Dynamics of Bemisia tabaci biotypes and insecticide resistance in Fujian province in China during 2005–2014

The whitefly Bemisia tabaci (Gennadius) is an important agricultural insect pest worldwide. The B and Q biotypes are the two most predominant and devastating biotypes prevalent across China. However, there are few studies regarding the occurrence of the Q biotype in Fujian Province, China, where high insecticide resistance has been reported in the B biotype. Differences in some biological characteristics between the B and Q biotypes, especially insecticide resistance, are considered to affect the outcome of their competition. Extensive surveys in Fujian revealed that the B biotype was predominant during 2005–2014, whereas the Q biotype was first detected in some locations in 2013 and widely detected throughout the province in 2014. Resistance to neonicotinoids (that have been used for more than 10 years) exhibited fluctuations in open fields, but showed a continual increasing trend in protected areas. Resistance to lambda-cyhalothrin, chlorpyrifos, and abamectin exhibited a declining trend. Resistance to novel insecticides, such as nitenpyram, pymetrozine, sulfoxaflor, and cyantraniliprole, in 2014 was generally below a moderate level. A decline in insecticide resistance in the B biotype and the rapid buildup of protected crops under global temperature increase may have promoted the establishment of the Q biotype in Fujian.

Resistance to insecticides in use since 2009. Toxicity of novel insecticides against B. tabaci is presented in Table 1. All of the four collected whitefly populations exhibited very low resistance to nitenpyram and from very low to low resistance to pymetrozine. The resistance factors for the tested populations for nitenpyram were no more than 6-fold, and for pymetrozine were no more than 12-fold. Although low and moderate resistance was found to cyantraniliprole, none to moderate resistance was detected to sulfoxaflor. The resistance factors to nitenpyram and sulfoxaflor were not always higher in open field populations than in protected area populations. However, for pymetrozine and cyantraniliprole, protected area populations always exhibited higher resistance factors than open field populations.

Discussion
Since the first detection of the Q biotype in the Yunnan Province in 2003 19 , the Q biotype has subsequently been found and become the prevalent biotype in the Yangtze River Valley 19,[32][33][34][35] and on crops in most locations in China 21 , although the south and southeastern coastal areas of China are dominated by the B biotype 20 . Our research also revealed that the B biotype was always the predominant biotype in Fujian during 2005-2014. In addition, the Q biotype was first detected in some areas in 2013 and widely detected throughout the province in  Table 2. Data for 2005 and 2009 partially derived from He et al. 30 and Zheng et al. 31 , respectively. Maps were created by using the R statistical program version 2.15.2 65 with four add-on packages 'maps' 66 , 'maptools' 67 , 'ggplot2' 68 and GISTools 69 . The data for creating the base map of Fujian Province, China was obtained from the Global Administrative Areas (GADM) database (Version 2.8, http://www.gadm.org/). 2014. Wang et al. 22 reported finding the Q biotype in Fuzhou and Zhangzhou cities on sampled cauliflower in October 2008. Although another record relevant to the detection of the Q biotype in Fujian occurs in GenBank (accession number FJ594434), we could not find any descriptions in the paper published by Teng et al. 36 .
Regardless, it appears that the Q biotype invaded Fujian, particularly in areas where the development of agriculture was advanced. However, population establishment of the Q biotype may have not been successful initially. We re-sampled whiteflies in the fields where Wang et al. 22 had in 2009, but we failed to detect the Q biotype using the specific mtCOI primers used in the present study (see section 2.2) or mtCOI sequencing as in our previous research 31 , implying that the Q biotype previously invaded Fujian, but failed to colonize it before 2012. Multiple secondary introductions, however, may be an important prerequisite for the establishment of the Q population in the fields 37 .
A stronger resistance to insecticide is often regarded as a major cause for the establishment and the dominance of the Q biotype 38,39 , although the B biotype has been shown to be more competitive than the Q biotype under untreated conditions 5,40,41 . However, if the B biotype could rapidly evolve a high resistance, it would supplant the Q biotype even when repeatedly exposed to insecticides 41 . Moreover, the B biotype, being slightly greater in number (30 vs. 10), was found to displace the Q biotype in six generations when exposed to a low dose of neonicotinoid insecticide 39 . Thus, the B biotype, with high resistance and abundance, would impede the establishment of the Q biotype. This is in agreement with the fact that only the B biotype was sampled before 2012 by our research team and Pan et al. 21 and the finding that the Q biotype was rare in Fujian, where high levels of resistance to  various insecticides in the B biotype have been found 30,31 . Similarly, Crowder et al. 41 concluded it was the ability of the B biotype to rapidly evolve resistance to insecticides in the United States but not Israel that resulted in the exclusion of the Q biotype from the field in the United States 42 , but co-occurrence with and dominance over the B biotype in Israel 5,15,43 . Moreover, the shift in the predominant biotype from the Q to B in 2003-2012 may have partially resulted from an increase in resistance to insecticides of the B biotype in Israel 38 .
