The first detections of Asian citrus psyllid, Diaphorina citri Kuwayama, in North America occurred almost simultaneously in Florida and in the Caribbean (Guadeloupe). Damaging populations on citrus have been reported in Florida but not in Puerto Rico where the psyllid was first detected in 2001, having probably arrived with its parasitoid, Tamarixia radiata Waterston. In an effort to identify additional sources of mortality, the relative abundance of coccinellid species was estimated on 180 citrus trees from early April to early July 2003. The most abundant species were Coelophora inaequalis F. (38.8%) and Cycloneda sanguinea limbifer L. (31.3%), and the least abundant were Cladis nitidula F. (5.9%), Coleomegilla innonata Mulsant (4.8%), Chilocorus cacti L. (2.1%), Scymnus sp. (5.9%), Hippodamia convergens Guerin (2.4%), and Cryptolaemus montrouzieri Mulsant (8.8%). These eight species were collected from citrus production areas in Puerto Rico for a laboratory study of feeding behavior. All eight consumed D. citri nymphs, with C. innonata consuming psyllids at a greater rate than C. cacti or Scymnus sp. Choice tests showed that C. inaequalis and C. s. limbifer preferred brown citrus aphid (BCA) Toxoptera citricida to D. citri as prey, whereas C. nitidula and C. cacti (both Chilochorini) preferred D. citri. Our results suggest that coccinellid species could play an important role as predators of the psyllid in Puerto Rico and contribute to its natural control.
The Asian citrus psyllid (ACP), Diaphorina citri Kuwayama, was first described from Taiwan in 1907. It has been present in Brazil since the 1940s (Lima 1942) but was not detected in Florida until 1998 (Halbert et al. 1998) and in Texas until September 2001 (French et al. 2001). In the Caribbean, the psyllid was first reported from Guadeloupe (Etienne et al. 1998) and subsequently detected in Puerto Rico in May, 2001 (Halbert & Nuñez 2004; P. Stansly, unpublished data). Diaphorina citri spread rapidly throughout commercial citrus areas in Florida and heavy populations were observed, especially in the south. In contrast, infestations in Puerto Rico and Texas are typically light (unpublished data and V. French, pers. comm.).
Diaphorina citri feeds and reproduces on citrus and additional hosts belonging to the subfamily Aurantioidea of the family Rutacea including jasmine orange, Murraya paniculata L., which is widely grown in Florida, Puerto Rico, and elsewhere as an ornamental plant. Feeding damage on citrus by large psyllid populations causes shoot distortion and abscission of the growing terminals. The psyllid also can vector the bacterium Candidatus Liberibacter spp., causal agent of citrus greening or ‘huanglongbing’ (Xu et al. 1988a). Citrus greening is considered to be the most serious disease of citrus in Asia; it has recently been reported from Brazil (Coletta-Filho et al. 2004), and poses a serious threat to citrus in Florida and Puerto Rico.
Two parasitoids have been imported and released in North America for control of D. citri, the endoparasitic encyrtid Diaphorencyrtus aligarhensis (Shafee, Alam & Agarwal) and the ectoparasitic eulophid Tamarixia radiata (Waterson). Tamarixia radiata established itself in Florida and apparently arrived spontaneously in Puerto Rico (unpublished data) and Texas (French et al. 2001).
Predators in the family Coccinellidae (Coleoptera) have been shown to be responsible for a considerable degree of natural control in Florida (Michaud 2004). Michaud (2001) also presented field data indicating that Olla v-nigrum Mulsant and Harmonia axyridis (Pallas), coccinellid species capable of feeding and reproducing on D. citri, increased in relative abundance in psyllid-infested citrus groves. The following study was undertaken to provide baseline data on relative abundance of coccinellid species in Puerto Rican citrus groves and to evaluate the ability of the most abundant species to feed on D. citri.
Materials and Methods
Relative Abundance of Coccinellids on Citrus
Observations were carried out from early April to early July 2003, in citrus groves at the Agricultural Experimental Station in Adjuntas (18’10”N, 66’29”W. 457 m). The citrus groves of Adjuntas are well established and cover a significant proportion of both sides of the valley in which the experimental station is located. Because of the altitude and cooler weather, new growth in citrus is largely restricted to the main flush in the first half of the year. This is the time that population outbreaks of herbivorous insects such as aphids occur, with the subsequent increase of lady beetle numbers. The sampling period was designed to encompass this period of greater lady beetle activity. Thirty citrus trees were randomly selected and inspected between 8.00 a.m. and 12.00 a.m. every other week for a total of 180 trees, based on direct counts of all coccinellid adults, larvae, or pupae encountered. The relative abundance of each species was calculated as a proportion of the total coccinellids counted.
