The biology of Psyllaephagus bliteus Riek (Hymenoptera: Encyrtidae), a parasitoid of the red gum lerp psyllid (Hemiptera: Psylloidea)

Biological Control 32 (2005) 228–235

www.elsevier.com/locate/ybcon

The biology of Psyllaephagus bliteus Riek (Hymenoptera: Encyrtidae), a parasitoid of the red gum lerp psyllid

(Hemiptera: Psylloidea)

Kent M. Daanea,¤, Karen R. Simea, Donald L. Dahlstena, John W. Andrews Jr.b,Robert L. Zuparkoc

a Center for Biological Control, Division of Insect Biology, Department of Environmental Science, Policy, and Management, University of California, Berkeley, CA 94720, United States

b Insectary and Quarantine Facility, College of Natural Resources, University of California, Berkeley, CA 94720, United Statesc Essig Museum, College of Natural Resources, University of California, Berkeley, CA 94720, United States

Received 27 July 2004; accepted 27 September 2004Available online 5 November 2004

Abstract

The red gum lerp psyllid, Glycaspis brimblecombei Moore (Hemiptera: Psylloidea), is native to Australia, where it feeds uponEucalyptus species. It Wrst appeared near Los Angeles, California, in 1998, and soon spread throughout the state. A biological controlprogram directed against the psyllid was initiated and Psyllaephagus bliteus Riek (Hymenoptera: Encyrtidae) was imported fromAustralia and released in California. During quarantine screening, the taxonomic status of Psyllaephagus quadricyclus Riek wasassessed by one of us (RLZ) and is proposed here as a new junior synonym for P. bliteus. The experiments discussed herein providebasic biological information on P. bliteus to supplement and improve the control program. We found that P. bliteus can oviposit intopsyllid nymphs of any age but prefers third and fourth instars. Observations of host-handling behavior suggest that the large lerps ofWfth instar psyllids increase host-handling time, thereby impeding oviposition and providing some protection from parasitism.Female P. bliteus were observed host-feeding on all psyllid nymphal development stages. Adults are relatively long-lived and, at con-stant temperatures of 17, 21, 23, 26, and 32 °C, longevity is a negative linear function of temperature. Females lived signiWcantlylonger than males. Adult females can live for several months, provided with hosts and held under glasshouse conditions (22 § 3 °C),however, maximum egg deposition occurred within 22 days after adult emergence. Studies of larval development show that P. bliteusis a koinobiont and larval development is not initiated until the host reaches the late fourth or early Wfth instar. 2004 Elsevier Inc. All rights reserved.

Keywords: Psyllaephagus bliteus; Psyllaephagus quadricyclus; Glycaspis brimblecombei; Red gum lerp psyllid; Eucalyptus; Biological control;Parasitoid biology

1. Introduction

Psyllaephagus Ashmead (Hymenoptera: Encyrtidae)is a cosmopolitan genus containing over 200 describedspecies with perhaps as many as 1000 in all (Noyes and

* Corresponding author. Fax: +1 559 646 6593.E-mail address: [email protected] (K.M. Daane).

1049-9644/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.biocontrol.2004.09.015

Hanson, 1996). Their greatest taxonomic diversiWcationoccurs in Australia, where the endemic Psyllaephagusspecies, like almost all others in the genus, attacknymphs of Psylloidea, and a few are reported as hyper-parasitoids attacking other Psyllaephagus species (Noyesand Hanson, 1996; Riek, 1962). Their psyllid hosts, inturn, are often specialist on Myrtaceae, mainly speciesgroups and subgenera of Eucalyptus (Moore, 1988; Yen,2002). The eVectiveness and speciWcity of the Australian

K.M. Daane et al. / Biological Control 32 (2005) 228–235 229

Psyllaephagus has made them especially attractive candi-dates for use in classical biological control programsthat target psyllid pests of Eucalyptus, and they havebeen used for programs in California, Mexico, and theBritish Isles (Chauzat et al., 2002; Dahlsten et al.,1998a,b; Paine et al., 2000).

