Diaprepes abbreviatus (L.) is a major pest of Florida citrus. Adult females lay eggs in masses sealed between leaves in the citrus canopy, and recently-hatched neonate larvae drop to the soil and feed on roots. The coccinellid species, Cycloneda sanguinea (L.), Harmoniaaxyridis Pallas, and Olla v-nigrum Mulsant, are generalist predators that consume a wide range of citrus pests, although they have not been observed preying on Diaprepes. We conducted experiments to determine whether these species would feed on Diaprepes egg masses and neonate larvae, and how an exclusive or partial diet of Diaprepes eggs would influence their development. The three predators responded very similarly in our tests. In laboratory assays, coccinellid larvae and adults readily consumed exposed Diaprepes eggs and neonates less than 48 h old; and coccinellid larvae preyed on 40% of intact egg masses laid between wax paper strips or citrus leaves, whereas adults preyed on 8.7%. In a greenhouse assay,coccinellid larvae located and preyed on 22.7% of intact egg masses laid between leaves on potted citrus trees. Although neonates might have relatively limited exposure to predation in the canopy before they drop to the soil, predation could be an important factor selecting for the timing of egg hatch, neonate escape from leaf envelopes, and neonate drop. The developmental assays indicate that Diaprepes eggs are less suitable prey for thesecoccinellid species than eggs of the flour moth, Ephestia kuhniella Zeller, but that they could be a highly acceptable component of a mixed diet. Our experiments indicate that these coccinellid species are potentially important predators of Diaprepes but the extent to which they contribute to the natural biological control of this weevil remains unknown.
The root weevil, Diaprepes abbreviatus (L.), apparently originated in the Caribbean and is now a major introduced pest of citrus, sugar cane, and ornamentals in Florida (Simpson et al. 1996). In citrus, larval feeding damages roots, reduces yield, and kills trees by girdling or by facilitating infection by plant pathogens such as Phytophthora spp. (Graham et al. 1996). The combination of Diaprepes and Phytophthora can lead to rapid tree decline and destroy groves within a few yearsof a weevil infestation. Adults are long lived and feed on foliage, especially new growth. Mating occurs in the canopy, and eggs are laid in masses between leaves that are glued together by an adhesive secreted by the female during oviposition. The larvae hatch, escape from the sealed leaf envelope, drop to the soil, and burrow down to the roots where they begin feeding. As they grow, the larvae move to larger roots, and pupate in the soil after 9-11 instars(Woodruff 1985; Quintela et al. 1998; McCoy 1999).
Sources of mortality for the various life stages of Diaprepes include numerous predators and parasites. Several egg parasitoids have been discovered in the Caribbean and introduced into Florida for biologicalcontrol (Hall et al. 2001). Eggs are also subject to predation by ants, crickets, earwigs, lacewings, and spiders. Newly-hatched neonate larvae are especially prone to predation on the soil surface where they are attacked by ants, earwigs, hemipterans, and spiders whereas larvae below ground are attacked by ants and entomopathogenic nematodes. Adults are preyed on by ants, birds, lizards, snakes and spiders(Whitcomb et al. 1982; Richman et al. 1983a;Richmanet al. 1983b; Tryon 1986; Jaffe et al. 1990; McCoyet al. 2000; Stuart et al. in press).
In developing an effective IPM program to control D. abbreviatus, it would be advantageous to maximize the effectiveness of as many of the natural enemies of this insect as possible. Several coccinellid species including Cycloneda sanguinea (L.), Harmonia axyridis Pallas, and Olla v-nigrum Mulsant, are common in Florida citrus groves where they prey primarily on various aphid species (Michaud 2000). Although these coccinellids have not been reportedpreying on Diaprepes, coccinellids are known to consume a wide range of prey including certain weevil larvae (Kalaskar & Evans 2001). Moreover, these coccinellid species coexist with Diaprepes in the citrus canopy and are likely to encounter Diaprepes egg masses and possibly neonate larvae, before the larvae escape from their sealed leaf envelopes and drop to the soil. Hence, coccinellids might contribute to the natural biological control of this weevil, andwe conducted laboratory experiments to determine whether these species could be effective predators on Diaprepes egg masses and neonates. We also conducted experiments to determine how an exclusive or partial diet of Diaprepes eggs would influence the development of these species.
