Western North American Naturalist

Volume 62 Number 2 Article 22

5-2-2002

Full Issue, Vol. 62 No. 2

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

Recommended Citation (2002) "Full Issue, Vol. 62 No. 2," Western North American Naturalist: Vol. 62 : No. 2 , Article 22. Available at: https://scholarsarchive.byu.edu/wnan/vol62/iss2/22

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 Western North American Naturalist by an authorized editor of BYU ScholarsArchive. For more information, please contact [email protected], [email protected]. Western North American Naturalist 62(2), © 2002, pp. 129–140

LATE PLEISTOCENE MOLLUSKS FROM THE SOUTHERN BLACK HILLS, SOUTH DAKOTA

Christopher N. Jass1,3, Jim I. Mead1,2, Amy D. Morrison2, and Larry D. Agenbroad1,2

ABSTRACT.—Pleistocene-age mollusks recovered from the Nelson-Wittenberg Site and Mammoth Site provide infor- mation on paleoenvironments of the southern Black Hills, South Dakota. New mollusk records for the Mammoth Site include Vertigo modesta, Catinella sp., and Gyraulus parvus. The presence of V. modesta, Columella columella alticola, and Pupilla muscorum at the Nelson-Wittenberg Site suggests cooler than modern conditions in the Black Hills during the Pleistocene. Although the majority of identified taxa are consistent with previous interpretations of an arid, shrub- steppe environment for the Black Hills during the Wisconsin Glaciation, V. modesta and C. c. alticola suggest the presence of at least some isolated, local mesic areas.

Key words: Black Hills, mollusks, Late Pleistocene, paleoenvironments, Mammoth Site, Nelson-Wittenberg Site.

The Black Hills of South Dakota, aptly and Holocene deposits. At present, only 2 Plei- described as “a forested island in a grassland stocene paleontological localities, Salamander sea” (Froiland 1990:1), occupy a unique phys- Cave (Mead et al. 1996) and the Mammoth iographic position within the northern Great Site (Agenbroad et al. 1990), are published Plains (Fig. 1). The Black Hills are situated from the Black Hills proper. Most paleoenvi- approximately 610–2100 m higher than the ronmental data for the northern Great Plains surrounding northern Great Plains, with the during the Late Pleistocene were derived from local vegetation containing representatives research at localities in adjacent geographic from all adjoining communities (Weedon areas outside the Black Hills. Using data from and Wolken 1990). Reconstructions of paleo- areas peripheral to the Black Hills, some re- environments for the northern Great Plains searchers have inferred that portions of the rarely address the Black Hills. Due to the var- Great Plains, including the Black Hills, were ied topographic and biotic nature of the Black covered with spruce (Picea) forest during the Hills, paleoenvironmental reconstructions based Late Pleistocene (see Mead et al. 1994 for dis- on data recovered from the surrounding north- cussion). Other interpretations, based on faunal ern Great Plains may not provide an accurate evidence, suggest steppe environments for por- picture of conditions within the Black Hills tions of the northern Great Plains during the during the Pleistocene (1.77 Ma [meg annum; Pleistocene (Taylor 1960, 1965, Voorhies and million years ago] to approximately 10,000 yr Corner 1985). B.P.; years before present). Although the Black Molluscan and vertebrate data from the Hills have no direct evidence of glaciation Late Pleistocene Lange-Ferguson Site (10,670 (Lemke et al. 1965), the close proximity of the yr B.P.), east of the Black Hills (Fig. 1), suggest Black Hills to the Laurentide ice cap (~240 more mesic conditions than modern, with km to the east) during the Late Pleistocene brush and/or woodlands occurring at the site surely influenced the composition of local biotic (Leonard 1982, Martin 1987). Whether Lange- communities. Ferguson fossils represent an overall “mesic” Data on Pleistocene environments in the regional trend is unknown; drier, grassland en- Black Hills are not abundant. Due to a lack of vironments may have occurred in areas away natural lakes and ponds, the potential for paly- from Lange-Ferguson (Martin 1987). nological study for the reconstruction of past To the west of the Black Hills, mammalian environments is limited. Fredlund (1996) pro- faunas are the primary source of paleoenviron- vides geomorphic evidence of latest Pleistocene mental reconstructions (Mead et al. 1994).

1Quaternary Sciences Program, Box 5644, Northern Arizona University, Flagstaff, AZ 86011-5644. 2Department of Geology, Box 4099, Northern Arizona University, Flagstaff, AZ 86011. 3Corresponding author. Present address: Department of Geological Sciences, Mail Code C1140, University of Texas at Austin, Austin, TX 78712.

129 130 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 1. Map depicting locations (triangles) of Graveyard Cave (GC), Mammoth Site of Hot Springs (HSMS), Nelson- Wittenberg Locality (NW), Salamander Cave (SC), and PaleoIndian sites (asterisks) containing mollusks, including Agate-Basin (AB) and Lange-Ferguson (LF).

Alpine tundralike conditions and more equable Basin are indicative of more mesic conditions climates may have dominated much of Wyo- than today, at least in the immediate area of ming at approximately 22,000–18,000 yr B.P. the locality (Evanoff 1982). (Walker 1987). At that time boreal forest envi- Reported Pleistocene localities from the ronments were more widespread than today Black Hills include Salamander Cave (Fig. 1), (Walker 1987). Immediately west of the Black which has a minimum age of 252,000 yr B.P. Hills, faunal remains from the Late Pleis- (Mead et al. 1996), and the Mammoth Site of tocene–Early Holocene Agate Basin Site (Fig. Hot Springs (Agenbroad et al. 1990, Agenbroad 1) indicate a steppe savanna for the terminal and Mead 1994). Using faunal and limited Pleistocene (Walker 1982). Mollusks from Agate pollen data from the Mammoth Site, several 2002] PLEISTOCENE MOLLUSKS OF THE BLACK HILLS 131 authors (Czaplewski and Mead 1994, Mead et duced additional fossil mollusk taxa which are al. 1994, and others) suggested the presence of reported here (see Table 1). Recovered mol- a cold steppe-grassland in the Black Hills at lusks come from bulk samples and cannot be 26,000 yr B.P. More mesic conditions and geo- strongly correlated to a particular time or morphic stability, relative to modern, existed stratigraphic unit within the depositional in parts of the southern Black Hills during the framework established by Laury (1994). transition from Late Pleistocene to Early Holo- NELSON-WITTENBERG SITE.—The Nelson- cene (14,000 yr B.P. to 9000 yr B.P.; Fredlund Wittenberg Site is a small alluvial deposit in 1996). An interpretation of more mesic than the southern Black Hills, located west of the modern conditions from the Late Pleistocene Mammoth Site (Fig. 1). The structure of this through the Early Holocene is consistent with small deposit shows uniform, yet limited, stra- paleoenvironmental data from a variety of local- tigraphy (Mead personal observation). The ities peripheral to the Black Hills (see sum- locality appears to be a remnant drainage mary in Fredlund 1996). Fossil mollusks from channel that might have originated from a a third Pleistocene locality within the Black spring deposit, but apparently not one con- Hills, the Nelson-Wittenberg Site, provide taining any pooled water. No spring occurs in evidence for the reconstruction of past envi- the area of the locality today. A sample of fine- ronments in the southern Black Hills. Here grained sediments from around a Mammuthus we present identifications of mollusks from tusk and Bison tooth produced a thermolumi- the Nelson-Wittenberg Site as well as addi- nescence date of 37,900 ± 2900 yr B.P. (W2611). tional specimens from the Mammoth Site. We We feel this is an approximate age but verifies use these mollusks to further assess environ- assignment of the deposit to the Late Pleisto- mental conditions in the southern Black Hills cene. If the thermoluminescence date is cor- during the Late Pleistocene. rect, the locality lies well within the most recent glacial phase (Wisconsinan Glaciation) STUDY AREA and dates older than the Mammoth Site. Microfaunal elements discussed herein come Locality Descriptions from salvaged bulk sediment samples (washed Currently, Pleistocene-age fossil sites in the through 1-mm sieve). Small mammals from Black Hills are known only from the southern the locality include Lemmiscus curtatus (sage- Black Hills. brush vole; see Bell and Mead 1998 for identi- MAMMOTH SITE.—The Mammoth Site of fication of characters) and Microtus sp. (vole). Hot Springs is a 26,000-year-old deposit located in the southern Black Hills (Fig. 1). Detailed Modern Environments descriptions of the geology, chronology, and As a result of an updoming event that be- vertebrate paleontology of the Mammoth Site gan during the Laramide Orogeny, the geologic are presented in Agenbroad et al. (1990) and structure of the Black Hills region is charac- Agenbroad and Mead (1994). The Mammoth terized by a “layer cake” or anticlinal pattern Site represents an artesian spring-fed pond in which the oldest rock formations are encir- environment contained within a sinkhole (Agen- cled by younger rock formations as one pro- broad 1994). The sinkhole was likely in-filled gresses outward from the core (Froiland 1990). with sediment over a time period of 175 to The 4 major physiographic regions of the 700 years (Laury 1994). Timing of cessation of Black Hills include the Crystalline Core, Lime- spring discharge into the sinkhole is not known stone Plateau, Red Valley, and Hogback Ridge. (Laury 1994). The Mammoth and Nelson-Wittenberg sites Various aspects of the molluscan and verte- are situated on erosional remnants of the brate fauna from the Mammoth Site are dis- Permo-Triassic sediments that make up the cussed in Agenbroad and Mead (1994). Mead Red Valley. In the southern Black Hills this et al. (1994) describe mollusks recovered from region is currently dominated by grassland, the first 20 years of excavation at the Mam- interspersed with an open ponderosa pine moth Site. From 1990 to 1999, sediments were (Pinus ponderosa) forest. Fredlund (1996) con- washed through 1-mm-mesh screens and were sidered the Red Valley an extension of the stored at the site for future analyses. Sorting of grasslands that surround the Black Hills. these sediments began in July 1998 and pro- Local artesian springs (some thermal) provide 132 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 1. Fossil molluscan faunas from the southern Black Hills. For fossil and subfossil taxa, X = present in this study (see text), 1 = Mead et al. (1994). Extant refers to the presence of fossil species as part of the modern snail fauna of the Black Hills. For extant, TR = this report, 2 = Frest and Johannes (1993). For a complete list of extant Black Hills mol- lusk taxa, see Frest and Johannes (1993). NW = Nelson-Wittenberg, MS = Mammoth Site, GC = Graveyard Cave. Provincial information for terrestrial gastropods is as follows: C = common, widespread, RM = Rocky Mountain, I = Interior, N = Northern, based on discussions in Bequaert and Miller (1973), Frest and Dickson (1986), Frest and Johannes (1993), and Woodman et al. (1996). Taxa NW MS GC Extant Province

SPHAERIIDAE Pisidium castertanum —X——— Pisidium compressum — X, 1 — — — Pisidium obtusale —1——— Pisidium walkeri — X, 1 — — — LYMNAEIDAE Fossaria parva —X——— Fossaria cf. dallia —1——— Fossaria sp. — 1 — — — PHYSIDAE Physella sp. — X, 1 — TR — PLANORBIDAE Gyraulus parvus — 1 —TR— Gyraulus sp.— X —TR— PUPILLIDAE Columella simplex — 1 — 2 I, N Columella c. alticola X — — — N, RM Gastrocopta armifera — — X TR, 2 I Gastrocopta procera — — X TR, 2 I Pupilla muscorum X X, 1 — — N Pupilla hebes-like X — — — RM Pupoides albilabris X—X 2 C Vertigo modesta XX—2 N Vertigo gouldi hannai X——— I VALLONIIDAE Vallonia gracilicosta X X, 1 X 2 RM Vallonia cyclophorella X — X TR, 2 RM SUCCINEIDAE Catinella sp. X X X TR — DISCIDAE Discus whitneyi —— X 2 C Discus sp. X — X — — ZONITIDAE Hawaiia minuscula X — X TR, 2 C Zonitoides arboreus —— X 2 C LIMACIDAE Deroceras laeve X 1 — TR, 2 C the potential for atypical microhabitats in the of climatic changes through the Pleistocene/ southern Black Hills. Riparian areas occur as a Holocene transition (Weedon and Wolken 1990). result of drainage from spring discharge and Today, mean annual temperature for the runoff, although perennial streams are not Black Hills is 45.6°F (7.5°C); average precipi- abundant. tation is 73.6 cm (29 inches) at high elevations, The modern vegetative character through- with lower amounts occurring in the adjacent out the Black Hills is complex. Elements of plains (Froiland 1990). Climates in the Black eastern deciduous forests, Rocky Mountain Hills are highly variable and differ from north coniferous forest, northern coniferous forest, to south. The southern Black Hills are charac- and northern Great Plains grassland all occur terized by less annual precipitation, warmer in the Black Hills (Froiland 1990, Weedon and summer temperatures, and warmer winter Wolken 1990). Many plant species have only re- temperatures than the northern hills (Froiland lictual populations in the Black Hills as a result 1990, Weedon and Wolken 1990). 2002] PLEISTOCENE MOLLUSKS OF THE BLACK HILLS 133

Until recently the modern molluscan fauna Unless otherwise indicated, species names of the Black Hills region was not well known. follow Turgeon et al. (1998). Mollusk identifi- Prior to 1993, four published surveys, primar- cations (in part), descriptions, and anatomical ily reporting specimens from the northern terminology are based on Pilsbry (1946, 1948), Black Hills, provided the only information on Herrington (1962), Clarke (1981), and Burch extant molluscan species of the Black Hills (1962, 1989). Comparisons of modern and fos- area (Over 1915, Henderson 1927, Roscoe 1954, sil specimens from Northern Arizona Univer- Hubricht 1985). The most comprehensive sity, Quaternary Sciences Program, were also report on extant Black Hills land snails is an used for identifications. Institutional abbrevia- unpublished United States Forest Service and tions are as follows: Northern Arizona Univer- United States Department of the Interior sity, Quaternary Sciences Program (NAUQSP); report (Frest and Johannes 1993). To our knowl- Mammoth Site of Hot Springs Laboratory edge, no information concerning locally extant (MSL); and Wind Cave National Park (WICA). aquatic taxa is published. Nelson-Wittenberg mollusks and extant mol- When considering Black Hills molluscan lusks reported here are curated at NAUQSP. fauna as a whole for comparisons of fossil fau- Mammoth Site mollusks are curated at MSL. nas with locally extant taxa, we refer to the Graveyard Cave mollusks are under the aus- complete list of known Black Hills taxa pre- pices of WICA but are curated at the National sented in Frest and Johannes (1993). To sup- Park Service Repository, Northern Arizona University, QSP. plement the data of Frest and Johannes (1993), in Table 1 we include records of taxa collected RESULTS by us from disturbed and undisturbed of the southern Black Hills. Mollusks from Table 1 provides a list of molluscan remains Graveyard Cave, a Late Holocene locality in recovered from the Mammoth Site and the the southern Black Hills (Fig. 1), are also pre- Nelson-Wittenberg Site. At least 13 fossil mol- sented in Table 1. Graveyard Cave mollusks luscan species are now known from the Mam- represent a modern portion of the Black Hills moth Site (Table 1). Vertigo modesta, Catinella malacofauna from outside the Red Valley, in sp., and Fossaria parva are newly reported open ponderosa pine forests of the adjacent taxa from this locality. Twelve species of mol- Limestone Plateau (see Fig. 1). lusks were recovered from the Nelson-Witten- Extant Black Hills mollusks have mixed berg Site (Table 1). For the most part, fossil affinities, with the majority of species associ- mollusks from the Mammoth Site and Nelson- ated with molluscan provinces to the north Wittenberg Site are extant in the southern (Northern Province) and west (Rocky Moun- Black Hills today. Identification of fossil taxa tain Province; Frest and Johannes 1993). The not found in the region today requires discus- Red Valley malacofauna is less diverse than in sion of identification methods; these are pre- other portions of the Black Hills, possibly due sented below. to the composition of the substrate (Frest and Pisidium walkeri Johannes 1993). This model of a depauperate Red Valley may change with further sampling Sterki, 1895 of microhabitats in the southern Black Hills. Specimens of this small clam (MSL 1469, 1471–1473) include both left valves (with well- METHODS preserved C2 and C4) and right valves (with C3). The C2 is short, highly curved, and char- Methods of fossil snail recovery are dis- acteristically bent in the middle, with a cussed, in part, under site descriptions. The rounded tip at the posterior end and tapering use of 1-mm sieves for mollusk retrieval at the to a point at the anterior end. Mammoth Site and Nelson-Wittenberg Site is Columella columella alticola possibly unfortunate, as some smaller taxa (i.e., Vertigo spp.) might not be recovered at (Ingersol, 1875) that mesh size. A 0.7-mm (700-µm) sieve is Columella columella alticola is the only typically more desirable for mollusk recovery North American form of the Holarctic species and is now employed in all of our studies. C. columella. Specimens (NAUQSP 11585) 134 WESTERN NORTH AMERICAN NATURALIST [Volume 62 that can be confidently referred to this species conform to the characters outlined in Pilsbry (1948). In particular, the spire is rounded, weak striae are present, and the last whorl is more expanded than the preceding whorl (see Pilsbry 1948: figs 535–536). The cylindrical form of identified specimens (Fig. 2) is consis- tent with Pilsbry’s (1948) description of C. c. alticola rather than the more tapering form of Columella edentula (Pilsbry 1948). Columella columella alticola is not reported in the afore- mentioned surveys of the modern Black Hills malacofauna. Several other specimens are probably C. c. alticola, but the state of preser- vation prevents an unequivocal identification. Vertigo modesta (Say, 1824) The majority of specimens (MSL 1370; NAUQSP 11568–11574) conform to characters of the species outlined in Pilsbry (1948). The shell is irregularly striate or smooth. A parietal fold is present but minute. The columellar fold is barely present. A lower palatal fold is pre- sent and small, with a smaller labrum fold as either an upper palatal or interpalatal fold. Two specimens from the Nelson-Wittenberg Site have a small, angular lamella present in the aperture. These specimens (NAUQSP 11575–11576) conform to Pilsbry’s (1948) de- scription of V. modesta parietalis and so are referred to that taxon. Vertigo gouldi hannai Pilsbry, 1919, Pilsbry, 1948 Four shells (NAUQSP 11565–11567) con- tain an upper sinulus on the aperture. Stria- Fig. 2. Columella columella alticola (NAUQSP 11585) tions are distinct but finely present. The angu- from the Nelson-Wittenberg Site. lar fold is distinct and small. The parietal fold is strong and distinct from the angular. The columellar fold is large and distinct and sepa- rated from the smaller, subcolumellar fold. characters listed above, would identify these The upper and lower palatal folds are strong specimens as belonging to Vertigo gouldi han- with a callus barely present to wanting. The nai. Some authors vary as to whether this longer, lower palatal fold is slightly inset deeper taxon is a form of V. gouldi or a distinct species. than the position of the upper. Shells are 1.8– Even Pilsbry varies as to its taxonomic place- 1.9 mm in length and 1.1 mm in diameter ment (1948:971 vs. 976). Both Pilsbry and within 4.25–4.50 whorls. Hubricht (1985:11) placed this morph as a Our specimens fit into the V. gouldi-group taxon restricted to Pleistocene-age deposits, as defined by Pilsbry (1948:971), who clearly with which we concur. We also agree with states that select members may or may not Pilsbry (1948:976) that “V. g. hannai is very contain an angular fold. Our specimens con- closely related to V. gouldi, but differs by hav- tain a subcolumellar fold (= basal fold of Pils- ing slighter striations [and] a well-developed bry [1948:971]) that, along with the other angular lamella. . . .” We see little reason based 2002] PLEISTOCENE MOLLUSKS OF THE BLACK HILLS 135 on published accounts to relegate a slightly information concerning extant mollusks distinct and incompletely understood morph of the Black Hills, we refer to Frest and to species level due solely to chronological Johannes (1993). These data provide a base- assignment. For this reason we have cautious- line for comparison of habitat preferences ly identified our specimens as V. gouldi hannai associated with particular modern or fossil taxa morph, with the hope that more specimens in the Black Hills (i.e., found only in moist can be found in the future to better under- areas). stand this apparently extirpated form. The geographic distributions of extant North American mollusks are fairly well known, Pupilla muscorum although the factors that control such distribu- (Linnaeus, 1758) tions are not completely understood. Specifi- Identification of Pupilla muscorum follows cally, the ecological factors that control mod- Pilsbry (1948). The aperture of some referred ern land snail distributions (e.g., moisture, vege- specimens (MSL 1350–1360, NAUQSP 11597– tation, temperature, etc.) are numerous and 11604) contains a thickened calcareous de- the relationships between those factors in lim- posit, which separates them from Pupilla hebes. iting geographic distributions are not well Pupilla muscorum is not known to be extant in known (Goodfriend 1992). Nonetheless, pre- the Black Hills. liminary statements on the paleoenvironmen- A large number of specimens of Pupilla from tal meaning of fossil mollusks can be extrapo- Nelson-Wittenberg are “Pupilla hebes-like” lated using available data on extant mollusk (NAUQSP 11590–11596). The primary charac- distributions and their associated habitat pref- ter used to separate P. muscorum and P. hebes erences. is the lack of a thickened calcareous deposit in The Mammoth Site, dating to 26,000 yr B.P., the aperture of the latter. Separation of P. was previously sampled for mollusks, and a hebes from P. muscorum in the fossil record is preliminary account of recovered species was problematic, and we are not overly confident presented in Mead et al. (1994). Sediments in our ability to distinguish fossil or modern and mollusks from the site indicate the pres- shell specimens of these species. Although size ence of abundant, open water. The aquatic variation, paralleling climate change, in fossil genus Physella is by far the most abundant Pupilla muscorum is recognized (Rousseau component of the Mammoth Site molluscan 1989), there are currently no published diag- fauna (Table 2). Previous interpretations by noses on natural variation in anatomical shell Agenbroad and Mead (1994) state that the characters of fossil or modern P. muscorum ponding water may have been warm and that and P. hebes. We identify a large number of the local environment was treeless. Having a Pupilla specimens as “Pupilla hebes-like” based molluscan fauna highly depauperate in diver- on the lack of a thickened calcareous deposit sity but dominated by physid snails was con- in the aperture (Table 1). sidered consistent with a warm-water regime. The additional taxa reported here do not clar- DISCUSSION ify this interpretation. The Nelson-Wittenberg Site, possibly dat- Knowledge of the biology/ of extant ing to about 37,900 yr B.P., most likely repre- animals allows us to address the meaning of sents drainage from a nearby spring. There is their remains in Pleistocene fossil deposits. no indication of a warm-water spring at Nel- That is not to say that all fossil species have a son-Wittenberg. Unlike the Mammoth Site, modern analog or that extant animals could there is no indication either in sediments or not have existed under slightly differing eco- molluscan species that the spring supported a logical conditions in the past. Rather, studies body of open, ponded water at, or adjacent to, of the habitat preferences, , and this locality. All molluscan taxa are terrestrial ecology of extant mollusks, as well as verte- forms. Catinella sp. is the most abundant snail brates, provide us a starting point in address- from Nelson-Wittenberg (Table 2) but is not of ing paleoenvironmental questions. great utility for paleoenvironmental interpre- Most data on habitats associated with par- tation. The affinities of snails from Nelson- ticular molluscan taxa come from regional and Wittenberg are primarily with the Rocky local surveys of modern malacofaunas. For Mountain or Northern molluscan provinces. 136 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 2. Relative of mollusks reported here from the Mammoth Site and Nelson-Wittenberg Site, to near- est 0.1%. MSL 1300–1478 are included here as representative of the Mammoth Site. Ongoing excavation at the Mam- moth Site will likely produce additional specimens. The value of calculated relative abundance is limited due to sam- pling methods. NISP reflects the number of individual specimens of each species. For easily broken shells (Catinella, Physa, Fossaria), NISP was calculated by counting the apex in order to eliminate a bias toward less friable species. For others (Pupillidae), complete apertures were counted. Unless indicative of a distinct genus or species (Discus sp.), speci- mens referred to “sp.” (e.g., Pupilla sp.) were not counted. Nelson-Wittenberg total NISP = 829; Mammoth Site total NISP = 203.

______Nelson-Wittenberg ______Mammoth Site NISP Rel. Abund. NISP Rel. Abund. Pisidium casertanum — — 1 0.5 Pisidium compressum — — 2 1.0 Pisidium walkeri — — 4 2.0 Fossaria parva — — 1 0.5 Physella sp. — — 132 65.0 Gyraulus sp. — — 2 1.0 Columella columella alticola 4 0.5 — — Pupilla muscorum 90 10.8 14 6.9 Pupilla hebes-like 122 14.7 — — Pupoides albilabrisa 1 0.1 — — Vertigo modesta 202 24.4 1 0.5 Vertigo modesta parietalis 2 0.2 — — Vertigo gouldi hannai 4 0.5 — — VALLONIIDAE Vallonia gracilicosta 145 17.6 31 15.3 Vallonia cyclophorella 4 0.5 — — SUCCINEIDAE Catinella sp. 216 26.1 15 7.4 DISCIDAE Discus sp. 3 0.4 — — ZONITIDAE Hawaiia minuscula 6 0.7 — — LIMACIDAE Deroceras laeve 29 3.5 — — TOTALS 828 100.0 203 100.0 aAppears to be a modern contaminant.

Widespread species, such as the slug (Dero- the idea of a moist, local setting. Such a scenario ceras laeve) and the snail (Hawaiia miniscula), does not necessarily imply more mesic condi- are also present. The Nelson-Wittenberg mol- tions for the southern Black Hills as a whole, luscan fauna, as a whole, lacks some of the but may merely record the existence of a greater eastern and southern elements (e.g., Gastro- number of areas of spring discharge. copta procera, G. armifera) found in the Late The lack of modern Columella columella Holocene Graveyard Cave fauna and local alticola specimens in the Black Hills is puz- modern surveys (Table 1). zling considering the affinity of Black Hills mol- The presence of Vertigo modesta, Columella lusks with Rocky Mountain forms. Columella columella alticola, and Pupilla muscorum at columella alticola is typically a boreal form the Nelson-Wittenberg Site is noteworthy. from the Rocky Mountain molluscan province Vertigo modesta is common at the Nelson-Wit- (Bequaert and Miller 1973). At the more south- tenberg Site (Table 2). Frest and Johannes ern end of its range, C. c. alticola is usually (1993) reported V. modesta (as V. m. modesta) found at high elevations where moister condi- from wet, lowland areas of the central Black tions occur (Bequaert and Miller 1973). La Hills. Certainly, more mesic local conditions Roque’s (1970) data on the ecology of the than today existed in order to support species species indicate an affinity for moist areas. such as Vertigo modesta, Catinella sp., and Columella columella alticola is a rare species Deroceras laeve at the Nelson-Wittenberg Site. from the deposit; only 4 specimens were con- Indication of spring drainage further supports fidently assigned. 2002] PLEISTOCENE MOLLUSKS OF THE BLACK HILLS 137

Pupilla muscorum (considered a Northern steppe habitat from northwestern South Dakota Province species by Frest and Johannes 1993) to northern Colorado, eastern California, and is not reported in modern surveys of the Black southern Canada (Jones et al. 1985). The pres- Hills, and yet this taxon was relatively com- ence of sagebrush voles in the southern Black mon at the Nelson-Wittenberg Site, more so if Hills during the Pleistocene is consistent with the P. hebes-like forms are actually P. musco- the previous interpretations of a shrub-steppe rum (Table 2). As with Columella columella environment discussed by Mead et al. (1994) alticola, at the southernmost portion of its and others mentioned above. Shrub-steppe range (Arizona and New Mexico), P. muscorum environments may have surrounded the moister is usually found at high elevations (2043–569 and possibly more vegetated spring area at m [6700–12,000 feet]; Bequaert and Miller Nelson-Wittenberg. There appears to be some 1973). disparity between molluscan and mammalian Based on modern geographic and altitudi- environmental indicators at Nelson-Witten- nal distributions, we hypothesize that the berg. These differences suggest the presence presence of Pupilla muscorum and Columella of a mosaic of communities or microhabitats columella alticola at Nelson-Wittenberg implies for the Pleistocene in the southern Black Hills, cooler or more equable temperatures (i.e., a much in the same way that the modern Red lack of extremes, especially hot summer tem- Valley is predominantly grassland with inter- perature) than those existing today. Although spersed riparian and ponderosa pine micro- Black Hills climates are fairly mild today com- habitats. The presence of extralimital, boreal- pared to the surrounding plains, temperature adapted molluscan species may reflect the extremes are not uncommon (Froiland 1990). presence of a localized woodland or wet By themselves, mollusks from Nelson-Wit- meadow (spring discharge-based) tenberg do not support (or refute) previous interspersed with a regional shrub-steppe environmental interpretations for the southern community as indicated by Lemmiscus and Black Hills during the Pleistocene (see discus- other regional studies. sion above). The presence of Columella col- Paleoenvironmental indicators for the south- umella alticola and Vertigo spp. suggests that a ern Black Hills during the Late Pleistocene fairly well-vegetated area (possibly a wet hint at a more equable climate. The recovery meadow or woodland) was nearby, whether at of taxa typically associated with more boreal or adjacent to a spring. The mollusks do appear or northern habitats indicates some changes in to indicate cooler, possibly more equable, re- local environments from the Pleistocene to gional climates and a greater number of local recent. Certain species (e.g., Columella col- mesic microhabitats. Such an interpretation umella alticola, Pupilla muscorum) may not fits a pattern seen in the paleoenvironmental have been able to adapt to changes in the records from Agate Basin and Lange-Ferguson southern Black Hills environment through the (Evanoff 1982, Martin 1987). Although these Pleistocene/Holocene transition. The altither- sites are not chronologically equivalent to our mal (8000 yr B.P. to 4500 yr B.P.) in the Red data sets, they indicate the presence of (at Valley of the southern Black Hills was shown least) a higher number of localized mesic areas to be characterized by much different climatic in or near the Black Hills during the Pleis- conditions than the Pleistocene (Fredlund tocene. When the mollusks are compared with 1996). Presently, however, subtle local-level small mammal species from Nelson-Witten- changes in the southern Black Hills environ- berg for the purpose of paleoenvironmental ment and malacofauna do not appear as drastic interpretation, some unique differences are as seen in other areas of the Great Plains. evident. When compared with other Late Pleisto- The presence of Lemmiscus curtatus cene malacofaunas, one unique aspect of (NAUQSP 11629) at Nelson-Wittenberg is Mammoth Site and Nelson-Wittenberg Site notable given our fossil mollusk-based hypoth- mollusk assemblages is their overall similarity esis of a more vegetated, moist locality (rela- to the modern molluscan fauna of the Black tive to modern). The closest extant population Hills (Table 1). Both localities indicate that a of sagebrush vole is distributed north and west portion of the extant Black Hills molluscan of the locality by approximately 200–240 km. fauna may have been established by at least Sagebrush voles inhabit temperate-arid, shrub- the Late Pleistocene, during a glacial climatic 138 WESTERN NORTH AMERICAN NATURALIST [Volume 62 phase. The majority of species recovered from in controlling the distribution of C. c. alticola fossil deposits in the southern Black Hills are and P. muscorum in North America. More still extant in the region today (Table 1). This research on the physical and biotic factors that is in stark contrast to other localities in the control mollusk distributions is sorely needed. Great Plains where Pleistocene mollusk diver- sity was greater and representative of more ACKNOWLEDGMENTS biogeographic heterogeneity than modern (Baker et al. 1986, Frest and Dickson 1986). Collection permits and continued support Whether this will be a consistent trend in the were received from personnel of Wind Cave Black Hills or simply an artifact of sampling is National Park, especially Jim Nepstad, Ron presently unknown. Terry, and Rod Horrocks. We thank Jeff Nekola and an anonymous reviewer for helpful cri- CONCLUSIONS tiques of this paper. Carol Haden was instru- mental in the collection of materials from The assemblages under consideration are Graveyard Cave. Joe Muller, Kathy Anderson, distinct from one another in time and charac- and Judy Davids of the Mammoth Site helped ter. Neither locality contains a long strati- with several aspects of this project. The Board graphic record, and both appear to have short of Directors of the Hot Springs Mammoth Site depositional histories. The Mammoth Site mol- provided partial financial support. “Ole” Nelson luscan fauna is indicative of a unique local and Paul Wittenberg made it possible to re- microhabitat. Recovered mollusks are predom- cover fossils from the Nelson-Wittenberg Site. inantly aquatic, indicating abundant water. Special thanks to Chris Bell for his editorial The Nelson-Wittenberg mollusks, along with comments and identification of microtine recovered mammalian species, provide a fur- rodents from the Nelson-Wittenberg Site. The ther glimpse into Pleistocene environments in locality map was produced by Ron Redsteer the Black Hills. Recovered mollusk taxa sug- of the Research Imaging Laboratory, Bilby gest cooler or more equable temperatures Research Center, Northern Arizona Univer- than modern. Based on the recovery of extra- sity. Lynn Hetlet, Brian Smith, Doug Back- limital (Columella collumella alticola) and extra- lund, and Danny Walker assisted greatly in local (Vertigo modesta) taxa, a wet meadow or tracking down references. Jill Reeck gave woodland is inferred for portions of the south- directions for citation of Forest Service reports. ern Black Hills at approximately 37,900 yr B.P. Ray Middleton sorted and preliminarily iden- Remains of sagebrush vole indicate shrub- tified Graveyard Cave mollusks. Steve Jor- steppe in areas of the Red Valley. Together, gensen made editorial comments. We thank Pleistocene mollusks and mammals suggest a Steve Luken, Dave Moore, John Bayles, and slightly different mosaic of communities than Renata Brunner Jass for their help in the field. occurs in the southern Black Hills today. Future studies may provide greater accu- LITERATURE CITED racy in depicting long-term environmental and climatic changes or lack thereof. Addi- AGENBROAD, L.D. 1994. Geology, hydrology, and excava- tional molluscan records from the region will tion of the site. Pages 15–27 in L.D. Agenbroad and help further clarify small-scale environmental J.I. Mead, editors, The Hot Springs Mammoth Site: a decade of field and laboratory research in paleon- change (microhabitat-level) from large-scale tology, geology, and paleoecology. The Mammoth patterns of paleoenvironmental change or sta- Site of Hot Springs, SD. sis. Presently, the lack of reliable, stratigraphi- AGENBROAD, L.D., AND J.I. MEAD. 1994. The Hot Springs cally continuous records prohibits any inter- Mammoth Site: a decade of field and laboratory pretation beyond a localized, site-by-site basis. research in paleontology, geology, and paleoecology. The Mammoth Site of Hot Springs, SD. 457 pp. As localities of different ages are identified, AGENBROAD, L.D., J.I. MEAD, AND L.W. NELSON, EDITORS. the faunal record of the Black Hills will become 1990. Megafauna and man: discovery of America’s more useful in answering questions about heartland. The Mammoth Site of Hot Springs, South long-term paleoenvironmental change. Dakota, Scientific Papers 1:1–143. The validity of our hypothesis of “cooler” BAKER, R.G., R.S. RHODES II, D.P. SCHWERT, A.C. ASH- WORTH, T.J. FREST, G.R. HALLBERG, AND J.A. JANS- or more equable temperatures depends entire- SENS. 1986. A full-glacial biota from southeastern ly on the influence of specific ecological factors Iowa, USA. Journal of Quaternary Science 1:91–107. 2002] PLEISTOCENE MOLLUSKS OF THE BLACK HILLS 139

BELL, C.J., AND J.I. MEAD. 1998. Late Pleistocene micro- Plains. Pages 15–27 in H.E. Wright, Jr., and D.G. tine rodents from Snake Creek Burial Cave, White Frey, editors, The Quaternary of the United States. Pine County, Nevada. Great Basin Naturalist 58: Princeton University Press, Princeton, NJ. 82–86. LEONARD, A.B. 1982. Ecological and climatic implications BEQUAERT, J.C., AND W. B. M ILLER. 1973. The mollusks of of fossil mollusks at the Lange/Ferguson (39SH33) the arid Southwest with an Arizona check list. Uni- Clovis kill site. Page 64 in Programs and abstracts of versity of Arizona Press, Tucson. 271 pp. the forty-seventh annual meeting of the Society of BURCH, J.B. 1962. How to know the eastern land snails. American Archaeology. W.C. Brown, Dubuque, IA. 215 pp. MARTIN, J.E. 1987. Paleoenvironment of the Lange/Fergu- ______. 1989. North American freshwater snails. Malaco- son Clovis kill site in the Badlands of South Dakota. logical Publications, Hamburg, MI. 365 pp. Pages 314–332 in R.W. Graham, H.A. Semken, Jr., CLARKE, A.H. 1981. The freshwater molluscs of Canada. and M.A. Graham, editors, Late Quaternary mam- National Museum of Natural Sciences, National malian biogeography and environments of the Great Museum of Canada, Ottawa. 446 pp. Plains and prairies. Illinois State Museum, Scientific CZAPLEWSKI, N.J., AND J.I. MEAD. 1994. Late Pleistocene Papers, Volume 22. small mammals from Hot Springs Mammoth Site, MEAD, J.I., R.H. HEVLY, AND L.D. AGENBROAD. 1994. Late South Dakota. Pages 136–149 in L.D. Agenbroad Pleistocene invertebrate and plant remains, Mam- and J.I. Mead, editors, The Hot Springs Mammoth moth Site, Black Hills, South Dakota. Pages 117–135 Site: a decade of field and laboratory research in in L.D. Agenbroad and J.I. Mead, editors, The Hot paleontology, geology, and paleoecology. The Mam- Springs Mammoth Site: a decade of field and labora- moth Site of Hot Springs, SD. tory research in paleontology, geology, and paleoe- EVANOFF, E. 1982. Fossil nonmarine gastropods. Pages cology. The Mammoth Site of Hot Springs, SD. 357–359 in G.C. Frison and D.J. Stanford, The Agate MEAD, J.I., C. MANGANARO, C.A. REPENNING, AND L.D. Basin Site: a record of the PaleoIndian occupation of AGENBROAD. 1996. Early Rancholabrean mammals the northwestern High Plains. Academic Press, New from Salamander Cave, Black Hills, South Dakota. York. Pages 458–482 in K.M. Stewart and K.L. Seymour, FREST, T.J., AND J.R. DICKSON. 1986. Land snails (Pleis- editors, Palaeoecology and palaeoenvironments of tocene–Recent) of the Loess Hills: a preliminary late Cenozoic mammals: tributes to the career of survey. Proceedings of the Iowa Academy of Science C.S. (Rufus) Churcher. University of Toronto Press, 93:130–157. Toronto, Ontario. FREST, T.J., AND E.J. JOHANNES. 1993. Land snail survey of OVER, W.H. 1915. Mollusca of South Dakota. Nautilus 29: the Black Hills National Forest, South Dakota and 79–87, 90–95. Wyoming. Deixis Consultants Final Report Contract PILSBRY, H.A. 1946. Land Mollusca of North America #43-67TO-2-0054 to the USDA Forest Service Black (north of Mexico). Academy of Natural Sciences of Hills National Forest and USDI Fish and Wildlife Philadelphia Monographs, Volume 2, Number 3, Service South Dakota State Office. 156 pp. Part 1. FREDLUND, G.G. 1996. Late Quaternary geomorphic his- ______. 1948. Land Mollusca of North America (north of tory of lower Highland Creek, Wind Cave National Mexico). Academy of Natural Sciences of Philadel- Park, South Dakota. Physical Geography 17:446–464. phia Monographs, Volume 2, Number 3, Part 2. FROILAND, S. 1990. Natural history of the Black Hills and ROSCOE, E.J. 1954. Terrestrial gastropods from the Black Badlands. Center for Western Studies, Sioux Falls, Hills, Lawrence County, South Dakota, with a sum- SD. 225 pp. mary of published records from the state. Proceed- GOODFRIEND, G.A. 1992. The use of land snail shells in ings of the Utah Academy of Sciences, Arts, and Let- paleoenvironmental reconstruction. Quaternary Sci- ters 31:67–72. ence Reviews 11:665–685. ROSSEAU, D.D. 1989. Reponses des malacofaunes ter- HENDERSON, J. 1927. Some South Dakota Mollusca. Nau- restres Quaternaries aux contraintes climatiques en tilus 41:19–20. Europe Septentrionale. Palaeogeography, Palaeocli- HERRINGTON, H.B. 1962. A revision of the Sphaeriidae of matology, Palaeoecology 69:113–124. North America (Mollusca: Pelecypoda). Miscellane- TAYLOR, D.W. 1960. Late Cenozoic molluscan faunas from ous Publications of the Museum of Zoology, Univer- the High Plains. United States Geological Survey sity of Michigan 118:1–74. Professional Paper 337:1–94. HUBRICHT, L. 1985. The distributions of the native land ______. 1965. The study of Pleistocene nonmarine mol- mollusks of the eastern United States. Fieldiana lusks in North America. Pages 597–611 in H.E. (Zoology), New Series 24:1–191. Wright, Jr., and D.G. Frey, editors, The Quaternary JONES, J.K., JR., D.M. ARMSTRONG, AND J.R. CHOATE. 1985. of the United States. Princeton University Press, Guide to mammals of the plains states. University of Princeton, NJ. Nebraska Press, Lincoln. 371 pp. TURGEON, D.D., J.F. QUINN, JR., A.E. BOGAN, E.V. COAN, LA ROCQUE, A. 1970. Pleistocene Mollusca of Ohio. Ohio F. G . H OCHBERG, W.G. LYONS, P.M. MIKKELSEN, ET Geological Survey Bulletin 62:555–800. AL. 1998. Common and scientific names of aquatic LAURY, R.L. 1994. Paleoenvironment of the Hot Springs invertebrates from the United States and Canada: Mammoth Site. Pages 28–67 in L.D. Agenbroad and mollusks. 2nd edition. Special Publication 26, Amer- J.I. Mead, editors, The Hot Springs Mammoth Site: ican Fisheries Society, Bethesda, MD. 509 pp. a decade of field and laboratory research in paleon- VOORHIES, M.R., AND R.G. CORNER. 1985. Small mammals tology, geology, and paleoecology. The Mammoth with boreal affinities in late Pleistocene (Ranchola- Site of Hot Springs, SD. brean) deposits of eastern and central Nebraska. LEMKE, R.W., W.M. LAIRD, M.J. TIPTON, AND R.M. LIND- Institute for Tertiary-Quaternary Studies, TER-QUA VALL. 1965. Quaternary geology of northern Great Symposium Series 1:125–141. 140 WESTERN NORTH AMERICAN NATURALIST [Volume 62

WALKER, D.N. 1982. Early Holocene vertebrate fauna. Mead, and L.W. Nelson, editors, Megafauna and man: Pages 274–308 in G.C. Frison and D.J. Stanford, The discovery of America’s heartland. Scientific Publica- Agate Basin Site: a record of the PaleoIndian occu- tions of the Hot Springs Mammoth Site 1, SD. pation of the northwestern High Plains. Academic WOODMAN, N., D.P. SCHWERT, T.J. FREST, AND A.C. ASH- Press, New York. WORTH. 1996. Paleoecology of subarctic faunal assem- ______. 1987. Late Pleistocene/Holocene environmental blages from the Woodfordian Age (Pleistocene: Wis- changes in Wyoming: the mammalian record. Pages consinan) Elkander Site, northeastern Iowa. Occasion- 334–392 in R.W. Graham, H.A. Semken, Jr., and M.A. al Papers of the Natural History Museum, University Graham, editors, Late Quaternary mammalian bio- of Kansas 178:1–33. geography and environments of the Great Plains and prairies. Illinois State Museum, Scientific Papers 22. Received 14 June 2000 WEEDON, R.R., AND P. M . W OLKEN. 1990. The Black Hills Accepted 6 March 2001 environment. Pages 123–135 in L.D. Agenbroad, J.I. Western North American Naturalist 62(2), © 2002, pp. 141–150

DOES THE PRESENCE OF WYETHIA MOLLIS AFFECT GROWTH OF PINUS JEFFREYI SEEDLINGS?

G.M. Riegel1, T.J. Svejcar2, and M.D. Busse3

ABSTRACT.—Regeneration of Pinus jeffreyi in the Sierra Nevada is often limited on sites dominated by Wyethia mol- lis. Allelopathic chemicals and for soil moisture have been suggested as possible mechanisms for limiting regeneration. We tested the hypothesis that soil chemical and microbial properties from sites in different stages of suc- cession influence seedling growth of Pinus jeffreyi. Soil was collected from an early-seral site dominated by Wyethia mollis, a mid-seral site dominated by the shrubs Arctostaphylos patula, Ceanothus prostratus, C. velutinus, and Purshia tridentata, and a late-seral site dominated by mature Pinus. These sites were compared for nutrient content, Pinus seedling growth capacity, and microbial . Soil (0–33 cm) from the early-seral site had the lowest C, micro- bial , and fungal and bacterial populations. There were no consistent trends in soil nutrient content among sites. The early-seral site had the lowest soil Ca and Mg contents but also had a lower C/N ratio and more than twofold greater P content than either the mid- or late-seral site. Pinus seedling growth and foliar nutrient concentrations were compared at 3 harvest dates (220, 314, and 417 days after germination) in a greenhouse bioassay. The treatment design was a 3 × 2 factorial with soil from each of the 3 sites either with or without Pinus seedlings. Pots without seedlings were used as controls to assess the effects of seedlings on microbial biomass. Seedling growth in the early-seral soil was ini- tially suppressed in comparison to growth in the mid-seral soil, but by the final harvest total seedling weight was similar between these 2 treatments. The most obvious treatment effect was a reduction in growth for seedlings planted in late- seral soil, probably due to a nutrient imbalance in the soil. Seedlings grown in late-seral soil had Fe and Al levels that were nearly twice those of seedlings grown in early- and mid-seral soils. Microbial biomass followed a temporal pattern similar to that found for seedling growth. Differences in microbial biomass between the early- and mid-seral soils, although initially large, were not detected by the final harvest. We interpret these results to indicate that allelopathy or soil nutrient deficiencies resulting from the presence of Wyethia are unlikely to be responsible for limited growth of Pinus seedlings in Wyethia-dominated stands.

Key words: allelopathy, Wyethia mollis, Pinus jeffreyi, succession, forest regeneration, seedling growth, foliar nutri- ents, soil chemical, soil microbial biomass.

In the eastern slopes of the Sierra Nevada, foliage, Wyethia was most likely an early suc- regeneration of Pinus jeffreyi is often limited cessional plant with low frequency and cover on sites dominated by Wyethia mollis (Yoder- prior to European arrival (Young and Evans Williams and Parker 1987, Parker and Yoder- 1979, Williams 1995). Effective suppression Williams 1989, Williams 1995). Wyethia occurs and exclusion of fire since the early 1930s as an understory species of pine forests (P. jef- have resulted in increased fuel loading and freyi, P. ponderosa, and P. washoensis) east of higher forest stand densities (Agee 1993). As a the Cascade Range from south central Oregon result, wildfires now burn with greater severity and northeastern California through the east- and intensity than the frequent, presuppres- ern Sierra Nevada of California and Nevada sion fires (3–20 years) and provide Wyethia (Hopkins 1979, Riegel 1982, Smith 1994, with a competitive edge for expansion and Sawyer and Keeler-Wolf 1995), as an associate (Rundel et al. 1977, Agee 1993, species in the montane and subalpine Arte- Barbour and Minnich 2000). Heavy livestock misia steppe throughout this range (Cronquist grazing of palatable from 1860 to the 1994), and as isolated populations in central 1930s, primarily by sheep, also resulted in Nevada (Kartesz 1988). An herbaceous, distur- increased Wyethia abundance (Coville 1898, bance-adapted, long-lived perennial (>50 Kennedy and Doten 1901, Leiberg 1902, USDA years) with a deep taproot and fleshy resinous Forest Service 1937, Olmstead 1957). By the

1USDA Agricultural Research Service, 920 Valley Road, Reno, NV 89512. Present address: USDA Forest Service, Pacific Northwest Region, Area Ecology Program, 1645 NE Highway 20, Bend, OR 97701. 2USDA Agricultural Research Service, HC 71 4.51 Highway 205, Burns, OR 97720. 3USDA Forest Service, Pacific Southwest Research Station, 2400 Washington Ave., Redding, CA 96601.

