glinski_stemflow
tested hypotheses
Test to see whether field observations of stemflow concentrations are greater that forest throughfall concentrations (or vice-versa). Samples are paired across 3 locations and 18 different sampling dates. Therefore 54 observations for each of 32 chemicals. However, many samples are non-detected.
Non-parametric Wilcoxon Rank Sum test implemented. Permutation tests also possible. Test requiring distribution assumptions (e.g., t-test) not practical due to non-detects and data skewness.
This page is published at: http://rpubs.com/puruckertom/glinski_stemflow2
figures needed
- scatterplot of stemflow concentration versus throughfall concentrations (fig #6)
- figure 6 as jpg (add units to x and y - ug/L)
- a single plot stacked surface water concentration versus time for all sites (all 10 tifton pond sites- metolachlor, tebuconazole) figure 3 redone
- figure 4 stacked clustered surface water (4L samples) by site*month (Ponds_4L.csv)
- a single plot concentration versus time plots for stemflow and throughfall by site, 6 different symbol/color/filled) figure 5 redone (metolachlor, tebuconazole))
- signs versus time plot to support wrs tests
- signs versus rainfall intensity (cumulative rainfall) to support wrs tests
possible additional things to do
- wrs tests for stemflow greater than throughfall by date
- wrs tests for stemflow greater than throughfall by site (metolachlor)
- signs versus volume(s)
- implement permutation tests
- test all 32 chemicals at once rather than one at a time (done)
- effects estimation approaches
general configuration
Install and load supporting libraries.
## nodename
## "DZ2626UTPURUCKE"## [1] "R version 3.3.2 (2016-10-31)"## Warning: package 'ggplot2' was built under R version 3.3.3## [1] "list of loaded packages: "## [1] "ggplot2" "knitr" "dplyr" "rmarkdown" "stats"
## [6] "graphics" "grDevices" "utils" "datasets" "methods"
## [11] "base"import experimental exposure and dehydration data
Load csv file with experimental dehydration data. The below may return false but still be OK if rstudio does not have privileges to data directory (e.g., attached drive).
## [1] "Root directory location: k:/git/glinski_stemflow/"## [1] "check to see if R can access files OK: FALSE"tifton data structure
Check out structure of imported data sets. sf v tf (32 chemicals, 3 paired sites, 52 storm events), pondwater (10 samples * 12 months)
summary(tifton)## Type Site Date Compound
## SF:1728 Min. : 1.000 1/17/2016 : 512 2-Phenylphenol: 228
## TF:1728 1st Qu.: 3.000 10/8/2015 : 512 Acetochlor : 228
## TP:3840 Median : 4.000 11/8/2015 : 512 Alachlor : 228
## Mean : 5.105 12/17/2015: 512 Anthraquione : 228
## 3rd Qu.: 7.000 5/22/2015 : 512 Atrazine : 228
## Max. :10.000 6/23/2015 : 512 Benfluralin : 228
## (Other) :4224 (Other) :5928
## Conc Volume
## Min. : 0.010 Min. : 0.132
## 1st Qu.: 0.170 1st Qu.: 2.000
## Median : 0.260 Median : 5.000
## Mean : 0.454 Mean : 9.266
## 3rd Qu.: 0.400 3rd Qu.:16.000
## Max. :10.500 Max. :25.000
## NA's :6502 NA's :4416colnames(tifton)## [1] "Type" "Site" "Date" "Compound" "Conc" "Volume"chems <- unique(tifton$Compound)
chems## [1] Metolachlor Tebuconazole
## [3] 2-Phenylphenol Flutolanil
## [5] Biphenyl Atrazine
## [7] Metalaxyl Acetochlor
## [9] Alachlor Anthraquione
## [11] Endosulfan Lactone DEA
## [13] Endosulfan ether Cyprodinil
## [15] Triadimefon Chlorothalonil
## [17] Diphenylamine Fipronil
## [19] Myclobutanil Oxyfluorfen
## [21] Bifenthrin Fludioxonil
## [23] Endosulfan sulfate Ethalfluralin
## [25] Pendimethalin Benfluralin
## [27] Diazinon Malathion
## [29] Oxadiazon Piperonyl butoxide
## [31] Propyzamide THPI (Tetrahydrophthalimide)
## 32 Levels: 2-Phenylphenol Acetochlor Alachlor Anthraquione ... Triadimefonmedia <- unique(tifton$Type)
media## [1] TP SF TF
## Levels: SF TF TP#coerce concentrations to numeric because NAs
tifton$Conc <- as.numeric(tifton$Conc)
min(tifton$Conc,na.rm=TRUE)## [1] 0.01max(tifton$Conc,na.rm=TRUE)## [1] 10.5colnames(rainwater)## [1] "event" "amount_inches"colnames(rainwater)[1] <- 'Date'
summary(rainwater)## Date amount_inches
## 1/17/2016 : 1 Min. :0.150
## 10/8/2015 : 1 1st Qu.:1.123
## 11/8/2015 : 1 Median :1.927
## 12/17/2015: 1 Mean :2.137
## 3/18/2015 : 1 3rd Qu.:2.556
## 4/16/2015 : 1 Max. :6.760
## (Other) :12#chem_merge <- merge(chem_merge, rainwater, by="Date")wilcox stat testing
Run Wilcoxon rank sum test on throughfall versus stemflow for each individual chemical that has detected values. There are three different commands in R that calculate the Wilcoxon signed-rank test; wilcox.test, wilcox.exact, and wilcoxsign_test. They will each yield a different p-value.
The nonparametric Wilcoxon signed-rank test seems to be the most appropriate test for these data. There are two different methods to calculate the signed-rank test. The first is by Wilcoxon (1945), who discards any tied data and then calculates the signed ranks. The second method incorporates tied values in the ranking procedure (see J.W. Pratt, 1959, Remarks on zeros and ties in the Wilcoxon signed rank procedure: Journal of the American Statistical Association, Vol. 54, No. 287, pp. 655-667). There are two commands in R that calculate the original method by Wilcoxon, wilcox.test and wilcoxsign_test (make sure to include the argument “zero.method = c(”Wilcoxon“)”). There are two other commands in R that incorporate ties in the signed-rank test, wilcox.exact and wilcoxsign_test (make sure to include the argument“zero.method = c(”Pratt“)”).
- wilcox.test() in the stats package This function automatically switches to using a Normal approximation when there are ties in the data:
wilcox.test(x, y, paired=TRUE)$p.value
(You can suppress the warning (due to ties) by specifying the argument ‘exact=FALSE’.)
This function also uses a continuity correction unless told not to:
wilcox.test(x, y, paired=TRUE, correct=FALSE)$p.value
- wilcox.exact() in pkg exactRankTests This function can handle ties (using the “Wilcoxon” method) with an ‘exact’ calculation:
wilcox.exact(x, y, paired=TRUE)$p.value
If we want the Normal approximation:
wilcox.exact(x, y, paired=TRUE, exact=FALSE)$p.value
- wilcoxsign_test() in pkg coin This is the most comprehensive of these functions. It is also the only one that offers the “Pratt” method of handling ties. It will default to this method and a Normal approximation:
pvalue(wilcoxsign_test(x ~ y))
pvalue(wilcoxsign_test(x ~ y, zero.method=“Pratt”, distribution=“asympt”))
You can get the results from wilcox.exact() with
pvalue(wilcoxsign_test(x ~ y, zero.method=“Wilcoxon”, distribution=“asympt”))
and
pvalue(wilcoxsign_test(x ~ y, zero.method=“Wilcoxon”, dist=“exact”))
For the moment we are using wilcox.test with ties (paired = TRUE, alternative = ‘greater’, exact = FALSE) but the test is not being run unless there are at least 4 detected values for a chemical.
