Mojave National Preserve National Park Service U.S. Department of the Interior
Science Newsletter
Responses of Mule Deer to Experimental Manipulation of Water Sources: Preliminary Results from the First Year
Providing water sources for wildlife in arid regions is a method of habitat improvement that has been utilized by managers for many years. At least 122 artificial water sources, known as guzzlers, were in existence when Mojave National Preserve was created in 1994. These are basically tanks that collect runoff and store it underground. A significant population of desert bighorn sheep (Ovis canadensis nelsoni) is supported in the Old Dad and Kelso Peaks area by artificial tanks – otherwise ideal sheep habitat that lacks natural water sources.
The Mojave Desert has a long history of cattle grazing going back to the late Two mule deer does, one with collar and ear tags, photographed by a motion-triggered nineteenth century. Numerous wells, camera at Deer Spring in the Cima Dome area. troughs, and pipelines were developed to support and distribute cattle on the was utilized by constructing shallow In response, Mojave National Preserve, landscape. Near surface groundwater horizontal tunnels for gravity flow and by the University of Nevada – Reno, and pumping. From 1998 through 2002, California Department of Fish and Game grazing allotments in Mojave National initiated research with support from Safari In this Issue: Preserve were purchased, retired, and Club International to investigate the effect • Page 1. Responses of Mule Deer to donated to the National Park Service. that water provision has on mule deer Experimental Manipulation of Water Approximately 12,000 cattle were (Odocoileus hemionus) fecundity, Sources: Preliminary Results from the First Year, N. Darby, J. Dungan, K. removed from 1,260,980 formerly grazed physical condition, and survival. The Stewart, V. Bleich, D. Hughson. acres along with about 4000 feral burros study area is the northern half of the (Equis asinus). Troughs, pipelines, and Preserve from Cima Dome to Barnwell, • Page 4. Mohave Tui Chub Recovery: pumps were removed with the cattle and divided into a control area with Hybridization and Invasion, D. Hughson, N. Darby, S. Parmenter, S. many wells were abandoned. Loss of permanent natural water sources, a Henkanaththegedara, J. Seigel. these developments generated treatment area where water is added, controversy among hunting and and a dry area. Water was provided at • Page 7. Tortoise on Roads in Mojave Preserve, D. Hughson, N. Darby. environmental organizations regarding Government Holes, Petit Well, Gold King the relationship between livestock Well, Eagle Well, and Hogaboom Well in • Page 10. Effects of small mammals on watering troughs and viable populations late August to early September of 2008. post-fire vegetation recovery in the Mojave Desert, P. Stapp. of wildlife species. The experimental design calls for six
Mojave National Preserve Science Newsletter April 2009, Number 1 This Science Newsletter:
The Mojave Desert is internationally known as a place to conduct scientific research on desert ecosystems. In fact Mojave National Preserve was designated in part to "retain and enhance opportunities for scientific research in undisturbed ecosystems" as stated in the California Desert Protection Act of 1994. Significant research is done through the Sweeney Granite Mountains Desert Research Center, part of the University of California Natural Reserve System, and the Desert Studies Center, operated by the California Desert Studies Consortium of California State Universities, both of which are located in the Preserve.
The purpose of this newsletter is Mule deer and water distribution on Cima Dome in Mojave National Preserve. A yearlong 95% three-fold. First, we would like to range utilization estimate was 14,268 acres for five deer combined on Cima Dome. This range highlight some of the research being included two of five known water sources. done by university scientists in the more wells to be added at the five year vegetation at water sources in Mojave Preserve and distribute this information to park staff and midpoint of the study. Radio collared National Preserve. management. Second, this periodical mule deer will be located throughout the will allow us to inform the public and year and photographed by remote Mule deer were captured by net gun from research community about science cameras. Survival probabilities will be a helicopter in January of 2008 and 2009. being done by Preserve staff or estimated using known fate analyses Physical condition of the captured deer funded through the National Park Service. And most importantly, we implemented in Program Mark (White and was measured by ultrasound along with would like to build collaboration Burnham 1999). Information theoretic standard lengths and weight. Blood, hair, between scientists and resource model selection approaches (Burnham dental, and fecal samples were collected. managers so that scientists are made and Anderson 2002) will be used to The animals were equipped with ear tags aware of the needs of managers and assess variation in survival. and radio and Global Positioning Satellite top quality science is brought to bear (GPS) collars. The results from this first on the problems facing resource managers. Vegetation and local precipitation will be year of this study, although preliminary monitored in both the treatment and and inconclusive, have already Our intention is to publish this twice control areas. Shrub cover will be significantly increased our understanding, per year, once in the spring and again estimated within 200 m of springs and as little is known about this population. in the fall, in print at our Visitor wells in both the treatment and control Centers and electronically on the web. Articles will range from non- areas by establishing ten 50 m transects Five mule deer were tracked in the Cima technical news stories to highly within 200 m of each well and spring site. Dome control area and appeared to technical research reports. All Shrub cover will be estimated annually range along the northwest side of the material in this newsletter has been within each transect using the line dome. There are five permanent water peer-reviewed by subject matter intercept method (Bonham 1989). Cover sources on Cima Dome: Kessler Spring, experts. In each issue we will discuss and biomass of herbaceous plants by Cut Spring, White Rock Spring, Deer a resource management issue or research need along with possible forage category (Bonham 1989) will be Spring, and Valley View Ranch. The means of obtaining funding. The estimated each season by clipping range of the deer encompassed Deer resource management concern vegetation inside and outside of Spring and Valley View. Visitation to Deer highlighted in this issue is invasive moveable exclosures using a one square Spring by the five collared deer was plant species, particularly Sahara meter plot frame. These vegetation sporadic and inconsistent. Only one mustard. studies will provide basic information on collared deer visited throughout the year Debra Hughson, Science Advisor the spatial and temporal dynamics of but disappeared for periods of up to 34
