Influences of impatiens pollen and exposure to Beauveria bassiana on bionomics of western flower thrips Frankliniella occidentalis

Biological Control 37 (2006) 186–195 www.elsevier.com/locate/ybcon

InXuences of impatiens pollen and exposure to Beauveria bassiana on bionomics of western Xower thrips Frankliniella occidentalis Todd A. Ugine a,¤, Stephen P. Wraight b, John P. Sanderson a a

b

Department of Entomology, Cornell University, Ithaca, NY 14853, USA USDA-ARS, US Plant, Soil, and Nutrition Laboratory, Ithaca, NY 14853, USA Received 8 July 2005; accepted 22 September 2005 Available online 4 November 2005

Abstract Adult female western Xower thrips, Frankliniella occidentalis, were exposed for 24 h to impatiens leaf disks treated with Beauveria bassiana at low and high application rates (ca. 100 and 1000 viable conidia/mm2) and subsequently maintained on impatiens leaf disks supplemented or not supplemented with impatiens pollen. OVspring production and mortality of insects were monitored daily. Exposure to B. bassiana at the low and high rates signiWcantly reduced thrips longevity by 3.9 and 4.0 days, reduced the ovipositional period by 3.4 and 3.0 days, and reduced lifetime fecundity by 22 and 46% at the low and high rates, respectively. Infection by B. bassiana resulted in a sublethal (pre-mortem) eVect on oVspring production, decreasing oVspring production on the day before death by 1.2 oVspring/female. Pollen supplements had no eVect on adult female thrips longevity, yet did signiWcantly increase both daily (2.3 and 3.8 times) and lifetime fecundity (2.1 and 3.6 times) compared to control insects in tests at the low and high rates, respectively. No signiWcant Beauveria £ pollen interactions were detected at the low rate; however, a marginally signiWcant pollen £ B. bassiana interaction was present in tests of both daily and lifetime fecundity. There was a signiWcant eVect of B. bassiana on lifetime oVspring production in the presence of pollen, but the eVect was not detectable in the no pollen treatment. The increase in both daily and lifetime oVspring production in the presence of pollen and the slow action of B. bassiana suggest that if B. bassiana is to be used successfully as a thrips management tool in impatiens crops, it must be applied before pollen becomes present and targeted against thrips immature stages to kill the insects before they reach reproductive maturity. © 2005 Elsevier Inc. All rights reserved. Keywords: Western Xower thrips; Frankliniella occidentalis; Bionomics; Pollen; Fecundity; Entomopathogenic fungus; Beauveria bassiana

1. Introduction Western Xower thrips (WFT), Frankliniella occidentalis (Pergande), is a key pest of greenhouse vegetable and ornamental crops around the world. Direct damage is caused when thrips feed on Xower buds, leaves, and fruit. Indirect damage is caused by the transmission of plant tospoviruses, vectored by thrips (Daughtery et al., 1997). Thrips damage to ornamental and food crops can be unsightly and lower the aesthetic appeal, and thus value of the crop. To combat thrips, growers typically use chemical insecticides (both synthetic and biorational). Application of certain chemical *

Corresponding author. E-mail address: [email protected] (T.A. Ugine).

1049-9644/$ - see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.biocontrol.2005.09.020

insecticides can quickly achieve successful thrips management (often within 1 day), compared to biological control agents whose Wrst eVects are often not detectable for several days. The low cost of traditional insecticides and their ease of use make these management options attractive. However, WFT has demonstrated a strong capacity to resist chemical insecticides (Herron et al., 1996; Immaraju et al., 1992; Zhao et al., 1994), and research on alternative management options is being pursued. Among the alternatives under investigation are insect pathogens. Thrips, which feed on plant cell contents using piercing/sucking mouthparts, are not highly susceptible to pathogens that invade per os. Thus, fungi, which infect by penetrating directly through the insect integument, are the most promising microbial biocontrol agents for these insects.

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

Entomopathogenic fungi are comparatively slow-acting biological control agents, and their successful use can depend, in part, on the insects’ reproductive capacity during the disease incubation period. The use of biological control agents against insects that have a high rate of population replacement is aVorded the best chances of success if applied prophylactically or before pest populations enter the rapid growth phase. Gerin et al. (1999) established that populations of the WFT in crops of garden impatiens, Impatiens wallerana (Hook), grow at a signiWcantly faster rate when Xowers are present. Previous research by Ugine (2004) demonstrated that female thrips strongly prefer impatiens Xowers that contain pollen. It is known that many species of thrips derive nutrients from plant pollens and that pollens can signiWcantly increase thrips fecundity and longevity (Hulshof et al., 2003; Kirk, 1985; Lesky et al., 1997; Teulon and Penman, 1991; Trichillo and Leigh, 1988). Numerous researchers have investigated the eVects of fungal infection on insect bionomics (Blanford and Thomas, 2001; Castillo et al., 2000; Ekesi and Maniania, 2000; Fargues et al., 1991; Hsiao et al., 1992; Noma and Strickler, 2000). Findings in these studies included signiWcant reductions in lifetime fecundity, eggs per clutch, and egg fertility. On the other hand, other investigators have reported no signiWcant pre-mortem eVects of fungal infection on fecundity of insect hosts (Lacey et al., 1997; Lord et al., 1987; Nielsen et al., 2005; Wang and Knudsen, 1993). To fully assess the potential of slow-acting biological control agents such as Beauveria bassiana (Balsamo) Vuillemin for control of WFT, it is necessary to determine the impact of B. bassiana infection on thrips longevity and fecundity on crop plants with and without pollen. In this study, a series of factorial experiments were conducted to investigate B. bassiana £ pollen £ WFT interactions. 2. Materials and methods 2.1. General methods 2.1.1. Insects Western Xower thrips adults collected from a Cornell University research greenhouse were used to establish a laboratory colony. Thrips were continuously reared on excised leaves of red kidney beans (Phaseolus vulgaris L.) in small plastic containers (16 cm £ 16 cm £ 6 cm) with snapon lids. A hole (10.5 £ 10.5 cm) was cut in the lid and covered with thrips-proof organdy (mesh openings of 95 m) (Sefar America, Kansas City, MO), to allow for ventilation. Containers were maintained at 28 § 1 °C under a 14:10 h light:dark regime. Adult female western Xower thrips (25–30) were added to a rearing container and allowed to oviposit for 24 h. Eleven days after initial infestation of a rearing container, the thrips population comprised primarily newly eclosed adult thrips (624 h old). From this population, pairs of adult male and female thrips were collected via aspiration

