General Information: Project Title: Comparing Populations of Western Corn Rootworm (Coleoptera: Chrysomelidae) in Regions With and Without a History of Injury to Cry3 Corn Data Title: County comparison data set Abtract: Data was collected between 2015 and 2018. In this study, we examined fields in counties where greater-than-expected injury to Cry3 (Cry3Bb1 or mCry3A) corn roots (>1 node) had previously been reported (problem counties) and counties where injury had not been reported (non-problem counties). Four to eight fields were sampled per county in 2015, 2016, and 2017 to quantify rootworm abundance, root injury, Cry3Bb1 resistance, and rootworm management strategies by farmers (i.e., cultivation of corn, type of corn that was grown, and insecticide use over the previous 6 years). Additionally, a comparison of root injury to Bt and non-Bt corn within fields was made. Files are organized by analysis of interest, and each data file is paired with a text file of accompanying SAS code, which was used for analysis. Analyses are: survival data from bioassays to test for resistance to Cry3Bb1, analysis of corrected survival in bioassays, analysis of proportion third instar larvae in bioassays, analysis of field history variables, analysis of root injury and adult rootworm abundance in fields, and an analysis comparing root injury to Bt and non-Bt corn. Authors: Coy R. St. Clair (Corresponding author) Iowa State University cstclair@iastate.edu Graham P. Head Bayer Crop Science graham.head@bayer.com Aaron J. Gassmann Iowa State University aaronjg@iastate.edu Associated publication: "Comparing Populations of Western Corn Rootworm (Coleoptera: Chrysomelidae) in Regions With and Without a History of Injury to Cry3 Corn" Journal of Economic Entomology, in press Collection Information: Field data collected summer 2015, 2016, and 2017. Bioassays were performed in spring 2016, 2017, and 2018. Fields were located in Delaware, Winneshiek, Fayette, Jasper, Greene, and Taylor counties, Iowa. FILES: NOTE: for all fields, 0=true zero (measurement). A period "." is null or missing data. File list: 1. Bioassay - corrected survival data.csv - Corrected survival raw data Number of variables:4 Variable name (identity):pop (population name), year (year), Type (population type), corrected (corrected survival measurement) Notes on variables: Years: 2015, 2016, 2017. Population types: Control, PC [problem county], NPC [non-problem county]. Corrected: numerical measurement. Corrected survival was calculated from survival (found in "Bioassay - survival data.csv") as the proportion of larvae surviving on Cry3Bb1 corn in a replicate ÷ mean proportion surviving larvae on non-Bt corn in all replicates of a population. The purpose of this was the account for variation in survival in control populations. 2. Bioassay - corrected survival SAS code.txt - Corrected survival SAS code for analysis 3. Bioassay - prop third instar data.csv - Proportion third instar larvae in bioassays raw data Number of variables:6 Variable name (identity):pop (population name), corn (corn type), Number (dentifier), #survived (number of total number larvae surviving assay), #3rd instar (number of larvae that survived to third instar), Prop (#3rd instar divided by #survived). Notes on variables: corn types: Iso [isoline], Bt 4. Bioassay - prop third instar SAS code.txt - Proportion third instar larvae SAS code for analysis 5. Bioassay - survival data.csv - Larval survival on bt and non-bt corn in bioassays raw data Number of variables:6 Variable name (identity): Population (population name), Fieldtype (Field type), Year (year), Corn type (corn type), Survival (number of larvae surviving the assay), Total infested (total number of larvae placed on the plant at the beginning of the assay). Notes on variables: Field types: Control, PC [problem county], NPC [non-problem county]. Year: 2015, 2016, 2017. Corn types: Iso [isoline], Bt 6. Bioassay - survival SAS code.txt - Larval suvival on bt and non-bt corn SAS code for analysis 7. Field history analysis data.csv - field history variables raw data Number of variables: 25 Variable name: Field Farm Year Corn nonBt Cry3 Cry3435 Pyramid FH1 FH2 FH3 FH4 FH5 FH6 FH7 FH8 FH9 FH10 FH11 FH12 Category Root WCR NCR Both Notes on variables: Farm: Identifies the farm where the field in located, as some fields were managed by the same farmer Year: 2015, 2016, 2017 Corn: Corn type planted in the field in the year of sampling, key: Non-RW Bt or Non-Bt: 0 Cry3Bb1: 1 Cry34/35Ab1: 2 mCry3A: 3 Cry34/35Ab1 + Cry3Bb1: 4 Cry34/35Ab1 + mCry3A: 5 Variables nonBt, Cry3, Cry3435, and Pyramid are binary (0=absent or 1=present) representation of information found in the "Corn" variable. "FH" is short for "Field History" over the past six