California and Nevada

The Ecology of Selerocactus polyancistrus (Cactaceae) in California and Nevada

Item Type Article

Authors May, Richard W.

Publisher University of Arizona (Tucson, AZ)

Journal Desert Plants

Download date 11/10/2021 11:26:06

Link to Item http://hdl.handle.net/10150/554194 6 Desert Plants 11(1) 1994 Abstract The Ecology of Several key aspects of the ecology of Sclerocactus polyancistrus (Cactaceae) are described based on a 15 year study by the au- Sclerocactus thor. Highlights of the study, funded in part by grants from the Cactus and Succulent Society of America and the United States Navy (China Lake Naval Weapons Center), include a record of polyancistrus growth rates and the impact of predation and infestation. The study also includes the analysis of the carcass remains, identi- (Cactaceae) in fication of predators, the benefits and offsetting effects of predation, and the resulting morphological abnormalities.

California and Nevada Using microhystological analyses to determine the relative density of discerned fragments of scats found within the carcasses, Neotoma lepida was found to be the chief predator of this spe- Richard W. May cies at elevations below 1500 m (5000 ft). At higher elevations, 4210 Elderwood Dr. infestation primarily by the cerambycid beetle Moneilema Houston, Texas 77586 semipunctatum, is the dominant cause of mortality. Returning each spring over a 15 year period (1976 -1991) to six study sites in the Mojave Desert, the author recorded apical growth measurements of over 350 stems and correlated these growth rates to rainfall and other key climatological factors. The results show that microhabitat is a major factor in this species growth rate and that the oldest plants within a given population are in excess of 50 years in age. As expected, there is a direct correlation between seasonal rainfall and the growth rate and number of flowers produced.

