Abstract

Life tables were established for trichlorphon-resistant and susceptible Bactrocera dorsalis strains based on the laboratory observations. Trichlorphon-resistant B. dorsalis strain had longer pupal and preoviposition periods, and mean generation time compared to the trichlorphon susceptible strain. Lower fecundity, emergence rate, and probability of standard fecundity 𝑃𝐹, and shorter female and male longevity also were apparent in the trichlorphon resistant strain. Based on the life tables, the life population trend index (𝐼) of the resistant strain was 86.80, while that of the susceptible strain was 116.97. The net reproduction rate (𝑅0) and the intrinsic rate of increase (𝑟𝑚) of the resistant strain were 1565.33 and 0.0164, while those of susceptible strain were 2184.00 and 0.0173, respectively. The results from this research revealed that the resistant strain was at a reproductive and developmental disadvantage relative to the susceptible strain.

1. Introduction

The oriental fruit fly, Bactrocera dorsalis (Hendel) (Diptera: Tephritidae), is distributed throughout Southeast Asia and the Pacific [1]. Now it has also infested many areas of China and thus causes serious financial loss to orchards globally [2]. The loss cost incurred by B. dorsalis was estimated to be as high as 44.6~176.5 million U.S. dollars in California, and also led to 230 million U.S. dollars potential economic loss of stone fruit industry in California [3], 1.26 billion U.S. dollars in 1997 in Taiwan, and about 1.47 billion U.S. dollars in 2004 in the Fujian [4–6].

Using insecticides is the major way to control this pest, however, frequent use of common insecticides had resulted in insecticide resistance in several insecticides, including organophosphates [7]. Resistance to insecticides has become a major problem of many pest species, including the oriental fruit fly. Keiser used 73 insecticides to determine their toxicity to B. dorsalis in 1973 [8], and found most of these insecticides had high toxicity. Purcell tested 3 insecticides (Malathion, Benzyl Cypermethrin and Carbaryl) and found that Malathion was the most toxic in 1994 [9]. Resistance of B. dorsalis populations to insecticides in the field has become more and more serious since 2003 [2]. In 2004, it was reported that fly populations in Taiwan had resistance to ten insecticides (Naled, Fenthion, Trichlorphon, Fenitrothion, Formothion, Malathion, Methomyl, Cyfluthrin, Cypermethrin, and Fenvalerate), and that resistance to Malathion increased at the fastest rate, while Naled had the lowest rate [7, 10, 11].

In South China, the oriental fruit fly is distributed widely and its population density is very high, resulting in tremendous yield loss to many fruits. Chemical control strategies still play an important role in depressing oriental fruit fly density and reducing economic loss, therefore, it is very important to monitor the resistance of B. dorsalis to insecticides in the field [12, 13]. The development of resistance to organophosphate-based insecticides is a current and growing problem for the management of this pest species. A wide range of studies have focused on the elucidation of the molecular basis of this resistance [14, 15], and showed that flies exhibiting high levels of organophosphate resistance carried three specific mutations in the ace gene of this species. Trichlorphon has been used to control B. dorsalis in commercial orchards in China, and now there is evidence of high resistance to this insecticide as well [2, 16–18], however the biology and ecology of the resistant populations of B. dorsalis has not been well studied.

The ecology of resistant insects is essential to the selection of control methods and the understanding of resistance dynamics. This study presents population life tables and life-history traits of susceptible and resistant strains to determine whether the relative fitness of resistant strains increases or decreases [19]. In some resistant insects, because the relative fitness of resistant insects is decreased, rotation of different insecticides and temporarily discontinuing the use of certain insecticides can decrease the frequency of resistance genes, and restore their susceptibility to certain insecticides [20, 21]. However, if the relative fitness of resistant insects is not affected by the resistance, these practices can only delay resistance development, and it is difficult to restore their susceptibility to those insecticides [22].

Compared to susceptible insects, resistant insects usually have a lower fecundity [23]. To provide a scientific basis for resistance risk management and achieve successful control of resistant B. dorsalis, it is important to study the biological and ecological variations between the trichlorphon-resistant and -susceptible B. dorsalis populations. This study examined the following areas: (1) the basic susceptibility of B. dorsalis to trichlorphon; (2) the development of resistance to trichlorphon through exposure over several generations; (3) the biological and ecological features of trichlorphon resistant strain.