Psyllaephagus bliteus Riek was widely released in Cal-ifornia from 2000 through 2002 to control the Austra-lian red gum lerp psyllid, Glycaspis brimblecombeiMoore (Hemiptera: Psylloidea) (Paine and Millar, 2002;Paine et al., 2000), after quarantine studies indicated thatit would speciWcally attack this host (Dahlsten et al.,unpublished data). This psyllid was Wrst discovered inLos Angeles County in 1998 (Brennan et al., 1999) andhad spread throughout California by 2000 (Paine andMillar, 2002) and Mexico by 2002 (J. Guerra, pers.comm.). Like other Glycaspis spp. and some related psyl-lids, the nymphs of G. brimblecombei are characterizedby the shelters (lerps) they construct from excreted car-bohydrates and proteins (Ernst and Sekhwela, 1987;Gilby et al., 1976; Moore, 1961). The accumulation ofthe sticky lerps on leaves and underneath infested treescreates a nuisance, while heavy infestations lead to defo-liation, branch dieback, and occasionally tree death(Paine et al., 2000). The main host of G. brimblecombei inCalifornia, the river red gum (Eucalyptus camaldulensisDehnh.) (Brennan et al., 2001), is one of the most com-monly planted shade and windbreak trees in both urbanand rural environments and is also grown commerciallyfor fuel wood (Cockerham, 2004; Helms, 1988). The bio-logical control program with P. bliteus against the redgum psyllid has been largely successful in California’scoastal regions, but has provided, to date, only sporadiccontrol in some of the warmer interior regions (Dahlstenet al., unpublished data). We report herein on basic bio-logical information on P. bliteus that was collected tosupplement and improve the control program, especiallyinsectary operations and release strategies.

During quarantine screening, one of us (RLZ) alsoreviewed the taxonomic status of P. bliteus. Amongst thespecimens that emerged in quarantine from material col-lected in Australia were male and female Psyllaephagusspecimens: the females matched Riek’s (1962) descrip-tion of P. bliteus and the males that of Psyllaephagusquadricyclus Riek. These males and females wereobserved to mate and reproduce, indicating that the spe-cies descriptions required review. Riek (1962) describedP. bliteus from a series of females from Canberra (Aus-tralian Capital Territory), noting that it also occurred inSouth Australia, New South Wales and Queensland.Psyllaephagus bliteus had been reared from undeter-mined Glycaspis ( D Spondyliaspis) species on Eucalyptusrossii, E. blakelyi, E. leucoxylon and E. camaldulensis, aswell as from Creiis costatus (Froggatt) on E. blakelyi. Inthe same paper, he described P. quadricyclus from aseries of males from Canberra, also recording it from

South Australia and New South Wales. These had beenbred from undetermined Glycaspis species on Eucalyptusmelliodora, E. rossii, E. blakelyi, E. viminalis, and anundetermined Eucalyptus species. Thus, these specieswere essentially sympatric (one paralectotype of P. bli-teus and both paratypes of P. quadricyclus were collectedfrom the same site, “Black Mountain,” within a few daysof each other) and probably share the same hosts. Weexamined the lectotype (assigned by John Noyes) andthree paralectotypes of P. bliteus, and the holotype(male, not female as reported in Riek, 1962) and twoparatypes of P. quadricyclus. These specimens were con-speciWc with the females and males, respectively,imported to California—including F1 and F3 specimensfrom the importation cultures, as well as Weld-collectedspecimens. Additional veriWed P. bliteus specimens havebeen reared from G. brimblecombei on a Eucalyptus sp.from Sao Paulo, Brazil, in 2003, and from Ulupalakua,Maui, Hawaii, in 2004. The junior author (RLZ) hereinproposes the synonymization of P. quadricyclus Riekunder P. bliteus Riek as follows:

Psyllaephagus bliteus RiekPsyllaephagus bliteus Riek, 1962: 722–723 (Holotype

female, Australia, ANIC)Psyllaephagus quadricyclus Riek, 1962: 751–752 (Holo-

type male, Australia, ANIC) syn. nov.

2. Materials and methods

2.1. Insect and plant materials

Laboratory cultures of red gum lerp psyllids werederived from Weld-collected material on infested red gumeucalyptus in Alameda and Sacramento Counties, Cali-fornia. The psyllids were reared on potted (2.2 L) redgums that were either grown from seed (about 80% ofthe plants) or purchased from commercial nurseries. Redgums were maintained in a glasshouse at the Universityof California, Berkeley, Insectary and Quarantine Facil-ity, at 22 § 3 °C and with natural lighting. Small redgums (0.5–1 m) were inoculated with adult psyllids and,by manipulating the numbers and periods of ovipositingfemales, we were able to produce trees infested with 200–800 psyllids of the desired developmental stages. Weperiodically reintroduced Weld-collected psyllids toreduce inbreeding. These infested trees were used in alldescribed experiments.

The P. bliteus used in experiments were reared frompsyllids that were Weld-collected on red gum foliage inthe Ardenwood Regional Preserve (Alameda County,California). This parasitoid population had originatedfrom P. bliteus imported from southern Australia in1999, reared in quarantine and released throughoutCalifornia in 2000 and 2001 (Paine et al., 2000). Each

230 K.M. Daane et al. / Biological Control 32 (2005) 228–235

week, infested red gum foliage was collected and placedinside wood-sided sleeve cages (45 £ 45 £ 45 cm) thatwere provisioned with dilute honey (1:1 honey and waterby volume). Unless noted otherwise, adult P. bliteus werecollected daily as they emerged, placed in glass vials thatwere provisioned with dilute honey, and held in an incu-bator at 12 § 2 °C for 1–10 days until used. On each daywe collected all parasitoids found; therefore, we assumethe collected parasitoids were less than one day old.Voucher specimens of red gum lerp psyllids and P. bli-teus have been deposited in the Essig Museum, Univer-sity of California, Berkeley.

