(Season) Crop Residue Data for Establishment of ... - ACS Publications

Subscriber access provided by Fudan University

Article

Comprehensive analysis of the value of single vs multiple year (season) crop residue data for establishment of Maximum Residue Levels (MRLs) Carrie R Fleming, Val Gartner, Pablo Valverde-Garcia, Pieter W Geurs, and Carmen Tiu J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.6b05106 • Publication Date (Web): 03 Jan 2017 Downloaded from http://pubs.acs.org on January 5, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Agricultural and Food Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30

Journal of Agricultural and Food Chemistry

TITLE: Comprehensive analysis of the value of single vs multiple year (season) crop residue data for establishment of Maximum Residue Levels (MRLs) AUTHORSHIP: Carrie R Fleming, Val Gartner, Pablo Valverde-Garcia, Pieter Geurs, Carmen Tiu* Dow AgroSciences, 9330 Zionsville Road, Indianapolis, IN 46268 *317-337-4041, [email protected] KEYWORDS: pesticides, residues, seasonality, homogeneity, variability, maximum residue levels, proportionality.

Page 1

ACS Paragon Plus Environment


Page 2 of 30

1

ABSTRACT:

2

Disharmony currently exists in regulatory requirements regarding whether multiple seasons of

3

field residue trials are necessary. This analysis used historical residue data to evaluate whether

4

the year in which trials are conducted is a significant contributor to the overall variability in field

5

residue data. It was concluded that residue behavior is highly variable in nature, regardless of

6

the season, that variation of residue data compiled from multiple years is not statistically greater

7

than data resulting from trials conducted within any one year, and that variation across years

8

does not result in large systematic differences in residue values or resulting Maximum Residue

9

Limits compared to trials conducted in any single year. Field trials conducted at a variety of

10

locations across geographical regions will capture variability due to different environmental

11

conditions and agricultural practices and provide a robust estimate of the spread of residues

12

expected due to labeled use of a pesticide.

Page 2


Page 3 of 30


13

INTRODUCTION:

14

The purpose of the research presented here is to determine if there is a need for pesticide residue

15

data to be collected over more than one growing season. OECD and several regulatory agencies

16

allow pesticide residue data to be collected in a single year1-6; however, some other regulatory

17

agencies require residue data to be collected over more than one year or growing season7-11.

18

This research assessed 97 sets of residue data that have been generated on 20 active ingredients,

19

across four global regions, with pesticide residue trials (within same region) conducted over two

20

or three years. Statistical analysis was conducted to assess whether pesticide residues differ

21

between years and to determine the contribution of the year in which trials were conducted to the

22

overall variability in field trial residue data.

23

Pesticide residue field trials are conducted to determine the magnitude of the pesticide residues

24

in or on raw agricultural commodities (RACs), for the purposes of registration of a pesticide,

25

setting tolerances, or maximum residue levels (MRL) for the pesticide on RACs, and evaluating

26

the consumer risk related to consumption of foods derived from treated crops. Depending on the

27

crop and country, regulatory guidelines require 4-20 field residue trials to be conducted per

28

representative crop under a diverse range of climates and growing practices. The trials are

29

conducted under the Good Agricultural Practices (GAP) according to pesticide label instructions

30

that are expected to produce the highest residues (i.e., maximum application rate, maximum

31

number of applications, minimum re-treatment interval, and minimum period between treatment

32

and harvest of samples or pre-harvest interval).

Page 3



33

A wide variety of factors can influence the magnitude of pesticide residue concentrations in

34

agricultural commodities and can contribute to the variability observed within and across residue

35

field trials12. A number of factors can affect the final amount of pesticide residue on the day of

36

application to crops, even when the same GAP is followed, resulting in large variability between

37

trials. These include differences in the spray equipment used, calibration of that equipment, tank

38

mixing partners used (e.g, adjuvants), the crop growth stage at the time of application, etc.

39

Differences in crop variety, horticultural practices (e.g., differences in planting density or

40

pruning practices), and weather conditions can also lead to variability between trials.

