The biological control agent Tamarixia radiata (Waterston) (Hymenoptera: Eulophidae) is being mass-produced for field release in several countries, including Brazil (Parra et al. 2016), Costa Rica, Mexico, and the United States (Flores & Ciomperlik 2017; Kerr et al. 2017), to control Diaphorina citri Kuwayama (Hemiptera: Liviidae), a vector of Candidatus Liberibacter asiaticus, the putative bacterial pathogen that causes huanglongbing disease in citrus. Hypothetically, a higher number of the parasitoids relative to the psyllid host would increase parasitism, super-parasitism, and overall mortality of D. citri nymphs. However, this may not be the case because intraspecific competition can increase at higher parasitoid population levels (Visser et al. 1999; Harvey et al. 2013). Parasitoids compete for hosts under these conditions, prolonging the time required to locate and parasitize an acceptable host. As the relative number of parasitoids increases, more encounters with conspecifics occur and searching behavior is disrupted, leading to a decrease in searching efficiency and the rate of parasitism (Beddington 1975). Consistent with other parasitoid species (Ueno & Tanaka 1994), T. radiata females can discriminate between parasitized and unparasitized host nymphs due to marking by the females (Chen et al. 2016a). This differentiation includes nymphs parasitized by the same female or conspecific females, and super-parasitism is avoided by this capability. Single T. radiata females parasitized fewer third, fourth, and fifth instar D. citri nymphs in proportion to an increase in the overall number of nymphs (Sule et al. 2014), but no comparison was conducted in regard to multiple T. radiata females, as would be found in a mass-rearing system and in the field. The objective of this study was to compare the levels of parasitism and super-parasitism of D. citri fourth and fifth instar nymphs exposed to an increasing number of T. radiata females.

To compare the effects of varying release ratios of T. radiata to D. citri, while maintaining a 2:1 sex ratio (female:male) for T. radiata, treatments of 9 (6♀:3♂), 18 (12♀:6♂), or 45 (30♀:15♂) T. radiata were released into cages after adequate time was given for progeny, of a standardized ovipositing cohort of 60 adult D. citri, to develop based on the following methods. The study was conducted using 3 equivalent potted orange jasmine plants, Murraya paniculata (L.) Jack (Rutaceae), that were about 1 m tall, and had an ample number of young shoots with flush. The D. citri females oviposit exclusively on flush, rather than mature leaves or stems. To prepare each plant, 4 comparable shoots were selected and the remaining shoots were removed. Plants then were rinsed thoroughly with tap water, shaken dry, and placed into individual 24.5 W × 24.5 L × 63.0 cm H BugDorm insect rearing cages (BioQuip Products Inc., Rancho Dominguez, California, USA). The cages were held at 26 ± 1 °C, 50% RH, and 14:10 h (L:D) photoperiod in an environmental chamber (model number I36VLC8, Percival Scientific, Perry, Iowa, USA).

The D. citri adults were obtained from a stock colony collected in Florida, and maintained for more than 10 yr on orange jasmine plants at the Florida Department of Agriculture and Consumer Services, Division of Plant Industry, Biological Control Laboratory in Gainesville, Florida, USA (Kerr et al. 2017). An open 50 mL plastic vial containing 60 psyllid adults aspirated from the colony was placed upright on the potting media, and leaned against the stem of the orange jasmine plant in each experimental cage. The psyllids were given 3 d to ascend the plants and oviposit on the flush before being removed from the cage. The D. citri eggs and nymphs on the flush were held for 10 d to obtain the late fourth and fifth instar nymphs that are preferred by T. radiata for parasitism (Sule et al. 2014).

The T. radiata colony was established in 1998 at Florida Department of Agriculture and Consumer Services, Division of Plant Industry with specimens imported from Taiwan and southern Vietnam (Skelley & Hoy 2004). Adult parasitoids < 24 h old were aspirated at a 2:1 sex ratio (female:male) into 3 individual 50 mL glass vials. The vials were provisioned with a 0.6 cm × 5 cm strip of paper towel saturated with honey and blotted. Individual cohorts of 9 (6♀:3♂), 18 (12♀:6♂), or 45 (30♀:15♂) T. radiata were held in the capped vials for a 24 h mating period, and subsequently released into separate cages containing the D. citri nymphs that had been held 10 d to develop on the flush. The parasitoids were given 4 d to oviposit on the nymphs, after which they were removed from the cages.

