«Spider fauna of flooded rice field s in northern California. J. Arachnol., 16 :331-337. SPIDER FAUNA OF FLOODED RICE FIELD S IN NORTHERN CALIFORNI ...»
Oraze, M. J., A. A. Grigarick, J. H. Lynch and K. A. Smith. 1988. Spider fauna of flooded rice field s
in northern California. J. Arachnol., 16 :331-337.
SPIDER FAUNA OF FLOODED RICE FIELD S
IN NORTHERN CALIFORNI A
Michael J. Oraze, Albert A. Grigarick,
Joseph H. Lynch and Kirk A. Smith
Department of Entomology
University of Californi a
Davis, California 95616 US A
INTRODUCTIO NRice, Oryza sativa L., was introduced into California in 1912 and is now grow n annually on about 161,880 ha (400,000 acres). Rice production is a major industry in the Sacramento Valley where more than 90% of the rice acreage in th e state is located. A rice field is a complex agroecosystem, containing man y aquatic, semiaquatic, and terrestrial species. Spiders are well represented among the many predators found in this habitat. They feed mostly on insects and ma y contribute in reducing pest levels. Lower pest densities have been attributed to spider activity in rice fields of Asia (Kiritani 1979) and other agroecosystem s worldwide (Riechert and Lockley 1984).
Numerous surveys of spiders have been conducted in the rice growing region s of Asia (Barrion and Litsinger 1984). However, little is known about spider s associated with rice in the United States. Preliminary surveys have bee n conducted in Texas (Woods and Harrell 1976) and Arkansas (Heiss and Meisc h 1985), but no attempt has yet been made to formally describe the California ric e field fauna. This paper identifies the spiders collected from the levees and floode d paddies of several California rice fields over a three year period.
MATERIALS AND METHOD
Sampling began in 1983 at the following locations : the Beck ranch near Modesto (Stanislaus County), Van Dyke ranch near Natomas (Sutter County),
and in the Lattemore seed-field section of the Rice Expriment Station near Bigg s (Butte County). Sampling efforts in 1984-1985 were limited to the Biggs site afte r it was determined the three areas yielded nearly identical results with respect t o common species.
The California rice field habitat and associated common vegetation wer e described by Barrett and Seaman (1980). Notable differences in vegetation among sampling sites used in this study included dense populations of common cattail (Typha latifolia L.) and bermuda grass (Cynodon dactylon (L.) Pers.) on th e levees at the Modesto and Natomas sites, respectively, but not at Biggs.
Monochoria (Monochoria vaginalis (Burm. f.) Presl.) and toothcup (Rotala indica (Willd.) Koehne) were restricted to and abundant in the paddies at th e Rice Experiment Station.
Wide mouth Mason jars (11 cm deep and 7.5 cm in diameter) served as pitfall traps. They were inserted into plastic sleeves that were permanently buried in th e levees flush with the soil surface. Ten traps were installed at each site at ca. 8 m intervals in an alternating pattern (north side, center, south side, etc.) along the length of a levee in selected fields. One hundred and fifty ml of 95% ethylene glycol plus 5% liquid detergent was added to each trap. After seven days, th e traps were collected. The contents were filtered through a USA Standard Testin g Sieve No. 40 and stored in 70% EtOH. This procedure was repeated monthl y throughout the growing season (May-September).
Floating sticky traps were made by cutting white styrofoam into triangula r wedges 61 cm long, 4.5 cm high, with bases of 9 cm. A thin coat of Sticke m Special" (Seabright Enterprises Ltd. ; Emeryville, CA) was brushed on the upper surfaces. Five traps were placed in each field and positioned equidistant from on e another (ca. 34 to 92 m apart depending on field size) along a transect connectin g the NE and SW corners of the field with the end traps being placed 2 m from th e margins. They were held in place with green bamboo stakes in a manner that allowed the traps to move vertically so contact with the fluctuating water surfac e could be maintained. The stakes also served to mark the position of the traps.
