Introduction
The Asian citrus psyllid, Diaphornia citri Kuwayama (Hemiptera: Liviidae), is a global economic pest of citrus. It is a vector of the bacterium Candidatus Liberibacter asiaticus (CLas), the putative causal agent of the disease huanglongbing (HLB) (Halbert and Manjunath, Reference Halbert and Manjunath2004; Bové, Reference Bové2006; Gottwald, Reference Gottwald2010; Grafton-Cardwell et al., Reference Grafton-Cardwell, Stelinski and Stansly2013; Vázquez-García et al., Reference Vázquez-García, Velázquez-Monreal, Manuel Medina-Urrutia, de Jesús Cruz-Vargas, Sandoval-Salazar, Virgen-Calleros and Pablo Torres-Morán2013; Kanga et al., Reference Kanga, Eason, Haseeb, Qureshi and Stansly2016; Stockton et al., Reference Stockton, Pescitelli, Ebert, Martini and Stelinski2017). Vector suppression with insecticides is one of the primary current management practices for this disease (Grafton-Cardwell et al., Reference Grafton-Cardwell, Stelinski and Stansly2013; Kanga et al., Reference Kanga, Eason, Haseeb, Qureshi and Stansly2016; Chen et al., Reference Chen, Gill, Ashfaq, Pelz-Stelinski and Stelinski2018). The short generation time and high fecundity of D. citri have resulted in the development of resistance to several insecticides (Tsai and Liu, Reference Tsai and Liu2000; Kanga et al., Reference Kanga, Eason, Haseeb, Qureshi and Stansly2016; Chen and Stelinski, Reference Chen and Stelinski2017; Chen et al., Reference Chen, Gill, Ashfaq, Pelz-Stelinski and Stelinski2018).
The development of insecticide resistance is an increasing problem for citrus production (Grafton-Cardwell et al., Reference Grafton-Cardwell, Stelinski and Stansly2013; Chen et al., Reference Chen, Seo and Stelinski2017). Resistance has been documented for organophosphates (Tiwari et al., Reference Tiwari, Killiny and Stelinski2013; Vázquez-García et al., Reference Vázquez-García, Velázquez-Monreal, Manuel Medina-Urrutia, de Jesús Cruz-Vargas, Sandoval-Salazar, Virgen-Calleros and Pablo Torres-Morán2013; Kanga et al., Reference Kanga, Eason, Haseeb, Qureshi and Stansly2016), pyrethroids (Tiwari et al., Reference Tiwari, Killiny and Stelinski2013; Kanga et al., Reference Kanga, Eason, Haseeb, Qureshi and Stansly2016), neonicotinoids (Vázquez-García et al., Reference Vázquez-García, Velázquez-Monreal, Manuel Medina-Urrutia, de Jesús Cruz-Vargas, Sandoval-Salazar, Virgen-Calleros and Pablo Torres-Morán2013; Chen et al., Reference Chen, Gill, Ashfaq, Pelz-Stelinski and Stelinski2018), and carbamates (Kanga et al., Reference Kanga, Eason, Haseeb, Qureshi and Stansly2016). Neonicotinoid insecticides have been an important tool in pest control in citrus for well over a decade (Alyokhin et al., Reference Alyokhin, Dively, Patterson, Castaldo, Rogers, Mahoney and Wollam2007). The importance of these insecticides is in part due to their effectiveness against a broad spectrum of insect pests including D. citri. Neonicotinoids are highly systemic and mobile within citrus tissue and used as soil applications for young trees and foliar applications for mature trees. The insecticide resistance action committee classifies neonicotinoids within the chemical sub-group 4A, which acts on the nicotinic acetylcholine receptor (nAChR), thus hindering nerve impulse transmission as a result of a depolarizing effect within the central nervous system of insects (Elbert et al., Reference Elbert, Haas, Springer, Thielert and Nauen2008; Jeschke et al., Reference Jeschke, Nauen, Schindler and Elbert2011; Oliveira et al., Reference Oliveira, Schleicher, Buchges, Schmidt, Kloppenburg and Salgado2011; Salgado, Reference Salgado2016).