When surveying eastern Nevada and the Canyonlands National Park areas of Utah, the author has found similar predation and infestation in S. pare /orus and S. spinosior, which indicates that some of the data collected in this study may also be appli- cable to other small stem cacti of the southwestern deserts, par- Acknowledgments ticularly within the same genus. I thank the Cactus and Succulent Society of America and the Navy at China Lake for their financial assistance as well as the Introduction Navy environmental staff and now dear friends, Tom McGill The genus Sclerocactus includes 8 species occurring from the and Bev Novak for their help in the field. Thanks also for those Mojave Desert in California and Nevada to the Great Basin who provided their unselfish and free technical expertise that Desert, Utah, western Colorado, northern Arizona, and north- turned raw field data into meaningful results. This would in- west New Mexico. Several variants exist and, most recently, at clude John Doyen and Jerry Powell at U.C. Riverside, Bill Clark least two new species have been proposed from eastern Nevada. and Paul Blom at the University of Idaho, Tom O'Farrell at Most of the species have a prominent inflorescence and sinuous EGG (Boulder City), and W. P. Stephen at Oregon State for red -maroon (some hooked) and white spines. The plants usu- their help with insect identification. Also Dave McKay at The ally have a solitary growth habit and do not branch unless dis- National Aeronautics and Space Administration at the Johnson turbed or injured. S. polyancistrus is the largest species in the Space Center who provided me the use of the SEM for seed genus. studies. Thanks also to the California BLM staff and especially Texas A &M University and the University of Wichita soil labs Although there has been a growing interest in the genus for the analysis of soil samples. For the inspiration to continue Sclerocactus, a detailed study of Sclerocactus polyancistrus has and persevere as well as for his technical assistance, I thank not been performed to date. In 1976, the author became inter- Lyman Benson. Also thanks to Gerhard Haslinger in Austria ested in this particular species as a result of several published for his expertise and skill in germinating the seeds of the rare accounts and theories concerning its ecology. For example, there form of this species. Thanks to Doris Fredendall and Maryann have been several references in literature to the "mysterious Henry who were in the field when I wasn't able. Finally, a spe- boring grub" apparently responsible for its high mortality rate cial thanks to a special friend and field expert Matt Johnson and resulting low population densities. As a result of this study, who had the patience to help me in such tedious tasks as root some questions have been answered, and general information mapping and for recommending Desert Plants as a home for about this species has been generated. this work. Unless otherwise stated, all photos and illustrations are by the author. May Sclerocactus 7 Over 40 populations were visited between 1976 and 1991 and this genus in the east where summer rainfall is more common, six study sites were selected in California and surveyed yearly, the hilum of the S. polyancistrus seed does not have a crusty primarily during late April and early May, along with several protective "cap" which resists water imbibition. The hilum on remote sites on a more infrequent basis. The total study effort other species within this genus, have this structure which is included a record of population densities, the identification of advantageous, preventing premature summer germination and predators, pollinating agents, internal stem temperature mea- a resulting high seedling mortality rate. This explains why S. surements, morphological comparison studies, soil analyses, seed polyancistrus is not present further east. morphology and germination studies, reproduction strategies, and identification of threats, both natural and by man. Unpub- The northern extent of this species range is tied to climatologi- lished field data reports were generated yearly and submitted cal factors, particularly temperature extremes. In central Ne- at no cost to the California BLM, the Nature Conservancy and vada, below -18°C (0°F) readings are not uncommon and thermal other environmental agencies. The reports carry the name Desert tolerances of the S. polyancistrus stem, roots, and seedlings Research International, a non -profit, tax exempt organization may often be exceeded. The larger more exposed stem growth formed by the author. habit of this cactus does not fare well when temperatures are extremely cold. Smaller, globose cacti (such as Coryphantha) After a brief overview, this paper will focus on this species ecol- which have a more conservative growth habit closer to the ground ogy including growth rates and the impact and nature of preda- where they are protected by the snow, fare better in these more tion and infestation. northern regions. Perhaps unlike S. polyancistrus, Coryphantha has been shown to exhibit an ability to "cold harden" (Gibson Introduction to the Species and Nobel, 1986). S. polyancistrus (Figure 1, page 18) is thinly distributed over a large portion of the Mojave and Great Basin deserts of California Winter and spring, when the bulk of the precipitation in this and Nevada. Its range lies within the basin and range physi- region occurs, constitutes the growing season for this species as ographic province. It is found most frequently in plant communi- well as for most other flora in the Mojave Desert. Since this ties dominated by Larrea, Atriplex, Artemisia sp., Grayia, species is most conspicuous when in bloom, prior knowledge of Coleogyne, Haplopappus, Ephedra, and Hymenoclea shrub a population's peak blooming period during a field survey is complexes. It is most commonly known as the Mojave Fishhook advantageous. Anthesis varies from population to population Cactus, but has also been called the Hermit Cactus, the Pine- depending upon altitude, latitudinal (north -south) location, and apple Cactus, and the Devil's Claw. For a complete taxonomic other microclimatic factors as well as abnormal fluctuations or description, the reader is referred to Lyman Benson's Cacti of the extremes of weather. Generally first to bloom are those popula- United States and Canada (1982). tions situated in the Mojave River basin and El Paso Mountains of California which bloom in late April and early May. By mid - The plant favors porous, alluvial soils most frequently derived May, the plant is in bloom over most of its range. Larger stems from igneous (granite and rhyolite), some Jura -Trias at higher elevations have been found in bloom as late as mid - metavolcanics, and sedimentary (Pliocene/Pleistocene non -ma- June. The fruit are normally ripe by early July. rine) rocks (USGS survey data). It is rarely found on dark vol- canic soils and only occasionally in areas where basaltic rocks The red and white spines are also of interest and, like most are dominant. It is seldom found in flat areas or in areas of poor cacti, serve to benefit the thermal economy of the plant. During drainage but, instead, prefers the southern and western slopes the summer, desiccation of the stem causes the spines to form a of hills with slopes less than 45 degrees. Soils are typically al- dense mat, shielding the stem from the sun. This is particularly kaline (pH usually greater than 8.0) and in soils high in cal- the case in the southern populations and at lower elevations. cium. Elevations usually range from 730 -2300 m (2400 ft.- 7600 Spine measurements by the author reveal that the plants in these ft). southern populations have spines which average 2 cm longer than those in the northern or higher elevation populations. At Summer rainfall across the region is rare. Precipitation occurs the highest elevations (above 1700 m/5500 ft), the trend re- primarily during winter and early spring when Pacific storms verses and the red spine length again becomes longer. The width and frontal systems take a more southerly track. Rainfall is el- of the long white ribbon -like central spine likewise varies with evation dependent and within S. polyancistrus populations geographical /physiographical location, averaging 0.6 mm wider the average annual precipitation (some in the form of snow) in the southern populations. usually ranges from 100 to 200 mm (4 -8 in). Temperature extremes recorded during the study period (George AFB) ranged A rare yellow -spined form also exists (Figure 2, page 18) and has from - 10° C (14° F) to 43° C (109° F). The extent of this species been found in two populations. The first population found by range appears to be linked closely to winter temperatures and the author was in the El Paso Mountains of California where it summer rainfall. coexists in large numbers with the common form within a habitat which can best be described as typical for the species. Eleva- The eastern limits of its range (Nevada Test Site) corresponds tions range from 975 -1372 m (3500 -4500 ft). The yellow spined closely to the one inch summer rainfall isopluvial. The repro- form is quite distinct when compared to the common form. Both duction strategy of this species appears linked to the morphol- have the usual white radial and upper white central spines; how- ogy of the seed and the summer rainfall pattern which is nor- ever, instead of the usually hooked, 6 -8 red spines, the rare mally absent in the western Mojave. Unlike other species within 8 Desert Plants 11(1) 1994 form has pale yellow spines. The difference in spine color of choring roots reach a depth of approximately 16 cm (7 in) be- both the common and the rare form are also easily distinguish- low the surface. The surface roots, which appear to be designed able on the seedlings within three months after germination. for scant rainfall, are not more than 7.6 cm (3 in.) below the No additional morphological differences have been recorded to surface and fan out as far as 90 cm (3 ft). date. Cross pollination studies indicate that it is a recessive gene. A more complete description of this rare form was compiled by Transplant experiments were also conducted. Seven plants were the author (May, 1984). intentionally uprooted and relocated within two populations. A few were brought into cultivation and returned to the field. The flowers are normally magenta but often fade to pink after In all cases, with attention to transplant location selection (un- two or three days of sun exposure. The petals are usually obo- der a shrub), they survived and flowered 4 -6 years after trans- vate. The stigma lobes are normally red, but one population plant. Data such as this may be useful when populations are was found where 20 percent ofthe stems had green stigma lobes. threatened. On the China Lake Naval Weapons center, a population exists with almost totally white flowers, but such occurrences are rare. RAINFALL vs. BUD PRODUCTION The flower buds emerge from the apical region and are self -sterile and consequently the flowers require cross pollination. Flowers 300 are chiefly pollinated by non -social and social bees, but more so by the latter. Several pollen -eating beetles (usually Trichodes 250 GEORGE AFB ornatus) are also common visitors. Using wire screen cages to restrict the size of pollinating agents, 1982 studies by the author J200 indicate that some of these pollen eating beetles, which can carry J a substantial pollen load, also effect pollination through a primi- tive two stage "mess and soil" pollinating process. Transport- Z 150 ing pollen from flower to flower and depositing the pollen on the ¢ petals constitutes the first stage. Then, at night, the petals close, 100 depositing the pollen on the stigma lobes. This strategy may be useful in an unexpected spring cold snap when bees are normally 50 inactive. Hummingbirds have also been observed to frequent these flowers. 0 The fruits are dehiscent and naked, green initially, turning to 45 bright red when ripe. A typical fruit will contain approximately 40 (N = 8) 120 seeds. They are dispersed by birds and rodent activity. The seeds are relatively large in comparison to other species of cacti 35 and are the largest within this genus (3 - 3.5 mm). They have a 30 hard, black testa. Fresh seeds are difficult to germinate in culti- 25 vation and usually must be treated or "aged" before germination can occur. Seeds collected from old carcasses germinate 20 more readily. 15 10 Weather affects the time of anthesis. In general, anthesis occurs later (early May) after a cool, wet winter /spring, and earlier (early 5 - mid April) after hot/dry winter /spring. "Wet" winters are greater 0 + I oo C) O N N) than 100 mm, "hot" applies to April average maximum tempera- h r co m CO CO tures that exceed 25° C (78° F). 06 dr) O r- N ti ti W CO CO C) (3) C) C) C) rC) Figure 3 depicts the effect of rainfall on the number of buds for eight healthy stems in one population over a six year period. YEAR (JULY -JUNE) As one would expect, the fluctuations in number of flower buds Figure 3. are directly proportional to the amount of rainfall received. Growth and Growth Habits The plant also has extra -floral nectary glands in some of the "The next generation of studies on cactus growth habits will upper areoles which attract ants. The precise role, if any, of this have to consider many different structural and physiological structure is not yet known. The ants do not pollinate the plant. components ... our discussion is limited by a paucity of research on the subject." The mature plant does not have a clearly defined tap root. In- Gibson and Nobel, 1986 stead there are a series of 1 -3 entwined anchoring roots which enable the plant to remain erect. Root map studies were per- As Gibson and Nobel suggest, the data available on growth formed on a 14 cm (5.5 in) tall plant. Results indicate that an- rates and growth habits of cacti are scant. A major contribution May Sclerocactus 9 has been Steenbergh and Lowe's work at the Saguaro National sites on a random basis, but do not factor in significantly into Monument where quantitative growth data was collected for the growth data presented.It became apparent early on in the the giant Saguaro Cactus (Carnegiea gigantea) for the National survey that many of the original survey plants would either not Park Service (Steenbergh and Lowe, 1977). Although growth survive or become deformed during a 15 year time period. The habits for both large and small stem cacti have similarities, small number of plants included in the survey was increased as the stem cacti, such as S. polyancistrus, require separate treatment. survey advanced owing to the unexpected large turnover due to The effects of micro -habitat (slope, cover, etc.) are pronounced predation. and have far reaching effects when applied over the life span of the plant. Some small stem cacti may spend their entire life - Permanent bench marks were not installed at the base of each span under the cover of a shrub, for example, while other mem- plant. Stem heights, however, were consistently measured as bers of the same population may have full sun exposure. The accurately as possible, more than once, from the same point on age -height relationship consequently will vary from plant to the down slope side of the stem. The inaccuracies with this plant. method (estimated to be plus or minus 2 mm) should be considered when evaluating the data, but are deemed acceptable, be- On April 30, 1977, while surveying a newly found population cause of the long term nature of the study. Plants within each of S. polyancistrus near Red Rock State Park in California, the population were mapped and some were tagged to ensure proper author stooped over and took the first of over 5000 stem mea- identification. Relocation of previously mapped or tagged plants surements; and what began as an afterthought, has led to a fif- was never a problem. teen year growth rate study monitoring over 350 plants in six populations. The height, width, and number of buds /flowers were the primary data recorded annually for each stem. Other pertinent data Site Locations and Description such as overall health, vigor, disturbances, and change of cover Figure 4 depicts the location of the study sites where yearly was also recorded. As noted in the introduction, morphological growth data was collected. Site selection included considerations comparisons were also made with regard to spine length and for obtaining meteorological data. Elevations within the six width, variances in flower buds (e.g., stigma lobe color, petal primary study sites ranged from 850 - 1200 m (2800 - 4000 ft). size). Aside from what has already been discussed, the detailed Remote site locations were useful primarily for recording varia- results from these data are outside the scope of this paper. tions in morphology and predation/infestation rates across the range. Rainfall and other climatological data was obtained from the closest recording station. This included primarily George AFB, China Lake, and Barstow; but also Goldstone, Inyokern, Randsburg and the California Department of Water and Power recording Stations. Availability of such information was considered crucial in the study site selection.

