Thesis Data: Agroforestry v3.0

This R Markdown document contains all statistical analysis and visualization to date of Dari’s agroforestry project, as completed by Murray Sternberg.

One-way ANOVAs were performed to analyze the effects of agroforestry type (pecan vs. pine), cover crop type (for pecan, pine, and the two sites combined), and tillage (no-till vs. minimum tillage) on biological parameters. Two-way ANOVAs were perforemd to analyze the effects and interactions of tillage and agroforestry type, and tillage and cover crop type on biological parameters. For one-way ANOVAs with significant results, post-hoc testing of multiple comparisons between groups was done using Tukey’s HSD test.

Biological parameters (response variables) included microbial biomass carbon (MBC), microbial biomass nitrogen (MBN), 24-hour soil microbial respiration, enzyme activities, microbial necromass, total soil carbon (TC), and total soil nitrogen (TN). Statistical significance was calculated with a significance level of 0.05.

TC & TN data is not yet available but will be analyzed and incorporated when the instrument is available.

This document is broken into 5 main sections, each representing a parameter to be analysed (shown in the table of contents), followed by visualizations of summary statistics. Each section starts off with a summary of significant and non-significant results. Within each section, there are 6 main subsections, each representing a statistical comparison:

pecan vs. pine (with control),
the 10 cover crop types, combined (with control),
the 6 cover crop types used for pecan (with control),
the 4 cover crop types used for pine (with control),
no-till vs. minimum tillage, and
two-way ANOVAs (agroforestry type * tillage, cover crop type * tillage).

Microbial Biomass C

Significance Summary:

When considering all 3 seasons of data, the following analyses were found to be significant:

Agroforestry Tree Type

The following analyses were not significant:

Cover Crop Type - Combined Pecan & Pine
Cover Crop Type - Pecan
Cover Crop Type - Pine
Tillage Treatment
No interaction between tree type and tillage
No interaction between cover crop type and tillage

Agroforestry Type

# perform anova 
anova_MBCagrotype <- aov(MBIO_C ~ agro_type,  data = mbc_nocntrl)
summary(anova_MBCagrotype)

##             Df Sum Sq Mean Sq F value Pr(>F)  
## agro_type    1  58316   58316   5.676 0.0212 *
## Residuals   48 493170   10274                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 10 observations deleted due to missingness

The ANOVA revealed that there was a statistically significant difference in microbial biomass C between agroforestry tree type (pecan vs. pine) when considering all 3 seasons of data (p < 0.001).

Cover Crop Type (Combined)

# perform anova for combined for all seasons
anova_MBCCCtype <- aov(MBIO_C ~ cc_type,  data = mbio_all_avg)
summary(anova_MBCCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type     10 148966   14897   1.351  0.236
## Residuals   43 474185   11028               
## 10 observations deleted due to missingness

The ANOVA revealed that there was not a statistically significant difference in microbial biomass C between cover crop type when considering both sites combined (pecan and pine) with all 3 seasons of data.

Cover Crop Type (Pecan)

# perform anova for pecan for all seasons
anova_MBCpecCCtype <- aov(MBIO_C ~ cc_type,  data = mbcPEC)
summary(anova_MBCpecCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type      5  75900   15180   1.434   0.25
## Residuals   23 243549   10589               
## 7 observations deleted due to missingness

The ANOVA revealed that there was not a statistically significant difference in microbial biomass C between cover crop type (of the 6 types) for the pecan site when considering all 3 seasons of data.

Cover Crop Type (Pine)

# perform anova for pine for all seasons
anova_MBCpinCCtype <- aov(MBIO_C ~ cc_type,  data = mbcPIN)
summary(anova_MBCpinCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type      3  13010    4337   0.459  0.715
## Residuals   17 160711    9454               
## 3 observations deleted due to missingness

The ANOVA revealed that there was not a statistically significant difference in microbial biomass C between cover crop type (of the 4 types) for the pine site when considering all 3 seasons of data.

Tillage Treatment

# perform anova for all seasons
anova_mbcTIL <- aov(MBIO_C ~ tillage,  data = mbcTIL)
summary(anova_mbcTIL)

##             Df Sum Sq Mean Sq F value Pr(>F)  
## tillage      1  50458   50458   4.834 0.0328 *
## Residuals   48 501028   10438                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 10 observations deleted due to missingness

The ANOVA revealed that there was not a statistically significant difference in microbial biomass C between tillage treatments (minimum tillage vs. no-till) when considering all 3 seasons of data.

Two-Way ANOVAs

# Performing two-way ANOVA for MBIO-C using Agroforestry Type & Tillage
anova_MBC_agtl <- aov(MBIO_C ~ agro_type * tillage, data = mbc_nocntrl)
summary(anova_MBC_agtl)

##                   Df Sum Sq Mean Sq F value Pr(>F)  
## agro_type          1  58316   58316   6.141 0.0169 *
## tillage            1  49759   49759   5.240 0.0267 *
## agro_type:tillage  1   6616    6616   0.697 0.4082  
## Residuals         46 436796    9496                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 10 observations deleted due to missingness

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of agroforestry type and tillage.

# Performing two-way ANOVA for MBIO-C using Cover Crop Type & Tillage
anova_MBC_cctl <- aov(MBIO_C ~ cc_type * tillage, data = mbc_nocntrl)
summary(anova_MBC_cctl)

##                 Df Sum Sq Mean Sq F value Pr(>F)  
## cc_type          9 147226   16358   1.510 0.1895  
## tillage          1  39490   39490   3.646 0.0658 .
## cc_type:tillage  9  39862    4429   0.409 0.9203  
## Residuals       30 324908   10830                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 10 observations deleted due to missingness

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of cover crop type and tillage.

Microbial Biomass N

Significance Summary:

When considering all 3 seasons of data, the following analyses were found to be significant:

None

The following analyses were not significant:

Agroforestry Tree Type
Cover Crop Type - Combined Pecan & Pine
Cover Crop Type - Pecan
Cover Crop Type - Pine
Tillage Treatment
No interaction between tree type and tillage
No interaction between cover crop type and tillage

Agroforestry Type

# perform anova for all seasons
anova_MBNagrotype <- aov(MBIO_N ~ agro_type,  data = mbn_nocntrl)
summary(anova_MBNagrotype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## agro_type    1     37    36.7   0.039  0.844
## Residuals   43  40100   932.6

The ANOVA revealed that there was not a statistically significant difference in microbial biomass N between agroforestry tree type (pecan vs. pine) when considering all 3 seasons of data.