141 142 WESTERN NORTH AMERICAN NATURALIST [Volume 62 end of the 19th century, Coville (1897) de- tissue leachates exhibited an allelopathic effect scribed Wyethia as being “abundant” in south on Pinus seedlings. In a field bioassay Wyethia central Oregon where the Klamath Indians litter inhibited germination and reduced radi- had traditionally used the mashed roots as a cle elongation of Pinus seedlings. Allelopathy poultice to control swelling. Wyethia contin- is thought to depress germination and early ues to dominate sites in Lassen National Park growth by inhibiting cell division and elonga- that have been excluded from livestock graz- tion, disrupting membrane regulation, nutri- ing for more than 80 years (Oswald et al. ent uptake, respiration, and 1995). (Rice 1984). However, allelopathy can be diffi- Though there are many pathways succes- cult to demonstrate and remains controversial sion can take following in pine (Harper 1977, Radosevich and Holt 1984, forests, a generalized summary includes (1) Kelsey and Everett 1995, Dakshini et al. 1999). early seral: Wyethia dominance with associ- Depression of biomass production may be ated herbaceous species and some conifer re- caused by leaf extracts other than allelochemi- generation; (2) mid-seral: Arctostaphylos patula, cals. For example, plant litter high in labile C, Ceanothus prostratus, C. velutinus, and Purshia particularly carbohydrates, may enhance soil tridentata shrub dominance with pole-sized microbial growth and result in increased nutri- conifers; (3) and late seral: conifer dominance ent immobilization and decreased plant growth with some shade-tolerant herbs and shrubs. (Michelsen et al. 1995). Wyethia-dominated sites may suppress suc- In this study we tested the hypothesis that cession to mid-seral conifer regeneration for Pinus jeffreyi seedling growth is limited due to up to 100 years (Kennedy and Doten 1901, the presence of Wyethia in early-seral sites. Evanko 1951, Parker and Yoder-Williams 1989) Soils and litter were collected from early-seral and can persist in the understory of mid- and (Wyethia-dominated), mid-seral (shrub-domi- late-seral Pinus jeffreyi overstories as an infre- nated), and late-seral (depauperate understory quent, low-cover (<5%) associate (Smith 1994). with mature Pinus overstory) sites and used as Shrub dominance in the mid-seral phase growth media for greenhouse-grown Pinus should increase the soil nutrient capital with seedlings. time between fire and logging disturbance intervals that temporarily reduce the cover of METHODS N-fixing shrubs, Ceanothus and Purshia (Con- Study Area ard et al. 1985, Johnson 1995, Busse et al. 1996, The study area (39°53′N, 120°25′W) is 9.7 Busse 2000a, 2000b). km northeast of Beckwourth, Plumas County, Parker and Yoder-Williams (1989) proposed in the northern Sierra Nevada of California. 2 hypotheses to explain the dominance of Precipitation at nearby Clover Valley averaged Wyethia: (1) allelopathy and (2) competition 903 mm per year (s = 342) from 1959 through for water. However, the precise mechanism(s) 1991, approximately 85% of which was in the responsible for poor regeneration on sites form of snow falling in the months of Novem- dominated by Wyethia remains unclear. After ber through March (State of California, Re- a forest is removed through fire or logging, the sources Agency, Department of Water Re- developing shrub and herbaceous understory sources, personal communication, 1993). Soils often suppresses conifer regeneration as a are loamy-skeletal, mixed, frigid Ultic Agrixe- result of competition for soil moisture (Conard rolls (Sattley Family). They are moderately to and Radosevich 1981, 1982, McDonald 1983a, well drained and formed from weathered Lanini and Radosevich 1986, Shainsky and andesitic tuff breccia (USDA Forest Service Radosevich 1986, Parker and Yoder-Williams 1988). Horizon development includes a grav- 1989). Water-extractable allelochemicals from elly loam A1 (0–7.6 cm), a very cobbly loam Wyethia have also been suggested as an addi- A2 (7.6–20.3 cm), a cobbly loam Bt1 (20.3– tional factor suppressing Pinus jeffreyi regen- 33.0 cm), and a very cobbly clay loam Bt2 eration (Heisey and Delwiche 1983, Yoder- (33.0–117 cm) above bedrock. Williams and Parker 1987, Parker and Yoder- We selected 3 sites within a Pinus jeffreyi Williams 1989, Williams 1995). Yoder-Williams forest type (Sawyer and Keeler-Wolf 1995): and Parker (1987) hypothesized that Wyethia (1) early-seral site with Wyethia-dominated 2002] REGENERATION OF PINUS JEFFREYI 143 understory, (2) mid-seral, shrub-dominated lected from the A1 horizon at each site on 8 stand, and (3) late-seral, Pinus-dominated stand. June 1990, and culturable bacteria (grown on The sites have similar elevation (1829 m) and 1/10 strength tryptic soy agar) and fungi (grown slope (0–3%), and are within 2.4 km of each on malt extract agar) were enumerated within other in Crocker Creek drainage. Wyethia is 48 hours of collection. found in or adjacent to every site with a Pinus Greenhouse Experiment overstory. The early-seral site is dominated by Wyethia with scattered Artemisia tridentata The experiment was a completely random- ssp. vaseyana, Lithophragma parviflorum, Col- ized block design with 6 treatments, 3 harvest linsia parviflora, and Paeonia brownii; the dates, and 10 replications. Treatment design mid-seral shrub site is dominated by Arcto- was a 3 × 2 factorial, with soil from the 3 sites staphylos patula, Ceanothus prostratus, C. velu- in combination with 2 vegetation treatments: tinus, Purshia tridentata, Elymus elymoides (= (1) soil from early-seral, Wyethia-dominated Sitanion hystrix), Poa fendleriana, and Lupinus site + Pinus; (2) soil from mid-seral, shrub- spp.; and the late-seral site has a very depau- dominated site + Pinus; (3) soil from late- perate understory with scattered Symphori- seral, Pinus-dominated site + Pinus; (4) soil carpos rotundifolius, Rosa woodsii, Hydro- from Wyethia-dominated site without Pinus; phyllum capitatum, Vicia americana, and Sar- (5) soil from mid-seral, shrub-dominated site codes sanguinea, an obligate mycotroph. Pinus without Pinus; and (6) soil from late-seral, dominates the overstory at all sites with increas- Pinus-dominated site without Pinus. Treat- ing amounts of Calocedrus decurrens and Abies ments receiving no seedlings were included as concolor found at the mid- and late-seral sites. controls to determine whether Pinus root pro- Vascular plant nomenclature and taxonomy liferation counteracts the effect of Wyethia follow Hickman (1993). allelochemicals on soil biota. To avoid poten- Pinus diameter and age distribution by site tial microclimatic effects, we placed the pots within a 0.10-ha plot were as follows: early- on the greenhouse bench and randomly seral mean DBH = 41.6 cm, sx– = 6.93 cm, assigned them to each replication by block. age <50 years (n = 9); mid-seral mean DBH The 4 replicate A1 horizons were bulked by = 59.0 cm, sx– = 6.55 cm, age 60–100 years (n site (early-, mid-, and late-seral), and 4 Pinus = 9); and late-seral mean DBH = 88.3 cm, sx– jeffreyi seeds were placed in each pot filled on = 16.95 cm, and age 100–226 years (n = 4). 19 June 1990. Pinus seeds, which were col- Ocular estimates of Wyethia canopy cover by lected in the Toiyabe National Forest at 1982 m site within the 0.10-ha plot were early-seral in 1988, were stored dry (8% humidity) prior 40%, mid-seral 10%, and late-seral 5%. to a 30-day stratification at 1°C (R. Walker, Uni- versity of Nevada, Reno). Pot volume was Soil Properties 2625 cm3 and height was 15 cm. Four replicate soil samples were collected Pots were watered until the first sign of randomly at each site on 29 May 1990 from drainage: daily the 1st week, every other day the A1, A2, and Bt1 horizons, transported to the 2nd week, and twice weekly thereafter. greenhouses in Reno, Nevada, and sieved (<2 Plants were thinned to 2 per pot on day 56 (13 mm) prior to nutrient analyses. Organic matter August) and eventually to 1 per pot on day 71 was determined by weight loss on ignition (28 August). (Nelson and Sommers 1982) and converted to We collected current-year litter from each organic C by dividing weight loss by 1.742. site, and 10 g (dry weight equivalent which The maximum furnace temperature was 400°C approximated the volume/weight ratio found to avoid loss of clay-bound H2O. Total N was in the field) was placed on the soil surface in determined with a continuous flow analyzer each pot on day 71 (28 August) after germina- following Kjeldahl digestion (Bremner and tion. Litter at each site varied in species com- Mulvaney 1982). Extractable P was determined position and spatial distribution. Wyethia litter by dilute acid fluoride extraction (Olsen and was collected from the early-seral site; Arc- Sommers 1982). Microbial biomass was deter- tostaphylos, Ceanothus velutinus, and Purshia mined by the chloroform-fumigation-incuba- litter from the mid-seral site; and Pinus litter tion method (Jenkinson and Powlson 1976). from the late-seral site. Throughout the exper- Additional soil samples (triplicate) were col- iment we added new litter monthly (~2.5 g) to 144 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 1. Selected chemical and microbial properties of mineral soil from the early-, mid-, and late-seral field sites. error (n = 4) except for viable counts (n = 3). Different lowercase letters within a column and horizon indicate signifi-

Horizon Site pH C N C/N P K Ca Mg

–1 –1 –1 - - - - - g kg - - - - - mg kg ------cmolc kg ------A1 Early-seral 6.3 (0.1)a 33 (2)b 1.7 (0.1)a 19 (1)b 83 (11)a 1.8 (0.1)a 16 (1)b 3.0 (0.2)c Mid-seral 6.4 (0.2)a 47 (3)a 1.5 (0.2)a 34 (5)a 35 (8)b 2.3 (0.2)a 25 (1)a 4.8 (0.3)b Late-seral 6.5 (0.1)a 38 (4)ab 1.4 (0.2)a 27 (1)ab 33 (3)b 2.3 (0.2)a 18 (1)b 5.8 (0.2)a

A2 Early-seral 6.2 (0.1)b 28 (2)a 1.5 (0.1)a 19 (1)c 60 (12)a 1.9 (0.1)a 17 (1)b 2.9 (0.1)c Mid-seral 6.4 (0.1)ab 30 (2)a 0.8 (0.1)c 37 (1)a 17 (2)b 1.5 (0.1)b 22 (1)a 4.5 (0.3)b Late-seral 6.5 (0.2)b 31 (2)a 1.1 (0.1)b 29 (1)b 19 (2)b 2.0 (0.1)a 19 (2)b 5.4 (0.3)a

Bt1 Early-seral 6.2 (0.1)a 26 (1)a 1.2 (0.1)a 21 (1)c 26 (8)a 2.0 (0.2)a 18 (1)b 3.1 (0.1)b Mid-seral 6.3 (0.1)a 31 (3)a 0.7 (0.1)b 45 (2)a 12 (1)b 1.2 (0.1)b 22 (1)a 4.6 (0.2)a Late-seral 6.5 (0.1)a 34 (1)a 1.1 (0.1)a 31 (1)b 10 (4)b 1.9 (0.1)a 19 (1)b 5.2 (0.2)a each pot. Plants were harvested on 25 January, and C/N ratios were lower at the early-seral 24 May, and 4 September 1991. site compared to the mid- or late-seral sites. Mineral soil (50 g) was collected at each Microbial biomass and viable bacterial and harvest and tested for microbial biomass. After fungal populations in the upper horizon were washing soil from the Pinus seedlings, we sep- lowest at the early-seral site and highest at the arated them into aboveground and below- shrub-dominated, mid-seral site. Despite a ground components. The samples were dried smaller biomass, the microbial population at at 60°C for 48 hours and weighed. Needles the early-seral site had the highest C utilization were collected from the aboveground fraction efficiency among the 3 sites, as shown by and analyzed for nutrient content. Nitrogen lower qCO2 values (CO2 respired per unit bio- content was measured by Kjeldahl analysis mass; Table 1). (Bremner and Mulvaney 1982), and other Greenhouse Study nutrients were measured with an ICP spec- trophotometer (Issac and Johnson 1985) using PLANT BIOMASS.—There were no significant a nitric acid digest (Zarcinas et al. 1987). differences in Pinus germination between Data were analyzed with the General Lin- treatments, as nearly 95% of all seeds germi- ear Models procedure in SAS (SAS Institute nated. Pinus seedlings accumulated more total Inc. 1990). Mean separations were conducted mass when grown in soil from the mid-seral with the Waller-Duncan test at the P ≤ 0.05 site than from the late-seral site, but were not level. greater in total mass than the early-seral, Wyethia-dominated site at the 1st harvest (Table RESULTS 2). By the 2nd harvest the ranking for total plant mass was mid-seral > early-seral > late- Soil Properties seral. There was no difference in total weight Several differences in soil organic matter, of Pinus seedlings at the final harvest between nutrient status, and microbial characteristics mid-seral and early-seral treatments, but both were found between sites. The early-seral were greater than the late-seral treatment. site had the shallowest mean O horizon depth Root and shoot weights followed similar (early-seral = 1.85 cm (sx– = 0.47 cm); mid- trends (Table 2). seral = 3.78 cm (sx– = 0.72 cm); late-seral = FOLIAR NUTRIENT CONCENTRATION.—There 4.30 cm (sx– = 0.71 cm), in addition to the low- were no significant differences between treat- est C, Ca, and Mg contents at each mineral ment sites in foliar N, P, K, or Na (Table 3). soil depth (Table 1). The trend of lower nutri- Micronutrients were generally found in lowest ent status at the Wyethia-dominated, early- concentrations in foliage from the mid-seral seral site was not consistent, however. Phos- treatment (Table 3). Seedlings grown in soil phorus content was two- to threefold greater, from the late-seral site had higher levels of 2002] REGENERATION OF PINUS JEFFREYI 145

Depth of the upper horizons: A1 = 0–7.6 cm; A2 = 7.6–20 cm; Bt1 = 20–33 cm. Values in parentheses are standard cant differences (P < 0.05) between sites.

Microbial ______Viable counts biomass Cmicro/Ctotal qCO2 Bacteria Fungi mg kg–1 % ------× 106 g–1 ------688 (76)b 2.1 (0.2)a 0.21 (0.03)b 9.6 (0.1)c 0.3 (0.1)c 1217 (93)a 2.6 (0.3)a 0.38 (0.02)a 18.3 (3.0)a 2.0 (0.3)a 888 (64)b 2.4 (0.2)a 0.35 (0.02)a 12.0 (0.1)b 1.3 (0.1)b

545 (98)a 1.9 (0.4)a 0.23 (0.01)b 740 (90)a 2.3 (0.2)a 0.31 (0.06)a 691 (41)a 2.1 (0.1)a 0.37 (0.03)a

406 (86)a 1.6 (0.3)a 0.22 (0.02)b 557 (81)a 1.9 (0.3)a 0.31 (0.03)a 603 (17)a 1.8 (0.1)a 0.31 (0.08)a

Mg, Fe, and Al than the other treatments. Iron et al. 1989, 1992, Amaranthus et al. 1990, and Al concentrations were nearly twice those Borchers and Perry 1990). Results from this of the early- and mid-seral treatments. greenhouse study showed that seedling growth SOIL MICROBIAL BIOMASS.—The temporal in soil from the early-seral site was indeed response of microbial biomass was similar to lower than that of the mid-seral site during the pattern for total seedling growth. Micro- the 1st and 2nd harvests. By the 3rd harvest, bial biomass was greatest for the mid-seral however, differences between the 2 treatments treatment at the 1st and 2nd harvest (Table 4). were not significant. Thus, any nutrient limita- By the final harvest, the early- and mid-seral tions or allelopathic effects of Wyethia occurred treatments had comparable microbial biomass, only during the initial growth phase following and both were significantly greater than the germination and had no long-term effect on late-seral treatment. Root proliferation had lit- seedling growth. tle effect on microbial biomass; soils with or Leaves of Wyethia contain both water- without seedlings had the same microbial bio- extractable and volatile allelochemicals when mass at the 2nd harvest. A slight increase was collected soon after total leaf expansion in mid- found for soils without seedlings by the final to late spring (Heisey and Delwiche 1983). harvest, although the relative trend between These compounds include linoleic acid, isofla- sites (mid-seral = early-seral > late-seral) was vones, lanostene-type triterpenes, and some n- consistent, regardless of presence or absence alkalines (Waddell et al. 1982). Soil and litter of seedlings. were collected for the greenhouse study in late spring specifically for this reason, to approx- DISCUSSION imate the period of maximum leaf expansion We initially hypothesized that Pinus jeffreyi and allelochemical concentrations in Wyethia seedlings grown in soil from early-seral sites leaves. Below-average precipitation in the dominated by Wyethia mollis would produce year prior to soil collection may have also con- less total plant mass than seedlings grown in tributed to higher concentrations of allelo- soil from the shrub-dominated, mid-seral site. chemicals due to limited leaching losses. Given The basis for this hypothesis was potential these optimum conditions for allelochemicals, allelopathic effects of Wyethia (Yoder-Williams the lack of a sustained reduction in Pinus seed- and Parker 1987) and superior nutrient status ling growth in the Wyethia-dominated soil at the mid-seral site due to the beneficial suggests an unlikely role of allelopathy by effects of nitrogen-fixing shrubs (Conard et al. Wyethia in the early growth of Pinus, at least 1985, Johnson 1995, Busse et al. 1996, Busse via allelochemicals stored in the soils. Other 2000a, 2000b) and other microbial processes bioassays have shown that fresh Wyethia leaves purportedly associated with shrub-dominated are inhibitory to germination of Bromus mollis, sites (Amaranthus and Perry 1987, 1989, Perry Hordeum vulgare, and Lactuca spp. (Heisey 146 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 2. Total, shoot, and root dry mass of Pinus jeffreyi seedlings grown in soils from early-, mid-, and late-seral sites. Different lowercase letters within a row indicate significant differences (P < 0.05) between sites.

Days from ______Total plant mass (g) ______Shoot mass (g)______Root mass (g) sowing Early- Mid- Late- Early- Mid- Late- Early- Mid- Late- 220 2.01ab 2.45a 1.68b 0.68ab 0.82a 0.61b 1.33ab 1.63a 1.10b 314 7.80b 9.71a 6.31c 4.38a 4.77a 3.46b 3.42b 4.94a 2.85c 417 15.07a 15.72a 11.36b 6.52a 6.14a 4.43b 8.55a 9.58a 6.93b

TABLE 3. Foliar nutrient concentration of Pinus jeffreyi seedlings at 417 days after sowing. The seedlings were grown in soils from early-, mid-, or late-seral sites. Different lowercase letters within a column indicate significant differences (P < 0.05) between sites. Site N P K Ca Mg Mn Fe Zn Al B Cu Na ------mg g–1 ------µg g–1 ------Early-seral 8.56 1.93 6.94 3.83b 1.47b 160a 33b 96a 60b 31b 4.9a 37 Mid-seral 10.02 1.87 6.77 4.96a 1.53b 88b 33b 71b 45b 33ab 4.3b 50 Late-seral 8.78 1.75 7.10 4.05b 1.88a 137a 60a 77b 104a 37a 4.7ab 34

TABLE 4. Microbial biomass (mg C kg–1 soil) from the early-, mid-, and late-seral sites. Values are compared for pots with and without seedlings in order to separate the influence of seedling roots from site effects. Different lowercase let- ters within a row for with seedlings or without seedlings indicate significant differences (P < 0.05) between sites.

Days from ______With seedlings ______Without seedlings sowing Early- Mid- Late- Early- Mid- Late- 220 675b 913a 591b 348b 504a 392b 314 732b 875a 517c 740b 859a 513c 417 783a 799a 523b 858ab 928a 637b

and Delwiche 1983, Yoder-Williams and Parker 1985), and Abies concolor (Conard 1985) was 1987). Yoder-Williams and Parker (1987) found reported in soil-free bioassays. However, no reduced germination of Pinus jeffreyi seeds inhibitory effect of litter leachates was found that overwintered beneath Wyethia litter. In by Tinnin and Kirkpatrick (1985) when soil contrast, we found no evidence of allelopathic was used as the growth medium. It is doubt- inhibition of seed germination in soil from the ful, therefore, that litter from these shrubs Wyethia-dominated site. Germination of Pinus suppressed growth of Pinus in the mid-seral seeds was near 95% in our pot study, regard- treatment of our experiment. less of soil origin. Williams (1995) suggested We had anticipated an improved soil nutri- that direct interference by Wyethia is observed ent status, particularly soil N, at the shrub- as an allelopathic effect on germinating Pinus dominated site (mid-seral) compared to the seed in spring and in soil moisture utilization early-seral site due to the N fixation from acti- in late summer. These interacting processes norhizal symbionts associated with Ceanothus may be moderated during years of above-aver- prostratus, C. velutinus (Delwiche et al. 1965, age precipitation by leaching allelochemicals Conard et al. 1985, Busse 2000a, 2000b), and and increased soil moisture availability. Purshia tridentata (Webster et al. 1967, Busse Allelopathic inhibition from water-soluble 2000a, 2000b). Annual N fixation in forests leachates of Arctostaphylos and Ceanothus east of the Sierra Nevada and Cascade Range velutinus leaf litter has also been implicated in crest varies from 5 to 15 kg ha–1 for C. veluti- laboratory bioassays. Reduced radicle elonga- nus and 1 kg ha–1 for C. prostratus and Purshia tion of Bromus tectorum, Hordeum vulgare tridentata (Busse 2000a, 2000b). Busse et al. (del Moral and Cates 1971), Cucumis sativus, (1996) found increased soil C, N, and microbial Pseudotsuga menziesii (Tinnin and Kirkpatrick biomass in the upper horizon of a ponderosa 2002] REGENERATION OF PINUS JEFFREYI 147 pine forest due to long-term retention of (late-seral) Pinus stumps (Hopkins et al. 1988). shrubs. The shrub species in their study were The discrepancy between our findings and the same as those found at the mid-seral site observations from the field suggests that fac- (Ceanothus velutinus, Purshia tridentata, and tors other than allelochemicals or soil nutri- Arctostaphylos patula). Johnson (1995) also ents per se limit regeneration of Pinus on found improved soil N status at pine sites with Wyethia-dominated sites. Succession may be a dominance of C. velutinus. Results from our influenced by competition for soil moisture study indicate that the shrub-dominated site and changes in microclimate (Tilman and Wedin had greater soil C and microbial biomass, but 1991, De Pietri 1992). Previous research has slightly lower N than the early-seral site. Site shown that soil moisture availability on shrub- differences in soil P, Ca, Mg, and microbial dominated sites can be significantly greater characteristics had no measurable effect on than on Wyethia-dominated sites during the seedling growth in the greenhouse bioassay, latter part of the growing season (Williams however. Without knowing cover, density, and 1995). The shrub species associated with the duration of site occupancy of N-fixing shrubs, mid-seral stage compete with establishing it is difficult to interpret and compare the conifers for light, water, and nutrients (Conard results between other studies and ours. and Radosevich 1981, 1982, McDonald 1983a, The most striking effect we measured was a 1983b, Roy 1983, Radosevich 1984, Shainsky reduction in seedling mass for the Pinus-dom- and Radosevich 1986). A shrub canopy also inated, late-seral treatment. Although the provides a shaded microenvironment, how- mechanism(s) responsible for this observation ever, that reduces evaporative demand and is unclear, high Al and Fe concentrations in improves the water balance of trees, which foliage from the late-seral treatment suggest may be less physiologically stressful than com- either potential metal toxicity or nutrient im- petition for soil resources on hot, dry sites balance. Pinus litter has also been shown to (Conard and Radosevich 1982, Lanini and have allelopathic effects on plant growth (Lodhi Radosevich 1986). and Killingbeck 1980, 1982). It is not uncom- In conclusion, we could find only limited mon to find the area below large Pinus trees to evidence that allelochemicals from Wyethia have few species with low density and above- might suppress growth of Pinus seedlings. ground biomass. There is often a thick needle There was an initial reduction in growth of layer below large Pinus trees that may serve as seedlings in early- versus mid-seral soils, but a physical barrier to seed/soil contact for many after 417 days of growth the 2 treatments were species. Pinus litter generally has a high C/N similar. The more obvious response was the value (Hart et al. 1992), which can limit N reduction in seedling growth associated with availability. We found no evidence to support the late-seral, Pinus-dominated soil. We sug- this mechanism, however. Soil from the late- gest that research into the relationship of seral site had a lower C/N ratio than soil from microclimate and plant-soil water relations, the mid-seral site, and no significant differ- and an examination of the community compo- ences in foliar N concentrations were found sition of soil biota, may aid in explaining the between treatments in the greenhouse bioas- pattern of regeneration of Pinus jeffreyi. This say. Bever (1994) suggested that repeated information, coupled with an increased under- planting of a single species (analogous to the standing of larger scale climate effects (e.g., limited of the late-seral site) increased atmospheric temperature and may suppress plant growth due to negative drought cycles), may further our knowledge of feedback from soil biota or accumulations of the complex interactions involved in succes- specific pathogens, and/or change in composi- sion of Wyethia-dominated seral stages within tion of the community of mutalists. Root path- western pine forests. ogens were not observed in our short-term bioassay or in our field-collection sites. How- ACKNOWLEDGMENTS ever, Wyethia-dominated understories in Pinus ponderosa in south central Oregon are associ- We thank Dr. Roger F. Walker for his ated with higher rates of mortality of 10- to advice during the conception of this experi- 30-year Pinus regeneration from Heterobasid- ment and for the Jeffrey pine seeds. Dr. ium annosum root disease centered around old Michael P. Williams shared many stimulating 148 WESTERN NORTH AMERICAN NATURALIST [Volume 62 ideas and critiqued our hypothesis that was rhizal shrubs. Forest Ecology and Management 136: based on his original work. Sydney Smith gra- 85–95. BUSSE, M.D., P.H. COCHRAN, AND J.W. BARRETT. 1996. ciously shared her data and knowledge of north- Changes in ponderosa pine site follow- eastern California. Sara Prueitt Lovtang, Drs. ing removal of understory vegetation. Soil Science Robert F. Powers, John O. Sawyer, Walker, and Society of America Journal 60:1614–1621. Williams provided helpful editorial comments. CONARD, S.G. 1985. Inhibition of Abies concolor radicle growth by extracts of Ceanothus velutinus. Madroño Rick Schultz assisted in fieldwork, watered 32:118–120. and harvested the pines, and always kept us CONARD, S.G., AND S.R. RADOSEVICH. 1981. Photosynthesis, humored. Kara Paintner performed the sta- xylem pressure potential, and leaf conductance of tistical analyses. Dave Weixelman, Desiderio three montane chaparral species in California. Forest Zamudio, and Gene Lohrmayer assisted in Science 27:627–639. ______. 1982. Growth responses of white fir to decreased descriptive site measurements. Wayne Johann- shading and root competition by montane chaparral son provided soils and site location informa- shrubs. Forest Science 28:309–320. tion. Carol McDonald word-processed this CONARD, S.G., A.E. JARAMILLO, K. CROMACK, JR., AND S. manuscript. ROSE. 1985. The role of Ceanothus in western forest . U.S. Department of Agriculture, Forest Service, Pacific Northwest Forest and Range Experi- LITERATURE CITED ment Station, General Technical Report PNW-182, Portland, OR. AGEE, J.K. 1993. Fire ecology of Pacific Northwest forests. COVILLE, F.1897 Notes on the plants used by the Klamath Island Press, Washington, DC. 493 pp. Indians of Oregon. Contributions from the U.S. AMARANTHUS, M.P., AND D.A. PERRY. 1987. Effect of soil National Herbarium, Volume 5, No. 2. U.S. Depart- transfer on ectomycorrhiza formation and the sur- ment of Agriculture, Division of Botany. U.S. Gov- vival and growth of conifer seedlings on old, non- ernment Printing Office, Washington, DC. forested clear-cuts. Canadian Journal of Forest ______. 1898. Forest growth and sheep grazing in the Research 17:944–950. Cascade Mountains of Oregon. U.S. Department of ______. 1989. Interaction effects of vegetation type and Agriculture, Division of Forestry Bulletin 15, Wash- Pacific madrone soil inocula on survival, growth, and ington, DC. formation of Douglas-fir. Canadian Jour- CRONQUIST, A. 1994. Asterales. Page 17 in A. Cronquist, nal of Forest Research 19:550–556. A.H. Holmgren, N.H. Holmgren, J.L. Reveal, and AMARANTHUS, M.P., C.Y. LI, AND D.A. PERRY. 1990. Influ- P.K. Holmgren, Intermountain flora: vascular plants ence of vegetation type and madrone soil inoculum of the Intermountain West, U.S.A. Volume 5. New on associative nitrogen fixation in Douglas-fir rhizo- York Botanical Gardens, Bronx, NY. pheres. Canadian Journal of Forest Research 20: DAKSHINI, K.M.M., C.L. FOY, AND INDERJIT. 1999. Allelo- 368–371. pathy: one component in multifaceted approach to BARBOUR, M.G., AND R.A. MINNICH. 2000. California ecology. Pages 3–14 in Inderjit, K.M.M. Dakshini, upland forests and woodlands. Pages 161–202 in and C.L. Foy, editors, Principles and practices in M.G. Barbour and W.D. Billings, editors, North Amer- plant ecology: allelochemical interactions. CRC Press, ica terrestrial vegetation. 2nd. edition. Cambridge Boca Raton, FL. University Press, New York. DEL MORAL, R., AND R. CATES. 1971. Allelopathic poten- BEVER, J.D. 1994. Feedback between plants and their soil tial of the dominant vegetation of western Washing- communities in an old field community. Ecology ton. Ecology 52:1030–1037. 75:1965–1977. DELWICHE, C.C., P.J. ZINKE, AND C.M. JOHNSON. 1965. BORCHERS, S.L., AND D.A. PERRY. 1990. Growth and ecto- Nitrogen fixation by Ceanothus. Plant Physiology mycorrhiza formation of Douglas-fir seedlings grown 40:1045–1047. in soils collected at different distances from pioneer- DE PIETRI, D.E. 1992. Alien shrubs in a national park: can ing hardwoods in southwest Oregon clear-cuts. they help in the recovery of natural degraded forest? Canadian Journal of Forest Research 20:712–721. Biological Conservation 62:127–130. BREMNER, J.M., AND C.S. MULVANEY. 1982. Nitrogen— EVANKO, A.B. 1951. Response of Wyethia to 2,4-D. U.S. total. Pages 595–624 in A.L. Page et al., editors, Department of Agriculture, Forest Service, Northern Methods of soil analysis, part 2. 2nd edition. Agron- Rocky Mountain Forest Range Experiment Station omy Monographs 9. Agronomy Society of America Research Note 98, Fort Collins, CO. and Soil Science Society of America, Madison, WI. HARPER, J.L. 1977. Population biology of plants. Academic BUSSE, M.D. 2000a. Ecological significance of nitrogen Press, New York. fixation by actinorhizal shrubs in interior forests of HART, S.C., M.K. FIRESTONE, AND E.A. PAUL. 1992. Decom- California and Oregon. Pages 23–41 in R.F. Powers, position and nutrient dynamics of ponderosa pine D.L. Hauxwell, and G.M. Nakamura, technical coor- needles in a Mediterranean-type climate. Canadian dinators, Proceedings of the California Forest Soils Journal of Forest Research 22:306–314. Council conference on forest soil biology and forest HEISEY, R.M., AND C.C. DELWICHE. 1983. A survey of Cali- management. U.S. Department of Agriculture, Forest fornia plants for water-extractable and volatile in- Service General Technical Report PSW-178, Albany hibitors. Botanical Gazette 144:382–390. CA. HICKMAN, J.C., EDITOR. 1993. Jepson manual—higher ______. 2000b. Suitability and use of the 15N-istope dilu- plants of California. University of California Press, tion method to estimate nitrogen fixation by actino- Berkeley. 2002] REGENERATION OF PINUS JEFFREYI 149

HOPKINS, W.E. 1979. Plant associations of the Fremont NELSON, D.W., AND L.E. SOMMERS. 1982. Total carbon, National Forest. U.S. Department of Agriculture, For- organic carbon, and organic matter. Pages 539–579 est Service, Pacific Northwest Region, R6-ECOL- in A.L. Page et al., editors, Methods of soil analysis, 79-004, Portland, OR. part 2. 2nd edition. Agronomy Monographs 9. HOPKINS, W.E., D.J. GOHEEN, E.M. GOHEEN, AND K. FORRY. Agronomy Society of America and Soil Science Soci- 1988. Evaluation of annosus root disease on pon- ety of America, Madison, WI. derosa pine in the Fremont National Forest and Lake- OLMSTEAD, P. 1957. The Nevada-California-Oregon bor- view District, Bureau of Land Management. U.S. der triangle: a case study in sectional history. Mas- Department of Agriculture, Forest Service, Pacific ter’s thesis, University of Nevada, Reno. Northwest Region, Forest Pest Management Report OLSEN, S.R., AND L.E. SOMMERS. 1982. Phosphorus. Pages R6-88-04, Portland, OR. 403–430 in A.L. Page et al., editors, Methods of soil ISSAC, R.A., AND W.C. JOHNSON. 1985. Elemental analysis analysis, part 2. 2nd edition. Agronomy Monographs of plant tissue by plasma emission spectroscopy: col- 9. Agronomy Society of America and Soil Science laborative study. Journal of the Association of Offi- Society of America, Madison, WI. cial Analytic Chemists 68:499–505. OSWALD, V.H., D.W. SHOWERS, AND M.A. SHOWERS. 1995. JENKINSON, D.S., AND D.S. POWLSON. 1976. The effects of A flora of Lassen National Park, California. Califor- biocidal treatments on metabolism in soil, V. A method nia Native Plant Society, Sacramento. for measuring soil biomass. Soil Biology and Bio- PARKER, V.T., AND M.P. YODER-WILLIAMS. 1989. Reduction chemistry 8:209–213. of survival and growth of young Pinus jeffreyi by an JOHNSON, D.W. 1995. Soil properties beneath Ceanothus herbaceous perennial, Wyethia mollis. American and pine stands in the eastern Sierra Nevada. Soil Midland Naturalist 121:105–111. Science Society of America Journal 59:918–924. PERRY, D.A., M.P. AMARANTHUS, J.G. BORCHERS, S.L. KARTESZ, J.T. 1988. Pages 1474–1475 in A flora of Nevada. BORCHERS, AND R.E. BRAINERD. 1989. Bootstrap- Doctoral dissertation, Volume 3, University of Nevada, ping in ecosystems. BioScience 39:230–237. Reno. PERRY, D.A., T. BELL, AND M.P. AMARANTHUS. 1992. My- KELSEY, R.G., AND R.L. EVERETT. 1995. Allelopathy. Pages corrhizal fungi in mixed-species forests and other 479–549 in D.J. Bedunah and R.E. Sosebee, editors, tales of positive feedback, redundancy, and stability. Wildland plants: physiological ecology and develop- Pages 151–179 in M.G.R. Cannell, D.C. Malcom, mental morphology. Society for Range Management, and P.A. Robertson, editors, The ecology of mixed- Denver, CO. species stands of trees. Special Publication 11, British KENNEDY, P.B., AND S.B. DOTEN. 1901. A preliminary report Ecological Society, London, U.K. on the summer ranges of western Nevada sheep. RADOSEVICH, S.R. 1984. Interference between greenleaf Bulletin 51, Agricultural Experiment Station, Nevada manzanita (Arctostaphylos patula) and ponderosa State University, Reno. pine (Pinus ponderosa). Pages 259–270 in M.L. LANINI, W.T., AND S.R. RADOSEVICH. 1986. Response of Dureya and G.N. Brown, editors, Seedling physiol- three conifer species to site preparation and shrub ogy and reforestation success. Martinus Nijhoff/Dr. control. Forest Science 32:61–77. W. Junk Publishers, Dordrecht, Netherlands. LEIBERG, J.B. 1902. Forest conditions in the northern RADOSEVICH, S.R., AND J.S. HOLT. 1984. Weed ecology: Sierra Nevada, California. U.S. Geological Survey implications for vegetation management. John Wiley Professional Paper 8, U.S. Government Printing and Sons, New York. Office, Washington, DC. RICE, E.L. 1984. Allelopathy. 2nd edition. Academic Press, LODHI, M.A.K., AND K.T. KILLINGBECK. 1980. Allelopathic New York. inhibition of nitrification and nitrifying bacteria in a RIEGEL, G.M. 1982. Forest habitat types of the South ponderosa pine (Pinus ponderosa Dougl.) commu- Warner Mountains, Modoc County, northeastern Cali- nity. American Journal of Botany 67:1423–1429. fornia. Master’s thesis, Humboldt State University, ______. 1982. Effects of pine produced chemicals on Arcata, CA. selected understory species in a Pinus ponderosa ROY, D.F.1983. Natural regeneration. Pages 87–102 in T.F. community. Journal of Chemical Ecology 8:275–283. Robson and R.B. Standiford, editors, Management of MCDONALD, P.M. 1983a. Weeds in conifer plantations of the eastside pine type in northeastern California, northeastern California—management implications. proceedings of a symposium. Northern California Pages 70–78 in T.F. Robson and R.B. Standiford, edi- Society of American Foresters, SAF 83-06, Arcata, CA. tors, Management of the eastside pine type in north- RUNDEL, P.W., D.J. PARSONS, AND D.T. GORDON. 1977. eastern California, proceedings of a symposium. Montane and subalpine vegetation of the Sierra Northern California Society of American Foresters, Nevada and the Cascade ranges. Pages 559–599 in SAF 83-06, Arcata, CA. M.G. Barbour and J. Major, editors, Terrestrial vege- ______. 1983b. Climate, history, and vegetation of the tation of California. John Wiley and Sons, New York. eastside pine type in California. Pages 1–16 in T.F. SAS INSTITUTE INC. 1990. SAS/STAT users’ guide. Version Robson and R.B. Standiford, editors, Management of 6, 4th edition. Volume 2. SAS Institute Inc., Cary, NC. the eastside pine type in northeastern California, SAWYER, J.O., AND T. K EELER-WOLF. 1995. A manual of proceedings of a symposium. Northern California California vegetation. California Native Plant Soci- Society of America Foresters, SAF 83–06, Arcata, CA. ety, Sacramento. MICHELSEN, A., I.K. SCHMIDT, S. JONASSON, J. DIGHTON, SHAINSKY, L.J., AND S.R. RADOSEVICH. 1986. Growth and H.E. JONES, AND T.V. C ALLAGHAN. 1995. Inhibition water relations of Pinus ponderosa seedlings in com- of growth, and effects on nutrient uptake of arctic petitive regimes with Arctostaphylos patula seed- graminoids by leaf extracts—allelopathy or lings. Journal of 23:957–966. competition between plants and microbes? Oecolo- SMITH, S. 1994. Ecological guide to eastside pine plant gia 103:407–418. associations, northeastern California: Modoc, Lassen, 150 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Klamath, Shasta-Trinity, Plumas, and Tahoe National WILLIAMS, M.P. 1995. Inhibition of conifer regeneration Forests. U.S. Department of Agriculture, Forest Ser- by an herbaceous perennial, Wyethia mollis, in the vice, Pacific Southwest Region R5-ECOL-TP-004, eastern Sierra Nevada, California. Doctoral disserta- San Francisco, CA. tion, University of Washington, Seattle. TILMAN, D., AND D. WEDIN. 1991. Dynamics of nitrogen YODER-WILLIAMS, M.P., AND V. T. P ARKER. 1987. Allelo- competition between successional grasses. Ecology pathic interference in the seedbed of Pinus jeffreyi 72:1038–1049. in the Sierra Nevada, California. Canadian Journal TINNIN, R.O., AND L.A. KIRKPATRICK. 1985. The allelo- of Forest Research 17:991–994. pathic influence of broadleaf trees and shrubs on YOUNG, J.A., AND R.A. EVANS. 1979. Arrowleaf balsamroot seedlings of Douglas-fir. Forest Science 31:945–952. and mules ear seed germination. Journal of Range USDA FOREST SERVICE. 1937. Range plant handbook. Management 32:71–74. U.S. Government Printing Office, Washington, DC. ZARCINAS, B., B. CARTWRIGHT, L.R. SPOUNCER. 1987. Nitric ______. 1988. Plumas National Forest soil resource inven- acid digestion and multi-element analysis of plant tory. Final draft. Quincy, CA. material by inductively coupled plasma spectrome- WADDELL, T.G., M.H. THOMASSON, M.W. MOORE, H.W. try. Communications in Soil Science and Plant WHITE, D. SWANSON-BEAN, M.E. GREEN, G.S.VAN Analysis 18: 131–146. HORN, AND H.M. FALES. 1982. Isoflavone, wax, and triterpene constituents of Wyethia mollis. Phyto- Received 28 May 1999 chemistry 21:1631–1633. Accepted 5 March 2001 WEBSTER, S.R., C.T. YOUNGBERG, AND A.G. WOLLUM. 1967. Fixation of nitrogen by bitterbrush (Purshia triden- tata [Pursh] D.C.). Nature 216:392–393. Western North American Naturalist 62(2), © 2002, pp. 151–159

MOVEMENTS AND HOME RANGES OF SAN JOAQUIN KIT FOXES (VULPES MACROTIS MUTICA) RELATIVE TO OIL-FIELD DEVELOPMENT

Bruce W. Zoellick1,2, Charles E. Harris1,3, Brian T. Kelly1,4, Thomas P. O’Farrell1,5, Thomas T. Kato1,6, and Marni E. Koopman7,8

ABSTRACT.—We examined the effect of oil-field development on movements and patterns of spatial use of San Joaquin kit foxes (Vulpes macrotis mutica) on the Naval Petroleum Reserves in California (NPRC) in the San Joaquin Valley. To do this, we compared movements and home ranges of kit foxes from June 1984 to September 1985 in areas developed for petroleum production (30% of native habitat lost to production facilities) and areas with little develop- ment (3%). Distances traveled nightly by kit foxes did not differ between levels of petroleum development or between sexes (P > 0.2). Mean length of nightly movements during breeding (14.6 km) was longer than during pup-rearing (10.7 ± 2 km) and pup-dispersal (9.4 km) periods (P = 0.01). Mean size of home ranges was 4.6 0.4 (sx–) km (n = 21) and did not differ between levels of petroleum development and sexes (P > 0.2). Overlap of home ranges of foxes from the same social group (78 ± 4.3%) was greater than that of same-sex foxes (35 ± 7.8%) and males and females of different social groups (32 ± 8.0%, P < 0.01). Overlap of home ranges did not differ between kit foxes inhabiting developed and unde- veloped areas (P > 0.4). Despite extensive overlap of home ranges, kit foxes on NPRC maintained relatively exclusive core areas, particularly adjacent foxes of the same sex. Future studies should examine which levels of habitat conversion impact spatial use of kit foxes.

Key words: San Joaquin kit fox, Vulpes macrotis mutica, home range, movements, spatial organization, oil-field devel- opment, San Joaquin Valley.

San Joaquin kit foxes (Vulpes macrotis in prey availability but may be controlled by mutica) inhabit the Naval Petroleum Reserves long-term levels of prey biomass. of California (NPRC), an area of intensive Spacing patterns of home ranges, however, petroleum exploration and production in the may be influenced by short-term changes in San Joaquin Valley. Since 1979 the effects of prey abundance (White and Ralls 1993). Harris oil-field development on kit foxes have been (1986) determined that oil-field development investigated to identify and minimize adverse did not affect prey abundance on NPRC, but impacts to kit foxes and to ensure their contin- development altered local distribution of prey. ued existence on the NPRC. Oil-field devel- In addition to replacing native habitat with opment may affect patterns of space use of kit production facilities, development on NPRC foxes, including overlap and size of home ranges affected habitat composition and likely prey and use of habitats. availability by increasing shrub cover along Size of home ranges is influenced by body roads and pipelines (Warrick and Cypher 1998). size and resulting energetic needs (McNab The purpose of this study was to examine 1963, Harestad and Bunnell 1979, Gittleman kit fox movements, home range size, and spa- and Harvey 1982), but home-range size can be tial patterns of home ranges in developed and affected by food availability (Harestad and undeveloped portions of NPRC. Specific ob- Bunnell 1979). Other studies indicate that size jectives were to (1) determine if the length of of home ranges of kit foxes (White and Ralls nightly movements of kit foxes differed be- 1993) and red foxes (Vulpes vulpes; Macdonald tween areas of petroleum development and 1981) is not influenced by temporary changes undeveloped areas; (2) examine the length of

1EG&G Energy Measurements, Inc., PO Box 127, Tupman, CA 93276. 2Present address: U.S. Bureau of Land Management, Lower Snake River District Office, 3948 Development Avenue, Boise, ID 83705. 3Present address: Idaho Department of Fish and Game, 600 S. Walnut Street, PO Box 25, Boise, ID 83707. 4Present address: U.S. Fish and Wildlife Service, PO Box 1306, Albuquerque, NM 87103-1306. 5Present address: 611 H Avenue, Boulder City, NV 89005. 6Present address: Environmental Protections Division, Lawrence Livermore National Laboratory, PO Box 808, Livermore, CA 94550. 7EASI Endangered Species Program, PO Box 178, Tupman, CA 93276. 8Present address: Department of Zoology and Physiology, University of Wyoming, Laramie, WY 82071.

151 152 WESTERN NORTH AMERICAN NATURALIST [Volume 62 movements during breeding, pup-rearing, and Bakersfield, California, are 37°C in July and pup-dispersal periods; and (3) examine the 14°C in January (National Oceanic and Atmos- effect of oil-field development on the size and pheric Administration 1995). Mean minimum spatial organization of home ranges of kit temperatures are 20°C in July and 3°C in Jan- foxes. uary. Mean annual precipitation is 12.5 cm, most (90%) occurring between October and STUDY AREA April.

Located approximately 42 km southwest of METHODS Bakersfield, California, the Naval Petroleum Reserves encompass the moderately steep We trapped San Joaquin kit foxes in wire- slopes of Elk Hills and Buena Vista Hills, mesh live-traps (38 × 38 × 107 cm) baited with which are low foothills of the Temblor Range canned mackerel. Most foxes were trapped that extend southeast into the San Joaquin Val- during semiannual trapping sessions in July ley (Fig. 1). The 2 foothills are separated by 1984 and December 1984, conducted to esti- Buena Vista Valley. Elevations range from 88 mate size of the kit fox population on NPRC to 473 m above sea level. We studied kit foxes (Harris et al. 1987). Additional foxes were on 2 areas of NPRC: an area extensively trapped at dens during April–June 1984 and altered by oil-field development located on 1985, when we attempted to radio-collar foxes the southwestern slopes of the Buena Vista from all social groups on the study areas. We Hills and adjacent valley lands (Fig. 1), and an weighed, sexed, ear-tagged, and radio-collared area primarily unaltered by oil-field develop- each fox. ment located in Buena Vista Valley and por- Radiotelemetry tions of the southwestern slopes of Elk Hills (Fig. 1). We monitored kit foxes from July 1984 to Thirty percent of the developed study area September 1985. By taking simultaneous bear- is altered by petroleum development (well ings from 1 of 4 pairs of fixed-tower receiving stations (Fig. 1), we triangulated the location pads, sumps, roads, pipelines, pipe storage of foxes at night, using the null-signal method. yards, and other facilities). The town of Taft The receiving stations were located on hilltops covers approximately 5.1 km2 adjacent to the and consisted of two 3-element directional an- developed study area. An average of 3% of tennas mounted 2 m apart on a 7.6-m-tall mast undeveloped study area is altered by oil-field and connected to a null junction box and radio development (pipelines and roads). Exceptions 2 receiver. We located foxes every 15 minutes are 2 sections (2.6 km each) located on the during night-long (sunset to sunrise) monitor- northeastern edge of the area, of which 37% ing sessions. We monitored 10–15 foxes dur- and 21% are altered by oil-field facilities (pri- ing a session. Foxes in undeveloped habitat marily pipe storage yards and well pads). War- were monitored 4–5 nights (generally 1 night/ rick and Cypher (1998) provide a detailed de- week) per month from June to September 1984 scription of oil-field development on NPRC. and December 1984 to May 1985. Foxes in Vegetation of the study areas is dominated developed habitat were monitored 4–5 nights by red brome (Bromus madritensis) and red- per month from December 1984 to September stemmed filaree (Erodium cicutarium), charac- 1985 (exceptions were 2 foxes in developed teristic of the Valley Grassland vegetation type habitat that were monitored during June– (Heady 1977). Desert saltbush (Atriplex poly- September 1984 and December 1984–January carpa) is the most common shrub and grows in 1985). dense stands along washes and in disturbed We also located foxes at dens during the day areas. Other common shrubs include spiny 1–3 times per week, using hand-held anten- saltbush (Atriplex spinifera), cheesebush (Hy- nas. We considered foxes to be paired or be- menoclea salsola), matchweed (Gutierrezia long to the same social group if they frequent- bracteata), and bladderpod (Isomeris arborea). ly shared dens. Foxes that were radio-collared Annual weather patterns consist of hot, dry and occupied adjacent areas, but never shared summers and mild, damp winters. Mean maxi- dens, were considered to belong to different mum temperatures 42 km east of NPRC in social groups. 2002] HOME RANGES OF KIT FOXES 153

Fig. 1. Location of study areas on Naval Petroleum Reserves in California (NPRC) and area map of the location of NPRC.

To reduce error, triangulated locations com- Movements posed of bearings intersecting at angles <30° ° We calculated the distance that a fox moved or >150 were eliminated from all analyses. during a night by summing straight-line lengths We also discarded all outlier locations (those between successive locations taken every 15 >2 km from the previous location) identified minutes from sunset to sunrise. When the from visual inspection of locations obtained length of time a fox was monitored was 1–2 every 15 minutes during night-long monitor- hours less than the length of night (because of ing sessions. Using methods of Lee et al. (1985), gaps in the collection of successive 15-minute we evaluated the accuracy of the radio-track- relocations), we estimated the distance the fox ing system. Bearings taken from towers to sur- moved during the night by dividing the length veyed points had standard deviations of 1.0°– moved (km) by the time monitored (hr) and 1.7°. We calculated average 95% confidence then multiplying by the number of hours from ellipse areas for fox locations using an average sunset to sunrise. When the length of night standard deviation of 1.5° and the size of the was >2 hours longer than duration of the study areas monitored from each pair of re- monitoring period (due to logistic, weather, or ceiving stations (White 1985). Locations of equipment problems), we combined move- foxes in the undeveloped study area had aver- ments (of the same fox) from 2 different sam- age 95% confidence ellipse areas ranging from ple periods, obtained within a 2-week period, 0.3 to 0.4 km2 (accurate within 330–358 m). to estimate the length of nightly movement. Locations of foxes on the developed study area Mean lengths of nightly movements estimated had average 95% confidence ellipse areas from combined samples (n = 14) did not differ ranging from 0.2–0.3 km2 (accurate within from means estimated from single-night sam- 269–329 m). ples (n = 17) for either males or females during 154 WESTERN NORTH AMERICAN NATURALIST [Volume 62 the pup-dispersal period (t tests, P > 0.10). areas of non-use. We determined the number Lengths of 3 nightly movements were esti- of nightly sample periods or locations needed mated using distances moved by a fox during to adequately sample home range with area portions of 3 different sample periods. observation curves (Odum and Kuenzler 1955). We determined the length of nightly move- We examined the percent change in the size of ments during 3 biological periods: breeding MCP home ranges that occurred with the (December–mid-February), pup-rearing (mid- addition of locations collected during full- February–May), and pup-dispersal (June–Sep- night sample periods. By the 5th nightly sam- tember; Zoellick et al. 1987). We examined the ple, area observation curves increased <5%. length of nightly movements of foxes relative Therefore, we assumed that sequential loca- to level of development for pup-dispersal peri- tions from 4 to 5 nightly samples (172–215 ods of 1984 and 1985. We compared average locations) were sufficient to estimate size of length of nightly movements between level of home ranges. development, sex, and the interaction between Home ranges were calculated using all level of development and sex using analysis of locations collected within a 1-year period (gen- variance (ANOVA). In addition, for foxes in- erally November 1984–September 1985 for habiting developed areas, we compared length foxes in developed areas and June 1984–May of nightly movements among biological peri- 1985 for foxes in undeveloped areas). Esti- ods, between sexes, and the interaction be- mates of home range sizes were based on a tween biological period and sex using ANOVA. mean of 409 locations/fox (n = 9, range = To avoid pseudoreplication, lengths of nightly 271–820) in undeveloped areas and 550 loca- movements were treated as subsamples. Aver- tions/fox (n = 12, range = 165–1109) in devel- age distances traveled nightly by individual oped areas. foxes (for a given biological period and level of Using 2-way ANOVA, we compared sizes development) were used in ANOVAs examin- of home ranges and core areas between foxes ing the length of nightly movements. Repeated- inhabiting undeveloped and developed areas measures ANOVA was not used to compare and between sexes. Percent overlap of core length of movements among biological periods areas and home ranges was calculated for because most (8 of 10) foxes were present dur- foxes of the same social group and for foxes of ing only 1 or 2 of the 3 periods sampled. The different, but adjacent, social groups using the Tukey-Kramer HSD test was used to examine method of Macdonald et al. (1980). Percent differences in pairs of means. overlap of home ranges and core areas was arcsine transformed. We used t tests to exam- Home Ranges ine differences in overlap of core areas and We used program HOME RANGE (Acker- home ranges of foxes inhabiting undeveloped man et al. 1990) to estimate size of home and developed areas. ranges using the 100% minimum convex poly- gon (MCP) method (Hayne 1949), and to esti- Oil-field Development mate size of core home range areas using the We estimated the percentage of land area harmonic mean method (Dixon and Chapman affected by oil-field development (e.g., well 1980). Core areas were defined as the area pads, sumps, roads, pipelines, pipe storage within the 50% isopleth of the harmonic mean yards, production facilities) for each quarter (Spencer and Barrett 1984, White and Ralls section (65 ha) of NPRC by overlaying trans- 1993). For harmonic mean calculations, we parent dot grids (Mosby 1980) on 1:10,000 scale used a scale of 609.6 and a grid size of 72 × 32 aerial photographs taken in 1983. We consid- (Gallerani Lawson and Rodgers 1997). We also ered kit foxes to occupy developed habitat if estimated size of home ranges using the grid- their home ranges had >15% land alteration cell method (Rongstad and Tester 1969, Voight from oil-field development. and Tinline 1980, Laundre and Keller 1981) 2 from the number of 0.04-km (200 × 200-m) RESULTS grid cells (Zoellick and Smith 1992) entered by a fox. We calculated grid-cell home ranges Between June 1984 and May 1985, we to examine whether convex polygons overesti- studied 13 adult kit foxes (2 pairs, 2 females mate home-range size by including substantial cooperatively rearing pups, and 7 foxes from 2002] HOME RANGES OF KIT FOXES 155

± different social groups) on the undeveloped during the breeding period (14.6 1.1 [sx–] km, study area. Nine kit foxes (4 males, 5 females) n = 6) was greater than during pup-rearing were monitored from June to September 1984, (10.7 ± 1.0 km, n = 7) and pup-dispersal peri- while 4 kit foxes (3 males, 1 female) were mon- ods (9.4 ± 1.1 km, n = 6; P < 0.05). Lengths of itored from June 1984 to May 1985. Two kit nightly movements did not differ between foxes (1 male, 1 female) were assigned to pup-rearing and pup-dispersal (P = 0.66). developed habitat because their home ranges Home Ranges included areas of oil-field development in the northeastern corner of the study area. During HOME-RANGE SIZE.— MCP home range December 1984 to September 1985, we stud- size of kit foxes averaged 4.6 ± 0.4 km2 (n = ied 10 adult kit foxes (3 males, 7 females; all 21) and did not differ between developed (4.8 from different social groups) on the developed ± 0.7 km2, n = 12) and undeveloped areas (4.3 ± 2 study area. Additional adult kit foxes were 0.5 km , n = 9; F1,17 = 0.84, P = 0.37) or radio-collared on the study areas (4 on the between sexes (F1,17 = 1.61, P = 0.22). Size of undeveloped and 12 on the developed area), MCP home ranges averaged 5.2 ± 0.9 km2 (n but they were located too infrequently to esti- = 9) for males and 4.2 ± 0.4 km2 (n = 12) for mate home ranges or length of movements. females. Home ranges calculated with the grid- cell method averaged 4.3 ± 0.3 km2 (n = 21) in Movements size and did not differ from MCP estimates We determined the length of 52 nightly (F1,40 = 0.34, P = 0.56). Size of grid-cell home movements of 11 kit foxes (6 males, 5 females) ranges averaged 4.7 ± 0.6 km2 (n = 9) for inhabiting undeveloped areas, and of 73 males and 4.0 ± 0.3 km2 for females (n = 12). nightly movements of 10 foxes (4 males, 6 Size of core areas averaged 1.2 ± 0.1 km2 (n females) in developed areas (Table 1). Lengths = 21) and did not differ between foxes inhab- of nightly movements of kit foxes were sam- iting developed (1.2 ± 0.2 km2, n = 12) and ± 2 pled primarily during pup-dispersal, particu- undeveloped areas (1.2 0.2 km , n = 9; F1,17 larly for foxes inhabiting undeveloped areas = 0.02, P = 0.9) or between sexes (F1,17 = (Table 1). Lengths of nightly movements did 0.49, P = 0.49). Size of core areas averaged 1.3 not differ between undeveloped and devel- ± 0.1 km2 (n = 9) for males and 1.1 ± 0.1 km2 oped areas (F1,12 = 1.52, P = 0.24) or be- for females (n = 12). tween male and female foxes (F1,12 = 1.7, P = SPATIAL ORGANIZATION.—Overlap of home 0.3), and interaction between development and ranges calculated with the MCP method for sex was not significant (F1,12 = 1.41, P = 0.26). male and female foxes of different social Lengths of nightly movements of foxes in groups averaged 37.9 ± 5.2% (n = 30) and did developed areas (Table 1) differed among bio- not differ between developed areas (40.2 ± ± logical periods (F2,14 = 6.39, P = 0.01), but 6.6%, n = 22) and undeveloped areas (31.6 not between sexes (F1,14 = 0.36, P = 0.56). 8.0%, n = 8; t28 = 0.9, P = 0.4). Home ranges The interaction between sex and biological of same-sex foxes from different social groups ± period was not significant (F2,14 = 0.66, P = overlapped an average of 33.4 3.9% (n = 26) 0.53). Average length of nightly movements and did not differ between developed (32.9 ±

TABLE 1. Length of nightly movements (km) of San Joaquin kit foxes in undeveloped and developed areas of the Naval Petroleum Reserves in California during pup-dispersal, breeding, and pup-rearing periods; number of foxes (nf); and number of nightly movements sampled (nn), 1984–85.

