Environmental Entomology, 47(6), 2018, 1451–1458 doi: 10.1093/ee/nvy145 Advance Access Publication Date: 25 September 2018 Biological Control - Parasitoids and Predators Research

Comparative Study of Egg Parasitism by Gryon pennsylvanicum (Hymenoptera: Scelionidae) on Two Squash Bug tristis and (: )

Mary L. Cornelius,1,3 Jing S. Hu,1 and Bryan T. Vinyard2 Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021

1Invasive Biocontrol and Behavior Laboratory, USDA Agricultural Research Service, Beltsville Agricultural Research Center, Beltsville, MD 20705, 2Statistics Group, USDA Agricultural Research Service, Northeast Area Office, Beltsville, MD 20705, and 3Corresponding author, e-mail: [email protected]

Subject Editor: Zsofia Szendrei

Received 11 May 2018; Editorial decision 30 August 2018

Abstract This study evaluated how the size of the egg mass and the parasitoids prior exposure to eggs influenced parasitism rates by Gryon pennsylvanicum Ashmead (Hymenoptera: Scelionidae) on egg masses of two squash bug species, DeGeer and Anasa armigera Say (Hemiptera: Coreidae). G. pennsylvanicum is the primary egg parasitoid of A. tristis. There were no published reports available on egg parasitism of A. armigera. In choice tests, there was no difference in host acceptance by G. pennsylvanicum of egg masses of the two squash bug species. In no-choice tests, overall parasitism rates were significantly higher onA. armigera egg masses than on A. tristis egg masses. Naive parasitoids had significantly higher parasitism rates than experienced parasitoids on egg masses of both squash bug species. In a comparison of parasitism rates of field-collected and laboratory-testedA. tristis egg masses of different sizes, parasitism rates were similar in the field and in the laboratory, with the exception of egg masses with > 25 eggs. Only 17.9% of eggs were parasitized in the laboratory, compared with 36.4% in the field. Results of this study indicate that transient egg limitation prevents G. pennsylvanicum from ovipositing in every available host egg in large squash bug egg masses. The low parasitism rate of G. pennsylvanicum on large egg masses may limit its effectiveness as a biological control agent of squash bugs.

Key words: oviposition behavior, egg load, egg limitation, biological control, natural enemies

The squash bug Anasa tristis DeGeer (Hemiptera: Coreidae) is Several studies have examined the development, and survival of a serious pest of cucurbit crops causing plant wilt and acting as a A. tristis on different cucurbit species. Bonjour and Fargo (1989) found vector of a bacterial infection, cucurbit yellow vine disease, which that survival to the adult stage was significantly higher on cultivars of can devastate cucurbit crops (Doughty et al. 2016). Squash bugs Cucurbita pepo L. (Cucurbitales: Cucurbitacaeae), cv. pepo (70.0%), overwinter as adults in crop residue or other debris and emerge in cv. melopepo (49.0%), than on watermelon, Citrullus lanatus (Thunb.) the spring. After emergence, adults search for cucurbit plants. Once Matsum. & Nakai (Cucurbitales: Cucurbitaceae) (14.4%), cucum- plants are located, they feed and mate. Females begin depositing ber, Cucumis sativus L. (Cucurbitales: Cucurbitaceae) (0.3%), and eggs on the host plant within 7–10 d of emergence (Nechols 1987). muskmelon, Cucumis melo L. (Cucurbitales: Cucurbitaceae) (0%). Squash bugs complete their entire lifecycle within 6–8 wk (Beard Other studies have demonstrated that Cucurbita moschata 1940). Depending on location, squash bugs can have one to three Dusch (Cucurbitales: Cucurbitaceae). is more resistant to A. tristis than generations per year. In Virginia, there are one to two generations C. pepo (Novero et al. 1962, Vogt and Nechols 1993a). Additionally, with new adults emerging in late July to early August and entering A. tristis nymphs were not able to develop successfully on cucumber diapause in late September to early October (Doughty et al. 2016). (Cook and Neal 1999) and were more likely to oviposit on C. pepo The horned squash bug, Anasa armigera Say (Hemiptera: Coreidae), than on cucumber (Cornelius 2018). In contrast, A. armigera was is considered to be a minor pest of cucurbit crops (Drake and Harris reared successfully on both cucumber and C. pepo and females were 1926, Britton 1937, Gould 1944). There are no published reports on equally likely to oviposit on either host plant, regardless of which spe- the phenology of A. armigera. cies they were reared on (Cornelius 2018).

Published by Oxford University Press on behalf of Entomological Society of America 2018. This work is written by (a) US Government employee(s) and is in the public domain in the US. 1451 1452 Environmental Entomology, 2018, Vol. 47, No. 6

