo riverrivel during a 2 to 3 scoppettone et al 1981 whereas egg drift wk period whereas euicuieul ui were reported to densities of june sucker and observation dur- drift through the truckee rivelriver for nearlnearinearly 30 d ing both 1987 and 1988 indicated spawning differences between the two species in dura- occurred during all hours of the day and night tion of larval emergence and drift may result scoppettone et al 1983 reported peak emer- from a larger euicuieul ui spawning population gence ofcuiof cuieuieul ui larvae occurred 14 d after peak although the abundant numbers of june spawning differences between peak june sucker larvae produced in 1987 and 1988 are suekersuckersticker spawning and peak emergence varied surely less than histonehistoric numbers substantial between vealsyears fornhornfrombomm 19 d in 1987 tempera- numbers of larvae drifted into utah lake ture langeiangerange 13 15c15 C to 16 d in 1988 tempera- sigler et al 1985 suggested the decline in ture range 17 19c19 C the pyramid lake euicuieul ui population was due like euicuieul ui scoppettone et al 1986 and to fadurefailurebadure of natural reproduction based on other catostomids geen et al 1966 june the largelailalial ge numberss of larvae captured in the sucker larvae emigrate from spawning tribu drift despite the relareiarelativelytivel small population of tarvcarvtaital into receiving lake shortly after emer- adult june suckelsuckersueker insufficient spawning or gence drift activity of larval june suckelsuckerss was emergent success seemingly did not limit neailynearlyneally identical to that of cuieuieul ui most drift recruitment to utah lake instead factors oceuringcuring just prior to 0000 and declining to affecting survival after larval emergence such negligible numbers by 0600 in spite of large as nonnativenormative predatorspredatoistols seem likely numberss of larvae capturedcaptaicaptui ed in the drift rela- tively few were captured bby seiningbeining those acknowledgmentsACKNO ledcatfnts larvae seinedheined diningduring daylight hours were mostlvmostlymostov in poolpooi type habitats as portedreportedle bvby this stud was funded bv the utah both radant and hickman 1984 and shllieshirlev division of wildlife resources and the USU S 1983 few if any larvae remained in the fish and wildlife service randrandy radant 370 GREAT BASIN naturalist volume 54

assisted with all phases of the study and pro- radaeranaeRANTR ANT R D ANDND D K sakaguciiisakacSARAC LCIII 1981 utah likelake vided information on the history of june suck- fisheries inventinventoryorv united states bureau of er spawning in the provo reclamation contract 80740506348 07 40 50634 modificationlklodification river dennis 04 division of ildlifewildlife resources salt lake city shirley was instrumental in providing field utah coordination and logistical support at the scoppettonesc OPPETTONE G G 1988 growth and longevity of the study site roger mellinthinmellenthinMellinthin brad schmitz cuieuleui ui and longevity of other catostomids and and don archer provided field assistance cyprincypnmdscyprinidsids in western north america transactions of the american fisheries society 117 301 307 ScOPPETTOscoppettoneMb G G M COIFMACOLEMAN H birgeBLRGEBURGE AMAND G literature giteCITECITED WEDEMEIERWEDEMEYER 1981 cuiculgulGUI ui life history river phase annual report united states fish and wildlife bissoBisbissonsonsox P A J L NIELSEN R A palusoPAIUSOPALMASON AND L E service fisheries assistance office reno nevada croeGROEcrovegnoeGROVE 1982 A system for naming habitat hpesapestypes in 63 appp small streams with examples of habitat utilization byb scoppettone G G M COLEMAN ANDD G A WEDEMFR salmonsaimonsalmondssalmonidssalmomdsids during low stream flow pages 62 73 init 1986 life history and status of the endangered cuieuieulcm N B armantrout ed acquisition and utilization of ui of pyramid lake nevada fish and wildlife aquatic habitat inventory information western didiviulvii- research 1 1 23 sionslon of the american fisheries society bethesda scoppetioescoppermne G G G A whdemeifrwedenieyerwedenmeyer M golenCOLEUgolemanCOLEMAN marylandmarvmary land ANDND ifII11 BURGEBLRCE 1983 reproduction bby the endanbendan flfcnbergsFIENBERGS E 1980 the analysis of cross classified cate- ceredgered cuieuieul ui in the lower truckee river transactions gorical data MIT press cambridge massachusetts oftheodtheof the american fisheries society 112 788 793 GEEN G H T G NORTHCOTE G F HwHARTMANHARTriNMAN AND C C SHIRLEYSKIRLED D L 1983 spawning ecology and larval devel lindsetLINDSEILINDSEY 1966 life histories of two species of oimentopment of the june suckelsuckersueker proceedings of the catostomid fishes in sixteenmileSixteen mile lake british bonneville chapter american fisheries society columbia with particular reference to inlet stream 1983 18 36 spawning journal of the fisheries research board SKIERSIGLER W FE S vlotviceVIGG ANDD M BRES 1985 life history of of canada 23 1761 1788 the cuieuieul ui chasinistescha&mistes ctciousus cope in pyramid lake JORDAN D S 1891 report of exploration in colorado and nevada a review great basin naturalist 45 utah during the summer of 1889 with an account of 571 603 the fishes found in each of the nerriver basins exam- SDERSNYDER D E ADAND R T MLTHMUTH 1988 description mdand ined united states fisheries commission bulletin identification ofjuneof june utah and mountain sucker lar 9 1 40 vaeae and early juveniles contract 87289187 2891 division milmiiMILLER1 ER R R AND G R SMITH 1981 distribution and of wildlifeofwildlife resources siltsaltsait lake city utah evolution of chasinisteschasmistes pisces catostomidae in USU S FISH AND WILDLIFEwn DLIFE SERVICE 1986 endangered and western north america university of michigan threatened species listing and recoveryrecovervreco verv priority museum of zoology occasional papers 696 1 46 guidelines federal register 48 16756167759 RADANT R D ANDD T J HICKMAN 1984 status of the norus june sucker chasinisteschasmistes horus proceedings of the received 5 january5january 1994 desert fishes council 151983 277 282 accepted I111I1 april 1994 Great Basin Naturalist 54(4), © 1994, pp. 371-375

COMPARISON OF REPRODUCTIVE TIMING TO SNOW CONDITIONS IN WILD ONIONS AND WHITE-CROWNED SPARROWS AT HIGH ALTITUDE

Martin L. Mortonl

ABSTRACT.-Timing of reproduction was assessed for wild onions and White-crowned Sparrows in relation to snow conditions on the same subalpine meadow in the Sierra Nevada for 21 years. Flowering date and clutch 'initiation date were both highly correlated with snow conditions, being later as snowpack was deeper. Interannual variation in sched­ ule was 46 days for onions and 33 days for sparrows. There was nearly a fivefold difference in snowpack depth, and date ofsnow disappearance varied interannually by 72 days. Compensation for late-lying snows occurred in both species but was greater in sparrows than in onions because the nest-building behavior ofsparrows was flexible. In years of deeper snow, sparrows were able to lay eggs earlier because they built more nests than usual in trees and shrubs rather than waiting for grou~dcover to develop.

Key worM: Allium, Zonotrichia, snowpack, high altitude, proximate factors, reproduction.