______Developed areas ______Undeveloped areas ______Males ______Females ______Males ______Females a – b – – – Period x sx– nf nn x sx– nf nn x sx– nf nn x sx– nf nn Pup-dispersal 8.0 1.7 2 12 10.7 1.1 5 22 10.9 1.3 4 27 10.8 1.1 5 20 Breeding 14.6 1.5 3 6 14.6 1.5 3 5 15.5 2.6 2 2 Pup-rearing 10.9 1.5 3 12 10.5 1.3 4 16 13.6 0.9 2 3 aPup-dispersal = June–September, breeding = December–mid-February, pup-rearing = mid-February–May. bAverage distance traveled nightly by an individual fox (for a given biological period and level of development) was the sample unit used for statistical analyses. 156 WESTERN NORTH AMERICAN NATURALIST [Volume 62

4.6%, n = 20) and undeveloped areas (35.0 ± 7.8%, n = 6; t24 = 0.2, P = 0.8). Overlap of home ranges of males and females and same- sex foxes from different social groups, calcu- lated with the MCP method, averaged 35.8 ± 3.3% (n = 56). Home ranges calculated with the grid-cell method for those foxes over- lapped an average of 32.7 ± 2.8%. No kit foxes belonging to the same social group were ade- quately monitored in developed areas to cal- culate overlap of their home ranges. Overlap of home ranges of kit foxes in undeveloped areas calculated with the MCP method differed among foxes belonging to the same social group, same-sex foxes, and males and females from different social groups (F2,17 = 11.89, P = 0.001). Average overlap of home ranges was greater (P < 0.01) for foxes belong- ing to the same social group (77.7 ± 4.3%, n = 6) than between same-sex foxes from different Fig. 2. Spatial arrangement of the home ranges (dashed social groups (35.0 ± 7.8%, n = 6) and males lines) and core areas (solid lines) of 7 female kit foxes in and females from different social groups (31.6 developed areas of the Naval Petroleum Reserves in Cali- ± fornia, 1984–85. Home ranges were plotted using the 8.0%, n = 8). Overlap did not differ between minimum convex polygon method, and core areas were same-sex foxes and males and females from plotted using the 50% isopleth of the harmonic mean different social groups (P = 0.9). Overlap data method. for foxes belonging to the same social group included 2 females who cooperatively raised a litter of pups in 1984 and 2 known mated pairs. Core areas were relatively exclusive for kit (Zoellick et al. 1989). Visits by males to dens of foxes of the same sex (Fig. 2). Core areas of other pairs of kit foxes increased the length of male and female foxes from different social nightly movements during the breeding period groups often overlapped, with overlaps aver- of kit foxes in western Arizona (Zoellick et al. aging 14.4 ± 4.3% (n = 30). Overlap of core 1989). Kit foxes on NPRC were primarily noc- areas did not differ between foxes inhabiting turnal (Morrell 1972). The longer night length developed (12.8 ± 5.4%, n = 22) and undevel- during the breeding period (December to ± oped areas (19.0 6.4%, n = 8; t28 = 1.284, P February) also probably contributed to the in- = 0.26). crease in length of movements by allowing kit foxes to be active for a longer period of time. DISCUSSION Length of movements did not differ be- tween kit foxes inhabiting developed and un- Movements developed areas of NPRC likely because oil- The lengths of nightly movements of kit field development did not impact the abun- foxes from this study (11.5 ± 0.5 km for all bio- dance of prey species. Lagomorphs were the logical periods, n = 38) were 18% shorter than primary prey of kit foxes on NPRC during those of kit foxes in western Arizona (Zoellick 1984–85 (Cypher et al. 2000). Densities of et al. 1989). Available prey biomass was sub- lagomorphs did not differ between developed stantially lower in Arizona than in the San and undeveloped study areas, averaging 103/ Joaquin Valley (Zoellick and Smith 1992), which km2 during 1984–85 (O’Farrell et al. 1987). likely contributed to the difference in length Kangaroo rats (Dipodomys spp.) were also of movements between the 2 areas. Length of frequent prey of kit foxes during 1984–85 (ca nightly movements of kit foxes from this study 20% occurrence in diet; Cypher et al. 2000). was longer during the breeding period, similar Small mammals were common on both the to that of male kit foxes in western Arizona developed and undeveloped study area in 2002] HOME RANGES OF KIT FOXES 157

1984 (capture probabilities of 25–35%; O’Far- (O’Farrell et al. 1987). Fox density in Buena rell et al. 1987, Scrivner et al. 1987), but they Vista Valley during summer and winter 1984 may have been less abundant in 1985 because was estimated to be 1 fox/1.3 km2 and 1 fox/1.8 precipitation was lower in 1984 (13.4 cm) than km2, respectively, from closed population in 1983 (25.3 cm). Cypher et al. (2000) reported models (Harris et al. 1987). A minimum fox small mammal abundance on NPRC was re- density of 1 fox/1.8 km2 was estimated for lated to the previous year’s precipitation be- summer 1985 in Buena Vista Valley (Harris et tween 1985 and 1995. Any year differences in al. 1987). small mammal abundance would have been Large overlaps of home ranges of kit foxes additive to the effect of oil-field development; from different social groups on NPRC (averag- most data from kit foxes in developed areas ing 36%) indicated entire home ranges were were collected in 1985 when small mammals not defended from other kit foxes, either in likely were less abundant. Yet, no differences undeveloped areas or areas of oil-field devel- in movements and home range use were ob- opment. Morrell (1972) observed family groups served between levels of oil-field development, hunting in the same area of Buena Vista Valley, similar to the findings of Spiegel and Bradbury but not at the same time. Although core areas (1992) for an oil field near NPRC. of foxes in this study were fairly exclusive for same-sex foxes from different social groups, Home Ranges they overlapped an average of 14% for males Home ranges of kit foxes in this study and and females from different social groups. Home in western Utah (3.1 km2; O’Neal et al. 1987) ranges of San Joaquin kit foxes in nearby Mid- were significantly smaller than those of San way Valley (Spiegel and Bradbury 1992) and Joaquin kit foxes on the Carrizo Plain in Cali- on the Carrizo Plain (White and Ralls 1993) fornia (11.6 km2; White and Ralls 1993) and in also overlapped somewhat, but both studies western Arizona (11.2 km2, P < 0.001; Zoel- found that core areas were fairly exclusive for lick and Smith 1992). As discussed by White kit foxes of different social groups. and Ralls (1993) and Zoellick and Smith (1992), Because kit foxes on NPRC did not defend home-range size differs depending on prey their entire home ranges from other social availability. Lagomorph densities and overall groups, more than 1 kit fox or pair of kit foxes prey biomass on NPRC were substantially could use a local food source or other re- greater than in western Arizona and on the sources. This may partly explain why conver- Carrizo Plain in California (Zoellick and Smith sion of 30% of the native habitat to oil-field 1992, White and Ralls 1993), and similar to facilities had little effect on movements and those in western Utah (O’Neal et al. 1987). size of home ranges of kit foxes. Additionally, Home ranges of kit foxes in nearby Midway White and Ralls (1993) found that the size of Valley in California (6.1 ± 0.45 km2, n = 26; home ranges of kit foxes on the Carrizo Plain Spiegel and Bradbury 1992) were similar to in California did not change during periods of those of kit foxes on NPRC. Midway Valley drought-induced alterations in prey abundance. had plant communities similar to those of Instead, overlap of home ranges of adjacent, NPRC and presumably similar prey availabil- same-sex foxes decreased. In this study over- ity. Morrell (1972) also studied kit foxes in lap of home ranges did not differ between lev- Buena Vista Valley and estimated size of home els of oil-field development. ranges to be 2.6–5.2 km2, similar to this study. Overlap of home ranges of foxes from dif- Size of home ranges can be influenced by ferent social groups is dependent on resource differences in density and social structure availability (White and Ralls 1993). In produc- (Jewell 1966, Brown and Orians 1970, Schoe- tive, diverse habitats such as those on NPRC, ner 1981). Kit fox density was similar between kit foxes decrease the size of their home ranges developed and undeveloped areas during the and maintain less exclusive home ranges than period of study. Fox density on the developed in habitats with lower prey availability. Where study area was estimated at 1 fox/1.0 km2 dur- prey biomass and densities of kit foxes were ing winter 1984 using a closed population relatively low (1 fox per 4.1–6.5 km2; Zoellick model and at 1 fox/1.8 km2 during summer and Smith 1992, White and Ralls 1993), over- 1985 from the minimum population size lap of home ranges of adjacent same-sex foxes 158 WESTERN NORTH AMERICAN NATURALIST [Volume 62 was 2–3 times less than on NPRC. Addition- and Range Experimental Station, University of Idaho, ally, White and Ralls (1993) found increased Moscow. BROWN, J.L., AND G.H. ORIANS. 1970. Spacing patterns in spacing among kit foxes during a drought- mobile animals. Annual Review of Ecology and Sys- related decrease in prey abundance. tematics 1:239–262. The significantly smaller home ranges on CYPHER, B.L., ET AL. 2000. of San NPRC, compared to the size of home ranges Joaquin kit foxes at the Naval Petroleum Reserves in California. Wildlife Monograph 145:1–43. in other portions of the geographic range of DIXON, K.R., AND J.A. CHAPMAN. 1980. Harmonic mean the kit fox, indicate that NPRC lands provide measure of animal activity areas. Ecology 61: some of the highest quality habitat remaining 1040–1044. GALLERANI LAWSON, E.J,. AND A.R. RODGERS. 1997. Dif- for kit foxes in the San Joaquin Valley. Conver- ferences in home-range size computed in commonly sion of 30% of the native habitat on developed used software programs. Wildlife Society Bulletin portions of NPRC to petroleum-production 25:721–729. facilities has not reached a threshold where GITTLEMAN, J.L., AND P.H. HARVEY. 1982. home- range size, metabolic needs and ecology. Behavioral lagomorph numbers (Harris 1986, Warrick and Ecology and Sociobiology 10:57–63. Cypher 1998) and movements and home ranges HARESTAD, A.S., AND F.L. BUNNELL. 1979. Home range of kit foxes are impacted by habitat loss. and body weight—a reevaluation. Ecology 60: Increased spacing among social groups may 389–402. HARRIS, C.E. 1986. Comparison of line transects and road be the initial response of kit foxes to declines surveys to monitor lagomorph populations on Naval in prey abundance caused by more extensive Petroleum Reserve #1. U.S. Department of Energy levels of habitat conversion. Future studies Topical Report EGG 10282-2098. National Technical should examine which levels of habitat loss Information Service, Springfield, VA. HARRIS, C.E., T.P. O’FARRELL, P.M. MCCUE, AND T.T. K ATO. due to oil-field development alter movements 1987. Capture-recapture estimation of San Joaquin and use of home ranges by kit foxes. kit fox population size on Naval Petroleum Reserve #1, Kern County, California. U.S. Department of Energy Topical Report EGG 10282-2149. National ACKNOWLEDGMENTS Technical Information Service, Springfield, VA. HAYNE, D.W. 1949. Calculation of size of home range. The assistance of W.H. Berry, K.L. Buch- Journal of Mammalogy 39:190–206. inger, D.L. Garner, J.W. Johnson, N.E. Math- HEADY, H.F. 1977. Valley grassland. Pages 491–514 in ews, E.M. Rymills, M.C. Spencer, J.R. Thomp- M.G. Barbour and J. Major, editors, Terrestrial vege- tation of California. John Wiley & Sons, New York. son, and K. Timmerman with fieldwork is JEWELL, P.A. 1966. The concept of home range in mammals. greatly appreciated. E. Cross helped with data Symposium of the Zoological Society of London analysis. J. Ando, R. Elliott, and M. Probert 18:85–109. wrote and assisted with computer programs. LAUNDRE, J.W., AND B.L. KELLER. 1981. Home range use by coyotes in Idaho. Animal Behaviour 29:449–461. We thank M. McCoy for GIS analyses and LEE, J.G., C. WHITE, R.A. GARROTT, R.M. BARTMANN, AND preparation of figures, and B.L. Cypher, F.F. A.W. ALLDREDGE. 1985. Assessing accuracy of a Knowlton, E.M. Gese, and E.C. Hellgren for radiotelemetry system for estimating animal loca- reviewing earlier drafts of this paper. Permis- tions. Journal of Wildlife Management 49:658–663. MACDONALD, D.W. 1981. Resource dispersion and social sion to handle kit foxes was granted by the organization of the red fox, Vulpes vulpes. Pages U.S. Fish and Wildlife Service (permits PRT 2- 918–949 in J.A. Chapman and D. Pursley, editors, 4573 and PRT 683011) and California Depart- Proceedings of the worldwide furbearer conference. ment of Fish and Game (memorandum of University of Maryland Press, Frostburg. MACDONALD, D.W., F.G. BALL, AND N.G. HOUGH. 1980. understanding with EG&G Energy Measure- Evaluation of home range size and configuration ments, Inc.). This study was funded by the using radio tracking data. Pages 405–424 in C.J. U.S. Department of Energy, Naval Petroleum Amlaner, Jr., and D.W. Macdonald, editors, A hand- book on biotelemetry and radiotracking. Pergamon Reserves in California, and Chevron U.S.A. Press, Oxford, England. Production Company through the DOE Neva- MCNAB, B.K. 1963. Bioenergetics and the determination da Operations Office (contract DE-AC08- of home range size. American Naturalist 97:133–140. 83NV10282). MORRELL, S. 1972. Life history of the San Joaquin kit fox. California Department of Fish and Game 58:162–174. MOSBY, H.S. 1980. Reconnaissance mapping and map use. LITERATURE CITED Pages 277–290 in S.D. Schemnitz, editor, Wildlife management techniques manual. Wildlife Society, ACKERMAN, B.B., F.L. LEBAN, M.D. SAMUEL, AND E.O. GAR- Bethesda, MD. TON. 1990. User’s manual for program HOME NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION. RANGE. Technical Report 15. Forestry, Wildlife, 1995. Local climatologic data, 1995 annual summary 2002] HOME RANGES OF KIT FOXES 159

with comparative data, Bakersfield, California. Nation- Kern County, California. Transactions of the Western al Climatic Data Center, Asheville, NC. Section of the Wildlife Society 28:83–92. ODUM, E.P., AND E.J. KUENZLER. 1955. Measurement of VOIGHT, D.R., AND R.R. TINLINE. 1980. Strategies for ana- territory and home range size in birds. Auk 72: lyzing radio-tracking data. Pages 387–404 in C.J. 128–137. Amlaner, Jr., and D.W. Macdonald, editors, A hand- O’FARRELL, T.P., G.D. WARRICK, N.E. MATHEWS, AND T.T. book on biotelemetry and radiotracking. Pergamon KATO. 1987. Report of endangered species studies Press, Oxford, England. on Naval Petroleum Reserve #2, Kern County, Cali- WARRICK, G.D., AND B.L. CYPHER. 1998. Factors affecting fornia. U.S. Department of Energy Topical Report the spatial distribution of San Joaquin kit foxes. Jour- EGG 10282-2189. National Technical Information nal of Wildlife Management 62:707–717. Service, Springfield, VA. WHITE, G.C. 1985. Optimal locations of towers for trian- O’NEAL, G.T., J.T. FLINDERS, AND W. P. C LARY. 1987. gulation studies using biotelemetry. Journal of Wild- of the Nevada kit fox (Vulpes life Management 49:190–196. macrotis nevadensis) on a managed desert range- WHITE, P.J., AND K. RALLS. 1993. Reproduction and spac- land. Pages 443–481 in H.H. Genoways, editor, Cur- ing patterns of kit foxes relative to changing prey rent mammalogy. Plenum Press, New York. availability. Journal of Wildlife Management 57: RONGSTAD, O.J., AND J.R. TESTER. 1969. Movements and 861–867. habitat use of white-tailed deer in Minnesota. Jour- ZOELLICK, B.W., T.P. O’FARRELL, P.M. MCCUE, C.E. HAR- nal of Wildlife Management 33:366–379. RIS, AND T.T. K ATO. 1987. Reproduction of the San SCHOENER, T.W. 1981. An empirically based estimate of Joaquin kit fox on Naval Petroleum Reserve #1, Elk home range. Theoretical Population Biology 20 Hills, California, 1980–85. U.S. Department of Energy :281–325. Topical Report EGG 10282-2144. National Technical SCRIVNER, J.H., T.P. O’FARRELL, T.T. K ATO, AND M.K. JOHN- Information Service, Springfield, VA. SON. 1987. Diet of the San Joaquin kit fox (Vulpes ZOELLICK, B.W., AND N.S. SMITH. 1992. Size and spatial macrotis mutica) on Naval Petroleum Reserve #1, organization of home ranges of kit foxes in Arizona. Kern County, California, 1980–1984. U.S. Depart- Journal of Mammalogy 73:83–88. ment of Energy Topical Report EGG 10282-2168. ZOELLICK, B.W., N.S. SMITH, AND R.S. HENRY. 1989. Habi- National Technical Information Service, Springfield, tat use and movements of desert kit foxes in western VA. Arizona. Journal of Wildlife Management 53:955–96. SPENCER, W.D., AND R.H. BARRETT. 1984. An evaluation of the harmonic mean measure for defining carnivore Received 2 February 2000 activity areas. Acta Zoologic Fennica 171:255–259. Accepted 12 February 2001 SPIEGEL, L.K., AND M. BRADBURY. 1992. Home range characteristics of the San Joaquin kit fox in western Western North American Naturalist 62(2), © 2002, pp. 160–169

HELMINTH PARASITES OF SEVEN ANURAN SPECIES FROM NORTHWESTERN MEXICO

Stephen R. Goldberg1,3 and Charles R. Bursey2

ABSTRACT.—One hundred eighty-three specimens representing 7 anuran species were examined for helminth para- sites: Bufo kelloggi, B. mazatlanensis, Leptodactylus melanonotus, Pachymedusa dacnicolor, Rana forreri, R. magnaocu- laris, Smilisca baudini. The following species were found: 8 species of Trematoda, Cephalogonimus americanus, Clinosto- mum attenuatum (larva), Glypthelmins poncedeleoni, G. quieta, Gorgoderina attenuata, Haematoloechus complexus, H. longiplexus, and Megalodiscus temperatus; 2 species of Cestoda, Cylindrotaenia americana and Nematotaenia dispar; 13 species of Nematoda, Aplectana incerta, A. itzocanensis, Cosmocerca podicipinus, Cosmocercella haberi, Cosmocercoides variabilis, Foleyellides striatus, Oswaldocruzia pipiens, Rhabdias americanus, R. ranae, Subulascaris falcaustriformis, Physaloptera sp. (larva), Physocephalus sp. (larva), and Spiroxys sp. (larva); and 1 species of Acanthocephala (cystacanth). Helminth was 6.4 ± 2.4 s. Thirty-five new host records are reported.

Key words: Anura, helminths, Trematoda, Cestoda, Nematoda, Acanthocephala, Bufo kelloggi, Bufo mazatlanensis, Leptodactylus melanonotus, Pachymedusa dacnicolor, Rana forreri, Rana magnaocularis, Smilisca baudini, Mexico.

Although some 198 anuran species occur in Jalisco; Pachymedusa dacnicolor, southern Mexico (Flores-Villela 1993), reports of Sonora to the Isthmus of Tehuantepec. Three helminth parasite infections exist for only 13 have Neotropical distribution: Leptodactylus species: Bufo marinus Linnaeus, 1758, B. mar- melanonotus, Sonora through Central America moreus Wiegmann, 1833, B. valliceps Wieg- to central Ecuador; Rana forreri, southern mann, 1833, Leptodactylus melanonotus (Hal- Sonora along the Pacific coast to northwestern lowell, 1861), Rana berlandieri Baird, 1854, R. Costa Rica; Smilisca baudini, southern Texas dunni Zweifel, 1957, R. forreri Boulenger, (USA) and southern Sonora to Costa Rica 1883, R. trilobata (Mocquard, 1899) = R. (Frost 1985). megapoda Taylor, 1942, R. montezumae Baird, 1854, R. neovolcanica Hillis and Frost, 1985, MATERIALS AND METHODS R. vaillanti Brocchi, 1877, Spea multiplicata (Cope, 1863) = Scaphiopus multiplicatus (Cope, One hundred eighty-three specimens col- 1863), and Smilisca baudini (Duméril and lected 1959–1987 from the herpetological col- Bibron, 1841) (Baker 1987, Pérez-Ponce de lections at the University of Arizona (UAZ) or León et al. 2000). The purpose of this paper is Arizona State University (ASU) were exam- to report for the first time helminths harbored ined. Number and mean snout-vent length by 4 anuran species: Bufo kelloggi Taylor, (SVL) in mm for each species examined are 1938, B. mazatlanensis Taylor, 1940, Pachyme- given in Table 1. Museum accession numbers dusa dacnicolor (Cope, 1864), and Rana mag- are listed in Appendix 1. With the exception of naocularis Frost and Bagnara, 1976. Additional Rana forreri and 8 specimens of R. magnaocu- helminth records for Leptodactylus melanono- laris from Sinaloa, the anurans examined were tus, Rana forreri, and Smilisca baudini are also from the Mexican state of Sonora (Fig. 1). All reported. Four of these anurans are endemic anurans were from region 2, “tierra seca extra- to Mexico (Flores-Villela 1993): Bufo kelloggi, tropical” of Flores-Villela (1993). north central Sonora through Sinaloa to All specimens were originally fixed in 10% Nayarit; B. mazatlanensis, northern Sonora to formalin and stored in 70% alcohol. The body Colima; Rana magnaocularis, east central cavity was opened by a longitudinal incision Sonora through Sinaloa and Nayarit to central from vent to throat, and the gastrointestinal

1Department of Biology, Whittier College, Whittier, CA 90608. 2Department of Biology, Pennsylvania State University, Shenango Campus, Sharon, PA 16146. 3Corresponding author.

160 2002] HELMINTH PARASITES OF SEVEN ANURANS 161

TABLE 1. Number, mean snout-vent length (SVL) and range in mm, comparison of SVL (Kruskal-Wallis test), number infected, and comparison of infection rate between female and male specimens (chi-square test) for 7 anuran species from northwestern Mexico. Kruskal-Wallis Number Chi-square Anuran species N SVL, range test infected (%) test Bufo kelloggi female, 14 38.9 ± 5.2, 30–41 8.96, P < 0.001 8 (57) 0.27, P > 0.05 male, 15 33.6 ± 2.9, 32–49 10 (66)

Bufo mazatlanensis female, 6 67.7 ± 8.0, 59–80 4.95, P < 0.05 6 (100) 0 male, 14 59.8 ± 5.1, 50–68 14 (100)

Leptodactylus melanonotus female, 10 34.3 ± 5.5, 24–42 2.72, P > 0.05 10 (100) 0 male, 20 31.6 ± 3.3, 27–39 20 (100)

Pachymedusa dacnicolor female, 11 77.4 ± 11.2, 57–92 7.57, P < 0.001 9 (82) 0.50, P > 0.05 male, 13 65.7 ± 2.9, 61–70 9 (69)

Rana forreri female, 22 101.1 ± 17.7, 72–130 5.79, P < 0.05 19 (86) 3.81, P > 0.05 male, 17 88.5 ± 11.1, 73–114 10 (59)

Rana magnaocularis female, 13 65.0 ± 7.8, 44–78 11.28, P < 0.001 13 (100) 1.04, P > 0.05 male, 13 51.3 ± 7.3, 40–67 12 (92)

Smilisca baudini female, 5 69.8 ± 7.7, 59–80 6.24, P < 0.05 3 (40) 0.02, P > 0.05 male, 14 59.1 ± 3.5, 50–65 9 (64)

tract was removed by cutting across the tana incerta Caballero, 1949, A. itzocanensis esophagus and rectum. The lungs, esophagus, Bravo-Hollis, 1943, Cosmocerca podicipinus stomach, small intestine, large intestine, body Baker and Vaucher, 1984, Cosmocercella haberi cavity, liver, and urinary bladder of each speci- Steiner, 1924, Cosmocercoides variabilis (Har- men were examined separately. Each helminth wood, 1930), Foleyellides striatus (Ochoterena was removed and placed on a microscope and Caballero, 1932), Oswaldocruzia pipiens slide in a drop of undiluted glycerol, a cover- Walton, 1929, Rhabdias americanus Baker, slip added, and the slide set aside until the 1978, R. ranae Walton, 1929, Subulascaris fal- helminth became transparent. Nematodes and caustriformis Freitas and Dobbin, 1957, acanthocephalans were identified from these Physaloptera sp. (larvae), Physocephalus sp. slides. Trematodes and cestodes were stained (larvae in cysts), and Spiroxys sp. (larvae in with hematoxylin and mounted in balsam for cysts); and 1 species of Acanthocephala (cysta- further examination. Voucher helminths were canth larva in cyst). Although in all anuran deposited in the United States National Para- species except Leptodactylus melanonotus the site Collection (USNPC; Appendix 2). SVL of female specimens was significantly greater than in male specimens, there was no RESULTS difference in the infection rate between female and male hosts (Table 1). Because there We found the following species: 8 species was no difference in infection rates between of Trematoda, Cephalogonimus americanus female and male individuals of a host species, Stafford, 1902, Clinostomum attenuatum Cort, data for male and female individuals were 1913 (larvae in cysts), Glypthelmins pon- combined. Mean helminth species richness cedeleoni Razo-Mendivil and León-Règagnon, (summation of number of helminth species 2001, G. quieta (Stafford, 1900), Gorgoderina per particular host species divided by number attenuata (Stafford, 1902), Haematoloechus com- of host species) was 6.4 ± 2.4 s (range 4–10; plexus (Seely, 1906), H. longiplexus Stafford, Table 2). Prevalence (number of hosts infected 1902, Megalodiscus temperatus (Stafford, 1905); by a helminth species divided by the total 2 species of Cestoda, Cylindrotaenia ameri- number of hosts examined and expressed as a cana Jewell, 1916 and Nematotaenia dispar percentage), mean intensity (number of indi- (Goeze, 1782); 13 species of Nematoda, Aplec- viduals of a parasite species in a host species 162 WESTERN NORTH AMERICAN NATURALIST [Volume 62 23.9 ± s ± P I 10.7 — — 1.4 *58 26.8 ± ± s ± P I 46.2 — — — — 12.5 — — — — 0.9 —— —— ± ± ± s ± P I 22.8 — — *8 3.0 2.1 *31 1.7 ± ± s ± P I 0.7 —— —— —— —— 7.1 —— —— —— —— s ± ± ± P I 46.8 *7 2 *42 11.6 197 —— —— —— —— 53 *5 1.7 —— —— —— —— —— ± ± ± s ± P I ) and site of infection for helminths from 7 anuran species northwestern Mexico. 9.8 *25 59.6 s ± ± s (I ± s ± Bufo Bufo Leptodactylus Pachymedusa Rana Rana Smilisca kelloggi mazatlanensis melanonotus dacnicolor forreri magnaocularis baudini — — — — — — — — — — *12 8.7 — —— — — — — — — — 7 6.0 — — *31 34.3 — — — — — — — —*3 1— — *23 8.4 — —— — *40 3.5 *34 7.4 *7 4.5 *90 342 — ———— — — —— —— — — — —— *3 2 — — —— *3 2 — *3 2 — —— 2 — — — *4 *13 3.7 — —— — — — — — — — — — — 2. Prevalence (P), mean intensity small, large intestines small intestine cysts in muscle, skin small, large intestines large intestine small intestine small intestine small, large intestines small intestine urinary bladder lung lung ABLE Glypthelmins quieta Clinostomum attenuatum Glypthelmins poncedeleoni Nematotaenia dispar Aplectana itzocanensis T Cephalogonimus americanus Cylindrotaenia americana Aplectana incerta Gorgoderina attenuata Haematoloechus complexus Haematoloechus longiplexus Megalodiscus temperatus Site of infection P I EMATODA ESTODA REMATODA T C N Helminth ______2002] HELMINTH PARASITES OF SEVEN ANURANS 163 3.5 ± — — — — — — — — *11 3.5 — — — 0.8 — 11.7 — 0.7 — 6.4 5 3 ± ± ± ± — — — — *15 1*15 2.0 — — *85 9.7 *8 2.5 — — — 72.6 8 5.5 0.0 *4 1 — 1.2 —— —— ± ± ± — — — — — — — — — *8 1.0 — — — — — — 8 ——381 —— —— 2.2 15 35.2 ± ± — — — — — — — — — — — — 25 2.8 — 4 1 — 3.7 — 6.5 — — *13 2.0 1.9 — ± ± ± — — — — — — — — — *7 2— — 23 2.6 — 1.7 — 17.1 — ± ± — — — — — — — — *5 1— *17 3.4 — 40 9.6 — — 18.7 *5 2 — ± — — — — — — — — — — — *34 10.0 — — — —— — — — — — *97 7.6 — — *54 287 sp. (larva) — sp. (larva) 3 1 35 3.0 sp. (larva) — stomach large intestine under peritoneum stomach, small intestine lungs lungs small intestine cysts in peritoneum cysts in peritoneum cyst in peritoneum small, large intestines small, large intestines Physocephalus Cosmocercella haberi Cosmocerca podicipinus Cosmocercoides variabilis striatus Foleyellides Oswaldocruzia pipiens Rhabdias americanus Rhabdias ranae Subulascaris falcaustriformis Physaloptera Spiroxys oligacanthorhynchid cystacanth — *New host record CANTHOCEPHALA A 164 WESTERN NORTH AMERICAN NATURALIST [Volume 62

been reported from bufonids, ranids, and sala- manders; Glypthelmins quieta from bufonids, hylids, and ranids; Gorgoderina attenuata from bufonids, ranids, and salamanders; Haema- toloechus complexus from ranids; H. longi- plexus from bufonids and ranids; Megalodiscus temperatus from bufonids, hylids, ranids, and salamanders (Prudhoe and Bray 1982, Pérez- Ponce de León et al. 2000). Clinostomum attenuatum and Megalodiscus temperatus are additions to the list provided by Pérez-Ponce de León et al. (2000). Etges (1991) reported that Bufo marinus from northern Mexico (pre- cise location not given) harbor Clinostomum attenuatum. This is the 2nd report of C. atten- uatum from Mexico. Megalodiscus temperatus was originally reported as M. montezumae Travassos, 1934 from Rana montezumae col- lected in the state of Mexico (Bravo-Hollis 1941). Additional information on the life his- tory of digenean trematodes of Mexican amphibians is in Pérez-Ponce de León et al. (2000). Fig. 1. Approximate collection areas for 7 anuran species from northwestern Mexico: 1 = Bufo kelloggi, 2 = B. The cestode species found in this study mazatlanensis, 3 = Leptodactylus melanonotus, 4 = cannot be placed in the categories constructed Pachymedusa dacnicolor, 5 = Rana forreri, 6 = R. mag- above. Nematotaenia dispar is a variable species naocularis, 7 = Smilisca baudini. with a broad geographical range extending to all but the Australian and Ethiopian biogeo- graphical realms (Jones 1987). Host records divided by the number of infected hosts), suggest that N. dispar is primarily a parasite of range, and site of infection are given in Table bufonids, but hylids, ranids, salamanders, and 2. Thirty-five new host records are reported a varanid lizard have also been reported as (Table 2). hosts (Jones 1987). Cylindrotaenia americana is restricted to the Western Hemisphere, DISCUSSION where it is a parasite of bufonids, ranids, hylids, and leptodactylids (Jones 1987). This is Pérez-Ponce de León et al. (2000) have the 1st report of C. americana from Mexico. recently reviewed digenean fauna of some Nematotaenia dispar and C. americana are selected amphibians of central Mexico. They members of the Nematotaeniidae, which are distinguished 3 categories: helminths of anu- considered by Joyeux (1927) to have direct life rans that have Nearctic or Neotropical distri- cycles, that is, without intermediate hosts. butions and helminths of anurans endemic to Infection of a new host occurs through inges- Mexico. Of the trematode species found in tion of cestode eggs. Reports of cestodes in this study, all have been previously reported Mexican anurans are given in Table 3. from Mexico. All have Nearctic distributions The nematode species represented by with the exception of Glypthelmins ponce- mature individuals found in this study can be deleoni, currently known only from Mexico placed in the categories of Pérez-Ponce de (Razo-Mendivil and León-Règagnon 2001) León et al. (2000). The Nearctic species and considered a Mexican endemic by Pérez- include Aplectana incerta, A. itzocanensis, Ponce de León et al. (2000). All of these Cosmocercella haberi, Cosmocercoides vari- trematodes are generalists with the exception abilis, Oswaldocruzia pipiens, Rhabdias ameri- of Clinostomum attenuatum, a bird parasite canus, and R. ranae. Neotropical species (bitterns) that utilizes frogs as transport hosts include Cosmocerca podicipinus and Subulas- (Schell 1985). Cephalogonimus americanus has caris falcaustriformis. Foleyellides striatus is 2002] HELMINTH PARASITES OF SEVEN ANURANS 165

TABLE 3. Cestoda and Nematoda of species of Mexican anurans. Helminth Locality (State) Reference

CESTODA Cylindrotaenia americana Jewell, 1916 Leptodactylus melanonotus Sonora this study Nematotaenia dispar (Goeze, 1782) Lühe, 1899 Bufo marinus Nuevo León Martínez 1969 Bufo mazatlanensis Sonora this study Ophiotaenia magna Hannum, 1925 Rana tarahumarae Sonora Bursey and Goldberg 2001

NEMATODA Aplectana incerta Caballero, 1949 Bufo marinus Puebla Bravo-Hollis 1943 Chiapas Caballero 1949, 1954 Bufo marmoreus Jalisco Galicia-Guerrero et al. 2000 Bufo mazatlanensis Sonora this study Smilisca baudini Sonora this study Aplectana itzocanensis Bravo-Hollis, 1943 = Aplectana hoffmani Bravo-Hollis, 1943 Bufo kelloggi Sonora this study Bufo marinus Veracruz Caballero Deloya 1974 Bufo mazatlanensis Sonora this study Leptodactylus melanonotus Sonora this study Pachymedusa dacnicolor Sonora this study Rana magnaocularis Sonora this study Spea multiplicata Puebla Bravo-Hollis 1943 Smilisca baudini Sonora this study Aplectana sp. Bufo marinus Veracruz Guillén-Hernández 1992 Cosmocerca podicipinus Baker and Vaucher, 1984 Leptodactylus melanonotus Sonora this study Rana forreri Sinaloa this study Cosmocerca sp. Bufo marinus Veracruz Guillén-Hernández 1992 Cosmocercella haberi Steiner, 1924 Pachymedusa dacnicolor Sonora this study Cosmocercoides variabilis (Harwood, 1930) Travassos, 1931 Rana magnaocularis Sonora this study Cruzia morleyi (Pearse, 1936) Chabaud, 1978 Bufo marinus Yucatan Pearse 1936 Veracruz Caballero Deloya 1974 Falcaustra caballeroi Chabaud and Golvan, 1957 = Dibulgiber longispiculis Caballero, 1935 Rana montezumae Mexico Caballero 1935 Falcaustra inglisi (Anderson, 1964) Baker, 1980 Rana tarahumarae Sonora Bursey and Goldberg 2001 Falcaustra lowei Bursey and Goldberg, 2001 Rana tarahumarae Sonora Bursey and Goldberg 2001 Foleyellides striatus Caballero, 1935 = Chandlerella striata Ochoterena and Caballero, 1932 Rana forreri Sinaloa this study Rana magnaocularis Sonora this study Rana montezumae Mexico Caballero 1935 Rana tarahumarae Sonora Bursey and Goldberg 2001 Ochoterenella caballeroi Esslinger, 1987 Bufo marinus Chiapas Esslinger 1987 Ochoterenella chiapensis Esslinger, 1988 Bufo marinus Chiapas Esslinger 1988a Ochoterenella digiticauda Caballero, 1944 Bufo marinus Oaxaca Caballero 1944 Chiapas Caballero 1948 Chiapas Esslinger 1986 Jalisco Galicia-Guerrero et al. 2000 166 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 3. Continued. Helminth Locality (State) Reference

NEMATODA (continued) Ochoterenella figueroai Esslinger, 1988 Bufo marinus Chiapas Esslinger 1988b Ochoterenella lamothei Esslinger, 1988 Bufo marinus Chiapas Esslinger 1988b Ochoterenella nanolarvata Esslinger, 1987 Bufo marinus Chiapas Esslinger 1987 Ochoterenella sp. Bufo marinus Veracruz Guillén-Hernández 1992 Oswaldocruzia pipiens Walton, 1929 Bufo marinus Nuevo León Martínez 1969 Bufo mazatlanensis Sonora this paper Leptodactylus melanonotus Sonora this paper Rana magnaocularis Sonora this paper Rana tarahumarae Sonora Bursey and Goldberg 2001 Oswaldocruzia subauricularis (Rudolphi, 1819) Travassos, 1917 Bufo marinus Chiapas Caballero 1949, 1954 Oswaldocruzia sp. Bufo marinus Veracruz Guillén-Hernández 1992 Physaloptera sp. (larva) Bufo kelloggi Sonora this study Bufo marinus Jalisco Galicia-Guerrero et al. 2000 Bufo mazatlanensis Sonora this study Pachymedusa dacnicolor Sonora this study Rana forreri Sinaloa this study Rana magnaocularis Sonora this study Rana tarahumarae Sonora Bursey and Goldberg 2001 Smilisca baudini Sonora this study Physocephalus sp. (larva) Bufo marinus Jalisco Galicia-Guerrero et al. 2000 Bufo mazatlanensis Sonora this study Rhabdias americanus Baker, 1978 Bufo kelloggi Sonora this paper Bufo mazatlanensis Sonora this paper Smilisca baudini Sonora this paper Rhabdias fuelleborni Travassos, 1926 Bufo marinus Jalisco Galicia-Guerrero et al. 2000 Bufo marmoreus Jalisco Galicia-Guerrero et al. 2000 Rhabdias ranae Walton, 1929 Leptodactylus melanonotus Sonora this study Rana magnaocularis Sonora this study Rana tarahumarae Sonora Bursey and Goldberg 2001 Rhabdias sphaerocephala Goodey, 1924* Bufo marinus Veracruz Bravo-Hollis and Caballero 1940 Chiapas Caballero 1949, 1954 Nuevo León Martínez 1969 Veracruz Caballero Deloya 1974 Veracruz Guillén-Hernández 1992 Spiroxys corti Caballero, 1935 Rana montezumae Mexico Caballero 1935 Spiroxys sp. (larva) Leptodactylus melanonotus Sonora this study Subulascaris falcaustriformis Freitas and Dobbin, 1957 Rana magnaocularis Sonora this study Rana tarahumarae Sonora Bursey and Goldberg 2001

*Considered to be a Palearctic species only (Baker 1987). 2002] HELMINTH PARASITES OF SEVEN ANURANS 167 endemic to Mexico. Previous records of these et al. 1991). Elkins and Nickol (1983) and nematodes from Mexican anurans are given in Bolette (1997) consider reptiles in these Table 3. instances to be paratenic hosts. For those The above nematodes are generalists: acanthocephalans parasitic in terrestrial ani- Aplectana incerta is a common parasite of mals, the intermediate hosts are usually bufonids; A. itzocanensis of bufonids and a insects (Nickol 1985); thus, their occurrence in pelobatid; Cosmocercella haberi of hylids and could be anticipated. Because ranids; Cosmocercoides variabilis of frogs, these larvae occur in cysts, the possibility of toads, salamanders, lizards, turtles, and snakes; frogs as paratenic hosts must be considered. Oswaldocruzia pipiens of frogs, toads, a sala- Our knowledge of anuran helminths of mander, lizards, and turtles; Rhabdias ameri- Mexico is far too incomplete to make general- canus of bufonids; R. ranae of hylids and ranids izations about distribution patterns. But, as (Baker 1987). Cosmocerca podicipinus has been more hosts are studied, we believe they will reported from South American leptodactylids be shown to harbor generalist helminths and and Subulascaris falcaustriformis from a these generalists will fall into 1 of the 3 cate- Brazilian ranid (Baker 1987). Foleyellides gories erected by Pérez-Ponce de León et al. striatus is currently known only from Mexican (2000), i.e., Nearctic, Neotropical, endemic. ranids (Baker 1987). Aplectana, Cosmocerca, Cosmocercella, Cosmocercoides, Oswaldocru- ACKNOWLEDGMENTS zia, and Rhabdias have direct life cycles; Foleyellides are transmitted to new hosts by We thank Michael E. Douglas (Arizona haematophagous arthropods; and the life cycle State University) and Charles H. Lowe (Uni- of Subulascaris is not known (Anderson 2000). versity of Arizona) for permission to examine Adults of Physaloptera sp. occur in the Mexican anurans, and Irene S. Tan, Cynthia stomachs of mammals, snakes, and a few lizard Walser, and Cheryl Wong for assistance with species; larvae are common in amphibians and dissections. The figure was drawn by Peggy some lizard species (Anderson 2000). Adults of Firth. Physocephalus occur in the stomachs of swine, horses, cattle, and rabbits; infective larvae LITERATURE CITED have been recovered from dung beetles and ANDERSON, R.C. 2000. Nematode parasites of vertebrates. are found encapsulated in the tissues of am- Their development and transmission. 2nd edition. phibians, reptiles, birds, and mammals (Ander- CABI Publishing, CAB International, Wallingford, son 2000). Adults of Spiroxys sp. are common Oxon, U.K. 650 pp. parasites of the gastric mucosa of turtles in BAKER, M.R. 1987. Synopsis of the Nematoda parasitic in amphibians and reptiles. Memorial University of North America; larvae are found encapsulated Newfoundland, Occasional Papers in Biology, St. in the tissues of snails, fish, frogs, and newts John’s Newfoundland, Canada 11:1–325. (Anderson 2000). The life cycle is similar for BOLETTE, D.P. 1997. First record of Pachysentis canicola species in these 3 genera: embryonated eggs (Acanthocephala: Oligacanthorhynchida) and the occurrence of Mesocestoides sp. tetrathyridia (Ces- are passed with the feces of the definitive host toidea: Cyclophyllidea) in the western diamondback and hatch when eaten by intermediate hosts, rattlesnake, Crotalus atrox (Serpentes: Viperidae). various species of insects; infection is acquired Journal of Parasitology 83:751–752. by ingesting insects containing infective third- BRAVO-HOLLIS, M. 1941. Revisión de los géneros Diplo- stage larvae (Anderson 2000). Because larvae discus Diesing, 1836 y Megalodiscus Chandler, 1923 (Trematoda: Paramphistomoidea). II. Anales del Insti- of Physocephalus sp. and Spiroxys sp. were tuto de Biología, Universidad Nacional Autónoma de found in cysts and are not known to mature in México 12:643–661. anurans, the possibility of anurans as paratenic ______. 1943. Dos nuevos nemátodos parásitos de anuros hosts must be considered. Since larvae of del Sur de Puebla. Anales del Instituto de Biología, Physaloptera sp. were found in the digestive Universidad Nacional Autónoma de México 14:69–78. BRAVO-HOLLIS, M., AND E. CABALLERO C. 1940. Nemáto- tract and do not mature in anurans, we con- dos parásitos de los batracios de México. IV. Anales sider them to be pseudoparasites, a by-product del Instituto de Biología, Universidad Nacional of diet. Autónoma de México 11:239–247. Acanthocephalan cystacanths have been BURSEY, C.R., AND S.R. GOLDBERG. 2001. Falcaustra lowei n. sp. and other helminths from the Tarahumara frog, found in amphibians (Moore 1946), reptiles Rana tarahumarae (Anura: Ranidae), from Sonora, (Goldberg et al. 1998), and mammals (Radomski Mexico. Journal of Parasitology 87:340–344. 168 WESTERN NORTH AMERICAN NATURALIST [Volume 62

CABALLERO, C.E. 1935. Nemátodos parásitos del los batra- GALICIA-GUERRERO, S., C.R. BURSEY, S.R. GOLDBERG, AND cios de México. III. Cuarta contribucíon al G. SALGADO-MALDONADO. 2000. Helminths of two conocimiento de la parasitología de Rana montezu- sympatric toad species, Bufo marinus (Linnaeus) and mae. Anales del Instituto de Biología, Universidad Bufo marmoreus Wiegmann, 1833 (Anura: Bufo- Nacional Autónoma de México 6:103–117. nidae) from Chamela, Jalisco, Mexico. Comparative ______. 1944. Estudios helmintológicos de la región onco- Parasitology 67:129–133. cercosa de México y de la República de Guatemala. GOLDBERG, S.R., C.R. BURSEY, AND H.J. HOLSHUH. 1998. Nematoda, 1a parte. Filarioidea 1. Anales del Insti- Prevalence and distribution of cystacanths of an oli- tuto de Biología, Universidad Nacional Autónoma de gacanthorhynchid acanthocephalan from the long- México 15:87–108. nose snake, Rhinocheilus lecontei (Colubridae), in ______. 1948. Etudes helminthologiques sur la vallée du southwestern North America. Journal of the Hel- Rio Papaloapan (Mexique) II. Quelques filaires de minthological Society of Washington 65:262–265. batraciens et d’oiseaux. Annales de Parasitologie GUILLÉN-HERNÁNDEZ, S. 1992. Comunidades de helmintos Humaine et Comparée 23:323–333. de algunos anuros de “Los Tuxtlas,” Veracruz. Tesis ______. 1949. Estudios helmintologicos de la region onco- de Maestría, Facultad de Ciencias, Universidad Na- cercosa de México y de la República de Guatemala. cional Autónoma de México, D.F. 90 pp. Nematoda, 5a parte. Anales del Instituto de Biología, JONES, M.K. 1987. A taxonomic revision of the Nemato- Universidad Nacional Autónoma de México 20: taeniidae Lühe, 1910 (Cestoda: Cyclophyllidea). Sys- 279–292. tematic Parasitology 10:165–145. ______. 1954. Estudios helminthológicos de la región JOYEUX, C. 1927. Recherches sur la faune helmintholo- oncocercosa de México y de la República de gique Algérienne (cestodes et trématodes). Archives Guatemala. Nematoda, 8a parte. Anales del Instituto de l’Institut Pasteur d’Algerie 5:509–528. de Biología, Universidad Nacional Autónoma de MARTÍNEZ, V.J.M. 1969. Parásitos de algunos anfibios México 25:259–274. colectados en diferentes áreas de los Municipios de CABALLERO DELOYA, J. 1974. Estudio helmintológico de Escobedo, Pesquería y Santiago, Nuevo León, Méx- los animales silvestres de la Estación de Biología ico. Tesis Licenciatura, Facultad de Ciencias Biológi- Tropical “Los Tuxtlas,” Veracruz. Nematoda. I. Algunos cas, Universidad Autónoma de Nuevo León, Mon- nemátodos parásitos de Bufo horribilis Wiegmann, terrey. 51 pp. 1833. Anales del Instituto de Biología, Universidad MOORE, D.V. 1946. Studies on the life history and devel- Nacional Autónoma de México 45: 45–50. opment of Macracanthorhynchus ingens Meyer, 1933, ELKINS, C.A., AND B.B. NICKOL. 1983. The epizootiology with a redescription of the adult worm. Journal of of Macracanthorhynchus ingens in Louisiana. Jour- Parasitology 32:387–399. nal of Parasitology 69:951–956. NICKOL, B.B. 1985. Epizootiology. Pages 307–346 in D.W.T. ESSLINGER, J.H. 1986. Redescription of Ochoterenella Crompton and B.B. Nickol, editors, Biology of the digiticauda Caballero, 1944 (Nematoda: Filarioidea) Acanthocephala. Cambridge University Press, Cam- from the toad, Bufo marinus, with a redefinition of bridge, U.K. the genus Ochoterenella Caballero, 1944. Proceed- PEARSE, A.S. 1936. Parasites from Yucatan. Carnegie Insti- ings of the Helminthological Society of Washington tute of Washington, Publication 457:45–59. 53:210–217. PÉREZ-PONCE DE LEÓN, G., V. LEÓN-RÉGAGNON, L. GAR- ______. 1987. Ochoterenella caballeroi sp. n. and O. nano- CÍA-PRIETO, U. RAZO-MENDIVIL, AND A. SÁNCHEZ- larvata sp. n. (Nematoda: Filarioidea) from the toad ALVAREZ. 2000. Digenean fauna of amphibians from Bufo marinus. Proceedings of the Helminthological central Mexico: Nearctic and Neotropical influences. Society of Washington 54:126–132. Comparative Parasitology 67:92–106. ______. 1988a. Ochoterenella chiapensis n. sp. (Nematoda: PRUDHOE, S., AND R.A. BRAY. 1982. Platyhelminth para- Filarioidea) from the toad Bufo marinus in Mexico sites of the Amphibia. British Museum (Natural His- and Guatemala. Transactions of the American Micro- tory), Oxford University Press, London, U.K. 217 pp scopical Society 107:203–208. + 4 pp microfiche. ______. 1988b. Ochoterenella figueroai sp. n. and O. lamo- RADOMSKI, A.A., D.A. OSBORN, D.B. PENCE, M.I. NELSON, thei sp. n. (Nematoda: Filarioidea) from the toad AND R.J. WARREN. 1991. Visceral helminths from an Bufo marinus. Proceedings of the Helminthological expanding insular population of the long-nosed Society of Washington 55:146–154. armadillo (Dasypus novemcinctus). Journal of the ETGES, F.J. 1991. Clinostomum attenuatum (Digenea) from Helminthological Society of Washington 58:1–6. the eye of Bufo marinus. Journal of Parasitology RAZO-MENDIVIL, U.J., AND V. L EÓN-RÈGAGNON. 2001. 77:634–635. Glypthelmins poncedeleoni n. sp. (Trematoda: Macro- FLORES-VILLELA, O. 1993. Herpetofauna Mexicana: anno- deroididae) of amphibians from the Neotropical tated list of the species of amphibians and reptiles of region of México. Journal of Parasitology 87:686–691. Mexico, recent taxonomic changes, and new species. SCHELL, S.C. 1985. Handbook of trematodes of North Carnegie Museum of Natural History, Pittsburgh, America north of Mexico. University Press of Idaho, Special Publication 17:1–73. Moscow. 263 pp. FROST, D.R., EDITOR. 1985. Amphibian species of the world: a taxonomic and geographical reference. Received 12 April 2000 Allen Press, Inc., and the Association of Systematics Accepted 21 November 2000 Collections, Lawrence, KS. 732 pp. 2002] HELMINTH PARASITES OF SEVEN ANURANS 169

APPENDIX 1 APPENDIX 2 ZOOLOGY MUSEUM ACCESSION NUMBERS PARASITOLOGY MUSEUM ACCESSION NUMBERS Anurans examined from the herpetology collections of Arizona State University (ASU) or the University of Ari- Helminths deposited in the United States National zona (UAZ). Parasite Collection (USNPC). Bufo kelloggi (N = 29): Sonora UAZ 4457–4459, Bufo kelloggi: Cylindrotaenia americana 89794; 10192, 10289, 10290, 10293–10295, 11258, 11313, 11314, Aplectana itzocanensis 89795; Rhabdias americanus 89796; 23437, 31553, 31554, 31664, 31668–31670, 31674, 31686, Physaloptera sp. (larvae) 89797. 32036, 32037, 32040, 32043, 32044, 42822, 45212, 47068. Bufo mazatlanensis: Nematotaenia dispar 89798; Bufo mazatlanensis (N = 20): Sonora ASU 6109, 6111– Aplectana incerta 89799; Aplectana itzocanensis 89800; 6113, 6115–6117, 6119–6121, 6123, 6126, 6128–6134, Oswaldocruzia pipiens 89801; Rhabdias americanus 89802; 6194. Physaloptera sp. (larvae) 89803; Physocephalus sp. (larvae) Leptodactylus melanonotus (N = 30): Sonora UAZ 89804. 8215, 8216, 8219, 8223, 8225, 8229, 11301, 11302, 33324, Leptodactylus melanonotus: Glypthelmins poncedeleoni 33351, 36389, 38733, 39449, 39540, 42833, 42835, 44815, 89805; Gorgoderina attenuata 89806; Haematoloechus 45228–45230, 45232–45236, 45966, 50886, 50888–50890. longiplexus 89807; Megalodiscus temperatus 89808; Cylin- Pachymedusa dacnicolor (N = 24): Sonora UAZ 12877, drotaenia americana 89809; Aplectana itzocanensis 89810; 31597, 31717, 39250, 39251, 39451, 39452, 44787, 44788, Cosmocerca podicipinus 89811; Oswaldocruzia pipiens 45217, 45220, 45222, 45226, 45915, 45916, 45962, 46558, 89812; Rhabdias ranae 89813; Spiroxys sp. (larvae) 89814. 46688, 46689, 47423, 47493, 47777, 47778, 51577. Pachymedusa dacnicolor: Megalodisus temperatus Rana forreri (N = 39): Sinaloa UAZ 49726–49758, 89815; Aplectana itzocanensis 89816; Cosmocercella 49760–49765. haberi 89817; Physaloptera sp. (larvae) 89818; oligacan- Rana magnaocularis (N = 26): Sinaloa UAZ 51035, thorhynchid acanthocephalan cystacanth 89819. 51037–51043; Sonora UAZ 44740, 45103, 45941, 45949, Rana forreri: Cephalogonimus americanus 90728, 90729; 51031–51033, 51044–51047, 51049, 51051, 51053–51056, Clinostomum attenuatum 90730, 90731; Megalodiscus 51058. temperatus 90732, 90733; Cosmocerca podicipinus 90734; Smilisca baudini (N = 19): Sonora UAZ 31416, 31550, Foleyellides striatus 90735; Physaloptera sp. 90736; 31659–31661, 31840, 31998, 32020, 39247–39249, 45214, Rana magnaocularis: Glypthelmins quieta 89820; 45242–45244, 46691, 47424, 47491, 47492. Haematoloechus complexus 89821; Aplectana itzocanensis 89822; Cosmocercoides variabilis 89823; Foleyellides stria- tus 89824; Oswaldocruzia pipiens 89825; Rhabdias ranae 89826; Subulascaris falcaustriformis 89827; Physaloptera sp. (larvae) 89828. Smilisca baudini: Aplectana incerta 89829; Aplectana itzocanensis 89830; Rhabdias americanus 89832; Physa- loptera sp. (larvae) 89833. Western North American Naturalist 62(2), © 2002, pp. 170–187

FLOOD FLOWS AND POPULATION DYNAMICS OF ARIZONA SYCAMORE (PLATANUS WRIGHTII)

Juliet C. Stromberg1

ABSTRACT.—Platanus wrightii is a pioneer tree species of warm-temperate riparian deciduous forests in southwest- ern United States and northern Mexico. Dendrochronological analysis of populations in central and southern Arizona indicated that P. wrightii seedlings establish episodically. Long intervals (10–40 years) elapsed with no apparent estab- lishment. Seedling establishment years were positively associated with winter flood size and annual flow rate, and weakly negatively associated with summer flood size. Large floods sometimes preceded multi-year establishment waves. During the past 2 decades an abundance of winter floods and very wet springs has allowed frequent establishment of P. wrightii. For example, plants established at many sites during the winter flood years of 1993 and 1995, particularly along channels scoured and widened by flood waters. Platanus wrightii also reproduces asexually. Ramets were more abun- dant than genets in all populations, and ramets established more frequently than seedlings. Ramet density (mean num- ber per genet) varied widely between populations, from 2 (Huachuca Canyon) to 9 (Haunted Canyon), and increased with tree size within 4 of 9 populations. Population size structure varied across a gradient of watershed area. Populations along streams draining the largest watersheds had an abundance of small trees (mean trunk diameter of <10 cm), while those at headwater sites were dominated by mature trees with trunk diameter ≤190 cm and age ≤235 years. These observations suggest that population structure of P. wrightii is influenced by temporal as well as spatial differences in stream flow regimes.

Key words: Platanus wrightii, floods, riparian habitat, seedling establishment, age structure, ramet.

Population dynamics of many riparian tree well understood. Both P. fremontii and S. good- species are influenced by flood disturbance dingii disperse short-lived seeds in spring and (Duncan 1993, Hughes 1994, Scott et al. 1996, have narrow windows of opportunity for Cordes et al. 1997, Timoney et al. 1997, Rood seedling establishment (Horton et al. 1960, et al. 1998, Sakai et al. 1999). Patterns of ripar- Shafroth et al. 1998). Establishment of these ian tree mortality and establishment vary species depends on winter floods to erode or widely depending on frequency, magnitude, deposit alluvium, high spring stream flows to timing, and duration of floods (Poff et al. moisten seed beds, and slowly receding flood 1997). In arid regions flood patterns are highly waters to allow seedling roots to remain in variable between and within years, and flood contact with declining groundwater. Large flows can be very large relative to base flows. winter floods of long duration can extensively Large floods cause mortality of riparian trees widen channels and facilitate large-scale re- by eroding sediments from root zones and plenishment of Populus-Salix stands (Stromberg uprooting trees, breaking plant stems or bury- 1997, Stromberg et al. 1997, Mahoney and ing them with sediment, and causing mass Rood 1998). Along some perennial to intermit- wasting of flood plains. Floods also intermit- tent alluvial rivers in the Sonoran Desert, tently create conditions that allow establish- flood sequences that allow establishment of ment of pioneer species by changing channel P. fremontii occur approximately once every and flood plain geomorphology, clearing com- 7–10 years. These generally do not occur reg- peting vegetation, and moistening frequently ularly over time (Stromberg et al. 1991, dry flood plain surfaces. Stromberg 1998). Establishment frequency Hydrologic and geomorphic controls on varies among rivers, and some populations are establishment of Populus fremontii and Salix dominated by a single age cohort (Everitt 1995). gooddingii, the dominant pioneer species of Rates of population turnover appear to be Sonoran riparian deciduous forests, are fairly higher on lower reaches of rivers, as suggested

1Plant Biology Department, Arizona State University, Tempe, AZ 85287-1601.

