HOST-PLANT UTILIZATION BY PIERIS NAPI POPULATIONS IN CALIFORNIA (LEPIDOPTERA: PIERIDAE)

Oligophagous or polyphagous insects frequently exhibit ecologi-cally or geographically complex patterns o host-plant utilization. Such patterns have recently been documented or the butterflies Colias and Euphydryas editha Boisduval (Nyrnphalidae) (White and Singer, I974). This paper reports a similar situation among California populations o the Gray-Veined White, Pieris nail Linnaeus (Pieridae) and notes its potential significance in interspeciqc competition.

its potential significance in interspeciqc competition.
Pieris napi is a circumpolar species; in its extensive boreal and temperate range it has been recorded on a great variety o plants o the family Cruciferae. In California it is widely distributed in (  lower surfaces of the (appressed) cauline leaves (often near the point of attachment, as was also true on Barbarea verna), and on stems. One larva (3rd instar) was found, and two ovipositing females were observed. On 22 May, I6 additional ova and 5 larvae were ound at the same spot on the same plants, which were now mature and beginning to set fruit. Abundant evidence of larval feeding was found on the rosettes and lower cauline leaves, but the flowers and siliques were undamaged. Fourteen plants of flowering Barbarea orthoceras Ledeb. were found in a grassy meadow o.7 mile up the canyon, within the area where adult P. napi had been seen, but no ova, larvae, or eeding damage could be located on them. Lang Crossing, Nevada County. This population is at about 4500 feet in the South Yuba River canyon, at or near the upper altidudinal limit of Sierran P. napi; it is also univoltine. On 2 May I974, 3 ova of P. napi were found in a stand of 3o immature lrabis glabra and none on 2o flowering Barbarea orthoceras growing immediately adjacent to them. On z June the same stand was again searched, producing ova of P. napi, o of P. rapae Linnaeus, 3 of ,4. sara and of ,/lnthocaris lanceolata Lucas from ,4. glabra, and 8 of ,4. sara from B. orthoceras. An additional 34 ova of P. napi were collected from ,4. glabra elsewhere in the vicinity. Four larvae of P. napi were found feeding on rosette leaves of A. glabra, and two of A. sara on inflorescences and siliques of B. orthoceras. On 26 June four mature A. sara and two immature A. lanceolata larvae were found on siliques of A. glabra and one mature A. sara larva on siliques of B. ortkoceras. Many A. glabra plants showed feeding damage to the rosettes, but only seven P. raDae larvae were found; probably most naDi had already pupated. Pieris raDae, which is multivoltine, was flying in abundance at Lang Crossing on I9 July.
Despite the large number of Pierid species and individuals, the visible impact of feeding on the Crucifers at Lang Crossing was quite small. In particular, Arabis glabra plants often produced several hundred siliques on leafy, two-to four-foot stems.

DISCUSSION
There are ten species of obligate Crucifer-feeding Pierids in California, occurring in various combinations at different localities.
Some of these are spring-univoltine, a few are spring-bivoltine, and some are multivoltine. How they partition the available Crucifers among themselves may shed valuable light on the broad problems of niche differentiation, competitive exclusion, and species packing. Emmel (974) reported three species of inflorescence feeders (Pieris rotodice Boisduval and LeConte, ,4. sara and A. lanceolata) on drabis glabra in Riverside County, 2 June 973. Shapiro (974) described fiveand six-species Pierid assemblages in two high Sierran localities and concluded that competition is reduced by behavioral mechanisms (habitat selection).
Pieris napi occurs with up to six other Crucifer-feeding Pierids at the four localities described above. Based on adult collections, the Crucifer-feeding Pierid faunas of the four localities are (inflorescence feeders are marked "I"; others primarily leaf feeders): San Andreas Reservoir: P. napi, P. rapae, A. sara (I), Euchloe ausonides Lucas (I).
Not all of these breed in the same microhabitats. At Lang Crossing, for example, P. sisymbrii and E. hyantis are found only in exposed rocky situations and appear to breed only on 8trel)tanthus, and are thus not in competition with the woodland species. In examining the fauna of particular plants in particular habitats, the division of Pierids into a leaf-feeding and an inflorescence/silique feeding guild seems paramount. At Mix and Gates Canyons, where Barbarea is obviously in "short supply" and frequently completely defoliated, at least one species from each guild (P. na/)i, A. sara) can occur on this plant. At Lang Crossing the combined visible impact of four species two of each guild--on Arabis glabra is so small that it is tempting to speculate that the populations are regulated by other factors below the level at which interspecific competition would be significant.
Despite the potential for interspecific competition, at each of the study areas one plant received the bulk of the attention from Pierids while another appeared largely or wholly unutilized. It remains to be seen whether this reflects nutritional or toxicological unsuitability of certain Crucifer species. (In unpublished laboratory studies, Shapiro and F. Slansky (lers. comm.) have found variation in the suitability of native and weedy Crucifers as hosts of Pieris ra/)ae Psyche [September-December and other species.) It is also possible that host selection by ovipositing emales is closely tied to host plant phenology, and that flowering condition o the plants, and possible correlated changes in mustard oil concentrations, may determine the patterns observed.