Growth Stages Seed germination and seedling growth occur during the rainy season, between November and May. Seeds from both the previous years' and prior year harvests usually germinate during this time period. It was determined from a seed weathering ex- periment performed by the author using shielded and non - shielded test stands in the study sites that the seeds of S. polyancistrus will remain viable at least three years after ex- pulsion from the ripened fruit which occurs in late June or early CALIFORNIA July. o

SOPO RANGE & NUMBERED DATA COLLECTION STTEs The emergence of the seedling from the seed coat begins with

O PRIMARY SITES REMOTE SITES the appearance and elongation of the embryo (day 3), the fine (YEARLY DATA) 0 01- 2000(793m) 52-5600'(1671m) root hairs (day 4), and the emergence of the tap root (day 5). 02- 2600(í54m) 00-6000'(1325m) 03- 3200(976m) 07 3900(1100m) 00 3300(1006m) 25 4160'(1266m) The seedling has distinct cotyledons, areoles, and spines and Oa 2900'(954m) es. 3600'(1139m) the height of the seedling measures about 10 mm after day 12. By the end of the first winter - spring growing season, the seed- Figure 4. lings will undergo little vertical growth but, instead, will "fill out" and develop a prominent tap root. Seedlings at this stage Methodology will typically measure 10 -12 mm in height and 4 -6 mm in width. Two populations were identified for the study in 1977 (see Fig- The first four years are critical. Over 80% of the seedlings still ure 4). Two more were added in 1978, one in 1980, and a sixth alive at year four do not survive to maturity. Those that do will in 1981. Data was collected on a yearly basis during approxi- attain a globular form, about 20 mm in height and 25 mm in mately the same time period for all six of the populations (on or width at year four. Survival rates appear higher when seedlings about May 1st). Data was also obtained from several remote are within a carcass than those that are found adjacent to the 10 Desert Plants 11(1)1994 base of mature stems, although sufficient quantitative data was Growth Rate Data not obtained to validate this observation. To characterize the growth pattern for S. polyancistrus, several charts have been prepared and a brief description of the results The growth stages of S. polyancistrus are summarized on Fig- is provided. ure 5. Apical growth begins at the outset of the rainy season. Flower buds begin to emerge as early as February. Warmer spring Figure 6 depicts a typical growth record for 18 stems at site temperatures (10° C / 50° F) usually set in between late April or location 01. Variances in growth rate from stem to stem are early May and desiccation begins thereafter. Premature desicca- primarily attributable to variations in microhabitat and growth tion can occur, however, after an unusually dry winter (rainfall stage. Note the steeper slope of the growth lines for the stems less than 100 mm/4 in). between 5 and 25 cm (2 -10 in). The more globose growth habit during the first few years after germination accounts for the slow vertical growth during the first 8 -10 years. GROWTH STAGES

FORM AGE /YRS HEIGHT STAGE REMARKS Figure 7 (page 12) depicts the record of apical growth (change in (cm) overall height) for 21 stems in 3 populations over a nine year 0- 4 0 - 2 SEEDLING GLOBULAR (LATE) period. These plants were selected because they were considered the most reliable data points, primarily because of a lack of 2 4- 6 2 - 8 JUVENILE BLOOMS (LATE) any morphological or habitat disturbances which might affect the data. The results show that there is an apparent drop in the

3 6-15 8 - 15 MATURE CYLINDROID rate of longitudinal growth with age. From these and other data Il BLOOMS 2 41 accumulated during the study, it has been determined that the average rate of growth (all populations) during the "mature" 4 15- 30 15 - 30 LATE BLOOMS CAN MATURE EXCEED 10 growth stages (Figure 5) over the 15 year study period was approximately 1.3 cm per year. At site location 5, very large stems in excess of 50 cm (20 in) are present. Here, the rate of 30- 40 30 - 40 OLD STEM BEGINS LEAN growth was 1.8 cm (.7 in) per year and, as such, indicates that SIDE BLOOMS (RARE) these abnormally large stems are not necessarily the oldest found, they are just growing faster. It should be pointed out, however, that older, usually partially prostrate stems in excess of 40 cm 40- 50 40 - 60 VERY OLD PARTIALLY PROSTRATE SIDE BLOOMS COMMON (16 in) are more difficult to measure. Although the growing point remains active, there is some accompanying basal shrinkage which cancels any longitudinal gains. The apparent drop - T Qi...40- 50+ 40 - 60 SENESCENT DECLINE: PROSTRATE off in overall longitudinal growth noted for the older stems may STEM NARROWING actually be due to basal shrinkage (Figure 7). This would sug- LOSS OF VIGOR gest that the largest stems could be older than the curves por- tray. Figure 5. Figure 8 correlates rainfall to apical growth for N = 97 (max) Variables Affecting Growth stems. The stems are grouped by an arbitrary stem height cat- Growth rates vary from plant to plant and from year to year egory or classification (0 -60 mm, 61 -120 mm, etc.) to reflect within any given population. Consequently, age -height relation- variance in the rate of apical growth for each category. In addi- ships also differ. Based on the results of this study of small stem tion to the overall seasonal rainfall, the date of the last signifi- cacti, one cannot assess with accuracy the age of a plant by cant rainfall (prior to the survey date), the average maximum merely noting its size. Besides probable genetic variations, these spring (April) temperature, and the survey date (which was not differences in growth rates are primarily due to year to year always controllable by the author) impact the seasonal growth. climatic variations (particularly temperature and rainfall) as well Note for example the last two years of the survey. Despite a as microhabitat factors such as slope, cover, and substrate. Cli- yearly rainfall total of 46.4 and 45.9 mm (18 in) respectively, matic factors such as rainfall are dynamic, that is, they vary almost identical for these two years, the plants exhibited a nega- from year to year. Microhabitat factors (cover, slope) are fixed tive growth rate as recorded in April 1990. This can be attrib- and in most cases do not vary. The total annual apical growth uted to an abnormally warm dry April (27° C / 81° F)average for any given plant, therefore, is made up of two components; maximum) which resulted in premature desiccation of the plants. the genetic constant component plus the variable climatic com- Had the survey been performed in March, the apical growth would ponent. probably have been closer to the values obtained the previous year. Note also the 83 -84 season. Although temperatures were The resulting growth rate of any given plant is a complex equa- cool in April, there was no significant rain for over four months tion of interacting variables the development of which is be- preceding the survey date. Very few plants were in bloom and yond the scope of this work. The overall effects of these inter- most of the stems were desiccated. acting variables can be presented, however, as the apical growth is measured from year to year and correlated to the dynamic variables of rainfall and temperature. May Sclerocactus 11

TYPICAL GROWTH PATTERN, SITE 01, N =18

50 U 40 _ W 30 Apical growth S site 01 20 W ,'-- Figure 6. (1)1 0 X^ hi_Pt o :._:' ritil NiV.r M dt[) CD c0 \CO \CO \co \03 \CO \CO M N LO M CM N 11) t[) d' LO l)