41

Furthermore, variability in pesticide residues within the field may occur due to differences in

42

spray deposition (e.g., resulting from overlapping spray swaths or gusty winds) or crop growth

43

and maturation (e.g, due to gradients in soil quality within a field). Together, these factors can

44

lead to differences in the magnitude of pesticide residues from different trial sites, or even in

45

multiple samples taken from the same trial site.

46

A couple of previous analyses have established precedents for statistical methods to determine

47

the contribution of the different field related parameters into the typically high variability of crop

48

residue data. Brief summaries are presented below, as important precedents to the analysis

49

provided in this paper.

50

Proportionality of Residue Data to Application Rate

51

This was the first analysis using statistical means to compare residue data from different

52

pesticides, crops and regions13 and provided proof of evidence for the hypothesis that there is a

53

direct relationship between application rates and magnitude of residues. Based on this statistical Page 4


Page 4 of 30

Page 5 of 30


54

procedure, more data was validated and published by the Food and Agriculture

55

Organization/World Health Organization (FAO/WHO) Joint Meeting on Pesticide Residues

56

(JMPR) to conclude that a proportional relationship between pesticide application rate and the

57

resulting residues in harvested commodities exists within the range of 0.3 – 4 times the

58

maximum labeled application rate14, 15.

59

Homogeneity of Residue Data Across Regions

60

Analysis of variability in residue data across climatic regions (zones) has shown that there are no

61

systematic differences in pesticide residues from trials conducted at similar GAPs in different

62

regions, and that zone is a minor contributor to the overall variability. A joint OECD/FAO

63

project developed in 1999 through 2003 concluded that highest contributors to residue data

64

variability seemed to be the agricultural practices, while climatic zones did not appear to be a

65

significant contributor, though further analysis was considered necessary to confirm these

66

results12. A subsequent analysis on a global residue program using data from four continents and

67

five crops (apple, cabbage, grape, tomato, and wheat (grain, forage, hay, and straw)) concluded

68

that data is homogeneous across regions, and variability between regions is smaller than the

69

variability between trials within individual regions. This work was presented at the Global

70

Minor Use Summit meeting in Rome 2012 and published in the report from this meeting 16. To

71

further expand on and validate these conclusions, a joint effort was coordinated between the

72

United States Environmental Protection Agency (US EPA), the Canadian Pest Management

73

Regulatory Agency (PMRA) and Crop Life America (CLA) to evaluate more than 4,000 trials

74

generated on more than 70 crops in at least two climatic zones. A draft technical document Page 5



Page 6 of 30

75

summarizing this work and supporting the global exchangeability of field trial residue data was

76

appended to the OECD Crop Field Trials Guidance document17 and proceedings from the Codex

77

Committee on Pesticide Residues (CCPR)-4818.

78

Hypothesis for Homogeneity of Residue Data Across Seasons/Years

79

There is currently a lack of harmonization in global regulations concerning the need to conduct

80

residue trials in more than one year. OECD and several regulatory bodies consider residue data

81

generated all in a single growing season to be acceptable, as long as there are a sufficient number

82

of trial sites distributed across a wide enough geographical area to capture different climatic

83

conditions (e.g.,1-6). However, for other regulatory bodies, residue trials are required to be

84

distributed across more than one growing season7-11, leading to disharmonies in global data

85

requirements for establishment of pesticide MRLs. As part of a sequence of initiatives aimed at

86

harmonization of MRLs (e.g. comprehensive global programs – OECD 5093, exchangeability of

87

data across regions – CCPR 48 item 0.81 19, several documents of crop groupings – CCPR 4820),

88

harmonization of the requirement for trials from multiple growing seasons is the next step to

89

further align regulatory requirements for residue trials to facilitate the establishment of globally

90

harmonized MRLs, which are necessary for smooth global trade of food and feed commodities.

91

The purpose of this research was to address the disparity in regulatory requirements for field trial

92

data from either a single or multiple growing seasons by determining whether the growing

93

season contributes significantly to the variability captured in the overall package of field trial

94

data.

95

MATERIALS AND METHODS: Page 6


Page 7 of 30


96

Selection of Data

97

Pesticide residue data from existing regulatory studies were used to evaluate the homogeneity of

98

the data over two or three years. Residue data from trials conducted according to a similar Good

99

Agricultural Practices (GAP), and containing field trial data across more than one year, with at

100

least four trial sites per year, and with quantified residue values (i.e. all residues cannot be non-

101

detectable) were selected for inclusion in this analysis. The determination that trials were

102

conducted at a similar GAP was made according to the following criteria:

103

1.