Materials and Methods
Rearing
Laboratory colonies of Harmonia axyridis, Cycloneda sanguinea, and Olla v-nigrum were reared on frozen eggs of the flour moth, Ephestia kuhniella Zeller, and bee pollen as described by Michaud (in press).Coccinellid eggs were collected from ovipositing females and incubated as described by Michaud (2000) to produce larvae used in experiments. Diaprepesabbreviatus eggs were obtained from laboratory colonies where they were laid between strips of wax paper or between citrus leaves. Leaves that were presented to weevils for oviposition were held together with paper clips, and the leaf petioles were wrapped with wet cotton and parafilm to prevent desiccation. Diaprepes neonateswere obtained from egg masses laid between wax paper strips and were used in experiments within 48 h of their emergence from sealed egg masses. Aphis gossypii Glover were reared on cotton plants (Gossypium hirsutum, var. “SureGrow”) in growth chambers at 20 ± 1°C, 16:8 L:D h, 60-65% RH. Laboratoryexperiments were conducted in plastic Petri dishes (5.5 cm dia × 1.0 cm) on open laboratory benches at 24 ± 1°C under fluorescent lights unless otherwise noted.
Prey Acceptability
Tests were performed to determine the relative acceptability of Diaprepes eggs and neonates as prey for larvae and adults of the three coccinellid species. In the first test, individual first-instar coccinellid larvae (<24 h old) were transferred to plastic Petri dishes (as above) containing an excess of E. kuhniella eggs (∼0.1 gm) and a small D. abbreviatus egg mass of 10-19 eggs, which had been exposedby separating the wax paper strips between which the mass had been laid. We conducted 24-25 replicates per species, and the number of D. abbreviatus eggs consumed was recorded after 24 h. In the second test, 4-5 wk old adult coccinellids were similarly presented with excess E. kuhniella eggs and an exposed Diaprepes egg mass. We conducted 20 replicates per species, and predation was assessed after 24 h. In the third test, coccinellid larvae reared on E. kuhniella eggs to the third instar were presented with an excess of moth eggs and 10 D. abbreviatus neonates less than 48 h old. We conducted 18-20 replicates per species, andthe number of D. abbreviatus larvae consumed was recorded after 24 h. In the fourth test, individual coccinellid adults were starved for 16 h and then transferred to plastic Petri dishes that contained 10 A. gossypii (a mixture of 4th instars and apterous adults) and 10 D. abbreviatus neonates.We conducted 13-16 replicates per species, and the numbers of A. gossypii and D. abbreviatus larvae consumed were recorded after 30 min.
Prey Accessibility
Laboratory tests were conducted to determine whether coccinellid larvae and adults could access and prey upon intact D. abbreviatus egg masses that had been oviposited between strips of wax paper or between citrus leaves. The first test consisted of 18-20 replicates per species in which individual first instar coccinellid larvae thathad been fed E. kuhniella eggs for 24-48 h post eclosion were transferred to a plastic Petri dish containing a D. abbreviatus egg mass laid between wax paper strips. After 24 h, the wax paper strips were pulled apart and predation assessed. The second test consisted of 10-20 replicates per species and was similar to the first test except that the egg mass had been laid between citrus leaves. Predation was similarly assessed. The thirdtest used the same methodology and consisted of 19-20 replicates per species in which 4-5 wk old adult coccinellids from stock colonies were presented with egg masses that had been laid between wax paper strips. Predation was similarly assessed. Egg masses laid between wax paper strips were 2-5 days old at the time of testing whereas those between leaves were less than 24 h old.