There is a need to evaluate the effectiveness of biological con- on wild A. tristis egg masses collected in squash fields. Field-collected trol as a strategy for controlling pests in squash fields. The market A. tristis egg masses ranged in size from 1 to 44 eggs per egg mass with for organically grown produce is increasing (National Agricultural a mean (± SE) of 15.15 ± 0.69 eggs per egg mass (Cornelius et al. 2016). Statistics Service [NASS] 2010). In addition, squash bugs are difficult In parasitized egg masses with > 20 eggs, parasitoids never parasitized to control with insecticides (Nechols 1987, Olson et al. 1996). Egg 100% of the eggs in the egg mass, whereas they parasitized 100% of parasitoids can be effective biological control agents due to their the eggs in egg masses with ≤ 10 eggs 68% of the time. The inability of ability to prevent emergence of the pest, their relative host specificity, G. pennsylvanicum to successfully parasitize every egg in large squash and their high degree of host searching efficiency. bug egg masses limits its efficiency as a biological control agent. Gryon pennsylvanicum Ashmead (Hymenoptera: Scelionidae) is The lower rates of parasitism on larger egg masses could be the primary egg parasitoid of the squash bug, A. tristis (Olson related to transient egg limitation. Female parasitoids are expected et al. 1996, Cornelius et al. 2016, Doughty et al. 2016, Wilson and to maximize lifetime fecundity by balancing the risk of egg and time Kuhar 2017). It accounted for over 98% of parasitism in field sur- limitation (Rosenheim 1996, 1999; Rosenheim et al. 2008). For syn- veys conducted in Maryland and Virginia (Cornelius et al. 2016, ovigenic parasitoids, transient egg limitation can occur if the deple- Wilson and Kuhar 2017). Between June and September in 2013 tion of eggs exceeds the rate of egg maturation. If female parasitoids and 2014, G. pennsylvanicum emerged from 55.7% of squash bug deplete their supply of mature eggs before ovipositing in every host Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021 eggs in squash fields at the Beltsville Agricultural Research Center egg available, they must halt oviposition until they can replenish their in Maryland, with a peak emergence rate of 72.8% in the last week egg loads. Females in environments with a high host density may of July (Cornelius et al. 2016). In 2014 and 2015, over 50% of experience oviposition-mediated egg limitation (Rosenheim et al. squash bug eggs collected in Virginia were parasitized by G. penn- 2000). Egg limitation is mediated by egg maturation, oviposition, egg sylvanicum (Wilson and Kuhar 2017). These results indicate that resorption, and the factors influencing these processes Casas( et al. G. pennsylvanicum could be effective for suppressing squash bug 2009, Richard and Casas 2012). For instance, females of the soy- populations in commercial squash fields in Virginia and Maryland. bean aphid parasitoid Binodoxys communis (Gahan) (Hymenoptera: Currently, G. pennsylvanicum is being considered as a clas- Braconidae), were able to avoid egg limitation and maintain constant sical biological control agent of the invasive Western conifer seed egg loads throughout the day (Dieckhoff et al. 2014). bug Leptoglossus occidentalis Heidemann (Hemiptera: Coreidae) The objective of this study was to evaluate the oviposition in Europe (Sabbatini Peverieri et al. 2012, Peverieri et al. 2013, behavior of G. pennsylvanicum towards egg masses of different Roversi et al. 2014). When host egg age ranged from < 24 h old sizes to determine how many eggs a female will parasitize within to a day before hatching, parasitism rates on L. occidentalis eggs a 24-h period. Also, parasitism rates by naive parasitoids with no were not affected by age of the eggs (Peverieri et al. 2013). Because prior exposure to eggs were compared with parasitism rates by G. pennsylvanicum had high fecundity and long-lived adults, it was experienced parasitoids that had been previously exposed to eggs. considered to be a good candidate for classical biological control Parasitism rates with naive and experienced parasitoids were com- (Sabbatini Peverieri et al. 2012). pared because experienced parasitoids were likely to have a lower In a comparative study of G. pennsylvanicum with three species egg load due to their prior exposure to eggs. Because previous of Ooencyrtus, G. pennyslvanicum had the highest fecundity, net research determined that no oocyte resorption occurred for 3-d-old reproductive rate, and intrinsic rate of increase at 27°C of the four host-deprived G. pennsylvanicum (Vogt and Nechols 1993b), naive parasitoid species tested. Female G. pennsylvanicum had a mean (± 1–3-d-old females were expected to have their full egg loads. SD) fecundity of 80.7 ± 36 A. tristis eggs over a mean (± SD) ovi- In addition, this study evaluated the efficacy of parasitoids on positional period of 22.0 ± 6.7 d and their progeny had a percent eggs of A. armigera in both no-choice and choice tests. There are survivorship of 86% (Nechols et al. 1989). The daily number of no published reports concerning the natural enemies of A. armig- eggs laid by G. pennsylvanicum decreased with age. Peak oviposi- era. Therefore, it was not known how G. pennsylvanicum would tion occurred when females were 2–3 d old. Although there was a respond to A. armigera eggs or if the parasitoid would show any decline in parasitism rates as the host eggs aged, G. pennsylvanicum preference for A. tristis eggs over A. armigera eggs. was able to complete development in eggs that were 8 d old and This study further evaluated factors affecting the interaction contained fully developed unhatched nymphs (Nechols et al. 1989). between the egg parasitoid, G. pennsylvanicum and its host, the These results demonstrated that G. pennsylvanicum had the most squash bug, A. tristis. We examined whether the oviposition behav- potential to be an effective biological control agent of squash bugs ior of female squash bugs was influenced by the presence of the of the four parasitoid species tested. parasitoid. Female squash bugs deposit egg masses ranging in size Two studies examined the effects of different cucurbit cultivars from a single egg to > 40 eggs. It is possible that female squash bugs on the development of the host, A. tristis and the egg parasitoid, abort deposition of an egg mass when they detect the presence of G. pennsylvanicum (Vogt and Nechols 1993a,b). Negative effects the parasitoid, resulting in only a single egg or a small number of of resistant cultivars on the squash bug influenced the quality of the eggs deposited on a leaf. Experiments were conducted to evaluate the host egg for parasitoid development. There was a significant reduc- number of egg masses laid, the number of eggs per egg mass, and the tion in adult longevity of the parasitoid when its host was reared location of the egg mass in treatments where a mating pair of squash on resistant cultivars compared with susceptible cultivars of squash bugs was kept in a container with either no parasitoid, a single naive (Vogt and Nechols 1993a,b). In addition, fecundity was significantly parasitoid, or a single experienced parasitoid. lower in parasitoids that had been deprived of hosts for 3-d inter- vals than non-host-deprived parasitoids (Vogt and Nechols 1993b). Materials and Methods Therefore, availability and quality of host eggs could influence para- sitism rates in the field. Insect Rearing The efficacy ofG. pennsylvanicum as a biological control agent for Two species of squash bug, A. tristis and A. armigera, and the squash bugs depends on its ability to find and exploit host egg masses. egg parasitoid, G. pennsylvanicum, were collected in yellow sum- In a 2-yr field study in Maryland, there was a highly significant nega- mer squash, Cucurbita pepo L., fields at the Beltsville Agricultural tive correlation between number of eggs per mass and parasitism rates Research Center, Beltsville, MD (BARC) (39° 01′ 56.7′′ N, 76° Environmental Entomology, 2018, Vol. 47, No. 6 1453