chemical level analyses
chems_we_care_about <- c('Metolachlor', 'Tebuconazole', '2-Phenylphenol', 'Atrazine', 'Biphenyl', 'DEA', 'Flutolanil')
for(chem in chems_we_care_about){
print('=======================================================')
print(chem)
chem_data <- tifton[which(tifton$Compound==chem),]
chem_sf <- chem_data[which(chem_data$Type=='SF'),]
chem_tf <- chem_data[which(chem_data$Type=='TF'),]
chem_tp <- chem_data[which(chem_data$Type=='TP'),]
#merge on site, date, compound
chem_merge <- merge(chem_sf,chem_tf,by=c("Compound","Site","Date"))
chem_merge <- merge(chem_merge, rainwater, by="Date")
#1 plot by rainfall (metolachlor and tebuconazole)
plot(chem_merge$amount_inches, chem_merge$Conc.x, main=paste(chem,'stemflow'))
#1 plot by date (all together, with 3 different symbols/colors for site and filled/not filled for sf v tf)
plot(chem_merge$Date, chem_merge$Conc.y, title=paste(chem,'throughfall'))
plot(chem_merge$Date, chem_merge$Conc.x, main='stemflow')
#1 stacked barplots by month for all sites for tifton ponds
#conc.x is sf; conc.y is tf
#we want to test if stemflow is greater than throughfall (x > y)
x_nas <- sum(is.na(chem_merge$Conc.x))
y_nas <- sum(is.na(chem_merge$Conc.y))
x_n <- length(is.na(chem_merge$Conc.x))
y_n <- length(is.na(chem_merge$Conc.y))
x_dets <- x_n - x_nas
y_dets <- y_n - y_nas
if(x_dets>0){
if(x_nas==0){
x_min <- min(chem_merge$Conc.x,na.rm=TRUE)
}else{
x_min=0
}
x_max <- max(chem_merge$Conc.x,na.rm=TRUE)
}else{
x_min <- 0
x_max <- 0
}
if(y_dets>0){
if(y_nas==0){
y_min <- min(chem_merge$Conc.y,na.rm=TRUE)
}else{
y_min=0
}
y_max <- max(chem_merge$Conc.y,na.rm=TRUE)
}else{
y_min <- 0
y_max <- 0
}
### statistical analysis where all 3 locations are grouped
#colnames(chem_merge)
print(paste(chem,'stemflow detection frequency = ',x_dets,'/',x_n))
print(paste(chem,'stemflow concentration range =(',x_min,',',x_max,')'))
print(paste(chem,'throughfall detection frequency = ',y_dets,'/',y_n))
print(paste(chem,'throughfall concentration range =(',y_min,',',y_max,')'))
#change NAs to zeroes for conparisons and the Wilcox test
chem_merge$Conc.x[is.na(chem_merge$Conc.x)] <- 0
chem_merge$Conc.y[is.na(chem_merge$Conc.y)] <- 0
n_comparisons <- length(chem_merge$Conc.x)
n_tf_greater <- sum(chem_merge$Conc.y > chem_merge$Conc.x)
n_ties <- sum(chem_merge$Conc.y == chem_merge$Conc.x)
n_sf_greater <- sum(chem_merge$Conc.x > chem_merge$Conc.y)
comparisons <- paste('(throughfall_greater, ties, stemflow_greater) = ('
,n_tf_greater,',',n_ties,',',n_sf_greater,')')
print(comparisons)
#do the test for all sites
if((x_dets+y_dets)>4){
print(knitr::kable(chem_merge))
chem_wilcox <- wilcox.test(chem_merge$Conc.x, chem_merge$Conc.y,
alternative = 'greater', paired = TRUE, exact = FALSE)
max_conc <- max(x_max,y_max)
max_conc_vector <- seq(0,max_conc,20)
print(chem_wilcox)
plot(chem_merge$Conc.x, chem_merge$Conc.y,xlim=c(0,max_conc),ylim=c(0,max_conc),
xlab='Stemflow Concentration',ylab='Throughfall Concentration',
main=paste(chem,'(WRS p-value = ',round(chem_wilcox$p.value,4),')'),
sub = comparisons)
abline(0,1,col='red')
}else{
print(paste(chem,'does not have enough detected values for wilcox test'))
}
### statistical analysis for each location, not grouped
#do the test for each sites
}## [1] "======================================================="
## [1] "Metolachlor"
## [1] "Metolachlor stemflow detection frequency = 45 / 54"
## [1] "Metolachlor stemflow concentration range =( 0 , 2.63 )"
## [1] "Metolachlor throughfall detection frequency = 41 / 54"
## [1] "Metolachlor throughfall concentration range =( 0 , 2.98 )"
## [1] "(throughfall_greater, ties, stemflow_greater) = ( 19 , 5 , 30 )"
##
##
## Date Compound Site Type.x Conc.x Volume.x Type.y Conc.y Volume.y amount_inches
## ----------- ------------ ----- ------- ---------- --------- ------- ---------- --------- --------------
## 1/17/2016 Metolachlor 3 SF 0.2700000 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 Metolachlor 4 SF 0.1600000 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 Metolachlor 7 SF 0.2600000 25.000 TF 0.0000000 7.000 3.900
## 10/8/2015 Metolachlor 7 SF 0.4100000 17.000 TF 0.0000000 5.000 2.220
## 10/8/2015 Metolachlor 3 SF 0.2500000 20.000 TF 0.0800000 3.750 2.220
## 10/8/2015 Metolachlor 4 SF 0.3700000 25.000 TF 0.0000000 4.000 2.220
## 11/8/2015 Metolachlor 7 SF 0.4100000 5.000 TF 0.1200000 1.000 1.040
## 11/8/2015 Metolachlor 4 SF 0.2400000 15.000 TF 0.0000000 2.000 1.040
## 11/8/2015 Metolachlor 3 SF 0.0000000 NA TF 0.0000000 2.000 1.040
## 12/17/2015 Metolachlor 4 SF 0.3500000 25.000 TF 0.0000000 12.000 6.760
## 12/17/2015 Metolachlor 7 SF 0.3300000 25.000 TF 0.0000000 9.000 6.760
## 12/17/2015 Metolachlor 3 SF 0.0000000 25.000 TF 0.0000000 12.000 6.760
## 3/18/2015 Metolachlor 3 SF 0.2272727 0.132 TF 0.7559055 0.635 0.150
## 3/18/2015 Metolachlor 4 SF 0.4100000 1.000 TF 0.9090909 0.440 0.150
## 3/18/2015 Metolachlor 7 SF 0.0000000 0.230 TF 0.1600000 0.550 0.150
## 4/16/2015 Metolachlor 7 SF 0.4500000 12.500 TF 0.3400000 2.000 2.485
## 4/16/2015 Metolachlor 3 SF 0.0000000 0.670 TF 0.0000000 3.000 2.485
## 4/16/2015 Metolachlor 4 SF 0.3600000 13.000 TF 0.7100000 2.000 2.485
## 5/1/2015 Metolachlor 7 SF 1.1600000 20.000 TF 1.2000000 5.000 1.965
## 5/1/2015 Metolachlor 4 SF 0.8000000 5.000 TF 0.2700000 4.500 1.965
## 5/1/2015 Metolachlor 3 SF 0.0000000 16.000 TF 0.3200000 1.000 1.965
## 5/22/2015 Metolachlor 7 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 5/22/2015 Metolachlor 4 SF 1.3200000 25.000 TF 0.5300000 6.000 0.650
## 5/22/2015 Metolachlor 3 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 6/10/2015 Metolachlor 4 