Sampling and Variance in Measurements of Trifluralin Disappearance

3 Sampling and Variance in Measurements ofTrifluralinDisappearance from a Field Soil A. W. TAYLOR, J. H. CARO, H. P. FREEMAN, and B. C. TURNER

Downloaded by RUTGERS UNIV on December 30, 2017 | http://pubs.acs.org Publication Date: July 15, 1985 | doi: 10.1021/bk-1985-0284.ch003

Beltsville Agricultural ResearchCenter,Agricultural Research Service, U.S. Department of Agriculture, Beltsville,MD20705

Trifluralin was applied to separate field plots either as a spray of emulsifiable concentrate (EC) or by spreading a granular formulation (GF). Both formulations were incorporated and mixed to the 7.5 cm depth. Soil concentrations were measured by eightfold analysis of five field samples obtained from each plot (40 analyses per plot): elaborate precautions were taken to ensure complete mixing and random subsampling in the laboratory. Coefficients of variation were 10% in the EC data and 38% in the GF. Analysis of variance showed that the difference reflected irregular distribution of the herbicide in the laboratory subsamples due to localized high concentrations at granule sites. In later samplings at about 20 day intervals, this irregularity was reduced as the granules became mixed with the soil. Regression analysis indicated a 50% disappearance time of 81 days on the EC plot and 106 days on the GF, but this difference was not significant at the 90% probability level. Regression of a l l data from both plots gave a 50% disappearance time of 91 days with a 90% probability that the true value lay between 79 and 108 days. The implications of these observations in experimental design and sampling strategies in field studies and environmental work are discussed. The work reported i n t h i s paper was part o f a l a r g e r p r o j e c t whose p r i n c i p a l o b j e c t i v e was to measure the rate o f v o l a t i l i z a t i o n o f the h e r b i c i d e t r i f l u r a l i n to the atmosphere from a f i e l d s o i l a f t e r surface a p p l i c a t i o n and i n c o r p o r a t i o n to a depth o f 7.5cm by c u l t i v a t i o n with a d i s c harrow. In t h i s experiment i t was necessary to measure as a c c u r a t e l y as p o s s i b l e the amount o f the t r i f l u r a l i n residue i n the s o i l on s e v e r a l days during the growing season and to c a l c u l a t e the r a t e o f disappearance o f the This chapter not subject to U.S. copyright. Published 1985, American Chemical Society

Kurtz; Trace Residue Analysis ACS Symposium Series; American Chemical Society: Washington, DC, 1985.


26

TRACE RESIDUE ANALYSIS

residues. Two treatments were compared, one i n which the h e r b i c i d e was a p p l i e d as a spray o f water-based e m u l s i f i a b l e concentrate (EC) and a second i n which i t was spread as a dry granular formulation ( G F ) . Both p l o t s were otherwise treated and managed i d e n t i c a l l y . E a r l i e r i n v e s t i g a t i o n s o f the problems o f measurement and sampling o f p e s t i c i d e residues i n f i e l d s o i l s (1,2) have shown that a d e t a i l e d understanding o f the s i g n i f i c a n c e o f such data r e q u i r e s a sampling and a n a l y t i c a l scheme designed to measure the v a r i a n c e o f the o r i g i n a l data and permit estimates o f the degree o f confidence that can be placed on the f i n a l r e s u l t s . T h i s paper d i s c u s s e s the s t a t i s t i c a l a n a l y s i s and experimental s i g n i f i c a n c e o f the r e s u l t s obtained with the sampling scheme described below, which was i n i t i a l l y designed to o b t a i n the maximum information a v a i l a b l e that could be obtained with the resources a v a i l a b l e for t h i s part o f the work. D e t a i l s o f the work on t r i f l u r a l i n v o l a t i l i z a t i o n w i l l be published elsewhere. Experimental Design and Procedures F i e l d Layout and Management. The experiment was done on a l e v e l f i e l d o f about 5 hectares on the Harford Teaching and Research Center o f C o r n e l l U n i v e r s i t y at Harford, N . Y . The s o i l was a Howard g r a v e l l y loam with e s s e n t i a l l y uniform c h a r a c t e r i s t i c s over the experimental a r e a . The two formulations were a p p l i e d to separate r e c t a n g u l a r experimental p l o t s o f 2.97 hectares each, (244 x 122m) located s i d e by s i d e with an untreated 3m s t r i p between them on the long axes. The granular treatment was a p p l i e d on May 2nd, 1974 by spreading T r e f l a n 5G commercial formulation from a conventional g r a v i t y fed spreader at a nominal rate o f 0.7 kg/h o f a c t i v e ingredient. On May 8 t h , the emulsion-plot r e c e i v e d 0.66 kg/h t r i f l u r a l i n as a spray o f T r e f l a n 4EC e m u l s i f i a b l e concentrate d i l u t e d with water according to commercial p r a c t i c e . (Note: Mention o f a p r o p r i e t a r y product i s f o r purposes o f i d e n t i f i c a t i o n only and does not c o n s t i t u t e endorsement by the U . S . Department o f A g r i c u l t u r e over products o f a s i m i l a r nature not mentioned). This a p p l i c a t i o n was made between 0700 and 0830 a.m. i n c o o l ( a i r temp., 1 0 ° C ) c l e a r and windless weather. Both granular and emulsion formulations were incorporated to the 7.5cm depth with a d i s c c u l t i v a t o r w i t h i n 15 minutes o f a p p l i c a t i o n . A second c u l t i v a t i o n to the same depth, but at r i g h t angles to the f i r s t , was completed w i t h i n 2-3 hours. On May 20th both p l o t s were s p r i n g - t o o t h harrowed f o r seed-bed p r e p a r a t i o n f o r soybean p l a n t i n g on the same day. The only subsequent s o i l disturbance was a row c u l t i v a t i o n i n l a t e June to c o n t r o l a severe i n f e s t a t i o n o f quack-grass: c o n t r o l o f b r o a d l e a f weeds was e s s e n t i a l l y complete with the t r i f l u r a l i n . The f i r s t samples were taken from the GF p l o t on May 2nd, 4 hours a f t e r a p p l i c a t i o n and from the EC p l o t on May 8th, 7 hours