187

into modiWed 1-ml centrifuge tubes (bottoms removed and covered with screening). The vials, each containing a pair of adults, were then randomly assigned to the treatments associated with an assay. For additional details of the rearing/ selection processes, see Ugine et al. (2005a). 2.1.2. Plant material Garden impatiens plugs, I. wallerana var. “SuperelWn; white,” were purchased (Mid Atlantic Plant, Newark, DE) and were potted with Pro-Mix BX (Premier Horticulture, Quakertown, PA) into 10.2 cm pots. Plants were grown in a greenhouse without supplemental lighting at 24 § 3 °C for 1 month, which was suYcient for the plants to Xower and produce pollen for use in treatments. Plants were fertilized weekly with Excel 21-5-20 fertilizer (Scotts-Sierra Horticultural Products Company, Marysville, OH). Impatiens leaves and pollen-bearing stamens from newly opened Xowers were excised daily from randomly selected plants for use in the bioassays. 2.1.3. Fungus preparations and application Beauveria bassiana strain GHA, formulated as a claybased wettable powder (BotaniGard 22WP) was produced by Emerald BioAgriculture (Butte, MT) using proprietary solid-substrate culture methods and ingredients (Bradley et al., 1992; Wraight et al., 1999). The fungal product (lot #020813), containing 4.4 £ 1010 conidia/g, was stored at 4 °C before use. Conidial powder was added to distilled water in 50-ml plastic centrifuge tubes at concentrations of 0.2 or 1.0 mg/ml. After addition of 1 g of glass beads (2.0 mm diameter), each tube was agitated on a wrist action shaker (Model BT, Burrell ScientiWc, Pittsburgh, PA) set at maximum speed (6.7 oscillations/s) for 15 min. All fungal treatments were applied with a Burgerjon spray tower (Burgerjon, 1956), using the protocol described by Ugine et al. (2005a). The actual application rates (conidia/mm2) were quantiWed by counting conidia deposited on polystyrene petri dish lids (90 mm diam.), which were sprayed simultaneously with the leaf disks. The enumeration protocol was that described by Wraight et al. (1998). Conidial viabilities were determined by exposing a petri dish containing yeast extract agar (1%) to each spray application. The inoculated plates were sealed with ParaWlm and incubated for 24 h at 25 § 1 °C. The spray deposits were then stained with acid fuchsin, scanned under phase contrast microscopy (400£), and the Wrst 100 conidia encountered were scored for germination (793% in all cases). This procedure was repeated at Wve random locations per petri plate. Numbers of conidia applied in all assays were corrected for viability. 2.2. Bioassay protocol Aliquots of aqueous conidial suspensions were applied to the abaxial surfaces of all impatiens leaf disks (2.0 cm diam.) on a rotating turntable (33 rpm) in the bottom of the

188

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

spray tower. All leaf disks used in a single test were treated simultaneously to ensure that all insects were exposed to spray residues with equivalent conidial densities. Control leaf disks were treated with BotaniGard 22WP formulation blank provided by the manufacturer at application concentrations equal to those of each respective test; these treatments are hereafter referred to as carrier controls. Pairs of leaf disks with dry spray deposits were placed abaxial surfaces together between two pieces of Wlter paper (2.5 cm diam.). This leaf-disk “sandwich” was placed in the bottom of a 30 ml portion cup, and 35 l of water was added to each piece of Wlter paper to humidify the chamber and slow leaf desiccation (for a more detailed description of bioassay procedures, see Ugine et al., 2005a). Pollen on excised tissue from one freshly opened impatiens Xower was placed between the leaf disks in the appropriate treatments. A randomly selected pair of thrips was then added to each cup (tapped on top of Wlter paper). The treatments were thus: (1) impatiens leaf disks treated with B. bassiana, without impatiens pollen, (2) impatiens leaf disks treated with B. bassiana, with impatiens pollen, (3) impatiens leaf disks treated with carrier, without impatiens pollen, and (4) impatiens leaf disks treated with carrier, with impatiens pollen. Cups were sealed with ParaWlm and capped with the plastic snap-on lids provided by the manufacturer. Cups were held at 25 § 1 °C under a natural light regime (incubator with glass-door) in a laboratory with unshaded windows. A single replicate of the test was conducted at the low concentration (0.2 mg/ml) of B. bassiana formulation corresponding to a mean (§SE) application rate of 117 § 22 viable conidia/mm2, hereafter referred to as test 1. Because we were interested in investigating the eVects of B. bassiana at realistic concentrations (rates that kill 50% of a given thrips population in 65 days), we decided not to repeat the low concentration assay. Two replicate assays (2a and 2b) were conducted at a higher concentration (1.0 mg/ml) of B. bassiana formulation corresponding to respective mean (§SE) application rates of 1013 § 87 and 1003 § 111 viable conidia/mm2, hereafter referred to as test 2. Ten replicate cups of each treatment and carrier control were constructed in tests 1 and 2b, and 12 and 6 replicate cups of each treatment and carrier control-treated cups (respectively) were constructed in test 2a following a proportionally replicated design. Tests 1 and 2a were conducted over a period of 2 months and test 2b 18 months later. Thrips were exposed to B. bassiana- or carrier-treated foliage for only the Wrst 24 h. The pairs of thrips were removed (aspirated) from their cups every 24 h, at the same time each day, and transferred to new cups containing fresh untreated impatiens foliage with or without pollen. Separate aspiration vials were used for fungus exposed insects and control insects; aspiration vials were sterilized daily. Day of death was recorded for each female; dead males were replaced with new males (672 h old), if survived by the female. Cups from which adults were removed were incubated for an additional 5 days at 25 § 1 °C, and