For the experiments, egg masses oviposited by the weevils between wax paper strips were prepared by trimming the paper around the masses so that they would fit into the assay dishes, and by folding up the edges of the wax paper slightly to provide entry for the coccinellids without damaging the egg masses themselves or the integrity ofthe adhesive applied by the egg-laying weevil. Excess leaf material around egg masses laid by weevils between citrus leaves was also trimmed so that they would fit into assay dishes, again without damaging the egg masses themselves or the integrity of the adhesive sealing them between the leaves. Some of the egg masses laid by Diaprepes females between citrus leaves appeared to be poorly sealed, such that the leaves readily separated, exposing the eggs. Only well-sealed egg masses were used in our experiments.
A greenhouse experiment was conducted to investigate the accessibility of egg masses to predation by coccinellid larvae of the three species under more natural conditions. Fifty D. abbreviatus adults of unknown age were caged on each of three 3-year old Volkamer lemon trees for 24 h. The D. abbreviatus adults and cageswere then removed and 50 first instar coccinellid larvae (24-36 h old) of each species were released, one species per tree. Larvae were restricted to trees by an application of Tanglefoot® (The Tanglefoot Company, Grand Rapids, MI 40504) around the trunk. The larvae were removed after 24 h, D. abbreviatus egg masses located, and predation assessed.
Prey Suitability
Two developmental assays were performed to assess the suitability of D. abbreviatus eggs as food for H. axyridis, C. sanguinea, and O. v-nigrum larvae. In the first assay, we conducted 30-32 replicates per species in which individual coccinellid larvae (<24 h old) were transferred to plastic Petri dishes with water encapsulated in polymer beads (Entomos, LLC 4445 SW 35th Terrace, Suite 310, Gainesville, FL 32608),and approximately half were fed frozen E. kuhniella eggs (control diet) and the other half frozen D. abbreviatus eggs daily until pupation. Additional water beads were supplied every three days. Procedures for the second assay were identical to the first except that 36-39 larvae of each species were reared individually on the control diet for four days until they were late second or early third instar, whereupon approximately half were transferred to new dishes and fed D. abbreviatus eggs until pupation, whereas the other half continued receiving the control diet. Mortality, time to pupation, time to adult emergence, and adult dry weight wererecorded for each replicate in each experiment.
Data Analysis
Statistical analysis used the SAS System for Windows, release 6.12 (SAS Institute, Inc. 1990). Count data were transformed priorto analysis using square root transformations. Comparisons of means used PROC ANOVA, PROC GLM, and PROC TTEST. Comparisons of proportions were conducted using contingency table analysis and chi-square tests or Fisher’s exact tests using PROC FREQ. Untransformed data are reported in tables and figures as means and standard errors.Differences were considered significant at the P = 0.05 level, but sequential Bonferroni adjustments of critical values were used as indicated below to maintain error rates at the stated values when multiple comparisons were made within experiments (Rice 1989).
Results
Prey Acceptability
In the first choice test in which first instar coccinellid larvae were offered exposed Diaprepes eggs and an excess of wax moth eggs, C. sanguinea consumed Diaprepes eggs in 15 of 24 replicates, H. axyridis in 15 of 24 replicates, and O. v-nigrum in 13 of 25 replicates. There was no significant difference among coccinellid species in the proportion of replicates in which weevil eggs were consumed (χ2 = 0.749, df = 2, P = 0.688). Similarly, there was no significant difference among beetle species in the number of weevil eggs that were consumed (ANOVA, F = 1.610, df = 2, 70, P = 0.2074; Fig. 1).