Cover Crop Type (Combined)

# perform anova for combined for all seasons
anova_MBNCCtype <- aov(MBIO_N ~ cc_type,  data = mbn)
summary(anova_MBNCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type     10   7608   760.8    0.87  0.568
## Residuals   38  33238   874.7

The ANOVA revealed that there was not a statistically significant difference in microbial biomass N between cover crop type when considering both sites combined (pecan and pine) considering all 3 seasons of data.

Cover Crop Type (Pecan)

# perform anova for pecan for all seasons
anova_MBNpecCCtype <- aov(MBIO_N ~ cc_type,  data = mbnPEC)
summary(anova_MBNpecCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type      5   4290     858   1.709  0.184
## Residuals   18   9036     502

The ANOVA revealed that there was not a statistically significant difference in microbial biomass N between cover crop type (of the 6 types) for the pecan site considering all 3 seasons of data.

Cover Crop Type (Pine)

# perform anova for pine for all seasons
anova_MBNpinCCtype <- aov(MBIO_N ~ cc_type,  data = mbnPIN)
summary(anova_MBNpinCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type      3   2857   952.2   0.677  0.578
## Residuals   17  23917  1406.9

The ANOVA revealed that there was not a statistically significant difference in microbial biomass N between cover crop type (of the 4 types) for the pine site considering all 3 seasons of data.

Tillage Treatment

# perform anova for all seasons
anova_mbnTIL <- aov(MBIO_N ~ tillage,  data = mbnTIL)
summary(anova_mbnTIL)

##             Df Sum Sq Mean Sq F value Pr(>F)
## tillage      1    555   555.4   0.603  0.442
## Residuals   43  39582   920.5

The ANOVA revealed that there was not a statistically significant difference in microbial biomass N between tillage treatments (minimum tillage vs. no-till) considering all 3 seasons of data.

Two-Way ANOVAs

# Performing two-way ANOVA for MBIO-N using Agroforestry Type & Tillage
anova_MBN_agtl <- aov(MBIO_N ~ agro_type * tillage, data = mbn_nocntrl)
summary(anova_MBN_agtl)

##                   Df Sum Sq Mean Sq F value Pr(>F)
## agro_type          1     37    36.7   0.040  0.842
## tillage            1    547   547.4   0.603  0.442
## agro_type:tillage  1   2306  2305.7   2.538  0.119
## Residuals         41  37247   908.5

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of agroforestry type and tillage.

# Performing two-way ANOVA for MBIO-N using Cover Crop Type & Tillage
anova_MBN_cctl <- aov(MBIO_N ~ cc_type * tillage, data = mbn_nocntrl)
summary(anova_MBN_cctl)

##                 Df Sum Sq Mean Sq F value Pr(>F)
## cc_type          9   7183   798.2   0.777  0.639
## tillage          1    646   646.4   0.629  0.435
## cc_type:tillage  8   5584   698.0   0.679  0.705
## Residuals       26  26723  1027.8

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of cover crop type and tillage.

24-Hour Microbial Respiration

Significance Summary:

When considering all 3 seasons of data, the following analyses were found to be significant:

Agroforestry Tree Type
Cover Crop Type - Combined Pecan & Pine
Cover Crop Type - Pecan

The following analyses were not significant:

Cover Crop Type - Pine
Tillage Treatment
No interaction between tree type and tillage
No interaction between cover crop type and tillage

Agroforestry Type

# perform anova 
anova_CMINagrotype <- aov(`mg CO2/g dry weight` ~ agro_type,  data = cmin_nocntrl)
summary(anova_CMINagrotype)

##              Df Sum Sq Mean Sq F value   Pr(>F)    
## agro_type     1   71.0   71.02   32.99 3.93e-08 ***
## Residuals   178  383.1    2.15                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in microbial respiration between agroforestry tree type (pecan vs. pine) (p < 0.001).

Cover Crop Type (Combined)

# perform anova 
anova_CMINCCtype <- aov(`mg CO2/g dry weight` ~ cc_type,  data = cmin_allcc)
summary(anova_CMINCCtype)

##              Df Sum Sq Mean Sq F value   Pr(>F)    
## cc_type       9  108.1  12.009   5.899 3.71e-07 ***
## Residuals   170  346.1   2.036                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in microbial respiration between cover crop type when considering both sites (pecan and pine) combined (p < 0.001).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = `mg CO2/g dry weight` ~ cc_type, data = cmin_allcc)
## 
## $cc_type
##                                 diff         lwr         upr     p adj
## oat-clover-oat           0.256684648 -1.26815870  1.78152800 0.9999406
## oat-vetch-oat            0.792561578 -0.73228177  2.31740493 0.8125330
## rye-oat                 -0.765517050 -2.29036040  0.75932630 0.8418212
## rye-clover-oat           0.745738446 -0.77910490  2.27058179 0.8615394
## rye-vetch-oat            0.629976325 -0.89486702  2.15481967 0.9466521
## radish-oat              -0.873354322 -2.39819767  0.65148903 0.7111691
## radish-oat-oat          -1.455073291 -2.97991664  0.06977006 0.0751857
## radish-rye-oat          -0.759458756 -2.28430210  0.76538459 0.8480173
## radish-vetch-oat        -0.934395154 -2.45923850  0.59044820 0.6247918
## oat-vetch-oat-clover     0.535876930 -0.98896642  2.06072028 0.9812996
## rye-oat-clover          -1.022201698 -2.54704505  0.50264165 0.4957286
## rye-clover-oat-clover    0.489053798 -1.03578955  2.01389715 0.9901349
## rye-vetch-oat-clover     0.373291677 -1.15155167  1.89813503 0.9987167
## radish-oat-clover       -1.130038970 -2.65488232  0.39480438 0.3468704
## radish-oat-oat-clover   -1.711757939 -3.23660129 -0.18691459 0.0148558
## radish-rye-oat-clover   -1.016143404 -2.54098675  0.50869994 0.5045739
## radish-vetch-oat-clover -1.191079801 -2.71592315  0.33376355 0.2737073
## rye-oat-vetch           -1.558078628 -3.08292198 -0.03323528 0.0408033
## rye-clover-oat-vetch    -0.046823132 -1.57166648  1.47802022 1.0000000
## rye-vetch-oat-vetch     -0.162585253 -1.68742860  1.36225810 0.9999989
## radish-oat-vetch        -1.665915899 -3.19075925 -0.14107255 0.0203181
## radish-oat-oat-vetch    -2.247634869 -3.77247822 -0.72279152 0.0002024
## radish-rye-oat-vetch    -1.552020334 -3.07686368 -0.02717699 0.0423612
## radish-vetch-oat-vetch  -1.726956731 -3.25180008 -0.20211338 0.0133626
## rye-clover-rye           1.511255496 -0.01358785  3.03609884 0.0542436
## rye-vetch-rye            1.395493375 -0.12934997  2.92033672 0.1043124
## radish-rye              -0.107837271 -1.63268062  1.41700608 1.0000000
## radish-oat-rye          -0.689556241 -2.21439959  0.83528711 0.9093053
## radish-rye-rye           0.006058294 -1.51878505  1.53090164 1.0000000
## radish-vetch-rye        -0.168878103 -1.69372145  1.35596525 0.9999984
## rye-vetch-rye-clover    -0.115762121 -1.64060547  1.40908123 0.9999999
## radish-rye-clover       -1.619092768 -3.14393612 -0.09424942 0.0276932
## radish-oat-rye-clover   -2.200811737 -3.72565509 -0.67596839 0.0003072
## radish-rye-rye-clover   -1.505197202 -3.03004055  0.01964615 0.0562325
## radish-vetch-rye-clover -1.680133599 -3.20497695 -0.15529025 0.0184568
## radish-rye-vetch        -1.503330647 -3.02817400  0.02151270 0.0568577
## radish-oat-rye-vetch    -2.085049616 -3.60989296 -0.56020627 0.0008344
## radish-rye-rye-vetch    -1.389435081 -2.91427843  0.13540827 0.1077236
## radish-vetch-rye-vetch  -1.564371478 -3.08921483 -0.03952813 0.0392380
## radish-oat-radish       -0.581718969 -2.10656232  0.94312438 0.9677044
## radish-rye-radish        0.113895566 -1.41094778  1.63873891 1.0000000
## radish-vetch-radish     -0.061040832 -1.58588418  1.46380252 1.0000000
## radish-rye-radish-oat    0.695614535 -0.82922881  2.22045788 0.9047488
## radish-vetch-radish-oat  0.520678137 -1.00416521  2.04552149 0.9846579
## radish-vetch-radish-rye -0.174936397 -1.69977975  1.34990695 0.9999978