SURVEY DATE

300 - PRECIPITATION 250

200

150

100

MM 40 30 Figure 8. 20 1IIIJ 10 .' `. I 'mwr\\we:,,- x/ 1 ,rial 0 xMNAM \\ -10 " -20 HeightClassification 111I.-=MIMI 00.60mml]121-180mm \/ -30 x GROWTH 61-120mmX 181mm + - A -40 I i YR (Jul - Jun) 78 -7979 -80 80 -81 81 -82 82 -8383 -8484 -85 85 -8686-87 87 -8888 -8989 -90 Survey Date 5/235/19 5/2 5/24 5/3 5/2 4/27 4/5 5/3 5/3 4/22 4/14 Sig. Rain Date (1)5/2 5/1 4/2 4/11 4/30 12/253/27 3/16 3/15 4/23 3/25 2/18

Ave. April Max Temp.73 71 74 69 62 71 77 74 80 73 81 76

N 15 7 17 25 76 97 80 90 74 69 68 33

(1) Last significant rainfall greater than 25 mm 12 Desert Plants 11(1) 1994 that two huge 60 cm stems (24 in) found at site 04, for example, Figure 7. did not die from old age but rather they were consumed by predators during the extreme drought (1988). Predation is an impor- 14 - tant aspect, therefore, in understanding this species ecology. 9 YEAR APICAL GROWTH (1980-1989) Small Mammal Predation and Infestation 3 POPULATIONS Predation - An Introduction 12 - N = 21

x Many dead plants, at least half as many as live ones as seen in 1977, occur scattered across the northern Nevada Test Site and

xu Tonopah Test Range. These plants remain as intact skeletons of spines, suggesting that predation or disease has killed a sub- a - o stantial portion of the population. GROWTH - CM William A. Rhoads, 1978. s - It becomes immediately apparent to even the most casual ob- x x server that a large number of S. polyancistrus spine skeleton 4 - carcasses are usually present, even within relatively dense populations of this cactus. Through the careful examination and analysis of hundreds of carcasses as well as partially consumed 2 - and infested plant material and depositions at the base of such remains, it has been determined by the author that both small mammal predation and infestation are directly and indirectly

I f I I I responsible for the vast majority of S. polyancistrus deaths and o + I and resulting carcasses. The data also suggests that the preda- 0 5 10 15 20 25 30 35 tion process plays a significant role in the ecology of this plant 1980 STEM HEIGHT - CM and may, in fact, be advantageous from the standpoint of the species' long -term survival. The agents most often responsible This chart also shows that, as one would expect, the change in have been identified. overall height of the largest stem category (indicated by the "X ") responds more to rainfall and temperature. These plants The Carcass have more mass and water content and exhibit accordion -like The carcass and plant remains found within S. polyancistrus swelling and contraction facilitated by the ribs. populations offer clues as to the cause of death. These carcasses fall into two distinct classifications: "open" and "closed" (Fig- Figure 9 depicts the estimated age for a given stem height based ures 10 a and 10 b, page 18). Each of these two major classifica- on the data collected from 251 healthy stems at all six study site tions can be further sub -divided into three sub -classifications, locations. The calculated age value for each plant was estimated designated as type 1, type 2, or type 3. The Figure 11 table by a two step method. First, the number of years and resulting (page 14) summarizes these different classifications and distin- growth during the observed life of the plant was recorded. Added guishing characteristics and supports the following discussion. to this value was the calculated estimate of the plants age at the start of the observation period. This value was determined by As will be shown, open carcasses are the result of small mam- dividing the "start observation height" by the average apical mal predation. Open type 1 carcasses are most common. In this growth rate (cm/yr) of the stems within the same population case, the stem has been totally consumed leaving, at best, only which were less than or equal to the start height value. In other a portion of the spine skeleton. The remains are usually torn words, the age of the plant at the time when the measurements apart and fragments are often scattered about the ground within began was estimated based on the growth data collected from 2 to 3 meters (7 -10 ft) of the carcass. In the case of type 2 or the smaller stems. Since most of the stems within a particular type 3, the spine skeleton is virtually intact with the cortex again height category have apical growth rate within plus or minus .2 absent. An entry hole, 3 -4 cm (1 -1 1/2 in) in diameter located a cm per year of each other, the age of a 20 cm (8 in) plant may be few centimeters above ground level identifies a type 2 carcass. off approximately 2 1/2 years. The equation, simplified, for the Type 3 carcasses are the result of predation which has its origin curve in Figure 9 yields: below ground level, leaving the spine skeleton undisturbed. For all three types of open carcasses, the entire cortex, pith, and Height = -.58 + .505 (Age) + .0176 (Agee) most or all of the epidermal tissue is usually consumed, leaving an "open" interior within the skeleton. The results indicate that the oldest stems found were probably close to 50 years old. Allowing for error it is possible that the Closed carcasses are often the result of infestation. Insects ovi- oldest could be as much 60 years old. The odds are against the posit at various locations on the stem or basal area. The larvae plant living beyond 30 or 40 years, however, because of the bore into the stem and consume a portion of the fleshy cortex. constant threat of predation and infestation. The author observed Subsequent bacterial infection (rot) will usually kill the plant, May Sclerocactus 13

AGE vs. HEIGHT (N =251)

Figure 9.