3.

TAYLOR ET AL.

27

Measurements of Trifluralin Disappearance


a f t e r a p p l i c a t i o n . Subsequent samples were taken from both p l o t s on May 21st, June 25th, J u l y 3 0 t h and September 24th. On the GF p l o t these correspond to sampling times o f 0.16, 18.8, 53, 88 and 144 days, and 0.29, 1 3 . 3 , 48, 83 and 139 days on the EC p l o t . Sampling Procedure. A set o f twenty-five sampling p o i n t s was marked out on each p l o t on a rectangular 5 x 5 g r i d . The d i s t a n c e between p o i n t s was thus 40.5m on the long a x i s o f the g r i d and 20.2m between rows o f p o i n t s . Each sampling point was permanently i d e n t i f i e d by number. On each sampling, f i v e f i e l d samples were taken from each p l o t by combining s o i l samples from each o f f i v e sampling p o i n t s . The f i v e p o i n t s combined i n each sample were s e l e c t e d each day from a table o f random numbers. Since a new s e l e c t i o n o f p o i n t s was made on each day the i n d i v i d u a l f i e l d samples numbered I through V i n the Tables do not represent s o i l taken from the same points on the field. Each f i e l d sample can be regarded as composed c f s o i l from a randomly chosen set o f sampling p o i n t s that e q u a l l y represent the surface o f the e n t i r e f i e l d p l o t . The sample taken from each point was obtained by combining 16 separate c o r e s , a l l w i t h i n an area o f 50 x 50 cm and a l l taken with the same s p e c i a l l y designed c o r e r with an i n t e r n a l diameter of

4.4cm: a l l c o r e s were t a k e n

t o 15.0cm d e p t h .

Each f i e l d

sample

thus included s o i l taken from 80 cores r e p r e s e n t i n g a composite block o f s o i l o f 0.122m i n f i e l d area and 15.0 cm deep. On removal from the f i e l d t h i s sample was weighed and mixed by tumbling for 5 minutes i n a power d r i v e n p o r t a b l e concrete mixer. The whole sample, u s u a l l y weighing between 20 and 25kg, was then d i v i d e d and s u b - d i v i d e d by four passages through a r i f f l e r to o b t a i n a r e p r e s e n t a t i v e f i e l d subsample o f about 2.5kg weight f o r laboratory a n a l y s i s . These subsamples were returned to the l a b o r a t o r y w i t h i n 12 hours and stored at a temperature below 1 0 ° c u n t i l analyzed. On each sampling day the p r e c i s e p o s i t i o n s from which the cores were taken was marked with a stake. On the next day the cores were taken from a s i m i l a r area adjacent to the f i r s t so that a l l the f i v e sample s i t e s at each point were i n a l i n e along the long a x i s o f the f i e l d . This was a l s o the d i r e c t i o n i n which the sprayer and spreader moved during a p p l i c a t i o n . 2