Cover Crop Type (Pecan)

# perform anova for pecan
anova_CMINpecCCtype <- aov(`mg CO2/g dry weight` ~ cc_type,  data = cminPEC)
summary(anova_CMINpecCCtype)

##              Df Sum Sq Mean Sq F value Pr(>F)  
## cc_type       5  31.93   6.387   2.414 0.0411 *
## Residuals   102 269.87   2.646                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in microbial respiration between cover crop type (of the 6 types) for the pecan site (p < 0.05).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = `mg CO2/g dry weight` ~ cc_type, data = cminPEC)
## 
## $cc_type
##                              diff         lwr        upr     p adj
## oat-clover-oat         0.25668465 -1.31819119 1.83156049 0.9969662
## oat-vetch-oat          0.79256158 -0.78231426 2.36743742 0.6891349
## rye-oat               -0.76551705 -2.34039289 0.80935879 0.7197914
## rye-clover-oat         0.74573845 -0.82913740 2.32061429 0.7415541
## rye-vetch-oat          0.62997632 -0.94489952 2.20485217 0.8537166
## oat-vetch-oat-clover   0.53587693 -1.03899891 2.11075277 0.9206888
## rye-oat-clover        -1.02220170 -2.59707754 0.55267414 0.4170885
## rye-clover-oat-clover  0.48905380 -1.08582204 2.06392964 0.9451661
## rye-vetch-oat-clover   0.37329168 -1.20158416 1.94816752 0.9828667
## rye-oat-vetch         -1.55807863 -3.13295447 0.01679721 0.0542372
## rye-clover-oat-vetch  -0.04682313 -1.62169897 1.52805271 0.9999993
## rye-vetch-oat-vetch   -0.16258525 -1.73746109 1.41229059 0.9996659
## rye-clover-rye         1.51125550 -0.06362034 3.08613134 0.0676969
## rye-vetch-rye          1.39549338 -0.17938247 2.97036922 0.1132782
## rye-vetch-rye-clover  -0.11576212 -1.69063796 1.45911372 0.9999373

Cover Crop Type (Pine)

# perform anova for pine 
anova_CMINpinCCtype <- aov(`mg CO2/g dry weight` ~ cc_type,  data = cminPIN)
summary(anova_CMINpinCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type      3   5.13   1.711   1.527  0.215
## Residuals   68  76.19   1.121

The ANOVA revealed that there was not a statistically significant difference in microbial respiration between cover crop type (of the 4 types) for the pine site.

Tillage Treatment

# perform anova 
anova_cminTIL <- aov(`mg CO2/g dry weight` ~ tillage,  data = cminTIL)
summary(anova_cminTIL)

##              Df Sum Sq Mean Sq F value Pr(>F)
## tillage       1    0.4  0.3533   0.139   0.71
## Residuals   178  453.8  2.5494

The ANOVA revealed that there was not a statistically significant difference in microbial respiration between tillage treatments (minimum tillage vs. no-till).

Two-Way ANOVAs

# Performing two-way ANOVA for respiration using Agroforestry Type & Tillage
anova_CMIN_agtl <- aov(`mg CO2/g dry weight` ~ agro_type * tillage, data = cmin_nocntrl)
summary(anova_CMIN_agtl)

##                    Df Sum Sq Mean Sq F value   Pr(>F)    
## agro_type           1   71.0   71.02  33.196 3.65e-08 ***
## tillage             1    0.4    0.35   0.165   0.6850    
## agro_type:tillage   1    6.3    6.26   2.928   0.0888 .  
## Residuals         176  376.5    2.14                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of agroforestry type and tillage.

# Performing two-way ANOVA for respiration using Cover Crop Type & Tillage
anova_CMIN_cctl <- aov(`mg CO2/g dry weight` ~ cc_type * tillage, data = cmin_nocntrl)
summary(anova_CMIN_cctl)

##                  Df Sum Sq Mean Sq F value   Pr(>F)    
## cc_type           9  108.1  12.009   5.947 3.68e-07 ***
## tillage           1    0.4   0.353   0.175    0.676    
## cc_type:tillage   9   22.6   2.510   1.243    0.273    
## Residuals       160  323.1   2.020                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of cover crop type and tillage.