Fitness is a measure of the expected reproductive potential of an individual (Kilot and Ghanim, Reference Kilot and Ghanim2012). Differences in the biological parameters affecting the net reproductive rate are of particular interest for insecticide resistance management (Haunbruge and Arnaud, Reference Haubruge and Arnaud2001). It is critical to understand the fitness consequences associated with the development of insecticide resistance. This issue has fundamental implications for evolutionary responses to stress and immediate applications for efforts to manage resistance (Hollingsworth et al., Reference Hollingsworth, Tabashnik, Johnson, Messing and Ullman1997). Because resistant insects are typically not present at high frequency before frequent use of insecticides, it is generally assumed that resistance genotypes must have a reproductive disadvantage in the absence of selection pressure (Haubruge and Arnaud, Reference Haubruge and Arnaud2001). If resistance genes have associated negative consequences on fitness in the absence of insecticide, then the frequency of resistant phenotypes should decline when insecticide pressure is reduced (Arnaud and Haubruge, Reference Arnaud and Haubruge2002). Fitness costs associated with insecticide resistance have been reported for many classes of insecticides and insect species, including Nilaparvata lugens (Stål) (Hemiptera: Delphacidae) (Puinean et al., Reference Puinean, Denholm, Millar, Nauen and Williamson2010), Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae) (Feng et al., Reference Feng, Wu, Xu, Wang, Chang, Xie and Zhang2009), Plutella xylostella (L.) (Lepidoptera: Plutellidae) (Chen et al., Reference Chen, Sanada-Morimura, Yanagi and Nakasuji2006; Ribeiro et al., Reference Ribeiro, Wanderley-Teixeira, Ferreira, Teixeira and Siqueira2014), Musca domestica (L) (Diptera: Muscidae) (Abbas et al., Reference Abbas, Khan and Shad2015, Reference Abbas, Shah, Shad and Azher2016a, Reference Abbas, Shah, Shad, Iqbal and Razaq2016b), Spodoptera litura (Fabricius) (Lepidoptera: Noctuidae) (Abbas et al., Reference Abbas, Shad and Razaq2012, Reference Abbas, Shad, Razaq, Waheed and Aslam2014), Heliothis virescens (Fabricius) (Lepidoptera: Noctuidae) (Sayyed et al., Reference Sayyed, Ahmad and Crickmore2008), and Phenacoccus solenopsis (Tinsley) (Hemiptera: Pseudococcidae) (Afzal et al., Reference Afzal, Shad, Ayyaz and Walker2015).
Understanding the biological parameters associated with insecticide resistance can improve integrated resistance management; however, this information has been lacking for D. citri. Therefore, we compared fitness and morphological characteristics between three populations of D. citri from commercially managed citrus groves in central Florida that exhibited resistance to neonicotinoid insecticides (thiamethoxam and imidacloprid) and two susceptible populations reared either in the absence (CLas−) or presence (CLas+) of the CLas pathogen.
Materials and methods
Insects
A susceptible laboratory population of D. citri was reared in a greenhouse at the Citrus Research and Education Center (CREC), University of Florida, Lake Alfred, FL. The culture was established in 2000 from field-collected insects in Polk County, Florida (27°86′ N, 81°69′ W) prior to the discovery of HLB in the state. This strain has been reared without exposure to insecticides or subsequent input of field-collected D. citri for approximately 320 generations. CLas− D. citri were collected from a subculture free of CLas that was tested monthly using a quantitative real-time polymerase chain reaction (qPCR) assay for confirmation (Pelz-Stelinski et al., Reference Pelz-Stelinski, Brlansky, Ebert and Rogers2010). The colony was maintained on sweet orange (Citrus sinensis (L) Osbeck) ‘Valencia’ in a temperature controlled greenhouse at 27–28°C, with 60–65% relative humidity and a 14:10 h (light:dark) photoperiod. CLas+ D. citri were obtained from a population reared on CLas-infected C. sinensis ‘Valencia’ plants housed in a separate facility at the University of Florida, CREC. The infection status of these insects was confirmed as described below and the infection rate was 40%. All plants were 2–4 years of age.
The field populations of adult D. citri were collected from commercial citrus orchards in central Florida located in: Davenport (N: 28°09′933″; W: 81°37′907″), Clermont (N: 28°26′986″; W: 81°34′951″), and Lake Wales (N: 27°57′759″; W: 81°34′951″) in 2017. About 3000 adults were collected from each location. Growers sprayed insecticides monthly for psyllid control. Insects were collected using a D-Vac insect suction sampler (Rincon-Vitova Insectaries, Ventura CA). Both the Davenport and Lake Wales populations had a CLas infection rate of 17%, while the infection rate in the Clermont population was 20% as determined by the qPCR. Tested insecticides were commercial formulations and included Admire Pro 4.6F (imidacloprid, LC) and Actara (thiamethoxam, SG). Admire pro 4.6F was obtained from Bayer Crop Science, USA and Actara was obtained from Syngenta, USA.