2.2. Host-stage preference for oviposition

Infested trees were selected that had 300–500 psyllidnymphs, with the population comprised of all Wve devel-opment stages in similar proportions, as estimated bylerp size. For each replicate, 3–4 infested red gums wereplaced in cloth-sided tree cages (32 £ 45 £ 96 cm) andheld at 25 § 2 °C under Xuorescent lighting with a 16:8(L:D) photoperiod. At the start of each trial, 15–20female and 4–5 male P. bliteus were released into thecage. After 24 h, the parasitoids were removed and allpsyllids found were transferred to 70% ethanol. Todetermine whether or not the exposed psyllids were par-asitized and their development stage, the psyllids weretransferred to a clearing solution of chloralphenol (10 gphenol, 10 g chloral hydrate, and 3–5 ml distilled water).After 24 h in chloralphenol, the psyllid’s body becomestransparent and P. bliteus eggs are visible under a dis-secting microscope. At the same time, we determine thepsyllid development stage by counting the number ofantennal segments, as described by Moore (1961). Theparasitism rates for psyllids in each of the Wve nymphalstages were compared. There were Wve replicates, withnew parasitoids used for each replicate.

2.3. Host-handling behavior

We used direct observations of the host-handlingbehavior of experienced P. bliteus to further assess host-stage preference and oviposition success. An individualfemale was provided with a single red-gum leaf infestedwith 70–100 lerps, which were estimated (by size) tohouse psyllids in all Wve developmental stages and insimilar proportions. The leaf was placed in a transparentplastic cylinder (21 cm long £ 7 cm diameter), which wasventilated by a cloth mesh covering one end and a 4 cmdiameter hole in the side. The P. bliteus was added andher host-handling behavior was observed under a dis-secting microscope for 1 h, beginning when the parasit-oid Wrst touched a psyllid. The tested P. bliteus werediscarded if they showed no interest in the psyllids dur-ing the initial 10 min of observation. Oviposition andhost-feeding events and duration were recorded. During

each trial, the psyllid’s developmental stage was esti-mated by the lerp size. At the end of each trial, the lerpswere removed and presence or absence of psyllids wasnoted, as red gum psyllid nymphs often move from theirlerps (Moore, 1961). The psyllids were then transferredinto 70% ethanol, cleared in chloralphenol, and the pres-ence of P. bliteus eggs and the development stages ofparasitized nymphs were recorded, as described previ-ously. We completed 18, 1-h trials.

2.4. Adult longevity and fecundity

Adult longevity was determined at Wve temperatures(17, 21, 23, 26, and 32 °C). Males and females were testedseparately. Newly emerged (12–16 h) P. bliteus were col-lected, placed in 35 ml glass vials that were provisionedwith dilute honey–water, and randomly assigned to atemperature treatment. Thereafter, the vials werechecked every 24 h and the P. bliteus condition (live ordead) was recorded. The honey–water was refresheddaily and the vials were changed weekly. Temperaturecabinets maintained temperatures (T) at T § 1 °C, with a16:8 (L:D) photoperiod. We tested 20 females and 7–10males at each temperature.

Adult fecundity, as estimated by life-time egg deposi-tion, was also determined. Newly emerged (12–16 h)female and male P. bliteus were collected and heldtogether for 24 h. The mated females were individuallyisolated in clear plastic tubes (4 £ 8 cm) that eachenclosed a single infested leaf on a potted red gum tree inthe glasshouse (22 § 3 °C). Each leaf was infested with10–30 psyllids, primarily in the third instar developmentstage. The plastic tubes were covered with a nylon meshon the open end, to provide ventilation, and plastic foamon the other end, to wrap around the leaf petiole andenclose the leaf. The parasitoid was transferred to a newleaf every 2 days throughout her lifetime. After eachtransfer, the exposed psyllids were placed into 70% etha-nol, cleared in chloralphenol, and the presence of P. bli-teus eggs and the development stages of parasitizedpsyllids were recorded, as described previously.

2.5. Immature development time

The development time of P. bliteus immature stageswas assessed at three constant temperatures: 22, 26, and30 °C. For each trial, infested red gums were selected thathad 100–300 psyllids; all stages of psyllids were presenton each tree, but the populations were dominated bythird and fourth instars, which were identiWed by thehost-stage preference study as the preferred stages foroviposition. The trees were placed individually in thecloth-sided tree cages and 25–30 female P. bliteus wereplaced in each cage for a 48 h period (and then removed).This oviposition period occurred at 25 § 2 °C. The treeswere then randomly assigned to diVerent temperature

K.M. Daane et al. / Biological Control 32 (2005) 228–235 231

cabinets (treatment). The temperature cabinets usedmaintained temperatures at T § 1°C, with a 16:8 (L:D)photoperiod. Thereafter, 20 psyllids were collected every3–4 days, placed into 70% ethanol and later cleared inchloralphenol, as described previously. The presence ofP. bliteus eggs or larvae and the development stages ofparasitized psyllids were recorded. We completed fourtrials at each temperature.