For each of the three treatments 9 (6♀:3♂), 18 (12♀:6♂), or 45 (30♀:15♂) T. radiata, the 4 shoots with the D. citri nymphs and supporting flush were removed from each cage before adult parasitoid emergence at 10 to 12 d post-oviposition (Chien et al. 1991; Gomez-Torres et al. 2012). The nymphs on the flush were measured under a dissecting microscope (Leica S8 APO Stereomicroscope, Leica AG, Wetzlar, Germany), and classified to instar based on head capsule width: 0.40 to 0.49 mm and > 0.50 mm denoting fourth and fifth instar, respectively. The relatively few earlier instar nymphs were discarded. Each nymph was inverted to count the number of T. radiata eggs or larvae present on them to establish if it was unparasitized, parasitized, or super-parasitized (Chen et al. 2016a). The mean number of unparasitized, parasitized and super-parasitized nymphs for each treatment was divided by the total number of nymphs to determine the proportion in each category (5 replicates). Levels of parasitism were compared separately for fourth and fifth instar D. citri nymphs. Data were analyzed at a significance level of 0.05 with a 1-way ANOVA; when differences occurred, means were differentiated using Tukey's honestly significant difference (HSD) test

The mean number of unparasitized fourth and fifth instar D. citri nymphs remained similar across all treatments. However, the number of unparasitized fourth instar nymphs significantly increased when increasing the number of parasitoids released from 18 (12♀:6♂) to 45 (30♀:15♂) T. radiata (Table 1). The mean number of parasitized fourth and fifth instar nymphs decreased as the number of parasitoids relative to hosts increased. The highest level of parasitism of D. citri fourth and fifth instar nymphs resulted when 9 (6♀:3♂) T. radiata were released. The level of parasitism decreased almost proportionally, with 2.65, 2.34, and 3.40 × more parasitized fifth than fourth instar nymphs for 9 (6♀:3♂), 18 (12♀:6♂), and 45 (30♀:15♂) T. radiata, respectively. Super-parasitism of fourth instar nymphs took place only with 45 (30♀:15♂) T. radiata, and of fifth instar nymphs with 18 (12♀:6♂) and 45 (30♀:15♂) T. radiata. The percentage of parasitized D. citri nymphs for 9 (6♀:3♂), 18 (12♀:6♂), and 45 (30♀:15♂) T. radiata was 65%, 54%, and 13% for the fourth instar, and 92%, 89%, and 56% for fifth instar.

The relative number of T. radiata released into cages containing fourth and fifth instar D. citri nymphs, produced based on a standardized 60 D. citri adults, was increased to determine the amount of parasitism, super-parasitism, and its impact on mass rearing operations. The level of parasitism of the host nymphs decreased as the number of parasitoids relative to nymphs increased, probably resulting from hostfeeding and competition for nymphs suitable for oviposition (Chien et al. 1991; Chen et al. 2016b). More fifth instar nymphs were parasitized irrespective of the number of parasitoids released, indicating that instar was preferred by T. radiata females.

Our study expands on previous findings, and indicates that increasing the number of T. radiata beyond optimal parastoid:host ratios likely causes extreme competition among females for hosts. This competition could have resulted in super-parasitism, although it was infrequent. Tamarixia radiata is known to exhibit higher levels levels of super-parasitism when hosts are relatively scarce (Chen et al. 2016a). For mass rearing T. radiata, 1 female parasitoid per 45 nymphs was suitable; about 36% of the nymphs were parasitized, and 57% killed by host feeding (Skelley & Hoy 2004). A maximum of 60% parasitism has been achieved using 1 female for 80 nymphs (Chen et al. 2017). Experimentally, however, 73% parasitism has been accomplished using 1 T. radiata female per nymph (Chen et al. 2016b). We observed a maximum of 92% parasitism of fifth instar nymphs at the lowest number of females and, as nymphs became less abundant relative to T. radiata females, the level of parasitism decreased along with an increase in the occurrence of superparasitism.

Table 1.

Mean (± SD) number of unparasitized, parasitized, and super-parasitized fourth and fifth instar Diaphorina citri nymphs, and % parasitized nymphs, after 4 d exposure to a 2:1 sex ratio (female:male) of Tamarixia radiata ^a.

Partial funding was provided by a grant from the USDA, NIFA, Crop Protection and Pest Management Competitive Grants Program. Yasaman Moghaddassi (University of Tehran), Julio Medal (Florida Department of Agriculture and Consumer Services, Division of Plant Industry), and Mrittunjai Srivastava (Florida Department of Agriculture and Consumer Services, Division of Plant Industry) provided helpful reviews of the manuscript.