Diabroticina (Chrysomelidae: Galerucinae: Luperini) beetles were sampled under field conditions on two host plants of the family Cucurbitaceae, Cucurbita okeechobeensis ssp. martinezii L. Bailey (bitter, wild cucurbit) and C. moschata (Lam.) Poiret (non bitter, cultivated cucurbit). Seventeen species of Diabroticina were collected. Acalymma blomorum Munroe & Smith was the most abundant species on both host plants. The only parasitoid found was Celatoria compressa Wulp (Diptera: Tachinidae). This parasitoid attacked more beetle species on the cultivated cucurbit (65%) than on the bitter cucurbit (20%). However, the percentages of parasitism observed in all species were low (0.4% to 12.5%). These data suggest that host plant species might have an effect on parasitism.
Diabroticina (Chrysomelidae: Galerucinae: Luperini) are native to Mexico and Central America (Webster 1895). Few data, however, are published on host plant associations (Eben & Espinosa de los Monteros 2003) and natural enemies from this area (Eben & Barbercheck 1996).
In the Diabroticites (subtribe Diabroticina), the association of a number of species with plants in the family Cucurbitaceae is a well-known example for the effect of plant secondary chemistry on a plant-insect interaction (Chambliss & Jones 1966; Howe et al. 1976; Metcalf et al. 1982; Metcalf 1986). Host preferences of Diabroticites are strongly influenced by the presence of cucurbitacins (tetracyclic triterpenoids) in many wild cucurbit hosts. These non-volatile secondary compounds act as arrestants and feeding stimulants for these beetles (Chambliss & Jones 1966; Metcalf & Lampman 1989).
Furthermore, it has been proposed that Diabroticina species sequester cucurbitacins for their chemical defense. Studies of tritrophic effects demonstrated that cucurbitacins are deterrents for natural enemies such as mantids (Ferguson & Metcalf 1985), passerine birds (Nishida & Fukami 1990), the pathogenic fungus Metarhizium anisopliae (Moniliales: Moniliaceae) (Tallamy et al. 1998), and entomopathogenic nematodes (Barbercheck et al. 1995). On the other hand, no negative effects on general predators such as carabid larvae, mites, and centipedes (Brust & Barbercheck 1992) have been found. To date, no clear pattern has emerged from these studies. Field studies on larval host associations in the natural habitat are difficult due to the fact that Diabroticina larvae are root feeders.
Interestingly, although parasitoids are intimately associated with their hosts, and third trophic level effects of plant secondary compounds are described for a number of plant-insect associations (Gauld et al. 1992; Rowell-Rahier et al. 1995; Agrawal et al. 2002), no data exist for the cucurbit/Diabroticina/parasitoid system.
The objective of the present field study was to compare beetle abundance and diversity on two Cucurbita spp. that differed in presence, Cucurbita okeechobeensis ssp. martinezii L. Bailey, or absence, C. moschata (Lam.) Poiret, of secondary compounds. Both cucurbits, the bitter Cucurbita o. martinezii and the cultivated C. moschata, are the most common cucurbits in the study area. Their morphology is similar, but the bitter species has smaller and paler flowers, and smaller leaves. Furthermore, the bitter species produces secondary compounds characteristic for Cucurbitaceae, the cucurbitacins (Metcalf et al. 1982; Ventura 2002). Moreover, parasitoid incidence in adult beetles was monitored and compared between individuals collected from the two host plants.
Material and Methods
Study Area and Host Plants
All adult insects were collected in the central zone of the state of Veracruz, Mexico. Mean annual temperature fluctuates between 18 and 25°C, with three distinct seasons: a dry-cool season (November-March), a dry-warm season (April-May) and a wet-warm season (June-October). Annual rainfall varies between 800 and 2500 mm, with a peak in the second warm season (Soto & García 1989). Common crops in the area are sugarcane, coffee, corn, squash, and beans. The original vegetation at lower altitudes is deciduous forest, whereas remnants of tropical cloud forest are found at higher altitudes (Gómez-Pompa 1977).
Within the study area, six locations for each Cucurbita spp. were identified. These locations were separated by at least nine km (i.e., 12 sample areas in total). They differed in altitude and climatic conditions (Table 1). To avoid collections of beetles which might recently have moved between hosts, sites with coexistence of both cucurbits were not accepted. Due to the fact that beetle abundance is affected by the presence of flowers (pers. observ.), insects were collected once or twice per week on flowering plants only. When plants began to dry out, they were replaced by others in flowering stage within the same area. At each collection date we recorded the diversity and abundance of beetles found on both plants.
Areas of approximately 100 m2 covered by cucurbit vines were measured at each location to define the collection site.