2.6. Statistics

Results are presented herein as means per treatment(§SEM). Treatment eVects were analyzed using analysisof variance (ANOVA), with treatment means separatedusing Tukey’s HSD test (three or more treatments) or a ttest (two-way comparisons) at P < 0.05. We used regres-sion analysis to describe the relationship between P. bli-teus host-handling duration and psyllid developmentstage, and adult P. bliteus longevity and temperature.

3. Results and discussion

3.1. Host-stage preference for oviposition

From 2279 exposed and examined psyllids, we found164 P. bliteus eggs in 155 hosts, with an average7.9 § 2.7% parasitism across all trials. Percentage para-sitism was determined by the mean of means for eachreplicate, which ranged from 1.3 to 31.7%. We believe thewide range in percentage parasitism was an artifact ofthe variation in number of psyllids provided in each rep-licate, while the number of parasitoids remained thesame. Because the psyllids are covered by lerps, beforethe trial started we could only estimate the actual num-ber of live psyllids under the lerps; in some trials thereproved to be more empty lerps. Of the parasitized psyl-lids, signiWcantly more of the P. bliteus eggs were recov-ered from third (49.6 § 10.5%) and fourth (35.5 § 7.3%)instar psyllids than from the Wrst (0.23 § 0.23%), second(8.12 § 3.15%), and Wfth (6.60 § 2.93%) instars (Fig. 1).There was no signiWcant diVerence in the number of par-asitized Wrst, second or Wfth instar psyllids (Fig. 1); how-ever, only a single Wrst instar was parasitized. There wasno signiWcant diVerence in the numbers of each psylliddevelopmental stage provided (df D 4, 30, F D 1.789,P D 0.156) and there was an overabundance of psyllids ineach development stage, with the exception of one trialwhere no Wrst instars were found. Therefore, the encoun-ter frequency of the diVerent psyllid development stagesdid not impact oviposition decisions and we concludethat P. bliteus preferentially deposits eggs in third andfourth instar psyllids, although it will attack all nymphalstages in a host-choice test arena. A similar Wnding hasbeen reported for Psyllaephagus pulvinatus (Waterston)on the citrus psylla, Trioza erytreae (Del Guercio), which

strongly prefers to attack third instars (while theoVspring feed on and emerge from Wfths) (McDaniel andMoran, 1972), and for P. pilosus Noyes on the on bluegum psyllid, Ctenarytaina eucalypti Maskell (Dahlstenet al., 1998b) and P. euphyllurae (Masi) on Euphylluraolivine (Chermiti et al., 1986), though the latter twoappear to prefer older (fourth and Wfth instar) psyllids.In contrast, P. yaseeni Noyes typically attacks smaller(Wrst and second instar) psyllids (Patil et al., 1993).

We observed P. bliteus eggs deposited in the psyllidthorax and abdomen, without any apparent preferencefor either region. Most of the parasitized psyllids (94.5%)in this trial had a single egg deposited, while eight thirdinstars and a single Wfth instar had two eggs; however, inthis trial there was an overabundance of host-materialprovided. From Weld-collected material we commonlyfound two eggs per host and in a laboratory study withrestricted host-material provided we found up to 26 eggsper host. In both of these situations, the number of avail-able hosts was restricted and adult P. bliteus may havecompeted for host-material. Moreover, in all laboratorytrials and Weld collections, we have never reared morethan one P. bliteus adult per host nor found more thanone large P. bliteus larva per host. Therefore, we believethat hosts with more than one egg are superparasitizedand, although some solitary parasitoids lay multiple eggclutches (Rosenheim and Hongkham, 1996), we preferthe simpler explanation that competition for a limitedresource resulted in more than one egg per host. Regard-less, our experimental designs do not allow us to deter-mine whether psyllids containing more than one egg

Fig. 1. Psyllaephagus bliteus oviposition success in diVerent host devel-opmental stages, as indicated by the percentage egg deposition(§SEM) of parasitized red gum lerp psyllids, was signiWcantly diVerent(F D 12.48, df D 4, 25, P < 0.001). Above each bar, means followed bydiVerent letters are signiWcantly diVerent (Tukey’s HSD test, P < 0.05).

232 K.M. Daane et al. / Biological Control 32 (2005) 228–235

were attacked by one, or more than one, parasitoid. Thisobservation implies that, during host-handling, P. bliteushas either little or no ability to discriminate previouslyparasitized hosts. Several wasps probed empty lerps,which suggests that there is little recognition of host-condition under the lerp and that chemical cues on thelerp itself may initiate oviposition. We do not know, incases of superparasitism, if the numbers of parasitoidlarvae are reduced through resource or direct competi-tion.