2. Materials and Methods

2.1. B. dorsalis Strains

The trichlorphon resistant strain of B. dorsalis was collected in March 2003 from damaged carambola fruits, in Yangtao Park, Guangzhou City, Guangdong Province, China. Few insecticides had been used at this location because bags are used to protect carambola fruits from the flies. We collected about 2000 aging larvae from the damaged carambola, then placed the larvae into humid sand for pupation. After pupa emergence, flies were kept in the laboratory for 33 generations before using in the study. Bioassays were conducted every 3 generations, and new 24-h LC50 values were obtained and guided the trichlorphon concentration which was used to treat the next 3 generations.

The susceptible strain was collected in August 2003 from the same place as the trichlorphon resistant strain, and maintained without any exposure to insecticides for 30 generations in the same laboratory as the resistant strain collected [2].

2.2. Method
2.2.1. Concentration-Response Bioassays

In order to make sure that the 24-h LC50 value of the trichlorphon-resistant and -susceptible strains was stable, the populations of two strains from the field were both divided into 3 groups, and the 24-h LC50 value of each group was measured. Not only trichlorphon was used in the field, but also other organophosphorus pesticides which had systemic activity and the same mechanism as trichlorphon were used; so in the laboratory, we treated both larvae and adults with trichlorphon. The larvae of each generation were treated by topical application. In brief, aging larvae were treated with acetone diluted of trichlorphon whose concentration was the 24-h LC50 value of the last time measurement. Treated larvae were then placed in humid sand for pupation. Adult flies, 3–5 days old, were treated by vial residues of acetone diluted trichlorphon, whose concentration was the 24-h LC50 value of the last time measurement. Briefly, 5 mL of the acetone dilution was poured into a 250 mL triangular flask with constant shaking until the sides of the flask formed a uniform film. Excess diluent was discarded. When acetone evaporated, adults of B. dorsalis were then added to the flask, and supplied with a cotton ball saturated with 5% honey. The flask was sealed and placed at room temperature for 28 ± 2°C for 24 hours. Flies were considered dead when they could not crawl after continual prodding. The surviving adults were used for the next generation breeding [2, 24].

Two strains were both maintained at a temperature of 28 ± 2°C and in a photoperiod of 10 : 14 h [14]. The data from the regression equation, including 24-h LC50 values of treatments, 95% confidence intervals and correlation coefficient (𝑟), were analyzed by probit analysis. For each of 5 concentrations (treatments), there were 5 replicates, with each replicate using 30 flies. From the number of dead flies, we calculated the mortality and corrected mortality, translated corrected mortality into mortality probit. Then according to the logarithm of the concentrations and relevant mortality probits, we made the toxicity regression equation (LC-P line). Finally, according to the regression equation, we calculated the 24-h LC50 value, the confidence interval, chi-square value, and correlation coefficient. If the 95% confidence intervals overlapped, we concluded that trichlorphon sensitivity was not statistically different between the strains [25]. The resistance ratios (RRs) are given as the values of the resistant LC50 to trichlorphon/the susceptible LC50 to trichlorphon in the same bioassay method [10, 11]. Microsoft Office Excel (2003) was used for data analysis, the test of significance (ANOVA: 𝐹 test).

2.2.2. Duration of Life Stages

To determine the duration of the egg stage and hatching rate, we cut banana into small pieces and placed them in disposable plastic cups (150 mL), then we put the cups into the adult rearing cages (60×60×60 cm) of both strains to attract B. dorsalis female to lay eggs, respectively. The egg masses were removed after 2 hours, and 50 eggs from each strain were placed into a petri dishe (∅ 9 cm) with filter paper containing banana juice. Hatched larvae were removed every two hours and the total number of larvae was counted.

Larval duration and pupation rate were determined from 50 susceptible and 50 resistant strain larvae that simultaneously eclosed. Banana was used to feed the larvae in a plastic box (15×15×15 cm). Dead and aging larvae were removed every 4 hours. The larval stage was from egg hatch to aging, bouncing, larvae (larva of B. dorsalis have three instars, at the end of third-instar, they would bounce up 3–15 cm high and then fall to find a place to pupate). Total number of aging larvae was counted, and the duration of larvae and pupation rate were then calculated, respectively.