Prey Acceptability and Prey Preference
Psyllid nymphs used in the experiments were collected from a greenhouse colony established from field-collected individuals and reared on Murraya panniculata at the Río Piedras Agricultural Experiment Station. Nymphs of brown citrus aphid (BCA) Toxoptera citricida (Kirkaldy) were collected from infested citrus flushes at the Adjuntas Agricultural Experimental Station and used directly in the experiments. Coccinellid adults of the various species were collected from citrus groves at the Adjuntas Agricultural Experimental Station in Puerto Rico during the period from April to June, 2003 and maintained on Ephestia kuehniella Keller (Lepidoptera: Pyralidae) eggs at Adjuntas until needed. Chilocorus cacti, Cladis nitidula, Cryptolaemus montrouzieri, and Scymnus sp. could not be reared on E. kuehniella eggs, and so were collected directly from the field when needed for experiments.
The following eight species of cocinellids were tested for prey acceptance and prey preference in choice and no-choice tests: Cycloneda s. limbifer L., Coelophora inaequalis F., Cladis nitidula F., Chilocorus cacti L., Coleomegilla innonata Mulsant, Cryptolaemus montrouzieri Mulsant, Hippodamia convergens Guerin, and Scymnus sp. Individual adult coccinellids were starved for 24 h and then confined in Petri dishes (5 cm diam.) with one of the following prey configurations:10 psyllid nymphs only, 10 aphid nymphs only, or a combination of 5 psyllid nymphs and 5 aphid nymphs. The life stages of the two prey species were chosen to be of similar size (usually third-instars). Petri dishes were lined with white paper to assist with the visual assessment of feeding events. An experiment was judged to have been completed in the choice tests once the adult coccinellid had consumed all of one or the other prey species in the Petri dish. In the no-choice tests, the experiments were terminated after 7 h. Ten replicates of the tests were carried out and each adult coccinellid was tested only once. All experiments were carried out under controlled conditions at 25 ± 1°C and 75 ± 10% RH.
For determining the differences in coccinellid abundance, the square root transformation was used and then Tukey’s multiple comparison test applied in Proc GLM in SAS (1999). For the feeding data comparisons in the combined host experiments, normality was examined by the Shapiro-Wilk test and plot functions of Proc Capability (SAS 1999). The paired analysis was conducted based on the signed rank test in the same SAS procedure. For the consumption rate comparisons in the single host experiments, normality was determined by Proc Capability. Tukey’s multiple comparison test in Proc GLM was then used to compare coccinellid consumption.
Relative Abundance of Coccinellids
A total of eight species of coccinellids were found in citrus groves of the Adjuntas experimental research station during the course of the survey; C. inaequalis, C. s. limbifer, C. innonata, C. cacti, C. nitidula, Scymnus sp., H. convergens, and C. montrouzieri. The number of coccinellids observed was relatively constant over the 3-month study period (Fig. 1). Coelophora inaequalis and Cycloneda s. limbifer were the most common species found over the entire study period with the remaining species found on a regular basis, but in much lower numbers (Fig. 2).
Additionally, two species of Hymenoptera parasitized the coccinellids. Homalotylus sp. near terminalis Say (Hymenoptera: Encyrtidae) emerged from a C. s. limbifer pupa, while Oomyzus sp. near scaposus Thomson (Hymenoptera: Eulophidae) emerged from C. montrouzieri (larva) and C. s. limbifer (pupa) (determinations by M. W. Gates, pers. comm.).
All eight coccinellid species in the no-choice tests fed on both host species with no rejection of any offered prey type, although variation in the quantity and rate of prey consumption was observed. During the allotted interval there were few differences in relative amounts of prey consumed in no-choice tests, with the possible exception of C. nitidula, which ate twice as many psyllids as aphids (Table 1). Hippodamia convergens demonstrated the highest rate of aphid consumption whereas C. cacti, C. nitidula, and Scymnus sp. showed the lowest rate (Table 2). The highest rate of psyllid consumption was observed with C. innonata, although not significantly different from the rates of all others except C. cacti and Scymnus sp. (Table 3).
