RS Wk3 Final Project Assignment
Rick S
7/31/2020
R Week 3 Assignment:
Background: The codling moth (Cydia pomonella) is a pest which invades apples and many other fruit crops. This pest has not yet been detected in Japan. New Zealand grows and exports many apple cultivars. Methyl bromide is used to fumigate many crops necessary to exterminate the codling moth prior to export to Japan.
In 1988–1989 ‘Braeburn’, ‘Fuji’, ‘Granny Smith’, ‘Red Delicious’, ‘Royal Gala’, ‘Gala’ and ‘Splendour’ apples were infested with freshly laid eggs of codling moth, Cydia pomonella (L), and were fumigated at a range of methyl bromide doses.
In a World Trade Organization report on “Japan - Measures Affecting Agricultural Products (WTO, 1998, p. 30)” questions if pest control treatment varies by variety (cultivar).
Meaningful question for analysis: For a given treatment dose of methyl bromide, is there a difference between Cultivars on the level of high mortality of the codling moth in a sample of apple cultivars?
Sample Dataset:
Fumigation experiments with methyl bromide (MeBr) were carried out in New Zealand over several seasons. In 1988–1989 ‘Braeburn’, ‘Fuji’, ‘Granny Smith’, ‘Red Delicious’, ‘Royal Gala’, ‘Gala’ and ‘Splendour’ apples were infested with freshly laid eggs of codling moth, Cydia pomonella (L), and were fumigated at a range of methyl bromide doses.
Data are from trials that studied the mortality response of codling moth to fumigation with methyl bromide.
The research that generated these data was in part funded by New Zealand pipfruit growers.
# Read csv file
# BONUS: Place the original .csv in a github file and have R read from the link. This will be a very useful skill as you progress in your data science and career.
theURL <- "https://raw.githubusercontent.com/CUNYSPS-RickRN/CUNYSPS-Bridge/master/codling.csv"
Apple_MB_dose_df <- read.table(theURL, header=TRUE, sep = ",")Data Structure Exploration:
dim(Apple_MB_dose_df)## [1] 99 11head(Apple_MB_dose_df)## X dose tot dead pobs cm ct Cultivar gp year numcm
## 1 1 5 866 246 0.2841 0.2178 15.59417 ROYAL 1 1988 1676
## 2 2 8 911 220 0.2415 0.2178 20.26042 ROYAL 1 1988 1676
## 3 3 12 906 360 0.3974 0.2178 28.60292 ROYAL 1 1988 1676
## 4 4 16 712 271 0.3806 0.2178 32.68833 ROYAL 1 1988 1676
## 5 5 20 582 414 0.7113 0.2178 45.42708 ROYAL 1 1988 1676
## 6 6 24 1183 742 0.6272 0.2178 45.44292 ROYAL 1 1988 1676tail(Apple_MB_dose_df)## X dose tot dead pobs cm ct Cultivar gp year numcm
## 94 3529 30 2097 1968 0.9385 0.1879 62.51875 Splendour 16 1989 1474
## 95 3831 12 1430 531 0.3713 0.2481 34.92292 Splendour 17 1989 2648
## 96 3932 16 558 273 0.4892 0.2481 42.51292 Splendour 17 1989 2648
## 97 4033 23 1094 911 0.8327 0.2481 54.66375 Splendour 17 1989 2648
## 98 4134 24 1156 937 0.8106 0.2481 53.87000 Splendour 17 1989 2648
## 99 4235 30 795 788 0.9912 0.2481 64.13375 Splendour 17 1989 2648str(Apple_MB_dose_df) # examine data.frame structure## 'data.frame': 99 obs. of 11 variables:
