Visualize and Test data
# Box plots of each variable with CPUE (count)
boxplot(logCPUE ~ Vegetation.Type, data = Site.CPUE, xlab = "Vegetation Type", ylab = "logCPUE")
boxplot(logCPUE ~ Unit, data = Site.CPUE, xlab = "Unit", ylab = "logCPUE")
boxplot(logCPUE ~ Month, data = Site.CPUE, aes(x=factor(Month, level=c('May','June','July','August','October'), xlab = "Month", ylab = "logCPUE")))
# Testing normality of all CPUE data with Shapiro-Wilks
shapiro.test(Site.CPUE$logCPUE)
##
## Shapiro-Wilk normality test
##
## data: Site.CPUE$logCPUE
## W = 0.95506, p-value = 0.004245
# ANOVA - all variables, residuals
ANOVA_FULL = aov(logCPUE ~ Vegetation.Type + Unit + Month, data = Site.CPUE)
summary(ANOVA_FULL)
## Df Sum Sq Mean Sq F value Pr(>F)
## Vegetation.Type 5 26.20 5.239 3.137 0.01277 *
## Unit 4 24.77 6.193 3.708 0.00838 **
## Month 4 59.91 14.977 8.966 6.02e-06 ***
## Residuals 73 121.94 1.670
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
qqPlot(resid(ANOVA_FULL)) # Checking for normal distribution. If it does not meet assumptions, mention it in report
## [1] 85 9
summary(ANOVA_FULL)
## Df Sum Sq Mean Sq F value Pr(>F)
## Vegetation.Type 5 26.20 5.239 3.137 0.01277 *
## Unit 4 24.77 6.193 3.708 0.00838 **
## Month 4 59.91 14.977 8.966 6.02e-06 ***
## Residuals 73 121.94 1.670
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
# ANOVA of vegetation and count # tukey of ANOVA # Levene for normal variance in data (is variation significantly different?)
ANOVA_VEG = aov(logCPUE ~ Vegetation.Type, data = Site.CPUE)
summary(ANOVA_VEG)
## Df Sum Sq Mean Sq F value Pr(>F)
## Vegetation.Type 5 26.2 5.239 2.054 0.0797 .
## Residuals 81 206.6 2.551
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
qqPlot(resid(ANOVA_VEG))
## [1] 85 80
TukeyHSD(ANOVA_VEG)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = logCPUE ~ Vegetation.Type, data = Site.CPUE)
##
## $Vegetation.Type
## diff lwr upr p adj
## ME-Forest 1.65202679 -1.039624689 4.3436783 0.4768803
## OW-Forest 1.74011034 -0.457613889 3.9378346 0.2016105
## RB-Forest 1.88765349 -3.147965335 6.9232723 0.8823193
## SAV-Forest 2.02069337 0.001954754 4.0394320 0.0496341
## Typha-Forest 1.00685133 -1.510958089 3.5246607 0.8509461
## OW-ME 0.08808355 -2.109640684 2.2858078 0.9999968
## RB-ME 0.23562670 -4.799992130 5.2712455 0.9999931
## SAV-ME 0.36866657 -1.650072041 2.3874052 0.9946533
## Typha-ME -0.64517547 -3.162984884 1.8726339 0.9751303
## RB-OW 0.14754315 -4.642285771 4.9373721 0.9999991
## SAV-OW 0.28058302 -1.007947025 1.5691131 0.9879670
## Typha-OW -0.73325902 -2.714260872 1.2477428 0.8877685
## SAV-RB 0.13303987 -4.577350223 4.8434300 0.9999994
## Typha-RB -0.88080217 -5.825681694 4.0640774 0.9952425
## Typha-SAV -1.01384204 -2.794202152 0.7665181 0.5603715
leveneTest(CPUE ~ Vegetation.Type, data = Site.CPUE)