Montane settings are useful for the study of and, to my knowledge, there are none that environmental adaptation in organisms have compared plants and animals on the because their brief, sharply delimited growing same study area. Herein I present 21 seasons seasons and variable climates can be potent of data that index reproductive schedules of agents of natural selection. Diurnal and sea­ the wild onion (Allium validum) and the sonal cycles of abiotic factors, principally air Mountain White~crowned Sparrow (Zonotrichia temperature (Ta), moisture, and wind speed, leucophrys oriantha) at the same location in shift in level and amplitude as elevation the Sierra Nevada in relation to interannual increases (Rosenberg 1974). The resulting variations in snow conditions. decrease in mean Ta, high winds, decreased availability of soil moisture due to freezing, MATERIALS AND METHODS and variable snowpackcan greatly influence the phenology, distribution, and productivity The study site, Tioga Pass Meadow (TPM), of plants (Billings and Bliss 1959, Scott and is a subalpine meadow with an area of about Billings 1964, Weaver and Collins 1977, Ostler 50 ha (0.5 X 1.0 km) and elevation of 3000 m et al. 1982). Annual schep.ules ofreproduction located in the upper end of Lee Vining and surVival ofhibernating mammals (Morton Canyon, Mono County, California. Itis bound­ and Sherman 1978) as well as reproductive ed by Tioga Lake on the northern edge and success ofboth sedentary (Clarke and Johnson the boundary of Yosemite National Park at 1992) and migratory birds (Morton 1978, Tioga Pass on the southern edge. Allium Smith and Andersen 1985) also are known to validum, which grows in large clumps in wet be affected, especially by spring storms and meadows at elevations of 1200-3350 m in the snowpack depth. It follows that long-term Sierra Nevada, is usually 0.5-1.0 m in height studies of the annual rhythm of reproduction and has numerous small (6-10 mm) flowers of organisms at high altitude should provide organized into teIDlinal umbels (Munz 1970). valuable information for understanding the For 21 summers, 1968-70, 1973, 1976, and pathways and scope of adaptations to climatic 1978~93, when I was on TPM daily, I kept conditions and for determining the efficacy of notes on the flowering schedule ofone partic­ environmental variations to act as cues or ular, fairly compact patch ofA. validum that proximate factors in the·control of reproduc­ covered an area of about 0.1 ha (.2.5 X 40 m) tion schedules. Such studies are few, however, near the center of TPM and that usually

IBiology Department, Occidental College, Los Angeles, California 90041-3392.

371 312 GREAT BASIN NATURALIST [Volume 54 contained about 1200 umbels. Technically, RESULTS flowering includes the period from floral bud initiation through floral persistence (Rathcke During this study snow depths ranged from and Lacey 1985), but I use the term to describe a low of 79.0 cm in 1976 to a high of375.1 cm the date on which buds opened to reveal the in 1983. The earliest 0% snowcover date was 1 mass offlowers within. In any given year this June 1992, and the latest 11 August 1983, a date varied by 2 wk or more among the vari~ range of 72 d. The earliest date for onion flow­ ous patches of A. validum scattered across ering was 6 July 1968 and 1976, and the latest TPM, but flowering within a particular patch, was 21 August 1983, a range of46 d. The earli­ including the study patch, was highly synchro­ est mean date for clutch initiations, based on nous, occurring within a 3-d period for most the first 10 nests of the season, was 27 May individuals. The last day of this opening peri­ 1992, and the latest mean date was 29 June od was noted every year and is the datum 1983, a range of 33 d (Table 1). Thus, repro­ used in this analysis. ductive schedule in relation to snowpack was The primary focus of my field studies on affected more in onions than in sparrows. In TPM was the reproductive biology ofZ. l. ori~ both species, however, timing of reproduction maxi~ antha, a migratory finch that winters in was highly correlated (P < .001) with disappear~ Mexico and breeds in montane meadows of mum snowpack (Fig. 1). The final the western United States. Individuals arrive ance of snow (0% snowcover) was tightly cou­ at breeding areas in May and June and depart pled to snow depth, and flowering and clutch for wintering grounds in September and initiation schedules were related accordingly to time ofsnow disappearance (Table 2). Slope October. Wet subalpine meadows like TPM are a preferred breeding ground habitat, but values for the relationship offlowering to both measures of snow conditions (snow depth and more xeric locations at lower elevations are 0% snowcover) were about twice those sometime utilized (Morton and Allan 1990). observed for clutch starts in relation to these Nests are built only by females and are placed same two measures (Table 2). Slopes for both on the ground or in shrubs, such as willows comparisons were significantly different (t (Salix sp.), or in small trees. Data gathered on tests, P < .001). banded females included the date they laid their first egg ofthe season, i.e., the clutch ini­ DISCUSSION tiation date. Herein I use the mean date ofthe first 10 clutch initiations on TPM each season Reproduction was delayed by deep, late­ to indicate the onset of reproduction in Z. l. lying snow more so in A. validum than Z. l. oriantha. Most nesting data were obtained oriantha, but there was a compensatory mech­ from females that were frequently observed anism operating to lessen the temporal impact and trapped (each had a unique combination ofheavy snows even in the onion. When flow~ ofcolor bands), allowing me to follow changes ering date was regressed on snow disappear~ in their behaviors, body weights, and brood ance date, the slope was 0.53, far less than 1.0 patches. This was important to data quality (Table 2). How then does the onion adapt? because females quickly renested ifa nest was Only three major physical environmental fac­ lost from storms or predation. tors have been identified as cues that initiate Information on snow conditions was avail~ flowering: temperature, moisture, and photo­ able becaus'e, first, I estimated from direct period (Rathcke and Lacey 1985). Photoperiod­ observations the snowcover on TPM as the responsive or long-day plants are relatively season progressed up to the day when all unaffected by snowpack because they flower patches of snow had disappeared-the date of and are pollinated late in the summer. In con~ 0% snowcover. Second, information on snow trast, nonphotoperiod-responsive plants tend depth in TPM could be obtained because it is to bloom in spring or early summer, and a site traditionally used by the State of phenophases may be affected by as much as 6 California Department of Water Resources to wk by temperature and moisture conditions measure snow depth in order to predict water (Owen 1976). Some plants can "catch up," at storage and runoff. Maximum snowpack least somewhat, by condensing or telescoping occurs about 1 April, and this measurement is phenophases when delayed by overlying snow published in their bulletin 120. (Billings and Bliss 1959, Scott and Billings 1994] PLANT-ANIMAL REPRODUCTIVE SCHEDULES 373

TABLE 1. Twenty-one years ofdata on snow conditions, time offlowering inA. va1idum, and mean .of clutch initiations and nest locations in Z. 1. oriantha. The first 10 nests ofthe season were used to calculate mean date ofclutch initiation, and all nests found in a given season were used to calculate percentage of those built aboveground. Number of nests found per season: mean == 59.6, S.D. = 21.9, range = 18-100, total ofall years = 1252. Snow depth 0% snowcover Onion flowering Clutch initiation Aboveground Year (cm) (Julian day) (Julian day) (mean Julian day) nests (%) 1968 113.5 175 188 160.6 40.0 1969 342.1 215 233 170.1 72.3 1970 176.3 183 198 158.3 34.9 1973 204.2 181 207 163.1 59.3 1976 79.0 158 188 153.8 10.9 1978 263.4 215 222 177.2 50.8 1979 227.1 191 207 160.7 43.6 1980 262.6 214 217 177.0 52.4 1981 173.0 167 197 156.4 25.3 1982 294.4 216 225 169.0 43.8 1983 375.7 224 234 181.3 32.1 1984 205.0 197 208 158.3 44.6 1985 145.8 172 197 156.8 25.5 1986 243.3 213 213 160.9 56.9 1987 113.3 163 201 152.5 18.6 1988 121.2 170 198 161.3 ILl 1989 158.0 176 201 156.2 39.4 1990 90.9 172 196 157.4 30.0 1991 167.4 181 214 164.7 46.7 1992 108.2 152 201 146.8 37.0 1993 227.1 199 216 169.5 49.2