55′ 54.5′′ W) and maintained in laboratory colonies in an incuba- Experienced parasitoids varied in age and egg load. There were at tor (25° C; 16L: 8D). Adult squash bugs of the two species were least 12 replicates each for both naive and experienced females for originally collected in the summer of 2012 and were supplemented each egg mass size category for both squash bug species for a total with nymphs that hatched from field-collected eggs every summer of 338 replicates. from 2013 to 2017. Squash bugs were kept in cylindrical plastic A choice test for egg masses of the two squash bug species was containers (19.3 cm height by 20.0 cm diameter) (Pioneer Plastics, conducted in 90 × 20 mm Petri dishes. Two egg masses of the same Inc., Eagan, MN). Each container had a hole (approximately 14 cm size (5 eggs/egg mass), one of A. tristis and one of A. armigera, were diameter) in the lid covered with a fine mesh screen, a filter paper placed in each Petri dish. A single female was introduced into the on the bottom, and two cotton plugged glass shell vials filled with Petri dish. There were a total of 149 replicates (99 with naive parasi- water attached with a rubber band to prevent them from rolling. toids and 50 with experienced parasitoids). The number of replicates Both species were maintained on seedlings of yellow squash C. pepo with egg masses where eggs of only one squash bug species or both cv. Slick Pik YS26 grown in a greenhouse and fruit of either USDA species were parasitized was recorded. certified organically grown yellow squash or zucchini purchased at a local market. Parasitism Rates on Field-Collected A. tristis The laboratory colony of G. pennsylvanicum was collected from

Egg Masses Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021 A. tristis eggs in the summer of 2015 and kept in rectangular boxes Effects of host egg mass size (number of eggs per egg mass) on field (19.0 cm length by 10 cm width) (Pioneer Plastics, Inc., Eagan, MN) parasitism rates were evaluated. In the summer of 2016, 530 A. tris- with four air holes (approximately 2.5 cm diameter), one on each tis egg masses were collected from four squash (C. pepo cv. Slick Pik side, covered with a fine mesh screen. Parasitoids were provided with YS26) fields with two (10 m by 10 m) plots per field at BARC (39° 10% honey solution and egg masses of both A. tristis and A. armig- 01′ 56.7′′ N, 76° 55′ 54.5′′ W). Because wild egg masses were col- era (≤ 72 h). Naive parasitoids were aspirated after emergence and lected from plots weekly, exposure to parasitoids ranged from < 24 h kept in small cylindrical plastic containers (7.1 cm height by 6.6 cm up to 7 d. Field-collected egg masses were also potentially exposed diameter) (Pioneer Plastics, Inc., Eagan, MN) with two air holes to multiple G. pennsylvanicum females. Of those 530 egg masses, (approximately 2.5 cm diameter) on the sides covered with a fine 181 were parasitized by G. pennsylvanicum. The 530 egg masses mesh screen and provided with 10% honey solution and no access were divided into the six categories of egg sizes described previously. to egg masses. Access to containers was provided by a 2.5 cm hole in the lid plugged with a rubber stopper. Effect of the Presence of a Female Parasitoid on the Oviposition Behavior of A. tristis Parasitism by G. pennsylvanicum on Squash Bug There were three treatments in this experiment: 1) control with no Egg Masses parasitoid; 2) a single naive parasitoid was introduced; 3) a single For all laboratory experiments, reproductively active female para- experienced parasitoid was introduced. In each replicate, a mating sitoids were collected from containers by placing a freshly laid egg pair of squash bugs was placed in a container as described previ- mass into the parasitoid laboratory colony and aspirating females as ously. Squash bugs were provided with seedlings of yellow squash they probed the egg mass. After the experiment was completed, egg (C. pepo cv. Slick Pik YS26). Each container had a folded paper masses were maintained in an incubator (25°C; 16L: 8D) until para- towel taped to one side of the container to provide an additional ovi- sitoid emergence. The number of emerged parasitoids was recorded. position site. In replicates with a parasitoid, a vial containing 10% Intact eggs were dissected and the number of eggs with unemerged honey solution was taped to the side of the container to provide food adult parasitoids was also recorded. Parasitism was calculated based for the parasitoid. on the total numbers of eggs with emerged and unemerged adult After 1 wk, the number of egg masses, the number of eggs per parasitoids. The parasitism rate was calculated by dividing the total egg mass, and the location of each egg mass was recorded. The loca- number of eggs parasitized by the total number of eggs per egg mass. tion of the egg masses was recorded to determine if the presence of For both no-choice and choice tests, fresh egg masses (≤24 h) the parasitoid influenced the oviposition site of female squash bugs. were exposed to a single parasitoid for 24 h. For each replicate, a Location was recorded as leaf, stem, or container (for any egg masses single female parasitoid was introduced into a Petri dish containing a deposited on the paper towel, the lid, or the sides of the container). filter paper moistened with water and an egg mass and their activity Nymph hatch rates were calculated based on numbers of eggs with was observed and recorded, including the time where they probed nymph hatch per total number of eggs per egg mass. There were 12 and parasitized eggs. replicates of each treatment.

Parasitism by G. pennsylvanicum on Laboratory- Statistical Analysis Reared A. tristis and A. armigera Egg Masses All parasitism and nymphal survival rates, with 95% confidence inter- No-choice tests were conducted in 60 × 15 mm Petri dishes to evalu- vals, were estimated using SAS PROC GLIMMIX to fit generalized ate the effects of the host species (A. tristis or A. armigera), egg linear mixed effects ANOVA models (Stroup 2012, 2015) with neg- mass size (number of eggs per egg mass), and the parasitoid’s prior ative binomial distribution, log link function, and offset of log total exposure to eggs (naive or experienced) on parasitism rates. The eggs in the egg mass. Factors used to construct the ANOVA models egg mass sizes were divided into six categories: 1–5, 6–10, 11–15, were: squash bug species (A. tristis, A. armigera), the parasitoid’s prior 16–20, 21–25, and >25 eggs per egg mass. Tests were conducted exposure to eggs (experienced, naive), and/or egg mass size (1–5, 6–10, using mated, reproductively active naive and experienced females 11–15, 16–20, 21–25, and >25 eggs per egg mass) (SAS 2012). to evaluate the effect of egg load on parasitism rates. Naive parasi- In choice tests with egg masses of A. tristis and A. armigera, the toids were 1–3 d old and had no prior exposure to squash bug eggs. number of replicates with egg masses where eggs of only one squash Experienced parasitoids were collected from the laboratory colony bug species were parasitized were compared using a chi-square test and had previous exposure to both A. tristis and A. armigera eggs. (SAS 2012). 1454 Environmental Entomology, 2018, Vol. 47, No. 6

