1993. The Journal of Arachnology 21 :29—39 THE NATURAL HISTORY OF THE CALIFORNIA TURRET SPIDE R ATYPOIDES RIVERSI (ARANEAE, ANTRODIAETIDAE) : DEMOGRAPHICS, GROWTH RATES , SURVIVORSHIP, AND LONGEVITY Leonard S. Vincent: Division of Biological Sciences, Fullerton College, 321 E . Chapman Avenue; Fullerton, California 92632 USA ABSTRACT. A large and dense population of over 500 burrows of Atypoides riversi in a 2.0 x 3.2 m area was monitored for two years to indirectly determine demographics, growth rates, survivorship and longevity o f the spiders . Twelve size classes of spiders were designated by correlating spider size to burrow size. All size classes were present simultaneously throughout the year . Variable growth rates were recorded for spiders in eac h size class, and survivorship was lowest for spiders in the smallest size classes . It is estimated, based in large part on growth rates, that A . riversi can live at least 16 years in the field. The only long term comprehensive study of (1991) used a similar approach to study Geoly- the genealogy and demography of a large popu- cosa turricola . lation of mygalomorph spiders concerns the Aus - Additional natural history information on var- tralian ctenizid Anidiops villosus (Rainbow) (Main ious antrodiaetids can be found in Atkinso n 1978). In another study, Marples & Marple s (1886a, 1886b), Coyle (1971, 1986), Rivers (1972) observed a population of several specie s (1891), Smith (1908), Vincent (1980, 1985, 1986), of New Zealand ctenizids for six years . The dem- and Vincent & Rack (1982) . ographics of burrowing wolf spiders have bee n examined in detail by McQueen (1978, 1983) , METHOD S Humphreys (1976) and Miller & Miller (1991). Study sites.—Two study sites at the University Herein I describe the natural history of the of California's Blodgett Forest Research Station, fossorial mygalomorph spider Atypoides riversi located in the American River watershed on the O. P- Cambridge, the California turret building western slope of the Sierra Nevada in El Dorad o spider (Rivers 1892). Unlike the long term field County, approximately 10 miles west of George - studies of Main and the Marples, my objectives town and at an elevation 1275 m were chosen were to determine, in a two-year period, the dem - for their high density of burrows and uniform ographics, growth rate, survivorship, and lon- ground cover. The population dynamics of spi- gevity of A . riversi. Unlike Main (1978) and Mar- ders in both areas were similar; therefore, this pies & Marples (1972), who measured burro w paper reports on only one. Population data for and door diameters but did not correlate thes e the other study area and vegetation descriptions measurements to spider size, I measured and cor - for both are in Vincent (1980). related burrow entrance size to spider size . The study site measured 2.0 x 3 .2 m and con- McQueen (1978), Humphreys (1976) and Mille r sisted of 160 20 cm square quadrats formed by & Miller (1984) found positive correlations for a grid system composed of nylon string and certain burrowing wolf spiders . Decae et al. wooden stakes . The stakes were placed at 20 cm (1982), studying the burrow structure of a cten- increments around the perimeter of the plot, an d izid, also found a positive correlation between string was placed on or slightly above the groun d carapace length and burrow diameter but did no t connecting facing stakes. Ground cover was associate this with longevity. Using the correla- mostly pine and cedar needles with occasional tion and following all burrow size changes through pine seedlings present during the spring and sum - the two year period, I derived life history infor- mer months . A barbed-wire fence enclosed the mation comparable to following a single cohort study area to exclude deer and other large ani- ofA . riversi through its long life . Miller & Miller mals. 29 30 THE JOURNAL OF ARACHNOLOG Y Correlation of spider size and burrow entrance with silk]; (4) missing [burrow could not b e size.—A series of 15 non-metric ball bearings found]; (5) old [burrow was in a state of disrepair, ranging in diameter from 2/32 inch (1 .59 mm) the turret was stiff and/or torn or non-existent] . to 16/32 inch (12 .7 mm) in increments of 1/3 2 These and other relatively rare burrow condi- inch (0.79 mm) were hard-soldered to thin singl e tions are discussed in detail in Vincent (1980) . fiber wire "handles" . These ball bearings were Data collection dates.