VOR

VOR.null <- glmer(value ~ 1  + (1|Location/Pasture/year), 
                  data=subset(veg.s.stack, response=="Visual Obstruction Reading"), family = Gamma (link = "identity"))  
VOR.mgmt <- glmer(value ~ management + (1|Location/Pasture/year), 
                  data=subset(veg.s.stack, response=="Visual Obstruction Reading"), family = Gamma (link = "identity"))

anova(VOR.null, VOR.mgmt) #global statistic - report test statistic(chi2 value) and p value
## Data: subset(veg.s.stack, response == "Visual Obstruction Reading")
## Models:
## VOR.null: value ~ 1 + (1 | Location/Pasture/year)
## VOR.mgmt: value ~ management + (1 | Location/Pasture/year)
##          Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)    
## VOR.null  5 493.64 510.92 -241.82   483.64                             
## VOR.mgmt  8 467.75 495.40 -225.88   451.75 31.886      3  5.532e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCVOR <- glht(VOR.mgmt, linfct = mcp(management = "Tukey"))
summary(MCVOR) # 2x had signif greater cover than cont; no other mgmt types were signif different from one another
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.s.stack, response == "Visual Obstruction Reading"), 
##     family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0      -0.74263    0.41815  -1.776
## recently burned - continuous == 0          -1.28450    0.42233  -3.041
## twice-over - continuous == 0                0.31511    0.50414   0.625
## recently burned - not recently burned == 0 -0.54186    0.09259  -5.852
## twice-over - not recently burned == 0       1.05774    0.46571   2.271
## twice-over - recently burned == 0           1.59960    0.46980   3.405
##                                            Pr(>|z|)    
## not recently burned - continuous == 0       0.25350    
## recently burned - continuous == 0           0.00979 ** 
## twice-over - continuous == 0                0.91406    
## recently burned - not recently burned == 0  < 0.001 ***
## twice-over - not recently burned == 0       0.08846 .  
## twice-over - recently burned == 0           0.00282 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCVOR) # state the signif mgmt types (estimate/regression coeff = 3.14, 95% CI: 0.22-6.1)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.s.stack, response == "Visual Obstruction Reading"), 
##     family = Gamma(link = "identity"))
## 
## Quantile = 2.4969
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr     upr    
## not recently burned - continuous == 0      -0.7426  -1.7867  0.3014
## recently burned - continuous == 0          -1.2845  -2.3390 -0.2300
## twice-over - continuous == 0                0.3151  -0.9437  1.5739
## recently burned - not recently burned == 0 -0.5419  -0.7730 -0.3107
## twice-over - not recently burned == 0       1.0577  -0.1051  2.2206
## twice-over - recently burned == 0           1.5996   0.4266  2.7726

Litter

ltr.null <- glmer(value ~ 1  + (1|Location/Pasture/year), 
                  data=subset(veg.s.stack, response=="Cumulative Litter Depth"), family = Gamma (link = "identity"))  
ltr.mgmt <- glmer(value ~ management + (1|Location/Pasture/year), 
                  data=subset(veg.s.stack, response=="Cumulative Litter Depth"), family = Gamma (link = "identity"))