240 230 Onion Flowering R2 ~ 0.87 220 n ~ 21 years

Q) 210 • +.J 0 - 0 200 c 0 190 • • :3 180 o =-:> 8 170 160 Sparrow Clute h Initiation 150 :::>­ R2 == 0.67 o 140 ~-=-,---=...... J...~""""' ...l....,?--"""""'"",,"--~ .~="-_-~_-=J-----,~-__-_ ...... ~L--=--" ...... I ~ 50 100 150 200 250 400 Snow Depth (em)

Fig. 1. Date offlowering in A. validum and mean clutch initiation in Z. 1. oriantha as a function of snow depth on 1 April at Tioga Pass Meadow. 374 GREAT BASIN NATURALIST [Volume 54

TABLE 2. Slopes and coefficients ofdetermination (R~) for linear regressions involving date offlowering in A. va1idum and mean date of clutch initiation and percentage ofnests placed in sites off the ground in Z. 1. oriantha in relation to snow conditions (snow depth on 1 April and date of0% snowcover) at Tioga Pass Meadow. N = 21 years.

y X Slope R2(%) P

0% snowcover Snow depth 0.25 86.1 <.001 Flowering date Snow depth 0.15 86.9 <.001 Flowering date 0% snowcover 0.53 76.1 <.001 Clutch initiation date Snow depth 0.09 68.9 <.001 Clutch initiation date 0% snowcover 0.34 75.7 <.001 Aboveground nests Snow depth 0.12 40.2 .002 Aboveground nests 0% snowcover 0.47 42.4 .001

1964, Weaver 1974, Weaver and Collins 1977). The main point here is that when groundcover Alpine plants often store large quantities of was adequate for hiding nests, females seemed carbohydrate in underground organs; mobi­ to prefer nesting on the ground. When plant lization of these reserves to shoots can occur growth and development were impeded by even before snowcover is gone. Relatively high late-lying snow, they did not wait a long peri­ carbohydrate levels are then maintained in the od for this cover to develop, but.instead built a shoot portion until after fruiting, whereupon a greater proportion of their nests aboveground, return of peak reserve levels to underground usually in pines (Pinus sp.) and willows. Thus, parts occurs at the beginning offall dormancy behavioral plasticity in selection of nesting (Mooney and Billings 1960). The mechanism sites allowed Z. l. oriantha to proceed with whereby events in the growing season can be rearing young with less delay than might be accelerated has not been studied to my lmowl­ predicted from snow conditions or even from edge, but this seasonal cycle of transport and plant phenophases. utilization of stored energy must be a vital In summary; this correlative study presents constituent. temporal indices of reproductive schedules A puzzling aspect ofA. validum's flowering during 21 years in a plant and an animal occu­ response is that it was strongly affected by pying the same high~altitude environment, snow conditions, typical of nonphotoperiod~ thus permitting a comparison oftheir respons­ responsive plants that usually flower in May es to a proximate or environmental factor or June. A. validum flowers in July or August, experienced in common, namely interannual a time that is more typical of plants cued by variation in snowpack. Both organisms were long days (Owen 1976). Perhaps A. validum affected by this factor and both exhibited com­ has seasonal changes in its photoresponsivity pensatory adjustments of their schedules. The and is follOwing a mixed strategy energetically, adjustment was greater in the animal because using stored reserves early in the season and it possesses a basic trait, not present in the then later switching to a greater reliance on plant, that can be acted on by natural selec­ photosynthate of the current year, the latter tion, its behavior. being a trait common to photoperiodic species (Mooney and Billings 1960). LITERATURE CITED The lessened impact of snow conditions on reproductive schedule in Z. l. oriantha, as BILLINGS, W. D., AND L. C. BLISS. 1959. An alpine snow­ compared to A. validum, appears to occur bank environment and its effects on vegetation, because the bird has flexible nest-building plant development, and productivity. Ecology 40: habits. Only about 11% of all nests construct­ 388-397. ed at TPM in years, such as 1976, were CLARKE, J. A., AND R. E. JOHNSON. 1992. The influence of dry spring snow depth on White-tailed Ptarmigan placed in aboveground sites (Table 1). In wet breeding success in the Sierra Nevada. Condor 94: years, such as 1969, when snowpack was 622-627. unusually heavy, this increased to 72% (Table MOONEY, H. A., AND W. D. BILLINGS. 1960. The annual 1). The 1983 data seem anomalous, but hot carbohydrate cycle of alpine plants as related to spring weather induced rapid snowmelting, growth. American Journal ofBotany 47: 594-598. MORTON, M. L. 1978. Snow conditions and the onset of and more nests were placed on the ground breeding in the Mountain White-crowned Sparrow. than might have been otherwise expected. Condor 80: 285-289. 1994] PLANT-AN"IMAL REPRODUCTIVE SCHEDULES 375

MORTON, M. L., AND N. ALLAN. 1990. Effects of snowpack ROSENBERG, N. J. 1974. Microclimate. The biological and age on reproductive schedules and testosterone environment. John Wiley and Sons, New York. 315 levels in male White-crowned Sparrows in a mon­ pp. tane environment. Pages 235-249 in M. Wada, S. SCon; D., AND W D. BILLINGS. 1964. Effects of environ­ Ishii, and C. G. Scanes, eds., Endocrinology ofbirds: mental factors on standing crop and production of molecular to behavioral. Japan Scientific Society alpine tundra. Ecological Monographs 34: 243-270. Press, Springer Verlag, Tokyo. SMITH, K. G., AND D. C. ANDERSEN. 1985. Snowpack and MORTON, M. L., AND P. W SHERMAN. 1978. Effects of a variation in reproductive ecology of a montane spring snowstorm on behavior, reproduction and ground-nesting passerine, Junco hyemalis. Ornis survival of Belding's ground squirrels. Canadian Scandinavica 16: 8-13. Journal ofZoology 56: 2578-2590. WEAVER, T. 1974. Ecological effects of weather modifica­ MUNZ, P. A. 1970. A California flora. University of Cali­ tion: effects of late snowmelt on Festuca idahoensis fornia Press, Berkeley. 1681 pp. Elmer meadows. American Midland Naturalist 92: OSTLER, W K., K. T. HARPER, K. B. MCKNIGHT, AND D. C. 346-356. ANDERSON. 1982. The effects ofincreasing snowpack WEAVER, T., AND D. COLLINS. 1977. Possible effects of on a subalpine meadow in the Uinta Mountains, weather modification (increased snowpack) on Utah, U.S.A. Arctic andAlpine Research 14: 203-214. Festuca idahoensis meadows. Journal of Range OWEN, H. E. 1976. Phenological development of herba­ Management 30: 451-456. ceous plants in relation to snowmelt date. In: H. W Steinhoff and J. D. Ives, eds., Ecological impacts of Received 5April 1993 snowpack augmentation in the San Juan Mountains, Accepted 9 March 1994 Colorado. San Juan Ecology Project, Final Report, Colorado State University Publications, Fort Collins. RATHCKE, B., AND E. P. LACEY. 1985. Phenological pat­ terns ofterrestrial plants. Annual Review ofEcology Systems 16: 179-214. Great Basin Naturalist 54(4), © 1994, pp. 376-379

OPPORTUNISTIC BREEDING AFTER SUMMER RAINS BY ARIZONA TIGER SALAMANDERS

Linda J. Allison!, Paul E. Brunkow!, and James P. Collinsl

Key words: Ambystoma, Arizona, amphibians, opportunistic breeding.