The duration of the pupal stage and the adult emergence rate was determined from 50 aging larvae from the susceptible and resistant strains. Bouncing larvae were placed into a box (15×15×15 cm) containing sand maintained at 70% RH. The emergence of pupae was observed every 8 hours until all pupa emerged. Newly emerged adults were removed every 8 hours. The duration of pupa was designated from the time aging, bouncing, larvae entered the sand to the time of adults emergence. The total number of adults was counted; the duration of pupa and emergence rate were then calculated.

2.2.3. Life-History Traits

Newly emerged adults (<8 h) were reared in a plastic cage (60×60×60 cm) with artificial feed and water. Before adults began to lay eggs (emerged about 10 d at 28 ± 2°C), 20 successful mating pairs were put into a new cage (60×60×60 cm) to observe egg laying every day until all female adults were dead. The total number of eggs laid by each female was recorded. Net reproduction rate (Ro, the average fecundity of each adult by a generation), Intrinsic rate of increase (𝑟𝑚, an important proliferation potential parameter of populations under certain environmental conditions, a composite indicator of the survival rate, fecundity, growth rate on the impact of population changes), and the index of population trend (𝐼, the growth multiples of the next generation) were calculated using the following three formulas.

Formula of 𝑅0: 𝑅0=1𝑥𝑚𝑥(1) (see [26]).

Formula of 𝑟𝑚:𝑙𝑥𝑚𝑥𝑒−𝑟𝑚𝑥=1(2) (see [27]), where 𝑥 is age for insects; 𝑙𝑥 is the population of 𝑥 period 𝑚𝑥 is the number of female at next generation of each female of x period.

Formula of 𝐼: 𝐼=𝑆1𝑆2𝑆3⋯𝑆𝐾𝐹𝑃♀𝑃𝐹(3) (see [28–30]), where 𝑆1𝑆2𝑆3⋯𝑆𝐾 is the survival rate at each acting stage, respectively; 𝐹 is the standard egg number per female; 𝑃𝐹 is probability of standard fecundity; 𝑃♀ is female rate.

2.2.4. Adults Survival Lines and Relative Fitness

We used age as an independant variable, the survival rate as the dependent variable to make a survival line, and fit a model based on the Weibull distribution. When shape parameter 𝑐>1, then the age to which the vast majority of individuals in the population were able to achieve is its average life expectancy. When shape parameter 𝑐=1, then the population had the same mortality at different times. When shape parameter 𝑐>1, then the population had very high mortality in the prophase, and most were dead before average life expectancy. The formula was as follows: 𝑆𝑝(𝑡)=𝑒−(𝑡/𝑏)𝑐(4) (see [31]), where 𝑆𝑝 is survival rate of 𝑡 age; 𝑏 is scale parameter; 𝑐 is shape parameter.

Fitness costs associated with resistance to insecticidal agents can have substantial negative effects on many life history traits and have been reported in Diptera [32, 33]. The formula of Relative fitness (𝑅𝑓) was as follow: when 𝑅𝑓>1, it suggested the fecundity of resistant strain was enhanced; when 𝑅𝑓<1, it is suggested that the resistant strain possess a fitness defect 𝑅𝑓=𝑅0ofresistantstrain𝑅0ofsusceptiblestrain(5) (see [34, 35]).

3. Results

3.1. Establishment of Resistant Line

At the 8th generation, the resistant ratios (RRs) of the trichlorphon resistant strain exhibited up to a 41.42-fold increase (Table 1). After 14 generations of selection, the resistant ratios (RRs) rapidly went up to 84.55-fold. In order to understand the resistance stability, from the 14th generation to the 21st generation, no insecticide was used on the resistant strain, and the resistant ratios (RRs) rapidly declined to 19.49-fold; however, when trichlorphon was used again, after only 3 generations (i.e., the 24th generation in this experiment), the resistant ratios (RRs) went up to 54.60-fold. Then from 24th generation to 30th generation, the resistance of trichlorphon-resistant strain increased slowly, until the 33rd generation, the resistant ratios (RRs) reached 71.93-fold.