anova(ltr.null, ltr.mgmt) 
## Data: subset(veg.s.stack, response == "Cumulative Litter Depth")
## Models:
## ltr.null: value ~ 1 + (1 | Location/Pasture/year)
## ltr.mgmt: value ~ management + (1 | Location/Pasture/year)
##          Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)    
## ltr.null  5 1454.8 1472.1 -722.41   1444.8                             
## ltr.mgmt  8 1396.9 1424.6 -690.47   1380.9 63.889      3  8.669e-14 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCltr <- glht(ltr.mgmt, linfct = mcp(management = "Tukey"))
summary(MCltr) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.s.stack, response == "Cumulative Litter Depth"), 
##     family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0       -3.8178     2.3746  -1.608
## recently burned - continuous == 0          -10.3487     2.3919  -4.327
## twice-over - continuous == 0                -3.0722     3.2275  -0.952
## recently burned - not recently burned == 0  -6.5309     0.6742  -9.687
## twice-over - not recently burned == 0        0.7456     3.0117   0.248
## twice-over - recently burned == 0            7.2765     3.0282   2.403
##                                            Pr(>|z|)    
## not recently burned - continuous == 0        0.3411    
## recently burned - continuous == 0            <0.001 ***
## twice-over - continuous == 0                 0.7538    
## recently burned - not recently burned == 0   <0.001 ***
## twice-over - not recently burned == 0        0.9939    
## twice-over - recently burned == 0            0.0645 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCltr) 
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.s.stack, response == "Cumulative Litter Depth"), 
##     family = Gamma(link = "identity"))
## 
## Quantile = 2.5004
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr      upr     
## not recently burned - continuous == 0       -3.8178  -9.7553   2.1198
## recently burned - continuous == 0          -10.3487 -16.3293  -4.3681
## twice-over - continuous == 0                -3.0722 -11.1422   4.9978
## recently burned - not recently burned == 0  -6.5309  -8.2167  -4.8452
## twice-over - not recently burned == 0        0.7456  -6.7848   8.2760
## twice-over - recently burned == 0            7.2765  -0.2952  14.8482

POPR

POPR.null <- lmer(value ~ 1  + (1|Location/Pasture/year), 
                  data=subset(veg.c.stack, response=="Kentucky bluegrass"), REML=FALSE)  
POPR.mgmt <- lmer(value ~ management + (1|Location/Pasture/year), 
                  data=subset(veg.c.stack, response=="Kentucky bluegrass"), REML=FALSE)

anova(POPR.null, POPR.mgmt) 
## Data: subset(veg.c.stack, response == "Kentucky bluegrass")
## Models:
## POPR.null: value ~ 1 + (1 | Location/Pasture/year)
## POPR.mgmt: value ~ management + (1 | Location/Pasture/year)
##           Df    AIC    BIC  logLik deviance Chisq Chi Df Pr(>Chisq)    
## POPR.null  5 1824.6 1841.9 -907.32   1814.6                            
## POPR.mgmt  8 1801.5 1829.2 -892.77   1785.5 29.08      3  2.154e-06 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCPOPR <- glht(POPR.mgmt, linfct = mcp(management = "Tukey"))
summary(MCPOPR) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: lmer(formula = value ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "Kentucky bluegrass"), 
##     REML = FALSE)
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0         3.691      4.376   0.843
## recently burned - continuous == 0            -7.760      4.717  -1.645
## twice-over - continuous == 0                 10.594      5.378   1.970
## recently burned - not recently burned == 0  -11.451      2.224  -5.149
## twice-over - not recently burned == 0         6.903      4.763   1.449
## twice-over - recently burned == 0            18.353      5.078   3.614
##                                            Pr(>|z|)    
## not recently burned - continuous == 0       0.82263    
## recently burned - continuous == 0           0.33402    
## twice-over - continuous == 0                0.18504    
## recently burned - not recently burned == 0  < 0.001 ***
## twice-over - not recently burned == 0       0.44802    
## twice-over - recently burned == 0           0.00137 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCPOPR)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: lmer(formula = value ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "Kentucky bluegrass"), 
##     REML = FALSE)
## 
## Quantile = 2.5324
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr      upr     
## not recently burned - continuous == 0        3.6909  -7.3905  14.7723
## recently burned - continuous == 0           -7.7597 -19.7039   4.1844
## twice-over - continuous == 0                10.5936  -3.0247  24.2120
## recently burned - not recently burned == 0 -11.4506 -17.0825  -5.8188
## twice-over - not recently burned == 0        6.9027  -5.1601  18.9656
## twice-over - recently burned == 0           18.3534   5.4934  31.2133