10 20 30 50 AG E (YEARS) y =- 0.58226 + 0.50480x + 1.7607e-2x^2 R"2 = 0.835

leaving a type 1 or 2 closed carcass with an intact spine skeleton. Neotoma dens are a common feature within S. polyancistrus Brown (dead) epidermal tissue and some cortex remains are still populations. Studies of the habits of small mammals, particu- present and represent a major distinguishing characteristic of larly rodents, over recent years have established the dependency closed type carcasses (Figure 10b, page 18). Infestation is a much of some of these animals on cacti for survival. It has been found slower process in comparison to small mammal predation and that the desert woodrat must rely on cacti for food, water, mate- often takes many weeks to run its course. As a result, the stem rials for den construction, and a means for avoiding predation desiccates or collapses, leaving a spine skeleton in a compressed (Brown, et al. 1972). Cacti, in fact, comprise nearly one half of state, the spines packed tightly together. their yearly food intake (Schmidt- Nielsen, 1964). The desert woodrat must rely on succulent plant material when water is As depicted in Figure 11 (page 14), for 20 percent of the car- not present. It has also been established that desert woodrats casses examined to date, the cause of death could not be deter- have morphological and behavioral adaptations which enable mined. Intermediate carcass forms have also been found. Freeze them to negotiate the spines and digest the noxious chemical damage, for example, may account for some deaths particularly at substance oxalic acid found in some cacti (Schmidt- Nielsen, higher elevations. 1964). Small mammals without such adaptations would have great difficulty negotiating the dense mat of hooked spines It has also been found that open carcasses (and, hence, small present on S. polyancistrus. mammal predation) have a rather distinct geographical and elevational distribution and range. Open type 1 carcasses are It is likely that more than one predator (rabbits, for example) found predominantly in the southern, lower elevation popula- will consume parts of the same plant once the initial stem dam- tions (Figure 12, page 15). Closed carcasses occur almost exclu- age has been inflicted, the spines are pulled apart, and the fleshy sively in the northern and higher elevation populations. The stem interior is exposed. Evidence to support such occurrences available food supply for small mammals, which is less in the includes fecal depositions from more than one animal within south during hot summer months, may be a factor. the carcass remains of recently consumed stems. It is the initial predation, however, that sets the process in motion and to which Small Mammal Predation the focus is now directed. Once the protective spination and By examination and analysis ofpartially eaten stems, fecal matter epidermal tissue are destroyed, the plant is exposed to (scats) deposited at the base of open carcasses, and other mis- secondary predation and bacterial infection. This accounts for cellaneous data and observations, it has been determined by the the very low percentage (less than 10 percent) of stems the au- author that the vást majority of open carcasses are the result of thor has found that survive partial consumption. Field studies, predation by Neotoma lepida (desert woodrat). therefore, have focused on the identification of predator(s) responsible for the initial stem damage. CLASSIFICATION MAJOR CAUSE SUB -CLASS CARCASS©aß©aZZ DESCRIPTION CLAZ%0FOCATOOHt OCCURRENCEESTIMATED PREDATOR I.D. - FOUND MOST OFTEN IN SOUTHERN REMARKS TYPE i - SPINE SKELETONUSUALLY TORN ONAPART, ONE SIDE ONLY 75% NeotomaLepida - INITIAL ENTRY POINTMIDPOPULATIONS/LOWER OR USUALLY UPPER PORTION AT ELEVATIONS OF STEM SMALL - SPINE- CORTEX SKELETON ususALLY INTACT TOTALLY ABSENT - FOUND MOST OFTENPOPULATIONS IN NORTHERN / HIGHER ELEVATIONS OPEN PREDATIONMAMMAL TYPE 2 - 3-5 CM HOLE CHEWED ON SIDE/SPINES 20% POSSIBLE:SMALL RODENT UKELY, Neotoma lepida - INMAL ENTRY POINT(SOMETIMESSPINESIDE USUALLYOF SKELETON STEM, DIFFICULT ONLEAVING REMAINS TOHOLE DETECT) IN - SPINE SKELETON USUALLY INTACT - PREDATOR CONSUMESEXCISING TUBER, ROOTS; OFTEN THEN WORKS UP INTO TYPE 3 - ROOTS/TUBER- CORTEX REMAINS ABSENT EXCISED 5% ORSMALLUNIDENTIFIED; Thomomys RODENT 8otfae POSSIBLE; . CARCASS SOMETIMESENTRYSTEM TO FROMFOUND RODENT BELCMI AT DEN OR BURROW SPINE SKELETON INTACT, COMPRESSED - INSECTS OVIPOSITS AT BASE TYPE 1 --1 PORTIONS to 1 1/2 CM OF ENTRY TISSUEATCORTEX BASE HOLES STILL &OF EPIDERMAL CARCASS PRESENT 60% semipunctatumMona/lama PUPATECONSUMINGOF STEM: IN LARVAE CELLS CORTEX ATBORE BASE EXIT UP OF ATINTO STEMBASE; STEM INFESTATION - SPINEPUPAL SKELETON CELLS AT INTACT,BASE COMPRESSED CLOSED TYPE 2 - CORTEX & EPIDERMALENTRY HOLES TISSUE IN SIDE REMAINS OF STEM 20% OTHERDistopasta yumaeue, - LARVAE BOREINSECT INTO OVIPOSITS STEM FROM ON SIDE SPINES UNIDENTIFIED CAUSES TYPE 3 .- SPINE SKELETONPORTIONSTISSUE INTACT, OFSTILL CORTEX COMPRESSED PRESENT & EPIDERMAL 20% UNIDENTIFIED - POSSIBLE CAUSES: FREEZEUNIDENTIFIEDBACTERIAL DAMAGE INFECTION INFESTATION May Sclerocactus 15 a period of 5 days. Secondary predation may at times be a con- tributing factor as noted earlier, accelerating the process. Since actual consumption has never been witnessed, it is not clear whether the predator works alone. Successive observations on a daily basis and in localized areas indicate that type 1, predation occurs at night or at dusk and in localized areas suggesting the predator is nocturnal and territorial.

An attempt was made to determine the amount of biomass consumed by the predator(s) responsible. The use of live trappings to collect such data was ruled out. S. polyancistrus predation is seasonal (mid -late summer) and intensities fluctuate significantly from year to year and from location to location within any given population. Therefore, it would be impractical to ar- bitrarily set traps in an àttempt to capture the animal which RANGE OF SPECIES DEPICTING OCCURRENCE OF OPEN CARCASSES has recently consumed S. polyancistrus plant material. Stems TYPE i(Ope) CARDASSES PREDOMRMTE with fresh wounds must first be located (an element of luck!) ATYPE 283RARE: NPESTATION RARE and then traps can be set or direct observation techniques em- TYPE 2 (OPEN) CARCASSES PREDOMINATE: TYPE 183 CARCASSES OCCASIONAL; ployed. 11`4FES1AT ON FREQBif

'TYPE 3CARCASS maNTNQS Based on observations during July 1986, the average biomass consumed during predation was estimated to be 400 cm' (24.4 Figure 12. in') per day per host plant. Although this figure is based on only one series of observations, it is consistent with the estimated volume of plant material found missing (wounds) on other partially consumed stems. Inaccuracies may exist, however. It Examination of Partially Consumed Stems was found that the amount consumed each day increases some- A major source of data has been S. polyancistrus stems which what as consumption continues, suggesting, as noted earlier, were found partially consumed. In the examination of the wound that more than one animal may participate during the latter areas; location, size, width of the incisor marks, volume of the stages of consumption. The volume consumed on any given plant material consumed, and the sequence ofevents which occur day by any one predator may also be dependent on the volume during the course of the predation process are considered in of plant material consumed elsewhere which may include other predator identification. S. polyancistrus host plants.