Identifying factors influencing the number habitats (Houghton 1916). Some A. t. mavor­ oftimes organisms breed during a lifetime and tium populations breed both in spring and in the seasonal timing ofreproductive episodes is summer (Webb and Roueche 1971), and central to understanding the evolution of life Tanner et al. (1971) documented A. t. nebulo­ history traits (Stearns 1992). In this regard, am­ sum breeding in spring and summer in a con~ phibian reproductive cycles are often consid­ tinuously filled lake. In this studY:,we report ered adaptations to the seasonality of oviposi­ that A. t. nebulosum populatio~s can also tion opportunities Goly 1971, Lofts 1974, Salthe breed twice per year in ponds that fill with and Mecham 1974). This is seen in most north water during the winter, dry during the late temperate zone amphibians that breed in tem~ spring and early summer, and refill during porary aquatic habitats and have annual repro­ summer rains. ductive cycles (Bishop 1947, Wright and Arizona tiger salamanders (A. t. nebulosum) Wright 1949). These frogs and salamanders are found commonly at high elevations in typically come into reproductive condition montane Colorado and Utah (Stebbins 1985) once a year at the same time of year, often in and in Arizona between 1500 and 2900 m spring as ephemeral habitats predictably fill (Collins 1981). Aquatic habitats discussed in from snowmelt or winter rains. Where aquatic this study are in Rocky Mountain montane habitats fill unpredictably or irregularly, such conifer forest (Pase and Brown 1982). as in dry temperate or tropical regions, Arizona tiger salamanders breed regularly amphibians may have acyclical reproductive in late winter and spring following snowmelt periods allowing them to breed opportunisti­ (Sexton and Bizer 1978, Collins and Cheek cally (Salthe and Mecham 1974, van Beurden 1983, Holomuzki 1986, Jones and Collins 1979). Opportunistic breeding in ephemeral 1992). During the course of other fieldwork, habitats is commonly understood as an adapta_ we realized that a second breeding pattern tion for aVOiding predaceous fish (e.g., Webb also occurs. Here, we present observations that 1969, Heyer et al. 1915, Wilbur 1977, Collins led us to conclude that these tiger salamanders and Wilbur 1979). can breed "opportunistically," defined as any Tiger salamanders (Ambystoma tigrinum breeding outside the usual late winter and Green) range across North America and have early spring breeding period. an obligatory aquatic larval stage (Stebbins On 30 March 1990 we noted typical spring 1985). Like most ambystomatids (Bishop oviposition activity when we observed thou­ 1947), the six subspecies of tiger salamanders sands of A. t. nebulosum eggs in Horseshoe in western USA (californiense, mavortium, Lake (34°22'53/N, 111°14'38/W) on the nebulosum, diaboli, melanostictum, stebbinsi Mogollon Rim in central Arizona. We visited [Collins et al. 1980, Jones et al. 1988]) all the lake again on 6 August 1990, before· sum­ breed in late winter, spring, or even early mer rains began. It was completely dry at this summer at high elevations. Differences time, and we photographed thousands of des­ between populations in the timing of the pri­ iccated salamander larvae on the lake bed. We mary breeding period correspond to differ­ presumed this represented elimination of the ences in availability ofwater in their breeding spring 1990 cohort. We sampled this lake again

IDepartment ofZoology, Arizona State University, Tempe, Arizona 85287-1501.