BRIN

BRIN.null <- glmer(value + 0.3 ~ 1  + (1|Location/Pasture/year), 
                  data=subset(veg.c.stack, response=="smooth brome"), family = Gamma (link = "identity"))  
BRIN.mgmt <- glmer(value + 0.3 ~ management + (1|Location/Pasture/year), 
                  data=subset(veg.c.stack, response=="smooth brome"), family = Gamma (link = "identity"))

anova(BRIN.null, BRIN.mgmt) 
## Data: subset(veg.c.stack, response == "smooth brome")
## Models:
## BRIN.null: value + 0.3 ~ 1 + (1 | Location/Pasture/year)
## BRIN.mgmt: value + 0.3 ~ management + (1 | Location/Pasture/year)
##           Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)  
## BRIN.null  5 1718.4 1735.7 -854.22   1708.4                           
## BRIN.mgmt  8 1715.5 1743.2 -849.77   1699.5 8.8879      3    0.03082 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCBRIN <- glht(BRIN.mgmt, linfct = mcp(management = "Tukey"))
summary(MCBRIN) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.3 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "smooth brome"), family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0       -13.556      4.149  -3.267
## recently burned - continuous == 0           -15.572      4.260  -3.655
## twice-over - continuous == 0                -17.782      4.657  -3.818
## recently burned - not recently burned == 0   -2.016      1.332  -1.513
## twice-over - not recently burned == 0        -4.226      3.671  -1.151
## twice-over - recently burned == 0            -2.210      3.792  -0.583
##                                            Pr(>|z|)    
## not recently burned - continuous == 0       0.00504 ** 
## recently burned - continuous == 0           0.00102 ** 
## twice-over - continuous == 0                < 0.001 ***
## recently burned - not recently burned == 0  0.40179    
## twice-over - not recently burned == 0       0.63303    
## twice-over - recently burned == 0           0.93078    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCBRIN)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.3 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "smooth brome"), family = Gamma(link = "identity"))
## 
## Quantile = 2.5177
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr      upr     
## not recently burned - continuous == 0      -13.5561 -24.0023  -3.1099
## recently burned - continuous == 0          -15.5719 -26.2971  -4.8467
## twice-over - continuous == 0               -17.7819 -29.5073  -6.0565
## recently burned - not recently burned == 0  -2.0158  -5.3694   1.3377
## twice-over - not recently burned == 0       -4.2258 -13.4671   5.0156
## twice-over - recently burned == 0           -2.2099 -11.7579   7.3380

NatC3

NatC3.null <- glmer(value + 0.7 ~ 1  + (1|Location/Pasture/year), 
                  data=subset(veg.c.stack, response=="native C3"), family = Gamma (link = "identity"))  
NatC3.mgmt <- glmer(value + 0.7 ~ management + (1|Location/Pasture/year), 
                  data=subset(veg.c.stack, response=="native C3"), family = Gamma (link = "identity"))

anova(NatC3.null, NatC3.mgmt) 
## Data: subset(veg.c.stack, response == "native C3")
## Models:
## NatC3.null: value + 0.7 ~ 1 + (1 | Location/Pasture/year)
## NatC3.mgmt: value + 0.7 ~ management + (1 | Location/Pasture/year)
##            Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)    
## NatC3.null  5 1393.8 1411.0 -691.88   1383.8                             
## NatC3.mgmt  8 1376.5 1404.2 -680.27   1360.5 23.211      3   3.65e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCNatC3 <- glht(NatC3.mgmt, linfct = mcp(management = "Tukey"))
summary(MCNatC3) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.7 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native C3"), family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                             Estimate Std. Error z value
## not recently burned - continuous == 0       8.456505   0.004152  2036.5
## recently burned - continuous == 0           3.345839   0.004152   805.8
## twice-over - continuous == 0                2.017572   0.004152   485.9
## recently burned - not recently burned == 0 -5.110666   0.005872  -870.3
## twice-over - not recently burned == 0      -6.438933   0.005872 -1096.5
## twice-over - recently burned == 0          -1.328267   0.005872  -226.2
##                                            Pr(>|z|)    
## not recently burned - continuous == 0        <2e-16 ***
## recently burned - continuous == 0            <2e-16 ***
## twice-over - continuous == 0                 <2e-16 ***
## recently burned - not recently burned == 0   <2e-16 ***
## twice-over - not recently burned == 0        <2e-16 ***
## twice-over - recently burned == 0            <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCNatC3)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.7 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native C3"), family = Gamma(link = "identity"))
## 
## Quantile = 2.5354
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr     upr    
## not recently burned - continuous == 0       8.4565   8.4460  8.4670
## recently burned - continuous == 0           3.3458   3.3353  3.3564
## twice-over - continuous == 0                2.0176   2.0070  2.0281
## recently burned - not recently burned == 0 -5.1107  -5.1256 -5.0958
## twice-over - not recently burned == 0      -6.4389  -6.4538 -6.4240
## twice-over - recently burned == 0          -1.3283  -1.3432 -1.3134

