OUP user menu

Influences of Bacillus thuringiensis Berliner Cotton Planting on Population Dynamics of the Cotton Aphid, Aphis gossypii Glover, in Northern China

(CC)
Kongming Wu, Yuyuan Guo
DOI: http://dx.doi.org/10.1603/0046-225X-32.2.312 312-318 First published online: 1 April 2003

Abstract

The influence of Bacillus thuringiensis Berliner (Bt) cotton on population dynamics of cotton aphid, Aphis gossypii Glover, was investigated during 1999–2000 in northern China. The field experiments were conducted in plots of Bt cotton and conventional cotton that received no insecticide applications, and in plots of conventional cotton in which pyrethroid and organophosphate insecticides were used regularly for control of Helicoverpa armigera. The results indicate that resistance of cotton aphids to majority of insecticides used for control of H. armigera, and lower densities of predators in late June and early July caused by insecticide use, causes population densities of cotton aphids to become significantly higher in plots of insecticide-treated conventional cotton than in Bt cotton plots. These results suggest that Bt cotton planting not only played an important role in the control of H. armigera, but also efficiently prevented cotton aphid resurgence in response to insecticide use.

  • Bt cotton
  • Aphis gossypii
  • population dynamics
  • resurgence

Introduction

Helicoverpa armigera (HüBNER) IS a key pest of cotton, Gossypium hirsutum L., in China, and growers routinely use chemical insecticides for its control. Since 1990, resistance to chemical insecticides has been a serious problem, and is a major threat to cotton production in China (Wu et al. 1997).

Management of pest by use of genetically based host-plant resistance typically has significant environmental and economic advantages compared with use of chemical insecticide. Natural resistance of cotton to insects has been used in the development of germplasm lines and cultivars. However, no single germplasm line or cultivar released to date exhibits a level of resistance high enough to preclude the use of insecticides, especially under high insect infestations (Guo 1997). The availability of genetically modified cotton plants that carry genes for insect resistance from other organisms may increase the chance of developing germplasm lines with higher levels of resistance to insects (Jenkins et al. 1993, Halcomb et al. 1996, Adamczyk et al. 1998). Transgenic cotton, engineered to continuously express a δ-endotoxin from Bacillus thuringiensis Berliner (Bt), holds great promise in controlling Lepidopteran species (Wilson et al. 1992, Jenkins et al. 1993, Guo 1995, Flint et al. 1996, Halcomb et al. 1996). Bt cotton began to be commercially planted in northern China in 1997 ( ≈ 10,000 ha) and expanded rapidly to 1.067 million ha. in 2000 (≈90% cotton planting area in the region) (Wang et al. 1997, Wu and Guo 2000b, Qu et al. 2001). As a result of Bt cotton deployment, the amount of insecticide use has been drastically decreased, with the average number of pesticide applications per season declining from ≈20–7 in Hebei and Shandong Provinces (Pray et al. 2001).

Because nonlepidopterous pests are also involved in a comprehensive insect pest control strategy in cotton, it is necessary to understand field abundance of insect predators or parasitoids and other nontarget insect pests on transgenic Bt plants (Wilson et al. 1992, Rig-gin-Bucci and Gould 1997, Pilcher et al. 1997). Although the Bt protein is directly toxic to only a narrow spectrum of Lepidopteran species, the dynamics of other species may be indirectly affected. Effects on nontarget predatory and parasitic species may be positive because of the removal of disruptive pesticides, or negative because of the effective removal of lepidopteran prey.

Cotton aphid, Aphis gossypii Glover, is an important insect pest of cotton in northern China (Wu and Liu 1992). Insecticide use for H. armigera control disrupts aphid natural enemies is responsible for aphid resurgence in mid season (Wu and Liu 1992). In this paper, we report research results on the seasonal population dynamics of cotton aphid and its relationship with predators in Bt cotton and non-Bt cotton field plots in northern China.

Materials and Methods

Cotton Lines

A transgenic Bt cotton variety expressing the Cry1A gene (GK12) developed by the Biotechnology Research Center, Chinese Academy of Agricultural Sciences (CAAS; Beijing, China), and its conventional parental line without the Cry1A gene (Shimian3), were supplied by the Biotechnology Research Center, CAAS (Wu and Guo 2000b).

Field Experimental Design

Field experiments were conducted in 1999 and 2000 at the Langfang Experimental Station of CAAS, located in Hebei Province. Experiments consisted of four treatments, i.e., noninsecticide application treatments of GK12 and Shimian3, and two insecticide application plots of conventional cotton (Shimian3) treated with pyrethroid and organophosphorus (OP) insecticides, respectively. A randomized complete block design was used with three replicates of each treatment. Each plot within a block was ≈0.033 ha and was seeded at a rate expected to produce 45,000 plants per planted ha. A 3-m space was left between plots and among blocks to decrease insect dispersion among treatments. Cotton was maintained with standard agronomic practices used in northern China (Wu and Guo 2000b).