2 Mojave National Preserve Science Newsletter 2009, Number 1 days. The remaining four deer visited for one to three week intervals two or three times during the year. Telemetry location information indicated that at least two deer used the water source at Valley View, which did not have a camera station installed at the time, and that all deer wandered away from Deer Spring for varying lengths of time throughout the year. When deer were consistently around Deer Spring, defined as periods with visitation gaps less than 20 days, all five deer visited on average once every 5.23 days. The highest visitation rates occurred in August with visits once every two days. Deer visited the spring primarily during night-time hours. Marked deer were only photographed at Deer Spring on Cima Dome despite having cameras A Mojave National Preserve Wildlife Biologist releases a mule deer doe freshly equipped with a GPS collar, VHF radio collar, and ear tags near Kessler Spring. deployed at springs in the areas where mule deer were collared. Photographs season, only 14 does (50%) were collecting and analyzing photographs from cameras at three other springs carrying twins. We suspect that because from the motion-activated cameras. confirmed telemetry results in that no the fall of 2008 was much drier than 2007 collared deer visited those sites. the availability of forage was lower and This study of mule deer and water the does were not as able to carry twins. sources, planned for ten years in two five- We used photographs of collared deer These data are preliminary, but the year phases, will provide Mojave National taken with the infrared-triggered cameras correlation between late summer and fall Preserve’s management with much to estimate the number of mule deer weather events and the corresponding useful information on resource utilization utilizing Deer Spring. Separate Lincoln- response of vegetation likely influences by mule deer, mule deer reproduction, Peterson Index estimates were made pregnancy and twinning rates in these visitation at springs by mule deer and during 2 week periods from February 1, mule deer. As we collect more data other animal species photographed by 2008 through January 15, 2009 with 95% during this study we will be able to the cameras, and vegetation dynamics confidence intervals (CI). We estimated examine the effects of weather variation correlated with local precipitation that 26 deer (95% CI 18-44) were using as well as provision of water on physical variability and herbivory as it effects mule Deer Spring. condition and reproduction mule deer deer reproduction. Although this study is inhabiting this portion of the Mojave just beginning, experimental manipulation In January, 2008 we captured 18 mule Desert. of water in October of 2008 did result in a deer including 15 does, 2 bucks, and 1 detectable response when a mule deer yearling female. We looked at pregnancy A graduate student from the University of buck was photographed at water placed of those females and 15 of the 16 (94%), Nevada – Reno, Cody McKee, will be at Eagle Well. including the yearling were pregnant. Of working on this research for a Master of the 15 pregnant does, 11 (73%) were Science degree in wildlife ecology. Cody References carrying twins. During 2009, we captured is supervised by Dr. Kelley Stewart, Bonham 1989. Wiley-Interscience. 31 deer, 30 does and 1 6-month old doe Assistant Professor in the Department of Burnham and Anderson 2002. Springer- and again we examined pregnancy. We Natural Resources and Environmental Verlag, New York. examined 28 does for pregnancy, not Science and Principal Investigator for this White and Burnham 1999. Program including the juvenile and 2 does for research. Cody and several technicians MARK, Bird Study 46:120. which we were unable to collect data on will spend most of the summer from May reproduction. Of the 28 does 26 (93%) through August living in the Preserve, were pregnant, but during the 2009 monitoring vegetation, tracking the capture, which reflects the 2008 growing movements of the collared animals, and
3 Mojave National Preserve Science Newsletter 2009, Number 1 Mohave Tui Chub: Hybridization and Invasion
The Mohave tui chub is the only animal genetically pure AC in the Mojave River thus less likely to be swept downstream species in Mojave National Preserve and genetically pure MTC in the three to the Soda Lake terminal playa. In the federally listed as endangered under the extant populations (Zzyzx, Naval Air flood-prone Mojave River, Mohave tui Endangered Species Act. It was found Weapons Station at China Lake, and chubs were frequently washed surviving in two small ponds at Zzyzx, Camp Cady). Thus the viability of downstream to expire on the exposed California, on what is now the Preserve, Mohave tui chub – arroyo chub hybrids is terminal playa. A combination of low after being extirpated from the Mojave a research question that needs to be fertility of hybrids and greater tolerance of River in the early to mid-twentieth addressed prior to reestablishment of high temperatures and higher flow century, reportedly through hybridization MTC in the Mojave River basin, or velocities may explain the presence of with the non-native arroyo chub. establishment of recovery populations in pure arroyo chubs in the Mojave River Transplants from this relict population off stream waters that potentially could drainage. This project seeks to address were used to establish populations at be invaded by arroyo chubs. the issue of hybridization and competition China Lake near Ridgecrest and Camp between Mohave tui chub and arroyo Cady near Newberry Springs. The Abiotic factors may have contributed to chub in an aquarium and an artificial National Park Service is working with extirpation of MTC from the Mojave pond environment. A better the U.S. Fish and Wildlife Service and River. Castleberry and Cech (1986) understanding of the viability of hybrids California Department of Fish and Game argue that physiological responses to may be useful in selecting suitable to upgrade the status of this species to high temperatures, rapidly fluctuating habitats in the upper Mojave River threatened. To accomplish this we need temperatures, and high flow velocities drainage for recovery of the species. to establish three more populations and give AC a competitive advantage over maintain six geographically distinct MTC in predator avoidance, feeding, and The original objectives of this study populations each at a minimum level of escaping stressful environmental proposal were to address the following 500 adult fish. In this article we describe conditions. Mohave tui chub, on the other two research hypotheses. our work on experimental hybridization hand, may have a competitive advantage 1. Are Siphateles bicolor mohavensis and surveys of the Mojave River. in low temperature and/or hypoxic X Gila orcutti hybrids viable? conditions that typically occur in stratified 2. Will Siphateles bicolor mohavensis The Mohave tui chub (Siphateles bicolor lake environments, particularly during a out-compete Gila orcutti in a still mohavensis), (Pisces: Cyprinidae), was winter freeze. Arroyo chubs are better pond subject to seasonal conditions listed as endangered October 13, 1970. swimmers than Mohave tui chubs and of low temperature hypoxia? The Recovery Plan was published September 17, 1984. At the time of its listing, the extirpation of Mohave tui chub (MTC) from the Mojave River was thought to have resulted from hybridization with arroyo chub (AC) (Gila orcutti). The Recovery Plan states, “[t]he exotic species of Gila invaded the Mohave River and subsequently hybridized with the Mohave tui chub. By 1970 genetically pure Mohave tui chubs had been eliminated from the river by hybridization and subsequent introgression.” This statement was based on work (Hubbs and Miller 1943, Miller 1969) that assessed hybridization through morphometric comparisons such as scale counts, scale morphology, fin ray counts, and body proportions. More recent work (Chen et al. 2006) found Agency staff survey the Mojave River by electroshocking at the Upper Narrows.
