- Email: [email protected]
Effects ofPuccinia carduorumon Musk Thistle Herbivores
BIOLOGICAL CONTROL
6, 123–129 (1996)
Article No. 0015
Effects of Puccinia carduorum on Musk Thistle Herbivores L. T. KOK,* R. G. ABAD,*,1
AND
A. B. A. M. BAUDOIN†
*Department of Entomology and †Department of Plant Pathology, Physiology and Weed Science, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 Received November 30, 1994; accepted April 24, 1995
Puccinia carduorum Jacky, a rust fungus recently introduced into the United States for biological control of musk thistle (Carduus thoermeri Weinmann), may interact with three established thistle herbivores, Trichosirocalus horridus (Panzer) (Coleoptera: Curculionidae), Rhinocyllus conicus Froelich (Coleoptera: Curculionidae), and Cassida rubiginosa Mu ¨ ller (Coleoptera: Chrysomelidae). Tests were conducted to determine the effects of P. carduorum on feeding, oviposition, longevity, egg production, egg hatch, and larval development of the three insects. When offered both rust-infected and healthy musk thistle leaves, adults of C. rubiginosa and T. horridus consumed significantly more healthy than infected foliage. On infected leaves, feeding and oviposition were confined largely to pustule-free areas. The amount of rust-infected and healthy foliage consumed by C. rubiginosa larvae was not significantly different, but feeding on the infected leaves was again confined to rust-free leaf areas. Rust infection did not reduce oviposition by the three insects, and R. conicus oviposited only on healthy portions of the bracts. There were no significant differences in longevity, egg production, percentage of egg hatch, and larval development for any of the three insects fed continuously with both healthy and rust-infected leaves. Thus, effects of P. carduorum on the herbivores were slight, and among the interactions that were found, none is expected to be detrimental to biological control. q 1996 Academic Press, Inc. KEY WORDS: biological control; Carduus thoermeri; musk thistle control; Cassida rubiginosa; Trichosirocalus horridus; Rhinocyllus conicus; rust fungus; Puccinia carduorum; rust effects on thistle herbivores.
INTRODUCTION
Several natural enemies of musk thistle have been introduced into the United States over the years. Compatibility of Rhinocyllus conicus (a weevil that feeds 1 Present address: Philippine Coconut Authority, Davao Research Center, P. O. Box 295, Davao City 8000, Philippines.
in the thistle heads) and Trichosirocalus horridus (a rosette-feeding weevil) has been demonstrated (Kok and Trumble, 1979; Kok, 1981 a,b). In addition, Cartwright (1983) showed that Cassida rubiginosa, a foliage feeder, also blends well with the two weevils. The rationale advanced by these workers was that each herbivore occupies a distinct niche on the plant, thus avoiding direct competition for resources. Puccinia carduorum, a rust fungus from the Mediterranean area, was introduced into the United States in 1987 as an additional biological control agent. The rust has become established in the eastern United States (Baudoin and Kok, 1994), and musk thistle seed production in rust-inoculated plots was reduced compared with that in fungicide-protected plots (Baudoin et al., 1993). Since the insect herbivores have, in general, successfully suppressed musk thistle populations, we wanted to determine whether there was any risk of the rust interfering with the effectiveness of the insects. In addition, there are several literature reports of insects (Barbe, 1964; Lewis, 1979) or mollusks (Ramsell and Paul, 1990) showing preferential feeding on rust-diseased plant tissues, which raises the concern that the insect herbivores might diminish the activity of the rust. Contact of P. carduorum with all three insect herbivores is likely, as the pathogen infects most parts of the plant. As a foliage feeder, it appears that C. rubiginosa is most exposed to the rust. Next are the adult and egg stages of T. horridus. Eggs of R. conicus which are laid on the bracts of thistle heads may be occasionally exposed to the rust, but the adults, which feed minimally on leaves, would have little contact. Larvae of both weevils (R. conicus larvae feed within thistle heads and T. horridus larvae feed in the crown tissues) have the least contact with the rust. Transmission of the rust by all three herbivores is known to occur (Kok and Abad, 1994); however, what effect the rust has on these herbivores is not known. In this study, we determined the influence of the rust on the feeding behavior, longevity, oviposition, egg hatch, and larval development of each insect.
123
/ m3565$0376
01-10-96 13:43:56
bca
1049-9644/96 $12.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AP-Bio Con
124
KOK, ABAD, AND BAUDOIN
MATERIALS AND METHODS
Musk thistle plants were grown continuously in the greenhouse. About half of these were inoculated with a P. carduorum isolate from Turkey which was provided by Dr. William Bruckart of the Foreign Diseases– Weed Science Research Unit, United States Department of Agriculture, Frederick, Maryland. The plants were inoculated by misting with urediniospore suspensions of the fungus (0.5 mg/5 ml of water with a drop of Tween 20; polyoxyethylene sorbitan monolaureate) using a compressed-air garden sprayer and held overnight in a moisture-saturated atmosphere. To increase infection levels and maintain sufficient infected leaves or plants for the experiments, inoculations were repeated as necessary. The inoculated plants were kept in a separate room from the noninoculated plants. Preference Test
C. rubiginosa Feeding. Field-collected adults and larvae (third and fourth instars) were used in this study. For both stages, insects were confined in 1-liter plastic cages (10 cages with three individuals/cage) and maintained at room temperature (307C) with 14-h photophase. Each cage was provided with two musk thistle leaves, one rust-infected and one healthy. The leaves were removed after 3 days for the larval stage and 2 days for the adult stage. Leaf area consumed was calculated using the transparent grid technique (Cartwright, 1983). Leaf-area consumption by the insects on rustinfected leaves and healthy leaves was compared. The study was repeated five times. Since it was observed that the insects tended to avoid rust pustules during feeding, this was confirmed as follows: Leaf disks (21-mm diameter, containing 5–20 pustules) were cut from rusted leaves with a cork borer and placed in 6-cm petri dishes with the adaxial side down on moist filter paper. The pustules on each disk were counted and the pattern sketched, before and after feeding. One or two C. rubiginosa adults or larvae were placed on each disk and allowed to feed at room temperature for 1 to 2 days until approximately 25– 50% of the leaf surface had been eaten. Insects were then removed, and the leaf surface eaten by the insects was estimated by placing a transparent plastic sheet with dots on a 2.5-mm grid over the leaf disks and counting dots over eaten and noneaten areas. The observed number of pustules eaten was compared with the number expected on the basis of the leaf area eaten using a x2 test with one degree of freedom. This test was conducted three times. Oviposition. Adults were collected from the field in mid- to late May and brought to the laboratory. After an acclimatization period of 1 week, male and female
/ m3565$0376
01-10-96 13:43:56
bca