emerged nymphs were enumerated. This provided a tally of the daily oVspring production for each individual female. Dead females were placed onto 1% water agar to assess the cause of death. Insects that supported rapid fungal sporulation (within 24–48 h) were considered to have been killed by the fungus. The greenhouse-grown impatiens plants from which leaves were removed for use in bioassays typically had a lowlevel thrips infestation. To control for background numbers of thrips emerging from impatiens leaves, 10–15 control cups were constructed daily in the same fashion as treatment cups, except adult thrips were not added. These cups were incubated with the treatment cups and emerged nymphs were enumerated 5 days post-construction. Treatment counts were corrected for background levels of thrips, which were small in all cases (0.06–0.16 thrips/leaf/day), by subtracting the average number of emerged nymphs in adult-free cups on a given day from the respective counts in the treatments. 2.3. Statistical analysis Computations for all experiments were made using the statistical software package JMP version 4 (JMP, Version 4. SAS Institute Inc., Cary, NC, 2001). Two-way analysis of variance (ANOVA) was conducted on the daily oVspring production, total (lifetime) oVspring production, longevity, and oviposition duration data of all insects treated or not treated with B. bassiana in the presence or absence of pollen (2 £ 2 factorial). Total oVspring, longevity, and ovipositional duration data were transformed to log (x + 1) prior to ANOVA. Since the low and high application rates were applied in independent tests (tests 1 and 2, respectively), direct statistical comparisons among application rates were not conducted. Because the experiment was conducted twice at the high rate, an additional variable, date, was added to the model. Additional analyses were conducted on data sets that excluded insects that did not support fungal sporulation due to concerns that these insects might bias the results. However, results obtained from the second set of analyses were in all cases similar to those from tests including all insects. Given that there is a possibility of atypical (nonsporulating) fungal infections, and the fact that there was no change in the signiWcance of test results between the two data sets, the results from analyses conducted on the full data sets are reported herein. Additionally, there were a few instances (n D 3, all in diVerent experimental treatments) of insects dying within the initial 24 h of the experiment. We assume that these insects did not die from fungal infections (supported by plating on water agar) or from bacterial septicemia resulting from fungal penetration, which typically takes 724 h. These insects were excluded from analyses in each data set. Multiple comparisons were made, when appropriate, using the Tukey–Kramer test ( D 0.05). Daily oVspring production (oVspring production rate) per female thrips was calculated by dividing the total oVspring production (number of nymphs produced per female over

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

the entire experiment) by the number of days that the insect survived. Because rate data may not be normally distributed, we conWrmed the results obtained with log (x + 1) transformed data by running an analysis using the Box–Cox transformation (see Sokal and Rohlf, 1995). The results from this analysis gave virtually identical results. Two separate analyses were conducted to determine if B. bassiana infections had sublethal (pre-mortem) eVects on thrips fecundity. The Wrst entailed testing whether oVspring production by treated thrips was reduced relative to controls during the day before the day the insect was found dead due to mycosis (the last full day an insect was scored as being alive). The number of oVspring produced on the day before death of insects exposed to B. bassiana was compared to oVspring production by a randomly selected control insect on that same day (paired ANOVA). The second analysis was conducted only on insects treated at the high rate. The total number of insects that died each day from exposure to B. bassiana was determined for both the pollen and no pollen treatments. From this we identiWed the day of maximum mortality and thus the data set providing the greatest power for statistical comparison (largest sample sizes). Maximum mortality, in both groups, occurred on day 6, and thus day 5 was the last full day of oVspring production by these thrips. Numbers of nymphs produced on day 5 by these insects were then compared to the numbers produced by the control insects on day 5. The analysis included all control insects that lived for at least 6 days. For all statistical tests insigniWcant interactions were reported and then dropped from statistical models and analyses rerun; this leads to diVerences in the degrees of freedom error reported for test of interactions and main eVects among life stages when interactions are insigniWcant. 3. Results 3.1. Longevity Longevity of female thrips was not aVected by the presence of pollen at the low or high application rates of