Enzyme Activities

Significance Summary:

The following analyses were found to be significant:

Agroforestry Tree Type on BG, PHOS
Cover Crop Type - Combined Pecan & Pine on BG, PHOS
Cover Crop Type - Pecan on NAG, PHOS
Cover Crop Type - Pine on PHOS
Interaction between tree type and tillage for BG, NAG, PHOS
Interaction between cover crop type and tillage for BG, NAG, PHOS

The following analyses were not significant:

Agroforestry Tree Type on NAG
Cover Crop Type - Combined Pecan & Pine on NAG
Cover Crop Type - Pecan on BG
Cover Crop Type - Pine on BG, NAG
Tillage Treatment on BG, NAG, PHOS

Agroforestry Type

# perform anova for BG by agroforestry type
anova_ACT_BG_agrotype <- aov(BG ~ agro_type,  data = activity_nocntrl)
summary(anova_ACT_BG_agrotype)

##             Df  Sum Sq Mean Sq F value   Pr(>F)    
## agro_type    1 3675557 3675557   78.77 2.13e-12 ***
## Residuals   58 2706357   46661                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in activity of the enzyme BG between agroforestry tree type (pecan vs. pine) (p < 0.001).

# perform anova for NAG by agroforestry type
anova_ACT_NAG_agrotype <- aov(NAG ~ agro_type,  data = activity_nocntrl)
summary(anova_ACT_NAG_agrotype)

##             Df  Sum Sq Mean Sq F value Pr(>F)
## agro_type    1    7810    7810   0.374  0.543
## Residuals   58 1210224   20866

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme NAG between agroforestry tree type (pecan vs. pine).

# perform anova for PHOS by agroforestry type
anova_ACT_PHOS_agrotype <- aov(PHOS ~ agro_type,  data = activity_nocntrl)
summary(anova_ACT_PHOS_agrotype)

##             Df  Sum Sq Mean Sq F value   Pr(>F)    
## agro_type    1 5984926 5984926   40.77 3.13e-08 ***
## Residuals   58 8513787  146789                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in activity of the enzyme PHOS between agroforestry tree type (pecan vs. pine) (p < 0.001).

Visualizing all enzymes together on the same scale:

Cover Crop Type (Combined)

# perform anova for BG 
anova_ACT_BG_CCtype <- aov(BG ~ cc_type,  data = activity_nocntrl)
summary(anova_ACT_BG_CCtype)

##             Df  Sum Sq Mean Sq F value   Pr(>F)    
## cc_type      9 4096020  455113   9.955 1.48e-08 ***
## Residuals   50 2285893   45718                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in activity of the enzyme BG between cover crop type when considering both sites (pecan and pine) combined (p < 0.001).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = BG ~ cc_type, data = activity_nocntrl)
## 
## $cc_type
##                                diff         lwr         upr     p adj
## oat-clover-oat            87.792927  -320.85173  496.437589 0.9993178
## oat-vetch-oat             91.425614  -317.21905  500.070276 0.9990585
## radish-oat              -556.454702  -965.09936 -147.810041 0.0015050
## radish-oat-oat          -330.662999  -739.30766   77.981663 0.2104587
## radish-rye-oat          -332.644807  -741.28947   75.999854 0.2039012
## radish-vetch-oat        -393.654478  -802.29914   14.990183 0.0677776
## rye-oat                  224.351525  -184.29314  632.996187 0.7213328
## rye-clover-oat           176.494048  -232.15061  585.138709 0.9118013
## rye-vetch-oat             31.128892  -377.51577  439.773554 0.9999999
## oat-vetch-oat-clover       3.632687  -405.01197  412.277349 1.0000000
## radish-oat-clover       -644.247630 -1052.89229 -235.602968 0.0001413
## radish-oat-oat-clover   -418.455926  -827.10059   -9.811265 0.0407166
## radish-rye-oat-clover   -420.437734  -829.08240  -11.793073 0.0390394
## radish-vetch-oat-clover -481.447405  -890.09207  -72.802744 0.0098246
## rye-oat-clover           136.558598  -272.08606  545.203260 0.9818123
## rye-clover-oat-clover     88.701120  -319.94354  497.345782 0.9992594
## rye-vetch-oat-clover     -56.664035  -465.30870  351.980626 0.9999821
## radish-oat-vetch        -647.880317 -1056.52498 -239.235655 0.0001277
## radish-oat-oat-vetch    -422.088613  -830.73327  -13.443952 0.0376897
## radish-rye-oat-vetch    -424.070421  -832.71508  -15.425760 0.0361246
## radish-vetch-oat-vetch  -485.080092  -893.72475  -76.435431 0.0090079
## rye-oat-vetch            132.925911  -275.71875  541.570573 0.9848511
## rye-clover-oat-vetch      85.068433  -323.57623  493.713095 0.9994701
## rye-vetch-oat-vetch      -60.296722  -468.94138  348.347939 0.9999695
## radish-oat-radish        225.791703  -182.85296  634.436365 0.7141015
## radish-rye-radish        223.809895  -184.83477  632.454557 0.7240365
## radish-vetch-radish      162.800224  -245.84444  571.444886 0.9444764
## rye-radish               780.806228   372.16157 1189.450889 0.0000029
## rye-clover-radish        732.948750   324.30409 1141.593411 0.0000116
## rye-vetch-radish         587.583594   178.93893  996.228256 0.0006615
## radish-rye-radish-oat     -1.981808  -410.62647  406.662853 1.0000000
## radish-vetch-radish-oat  -62.991479  -471.63614  345.653182 0.9999559
## rye-radish-oat           555.014524   146.36986  963.659186 0.0015625
## rye-clover-radish-oat    507.157047    98.51239  915.801708 0.0052644
## rye-vetch-radish-oat     361.791891   -46.85277  770.436553 0.1240609
## radish-vetch-radish-rye  -61.009671  -469.65433  347.634990 0.9999663
## rye-radish-rye           556.996332   148.35167  965.640994 0.0014839
## rye-clover-radish-rye    509.138855   100.49419  917.783516 0.0050127
## rye-vetch-radish-rye     363.773699   -44.87096  772.418361 0.1196961
## rye-radish-vetch         618.006003   209.36134 1026.650665 0.0002907
## rye-clover-radish-vetch  570.148526   161.50386  978.793187 0.0010511
## rye-vetch-radish-vetch   424.783370    16.13871  833.428032 0.0355760
## rye-clover-rye           -47.857478  -456.50214  360.787184 0.9999958
## rye-vetch-rye           -193.222633  -601.86729  215.422028 0.8580500
## rye-vetch-rye-clover    -145.365155  -554.00982  263.279506 0.9725259

# perform anova for NAG 
anova_ACT_NAG_CCtype <- aov(NAG ~ cc_type,  data = activity_nocntrl)
summary(anova_ACT_NAG_CCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)  
## cc_type      9 295300   32811   1.778 0.0961 .
## Residuals   50 922734   18455                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme NAG between cover crop type when considering both sites (pecan and pine) combined.