years from the time the field was sampled. Key for these variables: FH1 = # years continuous corn FH2 = Proportion of years corn was planted in the field history FH3 = Was the field rotated in 6 yrs (1=yes, 0=no) FH4 = Was soil insecticide used in the year of sampling (1=yes, 0=no) FH5 = Proportion of years soil insecticide was used in the field history FH6 = Proportion of years non-Bt corn was planted FH7 = Proportion of years Cry3 corn was planted (either Cry3Bb1 or mCry3A) FH8 = Proportion of years Cry34/35Ab1 was planted FH9 = Proportion of years Bt corn was planted (any trait or pyramid) FH10 = Proportion of years a pyramid was planted FH11 = Proportion of years Bt corn was planted with soil insecticide FH12 = Proportion of years non-Bt corn was planted with soil insecticide Category variable is: 0=non-problem county, 1=problem county. Root: root injury measurement for the field WCR: abundance of western corn rootworm NCR: abundance of northern corn rootworm Both: total of western and northern corn rootworm 8. Field history analysis SAS code.txt - field history analysis in SAS 9. Root injury and abundance data.csv - Root injury and adult abundance from the field raw data Number of variables: 7 Variable names: Field Farm Year Category Root WCR Corrected Notes of variables: Field, Farm, Year, Category variables same as above. Root=root injury measurement for the field WCR=abundance of adult western corn rootworm in the field Corrected=corrected survival for those population used in bioassays 10. Root injury and abundance SAS code.txt - Root injury and adult abundance from the field SAS code for analysis 11. Root injury to bt and non-bt corn data.csv - Comparison of root injury to bt corn and injury to non-bt (refuge plants), raw data Number of variables: 5 Variable names: Field,year, type (field type), root(bt), root(refuge) Notes on variables: Year and Type are same as above. Root(Bt)=mean root injury to Bt plants in the field Root(refuge)=mean root injury to non-Bt (refuge) plants in the field. 12. Root injury to bt and non-bt corn SAS code.txt - Comparison of root injury to bt corn and non-bt, SAS code for analysis Methods and materials: Counties in Iowa, USA, were selected based on the presence or absence of fields where greater-than-expected injury (>1 node of injury [U.S. EPA 2011]) to Cry3Bb1 corn had been previously reported. “Problem counties” were chosen from the pool of counties where injury had been reported, primarily located in northeastern Iowa. “Non-problem counties” were chosen from counties in the central and southern areas of Iowa, where resistance issues had not been documented in previously published studies (Gassmann et al. 2011, 2012, 2014, 2016; Dunbar et al. 2016; Jakka et al. 2016). Resistance to Cry3Bb1 had also been confirmed in problem counties (see Dunbar et al. 2016, Gassmann et al. 2011, 2014). One problem county and one non-problem county were sampled in 2015, 2016, and 2017 (Fig. 1). The number of fields sampled in problem counties were N=6 (2015; Delaware County), N=8 (2016; Winneshiek County), and N=6 (2017; Fayette County). The number of fields sampled in non-problem counties were N=4 (2015; Taylor County), N=6 (2016; Jasper County), and N=7 (2017; Greene County). Fields studied in both problem and non-problem counties were identified by regional agronomists and local cooperators. None of the fields of this study were sampled in Gassmann et al. (2011, 2012, 2014, 2016) Dunbar et al. (2016), or Jakka et al. (2016). Data collection: Abundance of western corn rootworm in all fields was measured by placing 10 unbaited yellow sticky traps (Pherocon, Trécé Inc., Adair, OK) in each field. Traps were placed on plants at ear-level, in two parallel transects of five traps each (Dunbar et al. 2016). Individual traps were 30 m apart, with approximately 15 m between transects. All traps were placed a minimum of 30 m from any field border. Three sampling periods were conducted for each field, which ranged from 6 to 14 d (mean ± sd = 7.84 ± 1.85). Sampling periods began 21 July in 2015, 25 July in 2016, and 24 July in 2017. Root injury for all fields was assessed by digging roots of corn plants and rating them on a 0-3 node injury scale (Oleson et al. 2005). Ten roots were taken from rows adjacent to where the sticky traps were placed, and were washed and rated within 48 h. A sample of leaf tissue was taken from each plant and used to determine the presence and identity of Bt proteins targeting western corn rootworm using an ELISA-based test kit (Envirologix Inc., Portland, ME). Single-plant bioassays: Bioassays consisted of two corn types: Cry3Bb1 corn (DKC 43-48) and its non-Bt genetic isoline (DKC 43-46). Plants were grown to the V4-V5 growth stage (Abendroth et al. 