Adult feeding bioassay
An artificial diet (Hall et al., Reference Hall, Shatters, Carpenter and Shapiro2010; Langdon and Rogers, Reference Langdon and Rogers2017) was made by using 100 ml deionized water, 60 g sucrose (w/v; Fisher Scientific, Fair Lawn, NJ, Cat. No: S5-500), 0.2 ml green food dye (0.1% v/v; McCormick & Co., Inc. Hunt Valley, MD), and 0.8 mL yellow food dye (0.4% v/v; McCormick & Co., Inc. Hunt Valley, MD). The food solution was heated and mixed with a magnetic stirrer. When the sucrose was completely dissolved, deionized water was added to bring the final volume up to 100 ml. Aliquots of the stock 30% sucrose solution were then used to make serial dilutions of the insecticides. Caps were removed from 8 ml centrifuge tubes (Eppendorf Tubes, Hamburg, Germany, Cat. No: 033381D); the approximate dimensions of the 8 ml tubes were 1.3 cm × 5.5 cm. Five hundred and fifty microliters of the sucrose solution were added to each centrifuge tube cap and a 1.5 cm2 piece of Parafilm M® (Bemis®, Neenha, WI) was stretched in both directions and placed over the treatment cap. While stretching the Parafilm tight, the excess Parafilm was wrapped over the back side of the cap forming a feeding membrane. Four to six adult D. citri were aspirated into the individual centrifuge tubes and the treatment filled cap was then pressed onto the centrifuge tube to allow feeding. The tubes were placed upright in a tray and held at 25°C on a 14:10 h light:dark photoperiod for 72 h. Each insecticide was tested over a range of concentrations (0, 0.001, 0.01, 0.1, 1, 10, 100, and 1000 ng μl−1) and each concentration was replicated four times. Approximately 300 D. citri adults were tested from each resistant and susceptible population for each insecticide using the feeding bioassay. The mortality was determined 72 h after treatment. An insect was considered dead if there was no movement after being touched with a probe.
Effect of insecticide resistance on Asian citrus psyllid fertility and survivorship
One hundred and fifty to 200 mature nymphs reared from each test strain were collected and caged on potted plants. Adults that emerged within four days were aspirated into cylindrical cages (diameter: 110 mm; height: 130 mm). For each experiment, 100 mixed sex adults from the above colonies were transferred to a group of eight potted ‘Swingle’ C. aurantifolia (Chrism) plants (1 year old) with new leaf flush (12 cm × 11 cm) for a 24-h oviposition period. At the end of this period, the adults were removed from the plants and the number of eggs was counted using a stereomicroscope (Leica, Wild M3C, Leica Microsystems Inc, Buffalo Grove, IL, 6.4X). Following egg hatch, the fertility of each strain was determined at the onset of the experiment. The plants were returned to the growth chamber at 25°C, 60% RH, and a 14:10 h light:dark photoperiod. Thereafter, eggs and nymphs were counted daily. The nymphal instars were identified according to the size of the insect body and development of wing pads (Tsai and Liu, Reference Tsai and Liu2000). Survival of psyllids was recorded for each life stage: eggs, first-second-instar nymphs, third-instar nymphs, fourth-fifth-instar nymphs, and adults to determine cumulative mortality (K-value) of pysllids in response to pathogen infection and insecticide resistance. This was calculated as:
$$K\,= -{\rm ln}\,\lpar s\rpar \,{\rm and}\,K\,= \,\sum k_{\rm i}$$where k was the negative natural logarithm of survival (s) for each life stage and K was the sum of all k-values for the entire life cycle. The magnitude of the K-value reflects the risk of mortality for a treatment group, such that mortality of a group increases as k-values increase (Pelz-Stelinski and Killiny, Reference Pelz-Stelinski and Killiny2016).