NatC4

NatC4.null <- glmer(value + 0.7 ~ 1  + (1|Location/Pasture/year), 
                   data=subset(veg.c.stack, response=="native C4"), family = Gamma (link = "identity"))  
NatC4.mgmt <- glmer(value + 0.7 ~ management + (1|Location/Pasture/year), 
                   data=subset(veg.c.stack, response=="native C4"), family = Gamma (link = "identity"))

anova(NatC4.null, NatC4.mgmt) 
## Data: subset(veg.c.stack, response == "native C4")
## Models:
## NatC4.null: value + 0.7 ~ 1 + (1 | Location/Pasture/year)
## NatC4.mgmt: value + 0.7 ~ management + (1 | Location/Pasture/year)
##            Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)
## NatC4.null  5 944.11 961.39 -467.06   934.11                         
## NatC4.mgmt  8 947.33 974.97 -465.66   931.33 2.7881      3     0.4255
MCNatC4 <- glht(NatC4.mgmt, linfct = mcp(management = "Tukey"))
summary(MCNatC4) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.7 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native C4"), family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0       1.28916    0.69359   1.859
## recently burned - continuous == 0           1.27785    0.77578   1.647
## twice-over - continuous == 0                1.13084    0.99014   1.142
## recently burned - not recently burned == 0 -0.01131    0.45012  -0.025
## twice-over - not recently burned == 0      -0.15832    0.92466  -0.171
## twice-over - recently burned == 0          -0.14701    0.98799  -0.149
##                                            Pr(>|z|)
## not recently burned - continuous == 0         0.231
## recently burned - continuous == 0             0.334
## twice-over - continuous == 0                  0.647
## recently burned - not recently burned == 0    1.000
## twice-over - not recently burned == 0         0.998
## twice-over - recently burned == 0             0.999
## (Adjusted p values reported -- single-step method)
confint(MCNatC4)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.7 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native C4"), family = Gamma(link = "identity"))
## 
## Quantile = 2.5353
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr      upr     
## not recently burned - continuous == 0       1.28916 -0.46929  3.04762
## recently burned - continuous == 0           1.27785 -0.68900  3.24470
## twice-over - continuous == 0                1.13084 -1.37946  3.64114
## recently burned - not recently burned == 0 -0.01131 -1.15251  1.12989
## twice-over - not recently burned == 0      -0.15832 -2.50261  2.18598
## twice-over - recently burned == 0          -0.14701 -2.65186  2.35785

NatLeg

NatLeg.null <- glmer(value + 0.5 ~ 1  + (1|Location/Pasture/year), 
                   data=subset(veg.c.stack, response=="native legume"), family = Gamma (link = "identity"))  
NatLeg.mgmt <- glmer(value + 0.5 ~ management + (1|Location/Pasture/year), 
                   data=subset(veg.c.stack, response=="native legume"), family = Gamma (link = "identity"))