4 Mojave National Preserve Science Newsletter 2009, Number 1 We started this experiment with 10 Fork which ties in to the State Water of the first Afton Canyon crossing was laboratory spawned MTC, which had Project. also sampled on May 13. A riparian area been released into the Fire Pond at lies east of Afton Canyon road, south of Camp Cady on March 30, 2007 and This invasion of non-native fish led us to the turnoff to the Campground. About 0.5 which we obtained November 18, 2007 organize an interagency collecting km southwest of the campground is a when California Department of Fish and expedition in 2008 to update our small, shallow swamp/slough near the Game was in the process of transferring understanding of fish community main riverbed. Schools of very small fish MTC from the Fire Pond to Bud’s Pond, composition in the Mojave River. Voucher were noticed in the shallows and and with 10 presumptive AC which we specimens from this collection are collected with dipnets. No larger juveniles collected from the Mojave River at Afton permanently archived at the Natural or adults were observed. Fifty four small Canyon between December 4 and 11, History Museum of Los Angeles County. fish were collected, identified and 2007. We collected seven fish species from the vouchered at Los Angeles County Mojave River above the Upper Narrows Museum, all identified as arroyo chub. We attempted to spawn AC by near the Lewis Center for Educational manipulating photoperiod and water Research on May 12, 2008. They were: Three localities along Deep Creek, temperature in the aquarium but saw no mosquitofish (Gambusia affinis), Devil's Hole, Warm Springs, and above sign of spawning. These fish collected threespine stickleback (Gasterosteus Mojave Forks Dam, were qualitatively from Afton Canyon also appeared to be aculeatus), brown bullhead, and green electrofished and trapped on September infected with anchorworm (Lernaea sunfish (Lepomis cyanellus). Fifty four 22-23, 2008 to evaluate species cyprinacea). Upon close inspection, eight cyprinids were collected, vouchered and composition. Identification in the field of the ten fish collected in Afton Canyon identified on the basis of fin–ray, gill raker indicated rainbow trout (Oncorhynchus were found to be non-native hitch and pharyngeal tooth counts as 27 arroyo mykiss), arroyo chub, and threespine (Lavinia exilicauda), a minnow from chub, 21 hitch, and 6 hitch/arroyo chub stickleback at Devil's Hole in addition to Central California, and only two were hybrids (Jeffrey A. Seigel, personal the Eurasian snail, Radix auriculata. This arroyo chub. communication). These numbers may be mollusk is not a new or unusual find in meaningful in that they show the relative California. There were 384 specimens, Prior to this discovery in December, abundance of each minnow at that identified as arroyo chub, collected from 2007, nearly all fish collected from Afton locality. This at least provides us with a Devil’s Hole. Arroyo chub (279 Canyon had been reported as either baseline. It may be worthwhile to collect specimens collected) and black bullhead Mohave tui chub or arroyo chub or their there in the future to see how these were found at Warm Springs. Above the hybrids, although Timothy Brown did numbers change. Certainly hitch is now a Forks Dam black bullhead, hitch, green report 7 black bullhead (Ameiurus melas) dominant part of the river’s fish fauna; a sunfish, black bass (Micropterus in 1978. Twenty seven fish collected on new factor influencing any strategy to salmoides), arroyo chub, and prickly May 22, 2002 were all identified as arroyo reintroduce MTC. Twenty each of sculpin (Cottus asper) were found. chub with the possible exception of one mosquitofish, threespine stickleback, and (Jeffrey A. Seigel, Natural History hitch were sent to the California Nevada The apparent absence of hitch and/or Museum of Los Angeles County, Fish Health Center. The exotic parasite hybrids from the upper two sites suggests personal communication). Of another 72 Asian tapeworm (Bothripcephalus that hydraulic features between Warm collected from Afton Canyon on April 6-8, acheilognathi) was found in 30% of hitch Springs and the Forks may provide an 2005, 71 were identified as arroyo chub (N=6) and Schistocephalus spp. was effective natural barrier to upstream fish and one was identified as hitch. found in 10% of threespine stickleback migration. One possible barrier is a large (N=2). cataract about 1.7 miles upstream from Henkanaththegedara et al. (2004) noted Mojave Forks dam (Matthew Huffine, that brown bullhead (Ameiurus The minnow population in Afton Canyon Lewis Center for Educational Research, nebulosus) was the first reported non- has seemingly undergone a rapid change personal communication). The failure to native fish species in the Mojave River in the last 6 years. Of 107 fish collected collect trout and stickleback at the lower drainage. By 2002, 23 non-native fishes and vouchered from Afton Canyon on two sites may result from unsuitable were reported in the drainage (Marchetti May 13, 2008 there were 69 hitch, 18 water temperatures in mid-summer. The et al. 2004, Moyle 2002). Non-native arroyo chubs, and 20 hitch/arroyo chub numbers seem to indicate a shift in the species richness appears to be highest at hybrids. Additionally, a small slough predominate minnow in Deep Creek. The the Silverwood reservoir on the West south of the Campground and southwest upper Devil’s Hole and Warm Springs
5 Mojave National Preserve Science Newsletter 2009, Number 1 area is dominated by arroyo chub (663 Date Number Location Species Natural History Museum of Los Angeles County arroyo chub, no hitch) and the lower May 22, 2002 27 Afton Canyon Gila orcutti LACM 55990-1 Fork’s Dam area is dominated by hitch (1 April 6-8, 71 Afton Canyon Gila orcutti LACM 56881-1 arroyo chub, 37 hitch). It seems unlikely 2005 April 6-8, 1 Afton Canyon Lavinia exilicauda LACM 56881-2 that these numbers are an artifact of 2005 seasonality or collecting bias. May 12, 2008 27 Upper Narrows Gila orcutti LACM 56833-3 (10 specimens) LACM 56834-3 (11 specimens) LACM 56838-2 (6 specimens) Non-native fishes appear to be entering May 12, 2008 21 Upper Narrows Lavinia exilicauda LACM 56833-5, 4 specimens; the Mojave River drainage via the LACM 56834-6, 11 specimens; LACM 56838-5, 6 specimens California Aqueduct through Silverwood May 12, 2008 6 Upper Narrows Gila orcutti / LACM 56833-4, 1 specimen; Reservoir. Non-native fish species Lavinia exilicauda LACM 56834-5, 3 specimens; richness appears to be highest in Deep hybrids LACM 56838-4, 2 specimens May 13, 2008 69 Afton Canyon, Lavinia exilicauda LACM 56880-3 Creek near the Forks Dam, and in the ford Mojave River at the Upper Narrows. Hitch May 13, 2008 18 Afton Canyon, Gila orcutti LACM 56880-1 ford appears to be restricted in Deep Creek to May 13, 2008 20 Afton Canyon, Gila orcutti / LACM 56880-2 the lower reaches directly above the ford Lavinia exilicauda Forks Dam while arroyo chub becomes hybrids May 13, 2008 54 Afton Canyon, Gila orcutti LACM 56893-1 more common upstream. Black bullhead S of was found in the lower and middle campground September 22- 384 Devil's Hole Gila orcutti LACM 56868-1 reaches of Deep Creek whereas trout 23, 2008 was found only in the uppermost reach. September 22- 279 Warm Springs Gila orcutti LACM 56867-2 Arroyo chub and stickleback have been 23, 2008 September 22- 37 Above the Lavinia exilicauda (LACM 56864-3, 23 specimens; introduced in the upper watershed along 23, 2008 Forks Dam LACM 56865-3, 14 specimens) with trout (Gar Abbas, U.S. Forest September 22- 1 Above the Gila orcutti LACM 56864-2, 1 specimen Service, personal communication). Hitch 23, 2008 Forks Dam appears to be expanding in the Mojave River and hybridizing with and replacing arroyo chub. In order to better frame the recovery potential of Mohave tui chub, species composition above and below potential migration barriers in Deep Creek, the upstream extent of arroyo chub in the perennial headwaters, and the species composition in all Mojave basin reservoirs should be determined.
References Castleberry and Cech 1986. Ecology 67(4):912. Chen et al. 2006. 38th Meeting, Desert Fishes Council. Henkanaththegedara et al. 2008. 40th Meeting, Desert Fishes Council. Hubbs and Miller 1943. Michigan Acad. Sci., Arts, Lett. 28:343. Miller 1969. Cal-Nevada Wildl. Trans 1969:107. Marchetti et al. 2004. Ecological Mojave River watershed showing survey locations and Mohave tui chub sites mentioned in Applications 14(5):1507. the text. Moyle 2002. University of California Press.