The decrease in abundance, as well as the resistance of the B biotype to some extent, facilitated the establishment of the Q biotype.  Table 1). The Q biotype, being more resistant than the B biotype, superseded the B biotype when they were at the same initial number and exposed to a low concentration of a neonicotinoid insecticide 39 .
Another possible explanation for the establishment of the Q biotype in Fujian is the rapid buildup of protected agriculture across the province under climate warming. The Q biotype is more tolerant to heat stress than the B biotype 4,6 , which gives the Q biotype an advantage over the B biotype in environments with high temperature. The Q biotype that invaded Fujian must have encountered the B biotype under extremely hot temperatures in recent years. On one hand, rising temperatures characterized the recent and current climate 44 . In the summer of 2013, an extremely unusual heat wave hit central and eastern China. The average daily mean temperature, daily maximum temperature, daily minimum temperature, and number of days with daily maximum temperature that reached or exceeded 35 °C were all greater in 2013 than in 1960-2012 45 . Moreover, a continuing warming was observed in Fujian in the summer of the next year, when the average seasonal temperature increased from 27.4 °C in 2013 to 28.0 °C in 2014 46 . On another hand, the development of protected agriculture was encouraged and subsidized by the government since 2009, and its advancement was accelerated since 2013 when financial assistance increased substantially. Thus, high temperatures ≥ 35 °C, which appear in the open fields and often exist in protected areas, may have contributed to the results in this paper, specifically the colonization of the Q biotype and its proportional increase in whitefly population. Chu et al. 37 demonstrated that wide use of the protected agriculture in northern China promoted the colonization of the Q biotype in agricultural fields.
Although it is now dominated by the B biotype, a continuous increase of the Q biotype in protected areas in Fujian is probable. Kontsedalov et al. 15 and Hsieh et al. 16 both found that the B biotype dominated crops in open fields and the Q biotype dominated crops in protected agricultural areas. However, Kontsedalov et al. 15 speculated that the high fitness rate for the B biotype and the inefficiency of insecticide control contributed to the predominance of the B biotype in open fields, whereas Hsieh et al. 16 regarded temperature as the main impact factor. Although resistance to repeated insecticide spraying and high temperature stress contribute to the domination of the Q biotype in protected areas 15 , different kinds of insecticides in use may influence the B. tabaci biotype composition 16 . Other than insecticide, habitat type, and climate, the shift in biotype composition may also result from the infestation of B. tabaci by its symbionts and viruses 11,12,38 . Based on the resistance data tested in 2005, 2009, and 2014, we can infer that, in general, field populations of B. tabaci resistant to neonicotinoids underwent an increase in the first several years and a reduction during the subsequent 10-year period. Yearly population dynamics of B. tabaci in Fujian, as mentioned above, may reflect the variance of the resistance. Although a shift in the biotype composition for the ZZ population was detected in 2014, resistance may not be the reason for the decline because the Q biotype was more resistant to neonicotinoids than the B biotype 9,39,41 . However, neonicotinoid resistance in protected areas, such as greenhouses, net houses, plastic houses, etc., was completely different. Although a drop in neonicotinoid resistance was observed for open field populations in 2014, a continuous rise was found for LY and SM populations, both from plastic houses. The former contained the B biotype only, whereas the latter consisted of mixed B and Q biotypes, but was dominated by the Q biotype. Thus, B. tabaci in protected areas will increase in resistance regardless of biotype composition, because B. tabaci is prone to outbreak in these areas, and both B and Q biotypes have the capacity to evolve strong resistance to insecticides, including neonicotinoids 22,31,47 .