pairs were confined in plastic cages in the laboratory (10 cages, three pairs/cage). Each cage was supplied with one rust-infected and one healthy leaf. The leaves were removed and replaced by new sets after 3 to 5 days. Observations were continued with up to five leaf changes. The number of oothecae and eggs laid on the infected leaves were compared with those laid on healthy leaves. This oviposition study was supplemented with field sampling of plants exposed to the rust and thistle herbivores. C. rubiginosa egg masses were counted on 62 plants in fungicide-protected, insects-only plots, and 56 rust-infected plants in plots with insects and rust in June when rust infection was increasing and leaf senescence in the rust-infected plants became evident. T. horridus Feeding. Adults collected in late June were brought into the laboratory, placed in a cold room (15 { 17C), and fed with musk thistle leaf clusters. After 1 week, the insects were sexed, paired, and placed in plastic cages (10 cages, three pairs/cage) at 247C following the protocol of Trumble and Kok (1978). As with C. rubiginosa, each cage was supplied with one healthy and one rust-infected leaf. The leaves were replaced by new sets after 1 week. The number of T. horridus feeding holes on the rust-infected leaves was compared with those on the healthy leaves. The experiment was repeated five times. Avoidance of rusted areas was confirmed as described for C. rubiginosa, except that three or four T. horridus adults per leaf disk were used. Oviposition. The insects used in the feeding-preference test were supplied with their respective diets until the completion of the oviposition period of the insect. The number of eggs oviposited on infected leaves was compared with the number laid on healthy leaves. This test was repeated five times. R. conicus Oviposition. Since adult feeding on musk thistle is minimal, the study focused on oviposition preference. Fifty healthy and 50 rust-infected heads of fairly uniform size were arbitrarily selected from thistle plants adjacent to rust-inoculated field plots in June. R. conicus eggs were counted on each head. Effects of Continuous Exposure of the Rust on Thistle Herbivores
C. rubiginosa Longevity. Adults from fields without the rust fungus were collected in June and brought to the laboratory. They were placed in large holding cages together with musk thistle leaves. After 1 week, male and female pairs were transferred to smaller (39 cm3) plastic
AP-Bio Con
P. carduorum EFFECTS ON THISTLE HERBIVORES
cages (10 cages, three pairs/cage). Insects in five of these cages were continuously supplied with rust-infected leaves while the other five had healthy leaves (two leaves per cage). The insects were maintained in a growth chamber (50% RH, 14-h photophase with light:dark [L:D] temperatures of 30:257C). Food leaves in the rust group were misted with urediniospore suspensions of the fungus (at least 2 ml/cage from a preparation of 25 mg spores/500 ml of water) whenever necessary to ensure that insects were continuously exposed to fungus propagules even if they avoided feeding on rust pustules themselves. The cultures were maintained until the last test insect died. Mortality and date of deaths were recorded for each group. Feeding. Adults collected in spring were brought to the laboratory. The same procedure was followed as in the longevity experiment except that the number of cages was increased to 20 (10 cages/treatment). Leafarea consumption by the beetles in the rust and control groups was calculated using the transparent grid technique (Cartwright, 1983). Oviposition, egg hatch, and larval development. Adults collected in May and brought to the laboratory were subjected to the same procedure as in the feeding test. From the start to the end of an oviposition period of the insect, oothecae from each batch of leaves in a cage were transferred every 4 or 5 days to plastic cups (11.5 cm3) lined at the bottom with sponge and covered with lids. The oothecae were provided with freshly cut musk thistle leaves every 3 days (for 2 weeks) to provide humidity and food as the eggs hatched (Ward, 1976). The number of oothecae and eggs and the percentage of hatch were recorded for each treatment. Two successive batches of larvae in each treatment were reared to last instar (15 days after eclosion) using their respective diets. The prepupae from each treatment were weighed and then measured under a calibrated dissecting microscope.
125
maintained in this manner until all the insects in each cage died. The feeding holes in each batch of leaves were counted. Likewise, the number and date of adult mortality were recorded. Oviposition and egg hatch. Two experiments were conducted. The first experiment utilized the insects in the preceding longevity and feeding test. During the oviposition period of the insect, the eggs from every batch of leaves in each cage were collected, counted, and placed in petri dishes with moist filter paper lining the bottom. These petri dishes were stored in a refrigerator at 47C and the filter paper was kept moist (Kok and McAvoy, 1983). After several days, the petri dishes were exposed to higher temperatures (15 { 17C) to induce the eggs to hatch. After 12 to 15 days, the percentage of hatch was recorded and the means for the two groups were compared. The second experiment involved insects collected in June of the following year; procedures were the same. Larval development. Fifteen healthy and 15 rustinfected potted musk thistle plants in their rosette stage were moved from the greenhouse to a cold room (15 { 17C, 14-h photophase). Six laboratory-reared first instars (õ48-h old) were introduced onto the leaves and growth point of each plant by means of a capillary tube sealed at one end to prevent escape of introduced larvae (Sieburth, 1981). Small holes were pierced by means of a needle (or the sharp end of forceps) on the leaves or growth points where the larvae were introduced. In addition, to ensure immediate contact of rust with the larvae, the holes were misted with suspensions of rust spores scraped from infected leaves (at least six infected leaves/100 ml of water). After 28 days, larvae were removed from each plant to be weighed and measured (under a calibrated dissecting microscope). Differences in these data between the two groups were analyzed. If the original host died before the end of the 28-day period, the test larvae were transferred to a new potted plant (rust-infected or healthy).
T. horridus Longevity and feeding. The experimental procedure was similar to that for C. rubiginosa. Adults were collected in July and brought to the laboratory. The insects were initially maintained under room conditions (25–307C) and fed with healthy musk thistle leaves. The following month, the insects were sexed and paired. The pairs were placed in plastic cages (10 cages, five pairs/cage) inside a growth chamber (50% RH, 9-h photophase and L:D temperatures of 20:107C) following the protocol of Trumble and Kok (1978). The short photophase was used because it induces oviposition of the weevil (Ward and Kok, 1975). Five of these cages were supplied with two rust-infected leaves and the other five with two healthy leaves per cage. The food leaves were replaced after 7–10 days. The cultures were
/ m3565$0376
01-10-96 13:43:56
bca
R. conicus Longevity. This study was conducted in two consecutive years. Potted healthy and rust-inoculated musk thistle plants in the stem-elongation stage were used. In the first year, R. conicus adults were collected in May and brought to the laboratory. The following week, 10 adults (five pairs) were introduced to each plant. For each potted plant, the stem and leaves (with the insects) above the soil line were enclosed in a Mylar cage. The cages were maintained under greenhouse conditions (25–307C, 14-h photophase), until all the insects died. Plants that died during the period of study were replaced. The number and date of mortality of test insects in both treatments were recorded. This study was replicated 10 times.
AP-Bio Con
126
KOK, ABAD, AND BAUDOIN
TABLE 1 Preference of Feeding and Oviposition by Cassida rubiginosa on Rust-Infected and Healthy Thistle Leaves Means { SE
Observations Feedinga Larva (cm2 leaf/larva) Adult (cm2 leaf/adult) Degree of leaf infectiond Leaf size (cm): Length Greatest width Oviposition Laboratoryb Number of oothecae Number of eggs Degree of leaf infectiond Fieldc Number of egg masses/plant Degree of leaf infectiond
Healthy
Rustinfected
t teste
9.8 { 1.2 14.0 { 0.8 0
7.8 { 1.2 9.0 { 0.7 3.0 { 0.1
ns * —
4.2 { 0.2 2.4 { 0.2
4.0 { 0.3 2.2 { 0.1
ns ns
45.4 { 9.1 233.2 { 41.1 0
36.8 { 7.1 200.0 { 44.0 2.8 { 0.3
ns ns —
1.2 { 0.6 1.8 { 0.2
ns —
Analysis of Data
The SAS statistical program (SAS Institute, 1985) was used for data analysis. Differences in insect feeding, oviposition, percentage egg hatch, and growth of larvae between the rust-infected and healthy leaves or plants were analyzed using Student’s t test (Zar, 1984). RESULTS Preference Test
2.1 { 0.9 0.02 { 0.03
a
Means of five tests, each of 10 replicates/treatment, three individuals/replicate. b Means of five oviposition periods, 10 replicates/treatment, three pairs/replicate. c 68 egg masses from 56 plants in the rust-infected group; 130 egg masses from 62 plants in the healthy plants group. d Rating scale: estimated number of rust pustules (0–5) per 6.25 cm2 surface area. e Significance: *significant (P õ 0.05); ns, not significant (P ú 0.05), Student’s t test for infected versus healthy leaf; —, not determined.