Effect of insecticide resistance on fecundity and longevity
The objective of this experiment was to determine the effect of insecticide resistance or CLas infection on D. citri fecundity and longevity. Five colonies of D. citri were compared; two susceptible laboratory cultures (CLas− and CLas+) and three field populations exhibiting insecticide resistance and varying levels of CLas infection. Twenty pairs of virgin adults from each population were sexed and then transferred as male and female pairs onto CLas− ‘Swingle citrumelo’ (Citrus paradisi Macf. × Poncirus trifoliata L. Raf) plants. Citrus plants with psyllids were held in an environmental chamber at 25°C and 60% RH under a 14:10 h (light:dark) photoperiod to reflect typical field conditions. Egg production was determined by counting the total number of eggs laid by each female for 70 days. Eggs were counted and adults were transferred to new plants at 3–5 days intervals to promote feeding and oviposition on new flush throughout the experiment. To determine total egg deposition per female, leaf flush was removed with a sterile scalpel and the number of eggs was counted using a stereomicroscope.
Population growth and reproductive rate
Data were analyzed as stage dependent life tables (Birch, Reference Birch1948; Pelz-Stelinski and Killiny, Reference Pelz-Stelinski and Killiny2016; Chen et al., Reference Chen, Seo and Stelinski2017). Life tables were constructed from the cohort of eggs laid by the same females on the same day. Twenty females were used for each treatment. The net reproductive rate (R 0) was calculated as the number of female progeny produced per female per generation, assuming a 1:1 sex ratio, as per the following equation:
$$R_0\,= \,\sum \,\lpar 1_xm_x\rpar $$x: time (days); lx: proportion of females alive at time x, and; mx : age-specific fecundity (average daily number of eggs laid by females per treatment divided by 2 to compensate for the 1:1 sex ratio of progeny).
The intrinsic rates of population increase for CLas+ and CLas−, susceptible psyllids, as well as, field-collected, insecticide resistant populations of varying infection status were calculated as described by Birch (Reference Birch1948) and Chen et al. (Reference Chen, Seo and Stelinski2017) as: r m = ln R 0/T
T is the generation time (in days) calculated as:
$$T\,= \,\displaystyle{{\sum \,x{\rm l}_xm_x} \over {\sum l_xm_x}}$$The finite rate of population increase representing the number of females produced per female per day was calculated (Birch, Reference Birch1948; Pelz-Stelinski and Killiny, Reference Pelz-Stelinski and Killiny2016) as:
$${\rm \lambda }\,= \,{\rm exp}\,\lpar r_m\rpar $$
$$\lambda \,= \,\displaystyle{{ R_0^1 } \over {^{\sum x\,\lpar l_xm_x\rpar /R_0} }}$$The relative fitness (Rf) was calculated by the following method (Cao and Han, Reference Cao and Han2006; Afzal et al., Reference Afzal, Shad, Ayyaz and Walker2015):
Rf = R 0 of experimental population/R 0 of susceptible population.
Adult weight
The weight of male and female adults from the three field populations, as well as, the CLas− and CLas+, laboratory susceptible populations were recorded using a Mettler AE 160 (Mettler-Toledo Columbus OH USA) balance. D. citri were sampled at random from each population obtaining the following sample sizes: laboratory Clas− (male: 28; female: 41), laboratory Clas+ (male: 20; female: 20), Clermont (male: 68; female: 65), Davenport (male: 43; female: 45), and Lake Wales (male: 16; female: 6).
Morphological measurements
Morphological measurements for adult D. citri were made using a stereomicroscope. Insects were killed by freezing (−20°C for 30 min) and held using a piece of double-sided sticky tape. Body length, abdominal length, wing length, femur length, and head width were measured for two laboratory susceptible cultures and three field populations at 10 × magnification with a 1.2 mm ocular ruler. At least 40 individuals from each population were measured.