Beetles were collected from August to December 2001 and from May to November 2002. Collection dates were based on previous studies (Rodriguez & Magallanes 1994; Eben & Barbercheck 1996; Cabrera & Cabrera 2004) which found a clear seasonality for Diabroticinas with peak abundance from early summer to fall. Plants were visually inspected for Diabroticina adults. The sampling unit was the number of beetles collected per person in one hour. Field collected adults were separated by species, location, and collection date. In order to allow for parasitoid emergence, the colonies of adult field collected beetles were maintained in the laboratory (25 ± 3°C), with a photoperiod of 13:11 (L:D), in transparent plastic containers (15 cm diameter × 25 cm length), with a gaze cover for ventilation. Beetles were fed fruits of Cucurbita pepo L. (zucchini) and artificial diet (Branson et al. 1975). Abundance of Diabroticina were analyzed by one-way ANOVA (P < 0.05) with SigmaStat™ statistical software version 2.0 (Jandel Scientific 1992-1997), after square root transformation.
All cages with beetles were checked daily to collect and count parasitoid pupae (Eben & Barbercheck 1996). In addition, dead beetles with an entire abdomen were dissected to determine presence or absence of immature parasitoids. Percentage parasitism was calculated as the number of immature and adult parasitoids obtained for the total number of each beetle species, date, and location. Data were analyzed by a chi-square test (Zar 1999). Correlation between beetle abundance and percentage parasitism was analyzed by linear regression (Zar 1999).
In the study area, C. moschata is cultivated for human consumption. For this reason it was commonly found along road sides and in mixed corn-squash plots, mostly in direct sunlight. Cucurbita o. martinezii is grown in shadier places, with other plants as climbing structures. It was most abundant in and around coffee plantations.
We found 17 species of Diabroticinas from five genera (Table 2). All species were collected from C. moschata, 15 species were collected from C. o. martinezii. The abundance of three species, A. blomorum, D. balteata, and I. tetraspilota, differed between both cucurbit hosts (P = 3.5 × 109, P = 0.0011, and P = 0.048, respectively). The other 12 beetle species were not more abundant in any of the two host plants. Proportions of all species were different in the two cucurbits. In C. moschata, Acalymma blomorum was the most abundant beetle species, followed by Diabrotica balteata, D. scutellata, and D. viridula, and in C. o. martinezii, Isotes tetraspilota, A. fairmairei, and D. scutellata.
The most diverse genus was Diabrotica with six species in the fucata group, D. balteata, D. dissimilis, D. nummularis, D. sexmaculata, D. tibialis, D. undecimpunctata duodecimnotata, and three species in the virgifera group: D. porracea, D. scutellata, and D. viridula. Diabrotica balteata and D. scutellata were the most abundant species within either group. Cerotoma atrofasciata, Gynandrobrotica lepida and G. nigrofasciata were most common on the foliage of C. moschata (Table 2).
Incidence of Parasitism
The only parasitoid found was a tachinid species, Celatoria compressa Wulp. Parasitoids were obtained in June and July 2002 from beetles collected on the bitter cucurbit. No parasitoids were found in beetles collected on this plant in 2001. In beetles collected on the cultivated cucurbit, parasitoids were present throughout the collecting period in both years. In general, parasitoid pupae were obtained during the first 48 h after collecting the host beetle. Adult parasitoids emerged from all parasitoid pupae (n = 169). The presence of other parasitoids was not observed.
The tachinid parasitoid was found in three of the 15 beetle species collected on the bitter cucurbit (20%), and in 11 of the 17 beetle species collected on the cultivated cucurbit (65%). On the bitter cucurbit, the parasitoid attacked A. blomorum, A. fairmairei, and D. balteata at percentages of 0.9%, 7.7%, and 5%, respectively (Table 2). In the species collected on the cultivated cucurbit, A. blomorum, A. fairmairei, A. innubum, C. atrofasciata, D. balteata, D. porracea, D. scutellata, D. sexmaculata, D. tibialis, D. viridula, and G. nigrofasciata were parasitized. The percentage of parasitism in these species varied between 0.4% and 12.5% (Table 2). Highest percentages of parasitism were found in C. atrofasciata. No significant differences in parasitism between beetle species were detected. Also, no correlation between beetle abundance and percentage parasitism was found. We found, however, significantly higher numbers of parasitoids in beetles collected on the cultivated cucurbit (X2(1, 0.05) = 6.46). During the present study, parasitism was not observed in A. trivittatum, D. dissimilis, D. nummularis, G. lepida, and I. tetraspilota.