In the second year, the procedures were the same except that: (a) adults were collected in late April, (b) the cages were maintained in the laboratory (22–277C, 14-h photophase), (c) the adult insects in the rust group were initially submerged in a dish of spore suspensions of P. carduorum for 2 min (plain water in the case of the control) before introduction to their respective host plants, and (d) the test plants were replaced every 2 weeks or earlier when necessary. Oviposition. This aspect of the study was carried out by counting eggs every 2 weeks during change of test plants. The eggs from each test plant were collected and recorded as hatched or unhatched. The unhatched eggs were placed in a growth chamber (247C, 80% RH) and observed daily for hatch. Development of larvae. This field study was conducted using thistle stands adjacent to rust-inoculated plants in June. Fifteen heavily rust-infected musk thistle heads and 15 uninfected heads of fairly uniform size and age with R. conicus eggs were enclosed individually in paper bags. The bags were used to prevent contact with fresh rust-inoculum. Five weeks later, the bags were removed and the heads were harvested and
/ m3565$0376
brought to the laboratory. The thistle heads were dissected and the fourth instars or pupae were weighed. Newly formed pupae were obtained by dissections of additional uninfected and rust-infected heads bearing R. conicus eggs. The pupae were weighed and measured using a calibrated dissecting microscope. Weight and measurements of the pupae in the two categories (rust and control) were compared.
01-10-96 13:43:56
bca
C. rubiginosa Feeding. The areas of healthy leaves and infected leaves consumed by C. rubiginosa larvae were not different (Table 1). For adults, however, consumption was greater on healthy than on rust-infected leaves (P õ 0.05) (Table 1). Even on infected leaves, the insect preferred pustule-free surfaces of the leaf (Table 2). The results suggest that adults, which are more mobile, are better able to select and feed on an uninfected leaf while larvae merely avoid infected leaf areas. Oviposition. No differences were found for C. rubiginosa oothecae or eggs gathered from healthy leaves and rust-infected leaves (Table 1). As in the feeding test, the eggs were observed on pustule-free surface areas of infected leaves. Similar results were observed on plants in field plots. T. horridus Feeding. There were more T. horridus feeding holes recorded on healthy than rust-infected leaves (Table TABLE 2 Avoidance of Rust-Infected Leaf Areas of Musk Thistle during Feeding by Cassida rubiginosa and Trichosirocalus horridus Number of rust pustules
Insect
Stage
C. rubiginosa Adults C. rubiginosa Larvae T. horridus Adults a
Total 63 133 91
Observed Expecteda Probability of eaten eaten greater x2 4 6 1
Based on percentage of leaf area eaten.
AP-Bio Con
21 70 19
0.00001 õ0.00001 0.008
127
P. carduorum EFFECTS ON THISTLE HERBIVORES
TABLE 3 Preference of Feeding and Oviposition of Trichosirocalus horridus Adults on Rust-Infected and Healthy Thistle Leaves Means { SE Observations
Healthy
Feedinga Number of feeding holes/insect 14.9 { 6.3 Degree of leaf infectionb 0 Leaf size (cm): Length 4.1 { 0.1 Biggest width 2.4 { 0.1 Ovipositionc Number of eggs/female Degree of leaf infectionb
30.1 { 5.2 0
Rust-infected t testd
9.4 { 5.0 2.5 { 0.8
* —
4.0 { 0.1 2.3 { 0.1
ns ns
69.9 { 5.2 2.4 { 0.2
* —
a
Means of five feeding periods, 10 replicates/treatment, three pairs/replicate. b Rating scale: number of rust pustules (0–5) per 6.25 cm2 surface area. c Means of 10 replicates/treatment, three pairs/replicate. d Significance: *significant (P õ 0.05); ns, not significant (P ú 0.05), Student’s t test for rust-infected versus healthy leaves; —, not determined.
3). As with C. rubiginosa, feeding was largely confined to pustule-free areas of diseased leaves (Table 2). Oviposition. —Rust infection did not deter oviposition on rust-infected leaves, as there were significantly more eggs collected from diseased than healthy leaves (Table 3).
the rust-exposed group compared with only one pair in the control. Oviposition, egg hatch, and larval development. There were no differences in the numbers of oothecae and eggs oviposited, and percentage of egg hatch by insects continuously fed with healthy leaves and those fed with infected leaves (Table 4). No differences were noted in the sizes and weights of larvae that were continuously fed with infected leaves and those that were reared on healthy leaf diets, nor was there any difference in adult emergence. As previously observed, the larvae fed on pustule-free leaf areas of infected leaves. T. horridus Longevity and feeding. Adults continuously fed on rust-infected leaves and those fed on healthy leaves both survived for more than a year after field collection in July. Mortalities were 4.6 and 4.5 adults per month, respectively, for the rust and the control groups. After 14 months, there were two surviving adults in the former and three in the latter group. During the same period, the number of T. horridus feeding holes was not different between the two groups (Table 5). Oviposition, egg hatch, and larval development. There was no difference in the number of eggs ovipos-
TABLE 4 Effect of Continuous Exposure to Rust on Adult Feeding, Oviposition, and Larval Development of Cassida rubiginosa Means { SE
R. conicus Oviposition —No difference was noted in the mean number of R. conicus eggs collected from both rustinfected heads (43.2 { 3.2) and healthy heads (45.3 { 2.9). It was evident, however, that eggs were laid on healthy portions of the bracts. Effects of Continuous Exposure of the Rust on Thistle Herbivores
C. rubiginosa Longevity. Insects reared on both diets (healthy leaves and rust-infected leaves) survived for over 1 year, indicating that continuous exposure to the fungus did not affect the longevity of the insect. In fact, the last surviving female in the rust group still laid five egg masses after a year. Feeding. The leaf area consumed by C. rubiginosa continuously fed with infected leaves did not differ from the control (Table 4). As in earlier findings, feeding was confined to pustule-free surfaces of infected leaves. At the seventh month, three pairs of adults survived in
/ m3565$0376
01-10-96 13:43:56
bca
Observations
Control
Feedinga Leaf area consumed (sq cm/ insect) 60.7 { 7 Degree of leaf infectionb 0 Leaf size (cm) Length 4.1 { 0.2 Biggest width 2.1 { 0.1 Ovipositionc Number of oothecae/female 27.5 { 2.6 Number of eggs/female 167.0 { 17.4 Egg hatch (%) 80.0 { 2.3 Measurement of prepupaed Weight (mg) 18.6 { 0.6 Length (mm) 6.6 { 0.1 Width (mm) 3.2 { 0.04 Pupae emerging as adults (%) 91.0 a
Rust-infected t teste
52.2 { 4.0 1.9 { 0.5
ns —
4.1 { 0.1 2.2 { 0.1
ns ns
29.6 { 3.0 141.9 { 8.6 81.3 { 1.5
ns ns ns
17.8 { 0.6 6.3 { 0.1 3.2 { 0.04 88.0
ns ns ns ns
Means of 10 replicates, three pairs/replicate. Rating scale: number of rust pustules (0–5) per 6.25 cm2 surface area. c Means of seven oviposition periods (10 replicates/treatment, 3 pairs/replicate). d Means of 30 insects/treatment. e Significance: ns, not significant (P ú 0.05, Student’s t test) for rust-infected versus healthy leaf; —, not determined. b
AP-Bio Con
128
KOK, ABAD, AND BAUDOIN
TABLE 5 Effect of Continuous Exposure to Rust on Adult Feeding, Oviposition, and Larval Development of Trichosirocalus horridus Means { SE Observations
Control
Rust-infected
t testd
Feedinga Number of feeding holes/adult
400.7 { 58.0
431.1 { 48.6
ns
141.3 { 33.9
156.4 { 45.1
ns
79.4 { 3.8 76.9 { 3.4
74.5 { 3.4 77.2 { 2.8
ns ns
Oviposition Number of eggs/femaleb Egg hatch (%) 1st groupa 2nd groupb Measurement of larvaec Weight (mg) Length (mm) Width (mm)
9.5 { 0.4 5.0 { 0.1 2.0 { 0.03
9.1 { 0.3 4.9 { 0.1 2.0 { 0.04
ns ns ns
a
Means of five replicates, five pairs/replicate. Means of 10 replicates, three pairs/replicate. c Means of 23 larvae/treatment (28-day old larvae). d ns, not significant (P ú 0.05, Student’s t test) for rust-infected versus healthy leaf. b