## $ X : int 1 2 3 4 5 6 7 8 9 10 ...
## $ dose : int 5 8 12 16 20 24 5 8 12 16 ...
## $ tot : int 866 911 906 712 582 1183 603 640 627 788 ...
## $ dead : int 246 220 360 271 414 742 154 168 180 240 ...
## $ pobs : num 0.284 0.241 0.397 0.381 0.711 ...
## $ cm : num 0.218 0.218 0.218 0.218 0.218 ...
## $ ct : num 15.6 20.3 28.6 32.7 45.4 ...
## $ Cultivar: chr "ROYAL" "ROYAL" "ROYAL" "ROYAL" ...
## $ gp : int 1 1 1 1 1 1 2 2 2 2 ...
## $ year : int 1988 1988 1988 1988 1988 1988 1988 1988 1988 1988 ...
## $ numcm : int 1676 1676 1676 1676 1676 1676 1597 1597 1597 1597 ...class(Apple_MB_dose_df)## [1] "data.frame"#Rename some columns
names(Apple_MB_dose_df)## [1] "X" "dose" "tot" "dead" "pobs" "cm"
## [7] "ct" "Cultivar" "gp" "year" "numcm"colnames(Apple_MB_dose_df) <- c("Observation", "Dose_gm","T_Inchamber","Dead",
"PCT_dying","ControlMortality","ConcentrationTime",
"Cultivar","gp_factor","year_factor","T_ctrl_insects")
names(Apple_MB_dose_df) # see renamed columns## [1] "Observation" "Dose_gm" "T_Inchamber"
## [4] "Dead" "PCT_dying" "ControlMortality"
## [7] "ConcentrationTime" "Cultivar" "gp_factor"
## [10] "year_factor" "T_ctrl_insects"Data Exploration:
cat ("Number of Observations: ", as.character(count(Apple_MB_dose_df)))## Number of Observations: 99summary(Apple_MB_dose_df)## Observation Dose_gm T_Inchamber Dead
## Min. : 1.0 Min. : 5.00 Min. : 239.0 Min. : 96
## 1st Qu.: 25.5 1st Qu.:12.00 1st Qu.: 451.0 1st Qu.: 201
## Median : 50.0 Median :16.00 Median : 603.0 Median : 330
## Mean : 721.3 Mean :18.13 Mean : 734.8 Mean : 426
## 3rd Qu.: 599.5 3rd Qu.:24.00 3rd Qu.: 881.5 3rd Qu.: 533
## Max. :4235.0 Max. :30.00 Max. :2965.0 Max. :1968
## PCT_dying ControlMortality ConcentrationTime Cultivar
## Min. :0.2232 Min. :0.1506 Min. :15.59 Length:99
## 1st Qu.:0.3503 1st Qu.:0.1978 1st Qu.:32.85 Class :character
## Median :0.5461 Median :0.2160 Median :42.51 Mode :character
## Mean :0.5853 Mean :0.2178 Mean :43.30
## 3rd Qu.:0.8279 3rd Qu.:0.2368 3rd Qu.:53.70
## Max. :0.9973 Max. :0.2765 Max. :65.00
## gp_factor year_factor T_ctrl_insects
## Min. : 1.000 Min. :1988 Min. :1067
## 1st Qu.: 5.000 1st Qu.:1988 1st Qu.:1474
## Median : 9.000 Median :1988 Median :1773
## Mean : 8.919 Mean :1988 Mean :2066
## 3rd Qu.:13.000 3rd Qu.:1989 3rd Qu.:2333
## Max. :17.000 Max. :1989 Max. :4177Graphing Histograms:
These set of histograms will graph the frequency of “Percent of Dying”
ggplot(data=Apple_MB_dose_df) + aes(x=PCT_dying, fill=Cultivar) + geom_histogram( bins=10) + labs(x="Percent Dying")
ggplot(data=Apple_MB_dose_df) + aes(x=PCT_dying, fill=Cultivar) + geom_histogram( bins=10) + labs(x="Percent Dying") +
facet_wrap(~Cultivar)
hist(Apple_MB_dose_df$PCT_dying, main="Percent Dying Histogram", xlab="PCT_dying")
ggplot(data=Apple_MB_dose_df) + geom_histogram(aes(x=PCT_dying), fill="grey50",bins=10)
Graphing: Scatterplot featuring Cultivar
These set of graphs represent for each individual apple Cultivar the “Percent of Control Mortality” in relation to “Dose”.
Most all culivars achieved near complete mortality except for the Royal cultivar.
#Scatter plot by Cultivar ControlMortality
g <- ggplot(Apple_MB_dose_df, aes(x=ControlMortality,y=Dose_gm)) +
geom_point()
g + geom_point(aes(color=Cultivar)) + facet_wrap(~Cultivar)
g # display graph
#Scatter plot by Cultivar
g <- ggplot(Apple_MB_dose_df, aes(x=PCT_dying,y=Dose_gm)) +
geom_point()
g + geom_point(aes(color=Cultivar)) + facet_wrap(~Cultivar)
g # display graph
Graphing: Boxplot by Cultivar
This boxplot shows most all cultivars will have an effective mortality rate at a dose of 24 g m-3 while the cultivar Royal can achieve an effective mortality rate at a lower dose.
ggplot(Apple_MB_dose_df, aes(x=Cultivar, y=Dose_gm)) + geom_boxplot()
Graphing: Boxplot by Dose
This boxplot shows most an effective mortality rate at a dose of 24 g m-3.
ggplot(Apple_MB_dose_df, aes(x = Dose_gm,y = PCT_dying, group = Dose_gm)) + geom_boxplot()
Graphing: Simple Linear Regression
# Simple Linear Regression
ggplot(Apple_MB_dose_df, aes(x=Dose_gm, y=PCT_dying, fill=Cultivar)) + geom_point() + geom_smooth(method="lm") + labs(x="Dose", y="PCT Dying")## `geom_smooth()` using formula 'y ~ x'
Closer inspection of data at 24gm dose.