SF 0.9500000 25.000 TF 0.5400000 4.000 2.580
## 6/10/2015 Metolachlor 3 SF 0.6300000 15.000 TF 1.4000000 4.000 2.580
## 6/10/2015 Metolachlor 7 SF 0.3600000 20.000 TF 2.9800000 5.000 2.580
## 6/23/2015 Metolachlor 3 SF 0.6900000 NA TF 1.6700000 NA 1.810
## 6/23/2015 Metolachlor 7 SF 2.2000000 NA TF 2.7800000 NA 1.810
## 6/23/2015 Metolachlor 4 SF 1.7500000 NA TF 0.7100000 NA 1.810
## 7/10/2015 Metolachlor 7 SF 1.5300000 7.000 TF 0.5800000 2.000 1.370
## 7/10/2015 Metolachlor 4 SF 0.4300000 25.000 TF 0.2900000 4.000 1.370
## 7/10/2015 Metolachlor 3 SF 0.7400000 8.500 TF 0.6300000 3.000 1.370
## 7/17/2015 Metolachlor 7 SF 0.5300000 NA TF 0.9800000 NA 1.730
## 7/17/2015 Metolachlor 4 SF 0.6600000 NA TF 0.5300000 NA 1.730
## 7/17/2015 Metolachlor 3 SF 0.4500000 NA TF 0.9500000 NA 1.730
## 7/2/2015 Metolachlor 4 SF 0.5900000 0.675 TF 0.7700000 4.000 1.010
## 7/2/2015 Metolachlor 7 SF 1.5200000 3.000 TF 1.4100000 1.500 1.010
## 7/2/2015 Metolachlor 3 SF 0.0000000 NA TF 1.1300000 1.000 1.010
## 7/30/2015 Metolachlor 3 SF 0.4600000 20.000 TF 0.4700000 4.000 2.970
## 7/30/2015 Metolachlor 4 SF 0.5700000 25.000 TF 0.2700000 6.000 2.970
## 7/30/2015 Metolachlor 7 SF 0.9300000 20.000 TF 0.3800000 3.500 2.970
## 8/13/2015 Metolachlor 7 SF 0.8000000 15.000 TF 0.2900000 4.000 3.090
## 8/13/2015 Metolachlor 3 SF 0.3400000 13.000 TF 0.4700000 4.000 3.090
## 8/13/2015 Metolachlor 4 SF 0.4200000 20.000 TF 0.2500000 6.000 3.090
## 8/27/2015 Metolachlor 7 SF 0.8500000 25.000 TF 0.3600000 4.000 2.380
## 8/27/2015 Metolachlor 4 SF 0.2900000 25.000 TF 0.1400000 4.750 2.380
## 8/27/2015 Metolachlor 3 SF 0.2800000 17.000 TF 0.2900000 4.000 2.380
## 9/17/2015 Metolachlor 4 SF 0.6400000 25.000 TF 0.5500000 3.000 1.890
## 9/17/2015 Metolachlor 7 SF 2.6300000 14.000 TF 0.1300000 2.000 1.890
## 9/17/2015 Metolachlor 3 SF 0.0000000 2.500 TF 0.6400000 3.000 1.890
## 9/3/2015 Metolachlor 3 SF 0.7272727 0.220 TF 0.0289855 0.690 0.460
## 9/3/2015 Metolachlor 4 SF 0.3703704 0.540 TF 1.3300000 0.550 0.460
## 9/3/2015 Metolachlor 7 SF 0.4018547 0.647 TF 0.2068966 0.435 0.460
##
## Wilcoxon signed rank test with continuity correction
##
## data: chem_merge$Conc.x and chem_merge$Conc.y
## V = 701, p-value = 0.1907
## alternative hypothesis: true location shift is greater than 0
## [1] "======================================================="
## [1] "Tebuconazole"
## [1] "Tebuconazole stemflow detection frequency = 15 / 54"
## [1] "Tebuconazole stemflow concentration range =( 0 , 1.64 )"
## [1] "Tebuconazole throughfall detection frequency = 43 / 54"
## [1] "Tebuconazole throughfall concentration range =( 0 , 1.8 )"
## [1] "(throughfall_greater, ties, stemflow_greater) = ( 35 , 10 , 9 )"
##
##
## Date Compound Site Type.x Conc.x Volume.x Type.y Conc.y Volume.y amount_inches
## ----------- ------------- ----- ------- ---------- --------- ------- ---------- --------- --------------
## 1/17/2016 Tebuconazole 3 SF 0.0000000 25.000 TF 0.1800000 10.000 3.900
## 1/17/2016 Tebuconazole 4 SF 0.0000000 25.000 TF 0.1800000 10.000 3.900
## 1/17/2016 Tebuconazole 7 SF 0.2200000 25.000 TF 0.0000000 7.000 3.900
## 10/8/2015 Tebuconazole 7 SF 0.4700000 17.000 TF 0.2100000 5.000 2.220
## 10/8/2015 Tebuconazole 3 SF 0.0000000 20.000 TF 0.2200000 3.750 2.220
## 10/8/2015 Tebuconazole 4 SF 0.3300000 25.000 TF 0.3400000 4.000 2.220
## 11/8/2015 Tebuconazole 7 SF 0.0000000 5.000 TF 0.2600000 1.000 1.040
## 11/8/2015 Tebuconazole 4 SF 0.0000000 15.000 TF 0.3500000 2.000 1.040
## 11/8/2015 Tebuconazole 3 SF 0.0000000 NA TF 0.0000000 2.000 1.040
## 12/17/2015 Tebuconazole 4 SF 0.2900000 25.000 TF 0.2400000 12.000 6.760
## 12/17/2015 Tebuconazole 7 SF 0.0000000 25.000 TF 0.0000000 9.000 6.760
## 12/17/2015 Tebuconazole 3 SF 0.0000000 25.000 TF 0.0000000 12.000 6.760
## 3/18/2015 Tebuconazole 3 SF 0.0000000 0.132 TF 0.0000000 0.635 0.150
## 3/18/2015 Tebuconazole 4 SF 0.0000000 1.000 TF 0.0000000 0.440 0.150
## 3/18/2015 Tebuconazole 7 SF 0.0000000 0.230 TF 0.0000000 0.550 0.150
## 4/16/2015 Tebuconazole 7 SF 0.0000000 12.500 TF 0.1600000 2.000 2.485
## 4/16/2015 Tebuconazole 3 SF 0.0000000 0.670 TF 0.0000000 3.000 2.485
## 4/16/2015 Tebuconazole 4 SF 0.0000000 13.000 TF 0.0000000 2.000 2.485
## 5/1/2015 Tebuconazole 7 SF 0.0000000 20.000 TF 0.2100000 5.000 1.965
## 5/1/2015 Tebuconazole 4 SF 0.0000000 5.000 TF 0.2800000 4.500 1.965
## 5/1/2015 Tebuconazole 3 SF 0.0000000 16.000 TF 0.1900000 1.000 1.965
## 5/22/2015 Tebuconazole 7 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 5/22/2015 Tebuconazole 4 SF 0.2700000 25.000 TF 0.2200000 6.000 0.650
## 5/22/2015 Tebuconazole 3 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 6/10/2015 Tebuconazole 4 SF 0.0000000 25.000 TF 0.2000000 4.000 2.580
## 6/10/2015 Tebuconazole 3 SF 0.0000000 15.000 TF 0.1800000 4.000 2.580
## 6/10/2015 Tebuconazole 7 SF 0.0000000 20.000 TF 0.2600000 5.000 2.580
## 6/23/2015 Tebuconazole 3 SF 0.0000000 NA TF 0.1800000 NA 1.810
## 6/23/2015 Tebuconazole 7 SF 0.0000000 NA TF 0.1900000 NA 1.810
## 6/23/2015 Tebuconazole 4 SF 0.2400000 NA TF 0.2100000 NA 1.810
## 7/10/2015 Tebuconazole 7 SF 0.3600000 7.000 TF 0.2400000 2.000 1.370
## 7/10/2015 Tebuconazole 4 SF 0.0000000 25.000 TF 0.1900000 4.000 1.370
## 7/10/2015 Tebuconazole 3 SF 0.0000000 8.500 TF 0.2900000 3.000 1.370
## 7/17/2015 Tebuconazole 7 SF 0.0000000 NA TF 0.2800000 NA 1.730
## 7/17/2015 Tebuconazole 4 SF 0.0000000 NA TF 0.3500000 NA 1.730
## 7/17/2015 Tebuconazole 3 SF 0.0000000 NA TF 0.5900000 NA 1.730
## 7/2/2015 Tebuconazole 4 SF 0.0000000 0.675 TF 0.2300000 4.000 1.010
## 7/2/2015 Tebuconazole 7 SF 0.0000000 3.000 TF 0.3700000 1.500 1.010
## 7/2/2015 