In contrast to neonicotinoids, resistance of B. tabaci from open field and protected area to lambda-cyhalothrin and chlorpyrifos constantly decreased in the three test years, implying a declining trend from 2005 to 2014. The replacement with new chemicals, such as neonicotinoids, with novel modes of action relative to conventional insecticides, such as pyrethroids, organophosphates, and carbamates, will likely contribute to a dramatic decline in resistance to conventional insecticides 28,48 . Insecticide resistance in B. tabaci was demonstrated to be positively correlated with the frequency of insecticide application 49 . A shift from synergized pyrethroids to buprofezin and pyriproxyfen in B. tabaci control resulted in a significant reduction in pyrethroid use and recovered the susceptibility to synergized pyrethroids 50 . Similarly, rotation of old and new chemicals that prevents or delays the onset of resistance in whiteflies may have resulted in the susceptibility to lambda-cyhalothrin during 1992-2000 and chlorpyrifos during 1992-2007, whereas these two insecticides were commonly used in Pakistan 28,48 . In the present study, we also found that no resistance existed to abamectin. Abamectin appears efficacious in controlling not only the B biotype, but also the Q biotype, as clarified by Wang et al. 22 and Xie et al. 51 .
Nitenpyram and pymetrozine have been recommended for use by the agricultural department since 2009, when high levels of resistance to commonly used insecticides in Fujian Province were detected. Although high or very high levels of resistance to imidacloprid and thiamethoxam were detected in protected areas (Fig. 4), very low and low levels of resistance to nitenpyram and pymetrozine were found there, indicating these two insecticides could still be recommended for use.
Cyantraniliprole and sulfoxaflor were temporally registered in China in 2012 and 2013, respectively, and their commercial products were available on the market in 2013. Although cyantraniliprole, being an anthranilic diamide insecticide, acts exclusively on the ryanodine receptor in insects, it was thought to be highly efficacious against B. tabaci 52,53 . However, high LC 50  Sulfoxaflor targets insect nicotinic acetylcholine receptors (nAChRs) and functions discriminately with other insecticides acting at nAChRs, like neonicotinoids 54 . A lack of cross-resistance between sulfoxaflor and neonicotinoids has been demonstrated in B. tabaci and Trialeurodes vaporariorum 55,56 . In comparison with the recommended field concentration of 73.3 mg A.I. L −1 , a considerably low LC 50 value (0.64-12.61 mg A.I. L −1 ) along with slope values less than 1.5 were found in the four populations, suggesting that sulfoxaflor may be an important new tool for whitefly control.

Conclusions
The Q biotype was first detected in some areas in 2013, followed by frequent detection throughout Fujian Province in 2014. However, the whitefly population was predominated by the B biotype during 2005-2014, although the Q biotype appeared to be increased in proportion not only in open fields but also in protected areas. The abundant B biotype, with high levels of resistance, may prevent the colonization of the Q biotype even in cropping systems where insecticides were overused, although the Q biotype was thought to be more resistant. The establishment of the Q biotype, however, might have been facilitated by the decline in the abundance of the B biotype which resulted in a drop in insecticide resistance, as well as a warming environment, that is, the rapid buildup of protected agriculture in the context of global temperature rise.
Resistance to neonicotinoids used for more than 10 years exhibited fluctuations in open fields, but a continual increase in protected areas. Conversely, a continuous decline was observed for conventional insecticides, such as pyrethroids and organophosphates. Resistance variance reflects the change in intensity of insecticide utilization. Application reduction, therefore, contributes to some extent to the decline in resistance, which indicates the importance of rational rotation in insecticide resistance management. No resistance was found to abamectin during 2005-2014 demonstrating that it is a powerful insecticide against B. tabaci, regardless of biotype. Four novel insecticides, nitenpyram, pymetrozine, and sulfoxaflor, are suitable for controlling both biotypes of B. tabaci in open fields and protected agricultures, although cyantraniliprole should be used prudently. Judicious rotation of conventional and novel insecticides should be adopted in whitefly resistance management.  (Table 2 and Fig. 1). Adult whiteflies were sampled using a simple mouth suction apparatus. Samples used for resistance bioassays were first anesthetized by CO 2 and then transferred into plastic ventilated petri dishes (Φ = 3.5 cm) where cotton leaves, placed with their adaxial surface downwards onto the agar, were provided for whitefly population maintenance. For those used for biotype determination, samples were either anesthetized by CO 2 , stored in 95% ethanol, and taken back to the laboratory where they were frozen at − 20 °C or host plant leaves, along with whiteflies, were placed in plastic bags, and then taken back to the laboratory and frozen at − 20 °C.