170 2002] FLOOD FLOWS AND POPULATION DYNAMICS 171 by patterns of decreasing tree age with de- production has been shown to be stimulated creasing elevation (Stromberg and Patten 1992, by a variety of factors including disturbance Stromberg 1998). and changes in resource availability (Ischinger Less is known about establishment patterns and Shafroth 1995). of Platanus wrightii, a dominant species of My objective was to determine the in- warm-temperate riparian deciduous forests in fluence of temporal and spatial variability in the southwestern United States and northern stream flow regimes on establishment pro- Mexico (Brown 1994). Platanus wrightii occurs cesses and population size structure of P. in Arizona, New Mexico, western Texas, and wrightii. I hypothesized that P. wrightii seed- northern Mexico at elevations of approxi- lings would establish episodically in response mately 400–2000 m (Kearney and Peebles to flood disturbance and that establishment 1960). At lower elevations P. wrightii co-occurs would occur in years with large winter floods with Populus fremontii and Salix gooddingii. At and high spring flows, similar to patterns for higher elevations it grows with Populus angus- P. fremontii and S. gooddingii. I expected that tifolia and Alnus oblongifolia, among other frequency and abundance of species. Floods are a primary disturbance factor young trees would increase along a down- that structures these warm-temperate riparian stream gradient as drainage area, flood magni- plant communities (Campbell and Green 1968). tude, and channel instability increased. I also Floods occur during late summer after convec- hypothesized that ramets would establish tive thunderstorms, during winter and spring more frequently than seedlings and that pro- after Pacific frontal rainstorms and snow, and duction of ramets would increase during years occasionally in fall as a result of dissipating with above-average stream flow. tropical storms. Life history traits of P. wrightii suggest that MATERIALS AND METHODS it is adapted to disturbance. Each year trees Study Sites produce many small seeds (achenes) with tufted hairs that are suited to wind dispersal. Nine Platanus wrightii study sites were Seeds persist in clusters of 2–4 round balls selected along 7 rivers in southern and central (multiple achenes), each about 2.5 cm in diam- Arizona (Table 1). Each site spans a river eter, that gradually become detached from the length of 1–3 km. At 7 of the sites, data were tree throughout fall, winter, and spring. Seeds collected on tree age and population size ripen in November, germinate in April or May structure; at 2 sites (Sycamore Creek–Round following several weeks of winter stratifica- Valley and Huachuca Canyon), data were col- tion, and lose viability about 6 months after lected only on size structure. The Haunted maturation (Zimmerman 1969, Bock and Bock Canyon and Pinto Creek sites were combined 1989, Brock 1994). Platanus occidentalis, a for some analyses, as were the Sycamore species that occurs in eastern North America, Creek–Bushnell and Sycamore Creek–Work germinates at high rates on bare mineral soil Station sites. Some sites were selected because exposed to high light intensities (Fowells 1965, they allowed investigation of other biohydrol- Sigafoos 1976), and the same may be true of P. ogy relations of P. wrightii, including relation- wrightii. ship of groundwater level to bioproductivity Platanus wrightii reproduces vegetatively (Stromberg 2001a). Other sites were added to as well as by seed. Shanfield (1984) suggested provide a wide range of watershed sizes and that Platanus racemosa, a California species elevations. that is closely related to P. wrightii, may main- The streams mainly flow through alluvial tain its population size by relying on vegeta- sediments that are readily reworked by flood tive reproduction. However, factors that stim- flows; there also are areas within sites where ulate production of vegetative sprouts (ramets) channel morphology is controlled by exposed by P. wrightii are not known. Along one stream bedrock. Platanus wrightii is a dominant species in southern Arizona, Glinski (1977) found no at all sites. Riparian ecosystems at the low-ele- correlation between P. wrightii sprout densities vation study sites are bordered by Sonoran and several parameters including percentage desertscrub and interior chaparral; those at of canopy dieback, soil texture, and distance to higher elevations are bordered by desert stream channel. In other woody species ramet grassland, Great Basin conifer woodland, and 172 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 1. Description of study sites. Elevation Watershed Site name Land owner (m) area (km2) Oak Creek Red Rock State Park 1171 770 Sycamore Creek–Round Valley Tonto National Forest 718 230 Sycamore Creek–Bushnell Tonto National Forest 1040 133 Sycamore Creek–Work Station Tonto National Forest 1090 115 Pinto Creek Tonto National Forest 976 84 Haunted Canyon Tonto National Forest 988 43 Garden Canyon Fort Huachuca (DOD) 1600 30 Ramsey Canyon The Nature Conservancy 1739 13 Huachuca Canyon Fort Huachuca (DOD) 1650 13

Madrean evergreen woodland (Brown 1994). Ramsey Canyon is managed as a natural area Sycamore Creek is in the Mazatzal Mountains by The Nature Conservancy. Stream flow was of central Arizona. Study sites are on Tonto diverted from Garden Canyon in the early National Forest land. The Sycamore Creek– part of the 20th century for use by military Bushnell site experienced heavy recreational personnel prior to the shift to groundwater as use and was grazed by domestic livestock. The a water source. During the late 19th century grazing regime was changed from year-round and early 20th century, Ramsey Canyon har- to winter-only use in the early 1990s. Recre- bored a sizable, but short-lived, town during ational impacts (mainly camping and hiking) an active period of mining in southeastern were concentrated on higher flood plains and Arizona. These 3 headwater streams are in at river crossings. The Sycamore Creek–Work narrow mountain canyons and have narrower Station site is in a cattle grazing exclosure, but riparian corridors than the central Arizona it was used as a pasture for horses. Stream study sites. Garden Canyon has spatially inter- flow is spatially intermittent, with short seg- mittent stream flow and mean annual dis- ments of perennial flow interspersed with charge of 0.03 m3s–1. Stream flow is tempo- intermittently flowing segments. Mean annual rally intermittent at Ramsey and Huachuca discharge at Sycamore Creek is 0.5 m3s–1. canyons. Pinto Creek and its tributary Haunted Can- yon are in the Pinal Mountains of central Ari- Tree Age and Population zona. Study sites at these 2 rivers are on Tonto Size Structure National Forest land and were grazed by domestic livestock (cattle). Stream flow is spa- Slabs or increment cores were collected tially intermittent. from over 500 Platanus wrightii trees. Slabs Oak Creek, on the southern escarpment of and cores were collected when trees were the Colorado Plateau, is the largest of the dormant (1996, 1997, or 1998, depending on study streams. The study site is in Red Rock site) in 4 to 9 belt transects per site. The belts, State Park, which was established in 1986. approximately 10 m wide, extended from the The park lands are managed for natural area channel to the edge of the riparian zone. values, recreation, and environmental educa- There was age zonation at most sites, with tion. The area has not been grazed by live- young stands of dense trees close to the chan- stock since about the 1950s. Oak Creek has nel and sparse stands of older trees farther perennial flow and a mean annual daily dis- away. All live P. wrightii in the belts were mea- charge of 2.8 m3s–1. sured for diameter at breast height (dbh). Cores Garden Canyon, Huachuca Canyon, and or slabs were taken from trees in all apparent Ramsey Canyon are in the Huachuca Moun- size classes. If trees were <4 cm in stem diam- tains of southeastern Arizona. None of the eter, pruning clippers were used to clip slabs Huachuca Mountain study sites has been near the ground surface. For trees >4 cm grazed by domestic livestock for several dec- diameter, increment borers were used to col- ades. Garden Canyon and Huachuca Canyon lect cores at 0.1–0.2 m above the ground sur- are located on Fort Huachuca military base. face. One to 4 replicate cores were collected 2002] FLOOD FLOWS AND POPULATION DYNAMICS 173 per tree, with the larger numbers of cores classified as genets. Some of these may have being collected if first attempts did not pene- been ramets from a long-dead parent tree and trate the center of the tree. Very large trees may be incorrectly classified. Also, some trees were underrepresented in the samples. Some classified as genets may have arisen from of the largest trees had hollow interiors or buried branches of old trees (Bock and Bock extensive heart rot, and cores were not col- 1989). Analysis of covariance (tree age as lected. Interior portions of some of the col- covariate, tree dbh as dependent variable) was lected cores also showed evidence of decay conducted to determine if there were signifi- and were discarded. Those that did not have cant differences in size and thus growth rate heart rot and that had rings that extended to between genets and ramets. These analyses the center of the tree were retained for analy- were conducted within sites, stratified by sis. This group of usable cores and slabs stream flow reach type. These analyses indi- encompassed 104 from Garden Canyon, 42 cated that there were significant differences from Ramsey Canyon, 228 from Sycamore between groups, with ramets growing signifi- Creek, 114 from Haunted Canyon/Pinto Creek, cantly slower than genets (Table 2). and 92 from Oak Creek. Sample size was small In the laboratory, cores were glued to wood at Ramsey Creek because there were very few mounts and some slabs were re-sawn to pro- trees in small size classes. duce smooth surfaces (Stokes and Smiley 1968). Trees were classified as to seed origin Cores and slabs were sanded with a belt sander (genets) or sprout origin (ramets). Ramets were using a graded series of sandpaper (100–600 distinguished from genets based on their posi- grit) and viewed under a dissecting micro- tion (arising from the base of a larger tree or scope to mark (score) the annual rings. After within a few meters of the base), size, and loca- scoring, annual rings were measured for annual tion within the flood plain (i.e., there was appar- ring width with a Bannister-type automated ent age zonation at the sites, with seedlings measuring system. Site chronologies were gen- located along channel edges and larger trees erated by successively cross-dating within trees with associated small sprouts on flood plains (using the replicate cores) and then between trees within sites. Trees were aged by count- farther from and above the channel). In most ing annual rings. For the majority of young cases the origin of the trees was clear. At many trees (Figs. 1, 2, 3), the core contained the cen- sites there were stands of small, single- ter of the tree (as indicated by the presence of trunked trees, distributed in nonaggregated the persistent pith that is produced in the ini- fashion along the stream channel; these were tial growing season), allowing for precise aging. classified as genets. Some of these young trees In other cases the core was slightly off center. had smaller stems arising from their bases; These trees were aged by estimating distance these were classified as ramets. Some of the from the end of the core to the tree center (in larger trees were single-trunked and were cm), calculating average annual growth rate classified as genets. More often the large trees for the oldest section of the increment core, occurred in clumps or aggregations. The and adding to the annual ring count the esti- largest tree within the clump was classified as mated number of years to reach the center of a genet. Small-diameter stems arising from the the tree. Regression analyses showed high base of the larger tree or emerging from the correspondence between annual radial growth ground and forming a ring around the larger and annual stream flow rate and air tempera- tree were classified as ramets (Glinski 1977). ture, providing confirmation that rings were Most ramets arose directly from or just above correctly scored (Stromberg 2001b). Age of the the base of parent trees. Older genets gener- largest uncored trees at the sites was estimated ally were widely spaced on the flood plain and from reach-specific regression equations that surrounded by a ring of smaller ramets. In predicted age from dbh (Table 2). some cases, however, there was a circle of trees of relatively similar size. Tree aging indi- Data Analysis cated that in some cases one of the trees in the Forward-stepping multiple regression analy- circle was older than the others, in which case sis was used to identify hydrologic factors it was classified as a genet. In cases where all associated with seedling and sprout establish- trees in the circle were of similar age, all were ment years for the 3 populations (Sycamore 174 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 2. Regression equations predicting age (years) from trunk diameter (dbh in cm) of Platanus wrightii. All regres- sion models are significant at P < 0.05. Regression equation nr2 Ramsey Canyon genets* y = 1.22x + 10.5 16 0.67 ramets y = 1.79x + 16.1 26 0.45 Garden Canyon genets* y = 1.71x + 1.1 46 0.77 ramets y = 2.51x – 6.6 55 0.76 Oak Creek genets* y = 1.84x – 2.2 38 0.70 ramets y = 2.25x – 0.2 54 0.80 Haunted Canyon genets* y = 1.79x + 11.7 17 0.79 ramets y = 2.70x – 6.4 23 0.84 Pinto Creek (perennial) genets* y = 1.91x – 3.9 23 0.90 ramets y = 2.38x – 1.4 21 0.90 Pinto Creek (intermittent) genets* y = 2.32x + 3.9 12 0.83 ramets y = 3.13x – 1.6 18 0.95 Sycamore Creek–Work Station (intermittent 1) genets* y = 2.11x – 8.6 24 0.69 ramets y = 3.13x – 3.5 15 0.87 Sycamore Creek–Work Station (intermittent 2) genets y = 3.24x – 13.0 23 0.75 ramets y = 3.82x – 16.6 15 0.87 Sycamore Creek–Bushnell Tanks (perennial) genets* y = 1.61x – 6.0 22 0.77 ramets y = 2.15x – 3.6 20 0.84 Sycamore Creek–Bushnell Tanks (intermittent 1) genets* y = 1.84x – 5.7 24 0.81 ramets y = 2.05x + 1.5 17 0.75 Sycamore Creek–Bushnell Tanks (intermittent 2) genets* y = 3.11x – 15.8 32 0.89 ramets y = 3.54x – 5.2 36 0.85

*Significant difference between genets and ramets at P < 0.05, ANCOVA

Creek–Bushnell and Work Station, Oak Creek, site. The Sycamore Creek Sunflower gage and Haunted Canyon/Pinto Creek) located (09510150) is closer to the site, but stream near long-term United States Geological Survey flow data were collected only for 1962–1976. (USGS) stream gages. The independent vari- Mean annual flow rates and peak annual flood ables in the analyses were mean quarterly dis- sizes at the Fort Mcdowell and Sunflower charges (mean daily flow during October gages are highly correlated (r = 0.97 and r = through December, January through March, 0.94, respectively, n = 15 years of overlapping April through June, and July through Septem- record). Peak annual flows average about 25% ber), mean annual discharge during the water higher at Fort Mcdowell than at Sunflower. year, size of the largest winter flood (October Data for the Haunted Canyon/Pinto Creek through March), size of the largest summer analysis were from the nearby Pinal Creek flood (April through September), and size of Inspiration Dam gage (09498400). Oak Creek the largest respective winter and summer stream flow data were from the Cornville gage floods during the preceding 3-year period. (09504500), located 8 km downstream of the The following stream flow chronologies were study site. used in the analyses: 1962–1996 for Sycamore The dependent variable in the regression Creek, 1981–1996 for Haunted Canyon/Pinto analyses was categorical. A value of 1 was Creek, and 1949–1996 for Oak Creek. Data assigned to establishment years and a value of for the Sycamore Creek analysis were from zero to nonestablishment years. Number of the Sycamore Creek Fort Mcdowell gage trees per establishment year was not used as (09510200), 20 km downstream of the study the dependent variable to avoid confounding 2002] FLOOD FLOWS AND POPULATION DYNAMICS 175

Fig. 1. Size of largest recorded flood during winter Fig. 2. Size of largest recorded flood during winter (October–March) and summer (April–September) at Oak (October–March) and summer (April–September) at Pinal Creek (USGS gage 09504500), 1949–1996. Also shown are Creek (USGS gage 09498400), 1981–1996. Also shown are numbers of aged Platanus wrightii genets dating from this numbers of aged Platanus wrightii genets from nearby period. Black portions of the bars indicate trees for which Haunted Canyon and Pinto Creek. at least one core or slab penetrated the center of the tree; gray portions indicate trees with cores that were slightly off center. changes across a gradient of watershed area. Watershed area was presumed to be a surro- gate for flood magnitude, given that rivers sit- demographic effects such as age-dependent uated lower in the watershed collect runoff mortality processes. Regression analyses were over a much larger area (Ribeiro and Rous- conducted separately for genets and ramets. selle 1996). At each site stem diameter (dbh) of Analyses were conducted using only those all living P. wrightii was measured in a mini- trees that were clearly distinguishable as ram- mum of 4 belt transects spanning the riparian ets or genets. Two sets of analyses were con- corridor. The number of ramets per genet (i.e., ducted for each to test the robustness of the number of trees per cluster) was counted for a results. For one of the analyses, the data set minimum of 25 genets per site. was based on all of the aged trees. For the Pearson’s correlation coefficient was calcu- other a year was classified as an establishment lated to determine whether mean trunk diam- year only if at least 2 trees dated from that eter of the P. wrightii population, diameter of year. The analyses were based on 22 genets the largest measured tree, and mean ramet and 26 ramets that established during the 16- density per genet were significantly related to year period of analysis for Haunted Canyon/ log-transformed watershed area. Watershed Pinto Creek, 85 genets and 70 ramets from area was measured from USGS 15- or 7.5- Sycamore Creek (35-year period), and 12 genets minute topographic maps or obtained from and 45 ramets from Oak Creek (48-year period). published USGS records. Pearson’s correla- tion coefficient also was calculated to test for Population Size Structure significant relationship between tree size (dbh) Population size structure data were col- and ramet density within sites. SYSTAT 7.0 lected at all 9 sites to allow examination of and 8.0 were used for all statistical analyses. 176 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 4. Mean daily flow during the water years 1986– 1996 at Sycamore Creek (USGS gage 09510200).

den Canyon the 2 largest recorded floods dur- ing the short period of stream flow data collec- tion (1959–1965, 1993 to present) occurred in February 1995 (1.6 m3s–1) and July 1964 (1.6 m3s–1; data not shown). During years with large winter floods, such as 1993, stream flows tended to be high through- out the spring season (Fig. 4). Total annual dis- charge also was higher than average during Fig. 3. Size of largest recorded flood during winter (October–March) and summer (April–September) at years with winter floods. At Sycamore Creek, Sycamore Creek (USGS gage 09510200), 1962–1996. Also for example, peak winter flood size was signif- shown are numbers of aged Platanus wrightii genets. icantly correlated with January–March dis- charge (r = 0.90, n = 35), April–June discharge (r = 0.88), and mean annual discharge (r = RESULTS 0.86, n = 35, P < 0.05). Peak summer flood size Flood-flow Patterns was not significantly correlated with annual dis- charge (r = –0.18, n = 35). Annual discharge and size of winter and Summer floods generally were small or summer floods in the study rivers varied con- absent during years with large winter floods. siderably over time. At all sites there were Correlations between log-transformed sizes of several large winter floods in recent decades. the largest winter and largest summer flood Winter floods during the 1992–93 El Niño were significantly (P < 0.05) negative at Oak event were particularly large and long. At Oak Creek (r = –0.68, n = 48), Sycamore Creek (r Creek the 2 largest recorded floods occurred = –0.43, n = 35), and Pinal Creek (r = –0.59, in February 1993 (instantaneous peak dis- n = 16). Compared to winter floods, summer charge of 736 m3s–1) and February 1980 (748 floods tended to have low mean daily flow rate m3s–1), both of which resulted from rain on relative to their instantaneous peak flow rate, snow (Fig. 1). According to USGS records, a reflecting their tendency to be “flashy.” For very large flood also occurred in March 1938 example, the linear regression equation pre- in Oak Creek, for which discharge was not dicting mean daily discharge from instanta- recorded. At Pinal Creek the 1993 flood was neous peak flow rate in Pinal Creek has a the largest during the past 2 decades, with a lower slope for summer floods (y = 0.086x – peak flow rate of 161 m3s–1 in January (Fig. 2). 0.67; r2 = 0.54, n = 12) than for winter floods The largest recorded flood in Sycamore Creek (y = 0.63x – 12.8; r2 = 0.94). occurred in September 1970, with a flow rate of 685 m3s–1 (Fig. 3). The largest winter flood Seedling Establishment at Sycamore Creek occurred in March 1978, Patterns with a flow rate of 501 m3s–1. Peak flow during Platanus wrightii seedlings established winter 1993 ( January) was 399 m3s–1. At Gar- episodically (Figs. 1, 2, 3). Establishment events 2002] FLOOD FLOWS AND POPULATION DYNAMICS 177

TABLE 3. Multiple regression models predicting Platanus wrightii seedling establishment years from hydrologic vari- ables. For the conservative data set, establishment years were based on a minimum sample size of 2 trees per year. Direction of Model r2 Model variables correlation P value

SYCAMORE CREEK All genets 0.27 Annual flow rate + <0.01 Conservative data set 0.29 Winter flood size (3 yr) + <0.01 Summer flood size – 0.09 OAK CREEK All genets 0.36 Winter flood size (3 yr) + <0.01 Oct–Dec flow rate + 0.03 Conservative data set 0.21 Winter flood size (3 yr) + <0.01 HAUNTED CANYON/PINTO CREEK All genets 0.69 Winter flood size (3 yr) + <0.01 Oct–Dec flow rate + 0.01 July–Sept flow rate + 0.02 Conservative data set 0.81 Winter flood size (3 yr) + <0.01 July–Sept flow rate + 0.04 did not occur regularly over time. At all sites 1983. Seedlings at this site also established there were periods of over a decade with abundantly in 1991, 1993, and 1995, all of no apparent establishment: no establishment which were wet years with relatively large win- occurred, no survivors remained, or the trees ter or spring floods and small or absent sum- were too sparse to be detected in the sample mer floods (Fig. 4). The 1993 and 1995 saplings of cored trees. For all 3 populations analyzed, were abundant along channel margins that winter flood size (over the prior 3 years) was a had been eroded and widened by flood waters. significant component of the multiple regres- In some areas the 1993 and 1995 cohorts were sion models that predicted seedling establish- spatially intermingled. In other areas the 2 ment years (Table 3). Annual discharge, win- cohorts formed distinct bands, with the 1995 ter-season discharge, and summer-season dis- cohorts being closer to the channel. charge were components of the models for at Platanus wrightii also established fre- least a single population. Summer flood size quently at Sycamore Creek in the late 1920s (negative influence) was a component of 1 of and early 1930s and the 1860s through the the 2 seedling establishment regression mod- 1900 decade (Fig. 5). The oldest cored genet at els for Sycamore Creek. Sycamore Creek was approximately 207 years, During the 35-year period of hydrologic having established during the 1780s. The 3 analysis (1962–1996) at Sycamore Creek, there largest trees measured at the Sycamore Creek– were 13 establishment years (Fig. 3). In 12 of Bushnell and Work Station sites (111, 113, and these 13 years, there was a winter flood but no 131 cm dbh) had predicted ages of 175–200 summer flood. The winter flood was large (at years, based on extrapolation of age-dbh regres- least 150 m3s–1) in 7 of these years. All estab- sion equations (Fig. 7, Table 2). The largest lishment years occurred within 3 years of a and oldest P. wrightii at these sites ranged up large winter flood. Not all years with winter to 50 m from the stream channel. floods were establishment years: there were 2 At Haunted Canyon/Pinto Creek, 6 estab- years (1968 and 1978) with large winter floods lishment years occurred during the 16-year but no establishment. None of the years with period of hydrologic analysis (Fig. 2). Four of large or small summer floods were establish- these 6 years (1983, 1985, 1993, 1995) had ment years for P. wrightii at Sycamore Creek. winter floods but no summer floods. One At most sites there was a pattern for clus- (1984) had a winter flood and a summer flood. ters of establishment years to alternate with A small number of plants established in 1994, quiescent periods. This pattern was particu- which had a small summer flood and came larly strong at Sycamore Creek. A wave of 8 immediately after the 1993 flood-of-record. consecutive establishment years began in the Another active period of seedling establish- winter flood year of 1979 and extended through ment at Haunted Canyon and Pinto Creek 1986, with the largest establishment pulse in occurred in the 1900s through 1930s (Fig. 5). 178 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Many of the Oak Creek trees established in the 1930s and early 1940s (Fig. 5). The large spring flood of 1938 may have facilitated establishment of some of the trees in this re- generation pulse. The oldest trees cored at this site established in the 1830s (age of approximately 160 years). The largest uncored trees at the site (60–75 cm) had predicted ages of 100–135 years. At Garden Canyon several trees established during the 1970s and 1980s (e.g., 1973, 1978, 1983; Fig. 5), similar to patterns observed at the central Arizona sites. However, saplings from 1993 or 1995 were not present. Among older cohorts trees established fairly frequent- ly during the 1870s, 1880s, and 1890s. The oldest tree cored at Garden Canyon had a dbh of 122 cm and an estimated age of 175 years (established ca 1766). The largest measured tree at Garden Canyon had a dbh of 139 cm and a predicted age of 215 years. At Ramsey Canyon roughly 5 decades have elapsed since the last establishment year of 1952 (Fig. 5). Many of the trees at Ramsey Canyon established during the 1930s and be- tween 1880 and 1915. The oldest cored tree at Ramsey Canyon had a dbh of 162 cm and an age of approximately 235 years (established in the 1760s). The largest tree measured at Ram- sey had a dbh of 194 cm and a predicted age of 250 years. Fig. 5. Age structure histograms for Platanus wrightii genets at Oak Creek, Sycamore Creek, Haunted Ramet Establishment Patterns Canyon/Pinto Creek, Garden Canyon, and Ramsey Canyon. The x-axis extends from 1760 through 1996. Ramets established more frequently than seedlings (Fig. 6) and were more abundant than genets in all populations. In some cases The oldest cored genet was 200 years (estab- periods that were favorable for seedling estab- lished in the 1790s) and had a dbh of 92 cm. lishment were also favorable for ramet pro- Extrapolation of site-specific age-dbh regres- duction. In other cases ramets established sion equations yielded ages of 250–310 years during times when seedlings did not. For for the largest trees at these sites (127, 134, example, during the 1950s several ramets but and 160 cm dbh). few genets established at Haunted Canyon/ In the 48-year period of hydrologic analysis Pinto Creek and at Oak Creek. Neither ramets (Fig. 1) at Oak Creek, 7 establishment years nor genets established at Sycamore Creek in occurred. Six of these were in years with win- the 1950s. All recent recruitment at Ramsey ter floods, 5 of which (1952, 1967, 1979, 1993, Canyon has been due to ramets. 1995) had very large floods. No flood occurred Multiple regression models significantly in the establishment year of 1994. Only 1 (1985) predicted ramet establishment years for the 2 of the 7 establishment years had a summer populations (Table 4). However, ramet models flood, and it was small. Over 20 years had win- explained less variance than did seedling ter floods but no detected establishment. Many establishment models, and there were no con- of the establishment years at Oak Creek over- sistent trends between populations in terms of lapped with establishment years at the 2 other influential hydrologic variables. Among popula- central Arizona sites. tions, hydrologic correlates of ramet production 2002] FLOOD FLOWS AND POPULATION DYNAMICS 179

Fig. 7. Trunk diameter vs. age for Platanus wrightii at Fig. 6. Age structure histograms for Platanus wrightii Oak Creek, Sycamore Creek, Haunted Canyon/Pinto ramets at Oak Creek, Sycamore Creek, Haunted Canyon/ Creek, Garden Canyon, and Ramsey Canyon. Data are Pinto Creek, Garden Canyon, and Ramsey Canyon. The shown separately for genets and ramets. x-axis extends from 1760 through 1996.

sites as watershed size increased (r = 0.79, n included winter flood size, spring-season stream = 9, P < 0.05). flow rate, and summer dry-season stream flow Among the 9 sites sampled for population rate. size structure, mean number of live ramets per genet ranged from 1.9 to 8.6 (Table 5). Population Size Structure Ramet density was not significantly related to Mean stem diameter of the P. wrightii pop- watershed size (r = 0.38, n = 9, P > 0.05). ulation (inclusive of ramets and genets; Table Within sites, ramet density increased signifi- 5) declined as watersheds became larger (r = cantly with genet size for 4 of 9 populations. 0.78, n = 9, P < 0.05). At Oak Creek, which The relationship between ramet density and drains a watershed of >700 km2, most trees genet size appeared to be curvilinear at many were <10 cm in diameter (Fig. 8). At the 3 sites, with ramet density declining for the sites draining watersheds of ≤30 km2 (all of largest trees (Fig. 9). Trees of intermediate size which are in the Huachuca Mountains), the (40–80 cm dbh) generally produced the most mean diameter of P. wrightii stems was >30 ramets. Most ramets on ramet-rich trees were cm. At Ramsey Canyon (13-km2 watershed), small (<1 m tall) and young (1–3 years old; most trees were in the 41–50 cm dbh class, Fig. 10). Mortality rates for small ramets and only 7% had stems with diameters <10 appeared to be high, based on observations cm. Maximum tree size also declined among of large numbers of small, dead ramets. Two 180 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 4. Multiple regression models predicting Platanus wrightii sprout establishment years from hydrologic vari- ables. Direction of Model r2 Model variables correlation P value

SYCAMORE CREEK All ramets 0.25 Winter flood size + <0.01 Oct–Dec flows – 0.05 Conservative data set 0.31 Annual flow rate – 0.04 April–June flow rate + <0.01 OAK CREEK All ramets — Model not significant Conservative data set 0.13 Jan–Mar flow rate + 0.01 HAUNTED CANYON/PINTO CREEK All ramets 0.33 July–Sep flow rate – 0.02 Conservative data set — Model not significant

outliers are not shown on the graphs. One tree at Oak Creek with a dbh of 73 cm had 126 ramets. Another at Haunted Canyon (91 cm) had 137 ramets. Some trees with abundant ramets showed signs of canopy dieback.

DISCUSSION Seedling Establishment Processes Relationships between flood flow patterns and seedling establishment of P. wrightii were similar in many ways to those of dominant pio- neers of lower-elevation Sonoran riparian forests. Like Populus fremontii and Salix good- dingii, P. wrightii established episodically, main- ly during years with large winter floods and high spring-season stream flows. Large winter floods do the geomorphic “work” that creates seedling safe sites (sensu Harper 1997) by scouring vegetation, eroding sediments, de- positing fresh alluvium, or exposing the bare mineral soil that stimulates high rates of ger- mination (Sigafoos 1976). High spring stream flows that typically follow large winter floods serve to moisten seeds and provide adequate water for seedling growth, not always available in these small streams that vary widely in annual flow rates (Sacchi and Price 1992). Stream flows in 1993 and 1995, for example, remained high throughout spring and summer, presumably providing the moisture that allowed high seedling survivorship. At several study sites, including Sycamore Creek and Haunted Fig. 8. Size structure histograms (10-cm dbh increments) Canyon, P. wrightii saplings from 1993 and for 5 populations of Platanus wrightii. Values on the x-axis indicate the upper range of the size class (i.e., 20 indicates 1995 were intermingled with Salix bonplandi- trees with dbh of 11 to 20 cm). Plots are shown in order of ana, S. gooddingii, and P. fremontii in the descending watershed size. recruitment zones (Fig. 11), indicating a high 2002] FLOOD FLOWS AND POPULATION DYNAMICS 181

TABLE 5. Population size structure variables. Sites are listed in descending order of watershed size. Correlation Mean Mean Maximum coefficient, ramet trunk trunk ramet density density diameter diameter vs. genet size Site name (no./genet) (cm) (cm) (r value) Oak Creek 6.9 9 75 0.43* Sycamore Creek–Round Valley 1.9 8 23 0.13 Sycamore Creek–Bushnell 5.2 12 131 0.34 Sycamore Creek–Work Station 4.2 10 67 0.08 Pinto Creek 8.3 15 160 0.23 Haunted Canyon 8.6 10 134 0.49* Garden Canyon 3.6 35 133 0.32* Ramsey Canyon 2.4 60 194 0.07 Huachuca Canyon 2.3 31 186 0.34*

*Significant correlation between ramet density and genet size at P < 0.05 degree of similarity in regeneration niches among these species. The combination of large winter floods and cool wet springs also creates favorable regeneration conditions for Populus angustifolia, a member of the cool- temperate deciduous riparian forest assem- blage (Baker 1990). House and Hirschboeck (1995) describe the 1993 floods as one of the most severe winter flooding episodes in Arizona history, surpassed only by large floods in the 1890s. Regionally, floods were also very large in 1983 and the early 1900s. Large winter floods can exten- sively modify stream channels and flood plains (Hooke 1994). Floods that are large enough to extensively rework channel and flood plain sediments can facilitate a multi-year wave of P. wrightii establishment. Multi-year post-flood establishment also has been documented for Populus deltoides subsp. monilifera along the Missouri River in Montana (Scott et al. 1997). This differs somewhat from the pattern for P. fremontii along the larger alluvial desert rivers in Arizona, where high rates of sedimentation can quickly render the flood plain recruitment zones unsuitable for further seedling estab- lishment. This sedimentation process did not appear to be as evident in the P. wrightii study streams, allowing flood-scoured zones to remain suitable for establishment in ensuing wet years. Platanus wrightii establishment patterns showed a high degree of temporal variation, reflecting highly variable flood patterns both in terms of season and year. Decades with abundant establishment included the 1980s Fig. 9. Ramet density as a function of genet trunk diam- and 1990s, 1920s and 1930s, and those just eter for 5 populations of Platanus wrightii. before and after 1900. High variation in flood 182 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 10. Small (above) and large (facing page) ramets of Platanus wrightii arising from the patterns in semiarid regions, on decadal scales, shift has been a high frequency of El Niño is largely due to influences of El Niño–South- years during the 1980s and 1990s, character- ern oscillation (ENSO) weather patterns ized by wet winters and large winter floods. (Molles et al. 1992, Webb and Betancourt These hydrologic conditions have provided 1992). Swetnam and Betancourt (1998) note opportunities for frequent and extensive that “the post-1976 period shows up as an regeneration of P. wrightii, as well as Populus uprecedented surge in tree-ring growth fremontii and Salix gooddingii, in recent within millenia-length chronologies” of upland decades (Stromberg 1998). trees in the southwestern United States and Platanus wrightii did not establish in re- reflects a post-1976 shift to the negative phase sponse to summer floods. Monsoon-season of the Southern oscillation. The result of this floods conceivably could create conditions suit- 2002] FLOOD FLOWS AND POPULATION DYNAMICS 183

base of parent trees (photograph by J. Stromberg). able for seedling establishment in late summer tion. Compared to a winter flood of the same or the following spring. Optimum germination instantaneous discharge, summer floods cause temperature for some P. wrightii is 27°C, con- less geomorphic change (Everitt 1995) and pre- sistent with early- or late-summer germination sumably do not moisten soils for a long period (Siegel and Brock 1990). However, in a growth after the flood peak. Thus, summer floods chamber programmed to simulate field condi- probably do not frequently create suitable tions, all P. wrightii seeds germinated in April establishment sites for P. wrightii. or May (R. Richter personal observation). Many years, and even one or more decades, Summer floods arise due to localized convec- elapsed in which conditions apparently were tive thunderstorms and tend to have high unsuitable for establishment of P. wrightii. instantaneous peak flow rate but short dura- Some of these periods were characterized by 184 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 11. Flood-scoured stream channel with bands of Platanus wrightii and Populus fremontii seedlings that established in 1993 and 1995 (photograph by J. Stromberg). below-average stream flows. A severe drought However, although seedlings established dur- occurred in the American Southwest in the ing the 1993 and 1995 winter flood years at 1940s and 1950s (Shelton and Bahre 1993, central Arizona sites, they did not do so at Swetnam and Betancourt 1998). This drought southeastern Arizona sites (Ramsey and Gar- probably contributed to sparse regeneration den canyons). Regional differences in storm by P. wrightii (Fig. 8). tracks may have contributed to this pattern. Activity of convective summer thunder- Between-site differences in stream size and storms and floods was high in some parts of flood power also may have contributed. Popu- Arizona in the decades prior to the 1970s lations of P. wrightii along small headwater (Webb and Betancourt 1992), perhaps also fur- streams were dominated by old trees and had ther contributing to the paucity of cohorts dat- few young trees. Downstream decreases in ing from the middle decades of the 1900s. mean or maximum tree age also have been Summer floods have been reported to cause observed for Populus species (Stromberg and mortality of P. wrightii seedlings and trees by Patten 1992, Stromberg 1998). These patterns uprooting and scouring plants (Bock and Bock suggest that size of recruitment events and 1989). In this study there was a weak negative rates of population turnover are greater on correlation between summer flood magnitude larger streams. and P. wrightii seedling establishment for a Establishment processes of P. wrightii also single population. This pattern may directly may have been influenced by domestic live- reflect flood-caused seedling mortality, or may stock and native . In some circum- reflect a climatic tendency for summer floods stances, livestock grazing can favor pioneer to be small in years with large winter floods. woody plants by reducing abundance of herba- There was overlap among sites in seedling ceous competitors and maintaining the com- establishment years, reflecting the trend for munity in an early successional state. Heavy winter storms and floods to occur region-wide. livestock grazing also can inhibit establishment 2002] FLOOD FLOWS AND POPULATION DYNAMICS 185 of pioneer tree seedlings. Livestock grazing the contrast between Garden and Ramsey can- has been identified as a factor contributing to yons, both of which are headwater streams in low abundances of P. wrightii seedlings and the same ecoregion, suggests that some factor saplings (Glinski 1977, Rucks 1984), although may be reducing recruitment rates at Ramsey Bock and Bock (1989) attribute the negative Canyon. Deer browse, channel incision, dense effect to legacies of past heavy grazing (such as cover of Vinca major in the understory, and channel incision) rather than to direct effects canyon shade are factors that may be con- of browse or trampling. This study was not tributing to the lower abundance of young P. designed to assess effects of livestock grazing wrightii at Ramsey Canyon compared to Gar- on P. wrightii, in that there was no replication den Canyon. of grazed and ungrazed sites within stream Vegetative Reproduction flow regime or watershed size classes. Such a study would be valuable, given the large num- Platanus wrightii, like many riparian plant ber of riparian and upland sites on public and species that live in temporally unpredictable private land that are grazed in the American environments, reproduces by seed and vegeta- Southwest. This present study does highlight tively. Vegetative sprouts (ramets) comprised the need to consider a suite of factors that over half of the P. wrightii stems in all popula- influence seedling establishment when assess- tions. Ramet production may facilitate persis- ing the effects of livestock management changes tence of P. wrightii populations by allowing on riparian vegetation. Spatial controls in regeneration during long periods that can reference watersheds are important to allow occur in which seedling regeneration needs one to distinguish between land management are not met. The large root and stem masses changes and weather-related influences. With- that develop on old P. wrightii clones may pro- out control sites it is difficult to know whether vide a buffer against flood scour and perhaps increases or decreases in seedling numbers are increase survivorship over nonclonal species. related to livestock management changes or to Ramet production can be stimulated by changes in flood flow and low flow patterns. many factors, including disturbance, changes Riparian tree age structure, or at least pres- in resource availability, and disease. For 2 P. wrightii populations, ramet production was ence of seedlings and saplings, is one measure associated with winter flood years and/or wet of ecological health used by land managers to springs, although these trends were not robust. assess impacts from livestock grazing (Prichard Flood debris can trigger sprouting in riparian et al. 1998). If examined for a larger number of trees by causing wounding or stem removal sites, relationships between P. wrightii size or (Rood et al. 1994, Ischinger and Shafroth 1995). age structure and watershed area could pro- Increased ramet production in years with wet vide a reference against which managers could springs may be related to weather-induced assess a particular P. wrightii population for expression of anthracnose (Gnomonia platani). “healthy” structure. To be most useful to man- Anthracnose is a systemic fungal disease that agers, the relationships between size/age struc- is widespread in populations of P. wrightii and ture and watershed area should be developed other Platanus species (Svihra and McCain within ecoregions and within flow regime 1992). Cool, wet weather in spring increases types, thereby taking into account additional susceptibility to the , resulting in canopy site factors that influence population dynamics dieback. During this study I observed many such as the relative contribution of summer trees with canopy dieback and other symp- and winter flows to mean annual discharge. toms of anthracnose in 1995, a year with a For example, regeneration of P. wrightii may cool, wet spring. Many of these trees had an be inherently less frequent along streams in abundance of young ramets. Thus, seedlings ecoregions with larger and more frequent and ramets both may establish in years with summer floods. Results of this study suggest wet winters and springs, although due to dif- that dominance by old individuals is typical ferent mechanisms. for P. wrightii populations along some ungrazed Ramet density varied substantially among headwater streams. Thus, infrequent regener- populations but was not significantly related to ation of P. wrightii at headwater sites may not watershed area. However, the positive correla- necessitate management concern. However, tion coefficient between the 2 variables was 186 WESTERN NORTH AMERICAN NATURALIST [Volume 62 relatively high, warranting further investiga- BROCK, J. 1994. Phenology and stand composition of woody tion. To understand the causes of between- riparian plants in the southwestern United States. Desert Plants 11:23–31. population variability in ramet densities, a BROWN, D.E., EDITOR. 1994. Biotic communities: south- large number of sites should be studied to dis- western United States and northwestern Mexico. entangle effects due to tree size and age. Tree University of Utah Press, Salt Lake City. size influences ramet density and co-varies CAMPBELL, C.J., AND W. G REEN. 1968. Perpetual succes- sion of stream-channel vegetation in a semiarid with watershed size. Another factor that might region. Journal of the Arizona Academy of Science be examined is light availability. Ramet pro- 5:86–98. duction is higher in open, sunny sites for CORDES, L.D., F.M.R. HUGHES, AND M. GETTY. 1997. Fac- some, but not all, riparian trees (Shaw 1991, tors affecting the regeneration and distribution of Mishio and Kawakubo 1998). Ramet density riparian woodlands along a northern prairie river: the Red Deer River, Alberta, Canada. Journal of Bio- may be higher for P. wrightii growing in open geography 24:675–695. outwash plains or alluvial valleys than in nar- DUNCAN, R.P. 1993. Flood disturbance and the coexis- row shady canyons. tence of species in a lowland podocarp forest, South Between-site differences in herbivory rates Westland, New Zealand. Journal of Ecology 81: 403–416. may have contributed to some of the differ- EVERITT, B.L. 1995. Hydrologic factors in regeneration of ences in ramet density among populations. Fremont cottonwood along the Fremont River, Utah. Managers at Ramsey Canyon have observed Natural and Anthropogenic Influences in Fluvial deer browsing on the P. wrightii ramets and Geomorphology, Geophysical Mongraph 89:197–208. FOWELLS, H.A. 1965. Silvics of forest trees of the United noted that deer are very abundant in the Ram- States. USDA Forest Service Handbook 271. Wash- sey Canyon sanctuary. Deer browsing may ington, DC. have contributed to the low numbers of ram- GLINSKI, R.L. 1977. Regeneration and distribution of ets at Ramsey Canyon as well as to the lack of sycamore and cottonwood trees along Sonoita correlation between ramet density and genet Creek, Santa Cruz County, Arizona. USDA Forest Service General Technical Report RM-43:116–123. size at this site. HARPER, J.L. 1977. Population biology of plants. Academic Press, New York. ACKNOWLEDGMENTS HOOKE, J.M. 1994. Hydrological analysis of flow variation of the Gila River in Safford Valley, southeast Arizona. I thank the many people who collected or Physical Geography 15:262–281. HORTON, J.S., F.C. MOUNTS, AND J.M. KRAFT. 1960. Seed analyzed field data: Matt Chew, Caitlin Corn- germination and seedling establishment of phreato- wall, Joelle Don de Ville, Kim Fox, Ginny phyte species. USDA Rocky Mountain Forest and O’Connell, April Henry, Niki Herd, Jennifer Range Experiment Station Paper 48. Hickman, Heidi Johnson, Crystal Levine, Elena HOUSE, P.K., AND K.K. HIRSCHBOECK. 1995. Hydrological and paleohydrological context of extreme winter Ortiz-Barney, Susan Pierce, Troy Smith, and flooding in Arizona, 1993. Arizona Geological Sur- Ron Tiller. I thank Janet Johnson of Tonto vey Open-File Report 95-12. Tucson, AZ. National Forest, John Schreiber of Red Rock HUGHES, F.M.R. 1994. Environmental change, disturbance, State Park, and Sheridan Stone of Fort Hua- and regeneration in semi-arid floodplain forests. Pages 321–345 in A.C. Millington and K. Pye, edi- chuca for providing site information, and the tors, Environmental change in drylands: biogeo- following agencies for providing access to graphical and geomorphological perspectives. John study sites: Arizona State Parks, Tonto National Wiley and Sons, Ltd. Forest, Coronado National Forest, Fort Hua- ISCHINGER, L.S., AND P. B . S HAFROTH. 1995. Induced root- chuca, and The Arizona Nature Conservancy. suckering shows potential for reestablishing riparian trees (New Mexico). Restoration and Management I gratefully acknowledge the Arizona Water Notes 13:121. Protection Fund and United States Depart- KEARNEY, T.H., AND R.H. PEEBLES. 1960. Arizona flora. ment of Defense (Fort Huachuca) for provid- University of California Press, Berkeley. ing financial support. MAHONEY, J.M., AND S.B. ROOD. 1998. Streamflow require- ments for cottonwood seedling recruitment—an integrative model. Wetlands 18:634–645. LITERATURE CITED MISHIO M., AND N. KAWAKUBO. 1998. Ramet production by Mallotus japonicus, a common pioneer tree in BAKER, W.L. 1990. Climatic and hydrologic effects on the temperate Japan. Journal of Plant Research 111: regeneration of Populus angustifolia James along the 459–462. Animas River, Colorado. Journal of Biogeography 17: MOLLES, M.C., JR., C.N. DAHM, AND M.T. CROCKER. 1992. 59–73. Climatic variability and streams and rivers in semi- BOCK, J.E., AND C.E. BOCK. 1989. Factors limiting sexual arid regions. Pages 197–202 in R.D. Roberts and reproduction in Platanus wrightii in southeastern M.L. Bothwell, editors, Aquatic ecosystems in semi- Arizona. Aliso 12:295–301. arid regions: implications for resource management. 2002] FLOOD FLOWS AND POPULATION DYNAMICS 187

NHRI Symposium Series, Environment Canada, SHELTON, M.L., AND C.J. BAHRE. 1993. Twentieth-century Saskatoon, Saskatchewan. precipitation variability and directional vegetation POFF, N.L., J.D. ALLAN, M.B. BAIN, J.R. KARR, K.L. PRESTE- change in southeastern Arizona. APCG Yearbook 55: GAARD, B.D., RICHTER, AND J.C. STROMBERG. 1997. 143–161. The natural flow regime: a paradigm for river con- SIEGEL, R.S., AND J.H. BROCK. 1990. Germination require- servation and restoration. BioScience 47:769–784. ments of key southwestern woody riparian species. PRICHARD D., H. BARRETT, J. CAGNEY, R. CLARK, J. FOGG, Desert Plants 10:3–8. K. GEBHARDT, P.L. HANSEN, B. MITCHELL, AND D. SIGAFOOS, R.S. 1976. Relations among surficial materials, TIPPY. 1998. Riparian area management: process for light intensity, and sycamore-seed germination along assessing proper functioning condition. U.S. Bureau the Potomac River near Washington, D.C. Journal of of Land Management Service Center, Denver, CO. Research of the U.S. Geological Survey 4:733–736. RIBEIRO, J., AND J. ROUSSELLE. 1996. Robust simple scal- STOKES, M.A., AND T.L. SMILEY. 1968. An introduction to ing analysis of flood peaks series. Canadian Journal tree-ring dating. University of Chicago Press, Chicago, of Civil Engineering 23:1139–1145. IL. ROOD, S.B., C. HILLMAN, T. SANCHE, AND J.M. MAHONEY. STROMBERG, J.C. 1997. Growth and survivorship of Fremont 1994. Clonal reproduction of riparian cottonwoods cottonwood, Goodding willow, and salt cedar seed- in southern Alberta. Canadian Journal of Botany lings after large floods in central Arizona. Great Basin 72:1766–1774. Naturalist 57:198–208. ROOD, S.B., A.R. KALISCHUK, AND J.M. MAHONEY. 1998. ______. 1998. Dynamics of Fremont cottonwood (Populus Initial cottonwood seedling recruitment following fremontii) and saltcedar (Tamarix chinensis) popula- the flood of the century of the Oldman River, Alberta, tions along the San Pedro River, Arizona. Journal of Canada. Wetlands 18:557–570. Arid Environments 40:133–155. RUCKS, M.G. 1984. Composition and trend of riparian ______. 2001a. Biotic integrity of Platanus wrightii riparian vegetation on five perennial streams in southeastern forests in Arizona: first approximation. Forest Ecology Arizona. Pages 97–107 in R.E. Warner and K.M. and Management 142:249–264. Hendrix, editors, California riparian systems. Uni- ______. 2001b. Influence of stream flow regime and tem- versity of California Press, Berkeley. perature on growth rate of the riparian tree, Platanus SACCHI, C.F., AND P. W. P RICE. 1992. The relative roles of wrightii, in Arizona. Freshwater Biology 46:227–240. abiotic and biotic factors in seedling demography of STROMBERG, J.C., J. FRY, AND D.T. PATTEN. 1997. Marsh arroyo willow (Salix lasiolepis, Salicaceae). American development after large floods in an alluvial, arid- Journal of Botany 79: 395–405. land river. Wetlands 17:292–300. SAKAI, T., H. TANAKA, M. SHIBATA, W. SUZUKI, H. NOMIYA, STROMBERG, J.C., AND D.T. PATTEN. 1992. Mortality and T. K ANAZASHI, S. IIDA, AND T. N AKASHIZUKA. 1999. age of black cottonwood stands along diverted and Riparian disturbance and community structure of a undiverted streams in the eastern Sierra Nevada, Quercus-Ulmus forest in central Japan. Plant Ecol- California. Madroño 39:205–223. ogy 140:99–109. STROMBERG, J.C., D.T. PATTEN, AND B.D. RICHTER. 1991. SCOTT, M.L., G.T. AUBLE, AND J.M. FRIEDMAN. 1997. Flood Flood flows and dynamics of Sonoran riparian forests. dependency of cottonwood establishment along the Rivers 2:221–235. Missouri River, Montana, USA. Ecological Applica- SVIHRA, P., AND A.H. MCCAIN. 1992. Susceptibility of plane tions 7:677–690. trees to anthracnose and powdery mildew in Califor- SCOTT, M.L., J.M. FRIEDMAN, AND G.T. AUBLE. 1996. Flu- nia. Journal of Arboriculture 18:161–163. vial process and the establishment of bottomland SWETNAM, T.W., AND J.L. BETANCOURT. 1998. Mesoscale trees. Geomorphology 14:327–339. disturbance and ecological response to decadal cli- SHAFROTH, P.B., G.T. AUBLE, J.C. STROMBERG, AND D.T. matic variability in the American Southwest. Journal PATTEN. 1998. Establishment of woody riparian veg- of Climate 11:3128–3147. etation in relation to annual patterns of streamflow, TIMONEY, K.P., G. PETERSON, AND R. WEIN. 1997. Vegeta- Bill Williams River, Arizona. Wetlands 18:577–590. tion development of boreal riparian plant communi- SHANFIELD, A.N. 1984. Alder, cottonwood and sycamore ties after flooding, fire, and logging, Peace River, distribution and regeneration along the Nacimiento Canada. Forest Ecology and Management 93:101–120. River. Pages 196–202 in R.E. Warner and K.M. Hen- WEBB, R.H., AND J.L. BETANCOURT. 1992. Climatic vari- dricks, editors, California riparian systems. Univer- ability and flood frequency of the Santa Cruz River, sity of California Press, Berkeley. Pima County, Arizona. United States Geological Sur- SHAW, R.K. 1991. Ecology of the riverbottom forest on St. vey Water Supply Paper 2379. Mary River, Lee Creek, and Belly River in south- ZIMMERMAN, R.L. 1969. Plant ecology of an arid basin, western Alberta. Pages 79–84 in S.B. Rood and J.M. Tres Alamos–Redington Area. United States Geolog- Mahoney, editors, Proceedings of the Biology and ical Survey Professional Paper 485-D. Management of Southern Alberta’s Cottonwoods Conference. University of Lethbridge, Lethbridge, Received 27 June 2000 Alberta, Canada. Accepted 13 February 2001 Western North American Naturalist 62(2), © 2002, pp. 188–196

PATTERNS OF ENERGY ALLOCATION WITHIN FORAGERS OF FORMICA PLANIPILIS AND POGONOMYRMEX SALINUS

Peter Nonacs1

ABSTRACT.— workers of Formica planipilis and Pogonomyrmex salinus were collected at various distances from their colony by pitfall traps in the former species and by catching workers recruited to food at known locations in the latter. In F. planipilis the number of larger workers increased in proportional representation to distance from the nest. Larger workers also weighed relatively less with distance from the nest, indicating that energy or water allocated for maintaining all foragers is patterned such that resource loss with forager mortality is minimized. However, the small- est size class involved in foraging shows the opposite pattern, with heavier individuals being found farther from the nest. Thus, it is possible that smaller size classes function as mobile reserves of energy or water to maintain foraging activity of larger classes at a distance from the colony. In P. salinus all sizes of workers were equally likely to be found at any dis- tance. Foragers weighed significantly less than mound workers of the same head size, again suggesting resource conser- vation in provisioning foraging workers. Differences in patterns of energy allocation to foragers are discussed relative to the species’ .

Key words: foraging, ants, Formica, Pogonomyrmex, body weight.

Ant colonies forage upon a potentially enor- would follow from such a strategy. First, more mous array of food items that can have signifi- resources could be directly invested in colony cant mortality risks in their collection. Thus, growth or sexual production. Second, fewer natural selection may favor those diets that resources would be lost with the death of any best balance between collecting food and ex- forager. posure to mortality risks. At a community level The evidence suggests that foragers do, on this may lead to species partitioning their average, carry lower resource reserves than do habitat by prey size (Kaspari 1996) or activity nonforaging workers. Porter and Jorgensen patterns (Savolainen and Vepsäläinen 1988, (1981) found a consistent decline in body 1989). Within a species, combinations of mor- weight of Pogonomyrmex owyheei workers phological and behavioral responses are possi- from the interior of nests, to those working on ble. Many species have polymorphic worker the nest surface, and then to foragers. In Lep- castes that are distinctly different in size tothorax albipennis lipid reserves closely cor- (Hölldobler and Wilson 1991). Worker poly- relate with behavioral roles such that foragers morphism can allow foragers to specialize on are almost inevitably “lean” (Blanchard et al. particular types of foraging tasks, but such task 2000). Similarly suggesting that they carry less specialization according to caste is not univer- food, Myrmica rubra foragers kept without sally observed (Traniello 1989, McIver and food died faster than nonforagers (Weir 1958), Loomis 1993, Ferster and Traniello 1995). and old foragers in Formica rufa died faster In conjunction with morphology, behav- than young ones (Rosengren 1977). ioral adaptations can also increase the net ben- If it is adaptive for foragers to carry fewer efits from foraging. One such behavior is to reserves than interior workers, it may also be preferentially forage in those areas that maxi- adaptive to have a finer-scale subdivision of mize net gain (as measured by colony growth resources across foragers in relation to their minus worker mortality; Nonacs and Dill 1990, expected level of mortality risk. As an example, 1991). A 2nd adaptation is for foragers to main- risk could positively correlate with distance tain lower reserves of valued resources, such traveled from the nest. Although no relation- as food reserves (carbohydrates, proteins, or ship should be expected to be invariant, on lipids) and water in arid habitats. Two benefits average, it is probable that the farther an ant

1Department of Organismic Biology, Ecology and Evolution, University of California, Los Angeles, CA 90095.