ited by T. horridus that were continuously exposed to infected leaves and those raised on healthy foliage. Percentage of hatch of eggs from the two groups of insects was also not different. After 28 days, weights of larvae (offspring of parents which were also continuously fed rust-infected leaves) given a continuous diet of infected leaves did not differ from those fed with healthy leaves (Table 5). R. conicus Longevity. Mortalities of overwintered insects exposed to rust and those maintained on healthy musk thistle plants did not differ for the two groups in either year, indicating no effect of the presence of the fungus on the insect. Both groups had about 40% mortality after 3 weeks and 90% mortality after 5 weeks. This is not unusual as adults that emerge in spring die soon after they complete oviposition within 4–5 weeks. Oviposition, egghatch, andlarval development. Continuous exposure to the rust did not affect egg production and percentage of hatch of R. conicus. Also, the larval and pupal weights and sizes of R. conicus feeding on infected heads did not differ from those that were obtained from noninfected heads (Table 6). DISCUSSION
The overall results of this study indicated that exposure of the three herbivores to the rust had no significant effect on their life span, feeding, oviposition, egg hatch, and larval development. Exposure of the con-
/ m3565$0376
01-10-96 13:43:56
bca
fined insects and their various stages to the rust in most cases were abnormally continuous and long (6 months to over a year). Under natural conditions, the insects may be exposed to the rust for only a short period as they have freedom to feed and oviposit on musk thistle plants of their choice. There was a clear preference of C. rubiginosa adults and larvae, and T. horridus adults, for feeding on healthy leaves or pustule-free areas of infected leaves, in contrast to reports (Barbe, 1964; Lewis, 1979; Ramsell and Paul, 1990) of insects or mollusks preferentially feeding on rust-infected tissues. The same can be said generally for ovipositional preference. C. rubiginosa laid its eggs on pustule-free areas of infected leaves or rust-free leaves; T. horridus oviposited in the leaf midribs which were usually rust-free. Similarly, R. conicus laid its eggs on pustule-free portions of infected heads or healthy heads. The only exception was that more T. horridus eggs were recorded from rust-infected foliage than healthy leaves in the oviposition preference test. Apparently the rust affected only oviposition preference and not oviposition capacity since there was no difference in the continuous-exposure studies. Preference for healthy areas of the plant may be beneficial from the biological control point of view since this may mean minimal ingestion of rust by the thistle herbivores. Thus, because each of the biological control agents has its own niche, the use of all of the agents together will likely cause musk thistle populations to decline further and faster than the use of the weevils alone. These results are consistent with the data from our field tests (Baudoin et al., 1993) which showed synergism between the effects of the insect complex and rust on thistle seed production in one year, while there was no significant interaction in two other years.
TABLE 6 Effect of Continuous Exposure to Rust on Oviposition and Larval Development of Rhinocyllus conicus Means { SE Control
Rust-infected heads
t testb
Laboratorya Number of eggs/female Egg hatch (%)
44.2 { 10.5 68.3 { 5.2
38.6 { 11.4 79.9 { 4.6
ns ns
Field Prepupae (n) Weight (mg) Newly-formed pupae (n) Weight (mg) Length (mm) Width (mm)
16 31.0 { 0.1 25 23.5 { 0.9 6.2 { 0.1 3.2 { 0.04
30 28.6 { 0.8 30 25.7 { 0.6 6.7 { 0.1 3.1 { 0.03
Observations
a
ns ns ns ns
Means of 10 replicates, three pairs/replicate. ns, not significant (P ú 0.05, Student’s t test) for rust-infected versus healthy leaf. b
AP-Bio Con
P. carduorum EFFECTS ON THISTLE HERBIVORES
In conclusion, these studies suggest that the rust fungus does not interfere with the development and reproduction of the musk thistle herbivores. The rust is generally compatible with all three beetles. It has only a slight impact on the feeding behavior of C. rubiginosa adults and larvae and the adults of T. horridus, as they tended to avoid feeding on rust-infected foliage. The thistle insects attack areas on the plant not infected by the rust and thereby increase damage to the plant. ACKNOWLEDGMENTS This project was funded in part by a cooperative grant from the Foreign Disease and Weed Research Unit, USDA-ARS. We thank W. L. Bruckart, Foreign Disease-Weed Research Unit, USDA-ARS, Fort Detrick, Frederick, Maryland for the supply, advice, and technical information in handling the rust fungus.
REFERENCES Barbe, G. D. 1964. The relation of European earwig to snapdragon rust. Phytopathology 54, 369–370. Baudoin, A. B. A. M., and Kok, L. T. 1994. Rust disease introduced for biological control of musk thistle becomes widespread. In ‘‘Virginia Integrated Pest Management’’ (C. Laub and F. W. Ravlin, Eds.), pp. 29–30. Virginia Polytechnic Institute and State University, Blacksburg, VA. Baudoin, A. B. A. M., Abad, R. G., Kok, L. T., and Bruckart, W. L. 1993. Field evaluation of Puccinia carduorum for biological control of musk thistle. Biol. Control 3, 53–60. Cartwright, B. O. 1983. ‘‘Response of Carduus Thistles to Three Biological Control Agents.’’ Ph.D. Dissertation, Virginia Polytechnic Institute and State University. Blacksburg, VA. Kok, L. T. 1981a. Compatibility of Rhinocyllus conicus, Trichosirocalus horridus and 2,4-D for Carduus thistle control. In ‘‘Proceedings
/ m3565$0376
01-10-96 13:43:56
bca
129
of the V. International Symposium on Biological Control of Weeds,’’ 22–29 July 1980, Brisbane, Australia. (E. S. Delfosse, Ed.), pp. 441–445. CSIRO, Canberra, Australia. Kok, L. T. 1981b. Status of two European weevils for the biological control of Carduus thistles in the U.S.A. Acta Phytopathol. Acad. Sci. Hung. 16, 139–142. Kok, L. T. and Abad, R. G. 1994. Transmission of Puccinia carduorum by the musk thistle herbivores, Cassida rubiginosa (Coleoptera:Chrysomelidae), Trichosirocalus horridus and Rhinocyllus conicus (Coleoptera:Curculionidae). J. Entomol. Sci. 29, 186–191. Kok, L. T., and McAvoy, T. J. 1983. Refrigeration, a practical technique for storage of eggs of Trichosirocalus horridus (Coleoptera:Curculionidae). Can. Entomol. 115, 1537–1538. Kok, L. T., and Trumble, J. T. 1979. Establishment of Ceuthorrynchidius horridus (Coleoptera:Curculionidae), an imported thistlefeeding weevil, in Virginia. Environ. Entomol. 8, 221–223. Lewis, A. C. 1979. Feeding preference for diseased and wilted sunflower in the grasshopper Melanoplus differentialis. Ent. Exp. Appl. 26, 202–207. Ramsell, J., and Paul, N. D. 1990. Preferential grazing by molluscs of plants infected by rust fungi. Oikos 58, 145–150. SAS Institute Inc. 1985. ‘‘SAS User’s Guide: Statistics,’’ Version 5. SAS Institute, Cary, NC. Sieburth, P. J. 1981. ‘‘Impact of Trichosirocalus horridus (Panzer) (Coleoptera:Curculionidae) on Carduus Thistles and Factors Affecting its Mortality.’’ M. S. Thesis, Virginia Polytechnic Inst. and State Univ. Blacksburg, VA. Trumble, J. T., and Kok, L. T. 1978. Laboratory propagation of Ceuthorhynchidius horridus (Coleoptera:Curculionidae), an introduced weevil for biological control of Carduus thistles. Can. Entomol. 110, 1091–1094. Ward, R. H. 1976. ‘‘Biological studies on Cassida rubiginosa Mu¨ller, A Thistle-Feeding Shield Beetle.’’ Ph.D. Dissertation, Virginia Polytech. Inst. and State Univ., Blacksburg, VA. Ward, R. H., and Kok, L. T. 1975. Oviposition response of Ceuthorhynchidius horridus, an introduced thistle-weevil, to different photophases. Environ. Entomol. 4, 658–660. Zar, J. H. 1984. ‘‘Biostatistical Analysis.’’ Prentice–Hall, Englewood Cliffs, NJ.
