GENERAL INFORMATION Project Title: Linking land use patterns and pest outbreaks in Bt maize Data Title: Land use patterns data Abtract: Data was collected between 2015 and 2018. In this study, we used geospatial tools to examine land use patterns surrounding fields that had previously experienced greater-than-expected injury to Cry3 (Cry3Bb1 or mCry3A) corn roots (>1 node) between the years of 2009 and 2013 in Iowa, USA. Using information obtained from the National Agriculture Statistics Service (NASS) CropScape DataLayer and ArcGIS, we examined patterns of continuous maize cultivation at a distance of 1.6, 3.2, and 16.1 km from problem field centroids. We compared this information to points that were randomly placed on agricultural land in the state of Iowa. In addition, we calculated the proportion of the state that had been planted to third-year continuous maize for the years 2003 through 2018, and performed a correlation with the price of maize in the previous year to determine if maize price influenced continuous maize cultivation. 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: comparison of problem field history of continuous maize cultivation to controls, analysis of variance examining continuous maize cultivation at 1.6, 3.2, and 16.1 km in problem fields and controls, and the correlation of statewide third-year maize with maize price. Authors: Coy R. St. Clair (Corresponding author) Iowa State University cstclair@iastate.edu Aaron J. Gassmann Iowa State University aaronjg@iastate.edu Associated Publication: "Linking Land Use Patterns and Pest Outbreaks in Bt Maize," Ecological Applications (in review) Collection Information: Fields were originally sampled between 2009 and 2013 by other researchers in the Gassmann Lab at Iowa State University. All fields had a root injury rating of > 1 node on the node injury scale (Oleson et al. 2005). Publication history of the fields and field locations can be found below: Field Year County Publication Field name in publication 1 2009 Delaware Gassmann et al. 2011 P1 2 2009 Delaware Gassmann et al. 2011 P2 3 2009 Floyd Gassmann et al. 2011 P4 4 2009 Jackson Gassmann et al. 2011 P3 5 2010 Clayton Gassmann et al. 2012 S1 6 2010 Clayton Gassmann et al. 2012 S5 7 2010 Sioux Gassmann et al. 2012 S6 8 2010 Clayton Gassmann et al. 2012 S7 9 2010 Clayton N/A Unpub 10 2010 Fayette N/A Unpub 11 2010 Floyd N/A Unpub 12 2011 Hancock Gassmann et al. 2014 P1 13 2011 Howard Gassmann et al. 2014 P2 14 2011 Johnson Gassmann et al. 2014 P3 15 2011 Fayette Gassmann et al. 2014 P4 16 2011 Winneshiek Gassmann et al. 2014 P5 17 2011 Bremer Gassmann et al. 2014 P6 18 2011 Clayton Gassmann et al. 2014 P7 19 2011 Fayette Gassmann et al. 2014 P8 20 2011 Sac Gassmann et al. 2014 P9 21 2011 Clayton Gassmann et al. 2014 Supp 22 2011 Fayette Gassmann et al. 2014 Supp 23 2011 Fayette Gassmann et al. 2014 Supp 24 2011 Jones Gassmann et al. 2014 Supp 25 2011 Fayette Gassmann et al. 2014 Supp 26 2012 Sioux Jakka et al. 2016 P1 27 2012 Carroll Jakka et al. 2016 P2 28 2012 Franklin Jakka et al. 2016 P3 29 2012 Butler Jakka et al. 2016 P4 30 2012 Fayette Jakka et al. 2016 P5 31 2012 Jackson Jakka et al. 2016 P6 32 2012 Allamakee N/A Unpub 33 2012 Cass N/A Unpub 34 2012 Cherokee N/A Unpub 35 2012 Chickasaw N/A Unpub 36 2012 Clayton N/A Unpub 37 2012 Clayton N/A Unpub 38 2012 Clayton N/A Unpub 39 2012 Dubuque N/A Unpub 40 2012 Dubuque N/A Unpub 41 2012 Grundy N/A Unpub 42 2012 Hardin N/A Unpub 43 2012 Jones N/A Unpub 44 2012 Plymouth N/A Unpub 45 2012 Winneshiek N/A Unpub 46 2012 Winneshiek N/A Unpub 47 2012 Winneshiek N/A Unpub 48 2012 Winneshiek N/A Unpub 49 2012 Woodbury N/A Unpub 50 2012 Wright N/A Unpub 51 2013 Boone N/A Unpub 52 2013 Boone N/A Unpub 53 2013 Boone Gassmann et al. 2016 P3 54 2013 Franklin Gassmann et al. 2016 P4 55 2013 Wright Gassmann et al. 2016 P5 56 2013 Wright Gassmann et al. 2016 P6 57 2013 Pottawattamie Gassmann et al. 2016 P7 58 2013 Pottawattamie Gassmann et al. 2016 P8 59 2013 Boone Gassmann et al. 2016 P9 60 2013 Crawford N/A Unpub 61 2013 Bremer N/A Unpub 62 2013 Wright Dunbar et al. 2016 / Shrestha et al. 2018 Supp 63 2013 Wright Dunbar et al. 2016 / Shrestha et al. 2018 Supp 64 2013 Wright Dunbar et al. 2016 Supp Geospatial data on the fields and surrounding landscape was collected between 2015 and 2018. FILES: NOTE: for all fields, 0=true zero (measurement). A period "." is null or missing data. File List: 1. ANOVA DATA.csv - raw data for analysis of variance of maize Number of variables: 6 Variable name: Field (field identifier) Fieldtype (Control or past problem field) Year (Year corresponding to the year in which the relevant past problem fields experienced greater-than-expected injury to Cry3 corn; 2009, 2010, 2011, 2012, 2013) Buffer (A=1.6km, B=3.2km, C=16.1km) CC_Year (number of consecutive years that maize was planted; 1 through 9) Prop independent (the measured proportion of the area planted to maize, with area of the past problem fields removed) 2. ANOVA SAS CODE.txt - SAS code that uses File #1 above. 3. Problem fields vs controls DATA.csv - raw data for fields in which the randomly placed control points landed in a corn field and past problem fields, for comparing the history of corn cultivation between controls and problem fields Number of variables: 4 Variable name: Fieldtype (Control or past problem field) Year (Year corresponding to the year in which the relevant past problem fields experienced greater-than-expected injury to Cry3 corn; 2009, 2010, 2011, 2012, 2013) Field ID (field identifier) Years corn (number of years the field had been planted to corn consecutively). 