189

B. bassiana (low rate: F[1,33] D 0.03, P D 0.86; high rate: F[1,64] D 0.007, P D 0.93) (Table 1). Treatment with B. bassiana, however, signiWcantly reduced the longevity of adult female thrips by 3.9 and 4.0 days, respectively (low rate: F[1,33] D 16.6, P D 0.0003; high rate: F[1,64] D 26.9, P < 0.0001). There were no signiWcant pollen £ B. bassiana interactions (low rate: F[1,32] D 0.73, P D 0.40; high rate: F[1,63] D 1.0, P D 0.33). Large percentages of the female thrips exposed to B. bassiana ultimately showed signs of death by mycosis, as evidenced by sporulation when placed on water agar (low rate, 88% (16/19 individuals); high rate, 100% (40/40 individuals)). 3.2. Oviposition period The mean oviposition period (period of time from the Wrst to last day of oviposition) was not aVected by the presence of pollen in either test (low rate: F[1,33] D 0.69, P D 0.41; high rate: F[1,64] D 2.5, P D 0.12) (Table 1). B. bassiana applied at the low and high rates signiWcantly reduced the mean oviposition period by 3.4 and 3.0 days, respectively (low rate: F[1,33] D 9.4, P D 0.004; high rate: F[1,64] D 4.7, P D 0.03). There was not a signiWcant pollen £ B. bassiana interaction in either test (low rate: F[1,32] D 3.2, P D 0.08; high rate: F[1,63] D 2.4, P D 0.13). 3.3. Daily oVspring production Average daily oVspring production (total oVspring produced/days alive) varied signiWcantly as a function of pollen presence at the low application rate (F[1,33] D 17.2, P D 0.0002) (Table 2). When impatiens leaves were supplemented with impatiens pollen, oVspring production was increased 2.3-fold. Fungal infection resulting from the lowrate applications of B. bassiana did not aVect daily oVspring production (F[1,33] D 0.07, P D 0.80). Additionally there was no pollen £ B. bassiana interaction (F[1,32] D 0.45, P D 0.51). At the high B. bassiana application rate, the main eVects of pollen and B. bassiana were similar to those observed at

Table 1 Longevity and oviposition period of adult female western Xower thrips exposed to B. bassiana on impatiens leaf disks supplemented or not supplemented with impatiens pollen Longevitya

Test 1

c

Test 2c

Pollen

No pollen

Grand mean

Pollen

No pollen

Grand mean

B. bassiana Control

7.4 § 1.0 10.6 § 0.9

6.8 § 0.9 11.7 § 1.1

7.3A 11.2B

8.0 § 0.9 9.4 § 1.0

5.7 § 0.8 11.1 § 1.4

6.9A 10.3B

Grand mean

9.0a

9.2a

8.6a

8.4a

B. bassiana Control Grand mean

a

Oviposition period (days)

Treatments:b

5.0 § 0.5 9.8 § 1.1 6.8a

5.5 § 0.6 8.7 § 0.8 6.9a

5.2A 9.2B

5.0 § 0.5 9.6 § 0.9 6.8a

5.1 § 0.7 6.7 § 1.2

5.0A 8.0B

5.8a

Mean longevity in days (§standard error; test 1, n D 9; test 2, treatments n D 19–21; controls n D 13–15). Thrips exposed 24 h to leaf disks treated with an average of 117 and 1008 viable conidia/mm2 in tests 1 and 2, respectively. c Within each test, means within columns followed by the same uppercase letter and means within rows followed by the same lowercase letter are not signiWcantly diVerent (main eVects from F test,  D 0.05). b

190

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

Table 2 Total oVspring production and daily age-speciWc oVspring production of adult female western Xower thrips exposed to B. bassiana on impatiens leaf disks supplemented or not supplemented with impatiens pollen Daily oVspring productiona b

Test 1

Test 2b,d

Lifetime oVspring productiona

Treatments:c

Pollen

No pollen

Grand mean

Pollen

No pollen

Grand mean

B. bassiana Control

7.3 § 1.0 6.5 § 1.1

3.0 § 0.7 3.0 § 0.4

5.3A 4.6A

56.1 § 10.1 67.4 § 11.1

21.1 § 5.2 36.2 § 5.9

39.5A 50.9B

Grand mean

6.9A

3.0 B

61.1A

28.7B

B. bassiana Control

5.7 § 0.6a 6.9 § 0.7a

1.8 § 0.3b 1.4 § 0.4b

29.6 § 4.1a 66.9 § 8.6b

10.9 § 2.3c 13.5 § 3.6c

Grand mean

6.1A

1.6B

43.8A

12.1B

3.8A 4.0A

20.7A 38.3B

a

Mean number of nymphal thrips produced from eggs deposited in impatiens leaf disks § standard error. b Thrips exposed 24 h to leaf disks with an average of 117 and 1008 viable conidia/mm2 in tests 1 and 2, respectively. c Within each test, means within columns and means within rows followed by the same letter are not signiWcantly diVerent (main eVects from F tests;  D 0.05). d SigniWcant pollen by B. bassiana interaction term, means within columns and means within rows followed by the same lowercase letter are not signiWcantly diVerent (Tukey–Kramer test;  D 0.05).

the low application rate (F[1,64] D 95.8, P < 0.0001; F[1,64] D 0.26, P D 0.61, respectively); although the eVects of pollen appeared more pronounced, increasing daily oVspring production by 3.8-fold (Table 2). In this case a pollen x B. bassiana interaction was detected; however, signiWcance was marginal (F[1,63] D 3.9, P D 0.05), and no signiWcant eVect of B. bassiana was observed in the presence or absence of pollen, and the pollen eVects were signiWcant regardless of the presence or absence of B. bassiana (Table 2). Average oVspring production per surviving female thrips (age-speciWc rates of oVspring production) was plotted for each day for all treatments (Fig. 1), and cumulative means

are presented in Fig. 2. It is noteworthy that the decline in accumulation is caused by the fact that successive bars do not report the reproduction by the same females (as females die, their cumulative production is lost); the means are thus based on decreasing numbers of females over time. As expected, the data revealed a reduction in oVspring production with increasing age of the thrips. Maximum rates of oviposition were noted within the Wrst week after mating, and cumulative production tended to level oV after 8 or 9 days. Cohort-speciWc rates of reproduction are based on the total number of females initially treated and thus reXect

