Site Preparation Experiment on LodgePole Pine - Experiment (VII) (2)
This experiment is a soil scarification experiment with Lodegepole Pine in Västerbotten, Sweden. The treatments were either planting after soil scarification using deep plowing (deep), or planting without any site preparation (control). The size of the treatment plots were 30*30 m and the experiment was replicated in 4 blocks. The plantation was made in 1988 and the experiment was remeasured after several years, to 2012, when all trees in the treatment plots were cross calipered at dbh (mm).
Questions
For this data, I’ll be exploring the data and then calculating the basal_area/ha and stem_density/ha for every block and treatment. I seek to answer the following:
What effect have a radical soil scarification method such as deep plowing on:
survival and
growth of lodgepole pine on the site
In other words: Is there a significant effect of treatment if we test for:
stem density and
basal area.
# loading libraries
library(doBy)
library(dplyr)##
## Attaching package: 'dplyr'## The following object is masked from 'package:doBy':
##
## order_by## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(lattice)
library(ggplot2)
library(car)## Loading required package: carData##
## Attaching package: 'car'## The following object is masked from 'package:dplyr':
##
## recodelibrary(data.table)##
## Attaching package: 'data.table'## The following objects are masked from 'package:dplyr':
##
## between, first, lastlibrary(TukeyC)Importing the data
L_pine <- read.table('https://raw.githubusercontent.com/xrander/Slu_experiment/master/Data/Lab%207/deepcult_lodgepolepine.txt',
header = T,
sep = '\t',
na.strings = 'NA',
dec = '.',
strip.white = T)
head(L_pine)## plot block treenr treatment dbh1 dbh2 height dead
## 1 1 1 1 Control 110 112 NA NA
## 2 1 1 2 Control 101 102 NA NA
## 3 1 1 3 Control 109 112 NA NA
## 4 1 1 4 Control 81 85 NA NA
## 5 1 1 5 Control 115 117 NA NA
## 6 1 1 6 Control 100 106 NA NALittle Data Exploration
str(L_pine)## 'data.frame': 1799 obs. of 8 variables:
## $ plot : int 1 1 1 1 1 1 1 1 1 1 ...
## $ block : int 1 1 1 1 1 1 1 1 1 1 ...
## $ treenr : int 1 2 3 4 5 6 7 8 9 10 ...
## $ treatment: chr "Control" "Control" "Control" "Control" ...
## $ dbh1 : int 110 101 109 81 115 100 95 98 107 79 ...
## $ dbh2 : int 112 102 112 85 117 106 91 96 107 83 ...
## $ height : int NA NA NA NA NA NA NA NA NA NA ...
## $ dead : int NA NA NA NA NA NA NA NA NA NA ...I will change the data types of ‘plot’, ‘block’, and ‘treatment’ into factor data type
plot(L_pine$dbh1, L_pine$dbh2,
col = 'purple',
xlab = 'dbh1',
ylab = 'dbh2',
main = 'DBH vs DBH')
L_pine$plot <- as.factor(L_pine$plot)
L_pine$block <- as.factor(L_pine$block)
L_pine$treatment <- as.factor(L_pine$treatment)Survival
with(L_pine, table(treatment, dead))## dead
## treatment 1
## Control 40
## Deep 24subsetting only living trees
L_pine$dead <- ifelse(is.na(L_pine$dead) == TRUE,0 , L_pine$dead)L_pine1 <- subset(L_pine, dead== 0) # this returns values for only living trees, 1 means dead and 0 means alivesummary(L_pine1)## plot block treenr treatment dbh1
