Scaptomyza nigrita Wheeler (Diptera: Drosophilidae), a Leaf Miner of the Native Crucifer, Cardamine cordifolia A. Gray (Bittercress)

The biology of Scaptomyza nigrita on its host plant, a native crucifer (Bittercress) in the Rocky Mountains, is described. Development of each stage in the life history was studied both in the field and in the laboratory. This is the first documentation of a host for S. nigrita. We examined the activity of adult flies in two adjacent habitats, sun and adjacent willow shade. Adult flies were more abundant on bittercress plants in sunexposed versus in shaded areas, and were most active from mid-day to late afternoon. Female flies were significantly larger than male flies, but there were no differences in size of adults between the two habitats. Larval damage to bittercress is generally much greater on plants in sunny areas than on those in the shade, possibly due to the increased activity of ovipositing flies in sun-exposed areas. Spatial variation in damage by insect herbivores is common, both within host plant patches (Harcourt, 196 1 ; Jones, 1977; Thompson, 1978; Free and Williams, 1978,1979; Courtney and Courtney, 1982; Collinge and Louda, 1988a) and among patches (Janzen, 1975a, b, c; Stanton, 1975; Whitham, 1978, 1980; Parker and Root, 198 1; Faeth et al., 198 1; Louda and Rodman, 1983a, b). If this variation is consistent and repeated over time, with certain plants generally suffering higher levels of herbivore damage than others, then there are potentially significant implications for the population dynamics of both the insect herbivore (Janzen, 1975a, b, c) and the host plant (Louda, 1982a, b, 1988). Larval Scaptomyza nigrita Wheeler (Diptera: Drosophilidae) intensively mine the leaves of Cardamine cordifolia A. Gray (Bittercress, Cruciferae) at study sites near Rocky Mountain Biological Laboratory in west-central Colorado. In this interaction, and for insect herbivory on bittercress in general, there is a characteristic spatial pattern of feeding damage. Plants in the sun are generally much more heavily damaged by chewing and mining insects than are those in the adjacent willow shade (Louda and Rodman, 1983a, b; Collinge and Louda, 1988a; Louda, 1988). Since leaf miners are endophagous, the choice of oviposition site by the adult female will largely determine where the larvae feed. Most larvae remain on the same plant throughout development, although some late instar larvae migrate up the stem to mine new leaves (Hering, 195 1; Collinge, pers. obs.). Consequently, we examined the patterns of adult fly activity and fly abundance in sun-exposed and shaded habitats. We censused and conducted field observations on each life I Mailing address: Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138. Accepted for publication 24 August 1988. 2 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY history stage of Scaptomyza nigrita. In addition, we reared S , nigrita from bittercress leaves in the laboratory from the egg to the adult stage, providing the first definitive evidence on host plant relations for this species. Here we present our field and laboratory observations on the life history of this leaf-mining drosophilid, as well as its distribution on and damage to bittercress within and between habitats. Materials and Methods STUDY SITES: Our research was conducted in 1985 and 1986 at Rocky Mountain Biological Laboratory, Gothic, Colorado, 9.6 km northwest of Crested Butte, Gunnison County (38"57'56"N, 106O59'12"W). The vegetation in this area is montane and typical for the region (Langenheim, 1955, 1962). We used two specific study sites: Site 1 is at 3040 m, 1.2 km north of Gothic along Copper Creek ("Main Site" of Louda and Rodman, 1983a, b). Site 2, located at 2985 m, is 3 km west of Gothic on Forest Service Trail 40 1 ("ABP" of Collinge and Louda, 1988a, b, 1989). Both sites were large (30 x 10 m) snow melt seeps with dense stands of bittercress in and around them. Density at Site 2 was 355 stems/m2. Plant density was not measured at Site 1, however visually it was similar to Site 2. The laboratory portions of the research were completed at Rocky Mountain Biological Laboratory. LEAF MINER: Scaptomyza nigrita Wheeler was first described in 1952 from "several hundred" specimens caught in Pasadena, California (Wheeler, 1952). The flies were collected from a lawn containing grass and clover, and Wheeler believed that the host plant was clover. However, no larvae or mines were found, and he was unable to successfully rear any flies in the laboratory (Wheeler, 1952). In addition, "a few" specimens of S. nigrita were found in Wyoming, and in the mountains near Malad, Idaho (Wheeler, 1952). Later, the known distribution was extended by identification of these flies in collections from Alaska and Canada (Wheeler and Takada, 1 966). HOST PLANT: Cardamine cordifolia A. Gray (Bittercress) is a native, herbaceous perennial crucifer that grows in wet, often shaded areas along streams from Wyoming to Idaho, south to New Mexico and Arizona (Harrington, 