anova(NatLeg.null, NatLeg.mgmt) 
## Data: subset(veg.c.stack, response == "native legume")
## Models:
## NatLeg.null: value + 0.5 ~ 1 + (1 | Location/Pasture/year)
## NatLeg.mgmt: value + 0.5 ~ management + (1 | Location/Pasture/year)
##             Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)   
## NatLeg.null  5 1055.5 1072.8 -522.77   1045.5                            
## NatLeg.mgmt  8 1050.1 1077.8 -517.06   1034.1 11.403      3   0.009734 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCNatLeg <- glht(NatLeg.mgmt, linfct = mcp(management = "Tukey"))
summary(MCNatLeg) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native legume"), 
##     family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0        3.2920     0.8640   3.810
## recently burned - continuous == 0            3.4401     1.0077   3.414
## twice-over - continuous == 0                 6.7340     1.4156   4.757
## recently burned - not recently burned == 0   0.1481     0.6882   0.215
## twice-over - not recently burned == 0        3.4421     1.3500   2.550
## twice-over - recently burned == 0            3.2940     1.4391   2.289
##                                            Pr(>|z|)    
## not recently burned - continuous == 0       < 0.001 ***
## recently burned - continuous == 0           0.00304 ** 
## twice-over - continuous == 0                < 0.001 ***
## recently burned - not recently burned == 0  0.99623    
## twice-over - not recently burned == 0       0.04839 *  
## twice-over - recently burned == 0           0.09346 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCNatLeg)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native legume"), 
##     family = Gamma(link = "identity"))
## 
## Quantile = 2.5357
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr      upr     
## not recently burned - continuous == 0       3.29196  1.10110  5.48283
## recently burned - continuous == 0           3.44006  0.88493  5.99518
## twice-over - continuous == 0                6.73404  3.14450 10.32359
## recently burned - not recently burned == 0  0.14809 -1.59708  1.89326
## twice-over - not recently burned == 0       3.44208  0.01888  6.86528
## twice-over - recently burned == 0           3.29399 -0.35504  6.94301

IntLeg

IntLeg.null <- glmer(value + 0.5 ~ 1  + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="introduced legume"), family = Gamma (link = "identity"))  
IntLeg.mgmt <- glmer(value + 0.5 ~ management + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="introduced legume"), family = Gamma (link = "identity"))

anova(IntLeg.null, IntLeg.mgmt) 
## Data: subset(veg.c.stack, response == "introduced legume")
## Models:
## IntLeg.null: value + 0.5 ~ 1 + (1 | Location/Pasture/year)
## IntLeg.mgmt: value + 0.5 ~ management + (1 | Location/Pasture/year)
##             Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)  
## IntLeg.null  5 901.52 918.80 -445.76   891.52                           
## IntLeg.mgmt  8 898.20 925.85 -441.10   882.20 9.3153      3    0.02538 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCIntLeg <- glht(IntLeg.mgmt, linfct = mcp(management = "Tukey"))
summary(MCIntLeg) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "introduced legume"), 
##     family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                             Estimate Std. Error z value
## not recently burned - continuous == 0       2.050180   0.002781  737.26
## recently burned - continuous == 0           1.762071   0.002781  633.63
## twice-over - continuous == 0               -0.452036   0.002872 -157.39
## recently burned - not recently burned == 0 -0.288108   0.003933  -73.26
## twice-over - not recently burned == 0      -2.502216   0.003998 -625.91
## twice-over - recently burned == 0          -2.214108   0.003998 -553.83
##                                            Pr(>|z|)    
## not recently burned - continuous == 0        <2e-16 ***
## recently burned - continuous == 0            <2e-16 ***
## twice-over - continuous == 0                 <2e-16 ***
## recently burned - not recently burned == 0   <2e-16 ***
## twice-over - not recently burned == 0        <2e-16 ***
## twice-over - recently burned == 0            <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCIntLeg)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "introduced legume"), 
##     family = Gamma(link = "identity"))
## 
## Quantile = 2.5356
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr     upr    
## not recently burned - continuous == 0       2.0502   2.0431  2.0572
## recently burned - continuous == 0           1.7621   1.7550  1.7691
## twice-over - continuous == 0               -0.4520  -0.4593 -0.4448
## recently burned - not recently burned == 0 -0.2881  -0.2981 -0.2781
## twice-over - not recently burned == 0      -2.5022  -2.5124 -2.4921
## twice-over - recently burned == 0          -2.2141  -2.2242 -2.2040