Tebuconazole 3 SF 0.0000000 NA TF 0.4900000 1.000 1.010
## 7/30/2015 Tebuconazole 3 SF 0.0000000 20.000 TF 0.2100000 4.000 2.970
## 7/30/2015 Tebuconazole 4 SF 0.2500000 25.000 TF 0.2700000 6.000 2.970
## 7/30/2015 Tebuconazole 7 SF 0.0000000 20.000 TF 0.2100000 3.500 2.970
## 8/13/2015 Tebuconazole 7 SF 0.2400000 15.000 TF 0.3200000 4.000 3.090
## 8/13/2015 Tebuconazole 3 SF 0.0000000 13.000 TF 0.3000000 4.000 3.090
## 8/13/2015 Tebuconazole 4 SF 0.0000000 20.000 TF 0.2600000 6.000 3.090
## 8/27/2015 Tebuconazole 7 SF 1.2800000 25.000 TF 0.6500000 4.000 2.380
## 8/27/2015 Tebuconazole 4 SF 0.6500000 25.000 TF 1.8000000 4.750 2.380
## 8/27/2015 Tebuconazole 3 SF 1.6400000 17.000 TF 0.6500000 4.000 2.380
## 9/17/2015 Tebuconazole 4 SF 0.3300000 25.000 TF 0.7900000 3.000 1.890
## 9/17/2015 Tebuconazole 7 SF 0.0000000 14.000 TF 0.2600000 2.000 1.890
## 9/17/2015 Tebuconazole 3 SF 0.0000000 2.500 TF 0.4200000 3.000 1.890
## 9/3/2015 Tebuconazole 3 SF 0.0000000 0.220 TF 0.3333333 0.690 0.460
## 9/3/2015 Tebuconazole 4 SF 0.5370370 0.540 TF 0.7454545 0.550 0.460
## 9/3/2015 Tebuconazole 7 SF 0.6800618 0.647 TF 0.6206897 0.435 0.460
##
## Wilcoxon signed rank test with continuity correction
##
## data: chem_merge$Conc.x and chem_merge$Conc.y
## V = 160.5, p-value = 1
## alternative hypothesis: true location shift is greater than 0
## [1] "======================================================="
## [1] "2-Phenylphenol"
## [1] "2-Phenylphenol stemflow detection frequency = 15 / 54"
## [1] "2-Phenylphenol stemflow concentration range =( 0 , 0.68 )"
## [1] "2-Phenylphenol throughfall detection frequency = 23 / 54"
## [1] "2-Phenylphenol throughfall concentration range =( 0 , 0.545454545 )"
## [1] "(throughfall_greater, ties, stemflow_greater) = ( 21 , 23 , 10 )"
##
##
## Date Compound Site Type.x Conc.x Volume.x Type.y Conc.y Volume.y amount_inches
## ----------- --------------- ----- ------- ---------- --------- ------- ---------- --------- --------------
## 1/17/2016 2-Phenylphenol 3 SF 0.0000000 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 2-Phenylphenol 7 SF 0.0000000 25.000 TF 0.0000000 7.000 3.900
## 10/8/2015 2-Phenylphenol 7 SF 0.0000000 17.000 TF 0.0000000 5.000 2.220
## 10/8/2015 2-Phenylphenol 3 SF 0.0000000 20.000 TF 0.1900000 3.750 2.220
## 10/8/2015 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.1900000 4.000 2.220
## 11/8/2015 2-Phenylphenol 7 SF 0.3700000 5.000 TF 0.0000000 1.000 1.040
## 11/8/2015 2-Phenylphenol 4 SF 0.0000000 15.000 TF 0.2100000 2.000 1.040
## 11/8/2015 2-Phenylphenol 3 SF 0.0000000 NA TF 0.3000000 2.000 1.040
## 12/17/2015 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.0600000 12.000 6.760
## 12/17/2015 2-Phenylphenol 7 SF 0.0400000 25.000 TF 0.0000000 9.000 6.760
## 12/17/2015 2-Phenylphenol 3 SF 0.1700000 25.000 TF 0.2100000 12.000 6.760
## 3/18/2015 2-Phenylphenol 3 SF 0.0000000 0.132 TF 0.3622047 0.635 0.150
## 3/18/2015 2-Phenylphenol 4 SF 0.0000000 1.000 TF 0.5454545 0.440 0.150
## 3/18/2015 2-Phenylphenol 7 SF 0.0000000 0.230 TF 0.2300000 0.550 0.150
## 4/16/2015 2-Phenylphenol 7 SF 0.0000000 12.500 TF 0.0000000 2.000 2.485
## 4/16/2015 2-Phenylphenol 3 SF 0.1194030 0.670 TF 0.2200000 3.000 2.485
## 4/16/2015 2-Phenylphenol 4 SF 0.0000000 13.000 TF 0.0000000 2.000 2.485
## 5/1/2015 2-Phenylphenol 7 SF 0.0000000 20.000 TF 0.0000000 5.000 1.965
## 5/1/2015 2-Phenylphenol 4 SF 0.0000000 5.000 TF 0.2000000 4.500 1.965
## 5/1/2015 2-Phenylphenol 3 SF 0.0000000 16.000 TF 0.2000000 1.000 1.965
## 5/22/2015 2-Phenylphenol 7 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 5/22/2015 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.0000000 6.000 0.650
## 5/22/2015 2-Phenylphenol 3 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 6/10/2015 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.1900000 4.000 2.580
## 6/10/2015 2-Phenylphenol 3 SF 0.0000000 15.000 TF 0.0000000 4.000 2.580
## 6/10/2015 2-Phenylphenol 7 SF 0.0800000 20.000 TF 0.0000000 5.000 2.580
## 6/23/2015 2-Phenylphenol 3 SF 0.0000000 NA TF 0.0000000 NA 1.810
## 6/23/2015 2-Phenylphenol 7 SF 0.0000000 NA TF 0.0000000 NA 1.810
## 6/23/2015 2-Phenylphenol 4 SF 0.0000000 NA TF 0.0000000 NA 1.810
## 7/10/2015 2-Phenylphenol 7 SF 0.0000000 7.000 TF 0.0000000 2.000 1.370
## 7/10/2015 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.0000000 4.000 1.370
## 7/10/2015 2-Phenylphenol 3 SF 0.0400000 8.500 TF 0.0000000 3.000 1.370
## 7/17/2015 2-Phenylphenol 7 SF 0.0400000 NA TF 0.0000000 NA 1.730
## 7/17/2015 2-Phenylphenol 4 SF 0.0000000 NA TF 0.1900000 NA 1.730
## 7/17/2015 2-Phenylphenol 3 SF 0.0000000 NA TF 0.2800000 NA 1.730
## 7/2/2015 2-Phenylphenol 4 SF 0.0740741 0.675 TF 0.0000000 4.000 1.010
## 7/2/2015 2-Phenylphenol 7 SF 0.0000000 3.000 TF 0.0000000 1.500 1.010
## 7/2/2015 2-Phenylphenol 3 SF 0.0000000 NA TF 0.2000000 1.000 1.010
## 7/30/2015 2-Phenylphenol 3 SF 0.3700000 20.000 TF 0.2000000 4.000 2.970
## 7/30/2015 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.0000000 6.000 2.970
## 7/30/2015 2-Phenylphenol 7 SF 0.0000000 20.000 TF 0.1900000 3.500 2.970
## 8/13/2015 2-Phenylphenol 7 SF 0.0000000 15.000 TF 0.0000000 4.000 3.090
## 8/13/2015 2-Phenylphenol 3 SF 0.0000000 13.000 TF 0.0000000 4.000 3.090
## 8/13/2015 2-Phenylphenol 4 SF 0.0000000 20.000 TF 0.0000000 6.000 3.090
## 8/27/2015 2-Phenylphenol 7 SF 0.0000000 25.000 TF 0.0000000 4.000 2.380
## 8/27/2015 2-Phenylphenol 4 SF 0.0000000 25.000 TF 0.0000000 4.750 2.380
## 8/27/2015 2-Phenylphenol 3 SF 0.0500000 17.000 TF 0.0000000 4.000 2.380
## 9/17/2015 2-Phenylphenol 4 SF 0.0100000 25.000 TF 0.1900000 3.000 1.890
## 9/17/2015 2-Phenylphenol 