Fig. 1. Daily mean number of oVspring produced per surviving female thrips (age-speciWc rate of reproduction). Thrips exposed 24 h to two application rates of B. bassiana (117 and 1013 viable conidia/mm2) on impatiens leaf disks and subsequently maintained on impatiens leaf disks supplemented or not supplemented with impatiens pollen. Error bars represent standard errors of the means.

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

191

Fig. 2. Cumulative daily mean oVspring per surviving female thrips (age-speciWc rate of reproduction). Totals are derived from data presented in Fig. 1. Error bars represent standard errors of the means.

Fig. 3. Daily mean number of oVspring produced per treated female thrips (cohort-speciWc rate of reproduction). Thrips exposed 24 h to two application rates of B. bassiana (117 and 1013 viable conidia/mm2) on impatiens leaf disks and subsequently maintained on impatiens leaf disks supplemented or not supplemented with impatiens pollen. Error bars represent standard errors of the means.

the time-dependent decrease in oVspring production by the treatment cohort as a whole, due to both increasing age and mortality of the individuals comprising the cohort (each mean incorporating zero values for deceased individuals). Daily average cohort-speciWc oVspring per treated female and cumulative averages

were also plotted for each day (Figs. 3 and 4). Because B. bassiana did not cause signiWcant pre-mortem eVects (see below), daily reductions in oVspring production due to fungal infection were consistently evident only when reduced longevity was factored into the calculations (cf. Figs. 1–4).

192

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

Fig. 4. Cumulative daily mean oVspring per treated female thrips (cohort-speciWc rate of reproduction). Totals are derived from data presented in Fig. 3. Error bars represent standard errors of the means.

treated insects produced 22 and 46% fewer total oVspring compared to control insects in tests 1 and 2, respectively.

3.4. Lifetime oVspring production Lifetime oVspring production (total oVspring produced) per female thrips varied signiWcantly as a function of both pollen presence and B. bassiana treatment at the low rate (F[1,33] D 11.9, P D 0.002; F[1,33] D 4.2, P D 0.05) and the high rate (F[1,63] D 48.2, P < 0.0001; F[1,63] D 4.8, P D 0.03), respectively, and there was no signiWcant pollen £ B. bassiana interaction at the low rate (F[1,32] D 0.87, P D 0.36) (Table 2). At the high rate there was a signiWcant pollen by B. bassiana interaction (F[1,63] D 4.8, P D 0.03) (Table 2). Lifetime oVspring production increased 2.1 and 3.6 times with the addition of pollen in tests 1 and 2, respectively. B. bassiana-

3.5. OVspring production on the day before death due to mycosis There was no signiWcant reduction in oVspring production relative to controls on the penultimate day before death of thrips exposed to B. bassiana within either pollen treatment (with or without pollen) for either test (Table 3) in the paired analysis. In the second analysis (pooled controls), there was again no eVect of B. bassiana on oVspring production the day before death (F[1,34] D 0.11, P D 0.74).

Table 3 Mean number of oVspring produced on the penultimate and last full days before death of adult female thrips exposed to B. bassiana spray residues on impatiens leaf disks either supplemented or not supplemented with impatiens pollen OVspring production 1 day before deatha

Test 1c

Test 2c

a

OVspring production 2 days before deathb

Treatments:

Pollen

No Pollen

Grand mean

Pollen

No Pollen

Grand mean

B. bassiana Control

5.0 § 1.8 6.3 § 1.7

1.2 § 0.9 2.7 § 0.7

3.1 4.5

6.4 § 1.3 7.6 § 1.7

2.1 § 0.9 3.8 § 0.9

4.3 5.7

Paired F test

F[1,8] D 0.15, P D 0.71

F[1,8] D 3.61, P D 0.09

F[1,8] D 0.15, P D 0.71

F[1,8] D 3.61, P D 0.09

B. bassiana Control

5.5 § 1.1 7.6 § 1.1

1.7 § 0.5 1.9 § 0.5

6.7 § 0.9 7.6 § 0.7

1.6 § 0.4 1.1 § 0.4

Paired F test

F[1,13] D 1.3, P D 0.27

F[1,13] D 0.01, P D 0.91

F[1,13] D 1.3, P D 0.27

F[1,13] D 0.01, P D 0.91

3.6 4.8

4.3 4.4

Mean number (§SE; test 1: n D 9; test 2: n D 14–15) of nymphal thrips produced per female from eggs deposited on the day before the day of death due to mycosis (last data record from a full 24-h period). b Mean number (§SE) of nymphal thrips produced per female from eggs deposited on the penultimate day before the day of death due to mycosis. c Thrips exposed 24 h to leaf disks treated with an average of 117 and 1008 viable conidia/mm2 in tests 1 and 2, respectively.