6 Mojave National Preserve Science Newsletter 2009, Number 1 Tortoise on Roads in Mojave Preserve
The Desert Tortoise (Gopherus depend on where staff tends to go during agassizii) is federally listed as daily routines, but do give some threatened under the Endangered indication of where tortoises are seen. Species Act throughout its range in the Our observations as well as contours of Mojave Desert west and north of the the line distance data are shown in Colorado River. The tortoise is active in Figure 2. Mojave National Preserve and above ground from mid-March to June with a Observations by Preserve staff tend to secondary period of activity from confirm the area near the intersection of September to October. Tortoises are Ivanpah and Morningstar Mine roads as attracted to paved roads following having a concentration of tortoises and Desert tortoises are attracted to highways rainstorms by puddles of water where during rains by puddles of water. also provide supporting evidence for the they are susceptible to being killed by cluster along Essex Road. But our automobiles. Some of the heavily squares (bowties) to large polygons. This observations also indicate more tortoises traveled roads in the Preserve pass was done in an attempt to compensate in the area along Kelbaker Road than through high quality tortoise habitat. A for very low detection rates of tortoises. suggested by the line distance data and north - south stretch of roads is a major Live tortoise observations from the line fewer tortoises in the Goffs area. In fact thoroughfare connecting the high desert distance transects are indicated in Figure the tortoise population near Goffs may cities of Joshua Tree, Yucca Valley, and 1 by green diamond symbols. Contouring have declined significantly over the past Twentynine Palms to Las Vegas, of these observations revealed four areas few years (Berry 2000). Nevada. We initiated this study to in the Preserve of apparent population identify “hotspots” where high-speed concentrations: just north of Kelso, The sections of paved roads that connect traffic passes through more densely Ivanpah Valley near the northern park Interstate 40 to Kelso then to Morningstar populated tortoise habitat and to see if boundary, along Essex Road, and near Mine Road, Ivanpah Road, Nipton Road, warning signs might increase driver Goffs. and finally to Interstate 15 form a main awareness. traffic corridor between Las Vegas and Preserve staff conducted driving and populated areas of Southern California. We used line distance sampling data walking road surveys and collected We collected observations to compare collected by the U.S. Fish and Wildlife anecdotal observations simultaneously traffic intensity and driver behavior on this Service (USFWS) in Mojave National with observations of motorist response to heavily traveled thoroughfare, at Preserve for the years 2001, 2002, 2004, a model tortoise placed along the road Morningstar Mine Road, with the more and 2005 to identify potential edge. Preserve maintenance staff lightly traveled Essex Road that takes concentrations of tortoise. We illustrated conducted driving transects once per visitors to Mitchell Caverns and to the areas of high live tortoise observations week on all 164 miles of paved roads in Hole in the Wall Visitor Center. While using the kernel density function in the Preserve for the period from April 1 to most of the traffic on Morningstar Mine ArcMap 9.2 (Figure 1). Note that although June 30 and from September 1 to Road goes through the park from one this produces units of live tortoises per October 31, 2008 during the tortoise metropolitan area to another, most of the area it should not be mistaken for actual active season. Desert Studies Center traffic on Essex Road goes to the park tortoise densities. Density of tortoises is herpetologist Jason Wallace walked itself. calculated using the Line Distance sections of selected road segments in methodology for data aggregated by Ivanpah and Fenner Valleys surveying for We gauged motorist response to Recovery Unit and is available through tortoise and for all signs of road kill. tortoises on paved roads by placing a the USFWS Desert Tortoise Recovery Mojave National Preserve staff has medium sized model on the white line Office. reported anecdotal observations of desert along the edge of the pavement in the tortoise since 2001 but a significantly northbound lane. We observed motorist The polygons in Figure 1 show transects increased effort in 2008 resulted in 255 response to the model tortoise at a point conducted by contractors for USFWS. reports, 80 of which had GPS coordinates midway on Morningstar Mine Road and The sampling design changed over the and could be plotted. These are not on Essex Road. We are confident based years from small squares to paired scientific since they are opportunistic and on these observations that the model is
7 Mojave National Preserve Science Newsletter 2009, Number 1 difficult to distinguish from a real tortoise even up close. We collected 110 observations on Morningstar Mine Road May 5, 2008 and 104 observations on October 11, 2008. Given the fewer cars on Essex Road it took several days to collect a comparable number of observations. We collected 22 observations on May 21, 19 observations on May 22, 29 observations on May 24, 34 observations on May 26, and 13 observations on May 29, 2008. In the fall active season we collected 60 observations on October 12 and 49 observations on October 13, 2008.
These observations are count data of individual automobiles passing the model tortoise placed on the edge of the pavement. If several cars passed in a group only one observation (the lead car) was counted. Motorists that deviated from Figure 1. Live tortoise observations taken from U.S. Fish and Wildlife Service Line Distance normal travel either by obvious slowing Sampling data through 2005. that was noticed by both of two concealed observers, brake lights coming on, or most commonly, stopping to investigate were counted as reacting to the model tortoise. Several times we observed vehicles pass by without slowing only to turn around and return to investigate. We attached an informational brochure to the bottom of the model and overheard many comments about the issue and our study. Not counting park vehicles, Union Pacific vehicles, or motorcycles, our observations are summarized in the following contingency table.
Contingency Morningstar Essex Table Mine Road Road Reacted to 7 34 the model Did not 188 172 react to the model Figure 2. Park observations of desert tortoise in 2008 with contours from U.S. Fish and Wildlife Service Line Distance Sampling data. The difference in motorist behavior between Morningstar Mine Road and Essex Road was significant (χ2 = 16.9, P = 0.00004).
8 Mojave National Preserve Science Newsletter 2009, Number 1 The purpose of conducting the model tortoise observations is to test the hypothesis that road signs have an effect Cars vs. Speed FILENAME:
30 We are also interested in seeing if tortoise populations are depressed Cars 25 adjacent to paved roads and conducted 20 transects parallel to Morningstar Mine 15 Road and Essex Road this spring using the methodology of Boarman and Sazaki 10 (2006). 5
0 References 10 20 30 40 50 60 70 80 90 100 110 120 Berry 2000. U.S. Geological Survey, Speed Western Ecological Research Center, Riverside, California. Figure 3. Comparison of traffic speeds on Morningstar Mine Road and Essex Road starting Sunday morning, October 12, 2008 and ending Friday, October 17, 2008. Boarman and Sazaki 2006. Journal of Arid Environments 65:94.