188 2002] ENERGY ALLOCATION WITHIN FORAGERS 189 travels, the more likely it is to encounter preda- 1976, Rosengren and Sundström 1987, Traniello tors or aggressive competitors and the greater 1989, Fewell 1990, Gordon et al. 1992, McIver is its exposure to the elements. Also, the far- and Yandell 1998). Thus, it is possible for a ther an ant is from its nest entrance, the more forager to “know” how far it is going and how chances it will have of becoming lost and long it is likely to forage and then to adjust unable to return quickly to the nest if environ- either its initial energy or water reserves. From mental conditions deteriorate. In addition to a colony standpoint of conserving resources or differing levels of risk, the type of risk may workers, adjusting resource levels throughout also affect patterns of resource distribution in the territory or just at colony departure would the forage caste. If the preponderance of for- reflect the same overall goals. The former ager deaths is likely to result from starvation mechanism might allow finer partitioning in or dehydration, a “worker-conservative” strat- the patterns. egy might be favored. In such a scenario, work- Testing the above scenarios requires mea- ers traveling the farthest may carry relatively suring weights of unsuccessful foragers so as more food and water as insurance against get- to avoid confounding food reserves with food ting lost or otherwise delayed in replenishing to be returned to the colony. With this in mind, their resources. However, if risk or I used 2 experimental methods to sample pop- stochastic environmental events cause the pre- ulations of foragers at various distances from ponderance of mortality, then colonies may their colonies. In the first I set out pitfall traps employ a “resource-conservative” strategy in at specific distances and placed them in areas which those foragers most at risk carry the likely to trap insect-scavenging workers. In least reserves. the 2nd method I set out food baits at mea- Resources can be categorized in numerous sured distances and collected recruiting work- ways, but relevant to foraging, the important ers. The food (cracker bits) was attractive to distinction may be the extent to which the ants but could not be consumed on the spot. particular resource can be rapidly increased or The 2 methodologies differed in that the 1st exchanged across individuals. For example, trapped ants that may have been continuously lipids are energy rich but relatively slow in foraging for some time, and the 2nd generally being converted to metabolic needs; also, they captured ants that were recently recruited cannot be exchanged by trophollaxis between from the colony and probably not foraging replete and starving foragers. Conversely, aque- immediately prior to recruitment. Wet weights ous carbohydrates can be relatively quickly were measured as the best estimator of forag- metabolized or exchanged between workers ing reserves (water and carbohydrates dis- when needed. Thus, protein and lipids appear solved therein). to be shunted into colony growth, while colony respiration and foraging activities are fueled STUDY AREA AND SPECIES by carbohydrates (Brian 1973, Sudd 1987). Indeed, foragers seem to have minimal fat re- The experiments were conducted in May serves, which would conserve those particular 1999 at the Sierra Nevada Aquatic Research resources (Porter and Jorgensen 1981, Blan- Laboratory (SNARL) in the Great Basin chard et al. 2000). Additionally, in xeric habi- Desert Province (Franklin and Dyrness 1973). tats water could be a critical resource for The site is at approximately 2000 m elevation, which the above trade-off is important. With and the weather during the study was gener- more water, dessication deaths are minimized, ally warm (daytime highs 20–25°C) and sunny, but each death results in a greater loss of with no more than trace amounts of rainfall. resources. Vegetation at the site is dominated by sage- Patterns of food reserves among foragers brush (Artemisia tridentata) and sparse grass can be set and maintained by trophollaxis cover. between nestmates. Similar patterns could The study species were Formica planipilis also result without trophollaxis if workers left and Pogonomyrmex salinus (keyed from Wheeler the colony with energy reserves based on the and Wheeler 1983). Formica planipilis is a distance they expected to travel. Site speci- thatching ant that builds mounds of plant mat- ficity or allegiance in foragers is known from a ter as part of its nests. Colonies often have number of species (Rosengren 1971, Hölldobler trunk trails leading to foraging sites where 190 WESTERN NORTH AMERICAN NATURALIST [Volume 62 they collect honeydew. Individual foragers also around a colony until at least 20 foragers were scavenge arthropods (McIver and Loomis 1993). captured for each distance. However, for one Pogonomyrmex salinus is a seed-harvester that colony (#5), no foragers were collected at 10 builds low gravel mounds. The species recruits m. All collected workers were immediately foragers to food sites but does not form well- taken back to the laboratory, killed by freezing marked trunk trails like other species of seed- for 10 minutes, and then individually harvesters (personal observation). Both species weighed. are very common in the immediate vicinity of I also collected 12 F. planipilis workers from SNARL. each of the nest mounds in the morning be- For the study I marked out 11 nests of F. tween 0800 and 0900 hours. These were not planipilis and 8 nests of P. salinus that were random samples, as I tried to represent equally large, actively foraging, and had no conspecific a wide range of worker sizes. The ants were neighboring colony within 15–20 m in at least immediately taken to the laboratory and one direction. None of the F. planipilis colonies weighed. Thereafter, they were placed in plas- was within 200 m of a permanent water source. tic containers (2 ants per container) and put Thus, captured F. planipilis workers were not outside in the shade for the day. In the late involved in water collection, and differences afternoon (between 1600 and 1700 hours), the in wet weight would not reflect encounters ants were killed by quick freezing and weighed with standing water. Several P. salinus colonies again to measure weight loss over the length all were within 20 m of a running stream. of a foraging day. However, because only workers recruiting to a The pitfall trap method could not work for food source were collected, wet weight again the P. salinus colonies because of higher colony did not reflect water collection. densities (i.e., at 10 m traps were likely to con- tain foragers from more than a single colony) METHODS and lower aggression levels (i.e., known for- eign workers transferred to another mound Pitfall traps were laid out at distances of 1, were not always obviously attacked). Instead, I 4, 7, and 10 m from the center of the mound placed bright orange cracker crumbs at dis- for each of the 11 F. planipilis colonies. The tances of 5 or 10 m. Only one distance was traps were plastic trays with their sides coated presented to a colony at a time. All colonies with Fluon. Stones were piled in the center of located the crumbs at 5 m within 20–30 min- the tray to provide shade from the sun. Ants utes and foragers were actively recruited. Six were collected from the traps in the morning of 8 colonies also quickly found the crumbs at and late afternoon. Pitfall sites were located 10 m, but in 2 colonies the crumbs were away from trunk trails and bushes where for- always removed by other ant species before P. agers were actively collecting nectar or honey- salinus could recruit in numbers. Thus, for 2 dew. Thatching ants appear to form strong site colonies there are data only for 5 m. After allegiances that can be as restrictive as a single allowing the recruitment process to begin, I bush (McIver and Yandell 1998). Thus, it is collected between 18 and 30 foragers return- likely that most ants trapped were scavenging ing with crumbs at each distance. Thus, it is or otherwise patrolling the territory and not likely that most ants collected were inside the aphid tenders or nectar transporters. On occa- colony prior to collection rather than outside sion new recruitment trails would form near a foraging. trap, and many ants would be found in a sam- For P. salinus colonies I also collected 20– ple. I released these ants and moved the trap. 30 workers per nest that were working at the When emptying the pitfall traps at 7 and 10 m, mound (either moving pebbles from the nest I also collected some foragers from the imme- or patrolling the entrance). These samples diate vicinity and placed them on the pre- probably accurately reflect the population of sumed nest mound. If they were not attacked, mound workers, as I would collect the ants in I assumed the collected workers in the traps the order they exited the nest. All collected belonged to the marked colony. If attacked, I ants were immediately placed in 95% alcohol. released any caught ants and moved the traps. Although this may affect water and lipid con- Dead F. planipilis foragers found in the traps tent to some extent, I assumed the effects were were discarded. Traps were kept in place unbiased across ant sizes. (Note that since ant 2002] ENERGY ALLOCATION WITHIN FORAGERS 191 size was eventually found not to vary across distance, the results are robust even if the assumption does not hold.) Later the ants were briefly dried from the alcohol and then individually weighed. For both species morpho- metric size for each individual was estimated as its widest headwidth, which was measured with a calibrated dissecting microscope to the nearest 0.04 mm. For the majority of the statistical analyses, I combined headwidths into discrete size classes based primarily on where the mean mass increased more rapidly with headwidth. In F. planipilis, I categorized the distribution into 5 size classes that are in approximate agreement with McIver and Loomis’s (1993) 7 size classes, the difference being that I collected very few individuals in their 2 smallest size classes. In P. salinus, I categorized the ants into 4 size classes. Where needed, I ln-transformed body weight as weight increases cubically with lin- ear increases in headwidth.

RESULTS

For both species there is a strong positive correlation between headwidth and body weight (Fig. 1). In F. planipilis there is a skewed size distribution toward larger body size in foragers. In P. salinus intermediate body sizes are the most common. For both species I did ANOVAs on the per- Fig. 1. Mean weight (±s) of each headwidth class and centage of ants, with factors of size class and number of ants collected in each class for Formica planip- distance from nest. In F. planipilis there is a ilis (A) and Pogonomyrmex salinus (B). Data for F. planip- ilis are those only for ants collected foraging and not at the significant effect of size class (F = 124.1; df = nest mound. Dashed lines indicate where size class breaks 4,195; P < 0.0001), which reflects the relation- were added for ANOVA analyses. ship in Figure 1A of larger ants being more abundant in pitfall traps than smaller ants. There is no effect of distance of pitfall traps, class (i.e., physically larger ants weigh more: F but there is a significant interaction between = 188.7; df = 3,565; P < 0.0001) and distance size class and distance from nest (F = 3.199; (F = 8.834; df = 2,565; P = 0.0002). The lat- df = 12,195; P = 0.0003). This results from ter result comes from ants of all size classes the smallest size class being rarely trapped far weighing less when foraging than when work- from the colony and larger size classes becom- ing around the nest (Fig. 3). There was, how- ing relatively more common (Fig. 2A). For P. ever, no significant interaction between the 2 salinus there is a significant effect only of size factors, which indicates that all size classes class (F = 4.441; df = 3,76; P = 0.0063). Dis- behave similarly regarding their body weight. tance does not have a significant effect, and all A Scheffe-S post hoc test found body weights size classes are equally likely to work around of ants at 5 and 10 m are significantly less than the nest entrance or forage at distance (Fig. 2B). at 0 m (P < 0.012 for both comparisons). On Because there is no size class:distance re- average ants weighed more at 5 than at 10 m, lationship in P. salinus, I analyzed forager but the trend was not significant. The differ- weights directly. An ANOVA of ln-transformed ence in weights between ants at the nest and body weight found a significant effect of size foragers is unlikely to be due to foraging activity 192 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 3. Mean weight (±s) of each headwidth class col- lected at 0, 5, or 10 m from the nest mound in Pogono- myrmex salinus. Ants collected at 0 m were not foraging, and those at 5 or 10 m were foraging.

workers being relatively more abundant at greater distances. The ANOVA of residuals found a significant effect of distance (F = ± Fig. 2. Percentage of ants ( sx–) in each headwidth class 2.469; df = 4,1139; P = 0.0432), no effect of collected at set distances from the nest entrance for Formica planipilis (A) and Pogonomyrmex salinus (B). size class (F = 0.232; NS), and a significant interaction between size class and distance (F = 2.124; df = 2,1139; P = 0.0134). Thus, alone. Fewell (1988) estimated that a foraging workers are overall likely to weigh less farther trip of approximately 12 m in P. occidentalis from the nest, but this relationship is not the costs 0.088 J. If foragers use a liquid equiva- same for all size classes. As can be seen in Fig- lent of 20% sugar water as their energy source, ure 4A, individuals in the smaller size class this would translate into an estimated biomass (1.23–1.43 mm) tend to weigh more farther loss of 0.03 mg (using a value of 3.2 J of energy from the nest, while the larger size classes per mg of sugar water) to travel 10 m. In com- show the opposite relationship. parison, the mean weight differential between The basis of the distance effect on F. planip- ants at 0 m and those collected at 5 and 10 m ilis worker weights can be estimated by ad- is 0.33 and 0.43 mg, respectively. justing observed body weights with the weight Because there is a size class:distance rela- loss measured in the control series. When tionship in F. planipilis, forager weights cannot kept the entire day in a container without be analyzed directly. An effect of distance can access to food or water, the average ant loses come from either larger workers going differ- 6.31% (±1.5 s) of its body mass. A regression of ent distances or workers within any size class weight loss against headwidth was not signifi- carrying different amounts of food depending cant (F = 0.092; NS), indicating that ants of on the distance they forage. Therefore, I cal- different sizes do not lose relatively different culated a mean body weight for each measur- proportions of their body weight over time. able headwidth gradation. Subtracting these Therefore, I adjusted upward the weights of means from each individual ant’s weight gives all ants proportional to the distance they were a residual value that can be statistically ana- trapped from the colony. Ants trapped at 1 m lyzed. (Note that the smallest size class had to and 10 m had their weights increased by be dropped from this analysis because such 0.631% (= 6.31 × 0.1) or by 6.31% (= 6.31 × 1), ants were absent or rarely collected at 7 and respectively. Ants caught at intermediate dis- 10 m.) The residuals are not significantly tances were similarly adjusted proportionally. affected by headwidth (F = 0.003; NS), which The underlying assumption for these adjust- removes the effect of morphologically larger ments is that, on average, ants caught farther 2002] ENERGY ALLOCATION WITHIN FORAGERS 193

without the adjustment, but no effect with the adjustment, this would leave open the possi- bility that all foragers leave with about the same energy reserves and that the distance effect is due to differential energy expendi- ture. Finally, if after adjustment the distance effect is positive, this would imply that for- agers going farther leave the colony with pro- portionally more energy reserves. The data of unadjusted body weights in F. planipilis suggest that different size classes of foragers carry different levels of energy re- sources (Fig. 4A). Therefore, I regressed ad- justed residual weights against distance for- aged (1–10 m) separately for each size class. Only the smallest size class shows a significant effect of distance: workers are more likely to weigh more farther from the colony (Fig. 4B). For all other size classes, once weight was ad- justed upward, there was no significant effect of distance from colony. The previous analyses of F. planipilis data viewed individual ants as independent data points, although they were all members of only 11 colonies. To rule out any problems with pseudoreplication, I also correlated the residuals with distance at the colony level for ± Fig. 4. Mean residuals ( sx–) of Formica planipilis head- each of the 4 largest size classes. If there is no width classes collected on the nest mound or 1–10 m dis- weight-distance relationship, then the overall tant pitfall traps. In (A) recorded weights are given. In (B) mean of the correlation coefficients should not weights from pitfall traps have been adjusted upward as might be expected if foragers at more distant sites had be significantly different from zero. With the spent more time traveling (see text). Only the smallest unadjusted data, the 3 largest size classes all headwidth had a significant regression of residual values show a significantly negative relationship be- with distance from nest mound. tween body weight and distance from colony (Table 1). The smallest size class shows a posi- tive relationship that approaches significance from the colony have been foraging longer and (P = 0.0621). When weight is adjusted upward, thus may have lost more weight. The magni- there is a significant positive relationship be- tude of the adjustment is an approximation. tween weight and distance for the smallest Under natural conditions, ants probably move size class considered, but no significant rela- more than they do in containers, which would tionships in the 3 largest size classes (Table 1). lead to underestimating weight loss. However, Therefore, when the F. planipilis data are ana- this is likely to be balanced in the field by the lyzed at the colony level, results are consistent fact that foraging workers could get food from with the analysis at the individual level. others or return to the colony for more food. Reanalyzing the data after adjustment leads DISCUSSION to several possible outcomes, each of which would imply something about the foraging Foraging is a relatively dangerous activity process. If distance has a negative effect on for ants, and foragers can be considered body weight both without and with the adjust- expendable or even disposable (Porter and ment, it would imply that foragers going far- Jorgensen 1981). Estimates of mean life spans ther from the nest begin their trips with less for Pogonomyrmex foragers are about 14 days energy reserves than those foraging at closer (Porter and Jorgensen 1981), and few foragers distances. If distance has a negative effect of the thatching ant, F. obscuripes, were 194 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 1. One-sample t tests of correlation coefficients of residual value versus distance from colony across 11 colonies of F. planipilis. Nonadjusted tests have original data, and adjusted tests have data corrected for possible loss of weight due to activity (see text).

______Nonadjusted ______Adjusted Size class Mean t-value (df) P Mean t-value (df) P 1.23–1.43 0.261 2.100 (10) 0.0621 0.413 2.509 (9) 0.0334 1.44–1.64 –0.125 –2.843 (10) 0.0174 0.035 1.105 (10) NS 1.65–1.81 –0.211 –4.588 (10) 0.0010 –0.064 –1.110 (10) NS >1.81 –0.161 –2.524 (10) 0.0302 0.015 0.200 (10) NS observed for longer than 20 days (McIver and In contrast to larger size classes, smaller F. Yandell 1998). Given an expected high loss planipilis workers have a positive relationship rate of foragers, it is adaptive for colonies to between weight and distance from the nest minimize the resources lost with each worker. that is strengthened by weight adjustments for For foragers in xeric habitats, resources may assumed differences in activity. The pattern in equate to some combination of water and dis- these workers is most consistent with mini- solved carbohydrates (Brian 1973, Sudd 1987, mizing their loss rather than the resources Lighton et al. 1994). Porter and Jorgensen they carry. Thus, colony-level strategy seems (1981) found that P. owyheei foragers weighed diametrically opposite in small versus large less than those doing mound work, which cor- size classes of foragers. Four hypotheses may relates with their task-related mortality rates. account for this. First, smaller workers may Data from P. salinus show a similar pattern: get lost more often because of their relatively foragers weigh less than mound workers. In P. poorer orientation abilities (Bernstein and Bern- salinus there is weak evidence for further con- stein 1969, McIver and Loomis 1993). If they servation of resources in that the mean weight are more likely to make navigation errors, this of foragers traveling 10 m is less than for those could favor carrying extra resources. Second, traveling 5 m (Fig. 3). However, this effect is because smaller workers have a larger surface- not statistically significant. to-volume ratio, they may have to carry more There is stronger evidence for distance- water because of a relatively higher dessica- related foraging in F. planipilis. Overall, larger tion risk (Lighton et al. 1994). Lighton et al. workers formed a larger fraction of the work (1994) found that smaller workers lose water at force farther from the nest (Fig. 2A). This faster rates than do larger workers in Messor result differs from McIver and Loomis (1993), pergandei. However, there was no such size- who found a relationship of worker size with based effect in the control measurements with distance from colony in nectar-collecting F. planipilis. Smaller workers did not dessicate workers, but not scavenging workers. Differ- more rapidly than larger workers. Third, food ences between the studies may result from may move differentially between smaller and differences in behavior across species (F. pla- larger workers during transport to the colony. nipilis vs. F. obscuripes) or in collection tech- McIver and Yandell (1998) found that smaller niques (pitfall traps vs. collection by sight). A thatching ants tending aphids will transfer 2nd distance effect in F. planipilis is that body honeydew to larger ones, which transport it weights of the 3 largest size classes of workers back to the colony. However, for this to result decline significantly with distance from the in the observed patterns in F. planipilis, the nest. This pattern could result from workers exchanges would have to occur more fre- either maintaining lower reserves at greater quently on a per capita basis closer to the distances or losing proportionately more weight colony, and it is not obvious why this should on longer foraging trips. Both possible scenar- be the case. Moreover, pitfall traps were ios are consistent with minimizing resource located so as to avoid workers collecting hon- loss rather than worker loss. There is no sup- eydew; thus, it is likely that most workers port for large foragers increasing food or water were not involved in the immediate transport reserves, which would be expected if forager of food back to the colony. Fourth, food may deaths were due mostly to starvation or dehy- move differentially between smaller and larger dration. workers, but not in the context of transport to 2002] ENERGY ALLOCATION WITHIN FORAGERS 195 the colony. Instead, small workers may form a range in workers could result simply from resource pool from which larger foragers can developmental noise. Formica planipilis also draw resources when needed. Smaller workers specializes in nectar collection, while P. sali- form a smaller proportion of the work force as nus forages mostly on seeds. The combination distance from the nest increases (Fig. 1A), and of a larger evolved range of worker sizes and if they serve as “gas stations,” then the results access to an easily divisible food source may are consistent with both resource- and worker- favors more fine-scaled adjustments of forag- conserving strategies. Relatively fewer stations ing strategy in F. planipilis than in P. salinus. It should be found far from the nest where they is interesting to note that in the genus Formica are most at risk, and each station should have there is fairly consistent size-based specializa- relatively more fuel. tion in foraging tasks (Herbers 1979, Rosen- Comparing across F. planipilis and P. salinus, gren and Sundström 1987, Sundström 1987, there are both similarities and differences. In McIver and Loomis 1993, McIver and Yandell both species foragers as a class weigh less than 1998), while in seed-harvesting ants (Pogono- mound workers. In F. planipilis larger workers myrmex and Messor) results are far more mixed form an increasingly greater proportion of the and mostly negative (see reviews in Traniello work force with distance from the nest, and 1989, Ferster and Traniello 1995). there appear to be significant differences in In conclusion, variation in body weight of the amount of energy workers carry depend- ants in F. planipilis and P. salinus at least sug- ing on their class size and the distance they gests that all foragers are not equal in their expect to travel. In P. salinus there is no evi- distribution of energy reserves. The mecha- dence of the former and only weak support for nisms by which food is moved through and distance affecting foragers’ energy reserves. shared by a population of foragers are yet to These differences across the species may result be elucidated. The efficiency with which an from either differences in the foraging behav- ant colony collects food must depend on how ior sampled or in the species characteristics. much energy is allocated to each of the dispos- In F. planipilis the collected ants were most able elements and how well a colony balances likely on open-ended foraging trips that might between maximizing worker survival and re- continue for some time until they found prey source conservation. or returned to the colony. If foragers far from the nest are more at risk per trip, then signifi- ACKNOWLEDGMENTS cant modulation of energy reserves might be Ken Nagy kindly provided the energetic expected. In P. salinus most collected ants were calculations for the cost of a foraging trip. I probably recruited from within the colony. If thank Jeff Thomas for help in the field and foragers travel directly to the food patch and comments on the work, Dan Dawson and back, there may be relatively little difference SNARL for the use of their facilities, and in energy requirements (Fewell 1988) or expo- anonymous reviewers for helpful comments. sure time. If the mortality risk for foragers is more a factor of time spent foraging than dis- LITERATURE CITED tance traveled, then the ants recruited to food at 10 m will be only marginally more at risk BERNSTEIN, S., AND R.A. BERNSTEIN. 1969. Relationship than ants going 5 m. Thus, in recruited work- between foraging efficiency and the size of the head ers there may be only a weak relationship be- and component brain and sensory structures in the red wood ant. Brain Research 16:85–104. tween distance and resource reserve. BLANCHARD, G.B., G.B. ORLEDGE, G.M. REYNOLDS, AND Besides differences in sampling, the species N.R. FRANKS. 2000. Division of labour and seasonal- have different life history characteristics. For- ity in the ant Leptothorax albipennis: worker corpu- mica planipilis has polymorphic workers that lence and its influence on behaviour. Animal Behav- iour 59:723–738. show monophasic allometry (Hölldobler and BRIAN, M.V. 1973. Feeding and growth in the ant Myrmica. Wilson 1991). There is a significant propor- Journal of Animal Ecology 42:37–53. tional bias toward the larger size classes in for- FERSTER, B., AND J.F.A. TRANIELLO. 1995. Polymorphism aging (Fig. 1A) and size-based task specializa- and foraging behavior in Pogonomyrmex badius (Hy- menoptera: Formicidae): worker size, foraging dis- tion (McIver and Loomis 1993). Pogonomyr- tance, and load size association. Environmental mex salinus is monomorphic, and the size Entomology 24:673–678. 196 WESTERN NORTH AMERICAN NATURALIST [Volume 62

FEWELL, J.H. 1988. Energetic and time costs of foraging caste? Behavioral Ecology and Sociobiology 9: in harvester ants, Pogonomyrmex occidentalis. Behav- 247–256. ioral Ecology and Sociobiology 22:401–408. ROSENGREN, R. 1971. Route fidelity, visual memory and ______. 1990. Directional fidelity as a foraging constraint recruitment behaviour in foraging wood ants of the in the western harvester ant, Pogonomyrmex occi- genus Formica (Hymenoptera, Formicidae). Acta dentalis. Oecologia 82:45–51. Zoologica Fennici 133:1–105. FRANKLIN, J.F., AND C.T. DYRNESS. 1973. Natural vegeta- ______. 1977. Foraging strategy of wood ants (Formica tion of Oregon and Washington. USDA Forest Ser- rufa group) I. Age polyethism and topographic tradi- vice General Technical Report PNW-8. tions. Acta Zoologica Fennici 149:1–30. GORDON, D.M., R. ROSENGREN, AND L. SUNDSTRÖM. 1992. ROSENGREN, R., AND L. SUNDSTRÖM. 1987. The foraging The allocation of foragers in red wood ants. Ecologi- system of a red wood ant colony (Formica s. str.)— cal Entomology 17:114–120. collecting and defending food through an extended HERBERS, J.M. 1979. Caste-biased polyethism in a mound- phenotype. Pages 117–137 in J.M. Pasteels and J.L. building ant species. American Midland Naturalist Deneubourg, editors, From individual to collective 101:69–75. behavior in social insects. Birkhäuser Verlag, Basel. HÖLLDOBLER, B. 1976. Recruitment behavior, home range SAVOLAINEN, R., AND K. VEPSÄLÄINEN. 1988. A competition orientation and territoriality in harvester ants, hierarchy among boreal ants: impact on resource Pogonomyrmex. Behavioral Ecology and Sociobiol- partitioning and community structure. Oikos 51: ogy 1:3–44. 135–155. HÖLLDOBLER, B., AND E.O. WILSON. 1991. The ants. Belk- ______. 1989. of ant species within nap Press of Harvard University Press, Cambridge, territories of the wood ant Formica polyctena. Oikos MA. 732 pp. 53: 3–16. KASPARI, M. 1996. Worker size and seed selection by har- SUDD, J.H. 1987. Individual behaviour and mixed diet vester ants in a Neotropical forest. Oecologia 105: strategy in ants. Pages 81–92 in J.M. Pasteels and 397–404. J.L. Deneubourg, editors, From individual to collec- LIGHTON, J.R.B., M.C. QUINLAN, AND D.H. FEENER, JR. tive behavior in social insects. Birkhäuser Verlag, 1994. Is bigger better? Water balance in the poly- Basel. morphic desert harvester ant Messor pergandei. SUNDSTRÖM, L. 1987. Size allocation and task performance Physiological Entomology 19:325–334. in red wood ants (Formica). Page 120 in J. Eder and MCIVER, J.D., AND C. LOOMIS. 1993. A size-distance rela- H. Rembold, editors, Chemistry and biology of social tion in Homoptera-tending thatch ants (Formica insects. Verlag J. Peperny, München. obscuripes, Formica planipilis). Insectes Sociaux TRANIELLO, J.F.A.1989. Foraging strategies of ants. Annual 40:207–218. Review of Entomology 34:191–210. MCIVER, J.D., AND K. YANDELL. 1998. Honeydew harvest WEIR, J.S. 1958. Polyethism in workers of the ant Myr- in the western thatching ant (Hymenoptera: Formi- mica, part II. Insectes Sociaux 5:315–339. cidae). American Entomologist 44:30–35. WHEELER, G.C., AND J.N. WHEELER. 1983. The ants of NONACS, P., AND L.M. DILL. 1990. Mortality risk vs. food Nevada. Natural History Museum of Los Angeles quality trade-offs in a common currency: ant patch County. Los Angeles, CA. 138 pp. preferences. Ecology 71:1886–1892. ______. 1991. Mortality risk versus food quality trade-offs Received 29 June 2000 in ants: patch use over time. Ecological Entomology Accepted 20 February 2001 16:73–80. PORTER, S.D., AND C.D. JORGENSEN. 1981. Foragers of the harvester ant, Pogonomyrmex owyheei: a disposable Western North American Naturalist 62(2), © 2002, pp. 197–205

MOVEMENT, DISTRIBUTION, AND PREDATION: LEPIDOMEDA VITTATA AND NONNATIVE SALMONIDS IN EASTERN ARIZONA

Michael G. Sweetser1,2, Scott D. Bryan1, and Anthony T. Robinson1

ABSTRACT.—Nonnative rainbow trout (Oncorhynchus mykiss) are stocked into several reservoirs in the range of fed- erally threatened Little Colorado spinedace (Lepidomeda vittata), and so have the opportunity to negatively impact Little Colorado spinedace populations. We examined rainbow trout escapement from Nelson Reservior into Nutrioso Creek, critical habitat for L. vittata. We also examined movements of L. vittata and incidence of predation by rainbow trout on L. vittata. We detected no movement of rainbow trout out of Nelson Reservoir over 4 years of study. Lepidomeda vittata marked in 3 streams did not move much; but sample sizes were too small to make any meaningful conclusions regarding movement. Most L. vittata we captured during surveys subsequent to marking were unmarked, suggesting movement out of the study area, low tag retention, mortality, or failure to capture marked fish. Lepidomeda vittata co-occurred with O. mykiss, Salmo trutta, and Salvelinus fontinalis and were typically less than half the size of the sympatric nonnative salmonids. Consequently, they are potential prey fish for these species. We found fish remains in stomachs of 33% of S. trutta, 6% of O. mykiss and 25% of S. fontinalis examined, but remains of L. vittata were found only in a single S. trutta. Because S. fontinalis are rare in the streams examined, they probably do not pose a great threat to L. vittata. Salmo trutta, which are no longer stocked, had the highest piscivory level and may thus pose more of a threat to L. vittata than O. mykiss.

Key words: Lepidomeda vittata, Little Colorado spinedace, salmonids, distribution, movement, predation.

Native southwestern fishes have declined predation in the wild is unknown. Distribu- in part due to negative interactions with in- tions of L. vittata and O. mykiss overlap, and troduced species (Meffe 1984, Moyle 1986, historic distribution of L. vittata likely over- Minckley and Deacon 1991). Other factors lapped with native Apache trout (O. apache). that contribute to declines (Minckley 1973) However, O. mykiss may be more aggressive include drought, habitat loss and degradation, and piscivorous than O. apache, and stocking pollution, and poisoning (Miller 1961, 1963, may result in higher trout densities than natu- Minckley and Carufel 1967). Little Colorado rally occurred. spinedace (Lepidomeda vittata), a cyprinid Lepidomeda vittata was federally listed as endemic to the Little Colorado River basin in threatened, and critical habitat was designated eastern Arizona, is an example of a native fish in 1987 (USDI 1987). Critical habitats include that has declined since Europeans settled the Nutrioso Creek between Nelson Reservoir area in the late 1800s (Miller 1963, Minckley and the Little Colorado River, and East Clear and Carufel 1967, Minckley 1973). Creek from Blue Ridge Reservoir downstream Nonnative trouts likely impact L. vittata to Leonard Canyon. Because of suspected im- since these nonnatives are predaceous and pacts of O. mykiss stocking on L. vittata, Arizona have been extensively stocked. Rainbow trout Game and Fish Department (AGFD) began (Oncorhynchus mykiss), cutthroat trout (O. section 7 consultation with U.S. Fish and clarki), brook trout (Salvelinus fontinalis), and Wildlife Service in 1994. As a consequence of brown trout (Salmo trutta) have been stocked that consultation, AGFD altered its O. mykiss into the Little Colorado River basin since the management in the upper Little Colorado early 1900s (Miller 1961, Rinne and Janisch River basin; stockings within L. vittata range 1995), but only O. mykiss still is stocked. were restricted to reservoirs and limited to Oncorhynchus mykiss and S. trutta prey on L. spring–summer after dams ceased to spill vittata in stream enclosures (Blinn et al. 1993, from spring runoff. Fishing regulations were Rinne and Alexander 1995), but the extent of changed in 1998 to allow unlimited take of O.

1Arizona Game and Fish Department, 2221 W. Greenway Road, Phoenix, AZ 85023. 2Corresponding author. Present address: PO Box 1931, Springerville, AZ 85938.

197 198 WESTERN NORTH AMERICAN NATURALIST [Volume 62 mykiss and S. trutta from 1 September through northwest of Springerville, Arizona (Fig. 1). 1 May in streams designated as critical habitat Fishes here are L. vittata, R. osculus, P. discobo- for L. vittata (including Nelson Reservoir, lus, Little Colorado sucker (Catostomus sp.), P. Blue Ridge Reservoir, and Knoll Lake). promelas, O. mykiss, S. trutta, and L. cyanellus Another response to the consultation was (AGFD unpublished data). research to evaluate effects of O. mykiss on L. vittata. We report on 4 elements of that METHODS research. Objectives were to (1) determine if O. mykiss stocked into Nelson Reservoir move Oncorhynchus mykiss were marked with into Nutrioso Creek, (2) document L. vittata coded-wire tags (snout or below the adipose movements, (3) determine incidence of preda- fin; 1996 and 1997) or tetracycline (1997 to tion by trouts on L. vittata, and (4) document 2000) and stocked into Nelson Reservoir each L. vittata and trout distributions in Nutrioso spring immediately after reservoir overflow Creek critical habitat and adjacent study ceased (typically in May). Creel surveys were reaches. conducted 1996 through 1999 on Nelson Reser- voir to estimate percentage of salmonids re- STUDY SITES moved by angling. Nelson Reservoir was sampled with gill nets Nutrioso Creek, Rudd Creek, and Little (45-m; experimental) set overnight (April/May Colorado River in east central Arizona (Fig. 1) 1997 to 2000) to determine holdover rates were sampled from 1996 through 2000. Sam- from the previous year’s stocking. To deter- pling was on U.S. Forest Service and AGFD mine if L. vittata occupied the lake, we set 4 lands. Nutrioso Creek heads in coniferous for- trap nets (122 cm × 122 cm, 6-m middle wing) est and flows through a meadow from the perpendicular to and opening toward shore in town of Nutrioso to Nelson Reservoir, approxi- random locations at Nelson Reservoir for 2 mately 20 km from the headwaters. Nelson nights in summer 1997. Reservoir is managed by AGFD as a put- We sampled Nutrioso Creek twice each and-take O. mykiss fishery. The coniferous year (spring = pre-stocking, late summer to forest reach of Nutrioso Creek is occupied by early autumn = post-stocking) using a back- O. mykiss, S. fontinalis, and speckled dace pack electrofisher to assess escapement of O. (Rhinichthys osculus). Fishes in the meadow mykiss from Nelson Reservoir. Fish captured portion are L. vittata, R. osculus, bluehead were identified, measured (TL, mm), and sucker (Pantosteus discobolus), O. mykiss, S. weighed (g); location of capture (m below dam trutta, fathead minnow (Pimephales promelas), or above reservoir) also was recorded. All and green sunfish (Lepomis cyanellus; AGFD trouts were sacrificed and scanned for a coded- unpublished data). Nutrioso Creek is canyon- wire tag or tetracycline mark. bound below Nelson Reservoir, and the fish During 1996, 1999, and 2000, we surveyed assemblage is L. vittata, R. osculus, P. discobo- Nutrioso Creek below Nelson Reservoir by lus, P. promelas, O. mykiss, S. trutta, and S. fonti- electrofishing, making one pass through the nalis (AGFD unpublished data). The study reach. During 1997 and 1998, 18 random 50-m area extends 11.5 km below and 8.3 km above sites were sampled: 9 pre- and 9 post-stock- the reservoir (Fig. 1). ing. Block nets (3.2-mm mesh) were placed at Rudd Creek is approximately 13 km long up- and downstream boundaries, and 3 passes and joins Nutrioso Creek about 2 km down- were made. We similarly sampled twelve 50-m stream of Nelson Reservoir. A man-made fish sites in Nutrioso Creek above Nelson Reser- barrier (culvert through a cement dam with a voir, 1998 through 2000. 3-m-high waterfall) is 4.5 km above the conflu- To assess movements of L. vittata, we ence. The upper reach is occupied by O. my- marked fish (42–128 mm TL) prior to spring kiss and S. fontinalis, while L. vittata, R. oscu- runoff in selected reaches of each stream and lus, P. discobolus, O. mykiss, and S. fontinalis subsequently surveyed there as well as up- occupy the lower (AGFD unpublished data). and downstream (Table 1). Marking reaches We sampled only the lower 4.5 km (Fig. 1). had greater L. vittata densities than other Also sampled was 3.3 km of Little Colorado reaches as indicated by previous sampling. River within AGFD’s Wenima Wildlife Area, Reaches were separated by a distance equal to 2002] L. VITTATA: MOVEMENTS, DISTRIBUTION, PREDATION 199

Fig. 1. Map of the study area in east central Arizona. Shaded portions indicate reaches sampled.

TABLE 1. Stream reaches where L. vittata were marked (April). Surveys were conducted during May, July, and Sep- tember 1998; May and September 1999; and September 2000. Reaches are kilometers above Nelson Reservoir for upper Nutrioso, below Nelson Reservoir dam for lower Nutrioso, above the mouth for Rudd Creek, and above downstream property boundary of Wenima Wildlife Area for the Little Colorado River.

______Stream Year/reach type Upper Nutrioso Lower Nutrioso Rudd Little Colorado 1998 Marking reach 7.1–7.6 3–3.5 Survey reach 6.6–8.1 1.5–2.5 0–4.5 1999 Marking reach 6.3–6.8 3–3.5 0.997–1.047 7.3–7.8 5–5.5 2.184–2.234 7–7.5 2.734–2.784 3.181–3.231 Survey reach 5.8–8.3 0–11.5 0–4.5 0–3.3 2000 Survey reach 5.8–8.3 0–11.5 0–4.5 0–3.3 200 WESTERN NORTH AMERICAN NATURALIST [Volume 62 their length. Lepidomeda vittata were marked 2000. One (365 mm TL) coded-wire tagged O. with coded-wire tags in 1998 and with fluores- mykiss was captured in July, 5.7 km below cent elastomeres in 1999. Elastomeres were Nelson Reservoir, and another (248 mm TL) used so fish could be identified individually was captured in April, where Nutrioso Creek and movements could be approximated more enters Nelson Reservoir. The 5 O. mykiss accurately. Both types of tags have high, short- (198–230 mm TL) marked with tetracycline term retention rates (Blankenship and Tipping were caught approximately 250 m above Nel- 1993, Bonneau et al. 1995, Hale and Gray 1998). son Reservoir. No marked trout were captured During 1998 through spring 2000, we sur- in Rudd Creek or Little Colorado River. veyed for marked fish in May after spring Approximately 75,500 O. mykiss were runoff, in July during summer monsoons, and stocked into Nelson Reservoir from 1996 to in September after summer monsoons, 2000. Anglers removed 48–85% of these trout All trouts captured in Nutrioso Creek (all (Table 2). Nelson Reservoir spilled in spring years) and in Rudd Creek (2000 only) were 1997 (April 12 through 1st week of May), 1998 sacrificed to assess predation on L. vittata. In (March 29 through May 23), summer 1999 addition, Little Colorado River on Wenima (August), and periodically from winter 1999 Wildlife Area was sampled during spring, sum- through spring 2000, generally reflecting spikes mer, and autumn from 1997 through spring in Little Colorado River discharge (Fig. 2). 2000. We captured fish with a single pass of We gill-netted 3 O. mykiss/8 net nights in backpack and canoe electrofishers, recorded 1997, 4 O. mykiss/8 net nights in 1998, 6 O. location of capture, and sacrificed and eviscer- mykiss/4 net nights in 1999, and 18 O. mykiss/ ated trout for diet analysis. Stomach contents 4 net nights in Nelson Reservoir during 2000. were analyzed in the laboratory, and fish re- Nonnative black crappie (Pomoxis nigromacu- mains were identified to the lowest taxonomic latus) were more prevalent than O. mykiss level. (29/8 net nights, 1997; 7/8 net nights, 1998; Availability of L. vittata as prey was based 7/4 net nights, 1999; and 30/4 net nights, on the percentage of predator locations with 2000). Over 1000 nonnative P. promelas and a L. vittata present (number of sympatric loca- few nonnative L. cyanellus, but no L. vittata, tions divided by total number of locations were caught in the trap nets set in 1997. where the predator was captured). Species We marked 210 L. vittata and recaptured were considered sympatric at a location if they 71 (54–134 mm TL; Table 3); some fish may were collected within 10 m of each other in have been caught more than once. Typically the same habitat (i.e., pool, run, riffle). (13 of 17 surveys), more than 80% of the fish Distributions of trouts and L. vittata within critical habitat and nearby study reaches were captured during a survey in a given stream did assessed with data from all surveys. To assess not have a mark. Sixteen fish moved, 10 down- species ranges and overlap within the study stream and 6 upstream (Table 3). Timing and streams, only data from surveys of the entire direction of movement did not appear related study reaches are addressed within this paper to flow. Base flow occurred during the mark- (because Nutrioso Creek above the reservoir ing period each year (Fig. 2). Three flow spikes was sampled in its entirety only during 2000, were evident in Rudd Creek, 2 of which (May data distributions are not presented). 1998 and August 1999) were also evident in Flow data were obtained from the U.S. the Little Colorado River (Fig. 2). Geological Survey gauging station (09384000) Thirty-three percent of S. trutta (N = 24), on the Little Colorado River above Lyman Lake 6% of O. mykiss (N = 54), and 25% of S. fonti- and from an AGFD gauge on Rudd Creek at nalis (N = 4) had fish in their stomachs the fish barrier. Overflow of Nelson Reservoir (Table 4). Remains of one L. vittata were found dam was determined by observation. in an individual S. trutta from Little Colorado River during 1997. Lepidomeda vittata was RESULTS not in any other salmonid examined. Sixty-three percent of the sites with O. Seven marked O. mykiss were captured in mykiss (N = 19), 60% with S. trutta (N = 10), Nutrioso Creek, 2 with coded-wire tags in and 50% of the sites with S. fontinalis (N = 2) 1998 and 5 with tetracycline marks in May had L. vittata present (Table 4). At sites where 2002] L. VITTATA: MOVEMENTS, DISTRIBUTION, PREDATION 201

TABLE 2. Estimates of stocking and harvest (based on creel census) for O. mykiss in Nelson Reservoir, AZ.

______Year Information 1996 1997 1998 1999 Number stocked 16,042 19,897 20,000 19,546 Census period April 1996–April 1997 May 1997–April 1998 May 1998–April 1999 May 1999–Dec. 1999 Estimated harvest 7,723 13,943 11,956 16,529 Percent harvested 48 70 60 85

Colorado River). Nonnative P. promelas was typically the most abundant or 2nd most abun- dant species captured in Nutrioso Creek and Little Colorado River, but was rare in Rudd Creek. The other nonnative species (only in Nutrioso Creek and the Little Colorado River), L. cyanellus, typically comprised less than 1% of the catch. Unmarked trout were captured in all 3 streams, most during the 1st year (Figs. 3–5). Though not ample in numbers, Oncorhynchus mykiss was the most abundant and widely dis- tributed trout in Nutrioso and Rudd creeks, whereas S. trutta was the most abundant trout captured in the Little Colorado River.

DISCUSSION

Movement of O. mykiss out of Nelson Reser- voir is minimal, reducing the chance of nega- tive effects on L. vittata by nonnative trout. Despite reservoir spills every year (except 1996), post-spill sampling yielded only 1 Fig. 2. Flows in Little Colorado River, AZ, above marked O. mykiss below the dam and 5 Lyman Lake, 1996 through 1997 (A) and 1998 through marked O. mykiss above the mouth. Down- May 2000 (B); Rudd Creek, AZ, 1998 through May 2000 stream escapement may have been minimal (C). Dates when L. vittata were marked and surveyed are shown. because few O. mykiss were in the reservoir prior to spring runoff each year, as indicated by creel-take estimates and spring sampling trout and L. vittata were captured, all trout efforts. However, no escapement was detected were more than twice as long as L. vittata, after an extensive summer spill in August except at 4 sites where O. mykiss were larger 1999, when O. mykiss numbers in the reser- than, but not twice as large as, L. vittata. voir should have been relatively high. Escape- Lepidomeda vittata was throughout Nutrioso ment upstream is probably limited by an Creek (below Nelson Reservoir), Rudd Creek, extensive cattail-sedge bed at the mouth of and Little Colorado River (Figs. 3–5). Range Nutrioso Creek. Aside from that, the 5 marked in Nutrioso Creek and Little Colorado River O. mykiss obtained above the reservoir were was greater (Figs. 3–5) in 1999 and 2000 than smaller than O. mykiss (335–445 mm TL) con- in 1996 (Nutrioso Creek) or 1997 and 1998 currently captured in Nelson Reservoir and (Little Colorado River). Distribution overlapped were also relatively smaller than O. mykiss on broad scales with salmonids in each stream (average size approximately 230 mm TL) (Figs. 3–5). Other native species were P. dis- stocked in spring 1999 (AGFD unpublished cobolus, R. osculus, and the Little Colorado data). We therefore believe that these trout River sucker (Catostomus sp.; only in the Little came from a private pond upstream of Nelson 202 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 3. Number of L. vittata marked and captured in subsequent surveys (numerator = total fish captured, denomi- nator = number of recaptures), number of recaptures that moved, and distances moved (negative numbers indicate downstream and positive indicate upstream movement) by recaptured fish in 3 eastern Arizona streams, April 1998 through May 2000. Fish were marked with coded-wire tags (April 1998) or fluorescent elastomeres (April 1999).

______Year and month of survey ______1998 ______1999 ______2000 Tag type and stream Marked May July Sep May July Sep May

CODED-WIRE TAGS Rudd Creek (N fish) 66 40/12 45/17 90/12 20/2 8/1 41/0 81/1 No. marked that moved 1 0 0 1 1 0 0 Reaches moved (500 m) –2 –1 –3 Upper Nutrioso (N fish) 37 25/1 30/0 24/2 42/0 7/0 38/0 147/3 No. marked that moved 0 0 0 0 0 0 3, Reaches moved (500 m) –2, –3, –3

ELASTOMERE TAGS Rudd Creek (N fish) 22 20/3 8/3 41/1 81/3a No. marked that moved 3 1 1 0a Distance moved (m) –272, 268, 98 313 554 Upper Nutrioso (N fish) 13 42/2 7/2 36/1 147/0 No. marked that moved 1 1 1 Distance moved (m) 599 –190 –381 Little Colorado (N fish) 72 137/5 157/0 127/0 No. marked that moved 2 Distance moved (m) –115, 77 aFish were recaptured in reach they were marked in; actual meter where they were marked is unknown since complete mark could not be discerned upon recapture.

TABLE 4. Incidence of piscivory (number of and percentage [in parentheses] of individuals with fish in their gastroin- testinal tracts) by 3 salmonid species captured in Rudd and Nutrioso creeks and Little Colorado River, AZ, 1996–2000. Number of sites where salmonids were captured and number and percentage (in parentheses) of those sites with L. vit- tata are also given.

______Salmonid species O. mykiss S. trutta S. fontinalis (N = 54) (N = 24) (N = 4) Range of total length (mm) 105–495 212–552 162–302 Trout with fish in GI tract 3 (6) 7 (33) 1 (25) Trout with L. vittata in GI tract 0 1 (4) 0 Trout with other cyprinids in GI tract 3 (6) 7 (33) 1 (25) Number of sites with trout 43 17 4 Number of trout sites where presence- absence of L. vittata was recorded 19 10 2 Number of trout sites with L. vittata 12 (63) 6 (60) 1 (50)

Reservoir (at least one private pond upstream Lepidomeda vittata were not found within of the reservoir is known to be stocked with O. Nelson Reservoir, but sampling was limited mykiss from private hatcheries that are gener- and not a primary objective. Lepidomeda vit- ally known to use tetracycline-treated feed), tata are probably rare in Nelson Reservoir indicating that there was no O. mykiss move- because of the presence of nonnative piscivo- ment upstream from Nelson Reservoir. Inject- rous fishes and the fact that L. vittata gener- ing all AGFD-reared trout with coded-wire ally occupy stream habitats. tags before they are stocked would be a more Marked L. vittata did not move much, but precise method of detecting movement out of sample sizes were too small to make any mean- the reservoir. ingful conclusions regarding flood-induced 2002] L. VITTATA: MOVEMENTS, DISTRIBUTION, PREDATION 203

Fig. 3. Number of L. vittata and O. mykiss captured per Fig. 4. Number of L. vittata and S. trutta captured per 500 m below Nelson Reservoir in Nutrioso Creek, AZ, 500 m in Little Colorado River, AZ, within Wenima Wild- 1996–2000. life Area, 1997–2000. movement. Most L. vittata we captured dur- ing surveys subsequent to marking were un- marked, suggesting movement out of the study area, low tag retention, mortality, or failure to capture marked fish. We believe tag retention was high and our surveys extensive enough (particularly in 1999) to detect long-range movements. Unfortunately, we cannot rule out the possibility that fish moved outside our sur- vey area or suffered mortality. Lepidomeda vittata appear to be available prey fish for each salmonid species, and they co-occurred with each salmonid. All 3 salmon- ids consumed cyprinid fishes, and so they all may consume L. vittata. Because Salmo trutta were more piscivorous than O. mykiss or S. fontinalis, they may pose more of a threat to L. vittata. Fortunately, neither S. trutta nor S. fontinalis are still stocked into the Little Col- orado River basin, although natural reproduc- tion occurs within these waters. Blinn et al. (1993), based on experiments in stream enclosures, suggested that O. mykiss Fig. 5. Number of L. vittata and O. mykiss captured per limit the distribution of L. vittata. Our results 500 m in Rudd Creek, AZ, above the confluence of lend little support for this contention; both Nutrioso Creek, 1998–2000. 204 WESTERN NORTH AMERICAN NATURALIST [Volume 62 species were often caught within the same are reported to be. Because O. mykiss impact pools (during uninterrupted flow conditions), L. vittata populations and the nonnative sport and no predation was detected by O. mykiss fishery is to be retained, efforts to further on L. vittata. However, O. mykiss will likely reduce impacts should include (1) stocking O. prey on larval fish when they are available, as mykiss into lakes only, and only after dams suggested for other native-nonnative fish inter- cease to spill from spring runoff, (2) removing actions (Marsh and Langhorst 1988, Johnson nonnative salmonids where L. vittata occur, and Hines 1999). In laboratory settings they and (3) removing nonnative salmonids and re- readily preyed on larval L. vittata and P. establishing L. vittata in areas they historically promelas (unpublished data). However, since inhabited. O. mykiss tend to be largely insectivorous (Cada et al. 1987, Angradi and Griffith 1990, ACKNOWLEDGMENTS Metcalf et al. 1997), the impact of predation on larval stages on the overall population of L. This work was funded by a grant from Fed- vittata is unknown. Studies examining this eral Aid in Sport Fish Restoration Program interaction in natural situations (when various (Dingell-Johnson/Wallop-Breaux) and a match- prey types are available) would be valuable, ing state grant from the Arizona Game and but difficult, since larvae are quickly digested. Fish Department Heritage IIPAM Fund. We Sample size was relatively small for all thank L. Avenetti, M. Lopez, R. Dryer, and G. objectives within this study. Drought condi- Gonzales for their assistance in the field and tions likely affected the sample size of L. vit- their input of ideas. We also thank J. Novy, J. tata and trouts during our study. Annual pre- Burton, K. Young, and D. Blinn for their con- cipitation was below average each year during tributions to the study design and M. Brouder, our study, particularly during 1996 and 1997 T. McKinney, M. Childs, and M. Rabe for re- (data from the Western Regional Climate Cen- view of this manuscript. Finally, we thank the ter, Reno, NV). During summer and autumn U.S. Fish and Wildlife Service Fishery Resource months of 1996 and 1997, Rudd and Nutrioso office in Pinetop, Arizona, for use of their lab- creeks became intermittent. We estimated oratory facilities and equipment. that 50–75% of Nutrioso Creek below Nelson Reservoir and >50% of lower Rudd Creek LITERATURE CITED were dry in the summers of 1996 and 1997. ANGRADI, T.R., AND J.S. GRIFFITH. 1990. Diel feeding Such large decreases in available habitat cou- chronology and diet selection of rainbow trout pled with likely decreases in habitat quality (Oncorhynchus mykiss) in the Henry’s Fork of the (increased water temperatures and turbidity, Snake River, Idaho. Canadian Journal of Fisheries and decreased dissolved oxygen due to nonex- and Aquatic Science 47:199–209. BLANKENSHIP, J.M., AND H.L. TIPPING. 1993. Evaluation of istent flows) may have had effects on L. vittata visible implant and sequentially coded wire tags in and trout populations. It is also possible that sea-run cutthroat trout. North American Journal of electrofishing contributed to these issues as 13:391–398. well (Nordwall 1999). BLINN, D.W., C. RUNCK, D.A. CLARK, AND J.N. RINNE. 1993. Effects of rainbow trout predation on Little Efforts to prevent nonnative salmonid im- Colorado spinedace. Transactions of the American pacts on L. vittata include cessation of non- Fisheries Society 122:139–143. native salmonid stocking and nonnative fish BONNEAU, J.L., R.F. THUROW, AND D.L. SCARNECCHIA. removal from streams where L. vittata exist. 1995. Capture, marking, and enumeration of juve- nile brown trout and cutthroat trout in small, low- Native O. apache could be stocked rather than conductivity stream. North American Journal of nonnative O. mykiss (Rinne and Janisch 1995), Fisheries Management 15:563–568. since these waters are within historical range CADA, G.F., J.M. LOAR, AND D.K. COX. 1987. Food and of O. apache (Carmichael et al. 1995). How- feeding preferences of rainbow and brown trout in southern Appalachian streams. American Midland ever, before converting lake fisheries to native Naturalist 117:374–385. O. apache, we recommend studies examining CARMICHAEL, G.J., J.N. HANSON, J.R. NOVY, K.J. MEYER, the effects of O. apache on L. vittata and other AND D.C. MORIZOT. 1995. Apache trout manage- native fishes under a variety of habitat condi- ment: cultured fish, genetics, habitat improvements, tions and prey availability so these efforts would and regulations. American Fisheries Society Sympo- sium 15:112–121. not be refuted in the event that O. apache are HALE, R.S., AND J.H. GRAY. 1998. Retention and detection just as detrimental to L. vittata as O. mykiss of coded wire tags and elastomer tags in trout. North 2002] L. VITTATA: MOVEMENTS, DISTRIBUTION, PREDATION 205

American Journal of Fisheries Management 18: American Southwest. University of Arizona Press, 197–205. Tucson. JOHNSON, J.E., AND R.T. HINES. 1999. Effect of suspended MOYLE, P.B. 1986. Fish introductions into North America: sediment on vulnerability of young razorback suck- patterns and ecological impact. Pages 27–43 in H.A. ers to predation. Transactions of the American Fish- Mooney and J.A. Drake, editors, Ecology of bio- eries Society 128:648–655. logical invasions of North America and Hawaii. MARSH, P.C., AND D.R. LANGHORST. 1988. Feeding and Springer-Verlag, New York. fate of wild larval razorback suckers. Environmental NORDWALL, F. 1999. Movements of brown trout in a small Biology of Fishes 21:59–67. stream: effects of electrofishing and consequences MEFFE, G.K. 1984. Effects of abiotic disturbance on coex- for population estimates. North American Journal of istence of predator-prey fish species. Ecology 65: Fisheries Management 19:462–469. 1525–1534. RINNE, J.N., AND M. ALEXANDER. 1995. Non-native METCALF, C., F. PEZOLD, AND B.G. CRUMP. 1997. Food salmonid predation on two threatened native species: habits of introduced rainbow trout (Oncorhynchus preliminary observations from field and laboratory mykiss) in the upper Little Missouri River drainage studies. Proceedings of the Desert Fishes Council of Arkansas. Southwestern Naturalist 42:148–154. 26(1994):114–116. MILLER, R.R. 1961. Man and the changing fish fauna of RINNE, J.N., AND J. JANISCH. 1995. Coldwater fish stocking the American Southwest. Michigan Academy of Sci- and native fishes in Arizona: past, present, and future. ence, Arts, and Letters 46:365–404. American Fisheries Society Symposium 15:397–406. ______. 1963. Distribution, variation, and ecology of Lep- USDI, U.S. FISH AND WILDLIFE SERVICE. 1987. Endan- idomeda vittata, a rare cyprinid fish endemic to east- gered and threatened wildlife and plants; final rule ern Arizona. Copeia 1963:1–5. to determine Lepidomeda vittata (Little Colorado MINCKLEY, W.L. 1973. Fishes of Arizona. Arizona Game spinedace) to be a threatened species with critical and Fish Department, Phoenix. habitat. Volume 52, No. 179. MINCKLEY, W.L., AND L.H. CARUFEL. 1967. The Little Colorado River spinedace, Lepidomeda vittata, in Received 11 July 2000 Arizona. Southwestern Naturalist 12:291–302. Accepted 29 March 2001 MINCKLEY, W.L., AND J.E. DEACON, EDITORS. 1991. Battle against extinction; native fish management in the Western North American Naturalist 62(2), © 2002, pp. 206–209

NEW RECORDS AND RANGE EXTENSIONS OF SPECIES OF DIPOGON (HYMENOPTERA, POMPILIDAE) IN COLORADO

Howard E. Evans1 and David A. Leatherman2

ABSTRACT.—Four species of Dipogon are recorded from Colorado for the first time. These are hurdi Evans, pul- chripennis (Cresson), graenicheri Banks, and brevis (Cresson). All are inhabitants of wooded areas and are believed to nest in cavities in trees. Notes are presented on habitat and prey of these 4 species and of the recently described species kiowa Evans.