3.2. Host-handling behavior

Direct observations of host-handling revealed adiVerent pattern of host-stage preference as comparedwith that revealed by dissecting individuals for depositedeggs (Fig. 1). Here, we estimated the lerp size (and stageof the psyllid underneath) during the ovipositionattempt, and after clearing and examining the psyllid welater determined the nymphal stage and whether the ovi-position attempt resulted in successful egg deposition.This study showed that the P. bliteus made signiWcantlymore oviposition attempts in the larger psyllids (third toWfth instars) (Fig. 2), and that host-handling time (asmeasured by oviposition attempts either for host-feedingor oviposition) was a positive linear function of psylliddevelopmental stage (y D ¡1.37 + 1.29x, r2 D 0.96,df D 1, 4, F D 82.92, P D 0.003). The pattern is essentiallythe same as that found in the host-preference study withone exception—the largest lerps (Wfth instar nymphs)

Fig. 2. Psyllaephagus bliteus oviposition attempts in diVerent hostdevelopmental stages, as directly observed during a 1 h period(§SEM) in a host-choice arena of red gum lerp psyllids, was signiW-cantly diVerent (F D 11.32, df D 4, 53, P < 0.001). Means followed bythe same letter are not signiWcantly diVerent (Tukey’s HSD test,P < 0.05).

were the most commonly attacked. However, when theexposed psyllids were cleared and examined, we found P.bliteus eggs in third and fourth instar psyllids, but notWfth instars.

In the comparison of oviposition duration, we usedmeasurements from only those P. bliteus that ovipositedin both small and large lerps during the same 1 h period,to conduct a paired t test. Handling times for ovipositionwere signiWcantly shorter on the smaller (Wrst, second,third instar) than larger (fourth and Wfth instar) lerps,measuring 69.0 § 13.3 and 232.3 § 47.1 s, respectively(t D ¡3.801, P D 0.003). We believe the diVerence isexplained by the lerp structure. The psyllids have somefree space within the lerp, and the large lerps have both agreater external diameter and internal free space, relativeto the nymph size. During our observations, the psyllidscould be seen through the more translucent lerps, mov-ing within the lerp to avoid the parasitoid’s probing ovi-positor; the parasitoid, in turn, would move from side toside on top of the lerp and repeat its drilling attempts indiVerent locations. The greater internal free space there-fore allows the psyllid to move to avoid ovipositionattempts. For these larger lerps, successful ovipositiontypically occurred when the ovipositor was quicklyslipped under the edge of the lerp, which is loosely Wxedto the leaf surface and provides small gaps, rather thanthrough the lerp (as was reported by Moore, 1961). Thissame oviposition location was reported for Metaphycusanneckei Guerrieri and Noyes to oviposit in black scale,Saissetia oleae (Olivier), and was suggested as a methodto reduce oviposition time (Barzman and Daane, 2001).

We found that P. bliteus had signiWcantly greater ovi-position success in second and third instar psyllids (Fig.1), while there was a clear preference in ovipositionattempts for larger lerps (Fig. 2). We suggest this discrep-ancy is explained by the increased handling times andlower oviposition success in these larger nymphs. There-fore, one function or realized beneWt of the large lerps isprotection from P. bliteus. Moore (1961), citing the occa-sionally high rates of parasitism of Glycaspis species inAustralia, argued that the main function of the lerp wasnot to protect the nymphs from natural enemies but toeither reduce desiccation or provide a fortuitous methodof disposing of waste. Ours is the Wrst indication that thelerps can also play a role in deterring at least one naturalenemy.

Two modes of host-utilization for adult nutrientswere observed. In the Wrst, the wasps simply lapped atthe surface of the lerp without disturbing the nymphunderneath. Because the lerp is composed of excretedcarbohydrates and proteins (Ernst and Sekhwela, 1987;Gilby et al., 1976) it may provide some of the nutritionalrequirements gained through host-feeding (Jervis andKidd, 1986); however, as the lerp composition is primar-ily sugars, we suggest that lerp utilization is primarily foradult longevity rather than for increased egg load or

K.M. Daane et al. / Biological Control 32 (2005) 228–235 233

development (Godfray, 1994). In the second mode, P.bliteus females host-fed on the psyllid after either slidingtheir ovipositors under or drilling through the lerps.Only 7 of 18 P. bliteus host fed during the total observa-tion period (18 h). In these cases, the parasitoid probedbrieXy with her ovipositor, presumably piercing thenymphs, and then turned around and imbibed Xuids that

Fig. 3. Psyllaephagus bliteus adult female and male longevity was anegative linear function of Wve tested constant temperatures (17, 21,23, 26, and 32 °C). Longevity was signiWcantly diVerent among testedtemperatures for females (F D 41.52, df D 4, 89, P < 0.001) and males(F D 23.70, df D 4,43, P < 0.001); within each gender, means followedby the same letter are not signiWcantly diVerent (Tukey’s HSD test,P < 0.05). Female longevity was signiWcantly greater than male longev-ity at each tested temperature (17 °C, t test D 3.306; P D 0.003; 21 °C, ttest D 5.106; P < 0.001; 23 °C, t test D 2.789; P D 0.009; 26 °C, ttest D 5.307; P < 0.001; 32 °C, t test D 3.774, P D 0.001).