In the second choice test in which coccinellid adults were offered exposed Diaprepes egg masses and an excess of moth eggs, adults of each of the three species consumed Diaprepes eggs in 19 of 20 replicates. The number of eggs consumed was not recorded. In the third choice test in which third instar coccinellid larvaewere offered 10 D. abbreviatus neonates and an excess of moth eggs, all three species ate at least some Diaprepes neonates in all replicates, and there was no significant difference in the number of neonates consumed among coccinellid species (ANOVA, F = 0.10, df = 2, 55, P = 0.9055; Fig. 2). In the fourth choice test in which coccinellid adults were offered 10 A. gossypii and 10 D. abbreviatus larvae, all three species ate more Diaprepes larvae than aphids (Fig. 3). Based on ANOVA, the main effect for coccinellid species was significant (F = 4.60, df = 2, 82, P = 0.128) with O.v-nigrum consuming less than the other two species, which did not differ. The main effect for prey species was also significant (F = 24.15, df = 1, 82, P = 0.0001) with more Diaprepes being consumed than aphids; and the interaction between predator species and prey species was not significant (F = 1.95, df = 2, 82, P = 0.1489). However, in this experiment the Diaprepes larvae tended tostay on the bottom of the dishes whereas the aphids often moved up the sides and onto the lids, and this differential distribution of prey types might have contributed to the different consumption rates.
Prey Accessibility
When Diaprepes egg masses oviposited between wax paper strips were offered to first instar coccinellid larvae, C. sanguinea successfully penetrated between the strips and preyed on eggs in 4 of 20 replicates, H. axyridis in 10 of 20 replicates, and O. v-nigrum in 7 of 18 replicates. For egg masses between citrus leaves, C. sanguinea larvae successfully penetrated between the leaves and preyed on eggs in 3 of 10 replicates, H. axyridis larvae in 7 of 19 replicates, and O. v-nigrum larvae in 12 of 20 replicates. There were no significant differences among coccinellid larvae of different species in the proportion of egg masses that were preyed upon for wax paper (2 × 3 contingency table, χ2 = 3.978, df = 2, P = 0.137), leaves (2 × 3 contingency table, χ2 = 3.239, df = 2, P = 0.198), or for both substrates pooled (2 × 3 contingency table, χ2 = 5.255, df = 2, P = 0.072). Also, pooling the data for larvae of all three predator species, there was no significant difference for wax paper versus citrus leaves in the proportion of egg masses that were preyed upon (2 × 2 contingency table, χ2 = 0.835, df = 1, P = 0.361). Coccinellid larvae successfully preyed on 36.2% of egg masses sealed in wax paper, 44.9% sealed between citrus leaves, and 40.2% overall.
When coccinellid adults were offered egg masses sealed between wax paper strips, C. sanguinea penetrated between the strips and preyed on eggs in 0 of 20 replicates, H. axyridis in 1 of 20 replicates, and O. v-nigrum in 4 of 19 replicates, for an overall success rate of 8.5%. The largest difference in this experiment, that between C. sanguinea and O. v-nigrum, was borderline signi-ficant (Fisher’s exact test, P = 0.047), but this result was rendered nonsignificantwhen sequential Bonferroni adjustments of critical values were made with respect to the total number of comparisons within the experiment (n = 3; Rice 1989). Therefore, again, no significant difference among beetle species was evident.
In the greenhouse experiment, relatively few egg masses were laid on the lemon trees but first instar larvae of all three coccinellid species successfully preyed on some of them. C. sanguinea preyed on 1 of 5 egg masses, H. axyridis on 2 of 10 egg masses, and O. v-nigrum on 2 of 7 egg masses, for an overall predation rate of 22.7%. None of the differences among predator species in this experiment was significant (Fisher’s exact tests, P > 0.05). Thus, larvae of all three coccinellid species were capable of locating, accessing, and preying upon Diaprepes egg masses.
Prey Suitability
Coccinellids reared on Diaprepes eggs generally did worse than those reared on flour moth eggs. Significantly fewer larvae of each coccinellid species survived to adulthood when reared on an exclusive diet of Diaprepes eggs than when fed moth eggs, adult dry weight was significantly reduced, and developmentaltime was significantly extended (Table 1). Coccinellidlarvae of each species reared on Diaprepes eggs from the third instar onward survived to become adults as well as those reared entirely on flour moth eggs, but adult weight was still significantly lower and development time significantly longer (Table 1).