376 1994] NOTES 311 on 23 September 1990 after it had refilled fol­ larvae several centimeters larger than recently lowing summer rains and noted several hun~ hatched salamanders in this tank, suggesting dred small A. t. nebulosum larvae. Live mea­ salamanders produced a second cohort in this surements were taken from a sample oflarvae tank following rains in summer 1990. on 29 September and compared to measure­ Cottonwood Tank (34°08'43"N, 110 0 09'06''W) ments taken from projections of two close-up dried in June 199.2. Although we did not visit slides of some of the dead spring cohort. We this habitat later in 1992, presence of a large also visited other ponds in this region that had larva in March 1993 (before the spring cohort dried and refilled to gauge the extent and suc" hatched) suggests opportunistic breeding in cess ofthis breeding tactic. summer 1992. Larvae caught in Horseshoe Lake in Sep­ Late~season breeding may, however, fail. tember were either a second cohort that Johnnie Tank (34°10'06"N, 110 0 04'02''W) in hatched after summer rains, or animals from the the White Mountains is at the same elevation spring cohort that survived the lake drying, as 13 ponds in the surrounding 180 km2 . presumably by burrowing. Metamorphosed Breeding at all other ponds in this area was tiger salamanders burrow in soft lake mud completed by late March in 1992 and 1993. (Webb 1969), but this has not been reported Johnnie Tank dried in early spring 1992 and for larvae or branchiate adults. then refilled in late May after early monsoon Dead larvae photographed in August aver~ rains. Hatchlings produced by opportunistic aged 51.4 mm total length (SE = 1.20 mm, N breeding following this refilling were all killed == 35), and larvae collected in September when the tank dried again in late June. averaged 35.8 mm total length (SE = 0.78 Because the May oviposition does not overlap mm, N = 96). Ifanimals collected in Septem­ the usual breeding season in this area, we con" ber survived drying by burrowing, they should sider this to be opportunistic breeding. have been at least as large as the dead animals Other evidence suggests that summer breed­ observed in August. Animals in September ing is exhibited regularly in this subspecies. were, however, significantly smaller than Metamorphosed females with yolked follicles those photographed in August (t :::::: 10.53, P were recorded in Arizona by Durham (1956) <.0001). This is a conservative test since ani~ on 18 July on the Kaibab Plateau and by J. mals photographed in August were dried and Collins (unpublished observation) on 11 July therefore smaller than at death. . 1980 in the White Mountains. We do not consider summer breeding to be Tanner et al. (1971) reported sizes of A. t. the primary breeding event for this popula­ nebulosum larvae in Salamander Lake, a per­ tion. Because we observed several thousand manent lake in Utah. While following growth eggs in March, we conclude that there was a of larvae throughout July and August, they normal spring breeding. Typical larval densi­ recorded a small size class beginning in late ties (estimated by drop-box and seining July, which they interpreted as evidence of a through a measured volume of water) in this second breeding. A pattern of spring and fall part of Arizona in June are 2-75 salamanders breeding in permanent ponds with unpre~ m--3 (Pfennig et al' 1991). The low density of dictability in seasonal rainfall is also reported larvae we estimated in September, 0.3 sala­ for Triturus alpestris apuanus in Italy (Andreone manders m-3, supports the conclusion that and Dore 1992). Despite the fact that we reg­ there was a second reproductive episode in ularly visit a few dozen continuously filled which a small number of animals bred oppor­ ponds in Arizona, we have never observed a tunistically in Horseshoe Lake during 1990. second breeding in one. Our observations, Adult salamanders breeding in Horseshoe consistent with the hypothesis outlined below, Lake in late summer 1990 were taking advan­ emphasize that the natural history we are tage ofa newly filled habitat. At least some lar­ describing for A. t. nebulosum in Arizona dif­ vae in this second 1990 cohort overwintered fers importantly from that reported in Utah. successfully (M. Loeb personal observation, 4 Webb (1969) argued that an irregular May 1991). In 1990, Charco Tank (34°07'50"N, breeding pattern is among the traits that adapt 11OC07'32"W) in the White Mountains dried A. t. mavortium for life in the Chihuahuan following drought conditions that also dried Desert in southern New Mexico; i.e., this sub­ Horseshoe Lake. In spring 1991 we collected species reproduces whenever water fills the 378 GREAT BASIN NATURALIST [Volume 54 ephemeral ponds in which it commonly LITERATURE CITED breeds. A. t. mavortium breeds every year in ponds that fill in winter or spring, but may ANDREONE, E, AND B. DORE. 1992. Adaptation of the reproductive cycle in Triturus alpestris apuanus to also breed after summer rains. Om data sug­ an unpredictable habitat. Amphibia-Reptilia 13: gest A. t. nebulosum in Arizona has evolved a 251-261. similar life history tactic. In general, breeding BISHOp, S. C. 1947. Handbook of salamanders. Comstock occurs follOwing snowmelt at high elevations, Publishing Company, Ithaca, NewYork. 555 pp. but there are some conditions under which BURGER, W. L. 1950. Novel aspects ofthe life histories of two ambystomas. Journal ofthe Tennessee Academy individuals will breed opportunistically fol­ ofScience 25: 252--257. lowing summer rains. COLLINS, J. P. 1981. Distribution, habitats, and life history An amphibian larva from a spring cohort is variation in the tiger salamander, Ambystoma not guaranteed sufficient time to complete tigrinum, in east-central and southeast Arizona. development to metamorphosis in any aquatic Copeia 1981: 666-675. COLLINS, J. P., AND J. CHEEK. 1983. Effect of food and habitat that can dry unpredictably. This is an density on development of typical and cannibalistic explanation for iteroparity in most ambystom­ salamander larvae in Ambystoma tigrinum nebulo­ atids using temporary or "most nearly perma­ sum. American Zoologist 23: 77-84. nent" ponds (Wilbur 1977). It might be adap~ COLLINS, J. P., AND J. R. HOLOMUZKI. 1984. Intraspecific tive, however, for adults to take advantage of variation in diet within and between trophic morphs ponds whenever they refill. In contrast, late­ in larval tiger salamanders (Ambystoma tigrinum nebulosum). Canadian Journal of Zoology 62: season breeding in permanent aquatic habitats 168-174. is generally not advantageous since these COLLINS, J. P., AND H. M. WILBUR. 1979. Breeding habits habitats may harbor older larvae or fish that and habitats of the amphibians of the Edwin S. can prey on embryos and hatchlings (Burger George Reserve, Michigan, with notes on the local 1950, Reese 1968, Webb and Roueche 1971; distribution of fishes. Occasional Papers of the Museum of Zoology, University of Michigan 686: but see Dodson and Dodson 1971 and Collins 1-34. and Holomuzki 1984). Breeding after a habitat COLLINS, J. P., J. B. MITTON, AND B. A. PIERCE. 1980. Amby­ refills would be advantageous as the drying stoma tigrinum: a multi-species conglomerate? Copeia would eliminate fish or older larvae. A second 1980: 938-941. advantage ofbreeding opportunistically arises DODSON, S. 1., AND V. E. DODSON. 1971. The diet of Ambystoma tigrinum larvae from western Colorado. since metamorphosis is only possible after a Copeia 1971: 614-624. minimum size is achieved (Wilbur and Collins DURHAM, E E. 1956. Amphibians and reptiles of the 1973), and in temporary ponds like those we North Rim, Grand Canyon, Arizona. Herpetologica report here, this size may not be attained 12: 220-224. before the pond dries; however, larvae from HEYER, W. R., R. W. MCDIARMID, AND D. L. WEIGMANN. 1975. Tadpoles, predation and pond habitats in the summer cohorts that successfully overwinter tropics. Biotropica 1(2): 100-111. will have a growth advantage over larvae from HOLOMUZKI, J. R. 1986. Effect of microhabitat on fitness spring clutches. components of larval tiger salamanders. Oecologia A closer examination of the life histories of 71: 142--148. HOUGfITON, E E. 1976. Climates ofthe States. New Mexico. other subspecies oftiger salamanders found in Climatography ofthe United States, 60-15: 677-681. the arid and semiarid western USA might Gale Research Company, Detroit, Michigan. reveal that regular breeding in late winter and JOLY, J. M. J. 1971. Les cycles sexuels de Salamandra (L.). spring, with opportunistic breeding in sum­ 1. Cycle sexuel des males. Annales des sciences mer, also occurs in these subspecies. naturelles. Zoologie et Biologie Animale 13: 451-504. JONES, T. R., AND J. P. COLLINS. 1992. Analysis ofa hybrid ACKNOWLEDGMENTS zone between subspecies of the tiger salamander (Ambystoma tigrinum) in central New Mexico. We thank the White Mountain Apache Journal ofEvolutionary Biology 5: 375-402. Indian tribe for their cooperation and permis­ JONES, T. R., J. P. COLLINS, T. D. KOCHER, AND J. B. MITTON. 1988. Systematic status and distribution of Amby­ sion to work on their reservation (permits stoma tigrinum stebbinsi Lowe. Copeia 1988: 621-635. #90-04, #91~01, and #92~03). The Arizona LoITs, B. 1974. Reproduction. Pages 107-218 in B. Lofts, Game and Fish Department provided collect­ ed., Physiology of the amphibia. Academic Press, ing permit #CLNSOOOO1l8. This work was New York. PASE, C. P., AND D. E. BROWN. 1982. Rocky Mountain supported by NSF grant #BSR-8919901 to (Petran) and Madrean montane conifer forest. IPC. Desert Plants 4(1-4): 43-48. 1994] NOTES 379