Key words: Hymenoptera, Pompilidae, Dipogon, distribution, prey, spiders.

Wasps of the genus Dipogon are largely ponderosa) at Lake Isabel, Custer County, at confined to wooded areas where several of the about 9000 feet elevation, 8 July 1998. In size species are known to nest in cavities in trees. and structural details this specimen resembles Townes (1957) recorded 12 species, but 7 the type closely. Color is also similar except more have since been described (Wasbauer that the legs are darker, the coxae and fore and 1960, 1966, Evans 1974, 2000). These wasps middle legs being partially infuscated, while are rarely taken by conventional collecting the hind legs are wholly dark beyond the methods, and several of the species are known coxae. from only a few specimens. Clearly, some of HABITAT.—Lower montane forest charac- the species have much broader ranges than terized by mixed conifers: ponderosa pine, previously appreciated, and we report here on Douglas-fir (Pseudotsuga menziesii), and white range extensions for 4 of these. Ten species fir (Abies concolor). The tree from which the have now been recorded from Colorado, 6 of female was collected had been killed within them not included in a recent checklist of Col- the last 2 years by bark beetles of the genus orado Pompilidae (Evans 1997). We include Dendroctonus and still retained some comple- here notes on individual specimens and on the ment of brown needles. Recently fallen pon- habitat and behavior of 2 species of subgenus derosa pines, also killed by bark beetles, were Deuteragenia and 3 species of subgenus in the immediate vicinity. Dipogon. Fieldwork reported here was done Dipogon (Deuteragenia) by DAL in the course of surveys of trees pulchripennis (Cresson) throughout the state. This species is characteristic of woodlands Dipogon (Deuteragenia) of the northeastern United States and south- hurdi Evans eastern Canada, but it has also been taken in This species was described from mountains mountainous areas in Arkansas and Arizona. in the state of Durango, Mexico, based on a Thus, it is not surprising to discover it in the female taken with a spider of the genus Icius mountains of Colorado. A female was col- (Salticidae). An allotype was described from lected by DAL as it emerged from a wood- Portal, Arizona, at about 5000 feet elevation. boring beetle tunnel at the base of a live but Known range of this species is now extended fire-damaged ponderosa pine within the Black to the mountains of Colorado. A female was Tiger Fire area, 0.2 miles southeast of Sugar- taken by DAL on the trunk of a dead, stand- loaf Mountain, Boulder County, 22 August ing, large-diameter ponderosa pine (Pinus 1998. Another female was taken by DAL at

1Department of Bioagricultural Sciences and Pest Management, Colorado State University, Fort Collins, CO 80523. 2Colorado State Forest Service, Colorado State University, Fort Collins, CO 80523.

206 2002] RECORDS OF DIPOGON SPIDER WASPS 207

Beulah, Pueblo County, 23 August 2000, asso- Dipogon (Dipogon) ciated with a ponderosa pine killed by Den- brevis (Cresson) droctonus ponderosae. This individual was This species was recorded by Townes (1957) observed expelling boring dust from a bark from New England to Georgia, with speci- beetle exit hole, which presumably was the mens of his subspecies recalvus west to Michi- entrance to a nest within a wood borer tunnel. gan and Wisconsin. Specimens recently taken HABITAT.—The 1st female was taken in a in southeastern Colorado differ in no impor- ponderosa pine/Douglas-fir forest, which sus- tant structural details from brevis but match in tained a fire in summer 1990. The forest is color none of the 3 subspecies recognized by now quite open and characterized by mature Townes. We have 7 females, taken by DAL on live trees with basal fire scars on their north plains cottonwood and Siberian elm trunks in sides, dead standing snags, and fallen trees Lamar, Prowers County, 1 October 1997, 4–7 devoid of foliage and in moderate states of September 1998, and 15 September 2000. decay (bark beginning to sluff, wood softening These specimens vary in fore wing length and full of insect holes and cracks). The 2nd from 3.8 to 5.4 mm and have the close, silvery female was taken in a mature ponderosa pine to golden pubescence characteristic of brevis. forest sustaining considerable mortality from The body is black with ferruginous markings recent bark beetle attacks. as follows: antennae, clypeus, pronotum over Dipogon (Dipogon) its anterior 0.3 to 0.8, metapleura and sides of graenicheri Banks propodeum in 3 specimens, and a spot on the mesopleura in 1 specimen. The legs are dark, This species was described from South variably suffused with ferruginous. In the Miami, Florida, by Banks (1939). Townes (1957) series the metapleura vary from mat to shin- reported it from New Jersey to Florida and ing, and there is much variation in the spacing west to Louisiana. Hook (1998) recently of the punctures on the propodeum. These 2 reported it from Travis County, Texas. A single features were used by Townes to characterize female was taken by DAL on the trunk of a his subspecies recalvus. living, large-diameter Siberian elm (Ulmus Despite the fact that these females do not pumila) at Willow Creek Park, Lamar, Prowers fit well into presently recognized subspecies, County, 16 August 1998. Although Lamar is at we do not feel it is appropriate to add still one of the lowest elevations in the state (3600 another subspecies at this time. Wasbauer feet), the occurrence of graenicheri here was (1960) described a species from California unexpected. This female differs from south (leechi) that resembles brevis in many details, Florida specimens in having the frons lightly but it differs from our Colorado series in hav- alutaceous, less polished than in the type, and ing appressed hairs on the propodeum and 1st the eyes somewhat more convergent above. In metasomal tergite, as well as more weakly this specimen the basal 1.7 metasomal tergites banded wings. and all sternites are rufous; the amount of HABITAT.—Specimens were collected at rufous on the metasoma is variable in eastern both Willow Creek Park (described above specimens we have seen. under graenicheri) and the southeastern sec- HABITAT.—Typical city park with mature, tion of Lamar Community College grounds widely spaced trees, mostly Siberian elm and (similar to Willow Creek Park and directly plains cottonwood (Populus deltoides var. moni- adjacent to the undeveloped riparian wood- fera) with an irrigated bluegrass understory. land descibed under graenicheri). Many of the deciduous trees displayed bacter- ial wetwood disease, which has killed patches Dipogon (Dipogon) of trunk bark, exposing bare wood, borer tun- kiowa Evans nels, and cracks. The park is bisected by Wil- This species was described from 9 females low Creek, characterized along its banks by taken by DAL in Lamar, Prowers County, at thickets of coyote willow (Salix exigua). An sites described under the preceding 2 species undeveloped woodland immediately south of and at various dates July–September 1998– the park contains plains cottonwood, Siberian 1999. One female was taken on the ground elm, Russian-olive (Eleagnus angustifolia), and within 1 foot of the trunk of a tree, the others tamarisk (Tamarix ramosissima). on the trunks. The wasps were never found 208 WESTERN NORTH AMERICAN NATURALIST [Volume 62 prior to midmorning, and most were collected the North American sayi Banks and the Euro- on shaded sides of trunks of large deciduous pean variegata (L.) prey almost exclusively on trees in late afternoon and early evening. Two crab spiders (Thomisidae). Various other mem- of the females were taken with their spider bers of this subgenus are known to prey on prey: a female Tutelina harti (Emerton) and an Agelenidae, Amaurobiidae, Areneidae, Clubio- immature Phanias sp. (both Salticidae). nidae, Gnaphosidae, and Segestriidae. Useful Another female was reared by DAL on 24 reviews of the biology of the genus have been August 1999 from a bolt of dead Utah juniper presented by Richards and Hamm (1939) and ( Juniperus utahensis) trunk wood collected at Evans and Yoshimoto (1962). More recently, Crawford State Park, Delta County, Colorado, there have been detailed studies of sayi by on 5 August 1999. The primary hole maker Fye (1965), Krombein (1967), and Jennings found associated with this same wood was the and Parker (1987), as well as studies of several horntail wasp, Sirex areolatus (Cresson) (Siri- Palaearctic species by Gros (1997). Much cidae). Species of Rhyssa (Ichneumonidae) remains to be learned about the biology of and Ibalia (Ibaliidae) also emerged from the these reclusive insects. wood; presumably, these were parasites of the Sirex. This was a very different habitat from ACKNOWLEDGMENTS Lamar, comprising mostly scattered juniper and other aridland plants characteristic of the Spider determinations were made by Paula Colorado Plateau. The elevation is about 6200 Cushing of the Denver Museum of Nature feet. and Science and David B. Richman of New Mexico State University. Michael S. Kelley, of DISCUSSION the Museum of Comparative Zoology, Harvard With the exception of 2 female pulchripen- University, loaned us the type of Dipogon hurdi and several other specimens important nis, which were taken as they emerged from for identification of our material. presumed nest holes at the extreme bases of trees, and the Crawford female kiowa, all LITERATURE CITED Dipogon individuals mentioned were observed on the bark surface of large-diameter trees at BANKS, N. 1939. Notes and descriptions of native Psam- heights of 2–8 feet aboveground (or in one mocharidae. Canadian Entomologist 71:225–231. case on the ground near such a tree). Presum- EVANS, H.E. 1974. A review of the species of Dipogon occurring in Central America, Mexico, and extreme ably, they forage for prey outside this range, southwestern United States (Hymenoptera, Pompili- and their capture at this height is probably as dae). Transactions of the American Entomological much a function of convenient observation Society 100:29–51. height as a true reflection of their behavior. ______. 1997. Spider wasps of Colorado (Hymenoptera, Pompilidae): an annotated checklist. Great Basin Most trees where Dipogon individuals were Naturalist 57:189–197. observed also supported foraging ants, the ______. 2000. Three new species of Dipogon (subgenus resemblance to which was enhanced by the Dipogon) from central and western North America wasps’ banded wings. The wasps differed in (Hymenoptera, Pompilidae). Proceedings of the Entomological Society of Washington 102:1010–1013. foraging behavior only in that they occasion- EVANS, H.E., AND C.M. YOSHIMOTO. 1962. The ecology ally “jumped” across bark fissures or other and nesting behavior of the Pompilidae (Hymen- obstacles. Otherwise, they ran across the bark optera) of the northeastern United States. Miscella- and explored crevices in patterns and rates neous Publications of the Entomological Society of America 3:65–119. very similar to the ants with which they shared FYE, R.E. 1965. The biology of the Vespidae, Pompilidae, space. and Sphecidae (Hymenoptera) from trap nests in So far as known, members of the subgenus northwestern Ontario. Canadian Entomologist 97: Dipogon prey on jumping spiders (Salticidae). 716–744. GROS, E. 1997. Notes sur la biologie de quelques Pom- In addition to the records for kiowa presented pilides de la sous-famille des Pepsinae (Hymen- above, there are records of brevis preying optera, Pompilidae). Bulletin de la Société Ento- upon salticids of the genera Phidippus and Pel- mologique de France 102:345–354. lenes (Evans and Yoshomoto 1952, Kurczewski HOOK, A.W. 1998. New distribution records of aculeate wasps in Texas and adjacent states (Hymenoptera: and Kurczewski 1968). Several species of sub- Sphecidae, Pompilidae, Mutillidae). Southwestern genus Deuteragenia also utilize Salticidae, but Naturalist 43:249–255. 2002] RECORDS OF DIPOGON SPIDER WASPS 209

JENNINGS, D.T., AND F. D . P ARKER. 1987. Habitats and spi- TOWNES, H.K. 1957. Nearctic spider wasps of the subfam- der prey of Dipogon sayi sayi (Hymenoptera: Pom- ilies Pepsinae and Ceropalinae. United States Nation- pilidae) in Washington County, Maine. Great Lakes al Museum Bulletin 209. 286 pp. Entomologist 20:135–140. WASBAUER, M.S. 1960. Taxonomic and distributional notes KROMBEIN, K.V. 1967. Trap-nesting wasps and bees: life on some western spider wasps (Hymenoptera: Pom- histories, nests, and associates. Smithsonian Press, pilidae). Pan-Pacific Entomologist 36:171–177. Washington, DC. 570 pp. ______. 1966. A new spider hunting wasp of the subgenus KURCZEWSKI, F.E., AND E.J. KURCZEWSKI. 1968. Host Dipogon from western Nevada (Hymenoptera: Pom- records for some North American Pompilidae pilidae). Proceedings of the Biological Society of (Hymenoptera) with a discussion of factors in prey Washington 79:17–20. selection. Journal of the Kansas Entomological Soci- ety 41:1–33. Received 23 October 2000 RICHARDS, O.W., AND A.H. HAMM. 1939. The biology of Accepted 12 February 2001 the British Pompilidae (Hymenoptera). Transactions of the Society for British Entomology 6:51–114. Western North American Naturalist 62(2), © 2002, pp. 210–217

MESEMBRIOXYLON OBSCURUM, A NEW COMBINATION FOR ARAUCARIOXYLON? OBSCURUM KNOWLTON, FROM THE UPPER JURASSIC MORRISON FORMATION, WYOMING

David A. Medlyn1 and William D. Tidwell2

ABSTRACT.—A reinvestigation of the type specimen of Araucarioxylon? obscurum Knowlton (1900) from the Upper Jurassic Morrison Formation in the Freezeout Hills of Wyoming has determined that this fossil should be transferred to Mesembrioxylon Seward (1919) as M. obscurum comb. nov. (Knowlton) Medlyn and Tidwell. Knowlton (1900), who was uncertain of its appropriate generic disposition, tentatively referred the wood to Araucarioxylon because it had obscure growth rings and relatively low ray height. Reexamination of Knowlton’s slides demonstrates that the wood has uniseri- ate, rarely biseriate rays, podocarpoid pitting, and diffuse axial parenchyma, none of which Knowlton mentioned. Tra- cheary pitting is mostly separate, round, uniseriate, occasionally biseriate and, when biseriate, opposite to subopposite. The smooth-walled crossfields exhibit 1–3 thin-bordered podocarpaceous pits per field. All of these features are present in Mesembrioxylon Seward.

Key words: fossil wood, Araucarioxylon, Mesembrioxylon, Upper Jurassic, Morrison Formation, Wyoming.

When Knowlton (1900:418) placed the new GEOLOGICAL REVIEW species Araucarioxylon? obscurum from the Freezeout Hills, Carbon County, Wyoming, in Although best known for its dinosaur re- Araucarioxylon Kraus (1870), he stated: “The mains (Dodson et al. 1980), the Upper Jurassic placing of this wood in the genus, Araucari- Morrison Formation has yielded one of the oxylon, is open to more or less question. Yet as most abundant and diverse Upper Jurassic it approaches more closely to this genus, I petrified coniferous floras in the world (Med- have tentatively so referred it.” The species lyn and Tidwell 1975, 1979, 1992, Tidwell epithet refers to the obscure nature of the 1990, 1998, Tidwell and Medlyn 1992, 1993). growth rings which Knowlton noted is consis- The formation extends from northern Arizona, New Mexico, Oklahoma, and Texas in the tent with Araucarioxylon, but he later added south and east through Utah, Wyoming, Mon- that the species lacks other characteristics of tana, western South Dakota, and Nebraska to this genus. A review of the slides of A.? obscu- the north. During the Late Jurassic (Kimmer- rum shows that his description does not idgian through Portlandian), there were suc- entirely agree with the holotype. The growth cessive invasions of western North America by rings are weakly developed or “obscure” as shallow seas from the north. As the seas Knowlton (1900) described, a condition simi- regressed northward, fluvial sediments of lar to false ring development. However, the sandstone, variegated shale, volcanic ash, and specimen also has mostly uniseriate tracheary siltstone comprising the Morrison Formation pitting and low, uniseriate rays; lacks araucari- were deposited (Derr 1974, Dodson et al. oid pitting; and has axial parenchyma and 1980, Stokes 1986). Most fossil wood occur- podocarpoid crossfield pitting. All of these ring in the Morrison Formation was likely make placement of this specimen in Araucari- transported from its place of growth prior to oxylon untenable. Therefore, we propose plac- burial through the extensive river systems that ing Araucarioxylon? obscurum into Mesembri- deposited the formation. oxylon Seward (1919), a form genus consistent According to Ward (1900a), the type speci- with the nature of its xylotomy. men of Araucarioxylon? obscurum was collected

1Department of Geology, Utah State University, Logan, UT 84322-4900. 2Department of Botany and Range Science, Brigham Young University, Provo, UT 84602.

210 2002] MESEMBRIOXYLON OBSCURUM, COMB. NOV. 211 in the late 1890s from the northern part of the by latewood tracheids. Latewood 2–6 cells, Freezeout Hills in the so-called cycad beds, occasionally up to 12 wide; transition usually now placed in the Morrison Formation. The abrupt, sometimes gradual (all the above char- wood was associated with fossil “cycads” and acteristics may be present in one ring). Early- originally considered to be wood from one of wood tracheids large, square to roundish, some- these “cycads” (Ward 1900a). At the time of its what uniform in size and shape, radial diame- collection, silicified wood was very abundant ter 25–35 µm and tangential diameter 20–30 in the cycad beds, as well as in some of the µm; walls 3 µm thick; lumens mostly round, nearby ledges (Ward 1900a). The locality is in radial diameter 22–32 µm, tangential diameter sec. 13, T25N, R79W. At this locality the Mor- 17–27 µm. Latewood tracheids small, tangen- rison Formation rests conformably on the Sun- tially flattened, radial diameter 11–15 µm, tan- dance Formation and is overlain disconformably gential diameter 24–34 µm, walls 3 µm thick, by the Cloverly Formation (Baker 1965). In lumens small, radial diameter 7.5–10 µm, tan- this area the Morrison is approximately 124 m gential diameter 20–30 µm across; axial paren- (400 feet) thick and divisible into 2 major units. chyma diffuse, solitary, cells 30 µm across, The upper unit is similar to the Brushy Basin µ Member of this formation in Utah, Colorado, thin-walled, walls 2 m thick; some ray cells and New Mexico, whereas the basal member and tracheids resinous. apparently has no equivalency elsewhere (Baker Radial: Tracheids with occasional resin 1965). At this locality the Morrison Formation plugs; abundant tracheary pits, mostly uni- consists of fine, soft sandstones and white, red- seriate, occasionally partially biseriate, when dish, yellowish, or olive-gray calcareous shales biseriate opposite to subopposite, generally containing dark marls with dinosaur bones separate, occasionally close, sometimes con- and other vertebrate remains. tiguous; crassulae lacking; pits circular, 10–13 The Freezeout Hills, 25 miles north of Medi- µm in diameter; pit apertures circular to ellip- cine Bow, Wyoming, occupy an area of approx- tic, included, small, 2.5–4 µm across; rays imately 10 square miles. The highest of the hills homocellular, resinous; ray cells parenchyma- is Freezeout Mountain, so named because of a tous, rectangular, horizontal and tangential story about a party of men who froze to death walls smooth to slightly nodular; crossfield pit- in its immediate vicinity during the early ting 1–3, commonly 1 podocarpoid pit per European settlement history of the region. crossfield, 5–6 µm in diameter, border thin; pit The hills are near the famous Como Bluff apertures included, 2–3 µm in diameter at locality where so many Upper Jurassic dino- widest point, elliptical to oblong; axial paren- saur fossils have been found (Ostrom and chyma cells rectangular, 80–120 µm high, end McIntosh 1966). wall smooth, resinous. The fossil “cycads” from the cycad beds Tangential: Tangential pitting present, bor- were first described by Ward (1900a, 1900b) dered pits, 8–10 µm in diameter, circular, as 20 species in his new genus Cycadella. included, aperture circular to elliptic, border These species were eventually placed in thick, 2 µm in diameter; rays abundant, 60–70 Cycadeoidea as a single species, C. wyomin- ⋅ mm–2, typically uniseriate, sometimes biseri- gensis (Ward) Wieland, by Delevoryas (1960). ate; ray height variable, 1–15, rarely up to 21 (2–6 average) cells high; ray cells round tan- TAXONOMY gentially, 7–22 µm across, walls 3–5 µm thick. Division: Coniferophyta Order: Coniferales REPOSITORY.—U.S. National Museum, Smith- Mesembrioxylon obscurum (Knowlton) sonian Institution, Washington, D.C. (USNM comb. nov. 455260), 1900 Araucarioxylon? obscurum Knowlton, TYPE LOCALITY.—Freezeout Hills, Wyoming. pp. 418–419; pl. CLXXVIII. COMPARISONS (Figs. 1–4)

EMENDED DIAGNOSIS AND DESCRIPTION.— Mesembrioxylon is widely distributed both Transverse: Secondary xylem; growth rings geographically and geologically throughout absent or indistinct, mostly discontinuous; the Mesozoic era including the Jurassic of ring width 2–3 mm; outer margin delineated Poland (Gothan 1906, Seward 1919) and Korea 212 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 1. Mesembrioxylon obscurum, comb. nov. (all transverse sections): A, homogenous, secondary xylem showing a poorly developed growth ring (arrow; bar = 240 µm); B, enlargement of part of Figure A, demonstrating discontinuous nature of growth rings and resin-filled parenchyma (arrows; bar = 60 µm).

(Shimakura 1936); the Jurassic or Cretaceous Arctic Archipelago (Ash and Basinger 1991), of Syria (Edwards 1929); the Cretaceous of and an undescribed species of Mesembrioxy- Japan (Nishida and Nishida 1983), Australia lon from the Middle Jurassic Coon Hollow (Sahni 1920), Burma (Sahni 1937), India (Bhard- Formation of the Hells Canyon area, Oregon waj 1953, Ramanujam 1953, Jain 1964, Agashe and Idaho (Ash 1991), are the only presently 1968), Great Britain (Seward 1919), South reported species of Mesembrioxylon in North Africa (Bamford and Corbett 1994), and the America. United States (Thayn and Tidwell 1984, Ash Mesembrioxylon obscurum is characterized 1991, Ash and Bassinger 1991, Tidwell et al. by possessing occasionally thin-bordered, ellip- 1998). Mesembrioxylon stokesi Thayn and Tid- tic, podocarpoid crossfield pits; uniseriate to well (1984), M. carterii Tidwell, Britt and Ash occasionally biseriate, round tracheary pits, (1998), Mesembrioxylon sp. from the Canadian tangential pits, diffuse axial parenchyma, and 2002] MESEMBRIOXYLON OBSCURUM, COMB. NOV. 213

Fig. 2. Mesembrioxylon obscurum, comb. nov. (all tangential sections): A, section showing the relatively low, usually uniseriate rays (bar = 60 µm); B, illustration showing homogenous ray cell structure and uniseriate tangential pits (arrows; bar = 30 µm). the absence of tracheal septations and crassu- pseudo-bedfordense Nishida (1966), and M. lae. Mesembrioxlyon obscurum appears to rep- bedfordense (Stopes) Seward (1919) differenti- resent a distinct species. For instance, the ates them from M. obscurum. presence of septate tracheids in M. gothanii Although Dacrydioxlyon estherae Greguss (Stopes) Seward (1919) and M. nihei-takagii (1967) from the Lower Oligocene of Hungary Nishida (1966), the occurrence of crassulae in is somewhat similar, it differs in having thin M. woburnense (Stopes) Seward (1919), and growth rings with scattered parenchyma cells the lack of axial parenchyma in M. shanense in its rings, sometimes terminally, circular Sahni (1937), M. indicum Bhardwaj (1953), M. crossfield pits without borders, and tangential rajmahalense Jain (1964), and M. carterii Tid- pitting. Conversely, in M. obscurum, the growth well, Britt, and Ash (1998) distinguish them rings are absent or indistinct, its parenchyma from M. obscurum. The absence of tangential diffuse, its crossfield pits bordered, albeit pitting in M. godaverianum Sahni (1931), M. thinly, and it has tangential pitting. 214 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 3. Mesembrioxylon obscurum, comb. nov. (all radial sections): A, illustration of the mostly uniseriate and generally separate pits (bar = 60 µm); B, axial, resin-filled parenchyma cells (bar = 60 µm); C, partially biseriate, opposite to sub- opposite, tracheary pitting (bar = 30 µm).

DISCUSSION Podocarpoxylon crossfield pits as having verti- cal to steeply inclined apertures, whereas Seward (1919) instituted the genus Mesem- Phyllocladoxylon crossfield pits are character- brioxylon to replace Podocarpoxylon Gothan ized by oblique to elliptic apertures. Seward (1905) and Phyllocladoxylon Gothan (1905). (1919) reasoned that Gothan’s generic designa- Earlier, Stopes (1915) had recognized the diffi- tion implied affinities to extant genera which culties of separating Gothan’s 2 genera and he concluded could not be assumed. However, suggested combining them using Podocarpoxy- Seward’s description of Mesembrioxylon implies lon for podocarpaceous woods. Later Kräusel a relationship to the Podocarpaceae because (1949) concluded that Phyllocladoxylon and he indicated that this new genus agreed struc- Podocarpoxylon could be distinguished by dif- turally with the extant genera Podocarpus and ferences in their crossfield pitting. He defined Dacrydium. Many researchers (Bhardwaj 1953, 2002] MESEMBRIOXYLON OBSCURUM, COMB. NOV. 215

Fig. 4. Mesembrioxylon obscurum, comb. nov. (all radial sections): A, B, illustrations of the distribution and shape of the crossfield pits (arrows; bar = 30 µm).

Jain 1964, Nishida 1966, Thayn and Tidwell Other genera of petrified woods, including 1984) have subsequently accepted Seward’s Protophyllocladoxylon Kräusel (1939), Proto- view and used Mesembrioxylon for Mesozoic podocarpoxylon Eckhold (1923), Dacrydioxy- woods of this type rather than either Podocar- lon Greguss (1967), and Microcachryxylon poxylon or Phyllocladoxylon. Sahni (1937) con- Torres et al. (1991), have been postulated as sidered Mesembrioxylon to most likely repre- having podocarpaceous affinities. Emberger- sent a type of podocarpaceous conifer. Despite ixylon Lemoigne (1968) is similar but not yet Seward’s suggestion, some podocarpaceous allied with woods related to Podocarpaceae species of Mesozoic and Cenozoic age that (Medlyn and Tidwell 1975). Xenoxylon Gothan could be assigned to Mesembrioxylon have (1905) was established for woods without clear been placed in Phyllocladoxylon (Shimakura familial relationship, although some authors 1936, Nishida and Nishida 1983). have proposed a link with this genus to the 216 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Podocarpaceae (e.g., Arnold 1952). This con- DODSON, P., A.K. BEHRENSMEYER, A.K. BAKKER, AND J.S. nection has not been supported, however, by MCINTOSH. 1980. Taphonomy and paleoecology of the dinosaur beds of the Jurassic Morrison Forma- recent research (Phillipe and Thévenard 1996). tion. Paleobiology 6:208–232. Greguss (1955) characterizes the extant ECKHOLD, W. 1923. Die Hoftüpfel bei rezenten und fos- Podocarpaceae as having generally smooth or sile Coniferen. Jahrbuch Preussischen Geologishen only slightly thickened walls of the horizontal Landesanstait 42:472–505. EDWARDS, W.N. 1929. Lower Cretaceous plants from Syria ray cells. Usually, there is a single pit in the and Transjordania. Annals and Magazine of Natural crossfield in members of the family, except in History (10)4. Podocarpus minor Parl. and Michrocachrys GOTHAN, W. 1905. Zur Anatomie lebender und fossiler tetragona Hook, F. The single crossfield pit in Gymnospermen-Hölzer. Abhandlungen der Königlich Preussischen Geologishen Landesanstait, Berlin 44: Podocarpus L’Héritier and Dacrydium Sol. is 1–108. circular. Greguss (1955) further defines a ______. 1906. Fossile Hölzer aus dem Bathonien von Rus- podocarpoid crossfield pit as having a mostly sisch-Polen. Verhandlungen der Kaiserlich Russ. oblique or vertical aperture that is almost as Miner, Gesellshaft, St. Petersburg [ii] 44:435–458. GREGUSS, P. 1955. Identification of living gymnosperms on large as the pit and a very narrow border, a the basis of xylotomy. Akademiai Kiado, Budapest. distinguishing character present in Mesembri- ______. 1967. Fossil gymnosperm woods in Hungary: oxylon obscurum. from the Permian to the Pliocene. Akademiai Kiado, Budapest. JAIN, K.P. 1964. A new species of Mesembrioxylon, M. ACKNOWLEDGMENTS rajmahalense from the Rajmahal Hills, Bihar, India. Palaeobotanist 13:153–154. The authors are grateful to the U.S. National KNOWLTON, F.H. 1900. Description of a new species of Museum of Natural History, Smithsonian Insti- Araucarioxylon from the cycad bed of the Freezeout tution, for the loan of Dr. F.H. Knowlton’s Hills, Carbon County, Wyoming. Part 2, pages 418– 419 in L.F. Ward, editor, Status of Mesozoic floras of slides of Araucarioxylon? obscurum, and to Dr. the United States. U.S. Geological Survey Twentieth Sidney R. Ash of Albuquerque, New Mexico, Annual Report (1898–1899). for reviewing the manuscript. KRAUS, G. 1870. Bois fossiles de conifères. Part 2, pages 363–385 in W.P. Schimper, Traité de Paléontologie Végétale. F. Bailliére, Paris. LITERATURE CITED KRAÜSEL, R. 1939. Ergebnisse der Forschungsreisen Prof. E. Stromers in den Wüsten Ägyptens. 3. Die fossilen AGASHE, S.N. 1968. Studies on the fossil gymnosperms of Pflanzen Ägyptens. Abhandlungen der bayerischen India. Part I. A new species of Mesembrioxylon, M. Akademie der Wissenschaften m.-n. Abteilung (N.F.) mahabalei sp. nov. Palaeobotanist 17:312–316. 47:1–140. ARNOLD, C.A. 1952. Silicified plant remains from the Meso- ______. 1949. Die fossilen Koniferen-Hölzer (unter Auss- zoic and Tertiary of western North America. II. Some chluss von Araucarioxylon Kraus) II. Teil. Kritische fossil woods from northern Alaska. Papers of Michi- Untersuchungen zur Diagnostik lebender und fos- gan Academy of Science, Arts, and Letters 38:9–19. siler Koniferen-Hölzer. Palaeontographica B 89: ASH, S.R. 1991. A new Jurassic flora from the Wallow Ter- 83–203. rane in Hells Canyon, Oregon and Idaho. Oregon LEMOIGNE, Y. 1968. Un nouveu genre de structure ligneuse Geology 53(2):27–33. de type gymnospermien: Embergerixylon nov. g. ASH, S.R., AND BASINGER, J.F.1991. A high latitude Upper Extrait des Annales de la Société Geologique du Triassic flora from the Heiberg Formation, Sverdrup Nord 88:155–157. Basin, Arctic Archipelago. Geological Survey of Can- MEDLYN, D.A., AND W. D. T IDWELL. 1975. Conifer wood ada Bulletin 4112:101–131. from the Upper Jurassic of Utah, part I. Xenoxylon morrisonense sp. nov. American Journal of Botany BAKER, G.K. 1965. An environmental study of the Morri- 62:203–208. son Formation Freezeout Hills–Como Bluff Area, Car- ______. 1979. A review of the genus Protopiceoxylon with bon County, Wyoming. Unpublished master’s thesis, emphasis on North American species. Canadian University of Wyoming, Laramie. Journal of Botany 57:1451–1463. BAMFORD, M.K., AND I.B. CORBETT. 1994. Fossil wood of ______. 1992. Petrified conifers from the Upper Jurassic Cretaceous age from the Namaqualand Continental Morrison Formation in the western U.S.A. Organisa- Shelf, South Africa. Palaeontologia Africana 31:83–95. tion Internationale de Paléobotanique, IVeme Confér- BHARDWAJ, D.C. 1953. Jurassic woods from the Rajmahal ence, Paris, Résumés des Communications, p. 107. Hills, Bihar. Palaeobotanist 2:59–70. NISHIDA, M. 1966. On some petrified plants from the Cre- DELEVORYAS, T. 1960. Investigation of North American taceous of Choshi, Chiba Prefecture III. Botanical cycadeoids: trunks from Wyoming. American Journal Magazine, Tokyo 79:226–235. of Botany 47:778–786. NISHIDA, M., AND H. NISHIDA. 1983. Petrified plants from DERR, M.E. 1974. Sedimentary structure and depositional the Cretaceous of the Kwanto Mountains, central environments of paleochannels in the Jurassic Mor- Japan–I. Botanical Magazine, Tokyo 96:85–91. rison Formation near Green River, Utah. Brigham OSTROM, J.H., AND J.S. MCINTOSH. 1966. Marsh’s dino- Young University Geology Studies 21:3–39. saurs. Yale University, New Haven, CT. 2002] MESEMBRIOXYLON OBSCURUM, COMB. NOV. 217

PHILLIPE, M., AND F. T HÉVENARD. 1996. Distribution and Colorado, part II. Mesembrioxylon stokesi. Great palaeoecology of the Mesozoic wood genus Xenoxy- Basin Naturalist 44:257–262. lon: palaeoclimatological implications for the Juras- TIDWELL, W.D. 1990. Preliminary report on the megafos- sic of western Europe. Review of Palaeobotany and sil flora of the Upper Jurassic Morrison Formation. Palynology 91:353–370. Hunteria 2(8):1–11. RAMANUJAM, C.J.K. 1953. On two species of Mesembrioxy- ______. 1998. Common fossil plants of western North lon from the vicinity of Pondicherry, South India. America. 2nd edition. Smithsonian Institution Press, Palaeobotanist 2:101–106. Washington, DC. SAHNI, B. 1920. Petrified plant remains from the Queens- TIDWELL, W.D., B.B. BRITT, AND S.R. ASH. 1998. Prelimi- land Mesozoic and Tertiary formations. Queensland nary floral analysis of the Mygatt-Moore Quarry in Geological Survey Publication 267:7–38. the Upper Jurassic Morrison Formation, west-cen- ______. 1931. Revision of Indian fossil plants: Coniferales tral Colorado. Modern Geology 22:341–378. (b. Petrifactions). Memoirs Geological Survey of India, TIDWELL, W.D., AND D.A. MEDLYN. 1992. Short shoots Palaeontologia Indica (N.S.) 2:51–124. from the Upper Jurassic Morrison Formation, Utah, ______. 1937. A Mesozoic coniferous wood, Mesembrioxy- Wyoming, and Colorado, USA. Review of Palaeo- lon shanense sp. nov., from the southern Shan States botany and Palynology 71:219–238. of Burma. Records of the Geological Survey of India ______. 1993. Conifer wood from the Upper Jurassic of 71:380–388. Utah, USA, part II. Araucarioxylon hoodii sp. nov. SEWARD, A.C. 1919. Fossil plants. Volume IV. Cambridge Palaeobotanist 42:70–77. University Press, London. TORRES, T., B. CORTINAT, AND H. MEON. 1991. Microcachry- SHIMAKURA, M. 1936. Studies on fossil woods from Japan xylon gothani nov. gen. n. sp., Madera Cretacica do and adjacent lands, I. The Cretaceous woods from la Isla James Ross, Antarctica. Congreso Geologico Japan and Manchoukuo. Scientific Report of the Chilene 7:1702–1706. Tohoku Imperial University Series 2 (Geol.) 18: WARD, L.F. 1900a. Status of the Mesozoic floras of the 267–298. United States. U.S. Geological Survey Twentieth STOKES, W.L. 1986. Geology of Utah. Utah Geological and Annual Report (1898–1899), part 2:211–748. Mineral Survey, Department of Natural Resources, ______. 1900b. Description of a new genus and twenty Salt Lake City. new species of fossil cycadean trunks from the Juras- STOPES, M.C. 1915. Catalogue of Mesozoic plants in the sic of Wyoming. Proceedings of the Washington British Museum—the Cretaceous flora, part II. Academy of Sciences 1:253–300. Longman, London. THAYN, G.F., AND W. D. T IDWELL. 1984. Flora of the Lower Received 26 July 2000 Cretaceous Cedar Mountain Formation of Utah and Accepted 8 January 2001 Western North American Naturalist 62(2), © 2002, pp. 218–222

DO JAGUARS (PANTHERA ONCA) DEPEND ON LARGE PREY?

Carlos A. López González1,2 and Brian J. Miller1

ABSTRACT.—The jaguar (Panthera onca) has been classified as an opportunistic hunter that takes as many as 85 prey species, according to availability. In this study we analyzed jaguar food habits throughout its range to quantify the importance of small, medium, and large prey in the diet. Because peccaries (Tayassu) are present in most studies, we also tested their importance in relation to other prey items. We conclude that jaguars are equally using medium- and large-size prey, with a trend toward use of larger prey as distance increases from the equator. There was no significant difference between the importance of peccaries and other large prey.

Key words: jaguar, Panthera onca, food habits, prey.

The jaguar (Panthera onca) is one of the also analyze peccaries both as part of the large least studied of the large felids, with most prey category and separate from the other research carried out in the last 20 years (see large prey to gain an understanding of possi- review by Sunquist in press). Because jaguars ble evolutionary links. are considered opportunistic predators, feed- ing on as many as 85 different prey items (Sey- METHODS mour 1989), it can be difficult to predict which prey species are regionally important, a point To test for preferences in jaguar diets, we that is critical to conservation strategies. reviewed dietary studies of jaguars in 10 dif- In tropical areas where diets of jaguars have ferent geographic sites ranging from 25 degrees been studied, they generally eat medium- and South to 19 degrees North. Study sites included large-sized mammals (Schaller and Vasconce- (north to south) Jalisco, Mexico (Nuñez et al. los 1978, Mondolfi and Hoogestijn 1986, Rabi- 2000), Campeche, Mexico (Aranda and Sanchez- nowitz and Nottingham 1986, Emmons 1987, Cordero 1996), Belize (Rabinowitz and Not- Aranda 1994, Crawshaw 1995, Aranda and tingham 1986), Costa Rica (Chinchilla 1997), Sanchez-Cordero 1996, Chinchilla 1997, Taber Caatinga, Brazil (Olmos 1993), Peru (Emmons et al. 1997, Crawshaw and Quigley in press). 1987), Peru (Kuroiwa and Ascorra in press), In studies that analyzed more than 25 scats, 2 Paraguay (Taber et al. 1997), Argentina (Perovic reported a preference for peccaries (Tayassu in press), Iguazu, Brazil (Crawshaw 1995). A pecari and T. tajacu; Crawshaw 1995, Aranda summary of the data used for the present and Sanchez-Cordero 1996), 2 reported pref- analysis is presented in Table 1. Habitat type erence for deer (Mazama gouazoubira and was obtained from the original publication, Odocoileus virginianus; Taber et al. 1997, Nunez and when this description was absent we used et al. 2000), 1 reported preference for arma- the World Wildlife Fund classification. Human dillo (Dasypus novemcinctus; Rabinowitz and impact for each study site was classified as Nottingham 1986), and 1 reported preference low, medium, or high based on the description for reptiles (Emmons 1987). Four of these stud- in the original manuscript, and we included ies indicated use of large prey and 2 of those the status of the area as protected or not. reported a dominant use of peccaries. Pecca- To standardize data and reduce bias associ- ries have a range that coincides with the ated with frequencies of occurrence (Ciucci et jaguar, implying that those 2 species may have al. 1996), all data were converted to relative an evolutionary link (Aranda 1994). biomass consumed. Prey biomass consumed In this study we analyze the value of large-, per scat produced was regressed against live medium-, and small-sized prey to jaguars. We body weight of the prey animals to determine

1Department of Conservation Biology, Denver Zoological Foundation, 2300 Steele St., Denver, CO 80205. 2Corresponding author. Present address: Sonoran Jaguar Conservation Project, 2114 W. Grant #121, Tucson, AZ 85745.

218 2002] JAGUAR FOOD HABITS 219

TABLE 1. Database summary used to develop the present analysis. No. of Region (Latitude) scats Habitat type Human impact Source Jalisco (19°N) 47 Tropical dry forest Low, protected area Nuñez et al. 2000 Campeche (18°N) 37 Tropical seasonal Low, protected area Aranda and Sanchez- flooded forest Cordero 1996 Belize (17°N) 228 Tropical rainforest High, nonprotected area Rabinowitz and Nottingham 1986 Costa Rica (8°N) 22 Tropical rainforest Low, protected area Chinchilla 1997 Peru–Cocha Cashu (8°S) 25 Tropical rainforest Low, protected area Emmons 1987 Peru–Madre de Dios 13 Tropical flooded forest— Low, protected area Kuroiwa and Ascorra (11°S) riparian vegetation in press Brazil–Caatinga (11°S) 8 Tropical dry forest Low, nonprotected area Olmos 1993 Paraguay (20°S) 106 Tropical dry forest Low to high, nonprotected area Taber et al. 1997 Argentina (23°S) 246 Tropical dry forest Low to high, nonprotected area, Perovic in press fragmented Brazil–Iguazu (25°S) 73 Tropical rainforest High, protected area, Crawshaw 1995 fragmented

the relationship between body weight of prey 1984). Data from a given study site also were and scats produced. The resulting linear rela- plotted against latitude to search for prey class tionship (Y = 1.98 + 0.035X; the letter Y indi- patterns along a north–south gradient. P > 0.05 cates biomass of prey consumed per scat and was considered nonsignificant. X is prey body mass in kg) was then applied in the form of correction factor to convert fre- RESULTS quency of occurrence values for each taxon to a relative estimate of biomass of each con- Percent biomass consumed was similar sumed (Floyd et al. 1978, Ackerman et al. 1984). between medium- and large-prey categories, The linear model used to convert frequency of with a very low proportion of prey use from occurrence to biomass consumed was devel- the small category. A clinal pattern was evi- oped for pumas (Puma concolor; Ackerman et dent from the comparison of prey and latitude. al. 1984) and was applied to jaguars under the Jaguars living farther away from the equator assumption of similar digestive tracts. A simi- used larger prey more frequently, whereas lar analysis has been applied to jaguars (Nuñez jaguars living nearer the equator depended et al. 2000), leopards (P. pardus), and tigers (P. more heavily on medium-sized prey (Fig. 1A). tigris; Karanth and Sunquist 1995). Frequency Across all studies, average proportions of of occurrence and percent occurrence typi- prey consumed by jaguars was 4.32 ± 7.32% cally overestimate the importance of small for small-sized prey, 47.65 ± 26.84% for med- prey and underestimate the value of large prey ium-sized prey, and 48.03 ± 26.15% for large- in the diet (Ackerman et al. 1984, Karanth and sized prey. Relative percent biomass differed Sunquist 1995, Nuñez et al. 2000). Weights of among size classes (F = 26.076, df =2, P < most prey items were obtained from the re- 0.001). Small-prey consumption was signifi- spective studies; otherwise, we referred to cantly different from consumption of medium- Emmons (1997) and Reid (1997). sized prey (q = 8.905, P < 0.05) and large prey Prey were grouped into 3 categories: small (q = 8.783, P < 0.05). There was no difference (<1 kg), medium (1–10 kg), and large (>10 kg). in consumption of medium- and large-sized In a 2nd analysis prey were grouped into 4 prey (q = 0.122, P > 0.05). categories: small, medium, large excluding When peccaries were treated as a separate peccaries, and peccaries. Percent biomass for prey category from small-, medium-, and large- each prey category was arcsin transformed sized prey, percent biomass still differed among and compared for each study site and for data categories (F =10.435, df =3, P < 0.001). pooled across all study sites. These data were Although peccaries are present in most jaguar analyzed using a 1-way ANOVA, and signifi- diets (mean biomass = 25.7 ± 15.1%), their cant differences were detected using Student- value to the jaguar diet is not significantly dif- Neuman-Keuls multiple comparison tests (Zar ferent from the rest of the large-sized prey 220 WESTERN NORTH AMERICAN NATURALIST [Volume 62

(Branch et al. 1996). Average mass of medium- sized prey was 4.0 ± 2.2 kg, which should be enough to maintain a large cat (considering the energetic model by Ackerman et al. 1986). Large prey, however, may play a more impor- tant role when females have kittens (see Ack- erman et al. 1986). In the studies we reviewed, researchers had no reliable way of determining which individ- ual jaguar left a given scat, and this may have confounded analyses. Analyses may have been subject to pseudoreplication, with one individ- ual contributing more heavily to results (see Hurlbert 1984). Ross et al. (1997) reported that food habits of solitary female cats can be sig- nificantly different from those of males. Use of medium-sized prey by jaguars is also likely an artifact of human disturbance in a region. Unregulated harvest of large- and medium-sized prey by humans can signifi- cantly alter an (Redford and Robin- son 1987), and the disappearance of favored prey can force jaguars to prey upon livestock (Ackerman et al. 1986, Hoogesteijn et al. 1993). Prey declines also cause animals to move over greater distances, thereby increasing their vul- nerability (Woodroffe and Ginsberg 1998). We conclude that jaguars can use both medium- and large-sized prey that are avail- Fig. 1. North–south clinal variation in the percentage of able and behaviorally vulnerable (i.e., present biomass consumed by jaguars from prey categories. Data from Aranda and Sanchez 1996, Chinchilla 1997, Craw- in large groups, predictable distributions). We shaw 1995, Emmons 1987, Kuroiwa and Ascorra in press, do not conclude, however, that either medium- Nunez et al. 2000, Olmos 1993, Perovic in press, Rabi- or large-sized prey can be replaced adequately nowitz and Nottingham 1986, Taber et al. 1997. Graph A by the other category if prey in one category contains 3 categories: small, medium, and large prey. Graph B contains 4 categories: small, medium, large, and declines significantly. peccaries. From this review we could not conclude that peccaries, or any single prey species, were an important factor in jaguar evolution, items (q = 0.891, P > 0.05; Fig. 1B). The use largely due to the flexibility jaguars exhibit in of small-sized prey by jaguars was different prey acquisition (see Seymour 1989). In addi- from medium-sized prey (q = 7.862, P > 0.05), tion, humans have drastically altered the suite large-sized prey (q = 3.503, P < 0.05), and of available prey, making it difficult to draw peccaries (q = 4.394, P < 0.05). Use of pecca- evolutionary conclusions from recent informa- ries was significantly different from use of tion on prey selection. medium-sized prey (q = 3.469, P < 0.05). Jaguar fossils exist in North America from the mid-Pleistocene about 1.5 million years DISCUSSION ago (Seymour 1989, Turner 1997). During the mid-Pleistocene jaguars ranged over South and In our review of dietary studies, jaguars North America as far north as Washington, were not dependent on large prey and appar- Nebraska, and Maryland, but in the Recent ently can survive on medium-sized prey such Epoch, the northern limit has been southern as has been reported for leopards (Bothma and Arizona, New Mexico, and Texas (Seymour Le Riche 1986, Bailey 1993), cheetahs (Aci- 1989, Brown and López González 2000). nonyx jubatus; Laurenson 1995), and pumas Because of this range reduction, Kurten and 2002] JAGUAR FOOD HABITS 221

Anderson (1980) stated that jaguars in their Management in the Tropics Series, Columbia Uni- present range constitute a relict population of versity Press, New York. 429 pp. BOTHMA, J. DU P., AND E.A.N. LE RICHE. 1986. Prey pref- what was once a more widely distributed Hol- erence and hunting efficiency of the Kalahari desert arctic form (see also Seymour 1989). Indeed, leopard. Pages 389–415 in S.D. Miller and D.D. jaguar fossils in the north are older than those Everett, editors, Cats of the world: biology, conser- found in Central and South America, and vation, and management. National Wildlife Federa- tion, Washington, DC. North American fossils outnumber fossils of BRANCH, L.C., M. PESSINO, AND D. VILLAREAL. 1996. South America by 73 to 18 (Seymour 1989). Response of pumas to a population decline of the Turner (1997) has proposed that the jaguar plains vizcacha. Journal of Mammalogy 77:1132–1140. was likely driven from the more open habitat BROWN, D.E., AND C.A. LÓPEZ GONZÁLEZ. 2000. Notes on the occurrences of jaguars (Panthera onca) in Ari- in the northern part of its range by the later zona and New Mexico. Southwestern Naturalist 45: appearance of the lion (P. atrox) in North 537–542. America. CHINCHILLA, R.F.1997. La dieta del jaguar (Panthera onca), In contrast, the oldest peccary (Tayassu) rec- el puma (Felis concolor) y el manigordo (Felis pardalis) (Carnivora: Felidae) en el Parque Nacional Corcov- ords in the Americas are from the Late Pleis- ado, Costa Rica. Revista de Biologia Tropical 45: tocene to early Recent (Mayer and Wetzel 1223–1229. 1987), and the boundary between those 2 CIUCCI, P., L. BOITANI, E.R. PELLICIONI, M. ROCCO, AND epochs was about 11,000 years ago. Tayassu I. GUY. 1996. A comparison of scat-analysis methods probably underwent most of its evolution in to assess the diet of the wolf (Canis lupus). Wildlife Biology 2:37–48. Central and South America (Mayer and Wet- CRAWSHAW, P.G. 1995. Comparative ecology of ocelot zel 1987). Potential differences in time and (Felis pardalis) and jaguar (Panthera onca) in a pro- centers of evolution would reduce the likeli- tected subtropical forest in Brazil and Argentina. hood of co-evolution between peccaries and Doctoral dissertation, University of Florida, Gaines- ville. jaguars. It also indicates that although pecca- CRAWSHAW, P.G., AND H.G. QUIGLEY. In press. Habitos ali- ries are certainly important jaguar prey, the mentarios del jaguar y puma. In: R.A. Medellin, C. range overlap between those 2 species may be Chetkiewitz, A. Rabinowitz, K.H. Redford, J.G. correlated, but not causal. Robinson, E. Sanderson, and A. Taber, editors, El jaguar en el nuevo milenio: una evaluacion de su estado, deteccion de prioridades y recomendaciones ACKNOWLEDGMENTS para la conservacion de los jaguares en America. Fondo de Cultura Economica-Universidad Nacional This work was supported in part by Joe and Autonoma de Mexico, Mexico, D.F. EMMONS, L.H. 1987. Comparative feeding ecology of Valer Austin, Arizona Zoological Society, Eve- felids in a Neotropical rainforest. Behavior, Ecology lyn Delgado, Denver Zoological Foundation, and Sociobiology 20:271–283. Earthwatch Institute, Idea Wild Inc., Lincoln ______. 1997. Neotropical rainforest mammals: a field Zoo Neotropical Fund, Lyn Chase Wildlife guide. 2nd edition. University of Chicago Press, Foundation, Malpai Borderlands Group, Turner Chicago IL. 281 pp. FLOYD, T.J., L.D. MECH, AND P.A. JORDAN. 1978. Relating Foundation, and Wildlife Conservation Society. wolf scat content to prey consumed. Journal of Wild- life Management 42:528–532. HOOGESTIJN, R., A. HOOJESTIJN, AND E. MONDOLFI. 1993. LITERATURE CITED Jaguar predation and conservation: cattle mortality caused by felines on three ranches in the Venezuelan ACERMAN, B.B., F.G. LINDZEY, AND T.P. H EMKER. 1984. Llanos. Symposium of the Zoological Society of Lon- Cougar food habits in southern Utah. Journal of don 65:391–407. Wildlife Management 48:147–155. HURLBERT, S.H. 1984. Pseudoreplication and design of ______. 1986. Predictive energetics model for cougars. ecological field experiments. Ecological Monographs Pages 333–352 in S.D. Miller and D.D. Everett, edi- 54:187–211. tors, Cats of the world: biology, conservation, and KARANTH, U.K., AND M.E. SUNQUIST. 1995. Prey selection management. National Wildlife Federation, Wash- by tiger, leopard and dhole in tropical forests. Journal ington, DC. of Animal Ecology 64:439–450. ARANDA, M. 1994. Importancia de los pecaries (Tayassu KURTEN, B., AND E. ANDERSON. 1980. Pleistocene mammals spp.) en la alimentación del jaguar (Panthera onca). of North America. Columbia University Press, New Acta Zoologica Mexicana (ns):62:11–22. York. 443 pp. ARANDA, M., AND V. S ANCHEZ CORDERO. 1996. Prey spec- KUROIWA, A., AND A. ASCORRA. In press. Dieta y densidad tra of jaguar (Panthera onca) and puma (Puma con- de posibles presas del jaguar (Panthera onca) en las color) in tropical forest of Mexico. Studies in Neo- inmediaciones del Tambopata research centre (Zona tropical Fauna and Environment 31:62–64. reservada Tambopata-madre de Dios). In: R.A. BAILEY, T.N. 1993. The African leopard: ecology and Medellin, C. Chetkiewitz, A. Rabinowitz, K.H. Red- behavior of a solitary felid. Biology and Resource ford, J.G. Robinson, E. Sanderson, and A. Taber, edi- 222 WESTERN NORTH AMERICAN NATURALIST [Volume 62

tors, El jaguar en el nuevo milenio: una evaluacion REID, F.A. 1997. A field guide to the mammals of Central de su estado, deteccion de prioridades y recomenda- America and southeast Mexico. Oxford University ciones para la conservacion de los jaguares en Amer- Press, New York. 334 pp. ica. Fondo de Cultura Economica-Universidad ROSS, P.I., M.G. JALKOTSKY, AND M. FESTA-BLANCHET. 1997. Nacional Autonoma de Mexico, Mexico, D.F. Cougar predation on bighorn sheep in southwestern LAURENSON, M.K. 1995. Cub growth and maternal care in Alberta during winter. Canadian Journal of Zoology cheetahs. Behavioral Ecology 6:405–409. 75:771–775. MAYER, J.J., AND R.M. WETZEL. 1987. Tayassu pecari. SCHALLER, G.B., AND J. VASCONCELOS. 1978. Jaguar pre- Mammalian Species 293:1–7. dation on capybara. Zeitschrift Saugetierkunde 43: MONDOLFI, E., AND R. HOOGESTEIJN. 1986. Notes on the 296–301. biology and status of the jaguar in Venezuela. Pages SEYMOUR, K.L. 1989. Panthera onca. Mammalian species 125–146 in S.D. Miller and D.D. Everett, editors, 340:1–9. Cats of the world: biology, conservation, and man- SUNQUIST, M.E. In press. History of jaguar research in the agement. National Wildlife Federation, Washington, Americas. In: R.A. Medellin, C. Chetkiewitz, A. DC. Rabinowitz, K.H. Redford, J.G. Robinson, E. Sand- NUÑEZ, R., B. MILLER, AND F. G . L INDZEY. 2000. Food erson, and A. Taber, editors, El jaguar en el nuevo habits of jaguars and pumas in Jalisco, Mexico. Jour- milenio: una evaluacion de su estado, deteccion de nal of Zoology, London 252:373–379. prioridades y recomendaciones para la conservacion OLMOS, F. 1993. Notes on the food habits of Brazilian de los jaguares en America. Fondo de Cultura Eco- “Caatinga” . Mammalia 57:126–130. nomica-Universidad Nacional Autonoma de Mexico, PEROVIC, P.G. In press. Conservacion del jaguar en el Noro- Mexico, D.F. este de Argentina. In: R.A. Medellin, C. Chetkie- TABER, B.A., A. NOVARO, AND N. NERIS. 1997. The food witz, A. Rabinowitz, K.H. Redford, J.G. Robinson, habits of sympatric jaguar and puma in the Para- E. Sanderson, and A. Taber, editors, El jaguar en el guayan Chaco. Biotropica 29:204–213. nuevo milenio: una evaluacion de su estado, detec- TURNER, A. 1997. The big cats and their fossil relatives. cion de prioridades y recomendaciones para la con- Columbia University Press, New York. 233 pp. servacion de los jaguares en America. Fondo de Cul- WOODROFFE, R., AND J.R. GINSBERG. 1998. tura Economica-Universidad Nacional Autonoma de and the extinction of populations inside protected Mexico, Mexico, D.F. areas. Science 280:2126–2128. RABINOWITZ, A., AND B. NOTTINGHAM. 1986. Ecology and ZAR, J.H. 1984. Biostatistical analysis. 2nd edition. Pren- behavior of the jaguar (Panthera onca) in Belize. tice Hall, Englewood Cliffs, NJ. 718 pp. Journal of Zoology, London 210:149–159. REDFORD, K.H., AND J.G. ROBINSON. 1987. The game of Received 30 May 2000 choice: patterns of Indian and Colonist hunting in Accepted 12 February 2001 the Neotropics. American Anthropologist 89:650–667. Western North American Naturalist 62(2), © 2002, pp. 223–227

ASSESSMENT OF CRYPTOBIOTIC CRUST RECOVERY

Robert C. Rychert1

ABSTRACT.—Chlorophyll a and b analyses of cryptobiotic crusts from burned and unburned sites on the Snake River plain provided an estimate of crust recovery. In less than 3 years, chlorophyll a levels were approximately 50% of the unburned reference site. Chlorophyll a/b ratios indicated that the biotic composition of the recovering cryptobiotic crust consisted largely of eukaryotic algae and bryophytes.