DNA isolation and real time PCR assays
Deoxyribonucleic acid (DNA) was isolated from D. citri using a DNeasy Blood and Tissue kit (Qiagen, Valencia, CA) according to the manufacturer's protocol with a modification for the isolation of bacterial DNA from arthropods (Li et al., Reference Li, Hartung and Levy2006; Pelz-Stelinski et al., Reference Pelz-Stelinski, Brlansky, Ebert and Rogers2010). A Las specific 16S ribosomal DNA probe (5′-FAMAGACGGGTGAGTAACGCG-3BHQ-3′) and primers (LasF: 5′TCGAGCGCGTATGCAATACG-3′; LasR: 5′-GCGTTATCCCGTAGAAAAAGGTAG-3′) were used in qPCR assays to detect CLas (Li et al., Reference Li, Hartung and Levy2006; Pelz-Stelinski and Killiny, Reference Pelz-Stelinski and Killiny2016). In addition to target DNA, internal control primers were used in multiplex qPCR amplification of samples, as described previously (Pelz-Stelinski et al., Reference Pelz-Stelinski, Brlansky, Ebert and Rogers2010). Each reaction tube contained a primer and probe set to amplify the psyllid wingless gene (Wg) [(WgF: 5′GCTCTCAAAGATCGGTTTGACGG-3′; WgR: 5′-GCTGCCACGA ACGTTACCTTC-3′), Wg probe (5′-JOE-TTACTGACCATCAC TCTGGACGC-3BHQ2-3′)]. The quantitative PCR scheme for all assays consisted of 2 min at 50°C, 10 min at 95°C, and 3 min 40 cycles with 15 s at 95 and 60°C 1 min. Samples were considered positive for CLas if a product was amplified within the 40 amplification cycles used for reactions.
Statistical analysis
Mortality data of D. citri were subjected to Probit analysis using SAS software (SAS Institute Inc, Cary, 9.4, 2002–2012, NC, USA) for the determination of LC50 values and their 95% fiducial limits (FLs) (Finney, Reference Finney1971). The data were corrected for control mortality using Abbott's formula (Abbott, Reference Abbott1925). In order to determine the best response metric for calculating RRs, the relative precision of 95% FLs for LC50 values was calculated as the width of the 95% FL divided by the LC50 values obtained from the CLas− population. The RRs and their 95% FL were calculated for each field-collected and laboratory population according to Robertson and Preisler (Reference Robertson and Preisler1992) and Bilbo et al. (Reference Bilbo, Reay-Jones, Reisig and Greene2019) using the laboratory CLas− colony as the susceptible reference.
Insect survival was analyzed using the Kaplan–Meier method and pairwise comparisons between treatment groups were made using the log rank (Mantel Cox) test. The psyllid weight abdominal length, femur length, head width, and wing length were analyzed using an analysis of variance (ANOVA) with sex and colony as fixed effects (PROC GLM; SAS institute, 2002–2012).
Female longevity, number of eggs per female, finite rate of population increase, and net reproductive rate were compared between groups of the three insecticide resistant strains and the CLas− and CLas+ susceptible strains using a one-way ANOVA (α = 0.05) followed by Tukey's mean separation (PROC GLM; SAS Institute, 2002–2012). Population growth rate, reproductive rate, fecundity, fertility, and morphological measurements were dependent variables in linear regression models with the resistance ratio as the independent variable (PROC GLM; SAS Institute, 2002–2012). The models' assumption of equality of variance was assessed using a plot of residuals vs. predicted values. The assumption of normality in the residuals was assessed using a q–q plot. The independent variable was log transformed for ANOVA and both the dependent and independent variables were log transformed for regression.
Results
Susceptibility of laboratory and field populations to thiamethoxam and imidacloprid
All field populations exhibited more resistance than either of the laboratory susceptible colonies (Lab CLas−: 0.22–1.52 ng μl−1; Lab CLas+: 0.11–0.68 ng μl−1) except for the Clermont population to imidacloprid (0.9–1263 ng μl−1) (table 1 and fig. 1). The Lake Wales, Clermont, and Davenport populations exhibited resistance to imidacloprid, with resistance ratios ranging from 7.65–16.11. For thiamethoxam, resistance ratios in the field populations ranged from 26.79–49.09 compared with the susceptible CLas− laboratory population.
Figure 1. Dose-response curves for D. citri populations treated with serial dilutions of thiamethoxam or imidacloprid in water. Abbreviations identify geographic origins of each population (LB−, Laboratory CLas−; LB+, Laboratory CLas+; CM, Clermont; DP, Davenport; LW, Lake Wales).
Table 1. Susceptibility of D. citri adults to thiamethoxam and imidacloprid as measured by artificial diet feeding bioassay
a LC50 values generated by SAS exceed the maximum concentration while actual mortality was lower.
bResistance ratios calculated using LC50 from field populations or CLas+ colony divided by LC50 of CLas− susceptible laboratory colony.
Robertson and Preisler (Reference Robertson and Preisler1992), Bilbo et al. (Reference Bilbo, Reay-Jones, Reisig and Greene2019)
Fecundity and fertility
There was a significant relationship between resistance ratio and fecundity, with a reduction in eggs with increasing resistance ratio for both insecticides (table 2). There was a significant relationship between resistance ratio and fertility, with a reduction in egg survival with increasing resistance ratio for both insecticides (table 2).