# perform anova for PHOS
anova_ACT_PHOS_CCtype <- aov(PHOS ~ cc_type,  data = activity_nocntrl)
summary(anova_ACT_PHOS_CCtype)

##             Df   Sum Sq Mean Sq F value   Pr(>F)    
## cc_type      9 10555380 1172820   14.87 2.29e-11 ***
## Residuals   50  3943332   78867                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in activity of the enzyme PHOS between cover crop type when considering both sites (pecan and pine) combined (p < 0.001).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = PHOS ~ cc_type, data = activity_nocntrl)
## 
## $cc_type
##                                diff          lwr         upr     p adj
## oat-clover-oat            457.33191   -79.390259  994.054077 0.1575282
## oat-vetch-oat              40.94936  -495.772807  577.671530 0.9999999
## radish-oat                339.65369  -197.068478  876.375859 0.5395788
## radish-oat-oat           1104.51468   567.792512 1641.236849 0.0000005
## radish-rye-oat           1062.49458   525.772413 1599.216749 0.0000013
## radish-vetch-oat          898.14824   361.426074 1434.870410 0.0000467
## rye-oat                   574.68084    37.958667 1111.403004 0.0269435
## rye-clover-oat            295.26278  -241.459386  831.984951 0.7190649
## rye-vetch-oat            -129.12330  -665.845470  407.598867 0.9983380
## oat-vetch-oat-clover     -416.38255  -953.104715  120.339621 0.2596585
## radish-oat-clover        -117.67822  -654.400386  419.043950 0.9991978
## radish-oat-oat-clover     647.18277   110.460603 1183.904940 0.0074890
## radish-rye-oat-clover     605.16267    68.440504 1141.884840 0.0159591
## radish-vetch-oat-clover   440.81633   -95.905835  977.538501 0.1943207
## rye-oat-clover            117.34893  -419.373242  654.071095 0.9992155
## rye-clover-oat-clover    -162.06913  -698.791295  374.653042 0.9909802
## rye-vetch-oat-clover     -586.45521 -1123.177379  -49.733042 0.0220683
## radish-oat-vetch          298.70433  -238.017839  835.426497 0.7058304
## radish-oat-oat-vetch     1063.56532   526.843151 1600.287487 0.0000013
## radish-rye-oat-vetch     1021.54522   484.823051 1558.267387 0.0000032
## radish-vetch-oat-vetch    857.19888   320.476712 1393.921048 0.0001119
## rye-oat-vetch             533.73147    -2.990694 1070.453642 0.0524046
## rye-clover-oat-vetch      254.31342  -282.408747  791.035589 0.8565644
## rye-vetch-oat-vetch      -170.07266  -706.794831  366.649505 0.9873504
## radish-oat-radish         764.86099   228.138822 1301.583158 0.0007608
## radish-rye-radish         722.84089   186.118722 1259.563058 0.0017636
## radish-vetch-radish       558.49455    21.772383 1095.216719 0.0352425
## rye-radish                235.02714  -301.695023  771.749313 0.9048896
## rye-clover-radish         -44.39091  -581.113076  492.331260 0.9999998
## rye-vetch-radish         -468.77699 -1005.499160   67.945176 0.1353519
## radish-rye-radish-oat     -42.02010  -578.742268  494.702068 0.9999999
## radish-vetch-radish-oat  -206.36644  -743.088607  330.355730 0.9551738
## rye-radish-oat           -529.83384 -1066.556013    6.888323 0.0556903
## rye-clover-radish-oat    -809.25190 -1345.974066 -272.529730 0.0003059
## rye-vetch-radish-oat    -1233.63798 -1770.360150 -696.915814 0.0000000
## radish-vetch-radish-rye  -164.34634  -701.068507  372.375829 0.9900442
## rye-radish-rye           -487.81375 -1024.535913   48.908423 0.1040335
## rye-clover-radish-rye    -767.23180 -1303.953966 -230.509630 0.0007251
## rye-vetch-radish-rye    -1191.61788 -1728.340050 -654.895714 0.0000001
## rye-radish-vetch         -323.46741  -860.189575  213.254762 0.6063429
## rye-clover-radish-vetch  -602.88546 -1139.607628  -66.163291 0.0166086
## rye-vetch-radish-vetch  -1027.27154 -1563.993712 -490.549375 0.0000028
## rye-clover-rye           -279.41805  -816.140221  257.304115 0.7771198
## rye-vetch-rye            -703.80414 -1240.526305 -167.081969 0.0025602
## rye-vetch-rye-clover     -424.38608  -961.108252  112.336084 0.2368036

Visualizing all enzymes together on the same scale:

Cover Crop Type (Pecan)

# perform anova for BG for pecan 
anova_act_BG_pecCCtype <- aov(BG ~ cc_type,  data = act_agPEC)
summary(anova_act_BG_pecCCtype)

##             Df  Sum Sq Mean Sq F value Pr(>F)
## cc_type      5  217557   43511    0.76  0.586
## Residuals   30 1718148   57272

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme BG between cover crop type (of the 6 types) for the pecan site.

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = BG ~ cc_type, data = act_agPEC)
## 
## $cc_type
##                              diff       lwr      upr     p adj
## oat-clover-oat          87.792927 -332.4596 508.0455 0.9873361
## oat-vetch-oat           91.425614 -328.8269 511.6782 0.9848124
## rye-oat                224.351525 -195.9010 644.6041 0.5900028
## rye-clover-oat         176.494048 -243.7585 596.7466 0.7946582
## rye-vetch-oat           31.128892 -389.1237 451.3814 0.9999113
## oat-vetch-oat-clover     3.632687 -416.6199 423.8852 1.0000000
## rye-oat-clover         136.558598 -283.6939 556.8111 0.9180826
## rye-clover-oat-clover   88.701120 -331.5514 508.9537 0.9867362
## rye-vetch-oat-clover   -56.664035 -476.9166 363.5885 0.9983562
## rye-oat-vetch          132.925911 -287.3266 553.1785 0.9262587
## rye-clover-oat-vetch    85.068433 -335.1841 505.3210 0.9890166
## rye-vetch-oat-vetch    -60.296722 -480.5493 359.9558 0.9977887
## rye-clover-rye         -47.857478 -468.1100 372.3951 0.9992716
## rye-vetch-rye         -193.222633 -613.4752 227.0299 0.7275811
## rye-vetch-rye-clover  -145.365155 -565.6177 274.8874 0.8960663