9 Mojave National Preserve Science Newsletter 2009, Number 1 Effects of small mammals on post-fire vegetation recovery in the Mojave Desert
The frequency and intensity of fires is of Mojave National Preserve. In communities following catastrophic fire. increasing in the Mojave Desert driven particular, there are concerns that fire will by invasive annual plants, precipitation facilitate continued invasion by exotic I studied the effects of fire-mediated variability, and human-caused ignitions. plants (Lovich and Bainbridge 1999). changes in habitat and resources on the Meanwhile the role that rodents play in Herbivory by native ungulates and rabbits abundance and foraging activity of small vegetation dynamics is becoming and granivory by nocturnal rodents may mammals, and on their possible effects increasingly appreciated. The winter of also significantly influence the re-growth on post-fire vegetation recovery of 2004 and 2005 was the wettest on of vegetation and, potentially, the spread vegetation in sagebrush-juniper record in the Mojave Desert, followed by of exotic annual grasses such as Bromus vegetation in Mojave National Preserve. a dry summer and strong monsoon. rubens and Schismus spp., which out- The specific objectives of the project Lightning strikes in late June sparked a compete native plants and are the most were to determine: wildfire in Mid Hills and Hackberry important sources of fuels (Brooks 2002). Mountain that burned over 70,000 acres Compared to other arid and semiarid • relative abundance and community of Mojave National Preserve. Western systems (e.g., Schwilk and Keeley 1998, attributes of small mammals in National Parks Association funded Dr. Olson et al. 2003, Letnic et al. 2005), burned and unburned areas of Paul Stapp, Associate Professor in the however, the effects of fire on small sagebrush-juniper vegetation, and Department of Biological Science at mammal communities and the role of examine how these change across California State University – Fullerton, to small mammals in post-fire vegetation the transition (edge) between study the relationships between small recovery in the Mojave Desert are poorly burned and unburned vegetation; mammals and post-fire vegetation understood. For example, selective recovery. granivory can alter local plant • the pattern of seed removal by small communities and the trajectory of desert mammals across the burn edge to Although arid and semiarid ecosystems ecosystems (Brown and Heske 1990). estimate potential impact of are usually considered too unproductive Although rodent species are differentially granivores on the seed bank; to support high fuel loads, episodic affected by fire (McGee 1982, Schwilk wildfires can still have catastrophic and Keeley 1998, Esque et al. 2003), • the effects of small and large effects (Westerling et al. 2003, Brooks et individuals that survive the immediate herbivores and granivores on al. 2004, Letnic et al. 2005). In the effects of fire may accumulate in remnant vegetation recovery in burned areas Mojave Desert, large amounts of rainfall unburned patches, resulting in high compared to the burn edge. in the winter of 2004-05 resulted in population densities and a greater impact unusually high plant growth, especially at on local plants and seed resources. In Field work on the project began in May higher elevations. On 22 June 2005, dry theory, the high densities of rodents may 2006, 11 months after the Hackberry lightening strikes ignited a series of large also impose high foraging pressure on Complex Fire, and, except for the fires in the central Mojave National plants and seeds in the adjacent burned ongoing experimental studies of Preserve (MNP). When these fires, areas, which, depending on the source of vegetation recovery, was completed by collectively known as the Hackberry new seeds and the relative preference of August 2007. Most of the field work was Complex Fire (HCF), were finally rodents for exotic vs. native species, may conducted by a former California State contained on 27 June, some 31,580 ha result in very different plant communities University – Fullerton Master’s student had burned, mostly in mixed sagebrush near the edge than farther into the burn. (C. Moore). (Artemesia)-juniper (Juniperus) and other Understanding the effects of fire on woodland and scrub plant communities. species composition and relative The research site was in the Mid Hills abundance of small mammals, and the region of Mojave National Preserve, The Hackberry Complex Fire was the spatial pattern of foraging behavior and California (35º9' N, 115º23' W), at largest recorded for the eastern Mojave seed removal, may have important elevations between 1500-1600 m. Mean Desert, and is expected to have long- implications for the use of fire as a tool to annual precipitation at the nearby Mid lasting effects on the character, control exotic plant invasions and for the Hills Campground is 264 mm, with a management and ecology of this portion re-establishment of native desert plant mean winter daily minimum and summer
10 Mojave National Preserve Science Newsletter 2009, Number 1 daily maximum of 3.2°C and 34.3°C, respectively. Most precipitation falls during the winter months (November– March), although there are often summer thunderstorms as well. The dominant vegetation is big sagebrush (Artemisia tridentata) and Utah juniper (Juniperus osteosperma). Other abundant perennials include desert bitterbrush (Purshia tridentata), Nevada ephedra (Ephedra nevadensis), banana yucca (Yucca baccata), threadleaf snakeweed (Gutierrezia microcephala), and turpentinebroom (Thamnosma montana), as well as a number of forbs and grasses. Abundant annuals include western Figure 1. Layout of 1.60-ha plots in sagebrush-juniper vegetation in high-elevation Mojave tansymustard (Descuraina pinnata), Desert. The small black circles represent trap stations spaced 20 m apart (80 x 200 m grid), for a total of 55 Sherman live traps per plot. The larger gray circles indicate locations of birdnest buckwheat (Eriogonum seed trays placed 80, 40, 20, and 10 m into the burn, at the burn-unburned edge, and 20 and nidularium), storksbill (Erodium 80 m into the unburned vegetation.. cicutarium), Fremont’s phacelia (Phacelia on each plot from June–August 2006 and parallel transects, 20 m apart and fremontii), bristly fiddleneck (Amsinckia May–August 2007, for a total of 3,520 perpendicular to the burn edge, on each tessellata), desert globemallow trap-nights. On each plot, trapping of the 4 large study plots. Trays were (Sphaeralcea ambigua), desert allysum sessions were conducted during full and placed at 80, 40, 20 and 10 m into the (Lepidium fremontii), rose heath new moon periods to account for burn, at the edge of the burn (0 m), and (Chaetopappa ericoides), and cushion differences in rodent activity associated at 20 and 80 m into unburned sagebrush cryptantha (Cryptantha circumscissa). with moonlight. I used the number of vegetation (Figure 1), for a total of 21 unique individuals captured in each trays per plot. Seed trays were plastic In May 2006, four study plots were location (burned, 40-100 m, trap rows 1- plant saucers (36 cm diameter, 4 cm established at edge of the Hackberry 4; edge, 0, +20 m, traps 5-7; burned, 40- deep) filled with 2 L of pre-sifted sand. Complex Fire. Plots were 1.60 ha (80 x 100 m, traps 8-11) and plot during each Trays were set at dusk by mixing 4.00 g 200 m), with the long axis perpendicular trapping session, based on their location of irradiated millet into the sand, and to and centered on the edge of the burn. of first capture, as an estimate of relative smoothing the surface so that foraging Plots were at least 100 m away from abundance. In addition, four smaller grids activity could be detected. Trays were each other and roads. Trapping grids (5 x (5 x 4 stations at 20-m intervals) were collected at dawn and the sand was 11; 55 stations) were established on each established in an area of Joshua tree sieved to collect the remaining seed, plot to estimate relative abundance and woodland that also burned during the which was weighed to estimate the diversity of nocturnal rodents. One extra- Hackberry Complex Fire. Two plots were amount of seed removed. On each plot, long Sherman live-trap was placed at in burned vegetation and two were in trays were set for 2 consecutive nights each station, which were 20 m apart unburned vegetation, with plots during full and new moon periods in June (Figure 1). Traps were baited with approximately 100 m apart. These grids and July 2006. The difference between irradiated birdseed at dusk and checked were trapped twice between June and the first and second night reflect the time at dawn. Trapping sessions were 4 August 2006, with the goal of comparing needed for rodents to discover seed consecutive nights. Each individual rodent communities in burned and trays, which would be expected to be captured was marked with a uniquely unburned areas of Joshua tree longer on Night 1, i.e. fewer trays would numbered aluminum ear tag and woodlands to that in sagebrush be discovered, especially if moonlight released. All animal handling procedures vegetation. reduced aboveground foraging activity. were approved by the Institutional Animal Care and Use Committee at California To determine spatial patterns of seed To determine the effects of herbivores State University – Fullerton. removal across the sagebrush burn and granivores on vegetation recovery, in transition, foraging activity was measured March 2007 exclosures were established Two trapping sessions were conducted in artificial seed trays placed in three on 2 of the sagebrush study plots, as well