Although actual stem consumption has not been witnessed by Analysis of Fecal Matter the author, successive events which occur during the predation It is common to fmd fecal pellets (scats) deposited at the base of process have been recorded within hours of their occurrence. S. polyancistrus open carcasses. The scats measured 8 -10 mm Typically, for predation resulting from type 1 open carcasses, in length and were most likely from N. lepida. Size, shape, and the initial wounds are inflicted on the upper one third of the composition of these scats provided additional evidence for stem or, on smaller stems less than 10 cm (4 in) tall, in the predator identification. apical region. Some of the spines are chewed or torn off in the process. After the initial entry has been made, the remainder of Specimens collected from several locations within one popula- the cortex is consumed, the animal working basally. As the epi- tion were selected for analysis. In 1981, one batch was sent to dermal tissue and cortex is consumed and the animal activity Dr. T. O. O'Farrell, zoologist at EG &G - Boulder City. A sec- continues within the plant, the spines have a tendency to sepa- ond batch was sent to Colorado State University for a rate further, expanding the wound area, thus leaving the vast microhystological analysis. A third batch was examined in 1986 majority of type 1 carcass spine skeletons spread wide apart. by Dr. G. N. Cameron at the University of Houston. Both Often times, spine fragments can be found 2 -3 meters (7 -10 ft) Cameron and O'Farrell are noted in the field of mammalogy, and as much as 5 meters (17 ft)from the carcass remains. having performed a number of field studies within the Mojave - Spine dispersal occurs primarily by the wind. The predator may Great Basin desert region. also scatter some of the spines during its travels from the plant. Occasionally, the entire carcass remains will become dislodged The scats analyzed were typical in terms of size and shape and may be found many meters from the original site. to those observed in type 1 carcasses found elsewhere. The pellets measure 8 -10 mm in length and 3 -5 mm in Unless only a small stem is affected, the process described above width. Over 90 percent of type 1 carcasses from recently usually does not take place all at once but instead over a period consumed stems contain fecal depositions of this size and of time from several days to many weeks. Some stems have shape. The high percentage of occurrence indicates that been found to incur wounds periodically over the years, but have the predator which left these scats was also responsible never been totally consumed. In most cases, however, for stems for consuming these plants. The microhystological analysis sup- greater than 30 cm in height, the plant is totally consumed over ports this hypothesis. The percent relative density of discerned 16 Desert Plants 11(1)1994 fragments from the pellets were determined based on 20 fields Infestation by Moneilema semipunctatum per sample. S. polyancistrus plant material was provided for A significant number of S. polyancistrus carcasses show no comparison. It was found that the pellets contained both epi- signs of small mammal predation. These closed carcasses, as dermal tissue and spine fragments from S. polyancistrus. The already discussed, are intact, the spine skeletons are usually analysis also revealed that the fecal matter contains a large compressed, and they contain the decaying or dried remains of amount of vegetation. This tends to discount the possibility that the succulent portion of the stem. It is estimated by the author the scats are from heteromyids which primarily consume seeds that at least 60 percent of the closed carcasses are the result of an (Cameron, 1987). [It should be noted that fecal pellets of differ- infestation by the cerambycid beetle Moneilema semipunctatum ent size and shape from other animals have occasionally been (cactus beetle). This insect is normally associated with the ge- found, again suggesting secondary predation. Most of these nus Opuntia (Raske, 1966) and therefore, this fmding represents fecal deposits appear to be from jack rabbits (Lepus a significant host extension. Initial findings were published by californicus).] the author during a 1982 field survey of the China Lake Naval Weapons Center (R. May, 1982). Although the extent of this The size and shape of the scats, based on the analysis by O'Farrell infestation on populations of S. polyancistrus at this point can and Cameron, suggested three candidates: Neotoma lepida only be estimated, the form of infestation has been determined (woodrat), Ammospermophilus leucurus (antelope ground squir- by field collection and identification of the larvae and adult in- rel), and Thomomys bottae (pocket gopher). Neotoma was most sects from infested plants. Species identification was performed likely because: 1) the pellets are too large for A. leucurus (ground by the Department of Entomology at both the University of Idaho squirrel scats usually measure 6 -7 mrn in length), 2) A. leucurus and the University of California at Berkeley. cannot negotiate the spines and 3) T bottae droppings are usually found below ground level. Cameron and Rainey, in their 1965 According to Raske (1966) and Doyen (1982), adults of M study of woodrat habitat utilization (Cameron and Rainey, 1972), semipunctatum are nocturnal and are active at dusk. They typi- also found that the ground squirrel became tangled in cholla. cally remain on the plant at night where, during the spring and early summer, they also mate. Oviposition occurs at night, usu- Other Evidence ally between 10 and midnight. In this case, the female probably On several occasions, S. polyancistrus spine fragments and clus- lays her eggs at the base of the stem at the soil line. The number ters have been found within study plots amid the debris of of eggs oviposited is unknown, however, Raske (1966) reports woodrat dens, close to the den entry hole. They were found just that typically, under laboratory conditions, other species in this a few months after several S. polyancistrus stems, all within 30 genus may lay as many as 25 eggs which take about 35 days to meters (100 ft) of the den, had been consumed suggesting fur- hatch. Once the embryo is mature, it chews its way out of the ther that N. lepida is the predator. Additional evidence includes egg and then bores directly into the stem, tunneling up into the the analysis of the carcass remains which often times gives ad- cortex from below. After feeding on the succulent cortex, the ditional evidence as to the size of the predator. Based on the now much larger larvae exit at the base of the stem and form entry hole size and other geometrical considerations, the data pupal cells constructed with soil cemented together with mois- indicates that the predator is less than 20 cm (8 in) in length ture from the decaying plant and by their own secretions. This and can pass through cavities less than 6 cm (2 1 /4 in) in diameter. process usually occurs in the fall.

Infestation The exit holes produced by the larvae range from 10 to 15 mm It is attacked by an insect whose grub bores into the plant from in diameter. The pupal cells are usually 4 -6 cm (1 1/2 -21/2 in) in below length and approximately 3 cm (1 1/4 in) in diameter. If the stem is E.M. Baxter dislodged when the larvae are pupating, the pupal cells can still California Cactus, 1935 be found attached to the base of the stem. As many as four cells have been found on a single S. polyancistrus stem. ...Sclerocactus does not live well or long under cultivation, often because of rot that forms in the passages made by a boring Raske (1966) notes further that the larvae will overwinter in grub in the fleshy tissues of this species. the pupal cell, pupating in the spring. If the larvae do not pu- E. Yale Dawson pate in the spring, they will remain quiescent for another year, Cacti of California, 1971 pupating the following spring. Therefore, a small portion of the population may have a two year life cycle. The pupal stage Grubs have been seen in many dying S. polyancistrus, and in lasts about 15 or 20 days. The adult insect emerges in early some areas spine skeletons of this cactus are a common occur- spring but may remain in the pupal cell until late spring or rence. We do not know the identity of these grubs at this time, early summer. nor can we predict the final impact of this predation on major population sites. Massive damage to the cortex usually results from this infestation and the plant seldom recovers (Figure 13). In some cases, William A. Rhoads, et aí.,1979 the stem will become completely excised from the root system. If only partial damage has occurred, adventitious roots will sometimes emerge at the base of the damaged cortex (sometimes internally) and the plant will re -root. There is no quantitative data available on the percentage of stems that recover. Given May Sclerocactus 17 the likelihood of rot, as well as the possibility of further infesta- Other Infestation tion, the percentage is probably quite low. Based on the author's Although M semipunctatum accounts for the vast majority of own findings and that of MacDougal et al, (1915), most stems infestations, other infestations have been identified and are worth that do recover do not regain full vitality. noting. In 1980, larvae from the moth Distopasta yumaella (Keafott) were found in a stem within the Inyo National Forest Figure 14 depicts confirmed sitings of M semipunctatum, as east of Big Pine, California by the author. This insect is a scav- well as other infestations to be discussed below. An M enger and would normally enter a cactus after damage from semipunctatum infestation was discovered within the south- another agent. ernmost M semipunctatum study site shown in Figure 14. Eight confirmed and three suspected cases were found amongst the In 1987, infestation was found in two study plots in the south- 39 stems known to exist within the 16 hectare (1/4 square mile) ern Mojave. These two study plots, one near Red Rock State study plot. This was the first know infestation at this location Park, the other north of Victorville, have been under study by since the survey was begun here in 1980. It could not be deter- the author since 1976 with no previous recorded incidents of mined how extensive the infestation was within this population infestation. Specimens of larvae collected in 1989 were identi- as a whole. Affected stems had as many as 4 pupal cells at- fied by Dr. J. Powell at U.C. Berkeley (Powell, 1989) as the tached to the base of the plant remains. The pupating larvae parasitic Pyralid moth Yosemita graciella (Hulst). In 1984, Mark were in the latter stages of development. Two of the affected Donnell (1986) found a similar infestation near Barstow. These stems were dissected in the field. Specimen identification was recent fmdings are of great interest and may indicate that this performed by Dr. Paul Blom at the University of Idaho. form of infestation is spreading. Over one third of the stems within the author's two study plots were destroyed over a two year period. A third plot 96 km (60 miles) to the northwest was also found infested with this insect in 1988. Over 90 percent of the stems at this location were destroyed. Y graciella, according to Powell, is a species that is widespread over the southwestern deserts and has been recorded as feeding in Echinocereus and Coryphantha. Its recent appearance in the southern Mojave in association with S. polyancistrus poses new questions.