Xowed out from under the edge of the lerp. With theexception of two, relatively short (8 and 13 s) events,host-feeding involved repeated bouts of drilling, probingand drinking over the course of several minutes(833.2 § 224.3 s), which was signiWcantly longer in dura-tion than oviposition time on either small or large lerps(F D 11.245, df D 2, 28, P < 0.001). In each case, the P. bli-teus host-fed only 1–2 times, with psyllid nymphs of allstages attacked in the following frequency: 1, 4, 3, 3, and1 for Wrst, second, third, fourth, and Wfth instar psyllids,respectively. While there were more Wrst and secondinstar psyllids used for host-feeding than expected basedon egg deposition, observed host-feeding events were rel-atively rare and we could not clearly determine if P. bli-teus has concurrent and/or nonconcurrent host-feedingmechanism(s) (Heimpel and Collier, 1996).

3.3. Adult longevity and fecundity

We tested adult longevity under a range of tempera-tures that could be used in insectary operations. From 17to 32 °C, female and male P. bliteus longevity were nega-tive linear functions of temperature (Fig. 3), rangingfrom 40.8 § 3.3 days (at 17 °C) to 14.2 § 0.9 days (at32 °C). At each temperature, females lived signiWcantlylonger than males (Fig. 3).

Excluding two individuals that were lost during theWrst week, lifetime egg deposition was 125.7 § 24.6 eggsper female (range 34–302). Most eggs (88.1%) weredeposited during the initial 22 days, although one para-sitoid lived for 90 days and deposited eggs up to 70 daysafter emergence (Fig. 4). The average longevity of adultfemales, provided host material, was 60.4 § 6.4 days(under described glasshouse conditions). During the ini-tial 22 days, there were 7.0 § 0.8 eggs per female per daydeposited (range 0–39). The longevity and observed

Fig. 4. Psyllaephagus bliteus lifetime fecundity, as estimated by egg deposition under glasshouse conditions with an overabundant host supply.

234 K.M. Daane et al. / Biological Control 32 (2005) 228–235

host-feeding behaviors of adult P. bliteus indicate syno-vigenic egg production; however, the fact that most eggsare deposited during the initial 22 days indicates that P.bliteus may have a signiWcant proportion of its eggsavailable soon after emergence. Under the laboratoryconditions imposed, parasitoid longevity may have beengreatly overextended compared to longevity under Weldconditions. A truly synovigenic species would continueto develop and mature eggs during the entire lifetime,which rarely occurred in our experiment (Fig. 4). AsdiVerences between pro-ovigenic and synovigenic speciesoften represent a continuum of life history strategies(Jervis et al., 2001), we suggest that while P. bliteus issynovigenic, it cannot produce and mature eggsthroughout the long lifespan realized under laboratoryconditions.

These results have implications for insectary opera-tions and release strategies in classical biological controlprograms. Although adults may survive for long periods,most egg deposition occurs early in the adult’s lifetime.Insectary colonies should therefore be supplied with theneeded number of third or fourth instar psyllids for anoviposition period of 2–3 weeks. Because they are usedfor host-feeding, Wrst or second instars should be pro-vided as well. Rearing temperatures should range from21 to 24 °C. We suggest that for release programs, adultsshould be collected and released within 3–7 days afteremergence in order to provide highly fecund individuals.

3.4. Immature development time

There was considerable variation in the immaturedevelopment time, regardless of the temperature treat-ment. We found that P. bliteus, like other Psyllaephagusspp. studied in detail (McDaniel and Moran, 1972; Patilet al., 1993), is a koinobiont and the oVspring will delaydevelopment until the psyllid reaches the proper stage orsize (Godfray, 1994). In this trial, eggs were found inthird, fourth, and young Wfth instar psyllids (the relativeage of Wfth instars is indicated by the degree of scleroti-zation). Larvae and pupae were only found in Wfth instarpsyllids, however, and the psyllids appeared outwardlyunaVected until the Wfth instar, when the parasitoid larvapresumably begins to develop and feed. Eggs depositedinto younger nymphs that had still not reached the Wfthinstar were still eggs at the end of the trial, but larvaldevelopment was rapid when eggs were deposited intofourth or Wfth instars. These results indicate that P. bli-teus is a koinobiont that delays larval development untilthe host reaches the late fourth or early Wfth instar.