7 SF 0.0000000 14.000 TF 0.2300000 2.000 1.890
## 9/17/2015 2-Phenylphenol 3 SF 0.6800000 2.500 TF 0.4700000 3.000 1.890
## 9/3/2015 2-Phenylphenol 3 SF 0.1818182 0.220 TF 0.2898551 0.690 0.460
## 9/3/2015 2-Phenylphenol 4 SF 0.0370370 0.540 TF 0.0000000 0.550 0.460
## 9/3/2015 2-Phenylphenol 7 SF 0.0154560 0.647 TF 0.4827586 0.435 0.460
##
## Wilcoxon signed rank test with continuity correction
##
## data: chem_merge$Conc.x and chem_merge$Conc.y
## V = 100, p-value = 0.9982
## alternative hypothesis: true location shift is greater than 0
## [1] "======================================================="
## [1] "Atrazine"
## [1] "Atrazine stemflow detection frequency = 10 / 54"
## [1] "Atrazine stemflow concentration range =( 0 , 3.65 )"
## [1] "Atrazine throughfall detection frequency = 17 / 54"
## [1] "Atrazine throughfall concentration range =( 0 , 7.09 )"
## [1] "(throughfall_greater, ties, stemflow_greater) = ( 12 , 35 , 7 )"
##
##
## Date Compound Site Type.x Conc.x Volume.x Type.y Conc.y Volume.y amount_inches
## ----------- --------- ----- ------- ------- --------- ------- ------- --------- --------------
## 1/17/2016 Atrazine 3 SF 0.00 25.000 TF 0.00 10.000 3.900
## 1/17/2016 Atrazine 4 SF 0.00 25.000 TF 0.00 10.000 3.900
## 1/17/2016 Atrazine 7 SF 0.00 25.000 TF 0.00 7.000 3.900
## 10/8/2015 Atrazine 7 SF 0.00 17.000 TF 0.00 5.000 2.220
## 10/8/2015 Atrazine 3 SF 0.00 20.000 TF 0.00 3.750 2.220
## 10/8/2015 Atrazine 4 SF 0.00 25.000 TF 0.00 4.000 2.220
## 11/8/2015 Atrazine 7 SF 0.00 5.000 TF 0.00 1.000 1.040
## 11/8/2015 Atrazine 4 SF 0.00 15.000 TF 0.00 2.000 1.040
## 11/8/2015 Atrazine 3 SF 0.00 NA TF 0.00 2.000 1.040
## 12/17/2015 Atrazine 4 SF 0.00 25.000 TF 0.00 12.000 6.760
## 12/17/2015 Atrazine 7 SF 0.00 25.000 TF 0.00 9.000 6.760
## 12/17/2015 Atrazine 3 SF 0.00 25.000 TF 0.00 12.000 6.760
## 3/18/2015 Atrazine 3 SF 0.00 0.132 TF 0.00 0.635 0.150
## 3/18/2015 Atrazine 4 SF 0.00 1.000 TF 0.00 0.440 0.150
## 3/18/2015 Atrazine 7 SF 0.00 0.230 TF 0.00 0.550 0.150
## 4/16/2015 Atrazine 7 SF 0.00 12.500 TF 0.72 2.000 2.485
## 4/16/2015 Atrazine 3 SF 0.00 0.670 TF 0.00 3.000 2.485
## 4/16/2015 Atrazine 4 SF 0.00 13.000 TF 0.00 2.000 2.485
## 5/1/2015 Atrazine 7 SF 0.68 20.000 TF 1.55 5.000 1.965
## 5/1/2015 Atrazine 4 SF 2.44 5.000 TF 2.74 4.500 1.965
## 5/1/2015 Atrazine 3 SF 2.70 16.000 TF 7.09 1.000 1.965
## 5/22/2015 Atrazine 7 SF 0.00 NA TF 0.00 NA 0.650
## 5/22/2015 Atrazine 4 SF 3.65 25.000 TF 1.17 6.000 0.650
## 5/22/2015 Atrazine 3 SF 0.00 NA TF 0.00 NA 0.650
## 6/10/2015 Atrazine 4 SF 0.66 25.000 TF 0.52 4.000 2.580
## 6/10/2015 Atrazine 3 SF 2.69 15.000 TF 2.11 4.000 2.580
## 6/10/2015 Atrazine 7 SF 0.00 20.000 TF 0.37 5.000 2.580
## 6/23/2015 Atrazine 3 SF 1.01 NA TF 0.70 NA 1.810
## 6/23/2015 Atrazine 7 SF 0.00 NA TF 0.30 NA 1.810
## 6/23/2015 Atrazine 4 SF 0.00 NA TF 0.17 NA 1.810
## 7/10/2015 Atrazine 7 SF 0.00 7.000 TF 0.00 2.000 1.370
## 7/10/2015 Atrazine 4 SF 0.00 25.000 TF 0.00 4.000 1.370
## 7/10/2015 Atrazine 3 SF 0.87 8.500 TF 0.00 3.000 1.370
## 7/17/2015 Atrazine 7 SF 0.00 NA TF 0.23 NA 1.730
## 7/17/2015 Atrazine 4 SF 0.00 NA TF 0.00 NA 1.730
## 7/17/2015 Atrazine 3 SF 0.70 NA TF 0.24 NA 1.730
## 7/2/2015 Atrazine 4 SF 0.00 0.675 TF 0.16 4.000 1.010
## 7/2/2015 Atrazine 7 SF 0.00 3.000 TF 0.00 1.500 1.010
## 7/2/2015 Atrazine 3 SF 0.00 NA TF 0.25 1.000 1.010
## 7/30/2015 Atrazine 3 SF 0.00 20.000 TF 0.00 4.000 2.970
## 7/30/2015 Atrazine 4 SF 0.00 25.000 TF 0.00 6.000 2.970
## 7/30/2015 Atrazine 7 SF 0.00 20.000 TF 0.00 3.500 2.970
## 8/13/2015 Atrazine 7 SF 0.00 15.000 TF 0.00 4.000 3.090
## 8/13/2015 Atrazine 3 SF 0.42 13.000 TF 0.00 4.000 3.090
## 8/13/2015 Atrazine 4 SF 0.00 20.000 TF 0.00 6.000 3.090
## 8/27/2015 Atrazine 7 SF 0.00 25.000 TF 0.52 4.000 2.380
## 8/27/2015 Atrazine 4 SF 0.00 25.000 TF 0.00 4.750 2.380
## 8/27/2015 Atrazine 3 SF 0.00 17.000 TF 0.19 4.000 2.380
## 9/17/2015 Atrazine 4 SF 0.00 25.000 TF 0.00 3.000 1.890
## 9/17/2015 Atrazine 7 SF 0.00 14.000 TF 0.00 2.000 1.890
## 9/17/2015 Atrazine 3 SF 0.00 2.500 TF 0.00 3.000 1.890
## 9/3/2015 Atrazine 3 SF 0.00 0.220 TF 0.00 0.690 0.460
## 9/3/2015 Atrazine 4 SF 0.00 0.540 TF 0.00 0.550 0.460
## 9/3/2015 Atrazine 7 SF 0.00 0.647 TF 0.00 0.435 0.460
##
## Wilcoxon signed rank test with continuity correction
##
## data: chem_merge$Conc.x and chem_merge$Conc.y
## V = 81.5, p-value = 0.7134
## alternative hypothesis: true location shift is greater than 0
## [1] "======================================================="
## [1] "Biphenyl"
## [1] "Biphenyl stemflow detection frequency = 13 / 54"
## [1] "Biphenyl stemflow concentration range =( 0 , 1.34 )"
## [1] "Biphenyl throughfall detection frequency = 44 / 54"
## [1] "Biphenyl throughfall concentration range =( 0 , 0.07 )"
## [1] "(throughfall_greater, ties, stemflow_greater) = ( 35 , 10 , 9 )"
##
##
## Date Compound Site Type.x Conc.x Volume.x Type.y Conc.y Volume.y amount_inches
## ----------- --------- ----- ------- ------- --------- ------- ---------- --------- --------------
## 1/17/2016 Biphenyl 3 SF 0.00 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 Biphenyl 4 SF 0.00 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 Biphenyl 7 SF 0.00 25.000 TF 0.0200000 7.000 3.900
## 10/8/2015 Biphenyl 7 SF 0.00 17.000 TF 0.0200000 5.000 2.220
## 10/8/2015 Biphenyl 3 SF 0.00 20.000 TF 0.0300000 3.750 2.220
## 10/8/2015 Biphenyl 4 SF 0.00 25.000 TF 0.0200000 4.000 2.220
## 11/8/2015 Biphenyl 7 SF 1.34 5.000 TF 0.0000000 1.000 1.040
## 11/8/2015 Biphenyl 4 SF 1.12 15.000 TF 0.0500000 2.000 1.040
## 11/8/2015 Biphenyl 3 SF 0.00 NA TF 0.0700000 2.000 1.040
## 12/17/2015 