3.2. Growth and Development of Each Stage

Tables 2 and 3 showed that the duration of pupae, preoviposition period, longevity of female, longevity of male, fecundity per female, emergence rate, and probability of standard fecundity 𝑃𝐹 of trichlorphon resistant strain all had a significant difference to that of the susceptible strain. Results also clearly showed that the duration of egg, duration of larvae, the pupation rate, mean generation time, hatching rate, pupation rate, and female rate 𝑃♀ had no significant differences between susceptible strain and trichlorphon-resistant strain.

3.3. Life-History Traits

Analyses of the population life tables, the influence of eggs, larvae, pupae, and adults of the susceptible and resistant strains, and the ratio of adult females indicated that the population tendency index of two strains had significant differences. The population tendency index of susceptible was 116.97, but that of the resistant strain was only 86.80, which suggested that the succeeding generation of the resistant strain was smaller than that of the susceptible strain, and had certain propagation disadvantages.

Fecundity of the resistant and susceptible strains were both almost a parabolic shape; in the early oviposition period, the fecundity was small, and gradually increased, reaching a peak and then declining gradually. But there was a certain difference between the strains. The susceptible strain reached its peak oviposition time faster than that of the resistant strain. Fecundity at peak oviposition for the susceptible strain was more stable, while the resistant strain was more volatile. Based on the standard errors, the fluctuation range around the mean was larger in the resistant strain than that of susceptible strain (Figure 1).

3.4. Adults Survival Lines and Relative Fitness

The 𝑅0, 𝑟𝑚, and 𝑅𝑓 of susceptible and resistant strains showed significant differences (Table 4), suggesting that the fecundity of the resistant strain declined significantly after the populations of B. dorsalis developed resistance to trichlorphon. The fitness study showed that the relative fitness of resistant strain obviously declined (only 0.7167 of the susceptible strain), which suggested that the survival and development of the resistant strain had been significantly poorer than the susceptible strain.

According to formula 2, the fitting of the adult survival curve equation of adults of susceptible and trichlorphon-resistant strains (Table 5) and survival curve (Figure 2) showed that the values of shape parameter (𝑐) was 3.4637 and 2.6451. Both of them were greater than 1, in consistent with the type I of the basic model of survival curve, suggesting that the adults of both strains achieved their average life span and reached the inherent life to death. But the average survival rate of the susceptible strain was higher than that of resistant strain before 85 days.

4. Discussion

Insecticides can change the biology, ecology and other indicators of resistant insects. Many studies suggest that the resistant insects exhibit fitness costs or stimulate their proliferation [36, 37]. The variations of biology and ecology were closely related to different insecticides [38]. Studying the resistant population's relative fitness is the basis for understanding and resolving the problem of resistance [39, 40]. Relative fitness refers to the relative capacity of biological survival and reproduction, compared to the susceptible strain. It is generally believed that the biological characteristics, such as declined fecundity and prolonged growth period, change relative fitness. However resistance increases the pest survival rate under selective pressure, which can often lead to a fitness decline. Resistant individuals have shown a survival competitive disadvantage in the absence of selective pressure agents [41, 42].

Our results in this study indicated that the resistant ratios (RRs) of trichlorphon-resistant strain of B. dorsalis quickly reached 84.55-fold after 14 generations under trichlorphon selected pressure. However, the resistance was not stable, with the resistant ratios (RRs) rapidly declining when insecticide applications ceased for 7 generations. After insecticide pressure was resumed, RRs went up rapidly only after 3 generations. A resistant strain of RRs 71.93-fold was obtained after 33 generations. These results showed that in B. dorsalis, the rising speed of RRs of trichlorphon-resistant strain was far smaller than that of spinosad-resistant strain; LD50 of the selected line was 408 times greater compared with that of the untreated parental colony only after eight generations of selection [43]. And it was almost unanimous compared with rate of increase seen for other cases of insecticide resistance in B. dorsalis, including selection for resistance to six organophosphates, one carbamate, and three pyrethroids described previously in [7]. The results obtained here suggest that a rapid rise in trichlorphon-resistance may ultimately cause control failure after extended commercial use in the fields.