In chemical application treatments, H. armigera was controlled according to the thresholds described by Guo et al. (1985), which are defined as 350 eggs or 35 larvae per 100 plants for the second generation, and 80 eggs or 8 larvae per 100 plants for the third generation, respectively. Insecticides were applied with a spray nozzle held 0.3– 0.5 m above cotton plants. The sprayer type was a “Gongnong-18” knapsack sprayer with a tank capacity of 16 liters and a spray lance length of 0.8 m (Handan Sprayers Ltd., Handan, China). Spray volumes were 600 liters per ha before 10 July and 900 liters per ha after 15 July. Insecticide application details in 1999 and 2000 are shown in Table 1 and Table 2, respectively.

View this table:
Table 1.
View this table:
Table 2.

Sampling of Cotton Aphid and Predators

Sampling was undertaken once every 3–4 d from mid-June to early-September. Five sites were randomly chosen from every plot on each occasion with a total number of 100 cotton plants sampled per plot. For each sampled plant, a thorough whole-plant survey was conducted in the field to count the number and species of predators, which include lady beetles (Coccinella septempunctata Linnaeus, Leis axyridis (Pallas), and Propylaea japonica (Thunberg)), lacewing (Chrysopa sinica Tjeder, Chrysopa septempunctata Wesmael, Chrysopa shansiensis Kawa, and Chrysopa formosa Brauer), spiders (Erigonidium graminicolum (Sunde-vall) and Misumenopos tricuspidata (Fabricius)), and Orius similis Zheng. Because of its high density, cotton aphids were counted only on three leaves from upper, middle, and lower parts of main stem, respectively (Wu and Liu 1992).

Bioassay for Cotton Aphid Resistance to Insecticides

Cotton aphids were collected from the trial fields in Langfang in early July 1999 and were cultured with cotton seedlings in an environmental room held at 25°C, 70% RH, and 14:10 (L:D) photoperiod. A strain of A. gossypii, which had been maintained for several years in the laboratory, was used as the susceptible control.

A total of 13 insecticide formulations, which were major pesticides for control of H. armigerain northern China, were used in the contact toxicity test (Table 3). The commercial formulations were diluted with water according to the ratio of 1:100 to obtain the initial concentration dosage, from which further dilutions were made in a logarithmic series. Leaves from field collections containing 20–30 adult aphids per leaf were dipped in pesticide solutions or suspensions. Five concentrations of each insecticide were tested with three replicate leaves per concentration Three leaves treated with water were used as a control. Once treated leaves had dried, treated aphids were transferred to untreated leaves within small plastic containers in a constant environment room maintained under 25°C, 70% RH, and 14:10 (L:D) photoperiod. Mortality was assessed after 24 h, and LC50 values and 95% CL were calculated using Probit analysis (Finny 1971). Resistance factors were calculated as LC50 of a field clone divided by LC50 of the laboratory control clone.

View this table:
Table 3.

Statistical Analyses

Density of cotton aphids in each plot was estimated from 100 cotton plants and for each plant it was defined as the number of three leaves sampled from upper, middle, and lower parts of main stem, respectively. All data on population densities of insects from different treatments in the field were analyzed using analysis of variance (ANOVA) and means were separated using the protected least significant difference (LSD) test (SAS Institute 1988).

Results

Population Dynamics of Cotton Aphids and Predators

Population dynamics of cotton aphids and predators in four treatments in 1999 are presented in Fig. 1 and Fig. 2. Insecticides were used to control H. armigera on 24 June, 29 June, 2 July, and 7 July, respectively. There were no significant differences in cotton aphid densities among the four treatments before the first spray of insecticide to control H. armigera on 24 June (P > 0.05). After the second spray, the population density of cotton aphid in pyrethroid plots was significantly higher than those in other treatments (df =3, 6; F =4.93; P < 0.05). Aphid densities in the organophosphorus plots reached 14,042 aphids per 100 plants on 20 July, which was significantly lower than the 35,400 aphids per 100 plants in pyrethroid plots, but much higher than the density of aphids in Bt cotton and nonspray plots of conventional cotton (389 and 267 aphids per 100 plants, respectively) (df = 3, 6; F = 377.03; P < 0.01). However, after imidacloprid and monocrotophos were sprayed, respectively, in the pyrethroid and OP plots on 23 July, population densities of cotton aphids in insecticide application plots decreased drastically to <100 aphids per 100 plants toward late season.