Biphenyl 4 SF 0.00 25.000 TF 0.0200000 12.000 6.760
## 12/17/2015 Biphenyl 7 SF 0.28 25.000 TF 0.0300000 9.000 6.760
## 12/17/2015 Biphenyl 3 SF 1.16 25.000 TF 0.0300000 12.000 6.760
## 3/18/2015 Biphenyl 3 SF 0.00 0.132 TF 0.0472441 0.635 0.150
## 3/18/2015 Biphenyl 4 SF 0.00 1.000 TF 0.0681818 0.440 0.150
## 3/18/2015 Biphenyl 7 SF 0.00 0.230 TF 0.0300000 0.550 0.150
## 4/16/2015 Biphenyl 7 SF 0.00 12.500 TF 0.0300000 2.000 2.485
## 4/16/2015 Biphenyl 3 SF 0.00 0.670 TF 0.0600000 3.000 2.485
## 4/16/2015 Biphenyl 4 SF 0.25 13.000 TF 0.0400000 2.000 2.485
## 5/1/2015 Biphenyl 7 SF 0.00 20.000 TF 0.0200000 5.000 1.965
## 5/1/2015 Biphenyl 4 SF 0.00 5.000 TF 0.0300000 4.500 1.965
## 5/1/2015 Biphenyl 3 SF 0.00 16.000 TF 0.0300000 1.000 1.965
## 5/22/2015 Biphenyl 7 SF 0.00 NA TF 0.0000000 NA 0.650
## 5/22/2015 Biphenyl 4 SF 0.00 25.000 TF 0.0300000 6.000 0.650
## 5/22/2015 Biphenyl 3 SF 0.00 NA TF 0.0000000 NA 0.650
## 6/10/2015 Biphenyl 4 SF 0.00 25.000 TF 0.0200000 4.000 2.580
## 6/10/2015 Biphenyl 3 SF 0.10 15.000 TF 0.0300000 4.000 2.580
## 6/10/2015 Biphenyl 7 SF 0.00 20.000 TF 0.0000000 5.000 2.580
## 6/23/2015 Biphenyl 3 SF 0.00 NA TF 0.0300000 NA 1.810
## 6/23/2015 Biphenyl 7 SF 0.00 NA TF 0.0200000 NA 1.810
## 6/23/2015 Biphenyl 4 SF 0.00 NA TF 0.0200000 NA 1.810
## 7/10/2015 Biphenyl 7 SF 0.00 7.000 TF 0.0200000 2.000 1.370
## 7/10/2015 Biphenyl 4 SF 0.00 25.000 TF 0.0000000 4.000 1.370
## 7/10/2015 Biphenyl 3 SF 0.01 8.500 TF 0.0200000 3.000 1.370
## 7/17/2015 Biphenyl 7 SF 1.04 NA TF 0.0300000 NA 1.730
## 7/17/2015 Biphenyl 4 SF 0.00 NA TF 0.0000000 NA 1.730
## 7/17/2015 Biphenyl 3 SF 0.00 NA TF 0.0200000 NA 1.730
## 7/2/2015 Biphenyl 4 SF 0.00 0.675 TF 0.0200000 4.000 1.010
## 7/2/2015 Biphenyl 7 SF 0.00 3.000 TF 0.0300000 1.500 1.010
## 7/2/2015 Biphenyl 3 SF 0.00 NA TF 0.0200000 1.000 1.010
## 7/30/2015 Biphenyl 3 SF 0.68 20.000 TF 0.0200000 4.000 2.970
## 7/30/2015 Biphenyl 4 SF 0.00 25.000 TF 0.0200000 6.000 2.970
## 7/30/2015 Biphenyl 7 SF 0.00 20.000 TF 0.0000000 3.500 2.970
## 8/13/2015 Biphenyl 7 SF 0.00 15.000 TF 0.0200000 4.000 3.090
## 8/13/2015 Biphenyl 3 SF 0.01 13.000 TF 0.0200000 4.000 3.090
## 8/13/2015 Biphenyl 4 SF 0.00 20.000 TF 0.0200000 6.000 3.090
## 8/27/2015 Biphenyl 7 SF 0.00 25.000 TF 0.0200000 4.000 2.380
## 8/27/2015 Biphenyl 4 SF 0.02 25.000 TF 0.0200000 4.750 2.380
## 8/27/2015 Biphenyl 3 SF 0.00 17.000 TF 0.0300000 4.000 2.380
## 9/17/2015 Biphenyl 4 SF 0.02 25.000 TF 0.0300000 3.000 1.890
## 9/17/2015 Biphenyl 7 SF 0.00 14.000 TF 0.0300000 2.000 1.890
## 9/17/2015 Biphenyl 3 SF 0.49 2.500 TF 0.0300000 3.000 1.890
## 9/3/2015 Biphenyl 3 SF 0.00 0.220 TF 0.0000000 0.690 0.460
## 9/3/2015 Biphenyl 4 SF 0.00 0.540 TF 0.0500000 0.550 0.460
## 9/3/2015 Biphenyl 7 SF 0.00 0.647 TF 0.0459770 0.435 0.460
##
## Wilcoxon signed rank test with continuity correction
##
## data: chem_merge$Conc.x and chem_merge$Conc.y
## V = 359.5, p-value = 0.9451
## alternative hypothesis: true location shift is greater than 0
## [1] "======================================================="
## [1] "DEA"
## [1] "DEA stemflow detection frequency = 11 / 54"
## [1] "DEA stemflow concentration range =( 0 , 9.54 )"
## [1] "DEA throughfall detection frequency = 3 / 54"
## [1] "DEA throughfall concentration range =( 0 , 1.55 )"
## [1] "(throughfall_greater, ties, stemflow_greater) = ( 2 , 43 , 9 )"
##
##
## Date Compound Site Type.x Conc.x Volume.x Type.y Conc.y Volume.y amount_inches
## ----------- --------- ----- ------- ------- --------- ------- ------- --------- --------------
## 1/17/2016 DEA 3 SF 0.00 25.000 TF 0.00 10.000 3.900
## 1/17/2016 DEA 4 SF 0.00 25.000 TF 0.00 10.000 3.900
## 1/17/2016 DEA 7 SF 0.00 25.000 TF 0.00 7.000 3.900
## 10/8/2015 DEA 7 SF 0.00 17.000 TF 0.00 5.000 2.220
## 10/8/2015 DEA 3 SF 0.00 20.000 TF 0.00 3.750 2.220
## 10/8/2015 DEA 4 SF 0.00 25.000 TF 0.00 4.000 2.220
## 11/8/2015 DEA 7 SF 0.00 5.000 TF 0.00 1.000 1.040
## 11/8/2015 DEA 4 SF 0.00 15.000 TF 0.00 2.000 1.040
## 11/8/2015 DEA 3 SF 0.00 NA TF 0.00 2.000 1.040
## 12/17/2015 DEA 4 SF 0.00 25.000 TF 0.00 12.000 6.760
## 12/17/2015 DEA 7 SF 0.00 25.000 TF 0.00 9.000 6.760
## 12/17/2015 DEA 3 SF 0.00 25.000 TF 0.00 12.000 6.760
## 3/18/2015 DEA 3 SF 0.00 0.132 TF 0.00 0.635 0.150
## 3/18/2015 DEA 4 SF 0.00 1.000 TF 0.00 0.440 0.150
## 3/18/2015 DEA 7 SF 0.00 0.230 TF 0.00 0.550 0.150
## 4/16/2015 DEA 7 SF 0.00 12.500 TF 0.00 2.000 2.485
## 4/16/2015 DEA 3 SF 0.00 0.670 TF 0.00 3.000 2.485
## 4/16/2015 DEA 4 SF 0.00 13.000 TF 0.00 2.000 2.485
## 5/1/2015 DEA 7 SF 0.00 20.000 TF 0.00 5.000 1.965
## 5/1/2015 DEA 4 SF 9.54 5.000 TF 0.73 4.500 1.965
## 5/1/2015 DEA 3 SF 0.40 16.000 TF 1.55 1.000 1.965
## 5/22/2015 DEA 7 SF 0.00 NA TF 0.00 NA 0.650
## 5/22/2015 DEA 4 SF 0.72 25.000 TF 0.00 6.000 0.650
## 5/22/2015 DEA 3 SF 0.00 NA TF 0.00 NA 0.650
## 6/10/2015 DEA 4 SF 0.11 25.000 TF 0.00 4.000 2.580
## 6/10/2015 DEA 3 SF 0.50 15.000 TF 0.83 4.000 2.580
## 6/10/2015 DEA 7 SF 0.00 20.000 TF 0.00 5.000 2.580
## 6/23/2015 DEA 3 SF 0.71 NA TF 0.00 NA 1.810
## 6/23/2015 DEA 7 SF 0.00 NA TF 0.00 NA 1.810
## 6/23/2015 DEA 4 SF 0.00 NA TF 0.00 NA 1.810
## 7/10/2015 DEA 7 SF 0.00 7.000 TF 0.00 2.000 1.370
## 7/10/2015 DEA 4 SF 0.00 25.000 TF 0.00 4.000 1.370
## 7/10/2015 DEA 3 SF 0.47 8.500 TF 0.00 3.000 1.370
## 7/17/2015 DEA 7 SF 0.00 NA TF 0.00 NA 1.730
## 7/17/2015 DEA 4 SF 0.00 NA TF 0.00 NA 1.730
## 7/17/2015 DEA 3 SF 0.33 NA TF 0.00 NA 1.730
## 7/2/2015 DEA 4 SF 0.00 0.675 TF 0.00 4.000 1.010
## 7/2/2015 DEA 7 SF 0.00 3.000 TF 0.00 1.500 1.010
## 7/2/2015 DEA 3 SF 0.00 NA TF 0.00 1.000 1.010
## 7/30/2015 DEA 3 SF 0.08 20.000 TF 0.00 4.000 