The diversity of Diabroticina species on C. o. martinezii and C. moschata was similar. Nevertheless, Diabrotica scutellata and Isotes tetraspilota were more abundant on bitter C. o. martinezii, whereas Acalymma blomorum and D. balteata were more abundant on the cultivated C. moschata. Acalymma blomorum was the most abundant species in both Cucurbita spp. These results are similar to data obtained by Cabrera & Cabrera (2004) with respect to the abundance of Acalymma spp. in Cucurbitaceae in Argentina. Within the fucata group, D. balteata was the most abundant species, and within the virgifera group D. scutellata was dominant. These results agreed with data reported by Rodriguez & Magallanes (1994) for D. balteata in Tamaulipas and Veracruz, and by Eben & Barbercheck (1996) for D. scutellata in Veracruz. In the present study, Diabrotica dissimilis and G. nigrofasciata were not collected on the bitter cucurbit. The other 12 species had a continuously low abundance on both cucurbits.
Acalymma blomorum and D. balteata were collected most frequently in C. moschata, perhaps as a result of the high quantities of pollen in this cucurbit. We observed that the beetles visited these plants to feed on petals and flowers. Isotes tetraspilota was the only species that was found feeding on the leaves of C. o. martinezii, and it was never seen on the cultivated cucurbit C. moschata. Acalymma blomorum, A. fairmairei, C. atrofasciata, and D. balteata were found on the cultivated cucurbit even when the plants began to dry. Gynandrobrotica nigrofasciata was collected only in 2001 on the cultivated cucurbit and furthermore, in much higher numbers, on a leguminous plant (Pachyrhizus erosus (L.) Urban) growing in the vicinity. Fabaceae are reported as host family for this genus (Jolivet & Hawkeswood 1995).
Celatoria compressa was the only parasitoid species obtained. It was reported for the first time by Eben & Barbercheck (1996) in Diabroticite beetles. Previously, Celatoria bosqi Blanch. (Heineck-Leonel & Salles 1997) C. diabroticae Shimer and C. crawii Coquillett had been collected from some Diabrotica spp. (Chittenden 1905; Sell 1915; Gordon et al. 1987) and A. vittatum (Walton 1914). The differences in parasitism in the species collected on both cucurbits, three species in C. o. martinezii vs. 11 species in C. moschata, were notable and might suggest that floral odor, or secondary compounds sequestered by beetles from the bitter host plant, had an effect on the adult parasitoid. To date, no study has tried to corroborate this for parasitoids of Diabroticina beetles. On the other hand, Diabroticina beetles were more abundant on the flowers and leaves of C. moschata, perhaps because this species offered greater amounts of pollen than C. o. martinezii. Consequently, C. compressa might have simply responded to the abundance of host insects in C. moschata. In general, the incidence of the tachinid in the guild of beetle species collected in both cucurbits was low, with parasitism ranging from 0.4% to 12.5%. In an earlier study similar parasitism rates of 1.0% to 11.1% were found (Eben & Barbercheck 1996). Parasitism by C. bosqi in D. speciosa ranged from 0.1 to 30.2% (Heineck-Leonel & Salles 1997) and C. diabroticae parasitized D. u. howardi with rates of 3 to 15% (Meinke & Gould 1987; Elsey 1988).
Given our observations, it would be interesting to investigate if cucurbitacins sequestered by the beetles collected on C. o. martinezii function as a repellent for C. compressa. Our data suggest that a possible effect of plant secondary compounds is stronger on adult parasitoid behavior (i.e., host acceptance) than on immature physiology, since all parasitoid larvae that eclosed from beetles pupated and developed successfully into adults.
During the course of our study a larger number of isolated plants of the cultivated than the bitter cucurbit species was found. This situation was contrary to our observations in previous years. It might be the consequence of the rapidly declining number of coffee plantations in the central area of Veracruz. In this area, the main habitat of C. o. martinezii are coffee plantations, where it is found growing in a vertical fashion upon the coffee bushes as climbing structures. By contrast, C. moschata grows horizontally, covering bare areas exposed to plain sunlight. These differences in habitat and microclimate might have influenced the abundance and species composition of Diabroticina beetles present in both plants as well as the searching behavior of C. compressa.
Thanks to A. Martinez who helped with the statistical analysis. R. Ventura assisted with collections and maintenance of colonies. This study was financed by project CONACYT 35501-V to A. Eben and partially with the assistanship CONACYT 162421 to S. Gámez-Virués.
Climatic zones of the six collections areas in the state of Veracruz.
Diabroticina beetle abundance collected on Cucurbita okeechobeensis ssp. martinezii (A) and C. moschata; (B) (mean beetles collected per person per hour), and percentage parasitism in the total number collected of each beetle species in the years 2001 and 2002.