In our trial, while examining the cleared specimens,we identiWed newly parasitized psyllids by the presenceof the encyrtiform egg, attached to the host-integumentvia an aeroscopic plate. We were able to clearly distin-guish second and third instar P. bliteus, but we wereunable to locate Wrst instars and we suspect this develop-

ment stage may have been destroyed during the clearingprocess with chloralphenol. Across all temperatures (22,26, and 30 °C) and psyllid development stages (secondthrough Wfth), average P. bliteus development time, fromegg to pupa, was 18.3 § 0.9 days. With no regard to theinitial psyllid stage attacked, P. bliteus development was22.6 § 1.9, 18.0 § 1.1, and 12.6 § 1.2 days at 22, 26, and30 °C, respectively. When P. bliteus oviposited initiallyinto Wfth instar psyllids, development (egg to pupa) wasonly 7, 6, and 8 days at 22, 26, and 30 °C, respectively (novariation).

In summary, this study has several implications formass rearing and release of P. bliteus. Third-instarnymphs are the most suitable for oviposition and aresuitable for host-feeding as well. The development timeof oVspring, however, depends on the psyllid stageattacked as well as on temperature, decreasing withincreasing temperature. For the adults, high rates of ovi-position can be sustained for up to about 22 days afteremergence, at moderate temperatures, and because adultlongevity decreases sharply with temperature, high tem-peratures may decrease lifetime fecundity. This last set ofobservations has important consequences for the evalua-tion of the biological control program in California. Todate, about four years after the initial mass releases, P.bliteus remains less eVective, with lower parasitism rates,in the interior of California compared to coastal regions,where it rapidly established and lowered psyllid popula-tions (Dahlsten, unpublished data; Sime et al., 2004).While the coastal regions tend to be mild year-round, theinland Sacramento and San Joaquin valleys are charac-terized by hot summers, with temperatures commonlyreaching 40 °C. This study suggests that such tempera-tures would dramatically decrease adult longevity, tomuch less than the 22 days during which ovipositionrates are high. Lowered lifetime fecundity, in turn, wouldbe expected to slow and reduce the success of P. bliteusas a control agent. To test this hypothesis we are cur-rently investigating the longevity and fecundity of P. bli-teus in the Weld, comparing coastal and inland sites. Itmay prove necessary to import additional Psyllaephagusspecies, or diVerent populations of P. bliteus, to improveeVective biological control throughout California.

Acknowledgments

The University of California Exotic Pests andDiseases Research Program funded this research; whilefunding for the red gum psyllid biological controlprogram was provided by the University of CaliforniaIntegrated Pest Management Project, California Depart-ment of Food and Agriculture (through help with insec-tary operations), Disney Corporation, Action MulchInc., East Bay Regional Parks, the Los Angeles Zoo, andagencies in the cities of Los Angeles, Torrance, Hunting-

K.M. Daane et al. / Biological Control 32 (2005) 228–235 235

ton Beach, and Redwood City. We thank Luke Powell,Dave Rowney, Daniel Sullivan, and Marta Yamamotofor laboratory and Weld assistance; William Roltsch andtwo anonymous reviewers helped improve earlierversions of the manuscript. John LaSalle (AustraliaNational Insect Collection) provided loan of paratypesand Mach Fukada (Hawaii Department of Agriculture)provided parasitoid specimens from Hawaii.

References

Barzman, M.S., Daane, K.M., 2001. Host-handling behaviours in par-asitoids of the black scale: a case for anti-mediated evolution. J.Anim. Ecol. 70, 237–247.

Brennan, E.B., Gill, R.J., Hrusa, G.F., Weinbaum, S.A., 1999. Firstrecord of Glycaspis brimblecombei (Moore) (Homoptera: Psyllidae)in North America: initial observations and predator associations ofa potentially serious new pest of eucalyptus in California. Pan-PaciWc Entomol. 75, 55–57.

Brennan, E.B., Hrusa, G.F., Weinbaum, S.A., Levison Jr., W., 2001.Resistance of Eucalyptus species to Glycaspis brimblecombei(Homoptera: Psyllidae) in the San Francisco Bay area. Pan-PaciWcEntomol. 77, 249–253.

Chauzat, M.P., Purvis, G., Dunne, R., 2002. Release and establishmentof a biological control agent, Psyllaephagus pilosus, for eucalyptuspsyllid (Cenarytaina eucalypti) in Ireland. Ann. Appl. Biol. 141,293–304.

Chermiti, B., Hawlitzky, N., Boulay, C., Onillon, J.C, 1986. Quelquescaracteristiques du developement de l’endoparasite Psyllaephaguseuphyllurae (Hymenoptere, Encyrtidae) et exploitation de son hoteEuphyllura olivina (Homoptere, Psyllidae). Entomophaga 31, 351–361.