2.970
## 7/30/2015 DEA 4 SF 0.00 25.000 TF 0.00 6.000 2.970
## 7/30/2015 DEA 7 SF 0.00 20.000 TF 0.00 3.500 2.970
## 8/13/2015 DEA 7 SF 0.00 15.000 TF 0.00 4.000 3.090
## 8/13/2015 DEA 3 SF 0.08 13.000 TF 0.00 4.000 3.090
## 8/13/2015 DEA 4 SF 0.00 20.000 TF 0.00 6.000 3.090
## 8/27/2015 DEA 7 SF 0.00 25.000 TF 0.00 4.000 2.380
## 8/27/2015 DEA 4 SF 0.00 25.000 TF 0.00 4.750 2.380
## 8/27/2015 DEA 3 SF 0.02 17.000 TF 0.00 4.000 2.380
## 9/17/2015 DEA 4 SF 0.00 25.000 TF 0.00 3.000 1.890
## 9/17/2015 DEA 7 SF 0.00 14.000 TF 0.00 2.000 1.890
## 9/17/2015 DEA 3 SF 0.00 2.500 TF 0.00 3.000 1.890
## 9/3/2015 DEA 3 SF 0.00 0.220 TF 0.00 0.690 0.460
## 9/3/2015 DEA 4 SF 0.00 0.540 TF 0.00 0.550 0.460
## 9/3/2015 DEA 7 SF 0.00 0.647 TF 0.00 0.435 0.460
##
## Wilcoxon signed rank test with continuity correction
##
## data: chem_merge$Conc.x and chem_merge$Conc.y
## V = 51, p-value = 0.05977
## alternative hypothesis: true location shift is greater than 0
## [1] "======================================================="
## [1] "Flutolanil"
## [1] "Flutolanil stemflow detection frequency = 15 / 54"
## [1] "Flutolanil stemflow concentration range =( 0 , 0.38 )"
## [1] "Flutolanil throughfall detection frequency = 28 / 54"
## [1] "Flutolanil throughfall concentration range =( 0 , 0.63 )"
## [1] "(throughfall_greater, ties, stemflow_greater) = ( 23 , 26 , 5 )"
##
##
## Date Compound Site Type.x Conc.x Volume.x Type.y Conc.y Volume.y amount_inches
## ----------- ----------- ----- ------- ---------- --------- ------- ---------- --------- --------------
## 1/17/2016 Flutolanil 3 SF 0.0000000 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 Flutolanil 4 SF 0.0000000 25.000 TF 0.0000000 10.000 3.900
## 1/17/2016 Flutolanil 7 SF 0.0000000 25.000 TF 0.0700000 7.000 3.900
## 10/8/2015 Flutolanil 7 SF 0.0900000 17.000 TF 0.1000000 5.000 2.220
## 10/8/2015 Flutolanil 3 SF 0.0000000 20.000 TF 0.1300000 3.750 2.220
## 10/8/2015 Flutolanil 4 SF 0.1800000 25.000 TF 0.1400000 4.000 2.220
## 11/8/2015 Flutolanil 7 SF 0.0000000 5.000 TF 0.1700000 1.000 1.040
## 11/8/2015 Flutolanil 4 SF 0.0000000 15.000 TF 0.0800000 2.000 1.040
## 11/8/2015 Flutolanil 3 SF 0.0000000 NA TF 0.0000000 2.000 1.040
## 12/17/2015 Flutolanil 4 SF 0.2200000 25.000 TF 0.0000000 12.000 6.760
## 12/17/2015 Flutolanil 7 SF 0.0000000 25.000 TF 0.0000000 9.000 6.760
## 12/17/2015 Flutolanil 3 SF 0.0000000 25.000 TF 0.0000000 12.000 6.760
## 3/18/2015 Flutolanil 3 SF 0.0000000 0.132 TF 0.0000000 0.635 0.150
## 3/18/2015 Flutolanil 4 SF 0.0000000 1.000 TF 0.0000000 0.440 0.150
## 3/18/2015 Flutolanil 7 SF 0.0000000 0.230 TF 0.0000000 0.550 0.150
## 4/16/2015 Flutolanil 7 SF 0.0000000 12.500 TF 0.0000000 2.000 2.485
## 4/16/2015 Flutolanil 3 SF 0.0000000 0.670 TF 0.0000000 3.000 2.485
## 4/16/2015 Flutolanil 4 SF 0.0000000 13.000 TF 0.0000000 2.000 2.485
## 5/1/2015 Flutolanil 7 SF 0.0000000 20.000 TF 0.0000000 5.000 1.965
## 5/1/2015 Flutolanil 4 SF 0.0000000 5.000 TF 0.0000000 4.500 1.965
## 5/1/2015 Flutolanil 3 SF 0.0000000 16.000 TF 0.0000000 1.000 1.965
## 5/22/2015 Flutolanil 7 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 5/22/2015 Flutolanil 4 SF 0.0000000 25.000 TF 0.0000000 6.000 0.650
## 5/22/2015 Flutolanil 3 SF 0.0000000 NA TF 0.0000000 NA 0.650
## 6/10/2015 Flutolanil 4 SF 0.0000000 25.000 TF 0.0000000 4.000 2.580
## 6/10/2015 Flutolanil 3 SF 0.0000000 15.000 TF 0.0000000 4.000 2.580
## 6/10/2015 Flutolanil 7 SF 0.0900000 20.000 TF 0.0900000 5.000 2.580
## 6/23/2015 Flutolanil 3 SF 0.0000000 NA TF 0.0000000 NA 1.810
## 6/23/2015 Flutolanil 7 SF 0.0000000 NA TF 0.0000000 NA 1.810
## 6/23/2015 Flutolanil 4 SF 0.0000000 NA TF 0.0000000 NA 1.810
## 7/10/2015 Flutolanil 7 SF 0.1200000 7.000 TF 0.1300000 2.000 1.370
## 7/10/2015 Flutolanil 4 SF 0.0000000 25.000 TF 0.1100000 4.000 1.370
## 7/10/2015 Flutolanil 3 SF 0.1500000 8.500 TF 0.6300000 3.000 1.370
## 7/17/2015 Flutolanil 7 SF 0.0000000 NA TF 0.3000000 NA 1.730
## 7/17/2015 Flutolanil 4 SF 0.2300000 NA TF 0.2400000 NA 1.730
## 7/17/2015 Flutolanil 3 SF 0.2100000 NA TF 0.3500000 NA 1.730
## 7/2/2015 Flutolanil 4 SF 0.0000000 0.675 TF 0.0000000 4.000 1.010
## 7/2/2015 Flutolanil 7 SF 0.2400000 3.000 TF 0.1500000 1.500 1.010
## 7/2/2015 Flutolanil 3 SF 0.0000000 NA TF 0.2000000 1.000 1.010
## 7/30/2015 Flutolanil 3 SF 0.0000000 20.000 TF 0.1900000 4.000 2.970
## 7/30/2015 Flutolanil 4 SF 0.2100000 25.000 TF 0.0800000 6.000 2.970
## 7/30/2015 Flutolanil 7 SF 0.0000000 20.000 TF 0.1800000 3.500 2.970
## 8/13/2015 Flutolanil 7 SF 0.1400000 15.000 TF 0.3100000 4.000 3.090
## 8/13/2015 Flutolanil 3 SF 0.0000000 13.000 TF 0.0000000 4.000 3.090
## 8/13/2015 Flutolanil 4 SF 0.0000000 20.000 TF 0.2100000 6.000 3.090
## 8/27/2015 Flutolanil 7 SF 0.0000000 25.000 TF 0.3300000 4.000 2.380
## 8/27/2015 Flutolanil 4 SF 0.1300000 25.000 TF 0.1900000 4.750 2.380
## 8/27/2015 Flutolanil 3 SF 0.3800000 17.000 TF 0.2100000 4.000 2.380
## 9/17/2015 Flutolanil 4 SF 0.3000000 25.000 TF 0.3500000 3.000 1.890
## 9/17/2015 Flutolanil 7 SF 0.0000000 14.000 TF 0.2500000 2.000 1.890
## 9/17/2015 Flutolanil 3 SF 0.0000000 2.500 TF 0.3300000 3.000 1.890
## 9/3/2015 Flutolanil 3 SF 0.0000000 0.220 TF 0.0000000 0.690 0.460
## 9/3/2015 Flutolanil 4 SF 0.0000000 0.540 TF 0.1818182 0.550 0.460
## 9/3/2015 Flutolanil 7 SF 0.0927357 0.647 TF 0.2758621 0.435 0.460
##
## Wilcoxon signed rank test with continuity correction
##
## data: chem_merge$Conc.x and chem_merge$Conc.y
## V = 63, p-value = 0.9993
## alternative hypothesis: true location shift is greater than 0
global analyses
Two values with high concentrations are not shown on the all tifton data figure.