Discussion
This study demonstrates that coccinellid larvae and adults are potentially important predators of Diaprepes abbreviatus eggs and neonate larvae. The frequent consumption of Diaprepes eggs and neonates by coccinellid larvae and adults in these time-limited laboratory tests when other acceptable food was present and abundantindicates that these prey items are readily accepted by these predators. Moreover, young coccinellid larvae were especially capable of accessing and preying upon Diaprepes egg masses laid between wax paper strips or citrus leaves; and coccinellid adults exhibited this ability to a lesser extent. Diaprepes neonatesdrop to the soil soon after hatching (Woodruff 1985; McCoy 1999) and might have limited exposure to coccinellid predation in the canopy but predation pressure from coccinellids and other species could be an important factor selecting for the timing of egg hatch, neonate escape from leaf envelopes, and neonate drop.
The assays in this study provide little indication of differences among the three coccinellid species that might suggest that one is a more effective predator of Diaprepes eggs or neonates than the others. Only one of the experiments on prey acceptability, experiment 4, revealed any species differences. In this experiment,coccinellid adults were offered a choice between aphids and Diaprepes neonates, and O. v-nigrum ate less of both prey items than the other two species, which did not differ. No differences in preyaccessibility or suitability were detected. Similar experiments with larger sample sizes might reveal differences, but the present data suggest that these would be minor. Potential differences among these species in terms of their foraging behavior in citrus groves or their relative abundances might make one species more effective than the others under field conditions.
The differential predation success exhibited by coccinellids in this study when presented with exposed eggs versus intact egg masses sealed between wax paper strips or citrus leaves indicated that egg masses laid between leaves are protected to some extent from predation by coccinellids. However, the success of some coccinellid larvae and adults in penetrating intact egg masses and preying on eggs indicates that this protection is far from absolute. Indeed,coccinellid larvae penetrated and preyed on sealed egg masses in over 40% of replicates whereas coccinellid adults did so in 8.5%. The penetration of intact egg masses by coccinellids suggests that these predators are responding to chemical cues associated with egg masses. Similar responses have been documented for coccinellidstoward other prey (Obata 1986; Hattingh & Samways 1995).
Jaffe et al. (1990) noted that first instar Diaprepes larvae were somewhat repellent to various ant predators and that eggs less than 48 h old were often ignored by ants. Pavis et al. (1992) extracted and identified the chemicals apparently responsible for the repellency of the larvae and provided further evidence for their role in larvaldefense. Repellency of Diaprepes larvae to coccinellids was not observed in this study and was not evident in the experimental results. Moreover, as indicated by our leaf and greenhouse experiments,eggs less than 48 h old were not ignored by these predators. Thus, the chemical defenses of the larvae and apparent defenses of the eggs might only be applicable to certain taxa of predators. Defenses targeted toward ants would seem especially appropriate since ants are evidently the primary predators of various life stages of Diaprepes including egg masses and neonate larvae (Whitcomb et al. 1982; Richman et al. 1983a;Richman et al. 1983b; Tryon 1986; Jaffe et al. 1990; McCoy 1999; Stuart et al. 2001). Deterrence of coccinellids from attacking alfalfa weevil larvae has been attributed to the defensive wriggling of the larvae (Kalaskar & Evans 2001) but it is unclear whether wriggling might contribute to defense for Diaprepes larvae or whether alfalfaweevil larvae also possess chemical defenses.
