Source: UNIV OF MASSACHUSETTS submitted to NRP
EVOLUTION AND SPREAD OF INSECTICIDE RESISTANCE IN THE COLORADO POTATO BEETLE
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
Reporting Frequency
Annual
Accession No.
0193270
Grant No.
2002-35302-12578
Cumulative Award Amt.
(N/A)
Proposal No.
2002-02815
Multistate No.
(N/A)
Project Start Date
Sep 15, 2002
Project End Date
Sep 14, 2005
Grant Year
2002
Program Code
[51.2]- (N/A)
Recipient Organization
UNIV OF MASSACHUSETTS
(N/A)
AMHERST,MA 01003
Performing Department
PLANT, SOIL & INSECT SCIENCE
Non Technical Summary
The Colorado Potato Beetle is a major economic pest of potato and related crops worldwide, and has evolved resistance to most of the insecticides tried against it. Methods to slow the rate that resistance evolves, based on the establishment of pest `refuges' of untreated areas within treated crops, show promise. Susceptible beetles will spread from these refuges to mate with newly resistant beetles, imparting susceptibility to the offspring and forestalling spread of resistance. How well this works depends on several factors, including the genetic basis of resistance, the strength of selection favoring resistance, and the spacing of refuges within the field relative to the beetles' dispersal capabilities. Dispersal is the hardest to measure, but the most critical to the overall success of this strategy. Our goal is to measure dispersal in order to determine the optimal spacing. Our approach is to treat half of a field with insecticide and leave the other half untreated. Resistance evolves rapidly but only in the treated half, creating a sharp, step-like change in the proportion of resistant beetles at the boundary. However, dispersal of beetles within the field blurs this step in a predictable way. Using new mathematical and statistical methods we develop, we measure dispersal from the shape of the step. We then insert the dispersal measure into new models of the rate of spread of resistance in order to design the optimal size and spacing of refuges. This approach, once confirmed, can be used to manage resistance in a wide variety of crop pests.
Animal Health Component
50%
Research Effort Categories
Basic
50%
Applied
50%
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
21131101130100%
Goals / Objectives
0202185: Objectives Objective I: Develop and test a new model of evolution of polygenic traits in non-equilibrium resistance clines. We develop a modular mathematical model that accounts for resistance evolution in the critical life stages of the Colorado Potato Beetle. We then develop a likelihood-based statistical framework for testing hypotheses, and for estimating critical parameters including dispersal rate and the fitnesses of resistant and susceptible genotypes. This objective is largely met; only minor adjustments are needed. Objective II: Measure dispersal and selection in replicate clines in resistance in CPB Objective IIa: Create replicate clines in resistance. It is critical to replicate clines in different fields to measure variability. We address the following questions using likelihood-based analyses: (i) Are there differences in dispersal or selection parameters among replicate clines? (ii) .between MA and ME clines? (iii) .between years? (iv) Does the model that includes leptokurtic dispersal (k) give a better fit? (v) Do the non-equilibrium dispersal estimates fit better than the equilibrium cline approximations? Objective IIb: Calibrate adult and larval dose-response curves. Some of our tests rely on comparing dose-response curves between adult and larval stages, which have different LD50 doses. Once calibrated, common units will be used in all analyses of dose-response curves. Objective IIc: Test for treatment-specific movement rates of adults. It is reasonable to expect that movement might vary with exposure to imidacloprid. If this were true, we might inadvertently recommend suboptimal spacing for refuges. We estimate the difference using two ecological measures of movement: mass-marking and individual mark-recapture. If justified here, we will modify our non-equilibrium model to include treatment-specific movement factors. Objective IId: Measure the fitnesses of resistant and susceptible genotypes: We use cage experiments to test and validate the strength of selection inferred from the fitting the mathematical model. We measure fitness by calculating treatment-specific intrinsic rates of increase using a life table analysis, segregated by pesticide treatment. This captures any fitness consequences of variation in annual generations. The ratio of these rates of increase is the relative strength of selection. The life table analysis will provide important data on the costs of resistance that help maintain the resistance polymorphism. Objective 3: Test whether our estimates of dispersal generalize to other pesticide systems This objective has been dropped due to budget limitations. Objective 4: Use our results for better refuge design recommendations Our current, very rough recommendation is that CPB refuges covering 20% of a field can be spaced ~1200m apart, such that all treated plants are within ~600m of a refuge. We will determine how our recommendation is affected by (i) sampling error in our dispersal estimate; (ii) real variation in dispersal among fields, localities or years, (iii) leptokurtic dispersal, (iv) treatment-sensitive variation in dispersal, or (v) significant costs to resistance.
Project Methods
0202185: Research Approach Objective I: This is largely completed and in the proposal already. Objective IIa: The strategy is to measure dose-response curves before and after critical life stages, then estimate dispersal and selection parameters from the observed changes. We will plant two fields in Western MA and two in ME each season, each 5 acres, 750 meters long (~25 rows wide). We will use Btt sprays if necessary to keep densities below crop-loss thresholds. Sampling: We will collect 300 adults at emergence using drift fences and pitfall traps, and set pane traps and trap plants beyond the field edges to collect long-distance immigrants. We will sample eggs and larvae in each generation. In August, we will sample adults leaving the field using trap plants and drift fences with pitfall traps. Throughout, we will census densities of adults, eggs, and larvae at approximately 5-day intervals. Objective IIb: We use a split-brood design. 10 full-sib families from low and high resistance lines will be assayed for dose-response at 2nd instar and adult stages. The dose response curves of adults will be fit to cumulative normal distributions, then scaled in mean and variance to yield 2nd-instar larval equivalent doses. Objective IIc: We use mass-marking and individual mark-recapture (IMR). Mass-marking: We will release about 1000 adults and, after at the peak of egg laying, samples of 100-150 adults will be collected at each transect site. IMR: 2000 beetles will be uniquely marked early in the spring emergence and colonization period, and released where they were collected. Systematic daily searches of the entire 3-acre field will be made starting from the first release for one week. Analyses: For both methods, recaptures will be fit by likelihood to a diffusion model of spread from the release point, and dispersal-rate differences compared using likelihood ratio tests. Objective IId: We will cage egg masses from resistant and susceptible populations on treated and untreated plants in the field, and monitor mortality, growth rates and body sizes. Upon eclosion, we will mix marked adult survivors and monitor mating frequencies, fecundity and mortality schedules of these and their offspring. These schedules will be analyzed for intrinsic rates of increase using standard methods. We will measure diapause survival in situ using wire-mesh field cages. Objective 4: We will assess this two ways. First, we will modify a new model (in the proposal) to include a sensitivity analysis of each effect, especially simple inheritance patterns. Second, in Mathematica, we will run our non-equilibrium cline model through generations to determine the number of generations until resistance passes a threshold. By varying the starting conditions and dispersal and selection rates, we will repeat the sensitivity analysis using polygenic inheritance. We will then weight the maximum delay of resistance by its likelihood under average experimental cline conditions, to produce likelihood-based confidence limits on our optimal field dimensions.