A n a l y t i c a l Procedure. A f t e r thawing, each f i e l d sample was immediately d i v i d e d i n t o 4 quarters and subsamples o f approximately 100g o f moist s o i l taken from each for a n a l y s i s . s i n g l e subsample was o v e n d r i e d the water c o n t e n t .

a t 105 °C o v e r n i g h t

A

t o measure

Each a n a l y t i c a l sample was shaken f o r 30 minutes i n 250ml o f a n a l y t i c a l grade benezene c o n t a i n i n g 80ml o f i s o p r o p a n o l . After s e t t l i n g , an a l i q u o t o f the e x t r a c t a n t was washed with d i s t i l l e d water to remove the i s o p r o p a n o l and the t r i f l u r a l i n c o n c e n t r a t i o n measured by d i r e c t i n j e c t i o n i n t o a gas chromatograph equipped



28

T R A C E RESIDUE ANALYSIS

with an e l e c t r o n capture d e t e c t o r . T e s t i n g with samples spiked at the 0.1ppm l e v e l i n d i c a t e d that t r i f l u r a l i n r e c o v e r i e s with t h i s procedure were c o n s i s t e n t l y 95 to 100$. The l i m i t o f d e t e c t i o n o f t r i f l u r a l i n i n s o i l was l e s s than 10 ppb. S o i l concentrations were never l e s s than 50 times greater than t h i s . Four r e p l i c a t e s from each f i e l d sample were analyzed f o r t r i f l u r a l i n content, except f o r those from the f i r s t day when e i g h t r e p l i c a t e s were analyzed i n order to obtain more extensive data for the i n i t i a l s t a t i s t i c a l a n a l y s i s . The r e s u l t s obtained may therefore be c l a s s e d as a two-stage h i e r a r c h i c a l sampling s t r u c t u r e , with four (or e i g h t ) a n a l y t i c a l determinations on each of f i v e f i e l d samples each o f which can be regarded as p r o v i d i n g independent estimate o f the t r i f l u r a l i n content i n the e n t i r e field soil. Since the amount o f t r i f l u r a l i n i n each o f these f i e l d samples can be regarded as that i n a r e p r e s e n t a t i v e composite block o f s o i l o f 0.122m a r e a , a l l the data can be expressed as m i l l i g r a m s o f h e r b i c i d e per square meter o f s o i l area in the f i e l d . 2

Results The r e s u l t s o f the eight analyses o f the f i v e f i e l d samples taken from both f i e l d p l o t s on the day o f a p p l i c a t i o n are presented i n Table I together with means and standard d e v i a t i o n s . The values of the Grand Mean and i t s standard d e v i a t i o n were c a l c u l a t e d from the sum and v a r i a n c e o f the 40 i n d i v i d u a l samples. The grand mean of the EC data (Table I) has a c o e f f i c i e n t o f v a r i a t i o n o f 10%. Comparison with e a r l i e r experience (1_) suggests that t h i s approaches the best p r e c i s i o n that can be expected i n f i e l d experiments o f t h i s t y p e . Inspection o f the standard d e v i a t i o n s of the i n d i v i d u a l sample means r e v e a l s that (expressed as C . V ' s ) they range from 2.8% to 4.2%, i n d i c a t i n g a high degree o f r e p r o d u c i b i l i t y i n the l a b o r a t o r y subsampling and a n a l y t i c a l procedures. A d e t a i l e d a n a l y s i s o f v a r i a n c e (3.) g i v e s a value o f 14.6 as the component o f v a r i a n c e between f i e l d samples i n comparison with 2.0 for the component between l a b o r a t o r y subsamples. This shows that i n the EC p l o t the dominant source o f u n c e r t a i n t y i s due to v a r i a t i o n s between the f i v e f i e l d samples and not due to v a r i a t i o n s i n the l a b o r a t o r y subsamples or i n the a c t u a l chemical a n a l y s e s . A s i m i l a r c o n c l u s i o n can be drawn by i n s p e c t i o n o f the standard d e v i a t i o n s presented i n Table I . The samples from the p l o t r e c e i v i n g the granulated formulation (Table I) revealed a d i f f e r e n t s i t u a t i o n . The o v e r a l l v a r i a b i l i t y was much l a r g e r with a c o e f f i c i e n t o f v a r i a t i o n o f 39%, with increased v a r i a b i l i t y both between and w i t h i n the f i v e samples. A n a l y s i s o f variance gave components o f 140 between f i e l d samples, and 264 between l a b o r a t o r y subsamples. Since the a n a l y t i c a l procedures were i d e n t i c a l with those used i n the EC p l o t samples, where r e p r o d u c i b i l i t y was good, these r e s u l t s c l e a r l y i n d i c a t e a much greater i r r e g u l a r i t y o f the d i s t r i b u t i o n o f the h e r b i c i d e i n


3. TAYLOR ET AL.Measurements of Trifluralin Disappearance

29

the GF p l o t samples. This i r r e g u l a r i t y i n the i n d i v i d u a l analyses themselves probably r e f l e c t s the number o f granules present i n the 100-gram subsamples withdrawn for a n a l y s e s . The granular formulation consisted o f i r r e g u l a r g r a i n s between 0.6 and 1.0mm diameter which probably survived as i n d i v i d u a l u n i t s through the mixing and r i f f l i n g procedure i n the sample p r e p a r a t i o n .