Results In no-choice tests evaluating the effect of a parasitoid’s prior exposure to eggs, naive parasitoids had significantly higher parasit- Parasitism by G. pennsylvanicum on Laboratory- ism rates than experienced parasitoids for both squash bug species Reared A. tristis and A. armigera Egg Masses (A. tristis: F = 46.1; df = 1, 314; P < 0.0001; A. armigera: F = 12.2; In no-choice tests conducted within a 24-h period, there were signif- df = 1, 314; P = 0.0006). The parasitism rate for both experienced icant effects of the parasitoid’s prior exposure to eggs, squash bug and naive parasitoids was higher for A. armigera, than for A. tristis species, and the size of the egg mass on parasitism rates (Table 1). (Table 4). Overall, parasitism rates on egg masses of both squash bug species For A. tristis egg masses, there were significantly more eggs par- decreased gradually as the number of eggs per egg mass increased asitized by naive parasitoids than experienced parasitoids for egg with rates being significantly higher for egg masses with ≤ 15 eggs masses of all sizes, with the exception of the smallest (1–5 eggs) and per mass compared with egg masses with ≥ 21 eggs. Parasitism rates the largest (>25 eggs) egg masses (Table 4). For A. armigera, naive ranged from a mean (95% confidence interval) of 81.6% (69.5%, parasitoids only parasitized significantly more eggs per egg mass 95.8%) for egg masses with one to five eggs to 31.8% (28.6%, than experienced parasitoids for egg masses with 16–20 eggs and 35.4%) for egg masses with > 25 eggs (Table 2). >25 eggs per egg mass (Table 4). In no-choice tests, G. pennsylvanicum parasitized a significantly In a choice test where parasitoids were presented with egg Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021 higher number of A. armigera eggs than A. tristis eggs (Table 1). masses of each species, there was no difference in host acceptance There was no significant difference in parasitoid emergence on eggs of the two squash bug species. There were 48 replicates where the of the two squash bug species (F = 2.89, df = 1, 314, P = 0.09). parasitoid oviposited exclusively in A. tristis, 49 replicates where Parasitism rates were significantly higher onA. armigera egg masses the parasitoid oviposited exclusively in A. armigera, and 52 repli- with 11–15 eggs, 21–25 eggs, and >25 eggs per egg mass compared cates where the parasitoid oviposited in egg masses of both species. with A. tristis egg masses (Table 3). For egg masses with > 25 eggs, In a comparison of replicates where a female oviposited exclusively parasitism rates were more than twice as high on A. armigera than in an egg mass of one of the two species, parasitoids were equally on A. tristis. likely to choose egg masses of A. tristis or A. armigera (chi-square: 0.02; P = 1.0). Table 1. F statistics and P values from a generalized linear mixed effects ANOVA model with negative binomial distribution and off- Parasitism Rates on Field-Collected A. tristis set equal to log number of eggs in the egg mass, testing the effects of the parasitoid’s prior exposure to eggs (experienced or naive), Egg Masses the species of squash bug eggs (Anasa tristis or Anasa armigera), In 2016, 530 A. tristis egg masses were collected in squash fields. and the size of the egg masses (1–5, 6–10, 11–15, 20–25, >25 eggs The size of the egg masses ranged from 2 to 44 eggs per mass per egg mass) on the parasitism rate of egg masses by Gryon with a mean (± SE) of eggs per mass of 16.3 ± 0.32. Of the 530 pennsylvanicum in no-choice tests A. tristis egg masses, egg masses with 11–15 eggs per mass were Effect dfa F statistic P value the most common, accounting for 31.7% of egg masses collected, and egg masses with 1–5 eggs per mass were the least common, Exposure to Eggs 1 55.2 <0.0001* accounting for only 6.2% of egg masses (Fig. 1). The effect of Squash Bug Species 1 41.6 <0.0001* egg mass size on parasitism rates was evaluated for the 181 egg Egg Mass Size 5 54.2 <0.0001* masses that were parasitized by G. pennsylvanicum. Overall, Exposure to Eggs × Species 1 8.4 0.004* there was a significant effect of egg mass size on the parasitism Exposure to Eggs × Egg Mass Size 5 2.6 0.03* rates of field-collected egg masses versus egg masses of A. tristis Species × Egg Mass Size 5 9.7 <0.0001* exposed to parasitoids in laboratory bioassays (F =13.9; df = 5, Exposure to Eggs × Species × Egg 5 3,7 0.003* 333; P < 0.0001). However, parasitism rates on egg masses with Mass Size ≤ 25 eggs per egg mass were similar for field-collected and lab- oratory-tested egg masses. There was only a higher parasitism aDegrees of freedom for the numerator of the F statistic. The degrees of free- dom associated with the F statistic denominator is 314, for all model effects. rate on field-collected egg masses with > 25 eggs per egg mass *Indicates P < 0.05; generalized linear mixed effects model with negative than on egg masses exposed to parasitoids in laboratory tests binomial distribution and offset of log eggs in egg mass. (Table 5).