1954). Bittercress is very common in moist and shaded areas in our study region. If the soil is wet enough, bittercress can grow in the sun as well as in the shade of willows, Salix spp., or spruce, Picea engelmanni Parry (Chew, 1977; Louda and Rodman, 1983a, b). In this study, sun and adjacent shade represent habitat treatments at each field site. PROCEDURES: We estimated relative adult fly densities in the field by using tenminute observations on patches of bittercress (= 50 ramets) at Site 2. Observations were conducted in both sun-exposed and adjacent willow-shaded habitats (= 15 m apart) between 26-28 May 1986 ( n = 6/habitat). On 15 June 1986 we followed the daily activity pattern of adult flies in both open sun and willow shade. In this case, 5 minute observations were conducted every 2 hours from 0730-2030 on a patch of bittercress in each habitat. In addition, adults were collected in traps at both sites during the peak oviposition period in early June 1986. Yellow paper rectangles (6.5 x 11.5 cm) were spread with Tangletrap@ (Jones et al., 1986) and placed horizontally on a wire stake at 35 cm above the ground. Stakes with sticky traps were placed immediately adjacent to ramets of bittercress and left for 24 hours ( n = 5/habitat/date). Habitat densities were sampled twice at Site 1 (4 and VOLUME 6 2 , NUMBER 1 3 6 June) and three times at Site 2 (3, 5 and 7 June 1986). The sticky traps were returned to the lab and adult drosophilids were counted. Traps were also set out and checked every two hours on 15 June 1986 at Site 2 in conjunction with the observations on the activity pattern. A subsample (25 females, 7 males) of adult flies collected from the sticky traps in each habitat was examined using a dissecting microscope. We determined the sex of the adults, and for females we measured: thoracic width, as an indication of adult size; number of eggs; and length of eggs. For males we measured only thoracic width. Data from both field sites were pooled for analysis because of small sample sizes. We collected samples from bittercress plants (n = 15 ramets) from both study sites at 3 week intervals throughout both field seasons to quantify larval occurrence and feeding damage. Equal numbers of samples were taken from sun-exposed and adjacent willow-shaded habitats. Ramets were chosen from different plant clones of the same height and reproductive condition in both habitats. Ramets within clones were selected randomly and cut below the lowest leaf. After measuring the leaf area eaten by the leaf miners (see Louda, 1984), we preserved the larvae in 95% ethanol and later measured them using a dissecting microscope. Bittercress leaves containing leaf miner eggs or larvae were collected from sunexposed areas at both sites and brought to the laboratory for rearing. Forty leaves were collected in 1985 (between 22 June-9 July) and 79 leaves were collected in 1986 (between 4-1 8 June). We collected leaves earlier in 1986 to obtain leaves with eggs, but without developed mines. We recorded the number of eggs and larvae on each leaf and the leaf length. Rearing was accomplished by wrapping cotton around the upper leaf petiole and inserting the petiole and cotton into a two-dram vial filled with water. The vials were placed horizontally in a rack. A rearing chamber was made around the leaf blade by placing a small plastic bag over the leaf, and securing it to the top of the vial with a rubber band. This set-up maintained leaf turgidity and allowed continued larval feeding. As leaves senesced, they were replaced with new leaves and the larvae were transferred manually. Transfer did not harm larvae since soon after being placed on new leaves, the larvae formed new leaf mines. Upon pupation in the leaf mine, each pupa was moved to a 5 cm diameter covered petri dish containing a piece of moist filter paper and a fresh bittercress leaf. The filter paper was kept moist until the adults emerged (about 9 days later). STATISTICAL ANALYSES: Relative densities of adults in sun and shade habitats, as well as adult sizes and numbers of eggs in females, were analyzed using Student's t-test. The diurnal pattern of adult fly activity in sun and shade was compared using the non-parametric Sign Test (Campbell, 1974). The relationship between the number of individuals per leaf and leaf length was determined using Pearson's product-moment correlation. Results and Discussion ADULTS: Adult flies were at least twice as frequent on plants sampled in sunexposed areas than on plants in willow shade (Table 1A). This pattern was consistent between sites (Table 1A). Adults were most active from midday to late afternoon in midJune (Table 1B). The observations can be divided into three periods: early morning (0730-1 130), midday (1 130-1 530) and late afternoon 4 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY Table 1. Number of adult Scaptomyza nigrita observed in timed observations on bittercress and caught in sticky traps in the sunny habitat and in the adjacent shaded habitat. A. Habitat activity oattern (1986)