Table I .

Replicate analyses o f t r i f l u r a l i n content o f f i e l d samples o f s o i l r e c e i v i n g e m u l s i f i a b l e concentrate and granular f o r m u l a t i o n . (Samples taken on day o f application.) T r i f l u r a l i n Content

2

(mg/m )

Emulsion Sample

Granular Sample

I

II

III

IV

V

37 37 40 37 37 35 35 37 36.9 1.55

39 37 41 39 39 37 41 41 39 .3 1.67

34 34 35 37 37 34 35 36 35.3 1.28

35 35 35 35 39 37 37 37 36.3 1.49

46 47 44 46 44 44 44 44 44.9 1.25

Grand Mean and S . D . (of a l l samples) 38.5 + 3.80 A n a l y s i s o f Variance Component o f variance Between samples 14.6 Within samples 2.0

I 97 61 58 50 85 64 55 107 72.1 21.3

II

III

IV

V

46 36 22 17 75 47 35 20 37.3 19.1

22 82 46 53 32 48 43 43 46.1 17.5

39 50 33 49 37 40 67 44 44.9 10.7

52 57 40 48 45 50 39 60 48".9 7.5

49.9 +

19.4

140 264

In terms o f confidence l i m i t s the two Grand Means can be w r i t t e n as 38.5 + 6.4 mg/m for the EC p l o t and 49.9 + 32.7 mg/m for the GF p l o t at the 90% l e v e l . This statement emphasizes the extent to which sampling v a r i a b i l i t y can a f f e c t the confidence with which an a n a l y t i c a l r e s u l t i s known. Unless the sampling program i s designed to measure and i d e n t i f y the source o f the v a r i a b i l i t y much e f f o r t towards improvement o f the q u a l i t y o f the chemical analyses can be wasted (4). The d i f f i c u l t y o f improving the sampling procedures to reduce the v a r i a b i l i t y i s i l l u s t r a t e d by c a l c u l a t i o n o f the number o f samples that would have to be analyzed to o b t a i n estimates known to have an u n c e r t a i n t y l e s s than 10% at the 90% confidence l e v e l (4). This would r e q u i r e 106 analyses from the EC p l o t and 2140 from the GF. Both sample s i z e s 2


2

30


were u n r e a l i s t i c i n terms o f the resources a v a i l a b l e for the present work. The only v i a b l e course was therefore to perform the maximum number o f analyses p o s s i b l e and to subject the r e s u l t s to a f u l l s t a t i s t i c a l a n a l y s i s to quantify the u n c e r t a i n t y .


Change i n V a r i a b i l i t y With Time Table II contains the mean values o f the quadruplicate analyses o f each o f the f i v e samples taken from the EC p l o t on days 13, 48, 83 and 139, together with standard d e v i a t i o n s and the r e s u l t s o f the analyses o f v a r i a n c e . These may be compared with the data from the GF p l o t given i n Table I I I . The r e s u l t s i n Table II show t h a t , as i n the data o f Table I , the d i f f e r e n c e s between the f i v e f i e l d samples from the EC p l o t remained the main source o f v a r i a t i o n i n the data throughout the season: t h i s i s s t r i k i n g l y evident on the 139th day, where the v a r i a n c e between samples was 10.8 while the average value for the chemical analyses w i t h i n samples was only 0.4. Thus, although there was a d e c l i n e i n t r i f l u r a l i n content o f the samples with time there was no c o n s i s t e n t evidence o f r e d i s t r i b u t i o n o f the h e r b i c i d e to give a more uniform content o f the s o i l , and the f i e l d sampling was always the p r i n c i p a l source o f u n c e r t a i n t y . It may be noted that s i n c e the choice o f sampling p o i n t s i n the f i e l d was re-randomized each day, a p a r t i c u l a r number (I through V) does not represent the same set o f p o i n t s : temporal changes i n mean values (or standard d e v i a t i o n s ) a s s o c i a t e d with each sample number do not therefore have p h y s i c a l meaning.

Table I I .