Key words: cryptobiotic crusts, chlorophyll, fire recovery, Snake River plain.

Relatively undisturbed arid and semiarid increase in algal/cyanobacterial density of soil landscapes can exhibit a soil surface cover of crusts during the first 16 months following microphytes including mosses, lichens, algae, fire. It was also observed that immediately fol- fungi, and cyanobacteria (Lange et al. 1992). lowing fire the algal/cyanobacterial/moss crust, Termed cryptogamic or microbiotic (Johansen although dead, was still intact. Estimates of 1993), cryptobiotic (Belnap et al. 1994), or algal and cyanobacterial recovery time in terms microphytic crusts (Beymer and Klopatek of biomass and diversity range from 16 months 1991), these biological assemblages protect to 5 years (Johnasen et al. 1982, 1993). Recov- the soil from erosion (Harper and Marble 1988) ery appears to be highly dependent on precip- and function in carbon cycling (Beymer and itation patterns and composition of the crust, Klopatetk 1991) and nitrogen cycling (Rychert which are in turn influenced by soil character- et al. 1978). istics and climate (Johansen et al. 1993). There has been increasing interest in trying This study is a preliminary attempt to esti- to assess recovery rates of cryptobiotic crusts mate the recovery of cryptobiotic crusts on after disturbance (e.g., overgrazing, fire, agri- Idaho rangeland sites in the lower Snake River culture). Belnap (1993) found that the use of plains of Idaho following disturbance by fire. spectrophotometrically determined chlorophyll The use of chlorophyll determinations to pro- a as a measure of recovery of cryptobiotic vide both qualitative and quantitative estima- crusts was both time-efficient and reliable. tions of cryptobiotic crust recovery was evalu- Chlorophyll a levels increased at rates of 1.0– ated. 2.6% per year after disturbance at various Utah sites. MATERIALS AND METHODS Fire can destroy microbiotic crusts. Greene et al. (1990) observed reduction in coverage of Cryptobiotic crust samples were collected soil crusts following fire. Soil erosion rate and nearly 3 years after the 1996 Kuna Butte fire the amount of bare soil surface per unit area (location: southwest of Boise, Idaho, Township increased, and a decline in aggregate stability 2 North, Range 1 West, Section 4). The light- in the 0–1 cm horizon was also observed. ning-caused fire began on 26 August 1996 and Johansen et al. (1982, 1993) observed signifi- burned 5785 acres. The sampling area was for- cant reduction in the biomass of algal/cyano- merly a Wyoming big sagebrush community bacterial soil crusts after burning, although similar to an adjacent unburned area. The species diversity remained constant. burned area was dominated by cheatgrass Recovery rates of soil crusts following fire (Bromus tectorum) and tumblemustard (Sisym- can vary depending on the organisms involved brium altissimum). Most of the burned area and climatic conditions. In the lower Columbia still has good cover of perennial grasses, pri- Basin, Johansen et al. (1993) observed little marily Sandberg’s bluegrass (Poa secunda) and

1Biology Department, Boise State University, Boise, ID 83725.

223 224 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 1. Chlorophyll concentrations: burned vs. unburned sites. Chlorphyll a Chlorophyll b (mg m–2) (mg m–2) a/b ratio ______Standard Coefficient Standard Coefficient Standard Coefficient Site Meana deviation of variation Meanb deviation of variation Meanc deviation of variation

Unburned 130.163 50.690 38.943 53.259 18.146 34.071 2.493 0.808 32.418 Burned 65.159 26.258 40.298 39.647 14.421 36.372 1.658 0.369 22.274 at test, significantly different, P = 0.0020. bt test, not significantly different, P = 0.0797. ct test, significantly different, P = 0.0082. bottlebrush squirreltail (Elymus elymoides) Chlorophyll concentrations are expressed as amongst the cheatgrass. The fire was high mg m–2, extrapolated from the 1.73 × 10–3 m2 intensity, resulting in destruction of the bio- soil core(s). logical crust. Trace amounts of moss were Statistical analyses were run with SAS for observed on the soil surface during sampling. Windows, version 8 (SAS Institute, Inc. 1999). Observations in the nearby Poen fire suggest that crust cover does not become visibly RESULTS AND DISCUSSION apparent until approximately 3–4 years follow- ing a high-intensity fire. Chlorophyll a concentrations, chlorophyll b Vegetation in the adjacent unburned area is concentrations, and chlorophyll a/b ratios are Wyoming big sagebrush (Artemisia tridentata higher for the unburned than the burned site ssp. wyomingensis) with an understory of blue- (Table 1). The burned site had chlorophyll a bunch wheatgrass (Pseudoroegneria spicata) levels that are 50% of the unburned site. Mean and Thurber’s needlegrass (Achnatherum thur- chlorophyll a concentrations for both the berianum). Sandberg’s bluegrass and bottle- burned (65 mg m–2) and control (130 mg m–2) brush squirreltail are also present. The soil sites were much higher than the levels deter- surface in the unburned area has a nearly mined by Belnap et al. (1994) with disturbed complete cover of biological soil crust domi- cryptobiotic soil crusts in Utah (approximately nated by mosses (primarily Bryum spp.) and 10–50 mg m–2), but they are similar to lichens. The soil of the sampling area is classi- endolithic microphytic chlorophyll a levels fied as a fine-silty, mixed, mesic Xerollic Hap- reported by Bell and Sommerfield (1987; 87 largid. The texture is a silt loam. mg m–2) on the Colorado Plateau, and by Soil “cores” consisting of cryptobiotic crust Matthes-Sears et al. (1997; 73 mg m–2) in plus soil were collected with sterile plastic Ontario, Canada. Recovery of the cryptobiotic petri dishes serving as coring devices (1 cm crust probably does not occur as a linear func- deep and 47 mm in diameter). Ten samples tion; however, in less than 3 years, chlorophyll were collected from both the burned and un- a values are approximately half recovered. burned areas and kept on ice. Chlorophyll pig- Chlorophyll a/b ratios may provide some ments were extracted with dimethylsulfoxide indication of soil crust recovery after distur- (DMSO) the same day (12 March 1999), fol- bance. Table 2 lists some chlorophyll a/b ratios lowing the procedures of Ronen and Galun from a variety of algae, plants, and lichens. (1984; 60°C for 50 minutes). Following filtra- Cyanobacteria (blue-green algae) lack chloro- tion of the extracts through a Whatman #1 fil- phyll b (Shoaf and Lium 1976). So, if the soil- ter, absorbance readings at 750 nm, 665 nm, stabilizing, nutrient-cycling, and nitrogen-fix- 648 nm, 435 nm, and 415 nm were made with ing functions of cyanobacterial-lichen crusts a Spectronic 20 Genesys spectrophotometer. represent the undisturbed ecological condition Chlorophyll a (Ca) and chlorophyll b (Cb) for arid and semiarid soil surfaces (Rychert et concentrations were determined using the al. 1978, Buttars et al. 1998), higher chloro- equations of Barnes et al. (1992): phyll a/b ratios may represent a useful assess- ment parameter. In this study the burned site –1 Ca (in mg L ) = 14.85 A665 – 5.14 A648 exhibited a chlorophyll a/b ratio of 1.66, sug- –1 Cb (in mg L ) = 25.48 A648 – 7.36 A665 gesting that cryptobiotic recovery after fire 2002] CRYPTOBIOTIC CRUST RECOVERY 225

TABLE 2. Chlorophyll a/b ratios found in literature.

Biocomponent N Mean a/b s sx– Range Reference Vascular plants 10 2.38 0.56 0.24 1.51–3.23 Barnes et al. (1992) Moss 1 2.51 Sheridan and Rosentreter (1973) Chlorophyta 7 3.00 0.87 0.29 1.89–4.16 Shoaf and Lium (1976) Lichens, algae, and cyanobacteria (mixture) 1 3.69 Matthes-Sears et al. (1977) Lichens 8 6.09 2.50 0.41 2.84–10.18 Barnes et al. (1992)

involves mainly eukaryotic algae and bryo- cryptobiotic crust is changing. Temporal stud- phytes. The chlorophyll a/b ratio for the refer- ies exhibiting higher or increasing ratios ence site was 2.49, probably reflecting the would suggest development or return of cyano- predominant moss (Bryum spp.) cover, al- bacteria and lichens as components of the crust. though a few lichens were present. Chlorophyll a degradation to phaeophytin a ACKNOWLEDGMENTS occurs for a variety of reasons including ex- traction with solvents, air pollution, and expo- I thank Marcia Wicklow-Howard, Julie sure to air (oxidation). The state of degradation Kaltenecker, and Roger Rosentreter for their of chlorophyll a can be measured by the ab- helpful comments and Larry Taylor for his sorbancy at 435/415 nm ratio (Ronen and Galun technical assistance. This work was supported 1984). The mean A 435/415 in this study was by a challenge cost-share grant from the 0.933 for the unburned site (s 0.065, sx– 0.070) Bureau of Land Management. and 0.898 for the burned site (s 0.032, sx– = 0.40). This is very similar to the degradation LITERATURE CITED noted by Belnap et al. (1994). Based upon the mixtures tested by Ronen and Galun (1984), BARNES, J.D., L. BALAGUER, E. MANRIQUE, S. ELVIRA, AND A.W. DAVISON. 1992. A reappraisal of the use of just under 50% of the chlorophyll a has been DMSO for the extraction and determination of degraded. Freezing and thawing, characteris- chlorophylls a and b in lichens and higher plants. tic of the spring climate, would tend to reduce Environmental and Experimental Botany 32:85–100. chlorophyll a/b ratios due to the greater effect BELL, R.A., AND M.R. SOMMERFIELD. 1987. Algal biomass on chlorophyll a (Barnes et al. 1992). However, and within a temperate zone sandstone. American Journal of Botany 74:294–297. in this study the A 435/415 ratios are nearly BELNAP, J. 1993. Recovery rates of cryptobiotic crusts: identical; therefore, chlorophyll a/b ratios and inoculant use and assessment methods. Great Basin chlorophyll a concentrations for the burned Naturalist 53:89–96. and reference sites can be compared unam- BELNAP, J., K.T. HARPER, AND S.D. WARREN. 1994. Surface disturbance of cryptobiotic soil crusts: nitrogenase biguously. activity, chlorophyll content, and chlorophyll degra- This study represents the first analysis of dation. Arid Soil Research and Rehabilitation 8:1–8. chlorophyll a in cryptobiotic crusts from the BEYMER, R.J., AND J.M. KLOPATEK. 1991. Potential contri- Snake River plain. The crust recovery rate bution of carbon by microphytic crusts in pinyon- estimated from this study reflects crust assem- juniper woodlands. Arid Soil Research and Rehabili- blage and climatic differences from other sites tation 5:187–198. BUTTARS, S.M., L. ST. CLAIR, J.R. JOHANSEN, J.C. SPRAY, (Belnap 1993). Recovery rates may be site spe- M.C. PAYNE, B.L. WEBB, R.E. TERRY, B.K. PENDLE- cific. Selective sampling primarily of moss- TON, AND S.D. WARREN. 1998. Pelletized cyanobacte- and lichen-covered sites would provide impor- rial soil amendments: laboratory testing for survival, tant additional information on in situ chloro- escapability, and nitrogen fixation. Arid Soil Research and Rehabilitation 12:165–178. phyll a/b ratios. GREENE, R.S. B., C.J. CHARTRES, AND K.C. HODGKINSON. Monitoring chlorophyll a and b concentra- 1990. The effects of fire on the soil in a degraded tions can provide 2 important indicators of semi-arid woodland. I. Cryptogam cover and physi- cryptobiotic crust recovery. Chlorophyll a lev- cal and micromorphological properties. Australian els are an estimate of cryptobiotic crust bio- Journal of Soil Research 28:755–777. HARPER, K.T., AND J.R. MARBLE. 1988. A role for nonvas- mass, while chlorophyll a/b ratios are an indi- cular plants in management of arid and semiarid cator of how the biotic composition of the rangelands. Pages 135–169 in P.T. Tueller, editor, 226 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Vegetation science applications for rangeland analy- cliffs of the Niagara escarpment, Canada. Interna- sis and management. Kluwer Academic Publishers, tional Journal of Plant Sciences 158:451–460. Boston. RONEN, R., AND M. GALUN. 1984. Pigment extraction from JOHANSEN, J.R. 1993. Cryptogamic crusts of semiarid and lichens with dimethylsulfoxide (DMSO) and estima- arid lands of North America. Journal of Phycology tion of chlorophyll degradation. Environmental and 29:140–147. Experimental Botany 24:239–245. JOHANSEN, J.R., J. ASHLEY, AND W.R. RAYBURN. 1993. Effects RYCHERT, R.C., J. SKUJINS, D. SORENSEN, AND D. PORCELLA. of range fire on soil algal crusts in semiarid shrub- 1978. Nitrogen fixation by lichens and free-living steppe of the lower Columbia Basin and their subse- in deserts. Pages 20–30 in N.E. quent recovery. Great Basin Naturalist 53:73–88. West and J. Skujins, editors, Nitrogen in desert eco- JOHANSEN, J.R., A. JAVAKUL, AND S.R. RUSHFORTH. 1982. systems. Dowden, Hutchinson, and Ross, Strouds- Effects of burning on the algal communities of a burg, PA. high desert soil near Wallsburg, Utah. Journal of SHERIDAN, R.P., AND R. ROSENTRETER. 1973. The effect of Range Management 35:598–600. hydrogen ion concentrations in simulated rain on LANGE, O.L., G.J. KIDRON, B. BÜDEL, A. MEYER, E. KILIAN, the moss Tortula ruralis (Hedw.) Sm. Bryologist 76: AND A. ABELIOVICH. 1992. Taxonomic composition 168–173. and photosynthetic characteristics of the ‘biological SHOAF, W.T,. AND B.W. LIUM. 1976. Improved extraction of soil crusts’ covering sand dunes in the western chlorophyll a and b from algae with dimethylsulfox- Negev Desert. 6:519–527. ide. Limnology and Oceanography 21:926–928. MATTES-SEARS, U., J.A. GERRARTH, AND D.W. LARSON. 1997. Abundance, biomass, and productivity of endolithic Received 6 October 2000 and epilithic lower plants on the temperate-zone Accepted 10 April 2001 Western North American Naturalist 62(2), © 2002, pp. 227–229

THE EFFECT OF FLOODING ON NORTHERN BOBWHITES

Roger D. Applegate1, Christopher K. Williams2, and R. Scott Lutz2

Key words: Colinus virginianus, flooding, mortality, Northern Bobwhite.

Effects of flooding on mammals (Yeager floodplain in upland areas (termed upland and Anderson 1944) and reptiles (Stickel 1948) study areas, USA). have been documented, and Northern Bob- Bobwhites were captured 1 October–1 whites (Colinus virginianus) have been known December 1998 using bait-traps (Stoddard to drown (Schorger 1946, Mullan and Apple- 1931) and night-lighting (Labisky 1968). Upon gate 1969). However, on landscape and popu- capture, birds were sexed, aged (Rosene lation scales, the implications of mortality 1969), and weighed to the nearest gram. From from extensive flooding are undocumented for each covey we randomly selected 3 birds that birds. We document mortality of bobwhites, were fitted with a necklace-type radio-trans- possibly by drowning, because of flooding in mitter weighing <6 g (Burger et al. 1995). east central Kansas. Birds weighing <150 g were not radio-marked From 31 October through 2 November 1998, to avoid stress from radio-collars weighing 21 cm of rain fell in Lyon County, Kansas >5% of body mass (Samuel and Fuller 1994). (Kansas Precipitation Report ftp//oz.oznet. All other birds captured in the covey were leg- ksu.edu). Water levels in all rivers and streams banded. We immediately released all birds at in the area overflowed their banks for the 3rd the capture location. Radio-tagged individuals time in the decade (1993, 1995, 1998). The were located 5–7 times per week by homing Cottonwood and Neosho rivers and their trib- (White and Garrott 1990) until death, radio utaries flooded a minimum area of 130 km2 of failure, or 31 January. We recorded individual Lyon County. During this period we observed locations as Universal Transverse Mercator the effects of flooding on Northern Bobwhites. (UTM) coordinates with a resolution of 1.00 We conducted research on 12 independent ha (Exum et al. 1982). 259-ha parcels of private and public land. We monitored bobwhite survival during Study areas were separated by at least 1.6 km. the period of flooding (31 October–1 Decem- Distances between study areas were small ber 1998). Survival rates were calculated with enough to eliminate confounding climate and staggered entry additions, and all assumptions habitat heterogeneity effects on the popula- were met (Kaplan and Meier 1958, Pollock et tion and yet large enough to avoid bobwhite al. 1989). We allowed birds to adjust to radio- interchange between study sites. Study areas collars for 7 days before they were included in were located in eastern Lyon County, western survival analysis. We right-censored birds Osage County, and western Coffee County, when fate was unknown or there was radio Kansas. All study areas (1) were composed of failure or loss, emigration from the study area, habitat that was representative of east central or survival beyond 31 January. Additionally, Kansas and (2) allowed no hunting of bob- we estimated cause-specific mortality rates, whites. Additionally, 5 of the 12 study areas classifying mortality agents as flooding or nat- were entirely or partially within the floodplain ural mortality. Both survival and mortality of the Cottonwood and Neosho rivers (termed rates were calculated within the staggered floodplain study areas, FSA). The remaining 7 entry survival model between treatments with study areas were considered to be outside the log-rank tests (P ≤ 0.10; Pollock et al. 1989).

1Department of Wildlife and Parks, Box 1525, Emporia, KS 66801. 2Kansas Department of Wildlife Ecology, University of Wisconsin, Madison, WI 53706.

227 228 WESTERN NORTH AMERICAN NATURALIST [Volume 62

We measured the effect of flooding on whites had not been present in this neighbor- covey daily movement to investigate the pos- hood before the flood and have not been seen sible effect of migration as a means to avoid since 8 November when the covey consisted natural catastrophe stochasticity. We estimated of 10 birds. This covey was likely displaced movement by calculating a mean daily dis- from its usual range by the floodwaters. Yeager tance between radio-locations for the initial 2 and Anderson (1944) found that flooding dis- weeks after the flood. We used an indepen- places certain mammals. dent t test (P ≤ 0.10) to compare movement Assuming we did record the complete between treatments. All means are reported ± extinction of 5 coveys in FSA, we estimate that standard error. 0.39 coveys ⋅ km–2 would have been lost with- Between 31 October and 1 December in the floodplain. Assuming similar losses 1998, 66 individuals within 24 coveys in USA throughout the county impacted by the flood and 43 individuals within 11 coveys in FSA (130 km2), 50 coveys could have been lost in were used to estimate survival and cause-spe- Lyon County alone during the 31 October to cific mortality. After the floodwaters receded, 2 November flood event. In addition, an un- 5 radio-collared birds from 2 coveys were known number of coveys was likely displaced found under flood debris and silt in the FSA. to adjacent upland habitat during the flood. Five radio-collared birds from 3 coveys were Additional birds were likely lost during the never relocated. While we cannot be certain, smaller 30 November event. it is probable that these 5 coveys were deci- Flooding reduced populations of mammals mated in the flood. Additionally, after a 2.7-cm in some studies (Blair 1939, Yeager and rain event on 30 November 1998 (Kansas Pre- Anderson 1944, McCarley 1959, Turner 1966). cipitation Report ftp//oz.oznet.ksu.edu), we Mammal (Yeager and Anderson 1944, Stickel found 2 radio-collared birds, representing 2 1948, Wetzel 1958, Blem and Blem 1975, Ellis different coveys (1 in USA and 1 in FSA), et al. 1997) and reptile (Stickel 1948) numbers dead under flood debris and silt. Even this were not affected in others. The direct impact smaller flash flood event was not without of flooding on bobwhite populations may be as impact on the local bobwhite population. Final significant as severe winter weather (Rose- survival estimate between 31 October and 1 berry 1962, 1964) and could depress local bob- December 1998 on USA was 0.875 ± 0.048 white populations. Additional mortality due to (natural mortality = 0.104 ± 0.044, flood mor- displacement and concentration during floods tality = 0.021 ± 0.021). Final survival estimate is unknown. However, because natural mortal- on FSA was 0.417 ± 0.091 (natural mortality = ity in FSA was very high compared with USA, 0.364 ± 0.091, flood mortality = 0.218 ± 0.247). increased crowding on available dry ground χ2 possibly prompted a density-dependent feed- Survival was significantly lower on FSA ( 1 = 67.61, P < 0.01). Although flooding mortal- back. Regardless, our analysis indicates flood- χ2 ing in this region can provide significant envi- ity was higher on FSA ( 1 = 28.72, P < 0.01), interestingly, natural mortality was also higher ronmental stochasticity, and such variance χ2 should be considered in future predictive pop- on FSA ( 1 = 36.86, P < 0.01). It is possible that displaced coveys, as a result of the flood, ulation models. became more vulnerable to predation. Coveys on FSA that did not go extinct from This paper is a contribution in part of Kansas the flood moved their range to avoid flood- Department of Wildlife and Parks Pittman- waters. We found that coveys in FSA moved Robertson Federal Aid Project W-39-R. Jere more during the 2-week period after the flood Gamble, Terry Kostenek, and the staff of Flint ± ± (FSA: 200.90 29.50, USA: 149.93 12.61, t33 Hills National Wildlife Refuge cooperated in = 1.878, P = 0.69). Additionally, on 3 Novem- many ways. We are grateful to the private ber 1998, a covey of 12 Northern Bobwhites landowners who gave us access to their land was observed in a residential neighborhood of for study. Brian Flock assisted with GIS calcu- Emporia, Kansas, approximately 0.4 km from lation of floodplain area. N. Woffinden and an the nearest known bobwhite habitat in the anonymous reviewer suggested improvements flooded lowlands of the Neosho River. Bob- to the manuscript. 2002] NOTES 229

LITERATURE CITED the staggered entry design. Journal of Wildlife Man- agement 53:7–15. BLAIR, W.F. 1939. Some observed effects of stream-valley ROSEBERRY, J.L. 1962. Avian mortality in southern Illinois flooding on mammalian populations in eastern Okla- resulting from severe weather conditions. Ecology homa. Journal of Mammalogy 20:304–306. 43:739–740. BLEM, L.B., AND C.R. BLEM. 1975. The effect of flooding ______. 1964. Some responses of bobwhites to snow cover on length of residency in the white-footed mouse, in southern Illinois. Journal of Wildlife Management Peromyscus leucopus. American Midland Naturalist 28:244–249. 94:232–236. ROSENE, W. 1969. The bobwhite quail: its life and man- BURGER, L.W., T.V. DAILEY, E.W. KURZEJESKI, AND M.R. agement. Rutgers University Press, New Brunswick, RYAN. 1995. Survival and cause-specific mortality of NJ. Northern Bobwhite in Missouri. Journal of Wildlife SAMUEL, M.D., AND M.R. FULLER. 1994. Wildlife radio- Management 59:401–410. telemetry. In: T.A. Bookhout, editor, Research and ELLIS, L.M., M.C. MOLLES, JR., AND C.S. CRAWFORD. management techniques for wildlife and habitats. 1997. Short-term effects of annual flooding on a pop- The Wildlife Society, Bethesda, MD. ulation of Peromyscus leucopus in a Rio Grande ripar- SCHORGER, A.W. 1946. The quail in early Wisconsin. ian forest of central New Mexico. American Midland Transactions of the Wisconsin Academy of Sciences, Naturalist 138:260–267. Arts, and Letters 24:457–499. EXUM, J.H., R.W. DIMMICK, AND B.L. DEARDEN. 1982. STICKEL, L.F. 1948. Observations on the effect of flood on Land use and bobwhite populations in an agricul- animals. Ecology 29:505–507. tural system in west Tennessee. Proceedings of the STODDARD, H.L. 1931. The bobwhite quail: its habits, pres- National Bobwhite Quail Symposium 2:6–12. ervation, and increase. C. Scribner’s Sons, New York. KAPLAN, E.L., AND P. M EIER. 1958. Nonparametric estima- TURNER, R.W. 1966. Effects of flooding on the mouse Per- tion from incomplete observations. Journal of the omyscus leucopus. Transactions of the Illinois State American Statistical Society 53:457–481. Academy of Science 59:390–391. LABISKY, R.F. 1968. Nightlighting: its use in capturing WETZEL, R.M. 1958. Mammalian succession on midwest- pheasants, prairie chickens, bobwhites and cotton- ern floodplains. Ecology 39:262–271. tails. Illinois Natural History Survey Biological Notes WHITE, G.C., AND R.A. GARROTT. 1990. Analysis of wildlife 62. Urbana, IL. radio-tracking data. Academic Press, San Diego, CA. MCCARLEY, H. 1959. The effect of flooding on a marked YEAGER, L.E., AND H.G. ANDERSON. 1944. Some effects of population of Peromyscus. Journal of Mammalogy flooding and waterfowl concentrations on mammals 40:57–63. of a refuge area in central Illinois. American Mid- MULLAN, J.W., AND R.L. APPLEGATE. 1969. The drowning land Naturalist 31:159–178. of bobwhites in a large reservoir. Wilson Bulletin 81:467. Received 5 July 2000 POLLOCK, K.H., S.R. WINTERSTEIN, C.M. BUNCK, AND P. D . Accepted 19 December 2000 CURTIS. 1989. Survival analysis in telemetry studies: Western North American Naturalist 62(2), © 2002, pp. 230–233

OCCURRENCE OF COROPHIUM SPINICORNE STIMPSON, 1857 (AMPHIPODA: COROPHIIDAE) IN IDAHO, USA

Gary T. Lester1 and William H. Clark2

Key words: Amphipoda, Corophium, , Snake River.

We report a significant range and habitat by freshwater benthic taxonomists throughout extension for the amphipod Corophium spini- North America, treat most major groups of corne Stimpson, 1857, in Idaho, USA. The freshwater invertebrates in North America, species has been reported as an inhabitant of exclusive of insects. One author (GTL) deter- coastal estuaries, bays, lagoons, and adjacent mined that, due to the presence of osmo- river mouths from Alaska to California. Our branchia, the specimens were indeed aquatic; records add the inland state of Idaho to its dis- therefore, the possibility of a marine/brackish tribution and freshwater to its habitat. We species was considered. Specimens were sent report the species from 2 sites on the Snake to Dr. Kevin Li, King County Environmental River, which extend the range 754 km and 795 Lab, Seattle, Washington, for confirmation and km, respectively, upstream from the Pacific were identified as Corophium spinicorne Stimp- Ocean. son. Corophium spinicorne is known from Alaska The field methods used to collect these to Waddell Creek, Santa Cruz County, Califor- specimens are described in detail in Beneficial nia (Barnard 1954). This species has been Use Reconnaissance Project Technical Advi- found living in estuaries as a tube builder sory Committee (1998). In general, 3 macro- attached to debris on mud bottoms (Smith and invertebrate samples were collected at each Carlton 1975) and from American Pacific coastal site using either a Slack sampler or petite Ponar streams and pools (Eriksen 1968). Some water dredge, depending upon substrate conditions. quality and environmental data for Corophium Samples were collected at 3 riffle/run tran- sp. have been reported (Eriksen 1968, Fors- sects and were preserved in 70% ethanol in berg et al. 1977, Grosse et al. 1986). Several the field. EcoAnalysts, Inc. laboratory techni- reports of Corophium spinicorne as prey of fish cians combined the 3 samples in a 500-micron along coastal habitats in California, Oregon, sieve and spread the sample into a gridded and Washington give additional information Caton® subsampling tray. Technicians ran- concerning the distribution and habitats of domly selected each grid, removed all organ- this species (Forsberg et al. 1977, McCabe et isms, and placed the organisms in a properly al. 1993, Busby and Barnhart 1995, Swenson labeled vial containing 70% ethanol preserva- and McCray 1996). Although Corophium spini- tive. Once the target count of 500 organisms corne occurs in tidal freshwater, it has previ- was reached, the technicians sorted the last ously not been considered fully adapted to square to completion. Taxonomists then iden- freshwater (Hutchinson 1967). tified all organisms in the sample to the lowest While identifying contract macroinverte- practical level (usually genus or species). brate samples from Idaho waters, we encoun- Specimens of Corophium spinicorne were tered a crustacean that was unidentifiable encountered at 2 sampling stations on the using freshwater keys, including Thorp and Snake River. Both stations are upstream of Covich (1991) and Pennak (1989). These 2 tax- Lewiston, Idaho, a port located at river kilo- onomic references, which are frequently used meter 230, 752 km upstream from the Pacific

1EcoAnalysts, Inc., 105 East 2nd Street, Suite #1, Moscow, ID 83843. 2Department of Environmental Quality, 1410 North Hilton Street, Boise, ID 83706, and Orma J. Smith Museum of Natural History, Albertson College of Idaho, Caldwell, ID 83605-4432.

230 2002] NOTES 231

TABLE 1. Water quality and habitat variables for Corophium spinicorne Stimpson in the lower Snake River, Idaho, 2–3 September 1998. Variable Site 003 (Asotin) Site 004 (Grande Rond) Temperature (°C) 23.7 22.8 Dissolved oxygen (mg ⋅ L–1) 7.7 7.5 Conductivity (µS ⋅ cm–1) 320 333 pH (SU) 9.1 9.1 Habitat type run/riffle run/riffle Substrate embeddedness (%) 33 67

Ocean. Barge traffic from the ports of Lewis- were upstream of any barge traffic, it is likely ton, Idaho, and Clarkston, Washington, travel they persisted over time and populations have down the lower Snake River to the Columbia expanded upstream from Lewiston. Corophium River, where they deliver cargo to port cities spinicorne is a permanent resident of fresh- on the west coast of Oregon and Washington. water in North America and should be incorpo- Primary cargoes are timber and grain. The rated into future taxonomic treatments of fresh- collection localities are as follows: water macroinvertebrates.

(1) IDAHO, Nez Perce Co., Snake River (river We thank Luana McCauley and the Idaho kilometer 273.6). 6 specimens. 3.2 km above confluence w/Grande Rond River. T31N State Library for assistance in locating some of R05W S03 SW NE elev. 268 m. 3 Sept. the literature cited in this paper. The speci- 1998, K. Davis. 1998 BURP# RLEWP004. mens were collected by K. Davis. Dr. Kevin Li (2) IDAHO, Nez Perce Co., Snake River near verified the identification of Corophium spini- 10-Mile Creek (Asotin, Washington, river corne, and Russell C. Biggam (University of kilometer 233.4). 1 specimen. T34N R05W Idaho) assisted with unpublished distribution S08 NW NE SW elev. 229 m. 2 Sept. 1998, records. John Pfeiffer (EcoAnalysts, Inc.) ini- K. Davis. 1998 BURP# RLEWP003. tially brought the specimens to the primary author’s attention. Physical habitat and water quality parameters for these sites are presented in Table 1. A LITERATURE CITED complete list of macroinvertebrate taxa pres- ent at these sites is presented in Table 2. BARNARD, J.L. 1954. Marine Amphipoda of Oregon. Oregon Voucher specimens of Corophium spinicorne State Monographs, Studies in Zoology, Number 8. are deposited in the Orma J. Smith Museum Corvallis. 103 pp. of Natural History, Albertson College of Idaho, BENEFICIAL USE RECONNAISSANCE PROJECT TECHNICAL ADVISORY COMMITTEE. 1998. Beneficial use recon- Caldwell, Idaho; Monte L. Bean Life Science naissance project work plan—1998, river work plan. Museum at Brigham Young University, Provo, Idaho Division of Environmental Quality, Boise. 45 Utah; and EcoAnalysts, Inc. macroinvertebrate pp. laboratory in Moscow, Idaho. BUSBY, M.S., AND R.A. BARNHART. 1995. Potential food Although we have not examined specimens sources and feeding ecology of juvenile fall chinook from any other studies on the lower Snake salmon in California’s Mattole River lagoon. Califor- nia Fish and Game 81(4):133–146. River, Corophium spinicorne was reportedly ERIKSEN, C.H. 1968. Aspects of the limno-ecology of encountered throughout the lower Snake River Corophium spinicorne Stimpson (Amphipoda) and in Washington state (Russ Biggam personal Gnorimosphaeroma oregonensis (Dana) (Isopoda). commuication). We suspect that C. spinicorne Crustaceana 14:1–12. is well established throughout the lower Snake FORSBERG, B.O., J.A. JOHNSON, AND S.M. KLUG. 1977. Iden- tification, distribution, and notes on food habits of fish and Columbia rivers, both of which are fresh- and shellfish in Tillamook Bay, Oregon. Job Comple- water (nonbrackish) rivers. It is possible that tion Report, Oregon Department of Fish and Wild- damming of these 2 rivers has created favor- life, Tillamook. 117 pp. able habitat for C. spinicorne to colonize. It is GROSSE, D.J., G.B. PAULEY, AND D. MORAN. 1986. Species also possible that barges coming upstream profiles: life histories and environmental require- ments of coastal fishes and invertebrates (Pacific have carried specimens in ballast water and Northwest): Amphipods. Biological Report 82 introduced them as far upriver as Lewiston, (11.69), U.S. Fish and Wildlife Service, Washington, Idaho. Since the specimens collected in Idaho DC. 15 pp. 232 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 2. Invertebrate associates of Corophium spinicorne Stimpson in the lower Snake River, Idaho, 2–3 September 1998. Associated taxa Site 003 (Asotin) Site 004 (Grande Rond)

TRICLADIDA Dugesia tigrina XX NEMATODA XX OLIGOCHAETA Spirosperma ferox X Tubificidae X Limnodrilus hoffmeisteri X CRUSTACEA Corophium spinicorne XX INSECTA Ephemeroptera Acentrella insignificans XX Baetis tricaudatus XX Heptagenia sp. X Stenonema terminatum XX Tricorythodes sp. X Odonata Ophiogomphus sp.X Coleoptera Optioservus sp. X Zaitzevia sp. X X Trichoptera Brachycentrus occidentalis X Cheumatopsyche sp. X X Culoptila sp. X Glossomatidae X Hydropsyche sp. X X Hydrophychidae X Leucotrichia sp. X X Lepidoptera Petrophila sp. X X Diptera X Empididae X Chironomidae Cardiocladius sp. X X Chironomus sp. X Cladotanytarsus sp. X X Cricotopus (Isocladius) Type I X X Cricotopus sp. X X Dicrotendipes sp. X Eukiefferiella devonica group X Microtendipes pedellus group X X Nanocladius sp. X Orthocladius rivicola group X Orthocladius complex X X Paracladopelma sp. X Polypedilum sp. X Procladius sp. X Rheotanytarsus sp. X Synorthocladius sp. X Tanytarsus sp. X X MOLLUSCA Bivalvia Corbicula fluminea X Gastropoda Fisherola nuttali X Fluminicola sp. X

TOTAL TAXA 36 26 2002] NOTES 233

HUTCHINSON, G.E. 1987. A treatise on limnology. Volume SWENSON, R.O., AND A.T. MCCRAY. 1996. Feeding ecology II. Introduction to lake biology and the limnoplank- of the tidewater goby. Transactions of the American ton. John Wiley and Sons, Inc., New York. Fisheries Society 125:956–970. MCCABE, G.T., JR., R.L. EMMETT, AND S.A. HINTON. 1993. THORPE, J.H., AND A.P. COVICH, EDITORS. 1991. Ecology and Feeding ecology of juvenile white sturgeon (Acipenser classification of North American freshwater inverte- transmontanus) in the lower Columbia River. North- brates. Academic Press, New York. 911 pp. west Science 67(3)170–180. PENNAK, R.W. 1989. Freshwater invertebrates of the United Received 2 July 2000 States, Protozoa to Mollusca. 3rd edition. John Wiley Accepted 27 October 2000 and Sons, New York. 628 pp. Western North American Naturalist 62(2), © 2002, pp. 234–239

HERPETOFAUNA OF THE SOUTHERN SNAKE RANGE OF NEVADA AND SURROUNDING VALLEYS

Kirk Setser1, Jesse M. Meik1, and Daniel G. Mulcahy1,2

Key words: Snake Range, Great Basin, herpetofauna, distributions, habitat associations.

In comparison with other North American flanks the western face of the Snake Range deserts, the Great Basin has a depauperate and is higher in elevation than Snake Valley herpetofauna consisting primarily of wide- (~1750 m compared to ~1525 m at low points ranging, habitat-generalist species (Stebbins on the respective valley floors). Spring Valley 1985). Great Basin amphibians and reptiles lies outside the Bonneville Basin but was peri- are often placed into broad groups based on odically inundated during the Pleistocene by overall geographic distribution of the species, smaller, isolated pluvial lakes (Mifflin and such as northern versus southern (Macey and Wheat 1979, Hovingh 1997). Papenfuss 1991) or eastern versus western Survey crews recorded over 400 observa- (Hovingh 1997). Many species occur only in tions of reptiles and amphibians during the the periphery of the Great Basin Desert, where course of this study. Most observations were they are restricted to pockets of suitable habi- recorded during visual surveys in appropriate tat (Tanner 1978). Although general biogeo- habitat. Several additional observations resulted graphic patterns can be seen, the subtle distri- from driving roads at night and from inspect- butional nuances at local levels have yet to be ing beneath natural cover objects. We also elucidated for many species in the Great Basin attempted to listen for chorusing amphibians (but see Hovingh 1997, Zamudio et al. 1997, in wetland habitats; however, when using this Bos and Sites 2001). Distributions remain technique, we did not detect any animals. At poorly understood due to the relatively large least one specimen of each species captured area encompassed and the scarcity of inten- was taken as a voucher. These specimens are sive inventory studies (e.g., Vindum and Arnold deposited at the California Academy of Sci- 1997). Here we discuss the results of a her- ences, San Francisco. Crews searched lower petological survey of the southern Snake and middle elevations (below 2500 m) of 7 Range and surrounding valleys conducted 20– montane drainages of the southern Snake Range, 25 May 2000. concentrating efforts in the riparian zones and The study area is situated in and around adjacent pinyon-juniper woodland of peren- the southern Snake Range, including Great nial streams draining the eastern slope of the Basin National Park, in White Pine County, southern Snake Range. We also surveyed vari- Nevada, and adjacent Millard County, Utah ous lower-elevation vegetative communities in (Fig. 1). Physiography of the Snake Range is Snake and Spring valleys. typical of the Basin and Range Province, being We divided surveyed areas into 3 valley characterized by abrupt north–south trending and 2 montane habitat categories: low desert mountains and their associated interconnect- scrub, sagebrush shrubland, wetland, pinyon- ing valleys. The Snake Range is among the juniper woodland, and montane riparian, re- highest in the Great Basin with peaks exceed- spectively. Low desert scrub consists of com- ing 3000 m in elevation. Snake Valley lies to munities dominated by Artemisia arbuscula, the east of the Snake Range and was formerly Atriplex sp., Sarcobatus vermiculatus, and inundated by ancient Lake Bonneville (Mifflin Distichlis spicata. This habitat is prevalent at and Wheat 1979, Hovingh 1997). Spring Valley low elevations in Snake Valley. We recorded

1Department of Biology, Utah State University, Logan, UT 84322-5305. 2Corresponding author.

234 2002] NOTES 235

Fig. 1. Map of the southern Snake Range and surrounding valleys. Dots represent reptile and amphibian observation points, thus indicating areas surveyed. Thin lines represent major drainages, thick hatched lines represent roads, and the solid line is the Nevada-Utah state border.

Cnemidophorus tigris, Uta stansburiana, (Cowardin et al. 1978). Wetland complexes in Phrynosoma platyrhinos, Crotaphytus bicinc- Spring and Snake valleys include permanently tores, Pituophis catenifer, Masticophis taenia- flooded, intermittently exposed, semiperma- tus, Hypsiglena torquata, Crotalus viridis, and nently flooded, seasonally flooded, saturated, Spea intermontana from this habitat (Table 1). and temporarily flooded water regimes. These Sagebrush shrubland occurs at higher ele- palustrine wetlands also include open water vations in Snake Valley and is prevalent on areas with a mixture of unconsolidated bottom both the floor and lower foothills of Spring and aquatic bed class wetlands. Open water Valley. Artemisia tridentata typically domi- habitats are generally small and shallow, and nates in this habitat, but other taxa including they comprise a small portion of the total wet- A. arbuscula, Atriplex sp., Sarcobatus vermicu- land area. Juncus sp., Scirpus sp., Carex sp., latus, Ephedra sp., and various grasses are and various grasses characterize these areas. locally abundant. We observed Cnemidopho- We encountered 2 native species in low-eleva- rus tigris, Uta stansburiana, Sceloporus gracio- tion wetlands: Thamnophis elegans and Rana sus, S. occidentalis, Phrynosoma platyrhinos, P. pipiens. A single introduced species, R. cates- hernandesi, Crotaphytus bicinctores, Gambelia beiana, was also seen in wetland habitat (Table wislizenii, Pituophis catenifer, Thamnophis ele- 1). Both species of ranid frogs were seen only gans, Crotalus viridis, and Spea intermontana in low-elevation wetlands. in sagebrush shrubland (Table 1). On the upper foothills and in the canyons Several mixosaline wetlands occur in Spring of the Snake Range, woodlands of Pinus mono- and Snake valleys. These habitats chiefly con- phylla and Juniperus osteosperma replace sist of persistent palustrine emergent wetlands shrubland communities. Within most eastern 236 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 1. Number of observations of reptile and amphibian taxa by habitat in the southern Snake Range and surround- ing valleys. Low desert Sagebrush Pinyon-juniper Montane scrub shrubland wetland woodland riparian Total Person hours searched 32.5 33.5 10.25 51.5 51 178.75

TEIIDAE Cnemidophorus tigris 11 5 16 PHRYNOSOMATIDAE Uta stansburiana 61 21 21 103 Sceloporus graciosus 41 76 2 119 Sceloporus occidentalis 828137 Phrynosoma platyrhinos 16 2 18 Phrynosoma hernandesi 11 CROTAPHYTIDAE Crotaphytus bicinctores 11 2 Gambelia wislizenii 11 SCINCIDAE Eumeces skiltonianus 718 COLUBRIDAE Pituophis catenifer 916 3230 Masticophis taeniatus 2114 Hypsiglena torquata 112 Thamnophis elegans 14 43140 VIPERIDAE Crotalus viridis 24 2311 PELOBATIDAE Spea intermontana 32 5 RANIDAE Rana catesbeianaa 11 Rana pipiens 88 ALL TAXA 106 104 13 142 41 406 aIntroduced species slope drainages of the southern Snake Range, observed Sceloporus graciosus, S. occidentalis, these pinyon-juniper woodlands are relatively Eumeces skiltonianus, Pituophis catenifer, Mas- dense, dominated by P. monophylla, and con- ticophis taeniatus, Thamnophis elegans, and fluent with riparian vegetation along water- Crotalus viridis in montane riparian areas courses. We recorded the following species (Table 1); however, only T. elegans appeared from these pinyon-juniper woodlands: Scelo- concentrated in this habitat. This contrasts porus graciosus, S. occidentalis, Uta stansburi- with the observed trend among other snake ana, Eumeces skiltonianus, Pituophis catenifer, species, which were relatively unrestricted by Masticophis taeniatus, Hypsiglena torquata, habitat. No amphibian species were detected Thamnophis elegans, and Crotalus viridis (Table in montane riparian habitats. 1). Although we recorded U. stansburiana from Species richness is greatest in sagebrush pinyon-juniper woodland habitat, all observa- shrubland, followed by low desert scrub and tions occurred in the open, juniper-dominated pinyon-juniper woodland. Sagebrush shrub- woodland of a single montane drainage. land represents an interface between higher- Several perennial streams with well-devel- elevation montane woodland and low-eleva- oped riparian zones dissect the eastern slopes tion desert communities; most species en- of the southern Snake Range. Riparian plant countered in the latter habitats were also communities differ substantially between found in sagebrush shrubland. Xeric-adapted drainages. However, most drainages are domi- lizard taxa typical of more southern North nated by various combinations of Populus sp., American deserts, such as Cnemidophorus tigris, Salix sp., Betula occidentalis, and Prunus vir- Uta stansburiana, and Phrynosoma platyrhi- giniana. Abies concolor is also prominent in nos, were encountered more frequently in low higher-elevation riparian communities. We desert scrub than in any other habitat. Fewer 2002] NOTES 237 species were clustered in montane habitats. Snake Valley. Unsubstantiated reports of a Eumeces skiltonianus was observed only at Phrynosoma sp. exist from pinyon-juniper higher elevations, and we encountered Scelo- woodland on the eastern slope of the southern porus occidentalis primarily in montane habi- Snake Range (B. Hamilton personal communi- tats with only isolated observations from rock cation). In addition, fossil records of Phryno- outcrops at lower elevations in sagebrush soma douglasi (likely P. hernandesi, sensu shrubland. Zamudio et al. 1997) have been found on the Sceloporus graciosus is sympatric with S. eastern slope of the southern Snake Range occidentalis in montane habitats but also and from northern Snake Valley (Mead et al. occurs extensively in sagebrush shrubland. In 1989). These records suggest P. hernandesi areas of sympatry, S. occidentalis was encoun- likely occurs on the eastern slope of the south- tered only near rock outcrops and in areas ern Snake Range and may also inhabit higher with extensive rocky terrain, while S. gracio- elevations of Snake Valley. Pianka and Parker sus occupied more varied microhabitats. We (1975) noted that P. hernandesi and P. platyrhi- frequently observed Uta stansburiana in open, nos exhibit a complex distributional pattern at juniper-dominated woodland habitat within a another site in the eastern Great Basin. Inter- single montane drainage; however, this species actions between these congeners may also was not encountered in the relatively dense, influence their distributions and habitat affini- pinyon-dominated woodland of other montane ties in the valleys surrounding the southern drainages surveyed. Sceloporus graciosus is Snake Range. sympatric with U. stansburiana within the Differences between Spring and Snake Val- montane drainage occupied by U. stansburi- ley lizard assemblages cannot be attributed ana, but it is apparently absent from the floor solely to disparity in elevation. All species of Snake Valley, where the latter species was found in Snake Valley were observed at eleva- frequently encountered. Morrison and Hall tions comparable to, or higher than, the floor (1999) also observed that U. stansburiana is of Spring Valley. This is particularly evident associated with open, juniper-dominated areas with Uta stansburiana, which occurs in the within pinyon-juniper woodland at sites in the montane zone of at least one eastern slope White-Inyo Range of eastern California. drainage of the southern Snake Range. There We observed notable differences between are obvious differences in the vegetative com- the lizard assemblages of Snake and Spring munities of Snake and Spring valleys. The valleys (Table 2), despite their being separated floor of Snake Valley is largely covered by by <30 km. We frequently encountered Uta open, low desert scrub dominated by Artemisia stansburiana, Phrynosoma platyrhinos, and arbuscula and sparsely vegetated halophytic Cnemidophorus tigris on the floor of Snake plant communities. Spring Valley lacks the Valley. We also observed these species at extensive open desert vegetation characteris- higher elevations in sagebrush shrubland of tic of Snake Valley. Instead, the floor of Spring Snake Valley, albeit less frequently. Although Valley includes large expanses of sagebrush few individuals were located, we observed shrubland and dense, comparatively mesic Crotaphytus bicinctores and Gambelia wis- halophytic plant communities. The availability lizenii only in Snake Valley. In contrast to the of dense shrub cover afforded by these com- comparatively speciose lizard community ob- munities may explain the presence of Scelo- served in Snake Valley, only 2 lizard species porus graciosus and Phrynosoma hernandesi at were found in Spring Valley. We observed a low elevations of Spring Valley. single P. hernandesi from the floor of the valley Other dissimilarities exist between these and commonly observed S. graciosus at most valleys. Hotter and dryer than Spring Valley, sites surveyed in Spring Valley. Many taxa Snake Valley is not bordered by a discreet, seen in big sagebrush–dominated areas of high-elevation mountain range to the east. Snake Valley were not found in similar habitat Snake Valley contains rocky knolls with con- of Spring Valley (Table 2). siderable topographical relief and is confluent In contrast, Phrynosoma hernandesi was with the remainder of the Bonneville Basin. located only in sagebrush shrubland on the Spring Valley is a comparatively narrow valley floor of the Spring Valley and was not ob- bordered on both sides by high-elevation served in sagebrush shrubland habitats in ranges with a topographically uniform basin 238 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 2. Presence of reptile and amphibian taxa in the southern Snake Range and surrounding valleys based on sur- vey data. Snake Valley Spring Valley Snake Rangeb Person hours searched 58.75 17.5 102.5

TEIIDAE Cnemidophorus tigris X PHRYNOSOMATIDAE Uta stansburiana X1 Sceloporus graciosus XX6 Sceloporus occidentalis X6 Phrynosoma platyrhinos X Phrynosoma hernandesi X CROTAPHYTIDAE Crotaphytus bicinctores X Gambelia wislizenii X SCINCIDAE Eumeces skiltonianus 3 COLUBRIDAE Pituophis catenifer XX3 Masticophis taeniatus X1 Hypsiglena torquata X1 Thamnophis elegans XX4 VIPERIDAE Crotalus viridis XX2 PELOBATIDAE Spea intermontana XX RANIDAE Rana catesbeianaa X Rana pipiens X aIntroduced species bValues represent the number of montane drainages (of 7 surveyed) from which each taxon was recorded.

floor, as is typical of valleys in the interior We thank N. Darby, B. Hamilton, J. Sirotnak, Great Basin. Some interaction of elevation, and staff of Great Basin National Park for vegetation, temperature, and moisture likely invaluable expertise and assistance. We thank accounts for the apparent absence of xeric- T. Dahl, T. Grant, S. Newbold, C. Steele, C. adapted lizard species, such as Cnemidopho- Ustach, P. Ustach, N. Welch, A. Young, and K. rus tigris, Uta stansburiana, and Phrynosoma Young for field assistance. We also thank C. platyrhinos, in Spring Valley and their pres- Garrard for map assistance. The Southern ence in Snake Valley. Nevada Chapter of Trout Unlimited provided Many xeric-adapted taxa are widespread in funding. We thank the Brodie/Mendelson lab the lower-elevation Bonneville and Lahontan group and 2 anonymous reviewers for com- basins of the eastern and western Great Basin ments on earlier versions of this manuscript. but are restricted in (or absent from) the inte- rior of the Great Basin (Banta 1962, Stebbins LITERATURE CITED 1985). Although data on herpetofaunal distri- butions in the Snake Range region remain BANTA, B.H. 1962. Preliminary remarks upon the zoo- limited, we believe that the observed differ- geography of the lizards inhabiting the Great Basin ence between lizard taxa of Snake and Spring of the western United States. Wasmann Journal of valleys reflects a real difference in lizard com- Biology 20:253–285. munity composition. The Snake Range may BOS, D.H., AND J.W. SITES, JR. 2001. Phylogeography and mark the eastern edge of the high-elevation, conservation genetics of the Columbia spotted frog [Rana luteiventris; Amphibia, Ranidae]. Molecular interior Great Basin that lacks components of Ecology 10:1499–1513. the xeric-adapted herpetofauna typical of COWARDIN, L.M., V. CARTER, F.C. GOLET, AND E.T. LAROE. southern North American deserts. 1979. Classification of wetlands and deepwater habi- 2002] NOTES 239

tats of the United States. U.S. Fish and Wildlife Ser- STEBBINS, R.C. 1985. Western reptiles and amphibians. vice Report. FWS/OBS/-79/31. Washington, DC. 3rd edition. Houghton Mifflin Company, New York. HOVINGH, P. 1997. Amphibians of the eastern Great Basin 336 pp. (Nevada and Utah, USA): a geographical study with TANNER, W.W. 1978. Zoogeography of reptiles and amphib- paleozoological models and conservation implications. ians in the Intermountain Region. In: K.T. Harper Herpetological Natural History 5:97–134. and J.L. Reveal, editors, Intermountain biogeogra- MACEY, J.R., AND T.J. P APENFUSS. 1991. Amphibian and phy: a symposium. Great Basin Naturalist Memoirs reptile biogeography. Pages 293–303 in C.A. Hall, 2:43–54. Jr., editor, Natural history of the White-Inyo Range, VINDUM, J.V., AND E.N. ARNOLD. 1997. The northern alli- eastern California. University of California Press, gator lizard (Elgaria coerulea) from Nevada. Her- Berkeley. petological Review 28:100. MEAD, J.I., T.H. HEATON, AND E.M. MEAD. 1989. Late ZAMUDIO, K.R., K.B. JONES, AND R.H. WARD. 1997. Molec- Quaternary reptiles from two caves in the east-cen- ular systematics of short-horned lizards: biogeogra- tral Great Basin. Journal of Herpetology 23:186–189. phy and taxonomy of a widespread species complex. MIFFLIN, M.D., AND M.M. WHEAT. 1979. Pluvial lakes Systematic Biology 46:284–305. and estimated pluvial climates of Nevada. Nevada Bureau of Mines and Geology Bulletin 94:1–57. Received 18 January 2001 MORRISON, M.L., AND L.S. HALL. 1999. Habitat character- Accepted 20 December 2001 istics of reptiles in pinyon-juniper woodland. Great Basin Naturalist 59:288–291. PIANKA, E.R., AND W. S. P ARKER. 1975. Ecology of horned lizards: a review with special reference to Phryno- soma platyrhinos. Copeia 1:141–162. Western North American Naturalist 62(2), © 2002, pp. 240–242

MOVEMENTS OF A BULLSNAKE (PITUOPHIS CATENIFER) FOLLOWING PREDATION OF A RADIO-COLLARED NORTHERN POCKET GOPHER (THOMOMYS TALPOIDES)

Ray T. Sterner1, Brett E. Petersen, Stephen A. Shumake, Stanley E. Gaddis, Jean B. Bourassa, Todd A. Felix, Geraldine R. McCann, Kenneth A. Crane, and Abbe D. Ames

Key words: pocket gopher, open-hole index, bullsnake, predation, radio-tracking.