Table 2. Means, lower and upper 95% confidence intervals (LCL and UCL, respectively) of percent eggs parasitized for egg masses of dif- ferent sizes (1–5, 6–10, 11–15, 20–25, >25 eggs per egg mass) from both squash bug species combined in laboratory tests where a single female (naive and experienced combined) G. pennsylvanicum was exposed to an egg mass in a no-choice test for 24 h

Egg Mass Size # Replicates Number eggs per egg mass Mean % eggs parasitized

Mean (LCL, UCL)a Meanb (LCL, UCL)1

1–5 eggs 58 3.9 (3.4, 4.5) 81.6a (69.5, 95.8) 6–10 eggs 61 8.3 (7.6, 9.1) 69.4a (61.3, 78.7) 11–15 eggs 55 13.0 (12.1, 14.0) 66.5a,b (59.4, 74.5) 16–20 eggs 51 18.4 (17.2, 19.6) 52.5b,c (46.8, 58.9) 21–25 eggs 57 23.1 (21.9, 24.4) 43.0c (38.6, 47.9) >25 eggs 56 32.7 (31.2, 34.2) 31.8d (28.6, 35.4)

aMeans and confidence intervals estimated by a generalized linear mixed effects model with negative binomial distribution and offset of log eggs in egg mass. bMeans followed by different letters were significantly different (P < 0.05). Environmental Entomology, 2018, Vol. 47, No. 6 1455

Table 3. Comparison of means, lower and upper 95% confidence intervals (LCL and UCL, respectively) of percent eggs parasitized for egg masses of A. tristis versus A. armigera at each egg mass size category (1–5, 6–10, 11–15, 20–25, >25 eggs per egg mass) in laboratory tests where a single female (naive and experienced combined) G. pennsylvanicum was exposed to an egg mass in a no-choice test for 24 h

Squash bug species Egg mass size % Eggs parasitized F statistic P value*

Mean (LCL, UCL)a

A. tristis 1–5 eggs 79.9 (63.8, 100) A. armigera 1–5 eggs 82.8 (68.0, 100) 0.05 0.82 A. tristis 6–10 eggs 63.5 (54.3, 74.2) A. armigera 6–10 eggs 78.5 (67.2, 91.7) 3.60 0.06 A. tristis 11–15 eggs 58.3 (50.6, 67.0) A. armigera 11–15 eggs 75.0(65.9, 85.3) 6.78 0.0097* A. tristis 16–20 eggs 52.0 (45.3, 59.7) A. armigera 16–20 eggs 52.6 (46.2, 59.8) 0.01 0.91 A. tristis 21–25 eggs 34.2 (29.3, 39.8) A. armigera 21–25 eggs 48.2 (43.1, 54.0) 12.71 0.0004* Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021 A. tristis >25 eggs 17.8 (15.1, 21.0) A. armigera >25 eggs 43.5 (39.2, 48.3) 81.0 <0.0001*

aMeans and confidence intervals estimated by a generalized linear mixed effects model with negative binomial distribution and offset of log eggs in egg mass. *Mean percent eggs parasitized of A. tristis versus A. armigera egg masses compared for each egg size category were significantly different (P < 0.05).

Effect of the Presence of a Female Parasitoid on the with A. tristis, there were significantly more eggs parasitized by Oviposition Behavior of A. tristis naive parasitoids than experienced parasitoids for egg masses of all When a single female parasitoid was introduced into a container sizes, with the exception of the smallest (1–5 eggs) and the largest with a mating pair of squash bugs, the parasitoid’s exposure to eggs (>25 eggs) egg masses. In tests with A. armigera, naive parasitoids influenced parasitism rate F( = 5.1; df = 1, 85; P = 0.027) and nymph only parasitized significantly more eggs for egg masses with 16–20 hatch rate (F = 17.9; df = 2, 125; P < 0.0001). There was a signif- eggs and > 25 eggs. Behavioral bioassays have recently determined icantly higher parasitism rate between the treatment where a naive that G. pennsylvanicum spends an equal amount of time probing female parasitoid was introduced compared with the treatment with eggs of the two squash bug species, but that it takes females signif- an experienced female parasitoid. Naive parasitoids were able to icantly longer to oviposit in A. tristis eggs than in A. armigera eggs parasitize enough eggs to significantly reduce nymph hatch, but there (Cornelius, unpublished data). It is possible that the difference in the was no difference in nymph hatch rate between the control treat- time that it takes to complete oviposition in eggs of the two squash ment without a parasitoid and the treatment with an experienced bug species could contribute to the differences in overall parasitism parasitoid (Table 6). rates in no-choice tests. The number of egg masses laid by female squash bugs was not We examined how the presence of the parasitoid influenced the affected by treatment (F = 0.69; df = 2, 125; P = 0.51). However, oviposition behavior of female squash bugs. When a single female there was a marginally significant difference in the numbers of eggs parasitoid was introduced into a container with a mating pair of laid per egg mass by treatment (F = 3.1; df = 2, 125; P = 0.049). squash bugs, there was no evidence that female squash bugs depos- Squash bug females laid fewer eggs per mass in the treatment with a ited fewer egg masses. Female squash bugs did deposit slightly fewer naive parasitoid compared with the other two treatments. The loca- eggs per egg mass in the presence of a naive parasitoid, but not an tion of the egg masses (leaf, stem, container) had no effect on para- experienced parasitoid. Only the naive parasitoid exerted a signifi- sitism rate (F = 0.27; df = 2, 84; P = 0.76). cant effect on nymph hatch rates compared with the control treat- ment with no parasitoid. The location of the egg masses (leaf, stem, container) did not influence parasitism rates. Therefore, there was Discussion no advantage for female squash bugs to deposit egg masses in other We measured the parasitism rates of G. pennyslvanicum on eggs of locations in the container. Because the experimental design did not two squash bug species over a 24-h period. This is the first published allow female squash bugs an avenue to escape from parasitism pres- report of egg parasitism on A. armigera. In choice tests, G. pennsyl- sure, these results may not reflect the response of female squash bugs vanicum did not discriminate between egg masses of the two host to G. pennsylvanicum in the field. species in host acceptance. However, in no-choice tests, parasitism The primary objective of this study was to evaluate the parasitism rates on A. armigera eggs were significantly higher than onA. tristis rate of G. pennsylvanicum on squash bug egg masses of different sizes eggs overall. The percent of wasps emerging successfully from para- and measure the maximum rate of oviposition on large egg masses. sitized eggs was slightly higher on A. armigera eggs than on A. tristis As we expected, naive parasitoids had a higher parasitism rate than eggs, but the difference was not significant. In addition, G. pennsyl- experienced parasitoids. Experienced parasitoids would be expected vanicum consistently parasitized higher numbers of A. armigera eggs to have lower egg loads because they had been previously exposed to than A. tristis eggs in no-choice tests evaluating both egg mass size squash bug egg masses, whereas naive parasitoids had not been able to and the parasitoid’s prior exposure to eggs. The parasitism rate on oviposit prior to the experiment. Previous research demonstrated that egg masses with >25 eggs was more than twice as high for A. armig- G. pennsylvanicum deposits the most eggs on the first day of exposure era than for A. tristis. to egg masses. In a study of parasitism rates by G. pennyslvanicum There were also differences in how the parasitoid’s egg load on L. occidentalis, the average number of eggs parasitized per female influenced parasitism rates of the two squash bug species. In tests per day was 12.73 during the first 5 d of oviposition and the highest 1456 Environmental Entomology, 2018, Vol. 47, No. 6