Mean values (with standard d e v i a t i o n s ) o f four t r i f l u r a l i n analyses on f i v e f i e l d samples o f s o i l r e c e i v i n g e m u l s i f i a b l e concentrate, sampled from 13 to 139 days a f t e r a p p l i c a t i o n . T r i f l u r a l i n Content (mg/m*)

Age(days) 13 Sample Mean SD I II III IV V

38 29 27 29 33

31.2 Mean S.D. 4.4 Variance Between Within

1.4 4.2 0.5 2.6 1.0

48 Mean 23 21 16 30 32

83 SD 1.7 1.0 1.0 1.4 0.6

24.4 6.6 17 5.6

Mean 17 24 23 17 2.0

SD 1.3 0.5 2.1 1.4 4.3

8.3 13.0 7.7 10.0 16.0

0.2 0.8 0.6 0.7 0.6

11.0 3.5

20.2 3.3 41 1.4

139 SD Mean

9.4 5.3


10.8 0.4


3.

TAYLOR ET AL.

31

Measurements of Trifluralin Disappearance

The data for the samples from the GF p l o t presented i n Table I I I show a d i f f e r e n t p i c t u r e . Although the v a r i a b i l i t y i n these samples i s higher than that i n the EC p l o t s throughout the season, the s i t u a t i o n found i n the f i r s t sampling, where the dominant c o n t r i b u t i o n to the variance was due to the i r r e g u l a r d i s t r i b u t i o n of the h e r b i c i d e w i t h i n the l a b o r a t o r y samples d i d not p e r s i s t . On the eighteenth day the a n a l y s i s o f variance and the standard d e v i a t i o n presented i n Table I I I show that the between samples v a r i a t i o n was dominant and remained so throughout the season. This r e f l e c t s the breakdown o f the t r i f l u r a l i n granules and some r e d i s t r i b u t i o n throughout the s o i l . How f a r t h i s represents a c t u a l h e r b i c i d e movement i n the s o i l i s u n c e r t a i n because i t may merely r e f l e c t a decay o f the strength or the i n t e g r i t y o f i n d i v i d u a l granules so that they became more r e a d i l y mixed throughout the bulk o f the f i e l d s o i l sample as t h i s was s t i r r e d during the vigorous r o l l i n g a c t i o n i n the concrete mixer. The analyses o f variance summarized i n Tables II and I I I show that the i r r e g u l a r d i s t r i b u t i o n o f the t r i f l u r a l i n c o n t r i b u t e d to the higher o v e r a l l v a r i a b i l i t y i n the GF p l o t throughout the experiment, even though the v a r i a t i o n s between the means o f the f i e l d samples o f both emulsion and g r a n u l a r p l o t s were not d i s - s i m i l a r ( c o e f f i c i e n t s o f v a r i a t i o n o f 32% and 28% r e s p e c t i v e l y ) a f t e r about 140 days. Table I I I .

Mean values (with standard d e v i a t i o n s ) o f four t r i f l u r a l i n analyses on f i v e f i e l d samples o f s o i l r e c e i v i n g g r a n u l a r f o r m u l a t i o n , sampled from 18 to 144 days a p p l i c a t i o n . T r i f l u r a l i n Content

18 Age(days) Sample Mean

SD

Mean

I II III IV V

4.8 7.3 9.7 2.6 6.6

23 43 45 19 32

26 56 56 26 31

Mean 39.0 S.D. 15.7 Variance Between Within

88

53 SD 5.0 6.2 6.2 2.4 1.9

32.4 11.6 208 51

2

(mg/m )

Mean 27 30 16 29 23

SD 1.7 3.5 2.5 3.0 2.5

17 22 19 11 25

5.2 1.7 1.4 0.8 1.0

18.8 5.3

25.0 5.7 134 22

144 SD Mean

29 7.3

25 6.7

Estimation o f T r i f l u r a l i n Disappearance Rate. The t r i f l u r a l i n disappearance r a t e can be estimated by r e g r e s s i o n a n a l y s i s i n two ways, e i t h e r by using the f i v e i n d i v i d u a l sample means obtained