Field Identification Future field work relative to S. polyancistrus should, whenever possible, include information as to the presence and number of infested stems and "closed" carcasses within each population. During the months of March, April, and May, infested stems are more easily distinguished from healthy stems and manifest themselves in the following manner:

1) Loss of vigor - absence of apical growth, new spination, and buds during the peak of the growing period

2) Discoloration - the stem color is dull green; sometimes brown (when infestation is in the advanced stages)

3) Cortex damage - particularly at the base of the stem and/or the presence of one or more exit holes, 10 - 15 mm in diameter Figure 14. 4) Excision - stem excised from the root system In 1820, Thomas Say observed that the genus Opuntia was a host species for M semipunctatum, but recently Moneilema gi- 5) Pupal cell(s) - remnants of pupal cells in the soil, just below gas has been identified as a threat in the cultivation of Opuntia the base of the stem (These cells remain beneath the carcass phaeacantha and the South American Trichocereus spachianus several years after the death of the plant.) (Crosswhite, et al, 1985). The author has also identified a Moneilema sp. (probably M semipunctatum) within a 1986 field Strategies collected specimen ofSclerocactus spinosior in Utah. Pupal cells, But the cacti soon begin fighting back against their predators. typical of M. semipunctatum, were also observed at the base of Arthur Gibson & Park Nobel, 1986 Sclerocactus parvforus carcasses in Canyonlands National Park in 1988. This too represents a significant host extension. The Predation plays a major role in the ecology of Sclerocactus entire genus Sclerocactus may be a host for this insect, thus polyancistrus. Although predation impacts population den- explaining the presence of the many "closed" carcasses often sities and, in extreme cases, can be responsible for the found within other populations of this genus. expiration of the adult members of an entire segment or colony 18 Desert Plants 11(1) 1994

Figure 1. Figure 2.

Figure 10b.

Figure 10a.^ Figure 13. > May Sclerocactus 19

IMPACT OF PREDATION

1 PREDATION CONSUMED mum* SEED FRUIT DISPERSAL MIGRATION & REESTABLISHMENT 2 ,....10. SEEDLING CARCASSES SURVIVAL INCREASED REPRODUCTIVE CLUSTERS CAPABILITY (REESTABLISHMENT) DAMAGED STEMS 3 ABNORMAL STEM BRANCHING

4 HIGHER TURNOVER RESPONSE TO CHANGE (NATURAL SELECTION ENHANCED)

Figure 15. within a given population, the process in itself appears to have the spines and eventually fall to the ground in subsequent years. several short and long term benefits or offsetting effects. The Seed dispersal must also take place during transport and actual overall impact of predation is summarized in Figure 15. consumption of the fruit. The predator(s) responsible have not yet been identified, but rabbits and the woodrat are likely candi- Seed Dispersal dates. Birds are also a possible candidate and may disperse seeds The flower buds on the plant and the ripened fruit are con- over long distances.Isolated stems have been found by the sumed by the predator within 60 days following anthesis. The author within no apparent population. This process thus pro- bright red color of the fruit appears to be an attraction mecha- motes migration and reestablishment of the species. nism encouraging consumption and, consequently, seed dispersal. The Carcass -Seedling Relationship It is apparent that carcass remains are often found in associa- As soon as the fruit begin to set, partial or total consumption of tion with adult S. polyancistrus plants.There is strong evidence the still unripened fruit will occur. The number of unripened that the carcass plays a significant role in the ecology of this fruit consumed two weeks after anthesis was found to vary be- species by offsetting the impact of predation as well as ensuring tween 15 and 30 percent of the total yield. It would appear then successful reestablishment of subsequent generations of plants. that this is destructive, reducing the potential seed harvest. In 1986, a survey was conducted to determine the number of S. Typically, less than 10 percent of the stems are left with polyancistrus stems which are found in association with car- unconsumed fruit. By late summer, the surviving fruit dry and cass remains. Data was collected at five study plots (N =179) turn brown and by the following spring, less than one percent where evidence of moderate or intense predation was present. of the stems still have dried fruit. The still viable seeds from the It was found that a minimum of 40 percent of the stems were dried fruit which eventually reach the ground by this process adjacent to carcass remains. Carcasses will "decay" and disap- was calculated at one site to be less than 20 seeds per hectare pear after 25 years leaving some older plants without visible per year. carcass remains.

Seed dispersion is much more effective and is the first benefit The color of carcass spines fades as years pass, making it pos- of predation and occurs in two ways. The fruit of S. polyancistrus sible to estimate the ages of the carcasses by comparing them to are horizontally dehiscent, splitting horizontally at the base of spine colors of carcasses of known ages. It is not uncommon to the pericarpel. If the ripened fruit are detached from the areole, find carcasses of different ages adjacent to a stem suggesting several of the seeds will fall out. Similarly during the predation that more than one generation has existed at that site. process, approximately 10 percent of the seeds are left behind as the fruit are removed. This explains the usual deposition of Examination of recent (1 -3 year old) carcass remains will often fresh seeds in the apical region of the stems visible in late sum- reveal seed depositions and seedlings beneath or adjacent to the mer. These seeds become lodged in the ribs and areoles between spine skeleton. The establishment of these seedlings occurs in 20 Desert Plants 11(1) 1994 two ways. First, seeds trapped within the carcass remains even- Increased Reproductive Capability tually fall through the spine skeleton and reach the soil directly Approximately 10 -15 percent ofthe stems impacted by predation beneath the carcass remains. For open, type 1 carcasses which on any given year are only partially consumed. Most of these are torn apart, the seeds reach the ground during or immedi- surviving plants recover, but result in deformed or multi- headed ately after consumption of the plant. For closed carcasses with stems which have an increased reproductive capability. intact spine skeletons, the process of seed deposition is much slower, taking 1 -3 years. The carcass protects the juvenile plants The ability of a partially consumed stem to survive is depen- from intense sun and from predators. Consequently, lower soil dent upon several factors. Some have been found to survive temperatures provide a medium favorable for germination and with just 2 or 3 cm (tall) of the plant remaining above ground growth. level. Existing cover, bacterial infection, and the time of year that the wound is inflicted are factors which impact survival. Figure 16 depicts sequentially the growth stages of a plant from Some stems have repeated incidents of predation and may ulti- the seedling to juvenile stage as it emerges from within the mately be totally consumed over a 2 or 3 year period or longer. confines of an open carcass. Carcass remains adjacent to the wounded stems sometimes mini- mize recurring predation.

As the wound dries and scars over, "pups" or side branching appear. This new growth emerges from the first series of un- 4 damaged areoles adjacent to the wound area within 1 -2 yéars. APICAL 'CLUSTER As growth proceeds, the branches increase in size and the plant eventually becomes a multi- headed plant or cluster. These side .- / ,1 branches will then descend basally as is typical with the Cactacae 0 (see Buxbaum, 1950). Often, these "pseudoclusters" are mis- Q2 02 taken for separate, clustering stems. PSEUDODOUBLE It should be stated that S. polyancistrus normally does not ,--, branch. Only disturbances have been found to cause branching. Excessive or sudden sun exposure (stress), perhaps from the Li 522 cez Efè loss of a cover shrub, can burn sensitive epidermal tissue in the PSEUDOCLUSTER apical region and cause branching. The vast majority of stems, however, which exhibit this characteristic are the result of small mammal predation or (to a lesser extent) infestation. Infesta- 00 LI g tion also causes disfigured stems. LATERAL REGROWTH Abnormal stem branching is an important aspect of this species' ecology. The ability ofthe wounded stem to rapidly branch has significant adaptive advantages. Besides covering the exposed wound, the process serves to offset the effects that pre- ni dation has on population density. Multi- headed stems increase the reproductive capability of the plant (more heads = more CURL OVER (INFESTATION) flowers) and, consequently, a larger number of seeds are produced for reestablishment and population survival.