AP-Bio Con
6, 123–129 (1996)
Article No. 0015
Effects of Puccinia carduorum on Musk Thistle Herbivores L. T. KOK,* R. G. ABAD,*,1
AND
A. B. A. M. BAUDOIN†
*Department of Entomology and †Department of Plant Pathology, Physiology and Weed Science, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 Received November 30, 1994; accepted April 24, 1995
Puccinia carduorum Jacky, a rust fungus recently introduced into the United States for biological control of musk thistle (Carduus thoermeri Weinmann), may interact with three established thistle herbivores, Trichosirocalus horridus (Panzer) (Coleoptera: Curculionidae), Rhinocyllus conicus Froelich (Coleoptera: Curculionidae), and Cassida rubiginosa Mu ¨ ller (Coleoptera: Chrysomelidae). Tests were conducted to determine the effects of P. carduorum on feeding, oviposition, longevity, egg production, egg hatch, and larval development of the three insects. When offered both rust-infected and healthy musk thistle leaves, adults of C. rubiginosa and T. horridus consumed significantly more healthy than infected foliage. On infected leaves, feeding and oviposition were confined largely to pustule-free areas. The amount of rust-infected and healthy foliage consumed by C. rubiginosa larvae was not significantly different, but feeding on the infected leaves was again confined to rust-free leaf areas. Rust infection did not reduce oviposition by the three insects, and R. conicus oviposited only on healthy portions of the bracts. There were no significant differences in longevity, egg production, percentage of egg hatch, and larval development for any of the three insects fed continuously with both healthy and rust-infected leaves. Thus, effects of P. carduorum on the herbivores were slight, and among the interactions that were found, none is expected to be detrimental to biological control. q 1996 Academic Press, Inc. KEY WORDS: biological control; Carduus thoermeri; musk thistle control; Cassida rubiginosa; Trichosirocalus horridus; Rhinocyllus conicus; rust fungus; Puccinia carduorum; rust effects on thistle herbivores.
INTRODUCTION
Several natural enemies of musk thistle have been introduced into the United States over the years. Compatibility of Rhinocyllus conicus (a weevil that feeds 1 Present address: Philippine Coconut Authority, Davao Research Center, P. O. Box 295, Davao City 8000, Philippines.
in the thistle heads) and Trichosirocalus horridus (a rosette-feeding weevil) has been demonstrated (Kok and Trumble, 1979; Kok, 1981 a,b). In addition, Cartwright (1983) showed that Cassida rubiginosa, a foliage feeder, also blends well with the two weevils. The rationale advanced by these workers was that each herbivore occupies a distinct niche on the plant, thus avoiding direct competition for resources. Puccinia carduorum, a rust fungus from the Mediterranean area, was introduced into the United States in 1987 as an additional biological control agent. The rust has become established in the eastern United States (Baudoin and Kok, 1994), and musk thistle seed production in rust-inoculated plots was reduced compared with that in fungicide-protected plots (Baudoin et al., 1993). Since the insect herbivores have, in general, successfully suppressed musk thistle populations, we wanted to determine whether there was any risk of the rust interfering with the effectiveness of the insects. In addition, there are several literature reports of insects (Barbe, 1964; Lewis, 1979) or mollusks (Ramsell and Paul, 1990) showing preferential feeding on rust-diseased plant tissues, which raises the concern that the insect herbivores might diminish the activity of the rust. Contact of P. carduorum with all three insect herbivores is likely, as the pathogen infects most parts of the plant. As a foliage feeder, it appears that C. rubiginosa is most exposed to the rust. Next are the adult and egg stages of T. horridus. Eggs of R. conicus which are laid on the bracts of thistle heads may be occasionally exposed to the rust, but the adults, which feed minimally on leaves, would have little contact. Larvae of both weevils (R. conicus larvae feed within thistle heads and T. horridus larvae feed in the crown tissues) have the least contact with the rust. Transmission of the rust by all three herbivores is known to occur (Kok and Abad, 1994); however, what effect the rust has on these herbivores is not known. In this study, we determined the influence of the rust on the feeding behavior, longevity, oviposition, egg hatch, and larval development of each insect.
123
/ m3565$0376
01-10-96 13:43:56
bca
1049-9644/96 $12.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AP-Bio Con
124
KOK, ABAD, AND BAUDOIN
MATERIALS AND METHODS
Musk thistle plants were grown continuously in the greenhouse. About half of these were inoculated with a P. carduorum isolate from Turkey which was provided by Dr. William Bruckart of the Foreign Diseases– Weed Science Research Unit, United States Department of Agriculture, Frederick, Maryland. The plants were inoculated by misting with urediniospore suspensions of the fungus (0.5 mg/5 ml of water with a drop of Tween 20; polyoxyethylene sorbitan monolaureate) using a compressed-air garden sprayer and held overnight in a moisture-saturated atmosphere. To increase infection levels and maintain sufficient infected leaves or plants for the experiments, inoculations were repeated as necessary. The inoculated plants were kept in a separate room from the noninoculated plants. Preference Test
C. rubiginosa Feeding. Field-collected adults and larvae (third and fourth instars) were used in this study. For both stages, insects were confined in 1-liter plastic cages (10 cages with three individuals/cage) and maintained at room temperature (307C) with 14-h photophase. Each cage was provided with two musk thistle leaves, one rust-infected and one healthy. The leaves were removed after 3 days for the larval stage and 2 days for the adult stage. Leaf area consumed was calculated using the transparent grid technique (Cartwright, 1983). Leaf-area consumption by the insects on rustinfected leaves and healthy leaves was compared. The study was repeated five times. Since it was observed that the insects tended to avoid rust pustules during feeding, this was confirmed as follows: Leaf disks (21-mm diameter, containing 5–20 pustules) were cut from rusted leaves with a cork borer and placed in 6-cm petri dishes with the adaxial side down on moist filter paper. The pustules on each disk were counted and the pattern sketched, before and after feeding. One or two C. rubiginosa adults or larvae were placed on each disk and allowed to feed at room temperature for 1 to 2 days until approximately 25– 50% of the leaf surface had been eaten. Insects were then removed, and the leaf surface eaten by the insects was estimated by placing a transparent plastic sheet with dots on a 2.5-mm grid over the leaf disks and counting dots over eaten and noneaten areas. The observed number of pustules eaten was compared with the number expected on the basis of the leaf area eaten using a x2 test with one degree of freedom. This test was conducted three times. Oviposition. Adults were collected from the field in mid- to late May and brought to the laboratory. After an acclimatization period of 1 week, male and female
/ m3565$0376
01-10-96 13:43:56
bca
pairs were confined in plastic cages in the laboratory (10 cages, three pairs/cage). Each cage was supplied with one rust-infected and one healthy leaf. The leaves were removed and replaced by new sets after 3 to 5 days. Observations were continued with up to five leaf changes. The number of oothecae and eggs laid on the infected leaves were compared with those laid on healthy leaves. This oviposition study was supplemented with field sampling of plants exposed to the rust and thistle herbivores. C. rubiginosa egg masses were counted on 62 plants in fungicide-protected, insects-only plots, and 56 rust-infected plants in plots with insects and rust in June when rust infection was increasing and leaf senescence in the rust-infected plants became evident. T. horridus Feeding. Adults collected in late June were brought into the laboratory, placed in a cold room (15 { 17C), and fed with musk thistle leaf clusters. After 1 week, the insects were sexed, paired, and placed in plastic cages (10 cages, three pairs/cage) at 247C following the protocol of Trumble and Kok (1978). As with C. rubiginosa, each cage was supplied with one healthy and one rust-infected leaf. The leaves were replaced by new sets after 1 week. The number of T. horridus feeding holes on the rust-infected leaves was compared with those on the healthy leaves. The experiment was repeated five times. Avoidance of rusted areas was confirmed as described for C. rubiginosa, except that three or four T. horridus adults per leaf disk were used. Oviposition. The insects used in the feeding-preference test were supplied with their respective diets until the completion of the oviposition period of the insect. The number of eggs oviposited on infected leaves was compared with the number laid on healthy leaves. This test was repeated five times. R. conicus Oviposition. Since adult feeding on musk thistle is minimal, the study focused on oviposition preference. Fifty healthy and 50 rust-infected heads of fairly uniform size were arbitrarily selected from thistle plants adjacent to rust-inoculated field plots in June. R. conicus eggs were counted on each head. Effects of Continuous Exposure of the Rust on Thistle Herbivores