# perform anova for NAG for pecan 
anova_act_NAG_pecCCtype <- aov(NAG ~ cc_type,  data = act_agPEC)
summary(anova_act_NAG_pecCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)  
## cc_type      5 247391   49478   3.159 0.0208 *
## Residuals   30 469951   15665                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in activity of the enzyme NAG between cover crop type (of the 6 types) for the pecan site (p < 0.05).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = NAG ~ cc_type, data = act_agPEC)
## 
## $cc_type
##                              diff         lwr       upr     p adj
## oat-clover-oat         117.352561 -102.436667 337.14179 0.5898448
## oat-vetch-oat          106.580054 -113.209173 326.36928 0.6821303
## rye-oat                196.852704  -22.936524 416.64193 0.0997692
## rye-clover-oat         214.708682   -5.080545 434.49791 0.0585667
## rye-vetch-oat            6.995784 -212.793444 226.78501 0.9999987
## oat-vetch-oat-clover   -10.772507 -230.561735 209.01672 0.9999885
## rye-oat-clover          79.500143 -140.289085 299.28937 0.8774162
## rye-clover-oat-clover   97.356121 -122.433107 317.14535 0.7567825
## rye-vetch-oat-clover  -110.356778 -330.146006 109.43245 0.6501483
## rye-oat-vetch           90.272650 -129.516578 310.06188 0.8091463
## rye-clover-oat-vetch   108.128628 -111.660600 327.91786 0.6690888
## rye-vetch-oat-vetch    -99.584271 -319.373499 120.20496 0.7393208
## rye-clover-rye          17.855978 -201.933249 237.64521 0.9998602
## rye-vetch-rye         -189.856921 -409.646148  29.93231 0.1216047
## rye-vetch-rye-clover  -207.712899 -427.502127  12.07633 0.0724811

# perform anova for PHOS for pecan 
anova_act_PHOS_pecCCtype <- aov(PHOS ~ cc_type,  data = act_agPEC)
summary(anova_act_PHOS_pecCCtype)

##             Df  Sum Sq Mean Sq F value   Pr(>F)    
## cc_type      5 2334269  466854   11.74 2.42e-06 ***
## Residuals   30 1193389   39780                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in activity of the enzyme PHOS between cover crop type (of the 6 types) for the pecan site (p < 0.05).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = PHOS ~ cc_type, data = act_agPEC)
## 
## $cc_type
##                             diff         lwr        upr     p adj
## oat-clover-oat         457.33191   107.08757  807.57625 0.0050386
## oat-vetch-oat           40.94936  -309.29498  391.19370 0.9991725
## rye-oat                574.68084   224.43650  924.92517 0.0003184
## rye-clover-oat         295.26278   -54.98156  645.50712 0.1378947
## rye-vetch-oat         -129.12330  -479.36764  221.12104 0.8686687
## oat-vetch-oat-clover  -416.38255  -766.62689  -66.13821 0.0125790
## rye-oat-clover         117.34893  -232.89541  467.59326 0.9078541
## rye-clover-oat-clover -162.06913  -512.31346  188.17521 0.7223435
## rye-vetch-oat-clover  -586.45521  -936.69955 -236.21087 0.0002399
## rye-oat-vetch          533.73147   183.48714  883.97581 0.0008475
## rye-clover-oat-vetch   254.31342   -95.93092  604.55776 0.2636466
## rye-vetch-oat-vetch   -170.07266  -520.31700  180.17167 0.6809070
## rye-clover-rye        -279.41805  -629.66239   70.82628 0.1793919
## rye-vetch-rye         -703.80414 -1054.04847 -353.55980 0.0000142
## rye-vetch-rye-clover  -424.38608  -774.63042  -74.14175 0.0105509

Visualizing all enzymes together on the same scale:

Cover Crop Type (Pine)

# perform anova for BG for pine 
anova_act_BG_pinCCtype <- aov(BG ~ cc_type,  data = act_agPIN)
summary(anova_act_BG_pinCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)  
## cc_type      3 202906   67635   2.383 0.0998 .
## Residuals   20 567745   28387                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme BG between cover crop type (of the 6 types) for the pine site.

# perform anova for NAG for pine 
anova_act_NAG_pinCCtype <- aov(NAG ~ cc_type,  data = act_agPIN)
summary(anova_act_NAG_pinCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type      3  40099   13366    0.59  0.628
## Residuals   20 452783   22639

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme NAG between cover crop type (of the 6 types) for the pine site.

# perform anova for PHOS for pine 
anova_act_PHOS_pinCCtype <- aov(PHOS ~ cc_type,  data = act_agPIN)
summary(anova_act_PHOS_pinCCtype)

##             Df  Sum Sq Mean Sq F value Pr(>F)   
## cc_type      3 2236185  745395   5.421 0.0068 **
## Residuals   20 2749943  137497                  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in activity of the enzyme PHOS between cover crop type (of the 6 types) for the pine site (p < 0.01).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = PHOS ~ cc_type, data = act_agPIN)
## 
## $cc_type
##                              diff        lwr       upr     p adj
## radish-oat-radish        764.8610  165.65077 1364.0712 0.0094704
## radish-rye-radish        722.8409  123.63067 1322.0511 0.0146260
## radish-vetch-radish      558.4946  -40.71567 1157.7048 0.0732424
## radish-rye-radish-oat    -42.0201 -641.23032  557.1901 0.9972300
## radish-vetch-radish-oat -206.3664 -805.57666  392.8438 0.7710296
## radish-vetch-radish-rye -164.3463 -763.55656  434.8639 0.8680051

Visualizing all enzymes together on the same scale:

Tillage Treatment

# perform anova for BG
anova_act_BG_TIL <- aov(BG ~ tillage,  data = act_agTIL)
summary(anova_act_BG_TIL)

##             Df  Sum Sq Mean Sq F value Pr(>F)
## tillage      1   58738   58738   0.539  0.466
## Residuals   58 6323175  109020

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme BG between tillage treatments (minimum tillage vs. no-till).

# perform anova for NAG
anova_act_NAG_TIL <- aov(NAG ~ tillage,  data = act_agTIL)
summary(anova_act_NAG_TIL)

##             Df  Sum Sq Mean Sq F value Pr(>F)
## tillage      1    3749    3749   0.179  0.674
## Residuals   58 1214285   20936

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme NAG between tillage treatments (minimum tillage vs. no-till).

# perform anova for PHOS
anova_act_PHOS_TIL <- aov(PHOS ~ tillage,  data = act_agTIL)
summary(anova_act_PHOS_TIL)

##             Df   Sum Sq Mean Sq F value Pr(>F)
## tillage      1    23466   23466   0.094   0.76
## Residuals   58 14475246  249573

The ANOVA revealed that there was not a statistically significant difference in activity of the enzyme PHOS between tillage treatments (minimum tillage vs. no-till).