print('=======================================================')## [1] "======================================================="print('global')## [1] "global"tifton_sf <- tifton[which(tifton$Type=='SF'),]
tifton_tf <- tifton[which(tifton$Type=='TF'),]
#merge on site, date, compound
tifton_merge <- merge(tifton_sf,tifton_tf,by=c("Compound","Site","Date"))
#change NAs to zeroes for comparisons and the Wilcox test
tifton_merge$Conc.x[is.na(tifton_merge$Conc.x)] <- 0
tifton_merge$Conc.y[is.na(tifton_merge$Conc.y)] <- 0
n_tf_greater <- sum(tifton_merge$Conc.y > tifton_merge$Conc.x)
n_ties <- sum(tifton_merge$Conc.y == tifton_merge$Conc.x)
n_sf_greater <- sum(tifton_merge$Conc.x > tifton_merge$Conc.y)
comparisons <- paste('(throughfall_greater, ties, stemflow_greater) = ('
,n_tf_greater,',',n_ties,',',n_sf_greater,')')
print(comparisons)## [1] "(throughfall_greater, ties, stemflow_greater) = ( 172 , 1466 , 90 )"tifton_wilcox <- wilcox.test(tifton_merge$Conc.y, tifton_merge$Conc.x,
alternative = 'two.sided', paired = TRUE, exact = FALSE)
x_max <- max(tifton_merge$Conc.x)
y_max <- max(tifton_merge$Conc.y)
max_conc <- max(x_max,y_max)
max_conc_vector <- seq(0,max_conc,20)
print(tifton_wilcox)##
## Wilcoxon signed rank test with continuity correction
##
## data: tifton_merge$Conc.y and tifton_merge$Conc.x
## V = 21174, p-value = 0.001304
## alternative hypothesis: true location shift is not equal to 0plot(tifton_merge$Conc.x, tifton_merge$Conc.y,
xlim=c(0,4),ylim=c(0,4),
xlab='Stemflow Concentration',ylab='Throughfall Concentration',
main=paste('All Tifton Data (WRS p-value = ',round(tifton_wilcox$p.value,4),')'),
sub = comparisons)
abline(0,1,col='red')
date analyses
#break down by day
#rainwater
View(rainwater)
#tifton[which(tifton$Compound==chem),]
tifton_sf <- tifton[which(tifton$Type=='SF'),]
dim(tifton_sf)## [1] 1728 6tifton_tf <- tifton[which(tifton$Type=='TF'),]
dim(tifton_tf)## [1] 1728 6tifton_tp <- tifton[which(tifton$Type=='TP'),]
#merge on site, date, compound
tifton_merge <- merge(tifton_sf,tifton_tf,by=c("Compound","Site","Date"))
dim(tifton_merge)## [1] 1728 9tifton_merge <- merge(tifton_merge, rainwater, by="Date")
dim(tifton_merge)## [1] 1728 10unique_dates <- unique(tifton_merge$Date)
for(date in unique_dates){
tifton_merge_date <- tifton_merge[which(tifton$Date==date),]
#change NAs to zeroes for comparisons and the Wilcox test
tifton_merge_date$Conc.x[is.na(tifton_merge_date$Conc.x)] <- 0
tifton_merge_date$Conc.y[is.na(tifton_merge_date$Conc.y)] <- 0
print(date)
#print(tifton_merge_date)
n_tf_greater <- sum(tifton_merge_date$Conc.y > tifton_merge_date$Conc.x)
n_ties <- sum(tifton_merge_date$Conc.y == tifton_merge_date$Conc.x)
n_sf_greater <- sum(tifton_merge_date$Conc.x > tifton_merge_date$Conc.y)
comparisons <- paste(date,'(throughfall_greater, ties, stemflow_greater) = ('
,n_tf_greater,',',n_ties,',',n_sf_greater,')')
print(comparisons)
}## [1] "1/17/2016"
## [1] "1/17/2016 (throughfall_greater, ties, stemflow_greater) = ( 16 , 487 , 9 )"
## [1] "10/8/2015"
## [1] "10/8/2015 (throughfall_greater, ties, stemflow_greater) = ( 16 , 488 , 8 )"
## [1] "11/8/2015"
## [1] "11/8/2015 (throughfall_greater, ties, stemflow_greater) = ( 6 , 498 , 8 )"
## [1] "12/17/2015"
## [1] "12/17/2015 (throughfall_greater, ties, stemflow_greater) = ( 7 , 496 , 9 )"
## [1] "3/18/2015"
## [1] "3/18/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "4/16/2015"
## [1] "4/16/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "5/1/2015"
## [1] "5/1/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "5/22/2015"
## [1] "5/22/2015 (throughfall_greater, ties, stemflow_greater) = ( 14 , 493 , 5 )"
## [1] "6/10/2015"
## [1] "6/10/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "6/23/2015"
## [1] "6/23/2015 (throughfall_greater, ties, stemflow_greater) = ( 15 , 489 , 8 )"
## [1] "7/10/2015"
## [1] "7/10/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "7/17/2015"
## [1] "7/17/2015 (throughfall_greater, ties, stemflow_greater) = ( 20 , 481 , 11 )"
## [1] "7/2/2015"
## [1] "7/2/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "7/30/2015"
## [1] "7/30/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "8/13/2015"
## [1] "8/13/2015 (throughfall_greater, ties, stemflow_greater) = ( 8 , 501 , 3 )"
## [1] "8/27/2015"
## [1] "8/27/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"
## [1] "9/17/2015"
## [1] "9/17/2015 (throughfall_greater, ties, stemflow_greater) = ( 17 , 488 , 7 )"
## [1] "9/3/2015"
## [1] "9/3/2015 (throughfall_greater, ties, stemflow_greater) = ( 0 , 192 , 0 )"regressing versus rainfall amounts
We want to look at how the amount of rainfall influences the concentrations in stemflow and throughfall, in
print('=======================================================')## [1] "======================================================="print('global')## [1] "global"permutation stat testing
For many of the chemicals, some/most observations have identical/not detected values, hence the warning message thrown by R: unique ranks cannot be computed for all observations in the Wilcox test. Since there are ties, this prevents the computation of an exact p-value. With high numbers of tied data, Wilcoxon tests are not trustworthy. Moreover, in most non-parametric tests we assume that the sampled populations are symmetric and have the same dispersion or shape, which is not verifiable here.
A permutation test may be more appropriate in this case, for example permTS (perm), pperm (exactRankTests), or the coin package. Although these approaches do not take advantage of the paired design.
results
Mixed bag at the moment. Overall inference on paired data shows that throughfall concentrations exceed stemflow concentration 170 times versus 92 of the opposite exceedance. This corresponds to a p-value of 0.0019 with the Wilcoxon Rank Sum test.
There will be complaints from reviewers about the unequal volumes of samples for the two media, we need to be clear that the collection time is the same for these paired comparisons.
I guess we can also break out how often the exceedances are a detect versus a non-detect as opposed to a detect versus detect comparison. Can we assume that the actual detection limit is lower for the higher volume stemflow samples? (meaning that the volumes may bias the analyses towards stemflow exceedances and not throughfall).
Stemflow concentrations probably plagued by dilution versus first flush issues. We may want to think about how to use sample volume or observed rainfall totals for these dates to look at this and separate small from large rain events.
Having detection limits will help us characterize the magnitude of the difference between stemflow and throughfall (for example, ratio calculations that are not possible by treating non-detects as zeroes).