Figure 16. Response to Change It was noted earlier that in areas where predation is intense, there appears to be a higher turnover. Based on the Itis common to find numerous seedlings beneath carcasses author's 12 year study of S. polyancistrus growth rates within which accounts for numerous clusters of S. polyancistrus stems the study areas, it has been found that over 70 percent of the S. in areas where predation is intense. It is interesting when one polyancistrus stems do not live beyond 17 years of age (about draws an analogy between a population where predation is in- 20 cm (8 in) in height) due to light to moderate small mammal tense (large number of carcasses present) and populations where predation. In areas where predation is intense, this percentage predation is light. Intense predation appears to produce a greater appears to be higher, few living beyond the age of 10 years. In turnover. More seedlings survive to early maturity within the some of these populations, a relatively large number of juve- confines of carcasses, but more stems are ultimately eaten. nile, 3 -6 year old stems, less than 3 cm (1 1/4 in) in height, have Where predation is light, there are fewer carcasses and fewer been found capable of producing flower buds as compared to seedlings survive. Those that do survive, however, have a greater the norm of 6 -8 year old stems in areas with less predation. This longevity. The carcass, therefore, becomes a mechanism to off- may suggest that predation is selective, in this case enabling set predation by increasing the survival of the seedlings, and only those plants successful at producing buds /flowers at a juvenile plants. smaller, less visible age to survive. The author speculates that this apparent high turnover may also be beneficial to the species May Sclerocactus 21 by enhancing the natural selection process, enabling the species It has also been determined through the identification of field to respond more readily to change. collected specimens that the cerambycid beetle Moneilema semipunctatum is responsible for the majority of "closed" car- In summary, despite high mortality rates within some popula- casses found within populations of S. polyancistrus.Three tions, this species has probably evolved a balance in nature other less significant infestations have been noted and it is likely wherein light to moderate predation is beneficial. This is re- that others exist as well.Infestations occur most often in the flected in the seed dispersal process and, possibly, in the mecha- northern Mojave and are most readily identified in the field dur- nisms that the species has developed to relieve predation pres- ing the spring growing season. This species has probably devel- sures. Partially consumed plants form multi -headed stems which oped a balance in nature wherein light to moderate predation is increase the number of flower buds. Carcasses aid in seedling beneficial to the species survival. survival, ensuring successful reestablishment. Literature Cited Study Impact Baxter, E.M. 1935. California Cactus. Abbey San Encino As a product of this survey, an Area of Critical Environmental Press: Los Angeles, CA. p. 80. Concern (ACEC) within the Bureau of Land Management's California Desert Conservation Area was established in 1985 Benson, Lyman. 1982. The Cacti of the United States and for this species. The survey of the China Lake Naval weapons Canada. Stanford University Press, Stanford, CA. p. 747. center was also of great value. The study provided the Navy with a convenient reference source for the Navy in support of Brown, J.H. Gerald A. Lieberman, and W.F. Wengler. 1972. future environmental studies and land management studies. The Woodrats and Cholla: Dependence of a Small Mammal U.S. Navy at China Lake has already taken measures to avoid Population on the Density of Cacti. Ecology 53 (2): 310- impact to S. polyancistrus population areas. 313.

Summary Buxbaum, F. 1950. Morphology of Cacti. Abbey Garden It is hoped that the efforts described in this paper have contrib- Press, Santa Barbara. p. 51 uted to the understanding of this genus and species. In April, 1991, the author completed his fmal survey at the site where he Cameron, G.N. 1987. Personal correspondence. first found this species in habitat in 1976. At the base of "east hill ", one plant still remains from the original survey group. Cameron, G.N. and D.G. Rainey. 1972. Habitat Utilization by Gone was the dense Bromus rubens grass which used to cover Neotoma lepida in the Mojave Desert. Journal of the rocky slopes; gone were the Larkspur, Primrose and Desert Mammology 53 (2): 251. Asters and Marigold that had dotted the landscape. Severe drought and accompanying predation had taken its toll. But Crosswhite, Caròl D. and Frank S. 1985. Trichocereus as a beneath the carcass remains, scattered across the hillside, seed- Potential Nursery Crop in Southern Arizona with a lings of a new generation of S. polyancistrus emerge, awaiting Discussion of the Opuntia Borer (Cerambycidae: more favorable times. Moneilema gigas) as a Serious Threat to its Cultivation. Desert Plants 7 (4):195. Conclusion Several key aspects of the ecology of Sclerocactus polyancistrus Dawson, E. Yale. Cacti of California. 1971. University of have been discussed. The 15 year study included a record of California Press: Los Angeles, CA. p.84. growth rates for over 350 stems. Results indicate that, besides climatological factors such as rainfall, microhabitat is an Donnell, Mark S. 1986. Yosemita graciella: A Deadly important factor affecting growth. Some members of a given Parasite of Sclerocactus polyancistrus. Cactus and population are more than 50 years old. Average growth rates Succulent Journal 58:163. range between 1.3 -1.8 cm (.5 - .7 in) per year. Doyen, John T. 1982. University of California, Berkeley Carcasses provide valuable clues for the high mortality rates (Department of Entomology) Personal correspondence. and strategies for survival. Open carcasses are the result of small mammal predation which accounts for the vast majority of the Gibson, Arthur C. and Park S. Nobel. 1986. The Cactus S. polyancistrus deaths, particularly in the southern and lower Primer. Harvard University Press: Cambridge, MA p. 158 elevations of the Mojave. Predation also accounts for deformi- ties and abnormal appearances of some stems. Based on the MacDougal, D.T., E.R. Long, and J.G. Brown, 1915. End analysis of fecal depositions found in the carcasses, three can- Results of Desiccation & Respiration in Succulent Plants. didate predators were considered. Neotoma lepida, the desert Physiological Researches 1(6):289 -325. woodrat, appears to be most often responsible for initial stem damage. May, R.W. 1984. The Yellow -Spined Form of Sclerocactus polyancistrus. The Cactus & Succulent Journal 6:244. 22 Desert Plants 11(1) 1994

Continued from page 5 May, Richard W. 1982. Distribution and Status of Sclerocactus McPherson, E.G. and R.A. Haip. 1988. Tucson Arizona's polyancistrus on the Naval Weapons center - A Survey. urban vegetation: Past, present, and future. pp. 87 -91 Naval Weapons Center, China Lake, California Report # in: M. Pihlak (ed.), The city of the 21st century. NWC TP 6403. College of Agriculture and Environmental Design, Arizona State University, Tempe, AZ. Powell, Jerry A. 1989. University of California, Berkeley (Department of Entomology) Personal correspondence. Mullbarger, N. 1964. Ghosts of the adobe walls. Westernlore Press, Los Angeles, CA. 398 pp. Raske, A.G. 1966. Taxonomy and Bionomics of the Genus Moneilema. PhD Thesis University of California, Berke- Schaus, R. 1980. Arizona heritage: Richard F. Farrell, ley, California. 1847 -1931. Arizona cattle growers outlook. Vol 36(1): 24 -25. Rhoads, Willam, Susan A. Cochrane and Michael P. Will- iams. 1978. Status of Endangered and Threatened Plant Varney, P. 1980. Arizona's best ghost towns. Northland Species on Tonopah Test Range -a Survey. Part 2: Threat- Press, Flagstaff, AZ. 142 pp. ened Species. EG &G Inc., Goleta, CA. ReportNo. EGG 1183- 2356, p. 129. Wehrman, G. 1965. Harshaw: Mining camp of the Patagonias. Journal of Arizona History 6:21 -36. Rhoads, William, Susan A. Cochrane and Michael P. Will- iams. 1979. Status of Endangered and Threatened Plant Wheeler, C.C. (1858 -1941) Reminisces. Arizona Histori- Species on Tonopah Test Range - A Survey. EG &G Inc., cal Society Manuscript Collection, Tucson, AZ. Goleta, CA. Report No. EGG 1183 -2387, p. 81. Whitney, G.G. and S.D. Adams. 1980. Man as a maker Schmidt- Nielsen. 1964. Desert Animals Physiological of new plant communities. Journal of Applied Ecology Problems of Heat and Water. Oxford University Press, New 17:431 -448. York. pp. 144 -149.

Steenbergh,W.F. and G.H. Lowe. 1977. Ecology of the Saguaro:II. National Park Service. pp. 131 -165.