Progress 10/01/04 to 09/30/05

Outputs
In 2005 we manipulated the evolution of resistance in four fields, three in Freyberg ME and one in Water Mill NY, by treating half of each field with in-furrow imidacloprid at planting. Eggs were collected and hatched in early July. We assayed 25,800 2nd instar larvae for resistance using a 1 microliter drop of imidacloprid dissolved in acetone at doses in the range of 3.0e-6 to 2.0e-4 g/ml, to determine LD50 curves for each clutch at different distances from the treatment boundary. These values are fit by their geographic locations in the field to models of the evolution of resistance, to estimate dispersal and gene flow parameters. Resistance on the treated side of the field evolved to 1.4x the untreated sides in ME and 1.8x in NY, with considerable patchiness within fields. Treated and untreated sides were less resistant than in 2004. Data analysis for evolution of cline shape in the fields is in progress. To estimate overwintering costs of resistance, we used sperm precedence to compare resistance of stored sperm from fall matings in Long Island to sperm from naturally overwintered, spring-emerged males. Larvae of mated females were 2.4x as resistant as larvae of unmated females, suggesting that adults in the fall were 5.9 times as resistant as emergers the following spring. The overwintering cost of carrying a resistant genotype is therefore likely to be quite high. The following undergraduates were trained: Jeff Ahern (UMass), Nehal Galal (CUNY-Queens), Diana Ramroop (Louis Stokes Alliance for Minority Participation fellowship). Also trained: Bushra Wazed, Krystal Lum (Townsend Harris High School, NY).