PFENNIG, D. W, M. L. G. LOEB, AND J. P. COLLINS. 1991. the Australian water"holding frog, Cyclorana platy­ Pathogens as a factor limiting the spread ofcannibal­ cephalus. Herpetologica 35: 370-374. ism in tiger salamanders. Oecologia 88: 161-166. WEBB, R. G. 1969. Survival adaptations of tiger salaman­ REESE, R. W 1968. The taxonomy and ecology ofthe tiger ders (Ambystoma tigrinum) in the Chihuahuan salamander (Ambystoma tigrinum) of Colorado. Desert. Pages 143-147 in C. C. Hoff and M. L. Unpublished doctoral dissertation, University of Riedesel, eds., Physiological systems in semiarid Colorado, Boulder. environments. University of New Mexico Press, SALTHE, S. N., AND J. S. MECHAM. 1974. Reproductive and Albuquerque. courtship patterns. Pages 310-521 in B. Lofts, ed., WEBB R. G., AND W ROUECHE. 1971. Life history aspects Physiology of the amphibia. Academic Press, New ofthe tiger salamander (Ambystoma tigrinum mavor­ York. tium) in the Chihuahuan Desert. Great Basin SEXTON, O. J., AND J. R. BIZER. 1978. Life history patterns Naturalist 31: 193-212. of Ambystoma tigrinum in montane Colorado. WILBUR, H. M. 1977. Propagule size, number, and disper­ American Midland Naturalist 99": 101-118. sion pattern in Ambystoma and Asclepias. American STEARNS, S. C. 1992. The evolution of life histories. Naturalist Ill: 43-68. Oxford University Press, Oxford. 249 pp. WILBUR, H. M., AND J. P. COLLINS. 1973. Ecological STEBBINS, R. C. 1985. Western reptilj:ls and amphibians. aspects of amphibian metamorphosis. Science 182: Houghton Mifllin Company, Boston. 336 pp. 1305-1314. TANNER, W W, D. L. FISHER, AND T. J. WILLIS. 1971. WRIGHT, A. H., AND A. A. WRIGHT. 1949. Handbook of Notes on the life history ofAmvystoma tigrinum neb­ frogs and toads of the United States and Canada. ulosum Hallowell in Utah. Gre:at Basin Naturalist 31: Cornell University Press, New York. 640 pp. 213-222..., VAN BEURJ)EN, E. K. 1979. Gamete··development in rela" Received 19January 1993 tion to season, moisture, energy reserve, and size in Accepted 31 March 1994 Great Basin Naturalist 54(4), © 1994, pp. 380-383

VEGETATION RECOVERY FOLLOWING FIRE IN AN OAKBRUSH VEGETATION MOSAIC

Stephen E Poreda1,2 and Leroy H. Wullstein1,3

Ke.y words: fire, oak, secondary succession, soil erosion, shrub, grass, ecosystem.

Fire plays a role in maintaining ecosystem tata, Purshia tridentata, Chrysothamnus vis­ diversity, improving forage production, cidiflorus, Bromus tectorum, and Agropyron enhancing wildlife habitat, and recycling spicatum. nutrients in the soil (Wright and Bailey 1982). Climate of the Heber Valley area is charac­ Postfire succession in the Gambel oak (Quercus teristically continental. Annual precipitation at gambelii) type, however, has received less the Heber City weather station averages 39 attention in the literature than most major cm. Mean annual snowfall is 175 cm. Annual vegetation types. The most extensive work on average daily maximum and minimum tem­ secondary succession in the Gambel oak type peratures are 16.1°C and -2.6°C, respectively. was done by McKell (1950). The frost-free period is typically 70-80 days In August of 1990 an intense wildfire (USDA 1976). burned nearly 3000 ac ofoak-dominated vege­ One study site was selected within the burn, tation in the vicinity ofWasatch Mountain State and a similar nearby unburned area was chosen Park near Midway, Utah. This note reports on to provide comparison. Sites were selected for the first-year vegetation recovery in a vegeta­ similarity of elevation, aspect, slope, and soils. tion mosaic of oakbrush and sagebrush-grass The burned site (T3S R4E S33 SEl/4) within communities during the first year following an elevation range of 1711-1832 m consists of that fire. two opposing slopes, one with a generally Individual study sites are located in Heber east-faCing aspect, the other generally west­ Valley near the town ofMidway, Utah. All are in facing. Variability in topography allowedfor the lower foothill zone of the central Wasatch sampling across a full range (0-360°) ofaspect. Mountains. Gambel oak dominates the hillside The unburned site consists of a roughly cir­ vegetation ofHeber Valley and grows in a dis­ cular sampling transect centered on Memorial continuous belt extending from approximately Hill (T3S R4E S35 NEl/4). The sampling ele­ 1500 to 2600 m elevation. Study sites lie with~ vation ranges from 1740 to 1771 m. Slopes of in the ecotone near the lower margin of the both sites range from 20 to' 60%. Soils consist scrub-oak belt and the upper margin of the of a complex of Hennefer silt loam and foothill zone. The ecotone comprises a vegeta­ Hennefer cobbly silt loam (Pachic argixerolls; tion mosaic of open spaces and oak-clone USDA 1976). Runoffis rapid on Hennefer soils thickets. Major shrub or tree species associat­ and erosion hazard is considered high. ed with Gambel oak include Prunus virgini" Burned and unburned sites were identified ana, Acer grandidentatum, Symphoricarpos for intensive sampling and quadrat analysis. oreophilus, and Amelanchier alnifolia. Two supplemental sites ofsimilar slope, eleva­ Interspersed among oak-clone thickets are tion, and topography were selected for recon­ open spaces containing vegetation characteris­ naissance survey and identified as area C (T3S tic ofboth the mountain shrub community and R4E S21 NE1/4) and area D (T3S R4E S23 the foothill zone. The interspaces characteris" NEl/4). Comprehensive species checklists tically support populations ofArtemisia triden- were compiled for all sites to establish whether

1Department ofGeography, University ofUtah, Salt Lake City, Utah 84112. 2Present address: 839 E. Garfield Ave., Salt Lake City, Utah 84105. 3Address correspondence to this author.

380 1994] NOTES 381 the intensive study sites were representative The Mann~Whitney test was used for signifi~ ofvegetation ofthe vicinity as a whole (Poreda cance testing of mean differences. Percent of 1992). total cover (%TC) ofeach species was expressed Sampling was done in mid~June, early as a percentage of the summed maximum August, and late September during the 1991 cover of all species. Species identification fol­ growing season. For the reconnaissance sur­ lows Arnow et al. (1980). veys, species were recorded as encountered Total vegetative cover was substantially while the surveyor walked arbitrarily selected reduced for both communities on the bum transects within each of the four study areas. even after one full year compared to that of This permitted observation of additional the unburned site (Table 1). Changes in bare species not found in the quadrats. soil June through September on both unburned Quadrat sampling was conducted along sites are not statistically significant (P > .28), transect lines established on both the burned nor is the change between June and September and unburned sites. Quadrats (1.0 m2) were on burned shrub-grass sites (P > .4). On marked at 30~m intervals along each transect burned oak sites, however, the trend ofsignifi­ line. The burned site contained 171 quadrats, cantly (P = .0004) decreasing bare soil can be the unburned site 39. attributed to the dramatic increase in vegeta­ Cover for each species, total vegetative tive cover. cover, bare soil, rock, and litter were estimat­ Species count (Table 2) on burned shrub­ ed within each quadrat using a procedure grass quadrats (73) was 1.87 times the number slightly modified from Daubenmire (1959). on unburned quadrats (39). Total number of The modification consisted of adding one species on burned oak sites (61) was 1.33 extra cover class with limits of 0-1%. This times greater than the number on unburned modification provided a more accurate esti~ sites (46). The data strongly indicate that mate ofcover for small or subordinate species shrub-grass sites are richer in species than oak (Davis and Harper 1989). Plant densities were sites, and that this relationship holds even based on counts of individuals (by species) after fire. Moreover, species diversity is higher rooted within the 1.0-m2 quadrats. Species on both communities following fire. frequencies were the percentage of quadrats Many species declined in both frequency in which a species occurred. and cover following fire (Tables 3, 4). Others Values for cover-class and density were increased in both frequency and cover follow­ recorded for each species for each quadrat ing fire. A few showed little change. along with sampling date and community On shrub-grass sites ·species showing great~ type. Species mean percent cover (%C) was est decrease include Agoseris glauca, Agro­ computed separately for oak and shrub~grass pyron spicatum, Artemisia tridentata, Crepis communities (including aggregated communi­ acuminata, Lathyrus pauciflorus, and Loma­ ties) on both the burned and unburned sites. tium triternatum. Species increasing after