Life history and foraging data for the bull- of a bullsnake that we believe ingested the snake (Pituophis catenifer) include home range gopher. These serendipitous observations of 1–2 ha, density of <1 ⋅ ha–1, unimodal activ- were obtained during a field study to evaluate ity peak between April and September in most the use of chemical irritants in soil to expel temperate climes, and estimated longevity of gophers from burrows. ∼16 years (Gibbons and Semlitsch 1987, Parker The site is located in a 64.75-ha field and Plummer 1987, Hammerson 1999). Exten- (40°45′20″N and 105°02′30″W) near Welling- sive food habits data for the bullsnake show ton, Colorado; at the time of the study the that diverse species of rodents and lizards field had a 5+-year-old stand of alfalfa (Med- comprise the bulk of the diet (Hisaw and icago sativa L) and mixed grasses with center- Gloyd 1926, Brown and Parker 1982, pivot irrigation. The soil is characterized as a Reynolds and Scott 1982, Mushinsky 1987, sandy loam texture (62% sand, 26% silt, and Hammerson 1999), but observations of actual 12% clay), with 7.9 pH and 2.5% organic mat- predation events and of subsequent movements ter (Agvise® Laboratories, Northwood, ND). of these predators in the field are rare. Numerous mounds (∼1 mound or plug per 5–8 Carpenter (1982) reported on the stereo- m2) of northern pocket gophers were present typed, soil-excavation behaviors of the bull- in the field. snake (P. m. sayi), a behavior initially described The gopher (, 146 g) was live-trapped by Hisaw and Gloyd (1926). These authors (CO License 99-TR621A2) on 18 August 1999. inferred that bullsnakes dig to open and Within 6 hours of capture (trap last checked at search burrows for pocket gophers and to pre- ∼0900 hours MDT, capture detected at 1320 pare cavities for nesting. In laboratory trials hours MDT, and release at capture site accom- (Carpenter 1982), 7 of 8 bullsnakes were ob- plished at 1440 hours MDT), the gopher was served using their snouts to penetrate soil and anesthetized using a 2- to 3-minute inhalation head-neck flexures to move scooped soil pos- exposure to Metofane® (Mallinckrodt Veteri- terior from the dig sites. These trials revealed nary, Inc., Mundelein, IL) in a large glass jar. that an excavating bullsnake can move up to A 2.45-g radio-transmitter (Holohil Systems, 3400 cm3 of soil per hour. Moreover, soil ob- Ontario, Canada) containing a 1.5 × 1.0 × 0.7- tained from actual pocket gopher mounds and cm battery pack (40–50 day projected life) placed in test cages induced excavation behav- with a wire whip antenna (∼10 cm) was affixed iors in 11 snakes during 60% of trials; this snugly around the gopher’s neck using shrink- compared to bullsnake digging activity during tubing–covered wire, leader sleeves, and Duro only 7.5% of trials in control soil. Super Glue® (Loctite Corp., Newington, CT). We report here on the likely predation of a We visually located the ground-surface radio-collared northern pocket gopher (Tho- location nearest the transmitter using a hand- momys talpoides) and subsequent movements held, 3-element Yagi antenna and a portable

1Address of all authors: U.S. Department of Agriculture, Animal and Plant Health Inspection Service, National Wildlife Research Center, 4101 LaPorte Avenue, Fort Collins, CO 80521-2154.

240 2002] NOTES 241

Fig. 1. Schematic showing GPS transmitter locations between 18 and 26 August 1999. radio receiver (Advanced Telemetry Systems, pocket gopher movements (≤20 m from the Inc., Isanti, MN). Typically, we obtained these point of capture for the gopher). However, late locations during 4 daily tracking sessions: 0801– on 21 August 1999, the transmitter could not 1000, 1201–1400, 1601–1800, and 1901–2000 be detected during routine tracking. A more hours MDT. Geographic locations were marked extensive search of the area the next morning and determined using a Global Positioning yielded detection ∼63 m northeast of the origi- System (GPS) receiver (GeoExplorer®, Trim- nal gopher capture/release site. The transmit- ble Navigation, Sunnyvale, CA). Data were ter remained within 2 m of this location dur- differentially corrected via Pathfinder® soft- ing the subsequent 2 days (8 locations), before ware (Pathfinder®, Trimble Navigation, Sunny- returning to near the last 21 August location vale, CA) and imported into ArcView GIS soft- (∼16 m north of the capture/release site) for ware (ESRI, Redlands, CA). To estimate accu- <8 hours. Subsequently, the transmitter was racy, we always collected GPS data for a pre- located ∼60 m northwest of the capture site determined reference location prior to each and remained there for about 12 hours (25 series of transmitter locations; estimates of the August). Finally, on the morning of 26 August reported ground-surface locations were within (∼1000 hours MDT), the transmitter signal 2 m of true. was detected moving in a westerly direction Between 18 and 26 August 1999, the trans- on or adjacent to a field road along the north- mitter was located 29 times (Fig. 1). The first ern edge of the alfalfa field. The source of the 14 locations (3.5 days) were typical of northern signal was sighted—a 90- to 110-cm-long bull- 242 WESTERN NORTH AMERICAN NATURALIST [Volume 62 snake. The snake then entered a livestock pas- provided helpful reviews of the draft manu- ture to the north and later disappeared into a script. small rodent hole ∼20 m north of the fence line or 145 m west-northwest of the capture/ LITERATURE CITED release site (see Fig. 1). This was the final detection of the transmitter’s signal. BROWN, W.S., AND W. S. P ARKER. 1982. Niche dimensions and resource partitioning in a Great Basin Desert Based on our observations, it is apparent snake community. Pages 59–81 in N.J. Scott, Jr., edi- that the gopher and transmitter were ingested tor, Herpetological communities. Wildlife Research by the bullsnake on, or about, the afternoon of Report 13, U.S. Fish Wildlife Service, Washington, 21 August. We have rarely, if ever, recorded DC. movements of a northern pocket gopher >45 CARPENTER, C.C. 1982. The bullsnake as an excavator. Journal of Herpetology 16:394–401. m from the point of capture/release. Thirteen GIBBONS, J.W., AND R.D. SEMLITSCH. 1987. Activity pat- other radio-collared northern pocket gophers terns. Pages 396–421 in R.A. Siegel, J.T. Collins, and tracked in this same vicinity had a maximum S.S. Novak, editors, Snakes: ecology and evolution- movement distance (from the point of cap- ary biology. McGraw-Hill, New York. HAMMERSON, G.A. 1999. Pages 336–344 in Amphibians ture/release) of 45 m. Additionally, the subse- and reptiles in Colorado. University Press of Col- quent lack of movement for ∼72 hours is con- orado, Niwot. sistent with hypothesized restive/digestive HISAW, F.L., AND H.K. GLOYD. 1926. The bull snake as a activities of the bullsnake post-predation; but, natural enemy of injurious rodents. Journal of Mam- counter to Hammerson (1999), this snake did malogy 7:200–205. MUSHINSKY, H.R. 1987. Foraging ecology. Pages 302–334 not “sun-the-stomach” to aid digestion. We in R.A. Siegel, J.T. Collins, and S.S. Novak, editors, believe that the snake remained stationary and Snakes: ecology and evolutionary biology. McGraw- belowground from at least 21 to 24 August. Hill, New York. Considering that an adult bullsnake is active PARKER, W.S., AND M.V. PLUMMER. 1987. Population ecol- ogy. Pages 253–301 in R.A. Siegel, J.T. Collins, and for about 5.5 months per year in Colorado, S.S. Novak, editors, Snakes: ecology and evolution- weighs ∼1.0–1.5 kg, and ingests ∼1.0–2.0 kg of ary biology. McGraw-Hill, New York. prey per year (see Hisaw and Gloyd 1926, REYNOLDS, R.P., AND N.J. SCOTT, JR. 1982. Use of a mam- Mushinsky 1987), this predation event was malian resource by a Chihuahuan snake community. likely 1 of <20 feedings for the snake during Pages 99–119 in N.J. Scott, Jr., editor, Herpetological communities. Wildlife Research Report 13, U.S. Fish 1999 (i.e., assuming 80–150 g rodents/lizards/ Wildlife Service, Washington, DC. birds). Whether it survived the ingestion of the hardware is unknown. Received 14 August 2000 Accepted 13 March 2001 We thank James Anderson for his coopera- tion. Thomas Mathies and Kathy Fagerstone Western North American Naturalist 62(2), © 2002, pp. 243–245

GASTROINTESTINAL HELMINTHS OF THE BLACKNECK GARTER SNAKE, THAMNOPHIS CYRTOPSIS (COLUBRIDAE)

Stephen R. Goldberg1 and Charles R. Bursey2

Key words: Thamnophis cyrtopsis, Colubridae, helminths, Arizona.

The blackneck garter snake, Thamnophis mounts. Selected trematodes were rehydrated cyrtopsis, occurs from southeastern Utah in water, regressively stained in hematoxylin, through central New Mexico, central and dehydrated in ethanol, cleared in xylene, and southern Arizona, and central Texas to Guate- mounted in Canada balsam on glass slides. mala (Rossman et al. 1996) at elevations from Two species of trematodes, Ochetosoma sea level to around 2700 m (Stebbins 1985). ellipticum (Pratt, 1903) and Alaria sp. (meso- Three subspecies are recognized, T. c. cyrtop- cercariae only), and 3 species of nematodes, sis, T. c. collaris, and T. c. ocellatus; however, Ophidascaris labiatopapillosa Walton, 1927, intergrades between T. c. cyrtopsis and T. c. Abbreviata terrapenis (Hill, 1941), and Physa- ocellatus have been reported (Rossman et al. loptera sp. (larvae only), were found. Infection 1996). Our sample consisted of the subspecies sites, total number of helminths, prevalence T. c. cyrtopsis. There are, to our knowledge, (number of infected snakes divided by sample no reports of helminths from this species. The examined), mean intensity ± 1 s (number of purpose of this paper is to report host records helminths divided by number of infected for helminth parasites of T. c. cyrtopsis for the snakes), and range are given in Table 1. Voucher first time. helminths were deposited in the United States Twenty-three T. c. cyrtopsis (mean snout- National Parasite Collection (USNPC), Belts- vent length [SVL] = 490 mm ± 69 s, range = ville, Maryland: Ochetosoma ellipticum (90611), 366–653 mm) were borrowed from the her- Alaria sp. (mesocercariae) (90612), Ophidascaris petology collection of the University of Ari- labiatopapillosa (90614), Abbreviata terrapenis zona, Tucson (UAZ): COCHISE COUNTY 42342, (90613), Physaloptera sp. (larvae) (90615). 42478, 44290; GREENLEE COUNTY 42711, Ochetosoma ellipticum was originally de- 42712; PIMA COUNTY 26523, 26528, 26531, scribed by Pratt (1903) from specimens found 26542, 26552, 26556, 26558, 26559, 26562, in the mouth of an eastern hognose snake, 26564, 26567, 26569, 26570, 42713, 44976, Heterodon platirhinos (locality of collection 47141; SANTA CRUZ COUNTY 26561, 26565. not given), as Renifer ellipticus which was re- The snakes were initially fixed in 10% forma- assigned to its current taxonomic position by lin, washed in water, and stored in 70% iso- Yamaguti (1958). It has been reported from propanol. The body cavity was opened ven- other colubrid snakes of Central and North trally and the gastrointestinal tract removed America (Yamaguti 1958). The life cycle of O. by cutting across the anterior end of the ellipticum has not been studied. However, esophagus and the posterior end of the large Sogandares-Bernal and Grenier (1971) studied intestine. The gastrointestinal tract was then the life cycles of O. kansensis and O. latero- opened longitudinally and searched for hel- trema and found metacercariae encysted in minths using a dissecting microscope. The tadpoles of the leopard frog, Rana pipiens. body cavity and liver were also examined for Adult trematodes were recovered from the helminths. Nematodes were cleared individu- mouths of young cottonmouth snakes, Agki- ally in a drop of concentrated glycerol on a strodon piscivorus. Because T. cyrtopsis feeds glass slide and identified from these wet mainly on tadpoles and adult anurans in

1Department of Biology, Whittier College, Whittier, CA 90608. 2Department of Biology, Pennsylvania State University, Shenango Campus, Sharon, PA 16146.

243 244 WESTERN NORTH AMERICAN NATURALIST [Volume 62

TABLE 1. Infection site, total numbers of helminths, prevalence, mean intensity ± 1 s, and range for gastrointestinal helminths from 23 Thamnophis cyrtopsis cyrtopsis from Arizona. Infection Total Prevalence Mean Helminth site number (%) intensity ± 1 s Range

TREMATODA Ochetosoma ellipticum Esophagus 67 30 9.6 ± 11.1 1–32 Alaria sp. (mesocercariae) Esophagus 19 4 19.0 — NEMATODA Abbreviata terrapenis Esophagus 2 9 1.0 — Ophidascaris labiatopapillosa Stomach, small intestine 11 13 3.7 ± 3.8 1–8 Physaloptera sp. (larvae) Stomach, small intestine 15 17 3.8 ± 3.1 1–8

aquatic habitats (Rossman et al. 1996), it likely snake (see Goldberg and Bursey 1999). Walton acquires O. ellipticum by eating tadpoles (1937) found larvae of O. labiatopapillosa infected with metacercariae. Thamnophis cyr- encysted in the stomach wall, mesentery, and topsis is a new host record for O. ellipticum. muscles of the amphibians Amphiuma tridac- Likewise, mesocercariae of Alaria sp. de- tylum, Rana capito aesopus (= Rana aesopus), velop in the muscles of ranid and bufonid tad- R. catesbeiana, and R. sphenocephala. Tham- poles (Schell 1985). Garter snakes serve as nophis cyrtopsis may become infected with paratenic (transport) hosts (Schell 1985). This both A. terrapenis and O. labiatopapillosa by is the first report of mesocercariae of Alaria eating infected frogs. Thamnophis cyrtopsis is sp. from T. cyrtopsis. a new host record for O. labiatopapillosa. Abbreviata terrapenis was originally de- Third stage larvae of Physaloptera sp. were scribed from the ornate box turtle, Terrapene found in the stomach and intestines of 4 T. ornata, from Oklahoma (Hill 1945) and is the cyrtopsis (Table 1). The majority of the 16 only species of its genus known from North species of Physaloptera from North America America (Baker 1987). Abbreviata ranae, a lar- listed by Morgan (1941) are mammalian para- val form (considered a species inquirenda by sites. All species of Physaloptera require an Baker 1987), was described by Walton (1931) insect intermediate host (Anderson 2000), and from cysts in the stomach and intestine of a larvae of Physaloptera sp. could be expected bullfrog, Rana catesbeiana. However, the adult in any . Amphibians and reptiles form of A. ranae has not been described. Lar- harboring larvae of Physaloptera sp. but not val Abbreviata in cysts have been reported by adults are summarized in Goldberg et al. numerous authors (see Baker 1987) from North (1993). Because most of these larvae were American frogs, which are considered para- found in the intestinal lumen of the snakes tenic hosts by McAllister and Freed (1992). examined, we believe they are a byproduct of Whether A. ranae is the larval form of A. ter- diet and will be passed in feces. Thus, they are rapenis is yet to be resolved. We could find no not parasites of these snakes. This is the first morphological differences between our speci- report of physalopteran larvae in T. cyrtopsis. mens and the description by Hill (1945), nor In view of the extensive geographic distri- were differences found between our speci- bution of T. cyrtopsis (Rossman et al. 1996), mens and specimens of A. terrapenis acces- subsequent examination of different popula- sioned in the United States Parasite Collection tions will be needed before the helminth (USNPC 83537, 85038). This is the first report diversity of this snake is known. of A. terrapenis from a snake; T. cyrtopsis is a new host record. We thank Charles H. Lowe (University of Ophidascaris labiatopapillosa was first de- Arizona) for permission to examine T. cyrtopsis. scribed from an eastern racer, Coluber con- strictor, collected in Florida by Walton (1927) LITERATURE CITED and is the only member of its genus known to occur in the United States (Baker 1987). It has ANDERSON, R.C. 2000. Nematode parasites of vertebrates: their development and transmission. CABI Publish- been reported from 6 species of colubrids (but ing, CAB International, Wallingford, Oxon, U.K. 650 not Thamnophis) and 1 species of viperid pp. 2002] NOTES 245

BAKER, M.R. 1987. Synopsis of the Nematoda parasitic in SCHELL, S.C. 1985. Handbook of Trematodes of North amphibians and reptiles. Memorial University of America north of Mexico. University Press of Idaho, Newfoundland Occasional Papers in Biology 11: Moscow. 263 pp. 1–325. SOGANDARES-BERNAL, F., AND H. GRENIER. 1971. Life GOLDBERG, S.R., AND C.R. BURSEY. 1999. First reported cycles and host-specificity of the plagiorchiid trema- occurrence of Ophidascaris labiatopapillosa (Nema- todes Ochetosoma kansensis (Crow, 1913) and O. lat- toda: Ascarididae) in the red diamond rattlesnake. erotrema (Byrd and Denton, 1938). Journal of Para- California Fish and Game 85:181–182. sitology 57:297. GOLDBERG, S.R., C.R. BURSEY AND R. TAWIL. 1993. Gastro- STEBBINS, R.C. 1985. A field guide to western reptiles and intestinal helminths of the western brush lizard, amphibians. Houghton Mifflin Company, Boston. Urosaurus graciosus graciosus (Phrynosomatidae). 336 pp. Bulletin of the Southern California Academy of Sci- WALTON, A.C. 1927. A revision of the nematodes of the ences 92:43–51. Leidy collections. Proceedings of the Academy of HILL, W.C. 1945. Physaloptera terrapenis, a new nema- Natural Sciences of Philadelphia 79:49–163. tode from a tortoise. Transactions of the American ______. 1931. Note on some larval nematodes found in Microscopical Society 60:59–64. frogs. Journal of Parasitology 17:228–229. MCALLISTER, C.T., AND P. S . F REED. 1992. Larval Abbrevi- ______. 1937. The Nematoda as parasites of Amphibia. ata sp. (Spirurida: Physalopteridae) in introduced III. Studies on life histories. Journal of Parasitology Rio Grande chirping frogs, Syrrhophus cystigna- 23:299–300. thoides campi (Anura: Leptodactylidae), from Hous- YAMAGUTI, S. 1958. Systema helminthum. Volume I. The ton, Texas. Texas Journal of Science 44:359–361. digenetic trematodes of vertebrates. Part I. Inter- MORGAN, B.B. 1941. A summary of the Physalopterinae science Publishers, Inc., New York. 979 pp. (Nematoda) of North America. Proceedings of the Helminthological Society of Washington 8:28–30. Received 1 September 2000 PRATT, H.S. 1903. Descriptions of four distomes. Pages Accepted 15 February 2001 24–39 in Mark Anniversary Volume. Henry Holt, New York. ROSSMAN, D.A., N.B. FORD, AND R.A. SEIGEL. 1996. The garter snakes: evolution and ecology. University of Oklahoma Press, Norman. 332 pp. Western North American Naturalist 62(2), © 2002, pp. 246–248

COMPARISON OF NITROGEN ISOTOPE RATIOS IN FEATHERS FROM SEVEN SPECIES OF COLORADO BREEDING BIRDS

Sujay S. Kaushal1,2 and John J. Walsh2

Key words: birds, , isotopes, granivores.

The natural abundance of nitrogen isotopes We conducted a preliminary study to inves- in biomass has been useful in demonstrating tigate whether there is a relationship between trophic relationships of organisms in aquatic trophic level and nitrogen isotopes in feathers food webs (Hamilton and Lewis 1992, Hamil- from breeding birds of Colorado. We classified ton et al. 1992). The technique is capable of birds into 3 broad feeding groups: granivores establishing trophic positions of organisms (birds primarily eating seeds), insectivores that can feed on a variety of food sources and (birds primarily eating insects), and raptors whose feeding behavior cannot be observed (birds eating primarily small mammals and easily. The stable isotope technique is based birds). We used this approach because the on increase in the 15N content of biomass with actual trophic level of birds is very difficult to each trophic transfer. The 15N content of a determine. Feeding classifications for selected sample is often expressed as standard δ nota- species in this study were obtained from tion (parts per thousand deviation in the ratio Kingery (1998). We hypothesized that the δ15 of 15N and 14N in a sample from the ratio of mean N of feathers from granivores would 15N and 14N in air; Peterson and Fry 1987). It be significantly lower than those for insecti- is usually assumed that there is a 2.6–3.4 part vores and raptors (which represented birds in per mil increase in 15N content between suc- higher trophic levels). Tail feathers of adult specimens were ob- cessive trophic levels (DeNiro and Epstein tained from the museum at the University of 1981, Minagawa and Wada 1984, Owens 1987) Colorado, Boulder. In addition, tail feathers and that trophic fractionation can range as were also collected from a nestling that lived high as 5 parts per mil (Michener and Schell in a ponderosa pine (Pinus ponderosa) forest 1994). The exact value for isotopic fractiona- located in Boulder County, Colorado. Collection tion associated with an increase in trophic locations for museum specimens extended level may vary according to an animal’s age, throughout Colorado, but most specimens body size, metabolic rate, and the amount of were from Boulder County. In our study we protein in the animal’s diet (McCutchan 1999). assumed that the tail feathers were grown on Avian ecologists have used stable isotopes the breeding grounds in Colorado and that the to evaluate trophic relationships in songbirds isotopic ratios in the feathers reflected the diet (Hobson 1999) and seabirds (Hobson and at the time the feathers were grown. Feathers Montevecchi 1991, Thompson et al. 1999). were freeze-dried and stored until analysis in However, isotope data on songbirds are rela- a desiccator. Samples were placed in tin cups tively lacking compared to other taxa (Kelly and combusted in a Fisons elemental analyzer 2000). Results from these studies and other interfaced with a Finnigan dual inlet mass studies look promising, although the potential spectrometer. For a more detailed description of stable isotopes has not yet been explored of methodology, see Hamilton and Lewis (1992). fully. Isotopic composition values for primary The standard error of the mean for 3 repli- consumers (such as granivores) are particu- cates was <0.4 parts per mil. Both EDTA and larly scarce in the literature. glycine were used as standards.

1Center for Limnology, Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, CO 80309-0216. 2Department of Environmental, Population and Organismic Biology, University of Colorado, Boulder, CO 80309-0216.

246 2002] NOTES 247

Table 1. Nitrogen isotope ratios of bird feathers sampled from museum and field specimens obtained in Colorado. Locations given are counties in Colorado. Species δ15N Feeding group Type Location Dark-eyed Junco ( Junco hymalis caniceps) 6.4 Granivore Museum Boulder 5.7 Museum Boulder 6.3 Museum Boulder Lesser Goldfinch (Carduelis psaltria) 4.7 Granivore Museum Montrose 4.6 Museum El Paso Plumbeous Vireo (Vireo plumbeus) 6.3 Insectivore Museum Douglas 5.8 Museum Rio Blanco 6.1 Museum Rio Blanco Virginia’s Warbler (Vermivora virginiae) 5.2 Insectivore Museum El Paso 6.1 Museum El Paso 6.3 Museum El Paso Western Wood-pewee (Contopus sordidulus) 9.4 Insectivore Museum El Paso 7.6 Field Boulder Sharp-shinned Hawk (Accipiter striatus) 7.2 Raptor Museum Boulder 7.6 Museum Boulder 7.3 Museum Boulder Great-horned Owl (Bubo virginianus) 6.9 Raptor Museum Boulder 6.9 Museum Boulder 8.1 Museum Mineral

Table 1 summarizes our results. As hypoth- and Organismic Biology and the Graduate esized, feathers from granivores had a lower School at the University of Colorado. Rosanne δ15 mean N (5.5; sx– 0.4) than feathers from Humphrey provided access to specimens from feeding groups that represented higher the CU museum. Dr. W.M. Lewis, Jr., provided trophic levels. A 1-tailed t test indicated that suggestions that improved this manuscript. the mean δ15N of feathers from insectivores (6.6; sx– 0.5) was significantly higher than from LITERATURE CITED granivores (P = 0.05). In addition, a 1-tailed t test showed that the mean δ15N of feathers DENIRO, M.J., AND S. EPSTEIN. 1981. Influence of diet on the distribution of nitrogen isotopes in animals. Geo- taken from raptors (7.3; sx– 0.7) was significant- chimica et Cosmochimica Acta 45:341–351. 15 ly higher than the mean δ N of feathers from HAMILTON, S.K., AND W. M. L EWIS, JR. 1992. Stable carbon granivores (P = 0.01). and nitrogen isotopes in algae from the Results from this study support a growing Orinoco River floodplain, Venezuela. Geochimica et body of literature showing that the use of 15N Cosmochimica Acta 56: 4237–4246. HAMILTON, S.K., W.M. LEWIS, JR., AND S.J. SIPPEL. 1992. natural abundance in feathers has the poten- Energy sources for aquatic animals in the Orinoco tial to discriminate among different trophic River floodplain: evidence from stable isotopes. levels in birds. A possible source of error in Oecologia 89:324–330. our study could have been the different loca- HOBSON, K.A. 1999. Stable-carbon and nitrogen isotope ratios of songbird feathers grown in two terrestrial tions of our specimens and the assumption biomes: implications for evaluating trophic relation- that feathers were grown on the breeding ships and breeding origins. Condor 101:799–805. grounds. Previous work has shown that iso- HOBSON, K.A., AND W.A. MONTEVECCHI. 1991. Stable iso- tope signatures can be affected by local land- topic determinations of trophic relationships of Great Auks. Oecologia 87:528–531. scape attributes (Hobson 1999, Vander Zan- KINGERY, H.E. 1998. Colorado breeding birds atlas. Col- den 1999). To make trophic comparisons at a orado Division of Wildlife, Denver. finer resolution, the location of feather growth MCCUTCHAN, J.H. 1999. Carbon sources for macroinver- should be considered. Nonetheless, our pre- tebrates in St. Vrain Creek, Colorado. Doctoral dis- liminary work shows that nitrogen isotopes sertation, University of Colorado, Boulder. MICHENER, R.H., AND D.M. SCHELL. 1994. Stable isotope may be an important tool for investigating dif- ratios as tracers in marine aquatic food webs. Pages ferences in the reliance of western songbirds 138–157 in K. Lajtha and R.H. Michener, editors, on various food sources. Stable isotopes in ecology and environmental sci- ence. Blackwell Scientific Publications, London. MINAGAWA, M., AND E. WADA. 1984. Stepwise enrichment Funding was provided by grants from the of 15N along food chains: further evidence of the Department of Environmental, Population relation between δ15N and animal age. Geochimica 248 WESTERN NORTH AMERICAN NATURALIST [Volume 62

et Cosmochimica Acta 48: 1135–1140. Trophic relationships among six species of Icelandic OWENS, N.J.P.1987. Natural variations in 15N in the marine seabirds as determined through stable isotope analy- environment. Advances in Marine Biology 24: sis. Condor 101:898–903. 389–451. VANDER ZANDEN, M.J., AND J.B. RASMUSSEN. 1999. Primary PETERSON, B.J., AND B. FRY. 1987. Stable isotopes in eco- δ13C and δ15N and the trophic position of system studies. Annual Review of Ecology and Sys- aquatic consumers. Ecology 80:1395–1404. tematics 18:293–320. THOMPSON, D.R., K. LILLIENDAHL, JON SOLMUNDSSON, Received 27 April 2000 R.W. FURNESS, S. WALDRON, AND R.A. PHILLIPS. 1999. Accepted 21 November 2001 Western North American Naturalist 62(2), © 2002, pp. 249–252

OSPREY DIET ALONG THE EASTERN SIDE OF THE GULF OF CALIFORNIA, MEXICO

Jean-Luc E. Cartron1,2 and Manuel C. Molles, Jr.1

Key words: Osprey, Pandion haliaetus, Gulf of California, diet, Mugil cephalus, Tylosurus crocodilus.

A large resident population of Ospreys (Pan- sloping pebble beach near the rocky headland, dion haliaetus) has been documented in the Punta Tepopa. At the northern end of our Gulf of California and Baja California region study area, the 4-km stretch of coast north of of Mexico (Henny and Anderson 1979, Rei- El Desemboque (n = 10, 1995 and 1996) is therman and Storrer 1981, 1982, Judge 1983, located just south of the mouth of the inter- Danemann and Guzmán Poo 1992, Castel- mittent Rio San Ignacio. lanos and Ortega-Rubio 1995, Cartron 2000). The breeding season of Ospreys in the Gulf However, Osprey diet in this region has been of California is characterized by a high degree reported only for the Bahia de Los Angeles of asynchrony among pairs (Judge 1983, Islands (Judge 1981). To date, no published Cartron 2000). The onset of egg laying typi- information on diet exists for Ospreys nesting cally ranges from early January to early along the coast of Sonora on the eastern side March, with fledging occurring from April to of the Gulf of California. The Bahia de Los early June (Judge 1983). In 1995 prey remains Angeles Islands are small, rocky islands, where- were collected twice, mid-March and late as the coast of Sonora is characterized by May; in 1996 they were collected monthly extensive sandy beaches, shallow bays, and from January through May. estuaries (Molles 1978, Alvarez-Borrego 1983, We collected prey remains under nests and Alvarez-Borrego and Lara-Lara 1991, Thomson at the base of cardon (Pachycereus pringelei) et al. 2000). Because the Osprey is an oppor- cacti used by Ospreys for feeding. Nest distur- tunistic rather than selective fish-eating species bance was minimal (see Cartron 2000). A large (Poole 1989), its prey base is likely to differ number of collected prey items were suffi- between these 2 geographic areas. ciently intact for easy identification using In 1995 and 1996 we studied the diet of 39 Thomson and McKibbin’s (1976) guide to Gulf and 40 Osprey pairs, respectively, nesting along of California fish; they were subsequently the eastern side of the Gulf of California. Our used as references to identify other remains. study focused on 3 locations: Bahia Sargento, Some prey species were tallied using specific Mancha Blanca, and El Desemboque (Fig. 1). bones (e.g., jaw bones for Tylosurus crocodilus Each location provided distinctive foraging or Calamus brachysomus, opercula for Mugil environments and therefore potentially differ- cephalus). Other prey species (e.g., Balistes ent prey bases for Ospreys. Bahia Sargento (n polylepis), which were found in the form of = 22 occupied Osprey nests, 1995; n = 23, (nearly) complete skeletons with attached skin, 1996) is a wide, shallow bay dotted with scat- were simply counted. Scavenging of some prey tered rocky patch reefs. Except toward its remains by coyotes (Canis latrans) occurred in northern end, it is bordered by extensive the study area but appeared to be fairly con- beaches of fine sand. Along the south-facing sistent through time. side of the bay, a large negative estuary lined During the study period we recorded 1385 by mangrove encroaches upon the land. To the prey at the 3 sites; these prey represented at north of Bahia Sargento, Mancha Blanca (n = least 19 species from 18 families (Table 1). At 7, 1995 and 1996) is characterized by a steeply all 3 locations bullseye puffers (Sphoeroides

1Biology Department, University of New Mexico, Albuquerque, NM 87131. 2Corresponding author.

249 250 WESTERN NORTH AMERICAN NATURALIST [Volume 62

Fig. 1. Location of the study area on the eastern side of the Gulf of California. The 3 study sites (Bahia Sargento, Mancha Blanca, and El Desemboque) and all Osprey occupied nests are marked. annulatus) were occasionally found on the needlefish seemingly represented the bulk of ground next to Osprey prey remains, but these the Osprey diet, with Pacific porgy and poisonous fish never appeared to have been finescale triggerfish amounting to ≤5% of the consumed. Based on prey remains, the fish diet (Table 1). species most frequently consumed by Ospreys There was no significant (χ2 = 5.23, df = 3, at El Desemboque appeared to be Mexican P > 0.05) difference between 1995 and 1996 needlefish (Tylosurus crocodilus), followed re- in the proportion of the 4 main species (Mexi- spectively by striped mullet (Mugil cephalus), can needlefish, striped mullet, finescale trig- finescale triggerfish (Balistes polylepis), and gerfish, and Pacific porgy) recorded among Pacific porgy (Calamus brachysomus; Table 1). prey remains at El Desemboque. Similarly, no In 1995 and 1996 these species represented significant (χ2 = 0.18, df = 1, P > 0.05) differ- 83–87% of prey remains at that location. At ence was noted in the proportion of the 2 the other 2 sites striped mullet followed by main fish species (striped mullet and Mexican 2002] NOTES 251

TABLE 1. Number (%) of Osprey prey found at the 3 study sites during the 1995 and 1996 nesting seasons.

Bahia Sargento Mancha Blanca El Desemboque Common family ______Family name Species name 1995 1996 1995 1996 1995 1996

Rhinobatidae Guitarfishes Rhinobatos productus 0 0 1 (1) 0 1 0 Clupeidae Herrings Unidentified 0 0 1 (1) 0 0 0 Ariidae Sea catfishes Bagre panamensis 1 0 0 0 3 (3) 0 Synodontidae Lizardfishes Synodus scituliceps 0 00010 Belonidae Needlefishes Tylosurus crocodilus 78 (28) 85 (28) 20 (22) 45 (39) 45 (39) 183 (38) Fistularidae Cornetfishes Fistularia commersonii 0 00001 Scorpaenidae Rockfishes Scorpaena mystes 0 0 0 0 4 (3) 0 Triglidae Searobins Prionotus ruscarius 0 00001 Serranidae Sea basses Paralabrax maculofasciatus 1 00000 Unidentifieda 0 1 0 3 (3) 6 (5) 39 (8) Carangidae Jacks Seriola lalandi 0 0 1 (1) 0 0 1 Haemulidae Grunts Unidentified 0 0 1 (1) 0 0 0 Sparidae Porgies Calamus brachysomus 10 (4) 7 (2) 0 2 (2) 7 (6) 50 (10) Kyphosidae Seachubs Girella simplicidens 1 00000 Mugilidae Mullets Mugil cephalus 177 (63) 209 (69) 60 (66) 63 (55) 33 (29) 110 (23) Labridae Wrasses Unidentified 0 0 0 0 0 6 (1) Scombridae Mackerels and tunas Scomberomorus sierra 4 (1) 0 0 0 2 (2) 3 Pleuronectidae Righteye flounders Unidentified 0 0 0 0 1 0 Balistidae Triggerfishes Balistes polylepis 2 3 5 (5) 2 (2) 11 (10) 73 (15) Unidentified 5 2 2 (1) 0 1 11 (2) TOTAL 279 (100) 307 (100) 91 (100) 115 (100) 115 (100) 478 (100) aIncludes at least one other serranid species besides Paralabrax maculofasciatus.

needlefish) among prey remains at Bahia Sar- porgy. Overall, however, mullet or needlefish gento. At Mancha Blanca there was a signifi- was always the primary prey species at any cant (χ2 = 5.64, df = 1, P ≤ 0.05) difference in time in any collection of prey remains. the proportion of striped mullet and Mexican In 1995 and 1996 the proportion of needle- needlefish between 1995 and 1996. Yet, this fish and mullet in the diet of Ospreys appeared pattern may have been the result of missing to vary through time. In 1995 the relative pro- data for the month of February at that location portion of these 2 prey species was signifi- (prey remains were collected that month but cantly different between the 1st (March) and were lost before they were analyzed). Minor 2nd (May) samples we collected at El Desem- differences between 1995 and 1996 were noted boque (χ2 = 4.8, df = 1, P ≤ 0.05), Mancha at El Desemboque: stone scorpionfish (Scor- Blanca (χ2 = 14.29, df = 1, P ≤ 0.05), and paena mystes) and chihuil (Bagre panamensis), Bahia Sargento (χ2 = 12.88, df = 1, P ≤ 0.05). which seemingly represented 6% of the prey In 1996 the relative decrease in striped mullet base of Ospreys at that location in 1995, were and increase in Mexican needlefish was appar- not found among prey remains in 1996. ent as early as February and continued through Differences in diet among pairs were not April (Fig. 2). analyzed due to small sample size for nests Our results suggest that the diet of Ospreys along Bahia Sargento or at Mancha Blanca. At along the eastern side of the Gulf of California El Desemboque the high density of nests typi- is substantially different from that of Ospreys cally prevented us from identifying with cer- on islands of the Gulf. Ospreys nesting on tainty the pair that had fed on a fish whose islands in Bahia de Los Angeles rely most con- remains we found. One exception was a pair at sistently on Gulf opaleye (Girella simplicidens) the northern end of the study site. Between and sand basses and sea basses (Paralabrax mid-March and mid-April 1996, Pacific porgy and Epinephalis spp.; Judge 1981). In contrast, seemingly comprised 29% of its diet. During the diet of Ospreys in our study area appeared that same period the diet of all other pairs at to include chiefly striped mullet and Mexican El Desemboque appeared to consist of <5% needlefish. Locally, species such as the Pacific 252 WESTERN NORTH AMERICAN NATURALIST [Volume 62

ALVAREZ-BORREGO, S., AND J.R. LARA-LARA. 1991. The physical environment and primary productivity in the Gulf of California. Pages 555–567 in B.R.T. Simoneit and J.P. Dauphin, editors, The Gulf and peninsular province of the Californias. American Association of Petroleum Geologists Memoirs 47. CARTRON, J.-L.E. 2000. Status and productivity of Ospreys along the eastern coast of the Gulf of California: 1992–1997. Journal of Field Ornithology 71:298–309. CASTELLANOS, A., AND A. ORTEGA-RUBIO. 1995. Artificial nesting sites and Ospreys at Ojo de Liebre and Guerrero Negro Lagoons, Baja California Sur, Mex- ico. Journal of Field Ornithology 66:117–127. DANEMANN, G.D., AND J.R. GUZMÁN POO. 1992. Notes on the birds of San Ignacio Lagoon, Baja California Sur, Mexico. Western Birds 23:11–19. HENNY, C.J., AND D.W. ANDERSON. 1979. Osprey distribu- tion, abundance, and status in western North Amer- ica: III. The Baja California and Gulf of California population. Bulletin of the Southern California Academy of Science 78:89–106. JUDGE, D.S. 1981. Productivity and provisioning behavior of Ospreys (Pandion haliaetus) in the Gulf of Califor- nia. Master’s thesis, University of California, Davis. Fig. 2. Monthly variation in the proportion of Mexican ______. 1983. Productivity of Ospreys in the Gulf of Cali- needlefish vs. striped mullet in the diet of Ospreys at all fornia. Wilson Bulletin 95:243–255. 3 sites combined, January–May 1996. Sample size is only MOLLES, M.C., JR. 1978. Fish species diversity on model 2 (Bahia Sargento and El Desemboque) in February. and natural reef patches: experimental insular bio- Although declines in mullet relative to needlefish are geography. Ecological Monographs 48:289–305. observed at Bahia Sargento, Mancha Blanca, and El POOLE, A. 1989. Ospreys: a natural and unnatural history. Desemboque, their timing or magnitude varies somewhat Cambridge University Press, Cambridge, U.K. among these locations. REITHERMAN, B., AND J. STORRER. 1981. A preliminary report on the reproductive biology and ecology of the Whale Island Osprey (Pandion haliaetus) popula- tion, San Ignacio Lagoon, Baja California Sur, Mex- porgy or the finescale triggerfish became addi- ico. Western Foundation of Vertebrate Zoology, Los tional important prey species. Angeles. ______. 1982. The reproductive biology and ecology of the Whale Island Osprey (Pandion haliaetus) popula- We thank A. Thomas and R. Nugent for tion, San Ignacio Lagoon, Baja California Sur, Mex- providing assistance in the field. Reviews and ico. Western Foundation of Vertebrate Zoology, Los suggestions from C. Monson and H. Schaadt Angeles. helped to improve an earlier draft of this THOMSON, D.A., L.T. FINDLEY, AND A.N. KERSTITCH. 2000. Reef fishes of the Sea of Cortez: the rocky-shore paper. fishes of the Gulf of California. Revised edition. Uni- versity of Texas Press, Austin. LITERATURE CITED THOMSON, D.A., AND N. MCKIBBIN. 1976. Gulf of California fishwater’s guide. Golden Puffer Press, Tucson, AZ. ALVAREZ-BORREGO, S. 1983. Gulf of California. Pages 427– 429 in B.H. Ketchum, editor, Estuaries and enclosed Received 7 August 2000 seas. Elsevier, Amsterdam. Accepted 1 May 2001 Western North American Naturalist 62(2), © 2002, p. 253

MULE DEER GROUP KILLS COYOTE

Jack A. Wilkinson1 and John F. Douglass2

Key words: Canis latrans, Odocoileus hemionus, mobbing behavior, predator attacked by prey.

Coyotes (Canis latrans) are known to be frequented the hills east of the alfalfa field in predators of both young and adult mule deer the fall of 1996: they were observed feeding (Odocoileus hemionus; Cahalane 1947, Bekoff on small rodents in the open, and on several 1977, Bowyer 1987). Aggressive defense against occasions groups of mule deer were seen chas- coyotes by adult female mule deer, involving ing the coyotes away when the latter ventured chasing and vigorous kicking with the fore- too close. On the morning of 28 October 1996 legs, has been observed in does acting both the coyotes were heard howling more than singly and in groups (Cahalane 1947, Seiden- usual; it is possible that a deer kill had been sticker et al. 1973, MacConnell-Yount and made that morning. Immediately following the Smith 1978, Hamlin and Schweitzer 1979, death of the young coyote described above, 1/4 Truett 1979, Wenger 1981, Bowyer 1987, Grif- mile away, a group of 3 coyotes (2 adults and 1 fith 1988). The observation reported here yearling) was seen in the alfalfa field adjoining appears to be the first instance in which a a gully frequented by deer at night. coyote is known to have been killed by such an attack. We thank Darrylyn Douglass Garrett and At 0800 hours MST on 28 October 1996, Dr. Marc Bekoff for their help with this paper. the senior author was watching a group of LITERATURE CITED about 50 mule deer feeding in a 60-acre alfalfa field near the South Fork of the Shoshone BEKOFF, M. 1977. Canis latrans. Mammalian Species 79: River, 18 miles SW of Cody, Park County, 1–9. Wyoming. A large doe (number 1) was seen BOWYER, R.T. 1987. Coyote group size relative to preda- chasing a single coyote relentlessly around the tion on mule deer. Mammalia 51: 515–526. CAHALANE, V.H. 1947. A deer-coyote episode. Journal of field. As the doe tired, other does and fawns Mammalogy 28:36–39. joined the chase; following her periods of rest, GRIFFITH, B. 1988. Group predator defense by mule deer number 1 would join the chase again. After in Oregon. Journal of Mammalogy 69:627–629. HAMLIN, K.L., AND L.L. SCHWEITZER. 1979. Cooperation 10–15 minutes of this activity, the coyote by coyote pairs attacking mule deer fawns. Journal apparently tired while being pursued by num- of Mammalogy 60:849–850. ber 1, and the doe was observed to strike the MACCONNELL-YOUNT, E., AND C. SMITH. 1978. Mule coyote with a front hoof. Within seconds, 7 or deer–coyote interactions in north-central Colorado. 8 does surrounded the coyote, rearing up and Journal of Mammalogy 59:422–423. SEIDENSTICKER, J.C., M.G. HORNOCKER, W.V. WILES, AND striking it with their front hooves. Other deer J.P. MESSICK. 1973. Mountain lion social organization in the field stood watching this attack. Fifteen in the Idaho Primitive Area. Wildlife Monographs minutes later, as the deer began to move off, 35:1–60. the senior author approached and examined the TRUETT, J.C. 1979. Observations of coyote predation on mule deer fawns in Arizona. Journal of Wildlife Man- dead coyote. It appeared to be a yearling; the agement 43:956–958. body was badly broken by hooves and had been WENGER, C.R. 1981. Coyote–mule deer interaction obser- trampled into the ground. vations in central Wyoming. Journal of Wildlife The behavioral context of this attack is not Management 45:770–772. fully known, but the following observations Received 11 November 1999 seem worthy of note. A group of coyotes Accepted 5 March 2001

11612 Southfork Road, Cody, WY 82414. 23347 Airport Highway, Toledo, OH 43609. Corresponding author.

253 Western North American Naturalist 62(2), © 2002, p. 254

BOOK REVIEW

Bird Hand Book. Photos by Victor Schrager, might become a conversation piece for the text by A.S. Byatt. Graphis Inc., New curious visitor. York, New York. 2001. $60.00, hardcover; To be sure, however, if one enjoys historical 128 pages, 80 quadtone photographs. prose and poetry, some relating to birds, then ISBN 1-93124-104-X. the book may yet take on more interest. Byatt’s essays are a literary romp through 5000 years It is difficult to become overly excited about of bird writings. As examples, the text quotes this book. The title, as I see it, is a play on passages from an 1850 discussion by George words, inasmuch as the book features pho- Eliot about how things are named, gives a son- tographs of birds being held in hands that are net by Robert Frost about a bird’s voice, dis- themselves protruding from holes in sheets cusses the relationships between birds and which serve as backdrops. The publishers people during mediaeval times, explores some describe this somewhat differently: “[B]irds of the early social functions of birds to such are seen in a totally new way; the presence of well-known Christians as St. Francis, repeats a human hand, acting as a delicate pedestal, the poetry of Death of Cock Robin opposite provides the reader with a sense of scale for a picture of an American Robin, and describes the relationship between birds and humans, material from the journal of John James scale not present in any bird field guide.” Audubon. The actual photographed birds are usually Some biology is featured. A discussion of in some contorted position, and, even after anatomical structures of the skeleton, informa- careful searching, I could not find the delicate tion on the structure of feathers, and so forth, pedestal mentioned. Some are injured captive are opposite the photo of what is termed a birds, some have broken feathers, one lacks a hybrid pigeon. Fortified by quotes from tail, and one is a mounted passenger pigeon Rachel Carson’s Silent Spring, the author enu- lying on its side. Most birds are restrained in merates problems generated by agricultural some fashion; for example, a Brown Pelican chemicals. with a distorted beak is being restrained by an The book has a sturdy binding and is well arm coming out from behind the sheet. A put together. However, with the cost of books Gouldian Finch, with only the head seen clearly in general and the choices of excellent refer- between thumb and forefinger, is called a ence, historical, and prose/poetry books on Zebra Finch. Twelve photos are of domestic birds, I find this book too pricey for what one poultry. I think these comments give the flavor. gets. Perhaps it is a book for the public library, If one enjoys good photography with birds but I cannot recommend it. as the subject, then this is not the book to pur- chase. But, if one appreciates a unique pho- Clayton M. White tography style—perhaps I could call it “new Department of Zoology age”—then the book may have a bit more Brigham Young University appeal. Such a book lying on a coffee table Provo, UT 84602

254 Western North American Naturalist 62(2), © 2002, p. 255

BOOK REVIEW

The World of the Hummingbird by Robert mingbird in a nest with a second egg about to Burton. Firefly Books (U.S.) Inc., Buffalo, hatch, photo by Sid and Shirley Rucker; a New York. 2001. $40.00, hardback; 158 hummingbird drinking tree sap, photo by Jack pages, 151 color photographs. ISBN Wilburn; a White-bearded Hermit showing 1-55209-607-6. the great extent of the tongue, photo by Luisa Mazariegos; and a Magenta-throated Woodstar The World of the Hummingbird is a large, showing digested nectar passing out the cloaca 9 × 11 inches, coffee table sort of book that in as the bird feeds on a blossom, photo by addition to beautiful photographs also affords Michael and Patricia Fogden. great insights into the private lives of humming- I was fascinated by the chapter on human birds. It is printed on glossy, 100-lb., acid-free interactions with hummingbirds. The earlier paper. The author has accumulated color plates Peruvian cultures of 1500 years ago apparently from scores of internationally renowned wild- had a particular interest in them. A 300-foot life photographers, some of the finest humming- symbol on the ground of a hummingbird was bird photographs available in book form. No made by the Nazca of the Peruvian plateau. fewer than 88+ hummingbird species are The Nazca also made wood carvings of them. I shown, along with photographs of habitat and learned that the Taino people of the Caribbean, other critters that do much the same as hum- at the time of Columbus’s first encounter, called mingbirds, i.e., bats, Old World sunbirds, and their warriors colibri; they were impressed insects. with how hummingbirds fought to protect their Robert Burton has written several other territories. That term now forms a hummingbird books dealing with bird behavior, bird flight, genus name. and bird migration. In this book he deals with Because hummingbirds are so speciose, some interesting and complex topics relative many have small ranges often where unique to hummingbirds that make delightful reading habitats occur. Today, with the loss of habitats, for the scientist and lay person alike. He has the Hook-billed Hermit may consist of fewer divided the book into 7 chapters in which he than 250 individuals in a 40-square-mile frag- introduces hummingbirds and then discusses mented region of Brazil’s lowland Atlantic for- their flight, their relationships with plants and est; they were formerly abundant over a range humans, nesting behavior, migration, myths of 13,500 square miles. and legends about them, trade and commer- This book has much to offer both the scien- cial use of hummingbird feathers, and lastly tist and lay person, and the price makes it an current threats to their habitats, the latter of attractive purchase. It has a sturdy binding of great concern to conservation biology. lasting quality. I highly recommend it. Some photographs particularly caught my eye. First was a dorsal view sequence, by Clayton M. White Stephen Dalton, of 7 figures of the same hum- Department of Zoology mingbird showing flight and illustrating how Brigham Young University the wings make the more-or-less figure of 8 Provo, UT 84602 pattern; a newly hatched Black-chinned Hum-

255 256 WESTERN NORTH AMERICAN NATURALIST [Volume 62

CONTENTS

(Continued from back cover)

Movements of a bullsnake (Pituophis catenifer) following predation of a radio- collared northern pocket gopher (Thomomys talpoides) ...... Ray T. Sterner, Brett E. Petersen, Stephen A. Shumake, Stanley E. Gaddis, Jean B. Bourassa, Todd A. Felix, Geraldine R. McCann, Kenneth A. Crane, and Abbe D. Ames 240 Gastrointestinal helminths of the blackneck garter snake, Thamnophis cyrtopsis (Colubridae) ...... Stephen R. Goldberg and Charles R. Bursey 243 Comparison of nitrogen isotope ratios in feathers from seven species of Colorado breeding birds ...... Sujay S. Kaushal and John J. Walsh 246 Osprey diet along the eastern side of the Gulf of California, Mexico ...... Jean-Luc E. Cartron and Manuel C. Molles, Jr. 249 Mule deer group kills coyote ...... Jack A. Wilkinson and John F. Douglass 253

Book Reviews Bird Hand Book Victor Schrager and A.S. Byatt ...... Clayton M. White 254 The World of the Hummingbird Robert Burton ...... Clayton M. White 255