Material and Methods
Biotype determination. For whitefly samples collected in 2005 and 2009, direct sequencing of the mitochondrial cytochrome oxidase I (mtCOI) gene was used to identify the biotype of whitefly individuals. The methodology followed He et al. 57 . Briefly, whitefly DNA was extracted from individual adults following the method of Luo et al. 58 . The primers C1-J-2195 (5′ -TTGATTTTTTGGTCATCCAGAAGT-3′ ) and L2-N-3014 (5′ -TCCAATGCACTAATCTGCCATATTA-3′ ) described in Frohlich et al. 59 were used for PCR amplification. Total reaction volumes of 30 μ L were used with a final concentration involving 1.5 U of Taq polymerase, 2 mg mL −1 BSA, 0.25 mM of dNTPs, 2.5 mM MgCl 2 , 75 ng C1-J-2195, 75 ng L2-N-3014, and 4 μ L of DNA. Amplification was conducted in an Eppendorf thermocycler using the following parameters: 94 °C for 5 min, followed by 35 cycles at 94 °C for 1 min, 50 °C for 1 min, 72 °C for 1 min, and a final extension at 72 °C for 5 min. PCR products were electrophoresed in 1.2% agarose gel and purified by Promega DNA gel extraction kit and sequenced directly in two strands by Invitrogen (Shanghai, China). Biotype was determined for 24 individual adults for each population. The biotype of whiteflies collected in 2011, 2013, and 2014 was determined individually through the amplification of the mtCOI gene by PCR with the specific primers described in Shatters et al. 60 , that is, for the B biotype forward primer, 5′ -CTAGGGTTTATTGTTTGAGGTCATCATATATTC-3′ , and reverse primer, 5′ -AATATCGACGAGGCATTCCCCCT-3′ ; for the Q biotype, the forward primer, 5′-CTTGGTAACTCTTCTGTAGATGTGTGTT-3′ , and reverse primer, 5′-CCTTCCCGCAGAAGAAA TTTTGTTC-3′ . PCR reaction was performed in a volume of 20 μ L containing 4 μ L double distilled water, 10 μ L 2 × Long master mix (TianGen Biotech, Beijing, China), 0.5 μ L (10 μ M) forward primer for both biotypes, 0.5 μ L (10 μ M) reverse primer for both biotypes, and 4 μ L template DNA. PCR procedure consisted of an initial denaturation of 94 °C for 2 min, followed by 35 cycles at 94 °C for 0.5 min, 64 °C for 1 min, 72 °C for 1 min, and a final extension at 72 °C for 7 min. PCR products were separated by electrophoresis in a 1.2% agarose gel. Bemisia tabaci biotypes were easily identified based on the criterion of 478 bp for the B biotype and 303 bp for the Q biotype 21 . A total of 13-57, but typically 24 individual adults were identified for each population.
Bioassays. Samples collected from fields were reared on cotton plants in the laboratory at 25-28 °C, 65 ± 5% RH, and 14:10 h (L:D) photoperiod, and the F 1 generation adults were used for bioassays. Leaf dip bioassays with B. tabaci adults followed the methodology of He et al. 61 . Cotton leaf discs (Φ = 3.5 cm) were dipped for 5 s in 5-8 serial dilutions of insecticides or in water as the control. After air drying, leaf discs were placed with their adaxial surface downward onto agar in a petri dish (Φ = 3.5 cm). Twenty-five to thirty-five B. tabaci female adults were briefly anesthetized with CO 2 and gently transferred onto the leaf discs. Each petri dish was then covered with a perforated lid and placed inverted in a growth chamber at 25 ± 1 °C, 65 ± 5% RH, and 14:10 h (L:D) photoperiod. Each insecticide treatment was repeated three times and mortality was recorded after 48 h, with the exception of pymetrozine, which was examined after 96 h.  Statistical analysis. Data were corrected for control mortality using Abbott's formula 62 and analyzed by probit analysis 63 . Significant differences between LC 50 values were determined by the absence of the overlap of the 95% confidence limits. Resistance factors (RFs) were obtained by dividing LC 50 values by the corresponding values of the susceptible population (Lab-S) reared in laboratory without exposure to any insecticides since 2008. Based on Ahmad and Arif 64 , resistance level was ranked as none (RF < 2), very low (2 ≤ RF < 11), low (11 ≤ RF < 21), moderate (21 ≤ RF < 51), high (51 ≤ RF < 100), or very high (RF ≥ 100). Maps were created by using the R statistical program version 2.15.2 65 with four add-on packages 'maps' 66 , 'maptools' 67 , 'ggplot2' 68 and GISTools 69 .