Table 2. Regressions of fecundity and fertility against the thiamethoxam and imidacloprid resistance ratio.
a Models with fertility as dependent variable were log transformed. Models with fecundity as dependent variable were square root transformed; fertility: df = 1, 38; fecundity: df = 1, 98.
Survival
Survival of first (F = 0.33; df = 4, 35; P = 0.86), second (F = 0.33; df = 4, 35; P = 0.53), third (F = 0.53; df = 4; 35; P = 0.99), and fourth (F = 0.99; df = 4, 35; P = 0.42) instar nymphs did not differ between the three field populations and the susceptible populations (table 3). Survival of fifth instar nymphs from the Davenport population was significantly higher (F = 2.74; df = 4, 35; p = 0.044) than that observed in all of the other populations measured (table 3).
Table 3. Survival rate of immature stages of D. citri from three field resistant and laboratory susceptible populations.
a Means followed by the same letter within a row are not significantly different.
Population growth and reproductive rate
The net reproductive rate of the Lake Wales field population was significantly lower (F = 6.1; df = 4, 93; P < 0.001) than that for all other field and laboratory populations (table 4). Similarly, the finite rate of population increase was significantly lower (F = 5.67; df = 4, 93; P < 0.001) for the Lake Wales population as compared to all others measured (table 4). There were no significant differences between the populations for adult longevity (F = 0.86; df = 5, 58; P = 0.50); however, CLas− and CLas+ laboratory females laid significantly more (F = 3.26; df = 5, 58; P = 0.01) eggs than females from the Lake Wales field population (table 4).
Table 4. Fitness parameters (± SE) of thiamethoxam and imidacloprid resistant field populations of D. citri as compared with laboratory controls
a SS: susceptible line; RR: insecticide resistant line.
b Means followed by the same letter within a row are not significantly different.
c Rf = R 0 of experiment population/R 0 of laboratory susceptible CLas− population.
Adult weight
The source of collected insects (laboratory susceptible and field collected populations) and sex were significant independent variables in a model predicting ln(weight) (F = 11.2; df = 9, 342; P < 0.001). Adult weight (F = 9.23; df = 4, 342; P < 0.001) and sex (F = 58.25; df = 1, 342; P < 0.001) varied among colonies with a non-significant interaction term (F = 1.09; df = 4, 342; P = 0.362) that was removed from the model (fig. 2). While males were smaller than females and size differed between populations, the size difference between males and females did not vary depending on source population. Examining population differences for males using a Tukey test, we found that males from the CLas−, laboratory population weighed more than CLas+, laboratory males, and those from the Clermont population. There were no statistical differences in female weight between the populations examined.
Figure 2. Mean weight (±SE) of adult D. citri from laboratory susceptible and three insecticide resistant field populations (LB−, Laboratory CLas−; LB+, Laboratory CLas+; CM, Clermont; DP, Davenport; LW, Lake Wales) (SE < 0.0009).
Morphological characteristics
There were no significant differences between the populations for either head width or femur length (table 5). The CLas+ laboratory population had significantly shorter wings than the Clermont population, but there were no other differences in the wing length. Psyllids from Lake Wales had shorter body length than any other population, but D. citri from both Lake Wales and Clermont had shorter abdomens than the other four populations. There was a significant effect of sex for all measurements except femur length and in all cases, females were larger than males (table 5).
Table 5. Morphometric comparison of five adult body characteristics (± SE) of D. citri from three field resistant and two susceptible populations
Means for each colony followed by a different letter are significantly different by the Tukey test. ANOVA model P-values with sex and population as fixed effects are given below the LS means with df = 9, 200. The P-values for each variable follow.
