INSECTS ASSOCIATED WITH THE FLOWERS OF TWO SPECIES OF MALACOTHRIX { A ^ TEKACEKE ) ON SAN MIGUEL ISLAND

The insects associated with Malacothrix incana (Nutt.) T. & G. and M. implicata Eastwood on San Miguel Island were sampled as part of a general analysis of hybridization between the two species on the island (Davis and Philbrick, 1986). On San Miguel Island, M. incana is widely distributed on unstabilized and stabilized sand dunes on slopes near the ocean or on sandy substrate on the upper surfaces of the island including the slopes of San Miguel Peak and Green Mountain. In contrast, M. implicata is generally restricted to the slopes near the ocean or the walls of canyons above the ocean. Hybrid plants were found only where M. incana and M. implicata were growing in a common area and constituted less than 1% of the total number of the three forms in these areas. Hybrid plants were most frequent on the slopes above Cuyler Harbor and above Tyler Bight. Plants of M. implicata are spreading or erect perennials with large heads containing up to 80 florets. The ligules are white and have a purple stripe on the abaxial surface. Plants of M. incana are perennial and become mound-shaped after several years of growth. The large heads contain up to 100 florets with yellow ligules. The hybrid is also perennial and has large heads with pale yellow florets whose ligules often bear a reddish stripe on the abaxial surface. During our visit to San Miguel Island in May, 1984 a majority of the plants of M. incana, M. implicata and the hybrid were in full bloom. Numerous open heads were present on plants of both species and one plant of M. incana had over 200 open heads. Herbarium label data indicate that the blooming period of M. incana generally coincides with that of M. implicata on other Channel Islands as well as on San Miguel Island and that flowering of both species reaches a maximum in the months of May and June.

The insects associated with Malacothrix incana (Nutt.)T. & G. and M. implicata Eastwood on San Miguel Island were sampled as part of a general analysis of hybridization between the two species on the island (Davis and Philbrick, 1986).On San Miguel Island, M. incana is widely distributed on unstabilized and stabilized sand dunes on slopes near the ocean or on sandy substrate on the upper surfaces of the island including the slopes of San Miguel Peak and Green Mountain.In contrast, M. implicata is generally restricted to the slopes near the ocean or the walls of canyons above the ocean.Hybrid plants were found only where M. incana and M. implicata were growing in a common area and constituted less than 1% of the total number of the three forms in these areas.Hybrid plants were most frequent on the slopes above Cuyler Harbor and above Tyler Bight.
Plants of M. implicata are spreading or erect perennials with large heads containing up to 80 florets.The ligules are white and have a purple stripe on the abaxial surface.Plants of M. incana are perennial and become mound-shaped after several years of growth.The large heads contain up to 100 florets with yellow ligules.The hybrid is also perennial and has large heads with pale yellow florets whose ligules often bear a reddish stripe on the abaxial surface.
During our visit to San Miguel Island in May, 1984 a majority of the plants of M. incana, M. implicata and the hybrid were in full bloom.Numerous open heads were present on plants of both species and one plant of M. incana had over 200 open heads.Herbarium label data indicate that the blooming period of M. incana generally coincides with that of M. implicata on other Channel Islands as well as on San Miguel Island and that flowering of both species reaches a maximum in the months of May and June.D4Ö rsycne \ya\.^L in cultivation, plants of M. incana have been found to be selfcompatible and all have had a chromosome number of 2n = 14 while plants of M. implicata have all been self-incompatible and have had a chromosome number of 2n = 18.The one hybrid plant propagated from wild fruit was sterile and 16 univalents were found in meiosis.Tests using long and short wave UV light (366nm and 245nm) on flower heads of plants in cultivation indicate that the reflective pattern of M. incana is quite different from that of M. implicata, and that the hybrid has the same pattern as in M. incana.
Because of the differences between M. incana and M. implicata in flower color, breeding system, and UV reflectance it was clear that a study of the pollinators of the two species would contribute to a general study of the causes of interspecific hybridization.
The only mention in the literature of insects on Malacothrix on San Miguel Island are the records of six native bee species at flowers of M. implicata by Cockerell (1937), the recent synonomy of one of these bees by Rust (1984), and there were no data on pollinator choice among the three flower types (Miller and Menke, 1981;Miller, 1985b).

METHODS
The present preliminary report is based on collections made on San Miguel Island, 24-27 May 1984, anddata from Cockerell (1937), supplemented by earlier observations on this and other of the California Islands.On the May, 1984 trip most of the island was surveyed, with concentrated efforts on the slopes above Cuyler Harbor and Tyler Bight.Insect specimens are deposited in Santa Barbara Museum of Natural History, Natural History Museum of Los Angeles County, U.S. National Museum of Natural History, and California Department of Food and Agriculture collections.Our collecting emphasized sampling the diversity of insects present; time did not allow detailed observation of the abundance and activities of each species.Although our data are only preliminary, we feel that they are adequate to infer general patterns and suggest directions for future research, which should address seasonal and diurnal activity patterns of the insects, insect-flower interactions, and flower constancy of the insects (e.g.Hurd and Linsley, 1975;Hurd et al., 1980).