11 Mojave National Preserve Science Newsletter 2009, Number 1 as one new plot that crossed the burn locations of the Hackberry Complex analyses. Because sample size (n=5 per boundary of the Wildhorse Fire (WHF), Fire plots, and 5 exclosures of each type type) and therefore, power, was low for which burned sagebrush-juniper at the edge and burned locations on the Wildhorse Fire plots, I used α = 0.10 to vegetation in July 2002 in the same area. Wildhorse Fire plot. evaluate whether results were statistically Exclosures were constructed of 1.3 cm significant. Here, I focus on percent cover mesh hardware cloth and were 60 x 60 Vegetation in exclosures was measured and species richness as response cm (0.36 m2) and 30 cm high. The beginning in May 2007 and has continued variables. exclosure walls were buried 10 cm into intermittently since. Percent canopy the ground. Three different types of cover, number of individual plants and Heteromyid rodents were the most exclosures were used. One (N) had solid species richness was measured in a 40 x abundant species captured in sagebrush walls and a roof to exclude all herbivores, 40 cm (0.16 m2) quadrat in the center of areas, with D. panamintinus (DIPA) and including rodents, rabbits, deer and each exclosure. The effects of small and D. merriami (DIME) representing 65% cattle. Another (R) was similar to N but large herbivores was estimated by and 12% of the 428 individuals captured. had two 8 x 8 cm holes on each wall, comparing plant percent cover, plant Three species of Peromyscus which excluded larger herbivores but density and species richness between (maniculatus, crinitus, truei), here allowed access to large rodents exclosures and controls that were considered collectively (PESP), were the (Dipodomys panamintinus, Neotoma accessible to all herbivores. The most abundant cricetids (14% of lepida). The third type (C) was a control difference between N and C exclosures individuals), followed by N. lepida (6%). that was similar to R but without a roof, to reflected the effects of all herbivores on Perognathus longimembris (PELO) and permit access to all herbivores while vegetation; the difference between R and Onychomys torridus (ONTO) made up mimicking any effects of the exclosure C exclosures was used to estimate the the remaining 3% of individuals. The walls. effects of large herbivores only; and the rodent community in Joshua tree difference between N and R exclosures woodlands was similar, except that D. Exclosures were placed at the burn edge estimated the effects of rodents. To merriami was more abundant than D. and 100 m into the burn on each plot. account for local effects, differences were panamintinus, and that a single Five stations, each 20 m apart, were calculated for each set of exclosures Chaetodipus, identified in the field as C. established at each location. Each station (station) during each sampling period; penicillatus, was also captured. No had one of each type of exclosure, set at statistical analyses were conducted on voucher specimens were prepared. random locations approximately 2 m mean differences. Data from May 2007, By all three community metrics examined apart. Therefore, there were a total of 10 only ~1 month after exclosures were in (N0, N1, N2), rodent diversity was highest exclosures of each type at the edge and place, were excluded from statistical in unburned vegetation, intermediate at
Table 1. Mean abundance (N) and community diversity indices (species richness N0; Hill’s N1; Hill’s N2) on 4 1.60–ha study plots that spanned the transition from burned to unburned sagebrush in 2006 and 2007. For comparison, results from trapping in burned and unburned Joshua tree woodlands in 2006 are also included.
Vegetation/ year/variable Burn (40-100 m) Edge (+20 m) Unburned (40-100 m) Sagebrush (n=4 plots) 2006 N 10.25 (1.80) 7.75 (0.48) 10.75 (1.80) N0 3.00 (0.41) 3.00 (0.00) 4.25 (0.48) N1 1.91 (0.35) 2.63 (0.15) 3.49 (0.37) N2 1.60 (0.25) 2.47 (0.19) 3.12 (0.31) 2007 N 12.25 (0.48) 6.25 (0.75) 8.00 (0.82) N0 2.00 (0.71) 2.25 (0.25) 3.75 (0.25) N1 1.38 (0.22) 1.65 (0.19) 3.31 (0.28) N2 1.23 (0.13) 1.45 (0.16) 3.06 (0.29) Joshua tree (n=2 plots) 2006 N 25.50 (1.50) - 13.50 (0.50) N0 4.50 (0.50) - 4.00 (0.00) N1 2.38 (0.47) - 3.23 (0.15) N2 1.90 (0.38) - 2.89 (0.11)
12 Mojave National Preserve Science Newsletter 2009, Number 1 structure differed among vegetation types. Rodent diversity in unburned sagebrush was similar to that in unburned Joshua tree woodland (Table 1), but burned areas of Joshua tree woodlands retained higher rodent diversity than in burned sagebrush.
In both years, DIPA was the only species captured in high numbers in burned sagebrush areas (Figure 3), and their abundance generally declined toward the edge of the burn and in unburned vegetation, especially in 2007. DIME was most abundant at the edge and in unburned vegetation (Figure 3). Cricetids were more abundant and widespread in Figure 2. MDS plot showing the relationships between rodent communities in burned, edge 2006 than in 2007, and tended to be and unburned sagebrush in 2006 and 2007. Each point represents a different plot (a-d), restricted to unburned areas. taking the mean of numbers of individuals caught during trapping sessions within a year. Rates of seed removal were usually the edge, and lowest in the burn (Table types (P<0.020). Graphically, a non- highest on dark, new-moon nights than 1). For all three indices, diversity was metric multidimensional scaling (MDS) on full-moon nights, a pattern that was significantly higher in unburned plot, followed by an analysis-of-similarity particularly evident in the more open, vegetation than at the edge and burned (ANOSIM) in PRIMER (Clarke and Gorley burned vegetation (Figure 4). On full- areas (ANOVA, P<0.003), which were not 2006), indicated significant clustering of moon nights, the amount of seed significantly different from each other. rodent communities both by year removed was lowest in the burn, and Rodent diversity was significantly higher (P=0.011) and vegetation type (P=0.001; increased toward the burn edge. In in 2006 than 2007 for all three vegetation Figure 2), revealing that community contrast, on new-moon nights, rodents, presumably DIPA, consumed large amounts of seed in the burn (Figure 4). 7 7 DIPA DIPA 6 DIME 6 DIME On the first night of trials, significantly PELO 5 5 more seed was consumed on new- than
4 4 full-moon nights (ANOVA, P = 0.021), but 3 3 there was no significant difference 2 2 between distance categories (P = 0.145). Number of individuals Num ber of individuals 1 1 On the second night, when most trays 0 0 100 80 60 40 20 0 -20 -40 -60 -80 -100 100 80 60 40 20 0 -20 -40 -60 -80 -100 would have been discovered by rodents,
Distance from burn edge (m) Distance from burn edge (m) more seeds were eaten on new moon nights (P<0.001) and at the burn edge 7 7 PESP PESP (P=0.002). On Night 2, a significant 6 ONTO 6 NELE NELE interaction between moonlight and 5 5
4 4 distance (P<0.001) also indicated
3 3 significant differences between distances 2 2 on full-moon nights (Figure 4), but no Number of individuals Num ber of individuals 1 1 distance effects on new moon nights 0 0 100 80 60 40 20 0 -20 -40 -60 -80 -100 100 80 60 40 20 0 -20 -40 -60 -80 -100 (P>0.05).