The life-history traits of B. dorsalis were significantly influenced by trichlorphon. The trichlorphon resistant strain had significant fitness costs when compared to the susceptible strain. The duration of pupa, preoviposition period, and mean generation time of resistant strain were all prolonged, while the longevity of female and longevity of male were shortened, and mean fecundity per female, the emergence rate, and probability of standard fecundity 𝑃𝐹 were lower. The population tendency index of two strains had significant difference, the population tendency index of susceptible was 116.97, but that of the resistant strain was only 86.80 (Table 2). The 𝑅0,𝑟𝑚 of resistant strains was smaller than that of susceptible strain, and 𝑅𝑓 was only 0.7167 of the susceptible strain (Table 4). The average fecundity per female curves of two strains were both more or less a parabolic shape, but the susceptible strain reached its oviposition peak time in a shorter space of time than that of the resistant strain; the fecundity of oviposition peak time was more stable than that of resistant strains which was more volatile (Figure 1). Survival curves showed that the two strains were consistent with the I-type of the basic model of survival curve, but the susceptible strain reached its oviposition peak time in a shorter space of time than that of the resistant strain; the fecundity of the oviposition peak time was more stable; than that of resistant strains which was more volatile (Figure 2). The values of shape parameter (𝑐) was 3.4637 and 2.6451 (Table 5); they suggested that the adults of both strains achieved their average life span, and reached the inherent live to dead.

All the results could afford the important basics of resistance control of the resistant strain of B. dorsalis. In conclusion, the results obtained here show that in B. dorsalis, consistent with what has been seen in several species (Culex pipiens, Helicoverpa armigera, Drosophila melanogaster, Cydia pomonella) of insects in different orders, development of high levels of resistance to organophosphates will occur within a relatively short time after selection is applied, and had relative fitness costs [44–49]. The resistance gene had an adverse impact on the population reproduction and development. In general, arthropod resistance to insecticides often accompanies a fitness cost, which is the theoretical basis for using insecticide rotation, to govern pest resistance [50, 51]. For maintaining the effectiveness of this insecticide, any proposed management program must take into consideration the potential for resistance development seen here in the oriental fruit fly. If the major influencing factor of trichlorphon resistant strain was trichlorphon, when the resistance was found, it was proposed that trichlorphon and other organophosphate applications cease to allow the frequency of resistance genes to decrease. Because of the resistant strain's fecundity disadvantages and fitness costs, the population was relatively feeble, and without the presence of insecticides, the resistant population might slowly decline.

At the same time, in the nonresistant areas, rotation of trichlrophon and other different mechanism insecticides could thwart trichlorphon resistance development, increase the control effect, reduce environment pollution, and protect the environment, except for pyrethroids. The cross-resistance bioassays revealed, when the resistant ratios of trichlorphon resistant strain reached 69.21-fold, that some resistance to pyrethroids existed in trichlorphon resistant B. dorsalis strain. Pyrethroids showed about 30-fold cross-resistance to trichlorphon. So it is likely that oriental fruit flies already exhibiting higher resistance to trichlorphon also will develop middle-level resistance to pyrethroids. But abamectin showed low cross-resistance to trichlorphon, and we may use abamectin to rotate to trichlrophon (we will separately report this section study). Also we can use spinosad replace trichlorphon in the field where the population of B. dorsalis already had trichlorphon resistantce, because according to the report by Hsu et al. [14], the results showed the spinosad did not exhibit cross-resistance to ten insecticides, including six organophosphates (naled, trichlorphon, fenitrothion, fenthion, formothion, and malathion), one carbamate (methomyl), and three pyrethroids (cyfuthrin, cypermethrin, and fenvalerate) [43].

In this study, the biological and ecological indicators of two B. dorsalis strains were studied under laboratory conditions, although the dilution effect of gene flow to antagonistic alleles was avoided, the interference of other field factors outside was ignored, such as environment and natural enemies. So to be more accurate in our understanding of the dynamic resistance, and afford the most powerful evidence of resistance control, the biological and ecological indicators of trichlorphon resistant strain in the fields should be further studied.

Acknowledgments

The authors wish to acknowledge the significant improvements to this paper suggested by the editor and thank Dr. Julia W. Pridgeon (Agricultural Research Service, United States Department of Agriculture), Dr. Philip Leftwich; Dr. Gong HF (Oxitec Oxford Insect Technologies, Oxitec limited, England), and Professor Zeng FR (Plant Protection Institute of CAAS) for the correction of this paper. This paper was supported by the National Science and Technology Planning Project (no. 2006BAD08A14) and Scientific Research Project of Guangdong Province (no. 2004A20401002).