Discussion
We detected reduced susceptibility to neonicotinoids among the tested D. citri field populations (table 1 and fig. 1). In general, resistance was more pronounced with thiamethoxam than with imidacloprid and occurred at all sampled locations. Furthermore, the CLas+ susceptible colony was more sensitive to the insecticides evaluated than the CLas− laboratory population. This is consistent with earlier surveys of D. citri populations in Florida (Chen et al., Reference Chen, Seo and Stelinski2017; Langdon and Rogers, Reference Langdon and Rogers2017). Candidatus Liberibacter asiaticus infection increased the susceptibility of D. citri to insecticides relative to CLas− D. citri and this result is also consistent with previous investigations (Tiwari et al., Reference Tiwari, Lewis-Rosenblum, Pelz-Stelinski and Stelinski2010). nAChRs are ligand-gated ion channel receptor complexes that mediate fast cholinergic synaptic transmission. nAChRs play a central role in the mediation of fast excitatory synaptic transmission in the insect central nervous system and are also the targets of commercially important classes of insecticides (Salgado and Saar, Reference Salgado and Saar2004; Oliveira et al., Reference Oliveira, Pippow, Salgado, Büschges, Schmidt and Kloppenburg2010). Neonicotinoids are agonists that mimic the action of acetylcholine and therefore have widespread use against a broad spectrum of sucking and certain chewing insect pests (Oliveira et al., Reference Oliveira, Schleicher, Buchges, Schmidt, Kloppenburg and Salgado2011; Salgado, Reference Salgado2016). Following instances of misuse or overuse, nicotinic insecticide resistance can contribute to potentially reduced performance of these insecticides (Oliveira et al., Reference Oliveira, Schleicher, Buchges, Schmidt, Kloppenburg and Salgado2011; Salgado, Reference Salgado2016).
Several biological parameters were compared between D. citri that exhibited resistance to thiamethoxam and imidacloprid in the field vs. their counterparts from known laboratory susceptible cultures. There was a significant relationship between resistance ratio and fecundity with egg production reduced proportionally with increasing resistance ratio for both insecticides (table 2). As indicated in tables 3 and 4, both laboratory susceptible populations exhibited higher egg to adult survival than field resistance populations. Furthermore, the egg to adult survival rate in the Lake Wales population was between 2.6 and 3.4-fold lower than that of the susceptible populations. Higher mortality of resistant than susceptible D. citri is an indicator of a fitness disadvantage. In this case, both fecundity and egg survival were lower in populations exhibiting insecticide resistance. Collectively, our results indicate that resistance to thiamethoxam and imidacloprid in D. citri corresponded with a developmental disadvantage and suggest a trade off between the distribution of resources toward detoxification and fitness. Analogous fitness costs associated with insecticide resistance have been reported previously (Liu and Han, Reference Liu and Han2008; Feng et al., Reference Feng, Wu, Xu, Wang, Chang, Xie and Zhang2009; Abbas et al., Reference Abbas, Shad and Razaq2012; Kilot and Ghanim, Reference Kilot and Ghanim2012).
The finite rate of increase provides an estimate of insect population growth potential (Stark and Banks, Reference Stark and Banks2003), which provides broader insight than individual life history parameters. Net reproduction is not, however, the only component needed to assess the potential of population growth because the finite rate of increase depends on fecundity, percentage egg hatch, and growth (Stark and Banks, Reference Stark and Banks2003). In this study, the finite rate of increase in resistant populations was 23–43% lower than that of the laboratory susceptible (CLas−) population. This type of difference in net reproductive and finite rate of increase has also been observed between susceptible and resistant populations of Tetranychus urticae Koch (Acari: Tetranychidae) and is known to contribute the instability of milbemectin resistance (Nicastro et al., Reference Nicastro, Sato and da Silva2010). Similarly, imidacloprid resistance is associated with a lower net reproductive rate in Spodoptera litura (Fabricius) compared with laboratory susceptible counterparts (Abbas et al., Reference Abbas, Shad and Razaq2012). The decline in population growth of D. citri observed here appeared to be mainly due to the decreased fecundity and hatch rate.
Decreased Rf associated with insecticide resistance has been reported previously in various insect taxa and for various chemistries including thiamethoxam resistant B. tabaci (Rf = 0.53) (Feng et al., Reference Feng, Wu, Xu, Wang, Chang, Xie and Zhang2009); imidacloprid resistant S. litura (Rf = 0.38) (Abbas et al., Reference Abbas, Shad and Razaq2012) and two populations of N. lugens (Rf = 0.17 and 0.10) (Liu and Han, Reference Liu and Han2008) and acetamiprid resistant P. solenopsis (Rf = 0.22) (Afzal et al., Reference Afzal, Shad, Ayyaz and Walker2015). The present study indicated that the Rf of the resistant Lake Wales population was 0.56 of the laboratory susceptible (CLas−) population (table 4). Our results are thus congruent with the general convention that insecticide resistant populations exhibit reduced fitness compared with susceptible counterparts (Denholm and Rowland, Reference Denholm and Rowland1992).