## Warning in leveneTest.default(y = y, group = group, ...): group coerced to
## factor.
# ANOVA of month and count # tukey of ANOVA # levene test
ANOVA_MONTH = aov(logCPUE ~ Month, data = Site.CPUE)
summary(ANOVA_MONTH)
## Df Sum Sq Mean Sq F value Pr(>F)
## Month 4 61.64 15.410 7.383 3.96e-05 ***
## Residuals 82 171.17 2.087
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
qqPlot(resid(ANOVA_MONTH))
## [1] 85 50
TukeyHSD(ANOVA_MONTH)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = logCPUE ~ Month, data = Site.CPUE)
##
## $Month
## diff lwr upr p adj
## June-May 0.36206651 -0.9001131 1.6242462 0.9298306
## July-May 2.33797534 0.9601528 3.7157979 0.0000877
## August-May 1.49515614 0.2329765 2.7573358 0.0119970
## October-May 0.42757606 -0.9822593 1.8374114 0.9153183
## July-June 1.97590882 0.5564149 3.3954028 0.0018987
## August-June 1.13308963 -0.1744522 2.4406314 0.1210227
## October-June 0.06550954 -1.3850779 1.5160969 0.9999420
## August-July -0.84281919 -2.2623131 0.5766747 0.4665436
## October-July -1.91039928 -3.4626561 -0.3581425 0.0081240
## October-August -1.06758008 -2.5181675 0.3830073 0.2506608
leveneTest(CountOfSite.Number ~ Month, data = Site.CPUE)
# ANOVA of unit and count # tukey of ANOVA # levene test
ANOVA_UNIT = aov(logCPUE ~ Unit, data = Site.CPUE)
summary(ANOVA_UNIT)
## Df Sum Sq Mean Sq F value Pr(>F)
## Unit 4 32.16 8.039 3.285 0.015 *
## Residuals 82 200.65 2.447
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
qqPlot(resid(ANOVA_UNIT))
## [1] 85 80
TukeyHSD(ANOVA_UNIT)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = logCPUE ~ Unit, data = Site.CPUE)
##
## $Unit
## diff lwr upr p adj
## MN-MC -0.24567367 -1.7092113 1.2178640 0.9899684
## MS-MC -1.19549842 -2.6291648 0.2381680 0.1471773
## P1A-MC -1.11611791 -2.8750712 0.6428354 0.3979769
## SHR-MC 0.32879366 -1.1517624 1.8093497 0.9715864
## MS-MN -0.94982475 -2.2979366 0.3982871 0.2921422
## P1A-MN -0.87044424 -2.5603919 0.8195035 0.6059952
## SHR-MN 0.57446732 -0.8234070 1.9723417 0.7813912
## P1A-MS 0.07938052 -1.5847649 1.7435260 0.9999279
## SHR-MS 1.52429208 0.1577235 2.8908606 0.0210144
## SHR-P1A 1.44491156 -0.2597958 3.1496189 0.1357174
leveneTest(CountOfSite.Number ~ Unit, data = Site.CPUE)
## Warning in leveneTest.default(y = y, group = group, ...): group coerced to
## factor.
Is unit still significant when SHR is excluded?
##WITHOUT SHR
wetland_unit <- subset(Site.CPUE, Unit %in% c('MC','MN','MS','P1A'))
boxplot(logCPUE ~ Unit, data = wetland_unit, xlab = "Unit", ylab = "logCPUE")
# ANOVA of unit (without SHR) and count # tukey of ANOVA # levene test
ANOVA_UNIT_2 = aov(logCPUE ~ Unit, data = wetland_unit)
summary(ANOVA_UNIT_2)
## Df Sum Sq Mean Sq F value Pr(>F)
## Unit 3 18.7 6.233 3.307 0.0256 *
## Residuals 64 120.6 1.885
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
qqPlot(resid(ANOVA_UNIT_2))
## [1] 3 41
TukeyHSD(ANOVA_UNIT_2)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = logCPUE ~ Unit, data = wetland_unit)
##
## $Unit
## diff lwr upr p adj
## MN-MC -0.24567367 -1.460377 0.969029888 0.9505801
## MS-MC -1.19549842 -2.385410 -0.005587342 0.0485065
## P1A-MC -1.11611791 -2.576010 0.343774081 0.1926483
## MS-MN -0.94982475 -2.068728 0.169078016 0.1237256
## P1A-MN -0.87044424 -2.273063 0.532174614 0.3655428
## P1A-MS 0.07938052 -1.301823 1.460584081 0.9987453
leveneTest(CountOfSite.Number ~ Unit, data = wetland_unit)
## Warning in leveneTest.default(y = y, group = group, ...): group coerced to
## factor.