Two of the coccinellid species examined, C. s. limbifer and C. inaequalis, showed a strong preference for the brown citrus aphid, while C. nitidula, C. cacti, C. innonata, and C. montrouzieri showed significant preference for the Asian citrus psyllid (Table 4). The remaining two species, H. convergens and Scymnus sp. showed no preference for either prey species.
Of the seven species of Coccinellidae listed by Michaud (2004) as common in Florida citrus, only the exotic Coelophora inaequalis also was found in this survey. However, C. s. limbifer was considered by Gordon (1985) as a subspecies of C. sanguinea, the most commonly encountered ladybeetle in Florida citrus groves prior to the invasion of Harmonia axyridis Pallas (Muma 1953; Michaud 2001; Michaud 2002). Cycloneda limbifer and C. inaequalis, both aphid-feeders, were the two most abundant species encountered in this study and also the most abundant reported from Puerto Rican citrus groves by Michaud and Browning (1999). The predominance of aphid feeders followed by scale feeders are characteristics shared by the coccinellid fauna in citrus groves of Florida and Puerto Rico.
Our results showed that all ladybeetle species, with the exception of C. nitidula, consumed similar quantities of either host species presented separately. Cladis nitidula consumed more psyllids and also was the species showing greatest preference for D. citri when offered a choice between prey. Thus, choice and no-choice experiments were consistent in indicating C. nitidula’s preference for D. citri over T. citricida. A preference for T. citricida over D. citri was strongly expressed by Cycloneda s. limbifer, and to a lesser extent by Coelophora inaequalis, the two most abundant ladybeetles in this study, although both species consumed equal amounts of both prey when given no choice. Michaud & Olsen (2004) found that Cycloneda sanguinea also fed on D. citri as both larva and adult, but that larval development time was almost doubled compared to a diet of E. kuehniella and that female C. sanguinea stopped ovipositing following transferal to the D. citri diet. Further studies would be necessary see whether a D. citri diet negatively impacts the larval and reproductive performance of Cycloneda s. limbifer as it does C. sanguinea.
Prey suitability for Coccinellidae is a complex issue that goes beyond the scope of prey acceptance and prey preference studied here, and generalizations such as “aphid feeders” and scale feeders” are overly simplistic (Hodek 1996). Nevertheless, the species demonstrating the greatest degree of acceptance and preference for BCA, such as C. s. limbifer, C. inaequalis, and H. convergens were not surprisingly those considered to be aphid feeders. Those that preferred and/or most readily accepted ACP or showed no clear tendencies were either known to feed principally on other prey such as Diaspididae (C. cacti), and Pseudococcidae (C. montrouzieri), or had feeding habits that were largely undocumented (C. nitidula, Scymnus sp.).
Diaphorina citri was rarely encountered in Adjuntas during the course of this study, and ACP populations have remained generally low in Puerto Rico on both orange jasmine and citrus. Coccinellids typically respond to dense prey populations whereas parasitoids with narrow host ranges such as T. radiata are expected to track their host population at low densities. It is difficult to ascertain, in retrospect, why ACP never achieved high infestations in Puerto Rico as it has Florida; however, preliminary evidence suggests that T. radiata may be responsible for holding the psyllid in check in Puerto Rico (R. Pluke, unpublished data). In any case, the coccinellid guild present in Puerto Rican citrus, with its demonstrated ability to consume ACP and its similar mix of species in regard to feeding habits to the coccinellid guild in Florida, would likely respond positively to any future increase in psyllid numbers.
We thank the personnel of the Río Piedras and Adjuntas agricultural experiment stations for invaluable support and technical assistance, in particular, W. González, O. Cintrón, E. M. Pérez, F. Ortíz, F. Sherwood, and Dr. K. Nelo. We are grateful to M. Thomas of the Florida State Collection of Arthropods for identification of ladybeetles and M.W. Gates of USNM in Washington DC for identification of parasitoids. Research funded under grant number 51099 provided by the USDA--CSREES Tropical/Subtropical Agricultural Research Program. Florida Experiment Station Number R-10588.
Table 1. Mean percentage ± SD of 10 prey items consumed in no choice tests in 7 h.
Mean ± SD consumption rates of coccinellids on the brown citrus aphid in no choice tests.
Mean ± SD consumption rates of coccinellids on the Asian citrus psyllid in no choice tests.
Mean percentage (of 5 host individuals) eaten by coccinellids in choice tests.