RESULTS
In general, Malacothrix incana, M. implicata, and the interspecific hybrids are visited by the same suite of generalist bees and a few flies, as well as small beetles and true bugs (Table 1).Many of these species were seen on many other flowers in a variety of families including Convolvulaceae {Calystegia macrostegia subsp.macrostegia [Greene] Brummit) and Crassulaceae (Dudleya greenei Rose), in the Cuyler Harbor area.Because bees carry large pollen loads and move actively between flowers, they are probably the major pollinators of Malacothrix, although flies and beetles (as well as wasps which were not seen on flowers of Malacothrix on San Miguel Island) probably have some role in pollination.
It is probable that the most important pollinator is the native bee Agapostemon texanus Cresson, the relatively large metallic green females of which are abundant and active, especially in the early morning when the flowers of both species of San Miguel Island Malacothrix are open and receptive for pollination.Individuals of A. texanus were, in general, the most commonly observed bees on Malacothrix (although they were rnore obvious due to size and color than many of the small halictids), and were observed flying between Malacothrix flowers of the same and different colors.However, the relatively constant wind on the island made following individual bees for long periods impossible, so we have no data on flower preference or constancy.Pollen from the hind legs of eleven A. texanus females collected on Malacothrix was stained with cotton blue lactophenol and compared under a phase contrast light microscope with pollen of M. implicata and M. incana.The majority of the pollen grains from the bees were comparable to the pollen of Malacothrix in size, external wall structure, and other morphological characters although it was not possible to distinguish the pollen of M. incana from that of M. implicata.These observations are consistent with the biology of A. texanus in the laboratory (Roberts, 1969).
The insects found on flowers of Malacothrix on San Miguel Island in May, 1984 are listed in Table 1, which also includes records from Cockerell (1937).All of the taxa identified to species, except Lopidea nigridea hirta and the Hymenoptera, are San Miguel Island records first published here.All these species were

Psyche
[Vol.92  (Rust, in Miller, 1985a), but this may be due to the poor taxonomic condition of this difficult genus.Observations on other islands (Miller, unpubl.)and, specifically, from San Nicolas Island (Table 2) suggest that a similar suite of general pollinators is active on M. incana on other islands.Pollinator collections made on San Nicolas Island in May, 1984 were superficial compared to those on San Miguel Island, but consist of the same dominant elements.
Other parts of plants of Malacothrix were examined as well as flowers, and three families of Homoptera were found on stems and roots.The western aster root aphid.Aphis armoraceae Cowen (Aphididae) on the stem of a hybrid plant on the slopes above Cuyler Harbor, was associated with the introduced ant Lasius niger L. (Hymenoptera: Formicidae).The black scale, Saissetia oleae (Olivier) (Coccidae) was taken on stems of M. incana near Cabrillo Monument, Cuyler Harbor, in May, 1977 (by SEM) and on the slopes above Cuyler Harbor in May, 1984.An unidentified mealybug (Pseudococcidae) was taken from the root of a plant of M. implicata at Cuyler Harbor, but the specimen was lost in the mail during attempts to obtain an identification.

CONCLUSIONS
All the forms of Malacothrix on San Miguel Island are apparently visited by the same suite of generalist bees and a few flies, as well as small beetles and true bugs.Circumstantial evidence suggests that the bee Agapostemon texanus is probably the most important pollinator.Future research should examine the specific nature of the insect-flower interactions.The simplified nature (fewer species of both plants and pollinators compared to similar mainland sites) of insect-plant interaction systems on the California Islands presents excellent opportunities for testing hypotheses of insect-plant interaction and evolution (e.g.Cruden, 1972;Thorp, 1979).

Table I ,
(Van Duzee, 1921)ating and herbivorous insects known from flowers of Malacothrix on San Miguel Island (X = observed inMay, 1984; Ckll = observed by  Cockerell in July, 1937)., but some of them spend more time on leaves and stems than on flowers.The interspecific hybrid is somewhat more poorly represented with respect to its insects because fewer plants were seen.Lopidea nigridea hirta was described as a San Miguel Island endemic(Van Duzee, 1921), but the status oí hirta (as species, subspecies or synonym of a more widespread species) is unclear without revision of the very large genus Lopidea.Two tortricid moth larvae were reared from the flowers of the hybrid Malacothrix (SEM lot 84E1 A, B), yielding adult females of Argyrotaenia, ipxobab\y franciscana insulana; the various island populations once lumped under insulana are quite variable and need review.Lepidanthrax borius is here first recorded from San Miguel and San Nicolas Islands;Cockerell's (1940: 292)San Nicolas record of L. angulas Osten Sacken might be a misidentification of this.The five species of Dialictus were described by Cockerell (1937, in the genus Halictus) from San Miguel island.Dialictus cabrilli is still considered a San Miguel Island endemic

Table 2 .
Associates of flowers of Malacoihrix incana on San Nicolas Island (collected 11 and 31 May, 1984 by S. Junak from plants on stabilized sand dunes east of Corral Harbor, elevation about 20m).