C. rubiginosa Longevity. Adults from fields without the rust fungus were collected in June and brought to the laboratory. They were placed in large holding cages together with musk thistle leaves. After 1 week, male and female pairs were transferred to smaller (39 cm3) plastic
AP-Bio Con
P. carduorum EFFECTS ON THISTLE HERBIVORES
cages (10 cages, three pairs/cage). Insects in five of these cages were continuously supplied with rust-infected leaves while the other five had healthy leaves (two leaves per cage). The insects were maintained in a growth chamber (50% RH, 14-h photophase with light:dark [L:D] temperatures of 30:257C). Food leaves in the rust group were misted with urediniospore suspensions of the fungus (at least 2 ml/cage from a preparation of 25 mg spores/500 ml of water) whenever necessary to ensure that insects were continuously exposed to fungus propagules even if they avoided feeding on rust pustules themselves. The cultures were maintained until the last test insect died. Mortality and date of deaths were recorded for each group. Feeding. Adults collected in spring were brought to the laboratory. The same procedure was followed as in the longevity experiment except that the number of cages was increased to 20 (10 cages/treatment). Leafarea consumption by the beetles in the rust and control groups was calculated using the transparent grid technique (Cartwright, 1983). Oviposition, egg hatch, and larval development. Adults collected in May and brought to the laboratory were subjected to the same procedure as in the feeding test. From the start to the end of an oviposition period of the insect, oothecae from each batch of leaves in a cage were transferred every 4 or 5 days to plastic cups (11.5 cm3) lined at the bottom with sponge and covered with lids. The oothecae were provided with freshly cut musk thistle leaves every 3 days (for 2 weeks) to provide humidity and food as the eggs hatched (Ward, 1976). The number of oothecae and eggs and the percentage of hatch were recorded for each treatment. Two successive batches of larvae in each treatment were reared to last instar (15 days after eclosion) using their respective diets. The prepupae from each treatment were weighed and then measured under a calibrated dissecting microscope.
125
maintained in this manner until all the insects in each cage died. The feeding holes in each batch of leaves were counted. Likewise, the number and date of adult mortality were recorded. Oviposition and egg hatch. Two experiments were conducted. The first experiment utilized the insects in the preceding longevity and feeding test. During the oviposition period of the insect, the eggs from every batch of leaves in each cage were collected, counted, and placed in petri dishes with moist filter paper lining the bottom. These petri dishes were stored in a refrigerator at 47C and the filter paper was kept moist (Kok and McAvoy, 1983). After several days, the petri dishes were exposed to higher temperatures (15 { 17C) to induce the eggs to hatch. After 12 to 15 days, the percentage of hatch was recorded and the means for the two groups were compared. The second experiment involved insects collected in June of the following year; procedures were the same. Larval development. Fifteen healthy and 15 rustinfected potted musk thistle plants in their rosette stage were moved from the greenhouse to a cold room (15 { 17C, 14-h photophase). Six laboratory-reared first instars (õ48-h old) were introduced onto the leaves and growth point of each plant by means of a capillary tube sealed at one end to prevent escape of introduced larvae (Sieburth, 1981). Small holes were pierced by means of a needle (or the sharp end of forceps) on the leaves or growth points where the larvae were introduced. In addition, to ensure immediate contact of rust with the larvae, the holes were misted with suspensions of rust spores scraped from infected leaves (at least six infected leaves/100 ml of water). After 28 days, larvae were removed from each plant to be weighed and measured (under a calibrated dissecting microscope). Differences in these data between the two groups were analyzed. If the original host died before the end of the 28-day period, the test larvae were transferred to a new potted plant (rust-infected or healthy).
T. horridus Longevity and feeding. The experimental procedure was similar to that for C. rubiginosa. Adults were collected in July and brought to the laboratory. The insects were initially maintained under room conditions (25–307C) and fed with healthy musk thistle leaves. The following month, the insects were sexed and paired. The pairs were placed in plastic cages (10 cages, five pairs/cage) inside a growth chamber (50% RH, 9-h photophase and L:D temperatures of 20:107C) following the protocol of Trumble and Kok (1978). The short photophase was used because it induces oviposition of the weevil (Ward and Kok, 1975). Five of these cages were supplied with two rust-infected leaves and the other five with two healthy leaves per cage. The food leaves were replaced after 7–10 days. The cultures were
/ m3565$0376
01-10-96 13:43:56
bca
R. conicus Longevity. This study was conducted in two consecutive years. Potted healthy and rust-inoculated musk thistle plants in the stem-elongation stage were used. In the first year, R. conicus adults were collected in May and brought to the laboratory. The following week, 10 adults (five pairs) were introduced to each plant. For each potted plant, the stem and leaves (with the insects) above the soil line were enclosed in a Mylar cage. The cages were maintained under greenhouse conditions (25–307C, 14-h photophase), until all the insects died. Plants that died during the period of study were replaced. The number and date of mortality of test insects in both treatments were recorded. This study was replicated 10 times.
AP-Bio Con
126
KOK, ABAD, AND BAUDOIN
TABLE 1 Preference of Feeding and Oviposition by Cassida rubiginosa on Rust-Infected and Healthy Thistle Leaves Means { SE
Observations Feedinga Larva (cm2 leaf/larva) Adult (cm2 leaf/adult) Degree of leaf infectiond Leaf size (cm): Length Greatest width Oviposition Laboratoryb Number of oothecae Number of eggs Degree of leaf infectiond Fieldc Number of egg masses/plant Degree of leaf infectiond
Healthy
Rustinfected
t teste
9.8 { 1.2 14.0 { 0.8 0
7.8 { 1.2 9.0 { 0.7 3.0 { 0.1
ns * —
4.2 { 0.2 2.4 { 0.2
4.0 { 0.3 2.2 { 0.1
ns ns
45.4 { 9.1 233.2 { 41.1 0
36.8 { 7.1 200.0 { 44.0 2.8 { 0.3
ns ns —
1.2 { 0.6 1.8 { 0.2
ns —
Analysis of Data
The SAS statistical program (SAS Institute, 1985) was used for data analysis. Differences in insect feeding, oviposition, percentage egg hatch, and growth of larvae between the rust-infected and healthy leaves or plants were analyzed using Student’s t test (Zar, 1984). RESULTS Preference Test
2.1 { 0.9 0.02 { 0.03
a
Means of five tests, each of 10 replicates/treatment, three individuals/replicate. b Means of five oviposition periods, 10 replicates/treatment, three pairs/replicate. c 68 egg masses from 56 plants in the rust-infected group; 130 egg masses from 62 plants in the healthy plants group. d Rating scale: estimated number of rust pustules (0–5) per 6.25 cm2 surface area. e Significance: *significant (P õ 0.05); ns, not significant (P ú 0.05), Student’s t test for infected versus healthy leaf; —, not determined.
In the second year, the procedures were the same except that: (a) adults were collected in late April, (b) the cages were maintained in the laboratory (22–277C, 14-h photophase), (c) the adult insects in the rust group were initially submerged in a dish of spore suspensions of P. carduorum for 2 min (plain water in the case of the control) before introduction to their respective host plants, and (d) the test plants were replaced every 2 weeks or earlier when necessary. Oviposition. This aspect of the study was carried out by counting eggs every 2 weeks during change of test plants. The eggs from each test plant were collected and recorded as hatched or unhatched. The unhatched eggs were placed in a growth chamber (247C, 80% RH) and observed daily for hatch. Development of larvae. This field study was conducted using thistle stands adjacent to rust-inoculated plants in June. Fifteen heavily rust-infected musk thistle heads and 15 uninfected heads of fairly uniform size and age with R. conicus eggs were enclosed individually in paper bags. The bags were used to prevent contact with fresh rust-inoculum. Five weeks later, the bags were removed and the heads were harvested and
/ m3565$0376
brought to the laboratory. The thistle heads were dissected and the fourth instars or pupae were weighed. Newly formed pupae were obtained by dissections of additional uninfected and rust-infected heads bearing R. conicus eggs. The pupae were weighed and measured using a calibrated dissecting microscope. Weight and measurements of the pupae in the two categories (rust and control) were compared.