NatFrb

NatFrb.null <- lmer(value + 0.5 ~ 1  + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="native forb"), REML=FALSE)  
NatFrb.mgmt <- lmer(value + 0.5 ~ management + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="native forb"), REML=FALSE)

anova(NatFrb.null, NatFrb.mgmt) 
## Data: subset(veg.c.stack, response == "native forb")
## Models:
## NatFrb.null: value + 0.5 ~ 1 + (1 | Location/Pasture/year)
## NatFrb.mgmt: value + 0.5 ~ management + (1 | Location/Pasture/year)
##             Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)  
## NatFrb.null  5 1524.2 1541.5 -757.12   1514.2                           
## NatFrb.mgmt  8 1521.5 1549.2 -752.77   1505.5 8.7039      3     0.0335 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCNatFrb <- glht(NatFrb.mgmt, linfct = mcp(management = "Tukey"))
summary(MCNatFrb) 
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: lmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native forb"), REML = FALSE)
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0        11.721      2.409   4.865
## recently burned - continuous == 0            11.904      2.586   4.604
## twice-over - continuous == 0                 14.369      3.296   4.360
## recently burned - not recently burned == 0    0.183      1.186   0.154
## twice-over - not recently burned == 0         2.648      2.995   0.884
## twice-over - recently burned == 0             2.466      3.138   0.786
##                                            Pr(>|z|)    
## not recently burned - continuous == 0        <1e-04 ***
## recently burned - continuous == 0            <1e-04 ***
## twice-over - continuous == 0                 <1e-04 ***
## recently burned - not recently burned == 0    0.999    
## twice-over - not recently burned == 0         0.799    
## twice-over - recently burned == 0             0.850    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCNatFrb)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: lmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native forb"), REML = FALSE)
## 
## Quantile = 2.5276
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr     upr    
## not recently burned - continuous == 0      11.7209   5.6311 17.8108
## recently burned - continuous == 0          11.9039   5.3686 18.4392
## twice-over - continuous == 0               14.3694   6.0397 22.6992
## recently burned - not recently burned == 0  0.1830  -2.8137  3.1798
## twice-over - not recently burned == 0       2.6485  -4.9214 10.2185
## twice-over - recently burned == 0           2.4655  -5.4672 10.3982

IntFrb

IntFrb.null <- glmer(value + 0.5 ~ 1  + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="introduced forb"), family = Gamma (link = "identity"))  
IntFrb.mgmt <- glmer(value + 0.5 ~ management + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="introduced forb"), family = Gamma (link = "identity"))

anova(IntFrb.null, IntFrb.mgmt) #global statistic - report test statistic(chi2 value) and p value
## Data: subset(veg.c.stack, response == "introduced forb")
## Models:
## IntFrb.null: value + 0.5 ~ 1 + (1 | Location/Pasture/year)
## IntFrb.mgmt: value + 0.5 ~ management + (1 | Location/Pasture/year)
##             Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)  
## IntFrb.null  5 934.90 952.18 -462.45   924.90                           
## IntFrb.mgmt  8 934.62 962.27 -459.31   918.62 6.2798      3    0.09876 .
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCIntFrb <- glht(IntFrb.mgmt, linfct = mcp(management = "Tukey"))
summary(MCIntFrb) # 2x had signif greater cover than cont; no other mgmt types were signif different from one another
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "introduced forb"), 
##     family = Gamma(link = "identity"))
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0        1.3194     0.8672   1.522
## recently burned - continuous == 0            1.9276     0.9742   1.979
## twice-over - continuous == 0                 3.1436     1.1551   2.722
## recently burned - not recently burned == 0   0.6081     0.5202   1.169
## twice-over - not recently burned == 0        1.8241     1.0547   1.729
## twice-over - recently burned == 0            1.2160     1.1420   1.065
##                                            Pr(>|z|)  
## not recently burned - continuous == 0        0.4049  
## recently burned - continuous == 0            0.1823  
## twice-over - continuous == 0                 0.0294 *
## recently burned - not recently burned == 0   0.6287  
## twice-over - not recently burned == 0        0.2906  
## twice-over - recently burned == 0            0.6952  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCIntFrb) # state the signif mgmt types (estimate/regression coeff = 3.14, 95% CI: 0.22-6.1)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: glmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "introduced forb"), 
##     family = Gamma(link = "identity"))
## 
## Quantile = 2.5353
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate lwr     upr    
## not recently burned - continuous == 0       1.3194  -0.8791  3.5180
## recently burned - continuous == 0           1.9276  -0.5424  4.3975
## twice-over - continuous == 0                3.1436   0.2150  6.0721
## recently burned - not recently burned == 0  0.6081  -0.7107  1.9270
## twice-over - not recently burned == 0       1.8241  -0.8500  4.4982
## twice-over - recently burned == 0           1.2160  -1.6795  4.1115