## 8 :222 1:434 Min. : 1.000 Control:859 Min. : 27.00
## 2 :219 2:435 1st Qu.: 4.000 Deep :876 1st Qu.: 88.00
## 3 :218 3:432 Median : 8.000 Median :101.00
## 6 :218 4:434 Mean : 8.032 Mean : 99.33
## 4 :217 3rd Qu.:12.000 3rd Qu.:112.00
## 7 :216 Max. :15.000 Max. :153.00
## (Other):425
## dbh2 height dead
## Min. : 27.00 Min. : 72.0 Min. :0
## 1st Qu.: 88.00 1st Qu.: 94.0 1st Qu.:0
## Median :101.00 Median :102.0 Median :0
## Mean : 99.82 Mean :101.6 Mean :0
## 3rd Qu.:112.00 3rd Qu.:111.0 3rd Qu.:0
## Max. :157.00 Max. :133.0 Max. :0
## NA's :1606bwplot(dbh1~plot | treatment, data = L_pine1,
main = 'Dbh across plots for Treatments',
ylab = 'dbh(cm)',
xlab = 'plots')
getting the mean dbh
L_pine1$dbh <- (L_pine1$dbh1 + L_pine1$dbh2)/2000Evaluating basal area
L_pine1$ba <- (L_pine1$dbh/2)^2 * pisummary(L_pine1)## plot block treenr treatment dbh1
## 8 :222 1:434 Min. : 1.000 Control:859 Min. : 27.00
## 2 :219 2:435 1st Qu.: 4.000 Deep :876 1st Qu.: 88.00
## 3 :218 3:432 Median : 8.000 Median :101.00
## 6 :218 4:434 Mean : 8.032 Mean : 99.33
## 4 :217 3rd Qu.:12.000 3rd Qu.:112.00
## 7 :216 Max. :15.000 Max. :153.00
## (Other):425
## dbh2 height dead dbh
## Min. : 27.00 Min. : 72.0 Min. :0 Min. :0.02700
## 1st Qu.: 88.00 1st Qu.: 94.0 1st Qu.:0 1st Qu.:0.08800
## Median :101.00 Median :102.0 Median :0 Median :0.10100
## Mean : 99.82 Mean :101.6 Mean :0 Mean :0.09957
## 3rd Qu.:112.00 3rd Qu.:111.0 3rd Qu.:0 3rd Qu.:0.11200
## Max. :157.00 Max. :133.0 Max. :0 Max. :0.15350
## NA's :1606
## ba
## Min. :0.0005726
## 1st Qu.:0.0060821
## Median :0.0080118
## Mean :0.0080541
## 3rd Qu.:0.0098520
## Max. :0.0185057
## Basal area of only living trees
L_pineba <- summaryBy (ba + dbh ~plot+block+treatment,
data = L_pine1,
na.rm = T,
keep.names = T,
FUN = sum)
head(L_pineba)## plot block treatment ba dbh
## 1 1 1 Control 1.471621 19.733
## 2 2 1 Deep 2.121139 24.020
## 3 3 2 Control 1.392638 19.330
## 4 4 2 Deep 1.962766 23.030
## 5 5 3 Control 1.599848 20.393
## 6 6 4 Deep 1.918627 22.778Estimating basal area per hectare
L_pineba$baha <- L_pineba$ba * 10000/(30*30)Visualizing basal area per hectare
barchart(baha~block,
group = treatment,
data = L_pineba,
ylab = 'basal area per hectare (m3 ha-1)',
xlab = 'plots',
ylim = c(0, 30),
main = 'Basal area per hectare across plots',
auto.key = list(corners = c(0.05, 0.95)))
Stem Density
L_pinedens <- summaryBy(ba~plot+block+treatment,
data = L_pine1, FUN = length)
L_pinedens$dens_ha <- L_pinedens$ba.length * 10000 / (30*30) barchart(dens_ha~block,
group = treatment,
data = L_pinedens,
ylab = 'Density per Hectare',
xlab = 'plots',
ylim = c(0, 3000),
main = 'Density of Stand',
auto.key = list(corner = c(0.1, 0.95)))
Analysis of Variance
M_ba <- lm(baha~block+treatment,
data = L_pineba)
Anova(M_ba)## Anova Table (Type II tests)
##
## Response: baha
## Sum Sq Df F value Pr(>F)
## block 5.631 3 5.5025 0.0975220 .
## treatment 75.241 1 220.5562 0.0006624 ***
## Residuals 1.023 3
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1M_dens <- lm(dens_ha~block+treatment,
data = L_pinedens)
Anova(M_dens)## Anova Table (Type II tests)
##
## Response: dens_ha
## Sum Sq Df F value Pr(>F)
## block 293.2 3 0.0512 0.9820
## treatment 4459.9 1 2.3369 0.2238
## Residuals 5725.3 3Basal area was significantly higher in the deep plowing treatment but has no significant difference in stem density indicating that the primary reason to the good response of the deep plowing has been on growth more than on survival