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

B. bassiana-treated insects (n D 10) produced 4.0 § 0.9 oVspring versus control insects (n D 26) that produced 5.3 § 0.9 oVspring. However, inspection of the grand mean oVspring production 1 day before death (Table 3) reveals that controls consistently produced 25–30% more oVspring compared to B. bassiana exposed insects. Thus, ANOVA was performed on the log (x + 1) transformed mean oVspring production 1 day before death. There was a signiWcant eVect of pollen on oVspring production (F[1,5] D 120.7, P D 0.0001) as well as a signiWcant eVect of exposure to B. bassiana (F[1,5] D 10.3, P D 0.02) but no pollen x fungus interaction (F[1,4] D 0.10, P D 0.77). B. bassiana reduced oVspring production on the day before death by 1.2 oVspring/female thrips. Two days before death, there was a signiWcant eVect of pollen on oVspring production (F[1,5] D 2.7, P D 0.003) but not an eVect of B. bassiana (F[1,5] D 0.4, P D 0.55) or the pollen £ fungus interaction (F[1,4] D 0.002, P D 0.97). 4. Discussion A large increase in both lifetime oVspring production and the daily rate of oVspring production occurred when impatiens leaves were supplemented with pollen in tests conducted at two diVerent application rates of B. bassiana. This result was not due to an increase in longevity or ovipositional period. Pollens are known to have high protein content (Mound et al., 1980), and pollen supplements in thrips diets are known to increase egg production (Englemann, 1984; Hulshof et al., 2003; Trichillo and Leigh, 1988). Regardless of the level of B. bassiana (treated or control), insects whose diet was supplemented with pollen always produced more oVspring (daily and lifetime) than insects that did not receive pollen. This is despite the fact that insects exposed to B. bassiana died signiWcantly earlier than controls in tests 1 and 2 (3.9 and 4.0 days earlier, respectively). The signiWcant increase in lifetime oVspring production was similar to the Wndings obtained by Trichillo and Leigh (1988), who demonstrated that WFT oVspring production increased 4-fold when cotton foliage was supplemented with cotton pollen, and to the Wndings of Hulshof et al. (2003), who found that supplementing cucumber leaves with various pollens signiWcantly increased WFT oVspring production. The magnitude of the eVect of pollen on oVspring production, when compared across studies, is variable and seems dependent on the host plant and pollen species. Because the two tests (diVerent rates) were conducted independently, statistical comparisons between them could not be made. However, the results strongly suggest a B. bassiana concentration eVect on percent mortality and longevity and an associated eVect on lifetime oVspring production. At the low concentration (ca. 100 conidia/mm2), B. bassiana-treated insects lived an average of 7.3 days. When the concentration was increased 10-fold to 1000 conidia/mm2, insects died on average 2 days earlier and displayed greater

193

levels of mortality due to infection by the fungus (100% vs. 88%). Inspection of the mean daily oVspring production in the no fungus/no pollen treatment of each test (Table 2) reveals that insects in test 1 may have been more vigorous than insects used in test 2, producing 3 versus 1.4 oVspring/ day, respectively. Interestingly, pollen enhanced the vigor of the thrips as seen in the no fungus/pollen treatment, where the daily rate of oVspring production was 6.5 and 6.9 oVspring/day, respectively. The reduced vigor and daily handling of B. bassiana-exposed insects may have unknown eVects on their susceptibility; however, because insects exposed to fungus did not die more rapidly than we expected compared to similar assays we have conducted (data not shown), these transfers had negligible eVects. Pre-mortem eVects of fungal infection on the daily rate of oVspring or egg production in insects are highly variable. Fargues et al. (1991) showed that adult Colorado potato beetles, Leptinotarsa decemlineata (Say), surviving treatment with B. bassiana as newly molted fourth instars, laid signiWcantly fewer eggs/day compared to control insects over a 40-day period. Similarly, Noma and Strickler (2000) found that the oviposition rate of B. bassiana inoculated Lygus hesperus (Knight) was signiWcantly lower than control insects in a 6-day assay. Ekesi and Maniania (2000) investigated the eVects of Metarhizium anisopliae (MetchnikoV) on the fecundity of the legume Xower thrips, Megalurothrips sjostedti (Trybom), and found that adult thrips surviving larval infection produced fewer eggs/female/day versus control females. Contrary to these Wndings, however, Lord et al. (1987) found that chrysomelid beetles Cerotoma arcuata (Olivier), treated with B. bassiana did not vary in their oviposition rate compared to control insects. Similarly, Wang and Knudsen (1993) reported that the Russian wheat aphid, Diuraphis noxia (Kurdj.), exposed to B. bassiana, did not produce signiWcantly diVerent numbers of nymphs compared to control insects. In this study, we found no eVect of B. bassiana infection on the overall age-speciWc rate of daily oVspring production, and there seemed to be no diVerence in daily oVspring production trends with respect to application rate. The Wnding of no eVect of infection on daily oVspring production rate can be considered evidence of no sublethal eVect of infection on fecundity. However, analyses of the average daily oVspring production rate may overlook more subtle diVerences in the daily oVspring production rate such as might be observed only near the time of death. Further analyses (within each pollen treatment) comparing the rates of oVspring production of B. bassiana-exposed females to controls, on individual days before death, revealed that there was no eVect of infection even 1 day before the insect died. However, when the variation was removed from the data by taking the average across insects within each test, there was a signiWcant 25–30% reduction in oVspring production on the day before death due to mycosis. We anticipated Wnding a diVerence in oVspring production 1 or 2 days before death because thrips are synovigenic (adults do not emerge with fully developed eggs, and the eggs develop