2011) in 0.95 L plastic cups (Placon, Madison, WI); there were 12 replicates of each corn type for each population tested (N = 7 controls, N = 5 non-problem, N = 5 problem). Twelve neonate larvae (>24 h old) were added to each plant by placing them gently with a paintbrush on exposed root tissue. The plants were placed in environmental chambers (24°C, 16:8h L:D, 65% R.H). Larvae were allowed to feed undisturbed for 17 d, after which time the aboveground plant tissue was removed and the contents of the cups were placed on Berlese funnels for four days. Larvae were collected in vials filled with 85% ethanol and counted using a dissecting microscope (MZ6, Leica, Microsystems, Wetzlar, Germany). Head capsule width was measured using a microscope camera and imaging software (Moticam 2500, Motic Images Plus 3.0; Motic North America, Richmond, British Columbia, Canada), and this measurement was used to determine larval instar following Hammack et al. (2003). Bioassays began (i.e., larvae were added to plants) 19 April, 2016 for field populations collected in 2015, and the last vials were collected from Berlese funnels on 9 May, 2016. For 2016 populations, bioassays began 8 May, 2017, and ended 11 June. For 2017 populations, bioassays began 23 April, 2018, and ended 10 June. Field management history: Field management history was obtained from farmers and cooperators, which included Bt traits used (Cry3Bb1, mCry3A, eCry3.1Ab, Cry34/35Ab1, or a pyramid of multiple toxins) and soil insecticide use for a total of six years (the year of sampling and five previous years). In some cases, data on field history could not be obtained. The mean number of years for which Bt use could be obtained was 4.6 ± 1.7 (mean ± sd), and for soil insecticide use, the mean number of years was 5.1 ± 2.1 (mean ± sd) (Supp. Tables S1-S3). The total number of years that corn had been grown consecutively in each field was calculated using CropScape Data Layer (National Agricultural Statistics Service, United States Department of Agriculture; available at: https://nassgeodata.gmu.edu/CropScape) from 2003, the year Bt corn was first introduced for rootworm management. Using the Cropscape Data Layer information and the field management histories, the following metrics were calculated for each field: 1) the total number of years of continuous corn, 2) the proportion of years the field was planted to corn in the last six years, 3) the proportion of years soil insecticide had been used on corn out of the previous six years, 4) the proportion of years non-Bt corn was planted out of the previous six years 5) the proportion of years single-trait Cry3 corn was planted (Cry3Bb1 or mCry3A) out of the previous six years, 6) the proportion of years single-trait Cry34/35Ab1 corn was planted out of the previous six years, 7) the proportion of years pyramided Bt corn was planted in the last six years (either Cry34/35Ab1 + Cry3Bb1 or Cry34/35Ab1 + mCry3A), 8) the proportion of years soil insecticide was used on Bt corn in the last six years, and 9) proportion of years soil insecticide was used on non-Bt corn in the last six years. No fields in the experiment had planted any corn hybrids expressing eCry3.1Ab as part of the field management history. Data analysis: Data provided in this archive are raw (untransformed data). All transformations occurred in SAS, provided in the accompanying code. General methods for analytical approaches found below: All statistical analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC, USA). For rootworm abundance, values were calculated as the total number of adults caught/trap/sampling day. The values used in analyses are those of the single sampling period for a field when abundance was highest, and thus reflect peak measured abundance of western corn rootworm for the year within each field. A two-way analysis of variance was used (PROC GLM), with either root injury or abundance as the response, and field type (problem county or non-problem county), year of sampling, and field type × year as explanatory variables. Root injury and abundance were transformed using a square root transformation to improve normality of the residuals. A difference in mean injury to Bt versus non-Bt (refuge) plants within each field was examined using a paired t-test (PROC TTEST). Rootworm survival in plant-based bioassays was analyzed separately by year using analysis of variance (PROC GLM) with proportion survival as the response, and population, corn type (Cry3Bb1 corn vs non-Bt corn), and their interaction as explanatory variables. Proportion survival was transformed by the arcsine of the square root to improve normality of the residuals. Linear contrasts (CONTRAST statement) were used to compare survival in each field population to