NatWdy

NatWdy.null <- lmer(value + 0.5 ~ 1  + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="native woody"), REML=FALSE)  
NatWdy.mgmt <- lmer(value + 0.5 ~ management + (1|Location/Pasture/year), 
                    data=subset(veg.c.stack, response=="native woody"), REML=FALSE)

anova(NatWdy.null, NatWdy.mgmt) #global statistic - report test statistic(chi2 value) and p value
## Data: subset(veg.c.stack, response == "native woody")
## Models:
## NatWdy.null: value + 0.5 ~ 1 + (1 | Location/Pasture/year)
## NatWdy.mgmt: value + 0.5 ~ management + (1 | Location/Pasture/year)
##             Df    AIC    BIC  logLik deviance  Chisq Chi Df Pr(>Chisq)  
## NatWdy.null  5 1362.0 1379.3 -676.01   1352.0                           
## NatWdy.mgmt  8 1359.7 1387.3 -671.85   1343.7 8.3144      3    0.03994 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
MCNatWdy <- glht(NatWdy.mgmt, linfct = mcp(management = "Tukey"))
summary(MCNatWdy) # 2x had signif greater cover than cont; no other mgmt types were signif different from one another
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: lmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native woody"), REML = FALSE)
## 
## Linear Hypotheses:
##                                            Estimate Std. Error z value
## not recently burned - continuous == 0      0.008006   1.390542   0.006
## recently burned - continuous == 0          0.507787   1.560801   0.325
## twice-over - continuous == 0               5.680564   1.683809   3.374
## recently burned - not recently burned == 0 0.499781   0.895684   0.558
## twice-over - not recently burned == 0      5.672558   1.488336   3.811
## twice-over - recently burned == 0          5.172777   1.648527   3.138
##                                            Pr(>|z|)    
## not recently burned - continuous == 0       1.00000    
## recently burned - continuous == 0           0.98737    
## twice-over - continuous == 0                0.00411 ** 
## recently burned - not recently burned == 0  0.94121    
## twice-over - not recently burned == 0       < 0.001 ***
## twice-over - recently burned == 0           0.00861 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)
confint(MCNatWdy) # state the signif mgmt types (estimate/regression coeff = 3.14, 95% CI: 0.22-6.1)
## 
##   Simultaneous Confidence Intervals
## 
## Multiple Comparisons of Means: Tukey Contrasts
## 
## 
## Fit: lmer(formula = value + 0.5 ~ management + (1 | Location/Pasture/year), 
##     data = subset(veg.c.stack, response == "native woody"), REML = FALSE)
## 
## Quantile = 2.5412
## 95% family-wise confidence level
##  
## 
## Linear Hypotheses:
##                                            Estimate  lwr       upr      
## not recently burned - continuous == 0       0.008006 -3.525668  3.541680
## recently burned - continuous == 0           0.507787 -3.458555  4.474128
## twice-over - continuous == 0                5.680564  1.401633  9.959495
## recently burned - not recently burned == 0  0.499781 -1.776351  2.775913
## twice-over - not recently burned == 0       5.672558  1.890366  9.454751
## twice-over - recently burned == 0           5.172777  0.983506  9.362049