194

T.A. Ugine et al. / Biological Control 37 (2006) 186–195

gradually). Multiplication of fungal blastospores in the insect hemocoel consumes host nutrients, and infection would presumably deplete energy reserves needed for egg maturation. Given that most fungal infections, once initiated, progress in a standard fashion, we expected that all insects would experience similar pathogen loads before death and that a pre-mortem eVect would be found regardless of the application rate. The thrips in this study responded more slowly than expected to fungal infection at the concentrations applied. Previous studies investigating the eVects of B. bassiana residues on adult female western Xower thrips exposed to treated kidney bean leaf disks found that densities of 19 conidia/mm2 achieved 50% mortality in a 5-day assay (LC50 D 19 conidia/mm2) (Ugine et al., 2005b). In the present study, the percent mortality after 5 days of thrips exposed to B. bassiana conidia at densities of 117 and 1008 conidia/mm2 on impatiens leaf disks was 26% (5/19) and 65% (26/40), respectively. This suggests that thrips on impatiens foliage are less susceptible to infection than thrips on bean foliage. The mechanism behind this diVerence is unknown and could be related to several factors, including how eYciently thrips acquire conidia from the diVerent types of treated foliage, or host plant chemistry. When considering the use of slow-acting biological control agents against insects with high reproductive potential such as thrips, it is important to assess the impact of the biological control agent on oVspring production (total and daily rate of oVspring production) and longevity. If B. bassiana cannot kill adult female thrips before they produce a substantial percentage of their total oVspring, then foliar applications of B. bassiana against thrips may be ineVective in preventing pest population increases unless infection is achievable in a high percentage of nymphs (preventing these insects from reaching reproductive maturity). This represents a diYcult challenge, however, as nymphs are considerably less susceptible than adults because molting of the exoskeleton during development limits fungal infection (Ugine et al., 2005a; Vandenberg et al., 1998; Vey and Fargues, 1977). On the other hand, death due to B. bassiana mycosis is not so slow as to preclude signiWcant impacts on the overall rates of pest population increase. The high application rates ultimately reduced thrips lifetime fecundity by nearly 50% (Table 2). This level of population suppression could be useful in IPM systems and especially early in cropping cycles before plants begin to Xower and produce pollen. Also, western Xower thrips is primarily a pest of protected crops, and agents capable of slowing rates of pest population increase may be useful in short-cycle crops such as bedding plants, which do not remain in the greenhouse for protracted periods. Acknowledgments This research was funded in part through a SpeciWc Cooperative Agreement between the USDA/ARS Plant Protection Research Unit and the Cornell University

Department of Entomology, Ithaca, New York (SCA #581097-9-033) funded by the USDA/ARS, as part of the Floriculture and Nursery Research Initiative. References Blanford, S., Thomas, M.B., 2001. Adult survival, maturation, and reproduction of the desert locust Schistocerca gregaria infected with the fungus Metarhizium anisopliae var acridum. J. Invertebr. Pathol. 78, 1–8. Bradley, C.A., Black, W.E., Kearns, R., Wood, P., 1992. Role of production technology in mycoinsecticide development. In: Leatham, G.F. (Ed.), Frontiers in Industrial Mycology. Chapman and Hall, New York, pp. 160–173. Burgerjon, A., 1956. Pulvérisation et poudrage au laboratoire par des préparations pathogènes insecticides. Ann. I.N.R.A. Serie Epiphyties 2, 677–687. Castillo, M.-A., Moya, P., Hernández, E., Primo-Yúfera, E., 2000. Susceptibility of Ceratitis capitata Wiedemann (Diptera: Tephritidae) to entomopathogenic fungi and their extracts. Biol. Control 19, 274–282. Daughtery, M.L., Jones, R.K., Moyer, J.W., Daub, M.E., Baker, J.R., 1997. Tospoviruses strike the greenhouse industry: INSV has become a major pathogen on Xower crops. Plant Dis. 81, 1220–1230. Ekesi, S., Maniania, N.K., 2000. Susceptibility of Megalurothrips sjostedti developmental stages to Metarhizium anisopliae and the eVects of infection on feeding, adult fecundity, egg fertility and longevity. Entomol. Exp. Appl. 94, 229–236. Englemann, F., 1984. Reproduction in insects. In: HuVaker, C.B., Rabb, R.B. (Eds.), Ecological Entomology. John Wiley and Sons, New York, pp. 113–114. Fargues, J., Delmas, J.-C., Augé, J., Lebrun, R.A., 1991. Fecundity and egg fertility in the adult Colorado beetle (Leptinotarsa decemlineata) surviving larval infection by the fungus Beauveria bassiana. Entomol. Exp. Appl. 61, 45–51. Gerin, C., Hance, Th., Van Impe, G., 1999. Impact of Xowers on the demography of western Xower thrips Frankliniella occidentalis (Thysanoptera, Thripidae). J. Appl. Entomol. 123, 569–574. Herron, G.A., Rophail, J., Gullick, G.C., 1996. Laboratory-based, insecticide eYcacy studies on Weld-collected Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) and implications for its management in Australia. Austr. J. Entomol. 35, 161–164. Hsiao, W.F., Bidochka, M.J., Khachatourians, G.G., 1992. EVect of temperature and relative humidity on the virulence of the entomopathogenic fungus, Verticillium lecanii, toward the oat-bird berry aphid, Rhopalosiphum padi (Hom., Aphididae). J. Appl. Entomol. 114, 484–490. Hulshof, J., Ketoja, E., Vänninen, I., 2003. Life history characteristics of Frankliniella occidentalis on cucumber leaves with and without supplemental food. Entomol. Exp. Appl. 108, 19–32. Immaraju, J.A., Paine, T.D., Bethke, J.A., Robb, K.L., Newman, J.P., 1992. Western Xower thrips (Thysanoptera:Thripidae) resistance to insecticides in coastal California greenhouses. J. Econ. Entomol. 85, 9–14. JMP, Version 4. SAS Institute Inc., Cary, NC, 2001. Kirk, W.D.J., 1985. Pollen-feeding and the host speciWcity and fecundity of Xower thrips (Thysanoptera). Ecol. Entomol. 10, 281–289. Lacey, L.A., Mesquita, A.L.M., Mercadier, G., Debire, R., Kazmer, D.J., Leclant, F., 1997. Acute and sublethal activity of the entomopathogenic fungus Paecilomyces fumosoroseus (Deuteromycotina: Hyphomycetes) on adult Aphelinus asychis (Hymenoptera: Aphelinidae). Environ. Entomol. 26, 1452–1460. Lesky, T.C., Teulon, D.A.J., Cameron, E.A., 1997. EVects of temperature and sugar maple pollen on oviposition and longevity of pear thrips (Thysanoptera: Thripidae). Environ. Entomol. 26, 566–571. Lord, J.C., Magalhaes, B.P., Roberts, D.W., 1987. EVects of the fungus Beauveria bassiana on the behavior, oviposition and susceptibility to secondary infection of adult Cerotoma arcuata Oliver (Colepotera: Chrysomelidae). Anais Soc. Entomol. Brasil 16, 187–198. Mound, L.A., Heming, B.S., Palmer, J.M., 1980. Phylogenetic relationships between families of recent Thysanoptera (Insecta). Zool. J. Linnean Soc. 69, 111–141.