01-10-96 13:43:56
bca
C. rubiginosa Feeding. The areas of healthy leaves and infected leaves consumed by C. rubiginosa larvae were not different (Table 1). For adults, however, consumption was greater on healthy than on rust-infected leaves (P õ 0.05) (Table 1). Even on infected leaves, the insect preferred pustule-free surfaces of the leaf (Table 2). The results suggest that adults, which are more mobile, are better able to select and feed on an uninfected leaf while larvae merely avoid infected leaf areas. Oviposition. No differences were found for C. rubiginosa oothecae or eggs gathered from healthy leaves and rust-infected leaves (Table 1). As in the feeding test, the eggs were observed on pustule-free surface areas of infected leaves. Similar results were observed on plants in field plots. T. horridus Feeding. There were more T. horridus feeding holes recorded on healthy than rust-infected leaves (Table TABLE 2 Avoidance of Rust-Infected Leaf Areas of Musk Thistle during Feeding by Cassida rubiginosa and Trichosirocalus horridus Number of rust pustules
Insect
Stage
C. rubiginosa Adults C. rubiginosa Larvae T. horridus Adults a
Total 63 133 91
Observed Expecteda Probability of eaten eaten greater x2 4 6 1
Based on percentage of leaf area eaten.
AP-Bio Con
21 70 19
0.00001 õ0.00001 0.008
127
P. carduorum EFFECTS ON THISTLE HERBIVORES
TABLE 3 Preference of Feeding and Oviposition of Trichosirocalus horridus Adults on Rust-Infected and Healthy Thistle Leaves Means { SE Observations
Healthy
Feedinga Number of feeding holes/insect 14.9 { 6.3 Degree of leaf infectionb 0 Leaf size (cm): Length 4.1 { 0.1 Biggest width 2.4 { 0.1 Ovipositionc Number of eggs/female Degree of leaf infectionb
30.1 { 5.2 0
Rust-infected t testd
9.4 { 5.0 2.5 { 0.8
* —
4.0 { 0.1 2.3 { 0.1
ns ns
69.9 { 5.2 2.4 { 0.2
* —
a
Means of five feeding periods, 10 replicates/treatment, three pairs/replicate. b Rating scale: number of rust pustules (0–5) per 6.25 cm2 surface area. c Means of 10 replicates/treatment, three pairs/replicate. d Significance: *significant (P õ 0.05); ns, not significant (P ú 0.05), Student’s t test for rust-infected versus healthy leaves; —, not determined.
3). As with C. rubiginosa, feeding was largely confined to pustule-free areas of diseased leaves (Table 2). Oviposition. —Rust infection did not deter oviposition on rust-infected leaves, as there were significantly more eggs collected from diseased than healthy leaves (Table 3).
the rust-exposed group compared with only one pair in the control. Oviposition, egg hatch, and larval development. There were no differences in the numbers of oothecae and eggs oviposited, and percentage of egg hatch by insects continuously fed with healthy leaves and those fed with infected leaves (Table 4). No differences were noted in the sizes and weights of larvae that were continuously fed with infected leaves and those that were reared on healthy leaf diets, nor was there any difference in adult emergence. As previously observed, the larvae fed on pustule-free leaf areas of infected leaves. T. horridus Longevity and feeding. Adults continuously fed on rust-infected leaves and those fed on healthy leaves both survived for more than a year after field collection in July. Mortalities were 4.6 and 4.5 adults per month, respectively, for the rust and the control groups. After 14 months, there were two surviving adults in the former and three in the latter group. During the same period, the number of T. horridus feeding holes was not different between the two groups (Table 5). Oviposition, egg hatch, and larval development. There was no difference in the number of eggs ovipos-
TABLE 4 Effect of Continuous Exposure to Rust on Adult Feeding, Oviposition, and Larval Development of Cassida rubiginosa Means { SE
R. conicus Oviposition —No difference was noted in the mean number of R. conicus eggs collected from both rustinfected heads (43.2 { 3.2) and healthy heads (45.3 { 2.9). It was evident, however, that eggs were laid on healthy portions of the bracts. Effects of Continuous Exposure of the Rust on Thistle Herbivores
C. rubiginosa Longevity. Insects reared on both diets (healthy leaves and rust-infected leaves) survived for over 1 year, indicating that continuous exposure to the fungus did not affect the longevity of the insect. In fact, the last surviving female in the rust group still laid five egg masses after a year. Feeding. The leaf area consumed by C. rubiginosa continuously fed with infected leaves did not differ from the control (Table 4). As in earlier findings, feeding was confined to pustule-free surfaces of infected leaves. At the seventh month, three pairs of adults survived in
/ m3565$0376
01-10-96 13:43:56
bca
Observations
Control
Feedinga Leaf area consumed (sq cm/ insect) 60.7 { 7 Degree of leaf infectionb 0 Leaf size (cm) Length 4.1 { 0.2 Biggest width 2.1 { 0.1 Ovipositionc Number of oothecae/female 27.5 { 2.6 Number of eggs/female 167.0 { 17.4 Egg hatch (%) 80.0 { 2.3 Measurement of prepupaed Weight (mg) 18.6 { 0.6 Length (mm) 6.6 { 0.1 Width (mm) 3.2 { 0.04 Pupae emerging as adults (%) 91.0 a
Rust-infected t teste
52.2 { 4.0 1.9 { 0.5
ns —
4.1 { 0.1 2.2 { 0.1
ns ns
29.6 { 3.0 141.9 { 8.6 81.3 { 1.5
ns ns ns
17.8 { 0.6 6.3 { 0.1 3.2 { 0.04 88.0
ns ns ns ns
Means of 10 replicates, three pairs/replicate. Rating scale: number of rust pustules (0–5) per 6.25 cm2 surface area. c Means of seven oviposition periods (10 replicates/treatment, 3 pairs/replicate). d Means of 30 insects/treatment. e Significance: ns, not significant (P ú 0.05, Student’s t test) for rust-infected versus healthy leaf; —, not determined. b
AP-Bio Con
128
KOK, ABAD, AND BAUDOIN
TABLE 5 Effect of Continuous Exposure to Rust on Adult Feeding, Oviposition, and Larval Development of Trichosirocalus horridus Means { SE Observations
Control
Rust-infected
t testd
Feedinga Number of feeding holes/adult
400.7 { 58.0
431.1 { 48.6
ns
141.3 { 33.9
156.4 { 45.1
ns
79.4 { 3.8 76.9 { 3.4
74.5 { 3.4 77.2 { 2.8
ns ns
Oviposition Number of eggs/femaleb Egg hatch (%) 1st groupa 2nd groupb Measurement of larvaec Weight (mg) Length (mm) Width (mm)
9.5 { 0.4 5.0 { 0.1 2.0 { 0.03
9.1 { 0.3 4.9 { 0.1 2.0 { 0.04
ns ns ns
a
Means of five replicates, five pairs/replicate. Means of 10 replicates, three pairs/replicate. c Means of 23 larvae/treatment (28-day old larvae). d ns, not significant (P ú 0.05, Student’s t test) for rust-infected versus healthy leaf. b
ited by T. horridus that were continuously exposed to infected leaves and those raised on healthy foliage. Percentage of hatch of eggs from the two groups of insects was also not different. After 28 days, weights of larvae (offspring of parents which were also continuously fed rust-infected leaves) given a continuous diet of infected leaves did not differ from those fed with healthy leaves (Table 5). R. conicus Longevity. Mortalities of overwintered insects exposed to rust and those maintained on healthy musk thistle plants did not differ for the two groups in either year, indicating no effect of the presence of the fungus on the insect. Both groups had about 40% mortality after 3 weeks and 90% mortality after 5 weeks. This is not unusual as adults that emerge in spring die soon after they complete oviposition within 4–5 weeks. Oviposition, egghatch, andlarval development. Continuous exposure to the rust did not affect egg production and percentage of hatch of R. conicus. Also, the larval and pupal weights and sizes of R. conicus feeding on infected heads did not differ from those that were obtained from noninfected heads (Table 6). DISCUSSION