32


each day, or from the values o f the grand means o f a l l the data from each day. In p r a c t i c a l terms t h i s represents the use o f the r e s u l t s from a l l f i v e separate samples obtained each day or a s i n g l e determination that would have been obtained i f a l l f i v e were bulked and mixed before a n a l y s i s . The l a t t e r proceedure i s one that could be adopted to reduce the number and cost o f chemical a n a l y s e s . S t a t i s t i c a l l y the d i f f e r e n c e between these approaches i s that i n the f i r s t the f i n a l estimate of the u n c e r t a i n t y i n the r e g r e s s i o n includes the whole variance (except that due to a n a l y t i c a l v a r i a t i o n w i t h i n samples), while the second r e f l e c t s only the d e v i a t i o n o f the d a i l y grand means from the regression l i n e . Comparison o f the two c a l c u l a t i o n s r e v e a l s the amount o f information l o s t i n the second procedure. Comparison o f the regressions based on a simple l i n e a r equation P = ?^ - c . t (where Pj^ i s the i n i t i a l t r i f l u r a l i n c o n c e n t r a t i o n , t i s time i n days and c i s a constant)with r e g r e s s i o n upon a l o g a r i t h m i c curve, log P = log P^ - c . t , showed that the l a t t e r equation gave s l i g h t l y b e t t e r c o r r e l a t i o n c o e f f i c i e n t s i n both c a s e s . The data were therefore analyzed i n terms o f the l o g a r i t h m i c equation. I t must be emphasizedthat t h i s does not imply that the t r i f l u r a l i n disappearance followed f i r s t - o r d e r k i n e t i c s , but only that the data were best described by the l o g a r i t h m i c e q u a t i o n . The use o f more complex exponential decay curves was explored but no improvement i n c o r r e l a t i o n c o e f f i c i e n t s was found. The use o f such curves would introduce unnecessary complexities i n t o the a n a l y s i s . I t may a l s o be noted that r e g r e s s i o n a n a l y s i s u s i n g a l l the data p o i n t s i n s t e a d o f those from sample means gave no meaningful improvement. For s i m p l i c i t y , both sets o f data were normalized by s e t t i n g the i n i t i a l t r i f l u r a l i n c o n c e n t r a t i o n (?^) at 100%. This n o r m a l i z a t i o n was done by performing a r e g r e s s i o n on the raw data to determine the best value o f the i n t e r c e p t at t = 0, and then expressing a l l the analyses as a percentage f r a c t i o n o f t h i s f

Table I V . Regression and C o r r e l a t i o n C o e f f i c i e n t s (with Confidence L i m i t s ) for decrease i n t r i f l u r a l i n content o f f i e l d s o i l f o r 140 days a f t e r a p p l i c a t i o n , based upon the equalion log P = log (100%) - c . t . f

c» Emulsion

(+)

confidence l i m i t s

95%

D a i l y means Sample means

90%

r

80%

0.0037 0.0038

0.0006 0.0008

0.0005 0.0007

0.0004 0.0005

0.992 0.909

0.0028 0.0028

0.0005 0.0010

0.0004 0.0009

0.0003 0.0007

0.988 0.745

Granu l a r D a i l y means Sample means



33

value. These i n t e r c e p t s were 37 mg/m2 for the emulsion and 44 mg/m for the g r a n u l a r . This n o r m a l i z a t i o n reduces the data to a form where the two sets can be e a s i l y compared. The r e s u l t s o f the r e g r e s s i o n s o f the normalized data are presented i n Table I V . In both cases the values o f the decay constant are not a f f e c t e d when the f u l l set o f sample means are s u b s t i t u t e d for the d a i l y grand means. The values o f the r e g r e s s i o n c o e f f i c i e n t s however decrease c o n s i d e r a b l y , r e f l e c t i n g the l a r g e r s c a t t e r o f the i n d i v i d u a l sample r e s u l t s . The values o f the confidence l i m i t s to be assigned to the decay constant representing the l i m i t s w i t h i n which the "true" value can be expected to l i e at the 95, 90 and 80% p r o b a b i l i t i e s are a l s o wider for the i n d i v i d u a l sample d a t a . These d i f f e r e n c e s i l l u s t r a t e very c l e a r l y how an a r t i f i c i a l increase i n the confidence to be placed upon a r e s u l t can be obtained by the use o f averaged f i g u r e s that r e j e c t inherent v a r i a b i l i t y i n the o r i g i n a l d a t a . This i s p a r t i c u l a r l y important i n the present case because the most s t r i k i n g feature o f Table IV i s the d i f f e r e n c e between the decay constants o f the two f o r m u l a t i o n s . Presented i n terms o f " h a l f - l i v e s " these represent 81 days f o r the emulsion and 106 days for the granular f o r m u l a t i o n . I f r e a l , the d i f f e r e n c e o f 25 days represents an important e f f e c t o f formulation upon the environmental behavior o f t r i f l u r a l i n . I t i s e s s e n t i a l to enquire f u r t h e r i n t o the confidence that can be placed i n t h i s observation.