Distance from burn edge (m) Distance from burn edge (m) At the Hackberry Complex Fire burn Figure 3. Relative abundance (mean + 1 SE) of rodents with distance along a transition from sites, the consistently positive differences burned (positive values) to unburned vegetation (negative values) sagebrush vegetation in 2006 (left panels) and 2007 (right panels). between plant cover in N and C exclosures (Figure 5, top) indicated that
13 Mojave National Preserve Science Newsletter 2009, Number 1
All herbivores 4.0 Night 1 in the burn, but decreased it at the edge new moon 25 N-C dif f Burn full moon N-C dif f Edge 3.5 20 (P=0.012; Figure 6, bottom right).
3.0 15 2.5 10 Because field work on this project began 2.0 5 in May 2006, some 11 months after the percent coverpercent 1.5 0
seeds eaten (g) eaten seeds Hackberry Complex Fire, it is not possible 1.0 -5 to know how much mortality of small 0.5 -1 0 May07 Aug07 Sep07 Oct07 Nov07 Dec07 Mar08 Jul08 0.0 mammals was caused directly by the fire. 80 40 20 10 0 -20 -80 Session Nonetheless, despite the near absence of Distance from burn edge (m) Large herbivores only perennial vegetation, rodents inhabited 25 R-C dif f Burn Night 2 R-C dif f Edge the burned sagebrush areas in relatively 20 4.0 15 high numbers. Dipodomys panamintinus 3.5 10 was the only species captured 3.0 5 consistently far into the burn, although it
2.5 coverpercent 0 2.0 was also present in lower numbers at the -5 1.5 edge and in unburned areas. Assuming -1 0 seeds eaten (g) eaten seeds 1.0 May07 Aug07 Sep07 Oct07 Nov07 Dec07 Mar08 Jul08 that, prior to the fire, rodent communities 0.5 new moon Session full moon in the burn were similar to those in the 0.0 80 40 20 10 0 -20 -80 unburned vegetation, cricetids and Figure 5. Effects of excluding all herbivores Distance from burn edge (m) (top) and large herbivores only (bottom) on notably, N. lepida, likely suffered most percent cover of plants 100 m into the 2005 from the fire. This was particularly true in Figure 4. Amounts of seed removed (mean + HCF and at the fire’s edge. Effects were 1 SE) in 2006 from artificial seed trays with calculated as the difference between N and C 2007, when populations of most species distance from the edge of the HCF burn in exclosures, and R and C exclosures, and rodent diversity were much lower sagebrush vegetation. Trials were run for 2 respectively. Values are means + 1 SE. consecutive nights in each season on 4 study than in 2006. The difference between plots, with a total of n=9-12 trays per distance years may be due to worsening resource class for each moon phase. between plants. In contrast, rodents had conditions during the 2 years after the little effect on plant species richness in fire, or may simply reflect inter-annual herbivores collectively significantly either the burn or at the Hackberry variation in resource availability and reduced plant cover by approximately 7% Complex Fire edge (N-R, P=0.450; therefore rodent productivity. during the study period (paired t-tests, N- Figure 6, top right). C; P<0.05). The effects were similar in By all community indices examined, the the burn and at the edge (P=0.947; Herbivores, and especially rodents, had Hackberry Complex Fire significantly Figure 5). In contrast, excluding large markedly different effects at the older, reduced rodent diversity compared to the herbivores only (R-C) significantly Wildhorse Fire burn site. Excluding all unburned sagebrush. Diversity was increased plant cover in the burn herbivores significantly increased plant somewhat higher at the edge than 100 m (P=0.035), but not at the edge (P=0.151; cover and species richness vs. controls in into the burn, but the edge community Figure 5, bottom), and the difference in the burn (P<0.076), but not at the edge more closely resembled that in the burn plant cover between R and C exclosures (P>0.209). Excluding only large than in the unburned vegetation, where was consistently higher in the burn herbivores had no significant effect on cricetids were dominant and diversity was (P=0.030). This could have resulted from plant cover or species richness, either in consistently highest. Dipodomys higher foraging activity of large the burn or at the edge (P>0.10). merriami, a smaller kangaroo rat that herbivores in the burn; however, the Whereas the effect of rodents on plant would be expected to be competitively effect of rodents (N-R) was also cover was greater at the edge than in the subordinate to D. panamintinus, tended significantly higher at the edge than in the burn at the Hackberry Complex Fire to be most abundant at the edge, burn (P<0.001; Figure 6, top left), which (Figure 6, top left), at the Wildhorse Fire possibly responding to the decline in likely contributed to lower cover in both R site this pattern was reversed, with the abundance of its larger congener near and C exclosures. Excluding both large exclusion of rodents resulting in an the edge, especially in 2007. and all herbivores combined reduced increase in cover in the burn, but species richness in the burn relative to decreased cover at the edge (P=0.009; Although species composition clearly the edge (P<0.034), suggesting that Figure 6, bottom left). Similarly, excluding differed between the edge and burned herbivores may mediate competition rodents increased plant species richness and unburned areas, combining all
14 Mojave National Preserve Science Newsletter 2009, Number 1
HCF 2005 HCF 2005 4 N-R diff Burn 25 N-R diff Burn N-R diff Edge N-R diff Edge 20 3 that are only manifest over longer 15 10 2 periods. 5 1 0 -5 0 In conclusion, I found that granivorous percent cover -10 species richness -15 -1 rodents continued to live in high numbers -20 -2 -25 in the burned areas of the Hackberry May07 Aug07 Sep07 Oc t07 Nov07 Dec07 Mar08 Jul08 May07 Aug07 Sep07 Oc t07 Nov07 Dec07 Mar08 Jul08 Complex Fire and that, at least on dark Session Session nights, rates of seed removal were WHF 2002 WHF 2002 4 25 N-R diff Burn N-R diff Burn comparable to those near the edge of the N-R diff Edge 20 N-R diff Edge 3 unburned vegetation. This suggests that, 15 10 2 for restoration efforts, planting seeds far 5 0 1 into a burn does not necessarily make -5 0 percent cover them safer from granivores. Combining -10 species richness -15 -1 across nights, however, the higher rates -20 -2 -25 of seed removal at the edge, combined May07 Aug07 Sep07 Oc t07 Nov07 Dec07 Mar08 Jul08 May07 Aug07 Sep07 Oc t07 Nov07 Dec07 Mar08 Jul08
Session Session with the diversity of small mammals in remnant unburned patches, implies that Figure 6. The effect of small rodents on percent cover (left panels) and plant species richness (right panels) in the burn and at the burn edge in the sagebrush vegetation affected by the seeds and seedlings planted near the Hackberry Complex Fire (HCF) in 2005 and the Wildhorse Fire (WHF) 3 years earlier. The edge of a fire may be at greater risk of effects of rodents were calculated by subtracting values from wire cages excluding large herbivores (R) from those excluding all herbivores (N). Values are means + 1 SE. consumption or caching. Ultimately, rodents may have a facilitative effect on plant recovery, as suggested by results species, total rodent abundance varied The extent and mechanisms by which from the Wildhorse Fire exclosure little across the study plots. Presumably small mammals influence vegetation experiments, though it is not known if as a result of the high numbers of D. recovery following fires in high-elevation these effects benefit native or exotic plant panamintinus in the burn, there was sagebrush habitats are not yet clear, and species more. Unfortunately, data on the surprisingly little variation in seed removal this aspect of the project is ongoing. plant percent cover and density were not on dark, new-moon nights between However, results thus far from my collected separately for each species, burned, edge