Graphing: Scatterplot by Cultivar.
First scatterplot exhibits the control mortality percent. Second scatterplot exhibits the experimental mortality percent at 24 gm dose. All cultivars except Royal exhibited higher mortality rates at the 24 gm dose.
# Closer inspection of data at 24 gm dose
Dose_24gm <- Apple_MB_dose_df[Apple_MB_dose_df$Dose_gm == 24, ]
Dose_24gm## Observation Dose_gm T_Inchamber Dead PCT_dying ControlMortality
## 6 6 24 1183 742 0.6272 0.2178
## 12 12 24 1137 628 0.5523 0.2160
## 16 17 24 472 417 0.8835 0.2641
## 22 23 24 462 423 0.9156 0.2765
## 28 29 24 524 402 0.7672 0.2126
## 34 35 24 522 345 0.6609 0.1997
## 40 41 24 743 584 0.7860 0.2141
## 46 47 24 743 569 0.7658 0.2368
## 52 53 24 405 381 0.9407 0.1779
## 58 59 24 429 413 0.9627 0.1506
## 64 65 24 446 412 0.9238 0.1978
## 70 655 24 278 274 0.9856 0.2353
## 76 1311 24 650 535 0.8231 0.2523
## 82 2016 24 601 421 0.7005 0.1946
## 87 2722 24 1019 873 0.8567 0.2357
## 93 3428 24 2510 1729 0.6888 0.1879
## 98 4134 24 1156 937 0.8106 0.2481
## ConcentrationTime Cultivar gp_factor year_factor T_ctrl_insects
## 6 45.44292 ROYAL 1 1988 1676
## 12 53.71208 ROYAL 2 1988 1597
## 16 53.24750 BRAEBURN 3 1988 1662
## 22 54.53625 BRAEBURN 4 1988 2123
## 28 51.48250 FUJI 5 1988 1392
## 34 53.94625 FUJI 6 1988 1773
## 40 54.77875 GRANNY 7 1988 2284
## 46 53.51708 GRANNY 8 1988 3277
## 52 51.77208 Red Delicious 9 1988 1147
## 58 54.03728 Red Delicious 10 1988 1122
## 64 53.76875 Red Delicious 11 1988 1067
## 70 53.31667 Gala 12 1989 2214
## 76 53.68266 Gala 13 1989 3226
## 82 54.32658 Red Delicious 14 1989 4177
## 87 56.20417 Red Delicious 15 1989 2333
## 93 51.92708 Splendour 16 1989 1474
## 98 53.87000 Splendour 17 1989 2648#Scatter plot by Cultivar of Control Mortality
ggplot(Dose_24gm, aes(x=Cultivar, y=ControlMortality)) +
geom_point(aes(color=Cultivar))
#Scatter plot by Cultivar of PCT_dying
ggplot(Dose_24gm, aes(x=Cultivar, y=PCT_dying)) +
geom_point(aes(color=Cultivar))
Conclusion:
In this controlled experiment, fumigation using methyl bromide for pest control of the codling moth, Cydia pomonella (L), there is a difference in the level of effective mortality rate among various cultivars of apples. The treatment level of the Royal cultivar is different than the other cultivars. Therefore, product by product testing is necessary for pest control of the codling moth for apple export to Japan (WTO, 1998, p. 30). Each cultivar should require individual export approval to Japan based on data.
References:
Maindonald, J. H., Waddell, B. C., & Petry, R. J. (2001). Apple cultivar effects on codling moth (Lepidoptera: Tortricidae) egg mortality following fumigation with methyl bromide. Postharvest Biology and Technology, 22(2), 99-110. Retrieved from https://doi.org/10.1016/S0925-5214(01)00082-5
World Trade Organization (WTO). (1998). Japan – Measures Affecting Agricultural Products - Report of the Panel. Retrieved from http://www.worldtradelaw.net/reports/wtopanelsfull/japan-agproducts(panel)(full).pdf.download