2

An a n a l y s i s o f covariance (3.) using the d a i l y grand means gave an "F" value o f 5.37 with 1/6 degrees o f freedom: t h i s i n d i c a t e d that the "c" values are s i g n i f i c a n t l y d i f f e r e n t at the 90% but not the 95% p r o b a b i l i t y l e v e l . A s i m i l a r a n a l y s i s using the sample means however g i v e s F = 2.307 at 1/47 degrees o f freedom, i n d i c a t i n g s i g n i f i c a n c e at 75% but not at 90% p r o b a b i l i t y . Since the a n a l y s i s using the sample means takes i n t o account the f u l l variance present i n the o r i g i n a l data, and i n c l u d e s a f a r higher number o f degrees o f freedom, i t represents a much more c r i t i c a l test. I t must be concluded that at the 90% l e v e l there i s no s i g n i f i c a n t d i f f e r e n c e between the two estimates o f the disappearance rates o f the two t r i f l u r a l i n f o r m u l a t i o n s . As a f i n a l r e s u l t , a s i n g l e r e g r e s s i o n using the sample means f o r both emulsion and granulated p l o t s g i v e s the o v e r a l l r e g r e s s i o n equation log P = 2.00 - 0.0033 (+ 0 . 0 0 0 5 ) . t with confidence l i m i t s o f 90% p r o b a b i l i t y : t h i s represents a best estimate for the h a l f l i f e o f 91 days, with a 90% confidence that the "true" value i s between 79 and 108 days. In p h y s i c a l terms t h i s represents the decay rate o f t r i f l u r a l i n i n Harford s i l t loam during the growing season i n a c o o l and r a t h e r wet summer i n upstate New York: s i n c e the decay r a t e w i l l change r a d i c a l l y during the severe winter c o n d i t i o n s t y p i c a l o f t h i s l o c a t i o n , t h i s estimate cannot be extrapolated to o b t a i n an estimate o f the disappearance r a t e over a much more extended time p e r i o d .



34


Discussion The r e s u l t s presented and discussed i n t h i s paper are an example of how i r r e g u l a r i t i e s i n h e r b i c i d e d i s t r i b u t i o n i n a f i e l d s o i l cause sampling v a r i a t i o n s so great that they severely l i m i t our confidence i n how w e l l we can measure the amount o f h e r b i c i d e present i n the f i e l d at any time and i t s rate o f disappearance. Comparison o f the r e s u l t s from the two p l o t s i n the experiment reveals that t h i s i r r e g u l a r i t y o f d i s t r i b u t i o n can come from d i f f e r e n t causes or combinations o f causes. The most uniform, but far from p e r f e c t , d i s t r i b u t i o n was observed a f t e r the h e r b i c i d e was a p p l i e d as a spray o f water-based emulsion to the s o i l surface followed by a double t i l l a g e to incorporate the h e r b i c i d e i n t o the soil Here the i r r e g u l a r i t y probably r e s u l t s from i r r e g u l a r i t i e s in the o r i g i n a l spray p a t t e r n coupled with incomplete and i r r e g u l a r mixing i n t o the s o i l (5): other work with p e r s i s t e n t s o i l i n s e c t i c i d e s a p p l i e d i n the same way (1_) has revealed s i m i l a r i r r e g u l a r i t i e s i n d i s t r i b u t i o n with l a r g e v a r i a t i o n s i n i n s e c t i c i d e contents o f i n d i v i d u a l cores taken 35 to 40cm a p a r t . The r e s u l t s a l s o show much greater v a r i a b i l i t y i n a p l o t r e c e i v i n g a surface a p p l i c a t i o n o f granules 0.6 to 1mm i n diameter followed by i n c o r p o r a t i o n . Even through there was some improvement with time as the granules decayed and were more e a s i l y mixed throughout the whole s o i l during sample p r e p a r a t i o n , the v a r i a b i l i t y i n the data was much higher owing to the superimposed i r r e g u l a r i t y o f d i s t r i b u t i o n o f the h e r b i c i d e i n the l a b o r a t o r y subsamples. These d i f f e r e n c e s emphasize the advantages i n sampling s t r a t e g i e s which permit c a l c u l a t i o n o f confidence l i m i t s f o r the data obtained and the i d e n t i f i c a t i o n o f sources o f e r r o r . With t h i s information i t i s then p o s s i b l e to recognize that there i s a l i m i t to how w e l l the h e r b i c i d e content o f the s o i l i n the f i e l d can be measured i n p r a c t i c e and that increased e f f o r t may only lead to a b e t t e r measurement o f the u n c e r t a i n t y o f the estimate rather than the number i t s e l f . Attempts to increase the p r e c i s i o n of the a n a l y t i c a l procedure may a l s o be i r r e l e v a n t and the most important issue w i l l be to design a sampling procedure that w i l l represent the most e f f e c t i v e employment o f the a v a i l a b l e resources. Following the procedure o f Snedecor and Cochran (3.) i t may be estimated that the variance i n the f i n a l estimate o f the t r i f l u r a l i n content o f the f i e l d s o i l i n the f i r s t day could be reduced by about h a l f i n the EC p l o t and by about four i n the GF p l o t i f , on each sampling, twenty-five separate samples were taken and each analyzed t w i c e . This would represent a c o n s i d e r a b l y increased c o s t , i n c r e a s i n g the number o f determinations from f o r t y to f i f t y and, more i m p o r t a n t l y , i n c r e a s i n g f i v e - f o l d the number o f sampling operations such as mixing, d r y i n g and weighing. Whether t h i s i s j u s t i f i a b l e depends upon the o b j e c t i v e s o f the experiment. In the present case t h i s o b j e c t i v e was to estimate the disappearance rate o f t r i f l u r a l i n a p p l i e d i n a r e g u l a r