all susceptible control replicates from the same year, and this comparison was made for Cry3Bb1 corn and non-Bt corn. A one-tailed t-test (PROC TTEST) was used to test for a lower proportion of survival on Cry3Bb1 corn compared to non-Bt corn. The proportion of larvae surviving to the third, and final, instar on each plant (i.e. the number of larvae collected at the end of the assay that reached the third instar stage divided by the total number of larvae collected from the plant) was also examined. This metric was analyzed using a one-tailed t-test (PROC TTEST). Proportion survival and proportion of third instars were transformed by the arcsine of the square root to improve normality of the residuals. For all t-tests, pooled variances were used when variances were equal in the two groups, and the Satterthwaite method was used when variances were unequal. Equality of variances was assessed using the folded F-statistic and associated P-value (i.e., variances are unequal when P ≤ 0.05) as reported in the output of PROC TTEST. To account for variation in survival on non-Bt corn among the populations, a complementary analysis was performed using corrected survival. Corrected survival was based on Abbott (1925) and was calculated for each population as: proportion of larvae surviving on Cry3Bb1 corn in a replicate ÷ mean proportion surviving larvae on non-Bt corn in all replicates of a population. One-way analysis of variance (PROC GLM) was performed for each year, with corrected survival as the response variable and population as the explanatory variable. Linear contrasts (CONTRAST statement) were used to compare corrected survival in each field to the control replicates of the same year. An additional test was conducted using analysis of variance (PROC GLM) to test for a difference in corrected survival between problem county fields and non-problem county fields, which used corrected survival as the response and field type as a factor in the model. Corrected survival was transformed using an arcsine of the square root transformation to improve normality of the residuals. Field management in the year of sampling was examined using chi-square analysis (PROC FREQ) to compare fields in non-problem and problem counties. To examine differences in field management strategies between problem and non-problem counties, t-tests were conducted for past field management metrics (PROC TTEST). Proportion of soil insecticide used on non-Bt corn was not examined with a t-test, as the mean of this metric in non-problem counties was zero. Years of continuous corn was transformed by ln(y) to improve normality of the residuals, and all other metrics were transformed by the arcsine of the square root to improve homogeneity of variances. In cases where homogeneity of variances could not be achieved, the Satterthwaite method was used to calculate degrees of freedom and test statistics. Multiple regression was used to examine the effects of field management on rootworm abundance and root injury (PROC REG) (Sokal and Rohlf 1995). All current and historical field management metrics were analyzed using correlation analysis (PROC CORR) to determine the presence of collinearity among the variables (Pearson’s correlation coefficient >0.8). The field management metrics used as independent variables in multiple regression were: non-Bt corn planted in the year of sampling (0=no, 1=yes), Cry34/35Ab1 corn planted in the year of sampling (0=no, 1=yes), pyramided corn grown in the year of sampling (either Cry34/35Ab1 + Cry3Bb1 or Cry34/35Ab1 + mCry3A, 0=no, 1=yes), the total number of years the field was planted to corn continuously since 2003, the proportion of years the field was planted to corn in the previous six years, soil insecticide use, the proportion of years non-Bt corn was planted in the past six years, Cry3 corn grown, the proportion of years Cry34/35Ab1 corn was planted in the past six years, the proportion of years a pyramid hybrid of corn had been grown in the past six years (either Cry34/35Ab1 + Cry3Bb1 or Cry34/35Ab1 + mCry3A), and proportion of years soil insecticide was used on non-Bt corn in the past six years. Stepwise selection was used to select the best model, with the thresholds of P < 0.25 for entry into the model and P < 0.15 to be retained in the model, following Dunbar et al. (2016). Rootworm abundance and root injury were transformed using a square root transformation and years of continuous corn was transformed using ln(y) to improve normality of the residuals. Software: Name: SAS v9.4 Developer: SAS Institute Inc., Cary, NC, USA Licensing: This work is licensed under the Creative Commons Attribution (CC-BY) 4.0 International License. For more information visit: https://creativecommons.org/licenses/by/4.0