and unburned vegetation. exclosure experiments suggest that large information that is critical to determine if Even seeds 80 m from the unburned herbivores and granivores influence both there are differences in species refugia were discovered and consumed plant cover and species richness. composition, and especially in the on dark nights. However, indirect effects Excluding herbivores consistently led to abundance of exotic species, between of the Hackberry Complex Fire on rodent increases in plant cover in both the burn exclosures. Future sampling of plant foraging activity were apparent on full- and at the edge. The high rates of seed responses in exclosures will examine moon nights, when significantly fewer removal observed at the edge during possible differences in plant species seeds were removed in the burn. This seed tray trials may be translated into composition between exclosures, as well difference can probably be attributed to larger effects of rodent exclusion on plant as between the burn and edge. the loss of protective overhead cover as a cover near the Hackberry Complex Fire result of the fire, which could make edge than in the burn itself, though plant Acknowledgements rodents more sensitive to predation risk species richness was not different. In C. Moore collected field data in 2006 and on brightly lit nights (e.g. Price et al. addition, the effects of rodents were more 2007 and erected the exclosures. Y. 1984, Longland et al. 1991). Because pronounced on the 2002 Wildhorse Fire, Atallah and R. Fulton assisted with plant potential granivores were abundant at all suggesting that, given the low productivity identification. J. Kraft assisted with field distances from the burn edge, it seems of the Mojave Desert, it may take several work. Funding was provided by Western likely that the increasing seed removal years to observe significant changes in National Parks Association, California rates with decreasing distance into the plant responses to herbivores and State University, and the Department of burn reflects the tendency of rodents granivores. In particular, the relative Biological Science at California State living at the edge or in unburned decrease in cover and plant species University – Fullerton. vegetation to not wander far from the richness in rodent exclosures at the edge safety of the edge on full-moon nights. suggests some positive effects of rodents
15 Mojave National Preserve Science Newsletter 2009, Number 1 Resource Management Issue: Sahara Mustard
References Sahara mustard (Brassica tournefortii)
Brown and Heske 1990. Science has become a major threat to rare and endemic sand dune ecosystems in 250:1705 Mojave National Preserve following an Brooks 2002. Ecol. Appl. 12:1088 exceptionally wet spring in 2005. Sahara Brooks et al. 2004. Bioscience 54:677; mustard may increase fuel loads and fire Esque et al. 2003. SW Nat. 48:103 hazard in desert scrub (Brooks et al. Clarke and Gorley 2006. PRIMER-E, v. 2003). Dense stands appear to suppress native vegetation and wildflowers. In 6.1.5 some areas following a cool, wet spring Letnic et al. 2005. J. Mammal. 86:689 mustard can become so dense as to be Longland et al. 1991. Ecology 72:2261 difficult to walk through. Sahara mustard Lovich and Bainbridge. 1999. Environ. appears to be advancing into the Sahara mustard invasion in the sand dunes near Rasor, south of Zzyzx, Manage. 24:309 Preserve from the south and west California. Photo taken March 12, 2008. Ludgwig and Reynolds 1988. Statistical through the sand dunes of Devil’s Playground towards Kelso Dunes, one of research projects that will improve our Ecology, Wiley & Sons the most scenic geologic features in the understanding of the ecological effects of McGee. 1982. J. Range Manage 35:177 Preserve and habitat for rare and invasive species and lead to effective Olson et al. 2003. West. NA Nat. 63:50 endemic desert vegetation. management techniques. Although many Price et al. 1984. J. Mammal. 65:353 exotic species exist in the Mojave Desert, Schwilk and Keeley. 1998. SW Nat. Sahara mustard is native to semi-arid the rapid expansion of Sahara mustard and arid deserts of North Africa and the into sand dune areas with spectacular 43:480 Middle East, as well as Mediterranean native wildflower blooms has heightened Westerling et al. 2003. BAMS May:595. lands of southern Europe. It was concern. apparently uncommon in the deserts of This Science Newsletter is a publication of southern California until recently. Above Will predicted warming, drying, and Mojave National Preserve, 2701 Barstow Road, Barstow, CA 92311. Information average rainfall from 1977 through 1983 increased variability of precipitation reported here has been reviewed by subject may have caused a significant population resulting from climate change promote a matter experts. expansion and the species became shift from native annuals to exotic abundant in southeastern California annuals, such as Sahara mustard and Editor (Sanders and Minnich 2000). It is most red brome? Are there control methods Debra Hughson, Science Advisor common in wind-blown sand deposits that could take advantage of the early and disturbed sites. Because of its early seasonal development of this plant while Superintendent phenology, it is a winter annual forb that avoiding harm to native vegetation? Dennis Schramm responds to fall and winter cool season Mojave National Preserve resource rains, it appears to monopolize available managers would be interested in Contributors soil moisture as it builds canopy and sets proposals, ranging from annual Vern Bleich, California Department of Fish and Game, retired mature seed before many native species monitoring by students and field classes, Neal Darby, Mojave National Preserve have begun to flower. Dependence on or volunteer seasonal control efforts, to Jason Dungan, Mojave National Preserve the timing and magnitude of winter rain multiyear research efforts on chemical Steve Parmenter, California Department of results in highly variable and difficult to and biological control agents. Send Fish and Game predict distributions from year to year. research ideas and proposals to Debra Jeff Seigel, Natural History Museum of Los Hughson ([email protected]). Angeles County Mechanical treatments consisting of hand Paul Stapp, California State University -- pulling and hoeing are labor intensive, As a federal partner of the Cooperative Fullerton effective only on a small scale, and can Ecosystem Studies Units (CESU) Kelley Stewart, University of Nevada -- cause significant ground disturbance. (http://www.cesu.psu.edu/) we are able to Reno Brassica tournefortii has a taproot which work with academic scientists to seek
Reviewers can be pulled with no resprouting but is funding for research on resource related Gar Abbas, U.S. Forest Service difficult to hand pull when the plant is in conservation efforts. Contact Angie Todd Esque, U.S. Geological Survey its basal rosette stage. Plants must be Evenden ([email protected]) for Angie Evenden, Great Basin Cooperative bagged and removed if hand pulling is information about the CESU system. Ecosystem Studies Unit conducted during early to late seed pod Robert Fulton, Desert Studies Center development. Partial efforts that result in References David Graber, National Park Service – lower densities of plants, but not Brooks et al. 2003. USDA Rocky Pacific West Region complete eradication, may conversely Mountain Research Station. Kathy Longshore, U.S. Geological Survey lead to increased net seed production Sanders and Munnich 2000. In Bossard Richard Stevens, Louisiana State University (Trader et al. 2006). et al., University of California Press,
Berkeley 287-291. Contact: Debra Hughson, 760-252-6105 Mojave National Preserve is interested in Trader et al. 2006. Madrono 53(4):313 [email protected]
16 Mojave National Preserve Science Newsletter 2009, Number 1