35

commercial manner to a large experimental f i e l d . When the r e s u l t s obtained were compared with data from the remainder o f the p r o j e c t i t was c l e a r that t h i s o b j e c t i v e was achieved w i t h i n the framework o f the o v e r a l l p r o j e c t . A l l o c a t i o n o f increased resources to reduce the u n c e r t a i n t y i n the f i n a l r e s u l t c o u l d not therefore have been j u s t i f i e d . I f , however, the o b j e c t i v e had been to measure the disappearance rate o f the t r i f l u r a l i n with the g r e a t e s t p r e c i s i o n p o s s i b l e , i t i s c l e a r that an e n t i r e l y d i f f e r e n t experimental program, with i n t e n s i v e sampling o f smaller and more c a r e f u l l y c o n t r o l l e d experimental p l o t s would have been appropriate. Even i n more p r e c i s e measurements i t i s however important that the sampling program be designed to g i v e an estimate o f the confidence that can be placed upon the r e s u l t o b t a i n e d . Not only does the r e c o g n i t i o n and measurement o f the existence o f v a r i a b i l i t y and u n c e r t a i n t y i n f i e l d measurements add to t h e i r a u t h e n t i c i t y and v a l u e , but i t i s an e s s e n t i a l element i n the competent design and management o f an experimental program. This becomes important not only i n the choice o f a sampling p r o t o c o l , but i n the choice o f a sampling schedule. Without measurement o f the standard d e v i a t i o n resources may be wasted by too frequent samplings i f the i n t e r v a l between them i s not long enough to allow a s t a t i s t i c a l l y measurable change to have taken p l a c e . The choice o f sampling schedules w i l l o f course r e q u i r e some p r e l i m i n a r y estimate o f the disappearance r a t e , and sampling at shorter i n t e r v a l s w i l l be e s s e n t i a l for more unstable m a t e r i a l s . A m u l t i p l e range t e s t o f the r e s u l t s i n the present case r e v e a l that the d i f f e r e n c e s between samplings were s i g n i f i c a n t at the 5% l e v e l except between those o f the 53rd and 88th days on the EC p l o t . In other work with much more p e r s i s t e n t m a t e r i a l s (6,7) much longer i n t e r v a l s may be r e q u i r e d to f i n d s i g n i f i c a n t d i f f e r e n c e s . Literature Cited

1. Taylor, A. W.; Freeman, H.P.; Edwards, W. M. J . Agric. Food Chem. 1971, 19, p. 832-6 2. Caro, J . H.; Taylor, A. W. Proc Int. Conf. on Environmental Sensing and Assessment. 1976. Inst. Electrical and Electronic Eng.: Washington, D.C., pt 1-3, pp. 1-5. 3. Snedecor, G. W.; Cochran, W. G. "Statistical Methods"; 6th edit. Iowa State University Press: Ames, Iowa, 1967. 4. Kratochvil, B. The Role of Chemometrics in Pesticide/ Environmental Residue Analytical Determinations. 1984. ACS Symposium Series. (This volume). 5. Read, K.; Gebhardt, M. R.; Day, C. L. Trans. Am Soc. Agric. Eng. 1968, 11, 155-8. 6. Taylor, A. W.; Barrows, H. L. Proc 2ndIUPACCong Pesticide Chem. 1971, IV, 457-75. 7. Freeman, H. P.; Taylor, A. W.; Edwards, W. M. J . Ag. Food Chem. 1975, 6, 1101-5. RECEIVED March 25, 1985


Sampling and Variance in Measurements of Trifluralin Disappearance

Recommend Documents