Fig. 1.

Population dynamics of cotton aphid in different cotton fields, 1999, Hebei Province. Density of cotton aphid represents number estimated from 100 cotton plants, for each plant only calculated aphid number of three leaves sampled from upper, middle, and lower parts of main stem, respectively. Because of great differences in the values from different treatments, they were transformed with logarithm. Means from the same evaluation date followed by the same letter were not statistically different (P > 0.05; LSD test). Evaluation dates with means not followed by a letter were not significantly different from each other.

Fig. 2.

Population dynamics of predators in different cotton fields, 1999, Hebei Province. Means from the same evaluation date followed by the same letter were not statistically different (P > 0.05; LSD test). Evaluation dates with means not followed by a letter were not significantly different from each other.

There were no significant differences in predator densities among different treatments from 16 June to 24 June (P > 0.05). After the first insecticide application, predator numbers on 28 June in insecticide application treatments were significantly lower than in the Bt cotton and nonspray plots of conventional cotton (df = 3, 6; F = 4.91, P <0.05). After the third spray on 2 July, the population densities in the pyrethroid and OP treatments decreased to 0 and 2 per 100 plants, respectively These densities were significantly lower than those in Bt cotton and nonspray plots of conventional cotton (51.33 per 100 plants) (df = 3, 6; F = 11.66, P < 0.01). After the final spray on 7 July, predator densities increased rapidly in the OP and pyrethroid plots as a result of high densities of cotton aphids. By 20 July, predator density increased to 206.33 per 100 plants in the OP treatment, which was much higher than the density in the pyrethroid plots (31.30 predators per 100 plants), and the densities in Bt cotton and nonsprayed plots of conventional cotton (45.33 and 33.00 predators per 100 plants) (df = 3, 6; F = 20.18, P < 0.01).

Population dynamics of cotton aphids and predators in four treatments in 2000 are presented in Fig. 3 and Fig. 4. Insecticides were used to control H. armigera on 20 June, 23 June, 29 June, 3 July, 15 July, 19 July, 23 July, and 27 July, respectively. There were no significant differences among four treatments before the first spray of insecticide to control H. armigera on 20 June (df = 3, 6; F = 3.67; P > 0.05). After the second insecticide application on 27 June, the population density of cotton aphids in the pyrethroid plots was much higher than that in OP plots, and both were significantly higher than those in other treatments (df = 3, 6; F = 94.52; P < 0.01). By 26 July, population densities of cotton aphids in different treatments were 625 (GK12), 1508 (CK), 188,600 (pyrethroid), and 108,800 (OP) aphids per 100 plants, respectively. After late July, population densities of cotton aphids in both chemical treatments decreased gradually, but were still significantly higher than those in Bt cotton and nonsprayed conventional cotton plots until 29 August (P < 0.05).

Fig. 3.

Population dynamics of cotton aphid in different cotton fields, 2000, Hebei Province. Density of cotton aphid represents number estimated from 100 cotton plants, for each plant only calculated aphid number of three leaves sampled from upper, middle, and lower parts of main stem, respectively. Because of the great differences in values from different treatments, they were transformed with logarithm. Means from the same evaluation date followed by the same letter were not statistically different (P > 0.05; LSD test). Evaluation dates with means not followed by a letter were not significantly different from each other.

Fig. 4.

Population dynamics of predators in different cotton fields, 2000, Hebei Province. Means from the same evaluation date followed by the same letter were not statistically different (P > 0.05; LSD test). Evaluation dates with means not followed by a letter were not significantly different from each other.

There were no significant differences between predator densities in different treatments on 19 June (df = 3, 6; F = 0.67, P > 0.05). However, after the second pesticide application on 27 July, predator numbers in insecticide application treatments were significantly lower than those in Bt cotton and nonsprayed plots of conventional cotton (df = 3, 6; F = 4.91, P < 0.05). After the third spray, the population densities in the pyrethroid and OP treatments on 1 July decreased to 2.00 and 3.33 predators per 100 plants, respectively, which is significantly lower than the densities in Bt cotton and nonsprayed plots of conventional cotton (55.33 and 57.00 per 100 plants, respectively) (df = 3, 6; F = 121.75, P < 0.01). However, as population densities of cotton aphids in chemically treated plots increased rapidly, the predator numbers showed a tendency to increase although application of insecticides still killed most predators. By 17 August, population densities of predators in pyrethroid and OP plots increased to 195 and 220 predators per 100 plants, which was significantly higher than the densities in Bt cotton and nonsprayed plots of conventional cotton (df = 3, 6; F = 57.73; P < 0.01).