Table 4. Comparison of means, lower and upper 95% confidence intervals (LCL and UCL, respectively) of percent eggs parasitized by naive versus experienced female G. pennsylvanicum for egg masses of A. tristis and A. armigera at each egg mass size category (1–5, 6–10, 11–15, 20–25, >25 eggs per egg mass) in laboratory tests where a single female G. pennsylvanicum was exposed to an egg mass in a no-choice test for 24 h

Prior exposure to eggs Egg mass size % Eggs parzasitized F statistic P value*

Mean (LCL, UCL)a

A. tristis egg masses Naive 1–5 eggs 91.5 (67.3, 100) Experienced 1–5 eggs 69.8 (50.3, 97.0) 1.4 0.24 Naive 6–10 eggs 81.9 (64.9, 100) Experienced 6–10 eggs 49.2 (39.9, 60.6) 10.3 0.002* Naive 11–15 eggs 74.7 (62.0, 90.0) Experienced 11–15 eggs 45.4 (36.8, 56.0) 12.2 0.0006* Naive 16–20 eggs 65.5 (55.0, 78.0) Experienced 16–20 eggs 41.2 (33.3, 51.1) 10.9 0.001* Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021 Naive 21–25 eggs 58.0 (49.4, 68.0) Experienced 21–25 eggs 20.2 (15.5, 26.2) 46.3 <0.0001* Naive >25 eggs 17.8 (14.1, 22.6) Experienced >25 eggs 17.8 (14.2, 22.4) 0.0 0.99 A. armigera egg masses Naive 1–5 eggs 92.3 (71.3, 100) Experienced 1–5 eggs 74.2 (55.2, 99.7) 1.2 0.28 Naive 6–10 eggs 86.2 (70.3, 100) Experienced 6–10 eggs 7 1.5 (56.5, 90.4) 1.4 0.24 Naive 11–15 eggs 81.3 (67.7, 97.6) Experienced 11–15 eggs 69.2 (57.6, 83.0) 1.5 0.22 Naive 16–20 eggs 0.7 (50.7, 72.8) Experienced 16–20 eggs 45.5 (37.8, 54.7) 4.87 0.03* Naive 21–25 eggs 51.7 (44.2, 60.5) Experienced 21–25 eggs 44.9 (38.2, 52.8) 1.5 0.22 Naive >25 eggs 48.8 (42.7, 55.8) Experienced >25 eggs 38.8 (33.0, 45.6) 4.7 0.03*

aMeans and confidence intervals estimated by a generalized linear mixed effects model with negative binomial distribution and offset of log eggs in egg mass. *Mean percent eggs parasitized by naive versus experienced parasitoids compared for each egg size category for each squash bug species were significantly different (P < 0.05).

We compared our results from laboratory tests where a single parasitoid was exposed to an egg mass for 24 h with the parasitism rate of A. tristis egg masses that were collected weekly from squash fields. The field-collected egg masses may have been exposed to par- asitoids for as long as 7 d with possibly more than one parasitoid ovipositing in eggs from the same egg mass. Despite the differences in parasitoid exposure of laboratory and field-collected egg masses, parasitism rates in the field and in laboratory tests were similar for every egg mass size, with the exception of egg masses with > 25 eggs. Field parasitism rates (36.4%) on A. tristis egg masses with > 25 eggs exceeded the laboratory parasitism rate (17.9%), suggesting that multiple parasitoids may have visited the same egg mass or that a female returned to the same egg mass after replenishing her supply of mature eggs. However, an average parasitism rate of field-col- lected egg masses with > 25 eggs of only 36.4% demonstrates that G. pennsylvanicum frequently fails to parasitize every egg in larger egg masses in the field. Fig. 1. The distribution of 530 A. tristis egg masses collected in the field in In 2016, field-collectedA. tristis egg masses ranged in size from 2016 in the different egg mass size categories. 2–44 eggs per egg mass. The wide range in numbers of eggs per egg mass deposited by A. tristis could possibly reflect conflicting selec- mean parasitism rate was recorded on the first day of exposure to egg tive pressures from natural enemies. For squash bugs, depositing egg masses with 21.09 eggs parasitized per female (range of 8–34 eggs) masses with >25 eggs would most likely decrease the probability of (Sabbatini Peverieri et al. 2012). Naive females parasitized less than egg mortality due to parasitism by G. pennsylvanicum. Although half of the eggs in egg masses with > 25 eggs. Therefore, G. pennsyl- selective pressure from the primary egg parasitoid, G. pennsylvani- vanicum females seem to have transient egg limitations which prevent cum, would seem to favor egg masses with > 25 eggs per egg mass, them from fully utilizing an entire large egg mass within a 24-h period. the majority of egg masses deposited by A. tristis had ≤ 20 eggs per Environmental Entomology, 2018, Vol. 47, No. 6 1457

Table 5. Comparison of means, lower and upper 95% confidence intervals (LCL and UCL, respectively) of percent field-collected versus laboratory-tested A. tristis eggs parasitized at each egg mass size category (1–5, 6–10, 11–15, 20–25, >25 eggs per egg mass). Field para- sitism rate by G. pennsylvanicum of wild egg masses collected in the field in the summer of 2016 was compared to parasitism rate of egg masses in laboratory tests where a single female (naive and experienced combined) G. pennsylvanicum was exposed to an egg mass in a no-choice test for 24 h