Visualizing all enzymes together on the same scale:

Two-Way ANOVAs

# Performing two-way ANOVA for BG activity using Agroforestry Type & Tillage
anova_ACT_BG_agtl <- aov(BG ~ agro_type * tillage, data = activity_nocntrl)
summary(anova_ACT_BG_agtl)

##                   Df  Sum Sq Mean Sq F value   Pr(>F)    
## agro_type          1 3675557 3675557 111.898 5.73e-15 ***
## tillage            1   58738   58738   1.788    0.187    
## agro_type:tillage  1  808159  808159  24.603 6.91e-06 ***
## Residuals         56 1839460   32847                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

BG: A two-way ANOVA showed that there was a statistically significant interaction between the effects of agroforestry type and tillage (p < 0.001).

# Performing two-way ANOVA for BG activity using Cover Crop Type & Tillage
anova_ACT_BG_cctl <- aov(BG ~ cc_type * tillage, data = activity_nocntrl)
summary(anova_ACT_BG_cctl)

##                 Df  Sum Sq Mean Sq F value   Pr(>F)    
## cc_type          9 4096020  455113  38.392  < 2e-16 ***
## tillage          1   58738   58738   4.955   0.0317 *  
## cc_type:tillage  9 1752974  194775  16.430 7.38e-11 ***
## Residuals       40  474181   11855                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

BG: A two-way ANOVA showed that there was a statistically significant interaction between the effects of cover crop type and tillage (p < 0.001).

# Performing two-way ANOVA for NAG activity using Agroforestry Type & Tillage
anova_ACT_NAG_agtl <- aov(NAG ~ agro_type * tillage, data = activity_nocntrl)
summary(anova_ACT_NAG_agtl)

##                   Df Sum Sq Mean Sq F value   Pr(>F)    
## agro_type          1   7810    7810   0.453 0.503643    
## tillage            1   3749    3749   0.218 0.642752    
## agro_type:tillage  1 241210  241210  13.994 0.000433 ***
## Residuals         56 965265   17237                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

NAG: A two-way ANOVA showed that there was a statistically significant interaction between the effects of agroforestry type and tillage (p < 0.001).

# Performing two-way ANOVA for NAG activity using Cover Crop Type & Tillage
anova_ACT_NAG_cctl <- aov(NAG ~ cc_type * tillage, data = activity_nocntrl)
summary(anova_ACT_NAG_cctl)

##                 Df Sum Sq Mean Sq F value   Pr(>F)    
## cc_type          9 295300   32811   8.133 9.78e-07 ***
## tillage          1   3749    3749   0.929    0.341    
## cc_type:tillage  9 757617   84180  20.866 1.82e-12 ***
## Residuals       40 161368    4034                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

NAG: A two-way ANOVA showed that there was a statistically significant interaction between the effects of cover crop type and tillage (p < 0.001).

# Performing two-way ANOVA for PHOS activity using Agroforestry Type & Tillage
anova_ACT_PHOS_agtl <- aov(PHOS ~ agro_type * tillage, data = activity_nocntrl)
summary(anova_ACT_PHOS_agtl)

##                   Df  Sum Sq Mean Sq F value   Pr(>F)    
## agro_type          1 5984926 5984926  44.613 1.17e-08 ***
## tillage            1   23466   23466   0.175  0.67737    
## agro_type:tillage  1  977860  977860   7.289  0.00916 ** 
## Residuals         56 7512461  134151                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

PHOS: A two-way ANOVA showed that there was a statistically significant interaction between the effects of agroforestry type and tillage (p < 0.01).

# Performing two-way ANOVA for PHOS activity using Cover Crop Type & Tillage
anova_ACT_PHOS_cctl <- aov(PHOS ~ cc_type * tillage, data = activity_nocntrl)
summary(anova_ACT_PHOS_cctl)

##                 Df   Sum Sq Mean Sq F value   Pr(>F)    
## cc_type          9 10555380 1172820  40.944  < 2e-16 ***
## tillage          1    23466   23466   0.819    0.371    
## cc_type:tillage  9  2774085  308232  10.761 2.95e-08 ***
## Residuals       40  1145781   28645                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

PHOS: A two-way ANOVA showed that there was a statistically significant interaction between the effects of cover crop type and tillage (p < 0.001).

Necromass

Significance Summary:

The following analyses were found to be significant:

Cover Crop Type - Pecan (for Total, Fungal, and Bacterial)

The following analyses were not significant:

Agroforestry Type (for Total, Fungal, and Bacterial)
Cover Crop Type - Combined Pecan & Pine (for Total, Fungal, and Bacterial)
Cover Crop Type - Pine (for Total, Fungal, and Bacterial)
Tillage (for Total, Fungal, and Bacterial)
No interaction between tree type and tillage (for Total, Fungal, and Bacterial)

Agroforestry Type

Total Necromass

# perform anova for total necromass
anova_NECagrotype <- aov(MicrobialResidualC_mg_g ~ agro_type,  data = nec_nocntrl)
summary(anova_NECagrotype)

##             Df  Sum Sq  Mean Sq F value Pr(>F)
## agro_type    1 0.00434 0.004336   0.319  0.579
## Residuals   18 0.24457 0.013587

The ANOVA revealed that there was not a statistically significant difference in total necromass between agroforestry tree type (pecan vs. pine).

Fungal Necromass

# perform anova for fungal necromass
anova_NECFagrotype <- aov(FungalC_mg_g ~ agro_type,  data = nec_nocntrl)
summary(anova_NECFagrotype)

##             Df  Sum Sq  Mean Sq F value Pr(>F)
## agro_type    1 0.00001 0.000011   0.002  0.969
## Residuals   18 0.12798 0.007110

The ANOVA revealed that there was not a statistically significant difference in fungal necromass between agroforestry tree type (pecan vs. pine).

Bacterial Necromass

# perform anova for bacterial necromass
anova_NECBagrotype <- aov(BacterialC_mg_g ~ agro_type,  data = nec_nocntrl)
summary(anova_NECBagrotype)

##             Df   Sum Sq  Mean Sq F value Pr(>F)  
## agro_type    1 0.003907 0.003907   3.143 0.0932 .
## Residuals   18 0.022380 0.001243                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was not a statistically significant difference in bacterial necromass between agroforestry tree type (pecan vs. pine).

Cover Crop Type (Combined)

Total Necromass

# perform anova for combined
anova_NECCCtype <- aov(MicrobialResidualC_mg_g ~ cc_type,  data = necdata)
summary(anova_NECCCtype)

##             Df Sum Sq Mean Sq F value Pr(>F)
## cc_type     10 0.1914 0.01914   0.996  0.499
## Residuals   11 0.2114 0.01921

The ANOVA revealed that there was not a statistically significant difference in total necromass between cover crop type when considering both sites combined.