Bioassay for Cotton Aphid Resistance to Insecticide.

Bioassay results are presented in Table 3. Compared with the laboratory clone, pyrethroid resistance in the field strain was 88-fold to cyhalothrin, 153.18-fold to deltamethrin, 304.44-fold to fenvalerate, 389.75-fold to esfenvalerate, and 517.70-fold to beta-cypermethrin, respectively. Resistance levels to organophosphates were 3.85 to dichlorvos, 5.41 to monocrotophos, 10.44 to chlorpyrifos, and 17.44 to phoxim. Resistance to a carbamate insecticide, methomyl, was assessed at 3.45-fold the tolerance of the control clone. These toxicity data indicated that cotton aphids possessed different resistance to the major insecticides used for control of H. armigera.

Discussion

In China, the period during which the cotton aphid causes yield loss to cotton is restricted to the seedling stage of cotton plants. Before the 1970s, these aphids could easily be controlled by seed treatment with insecticide. In the mid 1970s, aphids became an important insect pest of cotton because of insecticide-induced resurgence during mid and late season. Since the 1980s, its damage to cotton has become more serious and frequent (Wu and Guo 2000a). It is reported that destruction of natural enemies in cotton system from insecticide use is a key factor for resurgence of the pest (Barlett 1968). Wu and Liu (1992) conducted a field investigation of the relationship between the population dynamics of cotton aphids, natural enemies, and insecticide application for control of H. armigera in early season of cotton growth. They found that insecticide sprays against H. armigera could kill most natural enemies, such as the lady beetle and lacewing that are major predators of cotton aphids in cotton fields. However, most cotton aphids could survive after spraying because of their high resistance to OP and pyrethroid insecticides. It was concluded that resurgence of cotton aphid in mid and late season in China was related to insecticide spray from control of H. armigera and insecticide resistance in the pest (Wu and Liu 1992).

In this study, our bioassay results demonstrate that resistance in the field strain of aphid to pyrethroid and OP insecticides was 88-fold to cyhalothrin, 517.70-fold to beta-cypermethrin, 3.85-fold to dichlorvos, and 17.44-fold to phoxm, respectively, implying that insecticide use for control of H. armigera should result in outbreak of cotton aphids. In comparison with the impacts of pesticides against H. armigera in the sprayed plots of conventional cotton, the aphid populations in unsprayed Bt cotton plots were suppressed efficiently by natural enemies. Therefore, as our experiments indicated, cotton aphids on transgenic Bt cotton plants remained at low densities with no use of insecticides in mid and late stages of cotton growth. It was suggested that Bt cotton not only could play an important role in control of H. armigera, but also could efficiently prevent cotton aphids from resurgence caused by insecticidal application against H. armigera.

It is reported that there are no adverse effects of Bt cotton on several beneficial insects including Hippo-damia convergens Gueréin-Méneville, Apis mellifera L., Nasonia vitiripennis (Walker), and Chrysopa carnea (Sims 1995). Population density of sweet potato whitefly, Bemisia tabaci (Gennadius), on Bt cotton is higher than on the control cultivars as a result of reduced leaf feeding damage by lepidopterous insects (Wilson et al. 1992). A study on the influences of Bt cotton plantings on the population dynamics of the mirids, Lygus lucorum Meyer-Dür, Adelphocoris fasci-aticollis Reuter, and Adelphocoris lineolatus (Goeze) (Hemiptera: Miridae), indicates that there are no significant differences between population densities of these bugs on unsprayed conventional cotton and un-sprayed transgenic cotton, but the mirid density on unsprayed transgenic cotton is significantly higher because of a reduced number of insecticide sprays against H. armigera compared with the number of sprays in the conventional cotton (Wu 2001).

The present results indicated that the use of Bt cotton in production agriculture of northern China should have the advantages of reducing the use of chemical insecticides for control of two key insect pests, H. armigera and cotton aphid. This should result in decreased environmental pollution and in lower cost for insect control in cotton. Additionally, by reducing the area sprayed with insecticides and the frequency of sprays, the rate of insecticide resistance evolution should decrease and there is an increase in the potential for natural and biological control of cotton insect pests.

Acknowledgments

We thank Fred Gould (North Carolina State University, Raleigh, NC) and Fengming Yan (Peking University, Beijing, China) for critical review and discussion of the manuscript. This research was supported by 973 Projects Grant (G2000016208) and 863 Projects Grant (2001AA212271) from Ministry of Science and Technology of China.

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/), which permits non-commercial reuse, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions{at}oup.com

References Cited

View Abstract