The overall results of this study indicated that exposure of the three herbivores to the rust had no significant effect on their life span, feeding, oviposition, egg hatch, and larval development. Exposure of the con-
/ m3565$0376
01-10-96 13:43:56
bca
fined insects and their various stages to the rust in most cases were abnormally continuous and long (6 months to over a year). Under natural conditions, the insects may be exposed to the rust for only a short period as they have freedom to feed and oviposit on musk thistle plants of their choice. There was a clear preference of C. rubiginosa adults and larvae, and T. horridus adults, for feeding on healthy leaves or pustule-free areas of infected leaves, in contrast to reports (Barbe, 1964; Lewis, 1979; Ramsell and Paul, 1990) of insects or mollusks preferentially feeding on rust-infected tissues. The same can be said generally for ovipositional preference. C. rubiginosa laid its eggs on pustule-free areas of infected leaves or rust-free leaves; T. horridus oviposited in the leaf midribs which were usually rust-free. Similarly, R. conicus laid its eggs on pustule-free portions of infected heads or healthy heads. The only exception was that more T. horridus eggs were recorded from rust-infected foliage than healthy leaves in the oviposition preference test. Apparently the rust affected only oviposition preference and not oviposition capacity since there was no difference in the continuous-exposure studies. Preference for healthy areas of the plant may be beneficial from the biological control point of view since this may mean minimal ingestion of rust by the thistle herbivores. Thus, because each of the biological control agents has its own niche, the use of all of the agents together will likely cause musk thistle populations to decline further and faster than the use of the weevils alone. These results are consistent with the data from our field tests (Baudoin et al., 1993) which showed synergism between the effects of the insect complex and rust on thistle seed production in one year, while there was no significant interaction in two other years.
TABLE 6 Effect of Continuous Exposure to Rust on Oviposition and Larval Development of Rhinocyllus conicus Means { SE Control
Rust-infected heads
t testb
Laboratorya Number of eggs/female Egg hatch (%)
44.2 { 10.5 68.3 { 5.2
38.6 { 11.4 79.9 { 4.6
ns ns
Field Prepupae (n) Weight (mg) Newly-formed pupae (n) Weight (mg) Length (mm) Width (mm)
16 31.0 { 0.1 25 23.5 { 0.9 6.2 { 0.1 3.2 { 0.04
30 28.6 { 0.8 30 25.7 { 0.6 6.7 { 0.1 3.1 { 0.03
Observations
a
ns ns ns ns
Means of 10 replicates, three pairs/replicate. ns, not significant (P ú 0.05, Student’s t test) for rust-infected versus healthy leaf. b
AP-Bio Con
P. carduorum EFFECTS ON THISTLE HERBIVORES
In conclusion, these studies suggest that the rust fungus does not interfere with the development and reproduction of the musk thistle herbivores. The rust is generally compatible with all three beetles. It has only a slight impact on the feeding behavior of C. rubiginosa adults and larvae and the adults of T. horridus, as they tended to avoid feeding on rust-infected foliage. The thistle insects attack areas on the plant not infected by the rust and thereby increase damage to the plant. ACKNOWLEDGMENTS This project was funded in part by a cooperative grant from the Foreign Disease and Weed Research Unit, USDA-ARS. We thank W. L. Bruckart, Foreign Disease-Weed Research Unit, USDA-ARS, Fort Detrick, Frederick, Maryland for the supply, advice, and technical information in handling the rust fungus.
REFERENCES Barbe, G. D. 1964. The relation of European earwig to snapdragon rust. Phytopathology 54, 369–370. Baudoin, A. B. A. M., and Kok, L. T. 1994. Rust disease introduced for biological control of musk thistle becomes widespread. In ‘‘Virginia Integrated Pest Management’’ (C. Laub and F. W. Ravlin, Eds.), pp. 29–30. Virginia Polytechnic Institute and State University, Blacksburg, VA. Baudoin, A. B. A. M., Abad, R. G., Kok, L. T., and Bruckart, W. L. 1993. Field evaluation of Puccinia carduorum for biological control of musk thistle. Biol. Control 3, 53–60. Cartwright, B. O. 1983. ‘‘Response of Carduus Thistles to Three Biological Control Agents.’’ Ph.D. Dissertation, Virginia Polytechnic Institute and State University. Blacksburg, VA. Kok, L. T. 1981a. Compatibility of Rhinocyllus conicus, Trichosirocalus horridus and 2,4-D for Carduus thistle control. In ‘‘Proceedings
/ m3565$0376
01-10-96 13:43:56
bca
129
of the V. International Symposium on Biological Control of Weeds,’’ 22–29 July 1980, Brisbane, Australia. (E. S. Delfosse, Ed.), pp. 441–445. CSIRO, Canberra, Australia. Kok, L. T. 1981b. Status of two European weevils for the biological control of Carduus thistles in the U.S.A. Acta Phytopathol. Acad. Sci. Hung. 16, 139–142. Kok, L. T. and Abad, R. G. 1994. Transmission of Puccinia carduorum by the musk thistle herbivores, Cassida rubiginosa (Coleoptera:Chrysomelidae), Trichosirocalus horridus and Rhinocyllus conicus (Coleoptera:Curculionidae). J. Entomol. Sci. 29, 186–191. Kok, L. T., and McAvoy, T. J. 1983. Refrigeration, a practical technique for storage of eggs of Trichosirocalus horridus (Coleoptera:Curculionidae). Can. Entomol. 115, 1537–1538. Kok, L. T., and Trumble, J. T. 1979. Establishment of Ceuthorrynchidius horridus (Coleoptera:Curculionidae), an imported thistlefeeding weevil, in Virginia. Environ. Entomol. 8, 221–223. Lewis, A. C. 1979. Feeding preference for diseased and wilted sunflower in the grasshopper Melanoplus differentialis. Ent. Exp. Appl. 26, 202–207. Ramsell, J., and Paul, N. D. 1990. Preferential grazing by molluscs of plants infected by rust fungi. Oikos 58, 145–150. SAS Institute Inc. 1985. ‘‘SAS User’s Guide: Statistics,’’ Version 5. SAS Institute, Cary, NC. Sieburth, P. J. 1981. ‘‘Impact of Trichosirocalus horridus (Panzer) (Coleoptera:Curculionidae) on Carduus Thistles and Factors Affecting its Mortality.’’ M. S. Thesis, Virginia Polytechnic Inst. and State Univ. Blacksburg, VA. Trumble, J. T., and Kok, L. T. 1978. Laboratory propagation of Ceuthorhynchidius horridus (Coleoptera:Curculionidae), an introduced weevil for biological control of Carduus thistles. Can. Entomol. 110, 1091–1094. Ward, R. H. 1976. ‘‘Biological studies on Cassida rubiginosa Mu¨ller, A Thistle-Feeding Shield Beetle.’’ Ph.D. Dissertation, Virginia Polytech. Inst. and State Univ., Blacksburg, VA. Ward, R. H., and Kok, L. T. 1975. Oviposition response of Ceuthorhynchidius horridus, an introduced thistle-weevil, to different photophases. Environ. Entomol. 4, 658–660. Zar, J. H. 1984. ‘‘Biostatistical Analysis.’’ Prentice–Hall, Englewood Cliffs, NJ.
AP-Bio Con