T.A. Ugine et al. / Biological Control 37 (2006) 186–195 Nielsen, C., Skovgård, H., Steenberg, T., 2005. EVect of Metarhizium anisopliae (Deuteromycotina: Hyphomycetes) on survival and reproduction of the Wlth Xy parasitoid, Spalangia cameroni (Hymenoptera: Pteromalidae). Environ. Entomol. 34, 133–139. Noma, T., Strickler, K., 2000. EVects of Beauveria bassiana on Lygus Hesperus (Hemiptera: Miridae) feeding and oviposition. Environ. Entomol. 29, 394–402. Sokal, R.R., Rohlf, F.J., 1995. Biometry: The Principles and Practice of Statistics in Biological Research, third ed. W.H. Freeman and Co, New York. Teulon, D.A.J., Penman, D.R., 1991. EVects of temperature and diet on oviposition rate and development time of the New Zealand Xower thrips, Thrips obscuratus. Entomol. Exp. Appl. 60, 143–155. Trichillo, P.J., Leigh, T.F., 1988. InXuence of resource quality on the reproductive Wtness of Xower thrips (Thysanoptera: Thripidae). Ann. Entomol. Soc. Am. 81, 64–70. Ugine, T.A., 2004. Evaluation of the entomopathogenic fungus Beauveria bassiana, for microbial control of the western Xower thrips, Frankliniella occidentalis, in greenhouse impatiens. Ph.D. Dissertation, Cornell Universtiy, 208 p. Ugine, T.A., Wraight, S.P., Brownbridge, M., Sanderson, J.P., 2005a. Development of a novel bioassay for estimation of median lethal concentrations (LC50) and doses (LD50) of the entomopathogenic fungus Beauveria bassiana, against western Xower thrips, Frankliniella occidentalis. J. Invertebr. Pathol. 89, 210–218.

195

Ugine, T.A., Wraight, S.P., Sanderson, J.P., 2005b. Acquisition of lethal doses of Beauveria bassiana conidia by western Xower thrips, Frankliniella occidentalis, exposed to foliar spray residues of formulated and unformulated conidia. J. Invertebr. Pathol. 90, 10–23. Vandenberg, J.D., Ramos, M., Altre, J.A., 1998. Dose–response and ageand temperature-related susceptibility of the diamondback moth (Lepidoptera: Plutellidae) to two isolates of Beauveria bassiana (Hyphomycetes: Monillaceae). Environ. Entomol. 27, 1017–1021. Vey, A., Fargues, J., 1977. Histological and ultrastructural studies of Beauveria bassiana infection in Leptinotarsa decemlineata larvae during ecdysis. J. Invertebr. Pathol. 30, 207–215. Wang, Z.G., Knudsen, G.R., 1993. EVects of Beauveira bassiana (Fungi: Hyphomycetes) on fecundity of the Russian wheat aphid (Homoptera: Aphididae). Envrion. Entomol. 22, 874–878. Wraight, S.P., Carruthers, R.I., Bradley, C.A., Jaronski, S.T., Lacey, L.A., Wood, P., 1998. Pathogenicity of the entomopathogenic fungi Paecilomyces spp. and Beauveria bassiana against the silverleaf whiteXy, Bemisia argentifolii. J. Invertebr. Pathol. 71, 217–226. Wraight, S.P., Carruthers, R.I., Jaronski, S.T., Bradley, C.A., Garza, C.J., Galaini-Wraight, S., 1999. Evaluation of the entomopathogenic fungi Beauveria bassiana and Paecilomyces fumosoroseus for microbial control of the silverleaf whiteXy, Bemisia argentifolii. Biol. Control 17, 403–417. Zhao, G., Liu, W., Knoles, C.O., 1994. Mechanisms associated with diazinon resistance in Western Xower thrips. Pestic. Biochem. Physiol. 49, 13–23.