—Burrows were initially then used to measure the internal diameter of censused 5—23 September 1976 . On 22—24 April spider burrow entrances. The internal diameter 1977 new burrows and burrow conditions for a was considered equivalent to the diameter of th e random sample (n = 153) of previously censuse d ball bearing that fit (or came the closest to fitting ) burrows of sizes 3 through 11 were recorded . the narrowest section of the tapered burrow en - Also, burrow conditions for all burrows of sizes trance. For convenience, burrow size classes wer e 12, 13, and 14 were recorded (n = 13). The ran- designated by the numerators that fit the entranc- dom-sample size for each size class was deter- es (sizes 2—16) . Attempts to measure burrow en- mined so that the standard deviation of the es- trances accurately with a caliper or ruler prove d timated proportion would be no greater than 0 .30. to be difficult and damaging to the flexible an d Confidence intervals for the true proportions wer e fragile entrance . calculated by a formula given in Bickel & Daksu n Six to 14 burrows (n = 128) representing each (1977, formula 5 .1 .13), and modified to account size class were arbitrarily chosen near the study for sampling without replacement (Cochran 1977 , site for measurement (Vincent 1980) . After each sec. 2.15). During 28—31 July 1977, new burrows burrow was measured, the resident spider wa s were mapped and recorded, and previously re - dug from its burrow, anesthetized by cooling with corded burrows were measured again . On 20 Au- crushed ice (large spiders) or CO2 (small spiders), gust 1977 a random sample of burrows presen t and measured . Spiders were measured with a on 28—31 July, 1977 was censused to confirm the stereomicroscope fitted with an ocular microm- presence of spiders in the burrows previously eter accurate to 0 .039 mm. Measurements of sized. Sample size and confidence intervals were maximum width of both the carapace and ster- determined as above for the April 1977 random num were correlated to the internal diameter o f sample. All burrows present on 28—31 July 197 7 the burrow entrance. were recensused 19—20 May 1978 to see if the y Observation platform.—A portable observa- contained spiders . Burrows that contained spi- tion platform consisting of a 2 x 1 .33 m sheet ders on 19—20 May 1978 were recensused and of plywood was supported approximately 12 c m remeasured 3—6 August 1978, and all burrows in over the plot by planks and blocks . The leading odd-numbered rows were observed at night wit h edge of the platform coincided with the trailing a dim unfiltered flashlight as a further check o n edge of the row being examined to allow a de- spider presence. If a spider was not immediately tailed view of one 20 cm square quadrat . After visible in its burrow, I waited several minute s examining all quadrats in a row, I advanced the for it to appear; if it still did not appear, I con- platform to the trailing edge of the next row, etc. sidered the burrow abandoned. Again, all new Since A. riversi, like some other fossorial my- burrows were recorded and mapped . galomorphs, is sensitive to vibrations, successful Survivorship calculations.—Survivorship of A . observations necessitated moving slowly on the riversi was indirectly determined by subtractin g platform. Adjusting the platform caused some from the initial number of burrows censused th e spiders to retreat temporarily down their bur - number of burrows missing or considered "old" rows. during each consecutive census . Survivorship of Burrow observations.—The position of each eggs and emerging spiderlings is unknown . burrow was noted and its entrance diameter was measured . Burrows with flexible and freshly silke d RESULTS AND DISCUSSIO N turrets were measured for size class designation . Correlation of spider size and burrow entranc e The following burrow conditions were recorded: size.—Burrow entrances ranged from 3/32 inc h (1) occupied [spider was seen in its burrow] ; (2) (2.38 mm) through 14/32 inch (11 .06 mm). A abandoned [burrow appeared in use, but no spi- regression of spider size (carapace widths) agains t der was detected during the immediate obser- burrow entrance size was highly significant (P < vation period, approximately five minutes] ; (3) 0.001) (Fig . 1). closed [entrance was folded closed and sealed Assumptions .—In estimating the following VINCENT—NATURAL HISTORY OF ATYPOIDES RIVERSI 3 1 2 4 6 8 10 12 1 4 Ball Bearing Siz e Figure 1 .—Regression of ball bearing size on carapace width (Y = 8 .3X -3 .1, R2 = 0.934, P < 0 .001) . demographics, survivorship, growth rates and during most of 1977, recovered only adult male longevity for A . riversi, I assume each burrow spiders and only during the fall mating season . has had only one occupant who had enlarged an d The above observations suggest that A. riversi maintained it over time, that burrows present does not leave its burrow to enter another burro w and in good shape contain a living spider, and to evict its resident in a competitive interactio n that missing burrows are a measure of spide r (as in Riechert 1978), to search out larger vacated mortality.