Seasonal cycles of coccinellid and Diaprepes abundance in Florida citrus indicate considerable periods of overlap when predation is likely to occur. The three coccinellid species are present in groves year-round but are most abundant during periods of flowering and flushing, especially spring and fall. Flowers are attractive as sources of pollen and nectar for all three species, and most of their primary homopteran prey species require newly expanding leaves for their growth and reproduction. Hot weather in summer is typically associated with low food availability anda majority of individuals of all three species aestivate during these periods (Michaud, unpublished). In central Florida, Diaprepes adults emerge in the spring and generally show a sharp peak in abundance sometime from April through June. However, emergence continues throughout the summer, and adults can remain abundant through mid November (McCoy, unpublished). Egg-laying occurs throughout this period, and neonate drop has been recorded from early July through early December (Nigg et al. in press). Diaprepes adults feed voraciously on newly-expanding leaves but lay egg masses between mature leaves (Woodruff 1985; McCoy 1999). Given their seasonal activity patterns and locations in trees, it is likely that coccinellids have frequent opportunities to prey on D. abbreviatus egg masses and perhaps on neonates as well.
The timing of Diaprepes egg hatch, neonate escape from sealed leaf envelopes, and neonate drop to the soil surface could have an important impact on the exposure of neonates to various predators. Jones & Schroeder (1983) found that a considerable period often elapsed between egg hatch and neonate escape from sealed leaf envelopes, estimated average larval age at the time of neonate drop to be about 48 h, and found that neonates dropped between 1100 and 2400 h. According to Richman et al. (1983), ant foraging on the soil surface during this time period is relatively low compared to early morning hours. Thus, the timing of neonate drop could be an adaptation to avoid peak ant foraging periods. The diurnal activity patterns of coccinellids in Florida citrusgroves have not been studied but these species appear most active during daylight hours (Michaud, unpublished), and neonate drop late in the day might also enable avoidance of these predators. However, since Diaprepes neonates appear to remain within sealed leaf envelopes for relatively lengthy periods, there is a strongpossibility that coccinellids will encounter neonates rather than eggs when penetrating leaf envelopes. Additional research on the activity patterns of predators, the factors that stimulate egg hatch and neonate drop, and the conditions that promote neonate survival in the canopy, on the soil surface, and below ground is necessary for a more thorough understanding of how these factors might shape D. abbreviatus life history and survival strategies.
Overall, our predation and developmental assays suggest that Diaprepes eggs and neonates could be a frequent and highly acceptable component of a mixed diet for these coccinellid species in nature. Diaprepes eggs were less suitable prey for these species than flour moth eggs but the costs in terms of reduced adult weight and extended developmental time were relatively small for larvae reared on this diet from the thirdinstar. Similar suboptimal developmental results have been obtained for various coccinellid species and some of the aphid species that they frequently consume as prey (Michaud 2000). Thus, the results of this study, especially for coccinellid larvae, indicate that Diaprepes egg masses and neonates will be preyed upon quite readily when encountered. However, the frequency of such encounters and the intensity of predation in nature remain unknown; and it is unclear to what extent coccinellids contribute to the natural biological control of this weevil.
Acknowledgments
We thank K. Crosby and the USDA-ARS (Ft. Pierce, FL) for providing some of the D. abbreviatus egg masses used in these experiments, H. Nigg for providing ovipositing Diaprepes for leaf and greenhouse assays, and R. Villanueva and A. Goldarazena for reviewing the MS. This work was supported by the Florida Agricultural Experiment Station and grants from USDA-APHIS-PPQ and the Florida Citrus Producers Research Advisory Council and approved for publication as Journal Series No. R-08580.
References Cited
Fig. 1.
Comparison of the number of Diaprepes eggs consumed by first instar coccinellid larvae when offered 10-19 D. abbreviatus eggs and an excess of flour moth eggs. The P-value is from an ANOVA (see text).
Fig. 2.
Comparison of the number of Diaprepes neonates consumed by third instar larvae of the three coccinellid species when offered 10 neonates and an excess of flour moth eggs. The P-value is from an ANOVA (see text).
Fig. 3.
Comparison of the number of Diaprepes neonates versus the number of A. gossypii consumed by adults of the three coccinellid species. ANOVA indicates that more Diaprepes were consumed than aphids, that O. v-nigrum consumed less than the other two predator species, which did not differ, and that the interaction between predator species and prey species was not significant (see text).