In contrast, there are instances where limited fitness costs have been associated with insecticide resistance (Haubruge and Arnaud, Reference Haubruge and Arnaud2001; Chen and Nakasuji, Reference Chen and Nakasuji2004; Baker et al., Reference Baker, Alyokhin, Porter, Ferro, Dastur and Galal2007; Bielza et al., Reference Bielza, Quinto, Gravalos, Abellan and Fernandez2008) or certain modifiers can compensate for resistance costs (Coustau et al., Reference Coustau, Chevillon and ffrench-Constant2000). The lack of quantifiable fitness costs related to resistance typically occurs under specific environmental conditions that would not be detected under experimental laboratory conditions (Bourguet et al., Reference Bourguet, Guillemaud, Chevillon and Raymond2004). Also, certain pleiotropic effects might not be detected with current methods to measure costs dependent on body size and reproduction (Fry, Reference Fry1993).
Given the interaction between D. citri and CLas and its associated consequences on insect fitness, the Rf costs associated with insecticide resistance cannot be investigated in D. citri without considering this additional interaction. Acquisition of phytopathogens by herbivorous insect vectors can have population scale fitness consequences. Several leafhopper vectors exhibit enhanced fitness following acquisition of plant pathogens. Beanland et al. (Reference Beanland, Hoy, Miller and Nault2000) and Ebbert and Nault (Reference Ebbert and Nault2001) reported that leafhoppers exposed to plant pathogens live longer than their uninfected counterparts. Pelz-Stelinski and Killiny (Reference Pelz-Stelinski and Killiny2016) indicated that fecundity of CLas+ D. citri is greater than that of uninfected counterparts and that there is a consequential trade off associated with CLas acquisition, which decreases life span compared with uninfected counterparts. Our results were congruent with previous investigations in that more adult female D. citri were produced per day by CLas+, as compared, with CLas− counterparts (Pelz-Stelinski and Killiny, Reference Pelz-Stelinski and Killiny2016). Propagative pathogens can impose metabolic or immune costs on their hosts associated with multiplication (Nault, Reference Nault1997). The hormonal regulation of immune function and metabolic allocation likely underlie life history trade-offs (Harshman and Zera, Reference Harshman and Zera2007). Our results confirm that acquisition of CLas by D. citri causes a fitness costs, which should be compounded by the negative fitness consequences associated with insecticide resistance in this species.
In summary, our investigation indicates that CLas infection increased susceptibility of D. citri to both thiamethoxam and imidacloprid. We then showed that field-collected populations of D. citri exhibited resistance to these insecticides. The net reproductive rate of resistant populations declined with increasing resistance ratio. An additional cost of resistance appears to be smaller body size in D. citri. The smaller size of resistant D. citri may result in reduced maximum dispersal distance and reduced reproduction (Langellotto et al., Reference Langellotto, Denno and Ott2000; Baguette and Schtickzelle, Reference Baguette and Schtickzelle2006; Lewis-Rosenblum et al., Reference Lewis-Rosenblum, Martini, Tiwari and Stelinski2015; Pelz-Stelinski and Killiny, Reference Pelz-Stelinski and Killiny2016; Ejsmond et al., Reference Ejsmond, McNamara, Søreide and Varpe2018; Villa et al., Reference Villa, Evans, Subhani, Altuna, Bush and Clayton2018). Further research is needed to identify cause and effect mechanisms vs. correlations between insecticide resistance in D. citri and fitness costs. Our investigation also suggests that the current widespread problem of insecticide resistance to neonicotinoids among populations of D. citri across Florida (Chen et al., Reference Chen, Gill, Ashfaq, Pelz-Stelinski and Stelinski2018) should be effectively mitigated by ceasing thiamethoxam and imidacloprid use for a period of at least 6 months (Chen et al., Reference Chen, Gill, Ashfaq, Pelz-Stelinski and Stelinski2018). Additional resistance management should include rotations of at least six insecticide modes of action to diversify selection forces and complicate the evolution of resistance (Chen and Stelinski, Reference Chen and Stelinski2017).
Acknowledgements
This project was supported by a grant from the Citrus Research and Development Foundation to LLS. The authors are grateful to Wendy L. Meyer, Angelique B. Hoyte, Kayla M. Kempton, and Rosa B. Johnson, Hunter K. Gossett, Kristin A. Racine for technical assistance.