After seven days, the traps were collected and the spiders identified (to specie s when possible) with the aid of a 10X hand lens. The sampling schedule was th e same as that for the pitfall traps.
Companion samples were taken in 1983 with a UC-VAC® suction device (Summers et al. 1984) to estimate absolute densities and determine the nature an d extent of any bias associated with pitfall and sticky-trap sampling. Ten sample s ca. 10 m apart were collected both on the levees and in the paddies. A circular unit-area-sampler, enclosing 0.093 m2 (1 ft. 2 ) and standing 38 cm (15 in.) high, was placed in the general vicinity of the pitfall and floating sticky traps. The enclosed substrate and vegetation were vacuumed for ca. 90 s. Sampling wa s conducted between 1200 and 1400 hours. Samples were immediately placed in a cooler with ice for transport, and later processed in Berlese funnels for 48 h.
More than 30,000 specimens were collected in the survey. Species that were taken at all sampling sites in every year—representing 11 families, 22 genera an d 28 species—are listed in Table 1. They have been ranked as 4 common, 2
Table 1.-Spiders collected in northern California rice fields (1983-85).
a = L, levee ; P, paddy. b = R = rare ( I%) ; 0 = occasional (1-5%) ; C = common ( 5%). Species frequencies determined b y averaging counts from UC-VAC (levee) and pitfall-trap samples.
occasional and 22 rare species. All 28 species were collected on the levee s (including vegetation) but, only 10 of the species were taken in the paddy.
Apparently many of the levee species were incapable of inhabiting an aquati c microhabitat. Only four species—Pardosa ramulosa (McCook), Pirata piraticu s (Clerck) and two linyphiid spp.—were common in the paddy. Other spider specie s occasionally found in the paddy were generally limited to the paddy margins lat e in the growing season after the crop canopy had filled in enough to allow plantto-plant movement and provide adequate sites for web attachment. The fou r common paddy spiders also dominated the levee fauna.
The number of taxa recorded are generally lower than those reported for othe r surveys (Paik and Kim 1973). This can be attributed in part to our exclusion o f
Table 2.-Relative abundance in percent composition of the major spiders in northern Californi a rice fields.
a = 1983 ; b = Pitfall or sticky-trap catches from 1983-1985.
transient species (species not collected at every sampling site in every year, usually represented by a single specimen). Even so, comparisons of this type may b e misleading, as great differences exist among the surveys in terms of the exten t and methods of sampling. For example, Woods and Harrell (1976) collected 75 2 specimens from a single 14.8 ha (37 acre) field during one growing season. I n contrast, Barrion and Litsinger (1984) collected 13,270 specimens from 1 7 localities over three years, and Okuma (1968) collected 1,487 spiders from 2 2 localities during a 10 day period. Furthermore, Heiss and Meisch (1985) sample d with an aquatic net and metal dipper but Okuma and Wongsiri (1973) utilized a sweep net and observations.
In spite of these differences, three families : Araneidae, including Tetragnathidae ; Linyphiidae, including the Erigoninae (Micryphantidae) and Lycosida e dominated the spider fauna in all but one of the surveys of rice fields cited in thi s paper. In addition, the relative abundances of these families changed wit h latitude. In semitropical rice-growing areas, such as Taiwan, Thailand and th e Philippines, araneids dominated (Okuma 1968 ; Chu and Okuma 1970 ; Okum a and Wongsiri 1973 ; Barrion and Litsinger 1984) while lycosids were mor e abundant in temperate regions such as Korea and the United States (Paik an d Kim 1973 ; Woods and Harrell 1976 ; Heiss and Meisch 1985 ; present study).
Lycosids and araneids were also abundant in the rice fields surveyed in Japan, although the fauna was dominated by two theridiid spp. (Paik and Kim 1973).