Impacts
We have trained 8 undergraduate research assistants and 2 high school students. We will publicize our field tests of the costs and benefits of refuges and mixed treatments. The results of our study will lead to specific recommendations for spatial scale of refuges and treated areas, predictions of resistance evolution in Potato beetle to imidacloprid and other insecticides, and a strong recommendation as to whether refuges must be planted adjacent to treated areas or can be planted significant distances from them.

Publications

  • No publications reported this period


Progress 09/15/02 to 09/14/05

Outputs
This research has generated valuable data for our understanding of the evolutionary dynamics of resistance to imidacloprid in Colorado Potato Beetle. These funds permitted us to create clines in pesticide resistance and allow us to measure microevolutionary parameters from the resulting cline shape. They also provided funds for the thorough study of mating preference, overwintering ability, and associated fitness costs for resistant beetles. These funds also supported development of analytical models, and associated computational software (that is still in development) for sophisticated analysis of microevolutionary forces. When data analysis is complete, we will develop the computational models for optimal refuge placement using those resulting dispersal and selection parameters. As this project terminates, we have accumulated large data sets on resistance evolution in 20 treated fields (some funded from other sources), of which 12 showed appropriate responses. In the remainder, 3 were unusable because the resident population had so little background resistance that the population was wiped out on the treated side; we were defrauded by the grower on 3 fields in South Deerfield, MA (who we confirmed had applied pesticide to the entire field, not half), and we believe the remaining 2 fields in ME also were fully treated. We are using other funding sources to increase the sample size of usable fields to 15 before initiating the computationally intensive data-analysis phase. The approach works: preliminary analysis indicates that resistance has evolved on several of these fields; others are ambiguous pending full analysis. We measured reproductive success in of males that had overwintered in diapause, and found significantly lower survivorship in resistant males. This approach takes advantage of the fact that spring females carry sperm from fall matings, and remate as well. Offspring of remated females were significantly less resistant, indicating higher mortality in resistant males. This paper is in review. We measured the inheritance of resistance; partly covered by these funds. We found resistance to be polygenic, and evidence that modifier genes had evolved between 1999 and 2004 that affected hatching success of resistant beetles. Inheritance is a key element in the microevolutionary models. This paper is in review. Our undergraduate honors student at UMass, Jeff Ahern, determined that the beetles were unlikely to be able to distinguish and avoid treated foliage, whether by taste or by the resulting physiological effects of ingestion. He is now a PhD student at Rice University, studying plant-insect interactions.

Impacts
This project has uncovered important fitness costs of resistance that significantly impact its microevolutionary dynamics of resistance. The expected measures of dispersal rates of resistant and susceptible beetles will permit us to parameterize models to optimize the placement of refuges to slow the evolution of resistance. The work trained several undergraduate students in scientific investigation, one of whom is now a PhD student in plant-insect interactions.

Publications

  • Baker, M. B., A. Alyokhin, S. R. Dastur, A. H. Porter & D. N. Ferro. 2005. Sperm precedence in overwintered Colorado Potato Beetle (Coleoptera: Chrysomelidae) and its implications for insecticide resistance management. Annals of the Entomological Society of America 98: 989-995.


Progress 10/01/03 to 09/30/04

Outputs
We planned experiments in replicate fields in ME and MA for this summer, with the pesticide bioassays from both areas carried out in MA. Work went according to schedule in ME, but due to circumstances beyond our control we fell behind in MA. Due to our setback, we will use a no-cost extension to fund the MA work in 2005. Progress towards Objectives: IA: Model development: The likelihood software is being written. IIA: Create replicate clines 4 clines were set up in Fryberg ME, 2 from NRI funds. We collected up to 50 clutches from 22 locations in each field, and assayed 14,582 larvae. The LD50 of treated fields was 1.45 times that of untreated fields. Clines were detectable by eye prior to in-depth analysis via likelihood. IIB: Calibrate dose-response curves. The data are collected and awaiting analysis. IIC: Treatment-specific adult movement These experiments were postponed for a year and will be carried out as a part of the extension. Fecundities (a measure of fitness) of resistant MA and ME populations were 36.8 (SE3.7) and 40.0(3.8) vs. 56.6(3.0) eggs/day in susceptible beetles, but hatching success did not differ. This predicts a stable polymorphism of resistant and susceptible beetles, and explains why resistance has not risen to 100% in MA. IV: refuge design recommendations Co-PI Baker has simulated refuge designs to test their sensitivity to biologically relevant parameters, including variation in field rotation, contiguous refuges versus those separated by varying amounts of low-host-density matrix, and the presence of dispersal polymorphisms. While much depends on individual parameters of the system, a general pattern emerged where optimal design depends on the ratio of dispersal rate to refuge spacing. Conditions that insulate refuges from treated areas sometimes enhance refuge success. The pattern is non-linear; even small deviations from optimal design have major negative impacts on resistance evolution. Refuges should therefore be designed from a risk-averse perspective.