104

All trials conducted in the same geographical region (e.g., EU southern zone, or North America)

105

2.

All trials measured the same analyte(s)

106

3.

All trials sampled from the same crop and crop matrices (e.g., tomato fruit, or

107

wheat straw)

108

4.

All trials had the same number of applications

109

5.

Application rate and pre-harvest interval (PHI) did not differ between trials

110

by more than a combined difference of 25%.

111

According to OECD Test Guideline 509, “In the case of up to 25% increases or decreases of the

112

active ingredient application rate, the number of applications, or the PHI, under otherwise

113

identical conditions, the residue results can be assumed to be comparable” 3. Therefore, the

114

selection criteria used in this analysis were conservative and would be considered to effectively

115

identify trials that should be considered comparable. For each of the 97 data sets included in the

116

analysis, the application rates at each trial site were within 17% of the mean application rate for Page 7



Page 8 of 30

117

their data set, with an average difference of 5%. The PHIs at each trial site were within 13% of

118

the mean PHI for their data set, with an average difference of 1%. Furthermore, when the

119

differences between the application rate and PHI for each trial site were compared to their data

120

set’s averages, the total cumulative difference was within 17% of the mean, with an average

121

difference of 3%.

122

There were 91 sets of residue data over two years and 6 sets of data over three years giving a

123

total of 97 sets of residue data included in the analysis. These data sets included residue results

124

from 20 active ingredients (eight herbicides, six insecticides, and six fungicides), 31 crops, and

125

four regions (Canada, European Northern Zone, European Southern Zone, and United States).

126

Table 1 presents the pesticide mode of action/crop combinations included in the analysis.

127

Of the 97 data sets available for analysis, 46 had replicate samples taken from each trial, i.e. two

128

independently collected samples from one field trial location analyzed for residues. For these 46

129

datasets, further evaluations were performed regarding the contribution of the trial site to the

130

overall variability vs. within trial residual variability (as described further below).

131

Assumptions for Data Analysis

132

Residue results less than or equal to the limit of quantitation (LOQ) were assigned the LOQ

133

value for analysis. For example, if the residue was reported as ND (non-detectable), and the

134

LOQ was 0.01 ppm, the residue value was adjusted to 0.01 ppm for the analysis. This is

135

consistent with the OECD guidance for MRL calculation, which requires the substitution of non-

136

detected residues by the LOQ values. 21 Page 8


Page 9 of 30


137

Based on common global regulatory pesticide residue definitions, where appropriate, metabolite

138

residue values were converted to stoichiometric active ingredient parent equivalents and added to

139

parent residue values; this total residue value was used in subsequent analyses.

140

Procedures and Methods for Data Analysis

141

From the procedural perspective, four different steps have been conducted:

142

1. Residue ratio comparison analysis,

143

2. Non-parametric and parametric statistical analysis to identify any systematic differences,

144

3. Variance components analysis,

145

4. Regulatory impact of exceptions on MRL calculations.

146 147

A brief description of the methods used for each of the 4 steps is described below. 1. Residue ratio comparison analysis

148

Residue values were log10 transformed to satisfy linear model assumptions (normality and

149

homogeneity of variance) for the parametric statistical analysis at step 2. The additional

150

property of the log10 transformation is that the difference in residue levels between two

151

years in the log10 scale is equivalent to the ratio of those residue levels in the physical

152

scale. For each dataset, differences between any pair of years were calculated as absolute

153

values, which provide ratios ≥ 1, and the maximum ratio between any pair of years was

154

reported.

155

2. Non-parametric and parametric statistical analysis to identify any systematic differences

Page 9



Page 10 of 30

156

Yearly subsets of each data package for the same combination of active and crops were

157

evaluated by non-parametric statistical tests. Non-parametric analyses used Wilcoxon

158

test for datasets with two years and Kruskal-Wallis test for datasets with more than 2

159

years (χ2 approximation, P

(Season) Crop Residue Data for Establishment of ... - ACS Publications

Recommend Documents