Pardosa ramulosa was dominant in numbers on the levees and in the paddies (Table 2). It comprised ca. 58 and 68% (UC-VAC samples) of the fauna in thes e respective areas. The two lycosids, Pardosa ramulosa and Pirata piraticus, together constituted ca. 80% (UC-VAC samples) of the paddy spiders. The y appeared to be well adapted to the water surface where they quickly ran about or remained motionless for long periods. They occasionally went underwater by crawling down emergent vegetation or debris. Other paddy spiders, although capable of limited locomotion on the water surface, spent most of their time o n vegetation or in webs constructed among the paddy plants. Linyphiids were als o seasonally abundant. However, their contribution to total spider biomass over th e growing season was relatively small, compared to that of the common lycosids, because of their small size and ephemeral occurrence.
Sticky and pitfall-trap samples probably overestimated Pardosa ramulosa while underestimated linyphiids and Pirata piraticus abundances compared to UC-VAC samples (Table 2). Although fewer spiders were collected with the UC-VAC, these data were probably more accurate in estimating relative abundances for the majo r species. The UC-VAC was not used more extensively because of the much greate r
relative time and effort it required. Foliage dwelling species such as Tibellus oblongus (Walckenaer) were only collected with the UC-VAC whereas nocturna l ground dwellers such as the dictynids and gnaphosids were limited to pitfall traps.
This illustrates the importance of utilizing multiple collecting techniques in faunal surveys of spiders.
The major paddy spider species exhibited a seasonal succession in relative abundance during the growing season (Fig. 1). The linyphiids dominated th e spider fauna in the paddies shortly after flooding. Their abundance wa s associated with the spring ballooning period when they arrived in massiv e numbers. Unlike the other two major paddy species, the linyphiids do not appea r to be specifically adapted for, or restricted to, aquatic environments.
Pirata piraticus is distributed throughout Europe and north of the 35 `h paralle l in North America. It is associated with swamps, marshes and the shores of lakes, ponds and streams (Wallace and Exline 1977). It became a major component o f the paddy fauna late in the growing season.
Pardosa ramulosa is found throughout California. Its range extends E throug h southern Nevada into the SW corner of Utah and S into northern Mexico. I n California it is one of the dominant lycosids at elevations below 300 m (Hydor n 1977). It is associated with mesic habitats such as salt marshes (Garcia and Schlinger 1972 ; Greenstone 1980), sewage oxidation ponds (Hydorn 1977), irrigated lawns (Van Dyke and Lowrie 1975) and irrigated crops (Leigh and Hunter 1969 ; Yeargan and Dondale 1974 ; Hickle 1981). The prevalence of P.
ramulosa in rice and other irrigated crops in California is probably related to th e seasonal compression of suitable habitat. As drying begins in the spring an d continues through the summer these spiders are probably forced to aggregat e where moist conditions persist. Irrigated cropland, particularly rice, which i s typically continuously flooded from May through September, offers such a refuge. Rice culture in California resembles the native habitat of some areas tha t existed before the advent of flood control and irrigation projects when man y parts of the Sacramento and San Joaquin Valleys were annually flooded fro m snowmelt. Rice fields probably represent the functional equivalent of th e numerous vernal ponds and marshes that were presumably utilized by P.
ramulosa in its pristine environment, but differ by extending moisture availability, which is essential for this species (Hydorn 1977), throughout the summer. For a large part of the growing season, this native natural enemy is actually favored by, and more abundant in, rice (an introduced annual crop) than in adjacent untilled
336 THE JOURNAL OF ARACHNOLOG Y
border areas (Oraze et al. unpublished data). Because of its abundance in, and preadaptation to, the rice field environment, Pardosa ramulosa appears to be th e spider most likely to contribute to a level of biological control of one or mor e insect pests in this agroecosystem. The impact of this spider on selected prey species in rice will be presented in a subsequent paper (Oraze and Grigarick 1988).
Our sampling did not include any wild rice (Zizania aquatica L.) fields. This crop supports more vegetative growth and is usually produced earlier in the year (February-July) than conventional rice. We suspect that these cultural difference s may cause minor differences in the respective spider faunas, and a comparative study would be of value.
We thank M. O. Way for suggesting the project ; O. G. Bacon, W. Johnson, an d B. Brandon for technical assistance ; R. Stein for collecting and counting samples ;