Egg mass Egg mass size % Eggs parasitized F statistic P value*

Mean (LCL, UCL)a

Lab-tested 1–5 eggs 80.3 (59.0, 100) Field-collected 1–5 eggs 71.0 (46.1, 100) 0.21 0.65 Lab-tested 6–10 eggs 59.7 (47.1, 75.6) Field-collected 6–10 eggs 62.7 (44.8, 87.8) 0.06 0.81 Lab-tested 11–15 eggs 59.0 (46.5, 74.8) Field-collected 11–15 eggs 50.0 (42.2, 59.3) 1.23 0.27 Lab-tested 16–20 eggs 53.5 (41.7, 68.8) Field-collected 16–20 eggs 53.0 (43.3, 64.9) 0.0 0.95 Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021 Lab-tested 21–25 eggs 39.3 (30.8, 50.2) Field-collected 21–25 eggs 42.9 (34.5, 53.3) 0.27 0.61 Lab-tested >25 eggs 17.9 (13.8, 23.3) Field-collected >25 eggs 36.4 (28.1, 47.2) 14.38 0.0002*

aMeans and confidence intervals estimated by a generalized linear mixed effects model with negative binomial distribution and offset of log eggs in egg mass. *Mean percent eggs parasitized from A. tristis field-collected versus laboratory-tested egg masses compared for each egg size category were significantly different (P < 0.05).

Table 6. Comparison of means, lower and upper 95% confidence the day (Dieckhoff et al. 2014). Additionally, analyses of their sugar intervals (LCL and UCL, respectively) of percent nymph hatch and profiles showed that females were feeding on sugar in the field at a eggs parasitized from three treatments: control (no parasitoid), an very high rate. The ability of B. communis females to compensate experienced parasitoid, or a naive parasitoid was introduced into a for oviposition by maintaining a high egg maturation rate may be container with a mating pair of A. tristis for a week related to their high consumption of sugar (Dieckhoff et al. 2014). Treatment % Nymph hatch % Eggs parasitized In the current study, we did not examine the nutritional status of G. pennsylvanicum females. In laboratory bioassays, females had 2 1 2 1 Mean (LCL, UCL) Mean (LCL, UCL) access to honey water prior to bioassays, but not during the 24 h bioassay. The nutritional status of G. pennsylvanicum females in the Control 93.7a (69.1, 100) - field is unknown. However, there is evidence thatG. pennsylvanicum Experienced 63.4a (46.2, 87.0) 27.2a (16.3, 45.6) Naive 25.5b (18.6, 35.0) 58.7b (37.8, 91.1) feeds on leaf trichome exudates of squash plants and that the exu- dates may provide an energy source for the parasitoid (Olson and Nechols 1995). It is possible that increasing the availability of sugar 1Means and confidence intervals estimated by a generalized linear mixed effects model with negative binomial distribution and offset of log eggs in sources for female G. pennsylvanicum could result in higher rates of egg mass. egg maturation. Increasing the rate of egg maturation would enable 2Means within a column followed by different letters were significantly dif- female parasitoids to better compensate for the depletion of eggs and ferent (P < 0.05). increase the percentage of eggs parasitized per egg mass. Egg limitation reduces reproductive opportunities for the parasi- egg mass. Predators of A. tristis eggs include Araneae, Cricetidae, toid. In a monoculture of squash, there is a high density of host egg Formicidae, and Gryllidae (Phillips and Gardiner 2016). Egg preda- masses. The reproductive strategy of G. pennsylvanicum does not tors, such as ants, often consume an entire egg mass. There may be appear to be well adapted to fully utilizing available host eggs for conflicting selective pressure to spread out the risk of egg mortality reproduction in conditions of high host density. Egg limitation also from predators by depositing smaller egg masses in multiple loca- reduces the parasitoid’s ability to suppress squash bug populations. tions. Therefore, female squash bugs deposit egg masses that range If factors such as nutritional status could increase rates of egg matu- widely in terms of numbers of eggs per egg mass in the field. ration in the field, the efficiency of G. pennsylvanicum as a biological When hosts are abundant, a synovigenic female parasitoid’s control agent could possibly be improved. egg load can be temporarily depleted, forcing the parasitoid to wait until additional eggs mature before resuming oviposition Acknowledgments (Rosenheim et al. 2000). Over a 22-d oviposition period, G. penn- sylvanicum was able to parasitize an average of 80.7 ± 36 A. tristis We would like to thank Madeline Topf for technical assistance with field col- eggs (Nechols et al. 1989). However, transient egg limitation pre- lections of egg masses and laboratory colony maintenance. We would also like vents G. pennsylvanicum from attacking every egg in large squash to thank Don Weber for revisions on an earlier draft of the manuscript. bug egg masses in both the field and the laboratory. The low para- sitism rate of G. pennsylvanicum on large host egg masses may References Cited limit its effectiveness as a biological control agent of squash bugs. Beard, R. L. 1940. The biology of Anasa tristis De Geer with particular refer- In a field study,B. communis females were able to mature eggs at ence to the tachinid parasite, Trichopoda pennipes Fabr. Conn. Agr. Exp. a rate that enabled them to maintain a constant egg load throughout Stn. Bull. 440: 597–679. 1458 Environmental Entomology, 2018, Vol. 47, No. 6