Cockerham, S.T., 2004. Irrigation and planting density aVect river redgum growth. Calif. Agric. 58 (1), 40–43.

Dahlsten, D.L., Hansen, E.P., Zuparko, R.L., Norgard, R.B., 1998a.Biological control of the blue gum psyllid proves economically ben-eWcial. Calif. Agric. 52 (1), 35–40.

Dahlsten, D.L., Rowney, D.L., Copper, W.A., Tassan, R.L., Chaney,W.E., Robb, K.L., Tjosvold, S., Bianchi, M., Lane, P., 1998b. Para-sitoid wasp controls blue gum psyllid. Calif. Agricul. 52 (1), 31–34.

Ernst, W.H.O., Sekhwela, M.B.M., 1987. The chemical composition oflerps from the mopane psyllid Arytaina mopane (Homoptera, Psy-llidae). Insect Biochem. Mol. Biol. 17, 905–909.

Gilby, A.R., McKellar, J.W., Beaton, C.D., 1976. The structure of lerps:carbohydrate, lipid, and protein components. J. Insect Physiol. 22,689–696.

Godfray, H.C.J., 1994. Parasitoids: Behavioral and Evolutionary Ecol-ogy. Princeton University Press, Princeton, New Jersey.

Heimpel, G.E., Collier, T.R., 1996. The evolution of host-feedingbehaviour in insect parasitoids. Biol. Rev. 71, 373–400.

Helms, J.A., 1988. Introduction of the eucalypts to California, theircurrent status and future prospects. In: Proceedings of the Interna-tional Conference for the Australian Bicentenniary. vol. 3. TheAustralian Development Institute, pp. 17.

Jervis, M.A., Kidd, N.A.C., 1986. Host-feeding strategies in hymenop-teran parasitoids. Biol. Rev. 61, 395–434.

Jervis, M.A., Heimpel, G.E., Ferns, P.N., Harvey, J.A., Kidd, N.A.C.,2001. Life-history strategies in parasitoid wasps: a comparativeanalysis of ‘ovigeny’. J. Anim. Ecol. 70, 442–458.

McDaniel, J.R., Moran, V.C., 1972. The parasitoid complex of the cit-rus psylla Trioza erytreae (Del Guercio) (Homoptera: Psyllidae).Entomophaga 17, 297–317.

Moore, K.M., 1961. Observations on some Australian forest insects. 8.The biology and occurrence of Glycaspis baileyi Moore in NewSouth Wales. Proc. Linnean Soc. New South Wales 86, 185–200.

Moore, K.M., 1988. Associations of some Glycaspis species (Homop-tera: Spondyliaspididae) with their Eucalyptus species hosts. Proc.Linnean Soc. New South Wales 110, 19–24.

Noyes, J.S., Hanson, P., 1996. Encyrtidae (Hymenoptera: Chalcidoi-dea) of Costa Rica: the genera and species associated with jumpingplant-lice (Homoptera: Psylloidea). Bull. Nat. Hist. Museum, Lon-don (Entomol.) 65, 105–164.

Paine, T.D., Millar, J.G., 2002. Insect pests of eucalyptus in California:implications of managing invasive species. Bull. Entomol. Res. 92,147–151.

Paine, T.D., Dahlsten, D.L., Millar, J.G., Hoddle, M.S., Hanks, L.M.,2000. UC scientists apply IPM techniques to new eucalyptus pests.Calif. Agric. 54 (6), 8–13.

Patil, N.G., Baker, P.S., Pollard, G.V., 1993. Life histories of Psyllaeph-agus yaseeni (Hym., Encyrtidae) and Tamarixia leucaenae (Hym.,Eulophidae), parasitoids of the leucaena psyllid Heteropsyllacubana. Entomophaga 38, 565–577.

Riek, E.F., 1962. The Australian species of Psyllaephagus (Hymenop-tera, Encyrtidae), parasites of psyllids (Homoptera). Aust. J. Zool.10, 684–757.

Rosenheim, J.A., Hongkham, D., 1996. Clutch size in an obligately sibl-icidal parasitoid wasp. Anim. Behav. 51, 841–852.

Sime, K.R., Daane, K.M., Dahlsten, D.L., Andrews, J.W., Rowney,D.L. 2004. Constraints on the eVectiveness of Psyllaephagus bliteus(Hymenoptera: Encyrtidae), a biological control agent for red-gumlerp psyllid (Hemiptera: Psylloidea) in California. In: Proceedingsof the Fourth California Conference on Biological Control, Berke-ley, CA, pp 141–144.

Yen, A.L., 2002. Short-range endemism and Australian Psylloidea(Insecta: Hemiptera) in the genera Glycaspis and Acizzia (Psylli-dae). Invert. Syst. 16, 631–636.