4. Problem fields vs controls SAS CODE.txt - SAS code that uses File #3 above. 5. Statewide 3-yr maize DATA.csv - raw data for the proportion of Iowa that was planted to 3rd-year continuous corn and the price of corn in the previous year between the years of 2004 and 2014. Number of variables: 3 Variable name: Year (Year) 3yr (proportion of the state of Iowa that was planted to third-year continuous corn, corresponding to the Year variable) Prev_Calendar (the average price of corn [$/bushel] in the year before the Year variable) 6. Statewide 3-yr maize SAS CODE.txt - SAS code used for analysis of File #5 above. METHODS AND MATERIALS Using geospatial tools (ArcGIS) and publicly available land-use data, we examined circular areas (buffers) surrounding fields that had previously experienced high levels of rootworm injury to Cry3Bb1 maize and rootworm resistance to Cry3Bb1 maize (>1 node; "problem fields"). We calculated the proportion of area inside each buffer planted to maize continuously for one to nine years, and compared these values to those for randomly selected control points throughout the state. We also calculated the proportion of the state planted to maize for at least three consecutive years for 2003 through 2018, and its relationship with the annual value of maize. Data Collection: Problem fields had previously been sampled (see above for relevant publications). Data collection for this project occured between 2015 and 2018. ArcGIS was used to create buffers around each problem field, calulated from the field centroid. Data from the National Agriculture Statistics Service (NASS) CropScape DataLayer were overlaid on the map in ArcGIS, and these date were used to calculate the area of each buffer that had been planted to corn for 1 to nine consecutive years. To make comparisons, control points were placed randomly on the map, and the same method was applied. This method was also applied to the entire state of Iowa to examine statewide trends in corn production. 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: Proportion of the landscape planted to maize was used as the measurement for statistical analysis because taking the proportion of area standardized the data such that comparisons could be made among buffer sizes. To test the hypothesis that problem fields were not randomly distributed in the state of Iowa, average nearest neighbor analysis was conducted in ArcGIS on a layer containing all 64 problem-field centroids. The null hypothesis was that problem fields were randomly distributed within the state, and the alternative hypothesis was that problem fields were spatially clustered in a non-random orientation. A t-test was performed (PROC TTEST) to determine if the number of consecutive years of maize planted differed between problem fields and the randomly-placed control points that landed in maize fields. The Satterthwaite method was used to calculate degrees of freedom due to unequal variances between the groups. A significantly higher number of years of consecutive maize cultivation in problem fields compared to controls would indicate that continuous maize cultivation within a field was associated with injury to maize by western corn rootworm. To analyze the proportion of the landscape planted to maize in control locations and problem-field locations, a mixed-model analysis of variance (ANOVA) was constructed using SAS 9.4 (SAS Institute Inc., Cary, NC, USA; PROC GLIMMIX). The response variable in the model was the proportion of buffer that was planted to maize. Fixed effects were location type (controls vs problem fields), buffer size (1.6 km, 3.2 km, or 16.1 km radius), number of consecutive years a field had been planted to maize (first-year maize through ninth-year continuous maize), and all possible interactions of these variables. Year of injury (2009 through 2013) and its interactions with the fixed-effect variables were included as random effects to account for year-to-year variation in the amount of maize that was planted in Iowa. To determine if the proportion of first-year maize differed between controls and problem fields at each buffer size, contrasts were conducted using the LSMESTIMATE statement for the interaction of location type × buffer size × consecutive years of maize, with a Bonferroni adjustment based on three comparisons (P ≤ 0.017). To examine whether differences were present between continuous maize in areas surrounding problem fields and control fields, within a buffer type (e.g. 1.6 km buffer) for each type of continuous maize (e.g., two or more consecutive years of maize cultivation, three or more consecutive years of maize, etc.) contrasts were conducted, with a Bonferroni correction based on 24 comparisons (P ≤ 0.002). To test the hypothesis that the area of Iowa that had been planted to three years of continuous maize would be influenced by the price of maize, a linear regression analysis was conducted using the proportion of third-year maize in Iowa as the dependent variable and the calendar year market average price of maize in the previous year as the independent variable (PROC REG). 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