Bonjour, E. L., and W. S. Fargo. 1989. Host effects on the survival and devel- tactics versus chemical control for squash bugs (Heteroptera: Coreidae) opment of Anasa tristis (Heteroptera: Coreidae). Environ. Entomol. 18: management on pumpkins. J. Econ. Entomol. 89: 631–639. 1083–1085. Peverieri, G. S., P. Furlan, D. Benassai, S. Caradonna, W. B. Strong, and P. Britton, W. E. 1937. Connecticut State Entomologist, Thirty-sixth Report 1936. F. Roversi. 2013. Host egg age of Leptoglossus occidentalis (Heteroptera, Bull. Conn. Agric. Exp. Stn. 36: 289–415. Peiper Press, Wallington, CT. Coreidae) and parasitism by Gryon pennsylvanicum (Hymenoptera, Casas, J. F. Vannier, N. Mandon, J. P. Delbecque, D. Giron, and J. P. Monge. Platygastridae). J. Econ. Entomol. 106: 633–640 2009. Mitigation of egg limitation in parasitoids: immediate hormonal Phillips, B. W., and M. M. Gardiner. 2016. Does local habitat management or response and enhanced oogenesis after host use. Ecology. 90: 537–545. large-scale landscape composition alter the biocontrol services provided to Cook, C. A., and J. J. Neal. 1999. Feeding behavior of larvae of Anasa tristis pumpkin agroecosystems? Biol. Control. 92: 181–194. (Heteroptera: Coreidae) on pumpkin and cucumber. Environ. Entomol. Richard, R., and J. Casas. 2012. A quantitative framework for ovarian dynam- 28: 173–177. ics. Funct. Ecol. 26:1399–1408. Cornelius, M. L. 2018. Ovipositional preferences of two squash bug species, Rosenheim, J. A. 1996. An evolutionary argument for egg limitation. Anasa tristis and Anasa armigera (Heteroptera: Coreidae), for different Evolution. 50: 2089–2094. cultivars and species of Cucurbitaceae. J. Insect Sci. 18: 30. Rosenheim, J. A. 1999. The relative contributions of time and eggs to the cost Cornelius, M. L., M. L. Buffington, E. J. Talamas, and M. W. Gates. 2016. of reproduction. Evolution. 53: 376–385. Impact of the egg parasitoid, Gryon pennsylvanicum (Hymenoptera: Rosenheim, J. A., G. E. Heimpel, and M. Mangel. 2000. Egg maturation, egg

Scelionidae), on sentinel and wild egg masses of the squash bug (Hemiptera: resorption and the costliness of transient egg limitation in . Proc. Downloaded from https://academic.oup.com/ee/article/47/6/1451/5106965 by guest on 29 September 2021 Coreidae) in Maryland. Environ. Entomol. 45: 367–375. Biol. Sci. 267: 1565–1573. Dieckhoff, C., J. C. Theobald, F. L. Wäckers, and G. E. Heimpel. 2014. Egg Rosenheim, J. A., S. J. Jepsen, C. E. Matthews, D. S. Smith, and M. load dynamics and the risk of egg and time limitation experienced by an R. Rosenheim. 2008. Time limitation, egg limitation, the cost of ovipo- aphid parasitoid in the field. Ecol. Evol. 4: 1739–1750. sition, and lifetime reproduction by an insect in nature. Am. Nat. 172: Doughty, H. B., M. Wilson, P. B. Schultz, and T. P. Kuhar. 2016. Squash bug 486–496. (Heteroptera:Coreidae): biology and management in cucurbitaceous Roversi, P. F., G. Sabbatini Peverieri, M. Maltese, P. Furlan, W. B. Strong, and crops. J. Integr. Pest Manag. 7: 1–8. V. Caleca. 2014. Pre-release risk assessment of the egg-parasitoid Gryon Drake, C. J., and H. M. Harris. 1926. Insect enemies of melons and cucumbers pennsylvanicum for classical biological control of Leptoglossus occidenta- in Iowa. Bull. Iowa Agric. Exp. Stn. 90: 1–12. lis. J. Appl. Entomol. 138: 27–35. Gould, G. E. 1944. Insect pests of cucurbit crops in Indiana. Proc Indiana Acad Sabbatini Peverieri G., P. Furlan, S. Simoni, W. B. Strong, and P. F. Roversi. Sci. 53: 165–171. 2012. Laboratory evaluation of Gryon pennsylvanicum (Ashmead) National Agricultural Statistics Service (NASS). 2010. Organic production sur- (Hymenoptera, Platygastridae) as a biological control agent of vey (2008), 2007 census of agriculture, volume 3, special studies, part 2 Leptoglossus occidentalis Heidemann (Heteroptera, Coreidae). Biol. (AC-07-SS-2). USDA-NASS, Washington, DC. Control 61:104–111. Nechols, J. R. 1987. Voltinisms, seasonal reproduction, and diapauses in the SAS® version 9.4. 2012. SAS/STAT PROC GLIMMIX©, SAS Institute, Inc., squash bug (Heteroptera: Coreidae) in Kansas. Environ. Entomol. 16: Cary, NC, USA. 269–273. Stroup, W. W. 2012. Generalized linear mixed models: modern concepts, meth- Nechols, J. R., J. L. Tracy, and E. A. Vogt. 1989. Comparative ecological stud- ods and applications. CRC Press, Boca Raton, FL. pp. 555. ies of indigenous egg parasitoids (Hymenoptera: Scelionidae; Encyrtidae) Stroup, W. W. 2015. Rethinking the analysis of non-normal data in plant and of the squash bug Anasa tristis (Hemiptera: Coreidae). J. Kans. Entomol. soil science. Agronomy Journal. 107: 811–827. Soc. 62: 177–188. Vogt, E. A., and J. R. Nechols. 1993a. Responses of the squash bug (Hemiptera: Novero, E. S., R. H. Painter, and C. V. Hall. 1962. Interactions of the squash Coreidae) and its egg parasitoid, Gryon pennsylvanicum (Hymenoptera: bug, Anasa tristis, and six varieties of squash (Cucurbita spp.). J. Econ. Scelionidae) to three Curcurbita cultivars. Environ. Entomol. 22: 238–245. Entomol. 55: 912–919. Vogt, E. A., and J. R. Nechols. 1993b. The influence of host deprivation and Olson, D. L., and J. R. Nechols. 1995. Effects of squash leaf trichome exu- host source on the reproductive biology and longevity of the squash dates and honey on adult feeding, survival, and fecundity of the squash bug egg parasitoid, Gryon pennsylvanicum (Ashmead) (Hymenoptera: bug (Heteroptera: Coreidae) egg parasitoid Gryon pennsylvanicum Scelionidae). Biol. Control 3: 148–154. (Hymenoptera: Scelionidae). Environ. Entomol. 24: 454–413, 458. Wilson, J. M., and T. P. Kuhar. 2017. A survey of the species of squash bug Olson, D. L., J. R. Nechols, and B. W. Schurle. 1996. Comparative evaluation (Hemiptera: Coreidae) egg parasitoids in Virginia and their distribution. J. of population effect and economic potential of biological suppression Econ. Entomol. 110: 2727–2730.