Impacts
We have trained 6 undergraduate research assistants and will train a similar number next year. We will publicize and test in the field the costs and benefits of refuges and mixed treatments. The results of our study will lead to specific recommendations for spatial scale of refuges and treated areas, predictions of resistance evolution in Potato beetle to imidacloprid and other insecticides, and a strong recommendation as to whether refuges must be planted adjacent to treated areas or can be planted significant distances from them.

Publications

  • No publications reported this period


Progress 10/01/02 to 09/30/03

Outputs
Our project creates clines in resistance to imidacloprid by planting fields consisting of imidacloprid treated and untreated sides. The clines will be analyzed to extract dispersal and gene flow data, and develop optimal refuge crop designs to slow the evolution of resistance to imidacloprid and other treatments in Colorado potato beetle and other insect pests. We developed our lab for assaying larger numbers (up to 2,000/day) of Potato beetle larvae and the rearing cages and facilities for maintaining 10 colonies of 150 adults or 400 larvae each. We surveyed imidacloprid resistance in Maine and western Massachusetts (9000 larvae from 5 fields in Maine and 19 fields in Massachusetts) and found similar levels of resistance as in 2001 in Massachusetts, a surprising result given the number of years imidacloprid has been in use. Resistance to imidacloprid in most areas in Maine was much lower than even untreated (with imidacloprid or related insecticides) fields in Massachusetts, but fields near the town of Freyburg, ME, have been located with high levels of resistance that will be ideal locations for creating resistance clines in 2004. We secured a 55-acre field in Massachusetts for the creation of imidacloprid resistance clines on 4 fields separated by roads or corn in 2004. We developed a model of resistance evolution in refuge based systems that incorporates costs of resistance, different crop rotation strategies (rotating crops versus rotating treatments), barriers to movement between the refuge and treated area, and several other factors, in each case asking how variation in dispersal influenced resistance evolution. We found that changing the additivity of resistance, the toxicity of the treatment, the size of the refuge, the costs of resistance, and the initial resistance frequency all greatly affected refuge success, but did not change the optimal spatial scale of treated areas relative to the dispersal of the insect. Different rotation strategies, and barriers to movement between refuges and treated areas, do affect the optimal spatial scale of refuges and treated areas, and these relationships are under continuing investigation. We used a sterile male technique to investigate mating competition and gene flow between resistant and susceptible beetles. Imidacloprid-resistant and susceptible males were placed with either virgin susceptible females on untreated plants, or virgin resistant females on treated plants, for 24 hours, and mating behavior and subsequent hatch rates were recorded. Resistance to imidacloprid does not appear to convey mating costs on untreated foliage. Resistant males mated just as frequently and sired as many offspring. Treated foliage is a strong barrier to gene flow, as resistant males were much more likely to mate and sired the vast majority of offspring when the contests were staged on treated plants.

Impacts
We have trained 6 undergraduate research assistants and will train a similar number next year. Our collaborations with growers will publicize and test in the field the costs and benefits of refuges and mixed treatments. The results of our study will lead to specific recommendations for spatial scale of refuges and treated areas, predictions of resistance evolution in Potato beetle to imidacloprid and other insecticides, and a strong recommendation as to whether refuges must be planted adjacent to treated areas or can be planted significant distances from them.

Publications

  • No publications reported this period


Progress 10/01/01 to 09/30/02

Outputs
Project is new.

Impacts
(N/A)

Publications

  • No publications reported this period