TABLE 1. Seasonal areal coverage (%c)a of vegetation, litter, bare soil, and exposed rock for burned and unburned sitesb. Shrub-grass Oak June Aug Sept June Aug Sept BURNED Vegetation 16.79 12.87 28.14 16.30 41.59 49.50 Litter 6.76 21.67 22.86 7.22 9.21 11.57 Bare soil 35.72 37.38 37.98 42.88 33.63 32.44 Rock 18.09 18.74 19.41 10.28 10.25 10.15 UNBURNED Vegetation 46.09 31.59 49.50 73.94 65.06 73.18 Litter 37.30 53.95 60.27 88.52 90.59 92.65 Bare soil 7.95 9.05 8.93 6.53 0.79 0.65 Rock 18.80 18.80 19.34 3.09 3.09 2.94

a%c expressed as mean cover based on m2 quad~J.ts. bn = 171 (burned); n= 39 (uuburned). 382 GREAT BASIN NATURALIST [Volume 54

TABLE 2. Cumulative number ofspeciesa observed in sampled quadrats. Shrub-grass Oak Aggregated # % # % # % BURNED Total 73 100 61 100 80 100 Annual 24 33 20 33 26 32 Forb 33 45 ~8 46 36 45 Grass 9 12 4 6 9 11 Shrub 7 10 9 15 9 11 UNBURNED Total 39 100 ~9 100 46 100 Annual 10 26 9 31 14 30 Forb 16 41 9 31 17 37 Grass 8 20 6 21 8 17 Shrub 5 13 5 17 7 15

"June through September 1991

TABLE 3. Frequency and cover for major species on shrub-grass sites. Unburned Burned Species Freq. %C %TC Freq. %C %TC Agoseris glauca 9.10 .05 .07 .00 .00 .00 Agropyron spicatum 63.60 10.31 16.04 9.80 .80 2.11 Amelanchier alnifolia .00 .00 .00 1.60 .05 .13 Artemisia tridentata 21.30 4.59 7.15 .00 .00 .00 Aster chilensis 9.10 .16 .25 3.30 .50 1.30 Bromus tectorum 68.20 6.84 10.64 70.50 6.51 17.14 Chenopodium album .00 .00 .00 24.60 1.41 3.72 C. leptophyllum .00 .00 .00 19.70 .50 1.32 Collinsia parviflora 18.20 .21 .32 26.20 .34 .88 Crepis acuminatum 54.50 2.16 3.36 4.90 .07 .17 Galium aparine 4.50 .02 .04 24.60 .52 1.38 Lathyrus pauciflorus 13.60 .84 1.31 3.30 .05 .13 Lomatium tritematum 27.30 .91 1.41 4.90 .02 .06 Machaeranthera canescens .00 .00 .00 9.80 .40 1.06 Poa pratensis 27.30 6.14 9.56 18.00 ~.39 6.28 Polygonum ramosissimum 4.50 .02 .04 19.70 .26 .69 Prunus virginiana .00 .00 .00 3.30 .30 .18 Quercus gamhelii 22.70 5.49 8.55 14.80 1.59 4.17 Solidago sparsiflora .00 .00 .00 4.90 .54 1.42 Symphoricarpos oreophilus .00 .00 .00 4.90 .11 .28 Verbascum thapsus .00 .00 .00 19.70 .90 2.37 Viguiera multiflora .00 .00 .00 36.10 2.53 6.67 burning include Chenopodium album, C. lepto­ suggesting only a minor effect in the first year phyllum, Collinsia parviflora, Galium aparine, following fire; relative importance of this Machaeranthera canescens, Polygonum ramosis­ species was, however, enhanced due to the simum, Solidago sparsiflora, Verbascum thap~ decline of most other species. Studies by sus, and Viguiera multiflora. Young and Evans (1978) suggest a potential Some species, such as Aster chilensis, may be explosive increase of B. tectorum in the sec­ somewhat less typical in that frequency was ond and third years after fire as the species lower on burned shrub-grass sites, yet cover rapidly colonizes space made available by fire. was actually greater than on unburned sites On oak~dominatedsites there was a similar (Table 4); suggesting a response of increased or greater reduction of those same species size and vigor ofthe surviving individuals. exhibiting a lowered frequency on shrub~grass Relative to the unburned, Bromus tectorum sites. In addition, Aster chilensis and Bromus showed httle difference in frequency or cover, tectorum (species showing substantial survival 1994] NOTES 383

TABLE 4. Frequency and cover for major species on oak-dominated sites. Unburned Burned Species Freq. %C %TC Freq. %C %TC Agoseris glauca 11.80 1.06 1.10 .90 <.01 .01 Agropyron spicatum 29.40 .59 .61 .00 .00 .00 Amelanchier alnifolia 23.50 1.09 1.13 8.10 .95 1.66 Artemisia tridentata 17.60 1.09 1.13 .00 .00 .00 Aster chilensis 5.90 .03 .03 .00 .00 .00 Bromus tectorum 23.50 4.17 4.33 18.90 .16 .28 Chenopodium album .00 .00 .00 24.30 1.24 2.15 C. leptophyllum 5.90 .03 .03 19.80 .39 .67 Collinsia parviflora 29.40 .15 .15 27.90 .32 .56 Crepis acuminata 47.10 1.38 1.44 3.60 .02 .03 Galium aparine 11.80 .21 .21 24.30 .69 1.20 Lathyrus pauciflorus 58.80 2.59 2.68 7.20 .30 .52 Lomatium tritematum 35.30 .77 .79 1.80 .01 .02 Machaeranthera canescens .00 .00 .00 .90 .03 .05 Poa pratensis 52.90 16.81 17.44 17.10 1.23 2.14 Polygonum ramosissimum .00 .00 .00 6.30 .03 .05 Prunus virginiana 29.40 8.38 8.69 10.80 .84 1.46 Quercus gambelii 100.00 51.12 53.03 92.80 38.36 66.68 Solidago sparsiflora .00 .00 .00 2.70 .19 .33 Verbascum thapsus .00 .00 .00 19.80 1.12 1.95 Viguiera multiflora .00 .00 .00 24.30 .86 1.49 on shrub-grass sites) were virtually eliminated DAVIS, J. N., AND K. T. HARPER. 1989. Weedy annuals and from oabdominated sites. Higher burn tem~ establishment of seeded species on a chained juniper-pinyon woodland in central Utah. veg~ peratures associated with oak-dominated Unpublished manuscript. etation were likely more damaging to these MCKELL, C. M. 1950. A study ofplant succession in the species. Shrubs most common to oak sites, oak brush (Quercus gambelii) zone after fire. Prunus virginiana and Amelanchier alnifolia, Unpublished master's thesis, University of Utah, both had decreased frequencies following fire. Salt Lake City. POREDA, S. E 1992. Vegetation recovery and dynamics fol­ Amelanchier alnifolia, however, exhibited lowing the Wasatch Mountain fire (1992), Midway, more vigorous resprouting. Although frequen­ Utah. Unpublished master's thesis, University of cy was lower on burned oak sites, cover ofA. Utah, Salt Lake City. 162 pp. alnifolia one year after the burn was only USDA. 1976. Soil survey of Heber Valley area, Utah. National Cooperative Soil Survey. slightly less on the burn site than on the WRIGHT, H. A., AND A. w: BAILEY. 1982. Fire ecology. John unburned sites. Wiley and Sons, New York. YOUNG, J. A., AND R. A. EVANS. 1918. Population dynamics LITERATURE CITED after wildfires in sagebrush grasslands. Journal of Range Management 31: 283-289. ARNow, L., B. ALBEE, AND A. WYCKOFE 1980. Flora ofthe central Wasatch Front, Utah. University ofUtah, Salt Received 2 July 1993 Lake City. Accepted19April 1994 DAUBENMIRE, R. 1959. A canopy-coverage method of veg~ etational analysis. Northwest Science 33: 43-64.

information FOR AUTHORS the great basin naturalist welcomes previously VOUCHER SPECIMENS authors are encouraged to unpublished manuscripts pertaining to the biologi- designate properly prepare label and deposit cal natural history of western north america pref- high quality voucher specimens and cultures docu- erence will be given to concise manuscripts of up to menting their research in an established permanent 12000 words simple species lists are discouraged collection and to cite the repository in publication SUBMIT manuscripts to richard W baumann references IN THE TEXT are cited by author and editor great basin naturalist 290 MLBM PO box date ege g martin 1989 or martin 1989 multiple 20200 brigham young university provo UT citations should be separated by commas and listed 84602020084602 0200 A cover letter accompanying the man- in chronological order use et al after name of uscript must include phone numbers of the author first author for citations having more than two submitting the manuscript and FAX number and authors emailE mail address when applicable the letter must acknowledgments under a centered main also provide information describing the extent to heading include special publication numbers when which data text or illustrations have been used in appropriate other papers or books that are published in press literature CITED also under a centered mainmaln submitted or soon to be submitted elsewhere heading lists references alphabetically in the fol- authors should adhere to the following guidelines lowing formats manuscripts not so prepared may be returned for revision mack G D and L D flake 1980 habitat rela- manuscript preparation consult vol 51 no tiontionshipsships of waterfowl broods on south dakota 2 of this journal for specific instructions on style stock ponds journal of wildlife management and format these instructions guidelinesclideGLIDE LINES FOR 4469544 695ggs 700 manuscripts SUBMITTED TO THE GREAT BASIN sousa W P 1985 disturbance and patch dynamics naturalist supply greater detail than those pre- on rocky intertidal shores pages 101 124 in sented here in condensed form the guidelines are S T A pickett and P S white eds the ecolo- printed at the back of the issue additional copies gy of natural disturbance and patch dynamics are available from the editor upon request also academic press new york check the most recent issue of the great basin nat- coulson R N and J A witter 1984 forest ento- uralist for changes and refer to the CBE style man- mology ecology and management john wiley ual ath5th edition council of biology editors one and sons inc new york 669 appp illinois center suite 200 illiiilii111 east wacker drive chicago IL 60601429860601 4298 USA PHONE 312 616- TABLES are double spaced on separate sheets and 0800 FAX 3126160226312616312 6160226616gig 02260996 24 designed to fit the width of either a single column TYPE AND DOUBLE SPACE all materials including or a page use lowercase letters to indicate foot- literature cited table headings and figure legends notes avoid hyphenated words at the righthandright hand margins photocopies OF FIGURES are submitted initially underline words to be printed in italics use stan- with the manuscript editors may suggest changes dard bond 22x28 cm leaving 252 5 cm margins on lettering on figures should be large enough to all sides withstand reduction to one or two column width SUBMIT 3 COPIES of the manuscript and the origi- originals must be no larger than 2222x28X 28 cm nal on a 5255235 25 or 353 5 inch disk utilizing WordwordperfectPerfect NOTES if the manuscript would be more appro- 424.24 2 or above number all pages and assemble each priate as a short communication or note follow the copy separately title page abstract and key words above instructions but do not include an abstract text acknowledgments literature cited appen- A CHARGE of 50 per page is made for articles dices tables figure legends figures published the rate for individual subscribers will TITLE PAGE includes an informative title no longer be 35 per page however manuscripts with com- than 15 words names and addresses of authors a plex tables andor numerous half tones will be running head of fewer than 40 letters and spaces assessed an additional charge reprints may be pur- footnotes to indicate change of address and author chased at the time of publication an order form is to whom coicoleolrespondencecorrespondence should be addressed if sent with the proofs other than the first author FINAL CHECK ABSTRACT states the purpose methods results cover letter explaining any duplication of and conclusions of the research it is followed by information and providing phone numbers 6 12 key words listed in order of decreasing FAX number and emailE mail address importance to be used for indexing 3 copies of the manuscript and WordwordperfectPerfect TEXT has centered mainmaln headings printed in all diskette capital letters second level headings are centered conformity with instructions in upper and lowercase letters third level head- photocopies of illustrations ings begin paragraphs issn0017ISSN 001736140017 3614 GREAT BASIN naturalist voivol 54 no 40019944 october 1994 CONTENTS articles mycorrhizal colonization hyphal lengths and soil moisture associated with two artemisia tridentata subspecies james D trent tony J svejcar and robert R blank 291 soil and vegetation development in an abandoned sheep corral on degraded sub- alpine rangeland james 0 klemmedson and arthur R tiedemann 301 Geogeostatisticalstatistical analysis of resource islands under artemisia dentatatritridentatatridentate in the shrub steppe jonathan J halvorson harvey bolton jr jeffrey L smith and richard E rossi 313 habitat preference and diurnal use among greater sandhill cranes donald E mcivormelvor and michael R conover 329 selenium geochemical relationships of some northern nevada soils stephen poole glenn gross and robert potts 335 status and distribution of the laridae in wyoming through 1986 scott L findholt 342 seed production in gentiana newberryinewberryi gentianaceae myra E bamesbarnes and richard W rust 351 use of a secondary nest in great basin desert thatch ants formica obscuripes forel james D mclvermcivermelvermeiver and trygve steen 359 spawning chronology and larval emergence of june sucker chasmistes horushofus

1 I timothy modde and neal muirhead 366 comparison of reproductive timing to snow conditions in wild onions and white crowned sparrows at high altitude martin L morton 371

notes opportunistic breeding after summer rains by arizona tiger salamanders linda J allison paul E brunkow and james PR collins 376 vegetation recovery following fire in an oakbrushoakbrush vegetation mosaic stephen FE poreda and leroy H wullstein 380

index to volume 54 385