Fungal Necromass

# perform anova for combined
anova_NECFCCtype <- aov(FungalC_mg_g ~ cc_type,  data = necdata)
summary(anova_NECFCCtype)

##             Df  Sum Sq  Mean Sq F value Pr(>F)
## cc_type     10 0.09917 0.009917   0.873  0.581
## Residuals   11 0.12502 0.011365

The ANOVA revealed that there was not a statistically significant difference in fungal necromass between cover crop type when considering both sites combined.

Bacterial Necromass

# perform anova for combined
anova_NECBCCtype <- aov(BacterialC_mg_g ~ cc_type,  data = necdata)
summary(anova_NECBCCtype)

##             Df  Sum Sq  Mean Sq F value Pr(>F)
## cc_type     10 0.02008 0.002008   1.701  0.198
## Residuals   11 0.01299 0.001181

The ANOVA revealed that there was not a statistically significant difference in bacterial necromass between cover crop type when considering both sites combined.

Cover Crop Type (Pecan)

Total Necromass

# perform anova for pecan
anova_NECpecCCtype <- aov(MicrobialResidualC_mg_g ~ cc_type,  data = necPEC)
summary(anova_NECpecCCtype)

##             Df  Sum Sq  Mean Sq F value Pr(>F)  
## cc_type      5 0.11489 0.022978   5.614  0.029 *
## Residuals    6 0.02456 0.004093                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in total necromass between cover crop type (of the 6 types) for the pecan site (p < 0.05).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = MicrobialResidualC_mg_g ~ cc_type, data = necPEC)
## 
## $cc_type
##                              diff          lwr         upr     p adj
## oat-clover-oat         0.26083978  0.006230636  0.51544892 0.0451230
## oat-vetch-oat          0.08341985 -0.171189287  0.33802899 0.7753391
## rye-oat               -0.05465415 -0.309263291  0.19995499 0.9449987
## rye-clover-oat         0.08214542 -0.172463717  0.33675456 0.7848482
## rye-vetch-oat          0.09494528 -0.159663859  0.34955442 0.6854201
## oat-vetch-oat-clover  -0.17741992 -0.432029063  0.07718922 0.1889825
## rye-oat-clover        -0.31549393 -0.570103067 -0.06088479 0.0191415
## rye-clover-oat-clover -0.17869435 -0.433303493  0.07591479 0.1848127
## rye-vetch-oat-clover  -0.16589449 -0.420503635  0.08871465 0.2310308
## rye-oat-vetch         -0.13807400 -0.392683145  0.11653514 0.3696387
## rye-clover-oat-vetch  -0.00127443 -0.255883571  0.25333471 1.0000000
## rye-vetch-oat-vetch    0.01152543 -0.243083712  0.26613457 0.9999568
## rye-clover-rye         0.13679957 -0.117809567  0.39140871 0.3773608
## rye-vetch-rye          0.14959943 -0.105009708  0.40420857 0.3053704
## rye-vetch-rye-clover   0.01279986 -0.241809282  0.26740900 0.9999276

Fungal Necromass

# perform anova for pecan
anova_NECFpecCCtype <- aov(FungalC_mg_g ~ cc_type,  data = necPEC)
summary(anova_NECFpecCCtype)

##             Df  Sum Sq  Mean Sq F value Pr(>F)  
## cc_type      5 0.06035 0.012071   5.152 0.0352 *
## Residuals    6 0.01406 0.002343                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in fungal necromass between cover crop type (of the 6 types) for the pecan site (p < 0.05).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = FungalC_mg_g ~ cc_type, data = necPEC)
## 
## $cc_type
##                               diff        lwr         upr     p adj
## oat-clover-oat         0.196318486  0.0036745  0.38896247 0.0461523
## oat-vetch-oat          0.068408469 -0.1242355  0.26105246 0.7214169
## rye-oat               -0.027873481 -0.2205175  0.16477051 0.9891946
## rye-clover-oat         0.074926009 -0.1177180  0.26757000 0.6527267
## rye-vetch-oat          0.059594166 -0.1330498  0.25223815 0.8094600
## oat-vetch-oat-clover  -0.127910017 -0.3205540  0.06473397 0.2186983
## rye-oat-clover        -0.224191967 -0.4168360 -0.03154798 0.0256357
## rye-clover-oat-clover -0.121392477 -0.3140365  0.07125151 0.2538653
## rye-vetch-oat-clover  -0.136724320 -0.3293683  0.05591967 0.1784237
## rye-oat-vetch         -0.096281950 -0.2889259  0.09636204 0.4390426
## rye-clover-oat-vetch   0.006517540 -0.1861264  0.19916153 0.9999898
## rye-vetch-oat-vetch   -0.008814303 -0.2014583  0.18382968 0.9999545
## rye-clover-rye         0.102799490 -0.0898445  0.29544348 0.3830977
## rye-vetch-rye          0.087467647 -0.1051763  0.28011163 0.5227219
## rye-vetch-rye-clover  -0.015331843 -0.2079758  0.17731214 0.9993222

Bacterial Necromass

# perform anova for pecan
anova_NECBpecCCtype <- aov(BacterialC_mg_g ~ cc_type,  data = necPEC)
summary(anova_NECBpecCCtype)

##             Df   Sum Sq   Mean Sq F value Pr(>F)  
## cc_type      5 0.009786 0.0019572   5.342 0.0325 *
## Residuals    6 0.002198 0.0003664                 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

The ANOVA revealed that there was a statistically significant difference in bacterial necromass between cover crop type (of the 6 types) for the pecan site (p < 0.05).

Results of Tukey HSD test and visualization:

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = BacterialC_mg_g ~ cc_type, data = necPEC)
## 
## $cc_type
##                               diff         lwr         upr     p adj
## oat-clover-oat         0.064521290 -0.01165648  0.14069906 0.0968423
## oat-vetch-oat          0.015011384 -0.06116638  0.09118915 0.9605475
## rye-oat               -0.026780670 -0.10295844  0.04939710 0.7285166
## rye-clover-oat         0.007219415 -0.06895835  0.08339718 0.9984424
## rye-vetch-oat          0.035351116 -0.04082665  0.11152888 0.5036867
## oat-vetch-oat-clover  -0.049509906 -0.12568767  0.02666786 0.2327084
## rye-oat-clover        -0.091301960 -0.16747973 -0.01512419 0.0223758
## rye-clover-oat-clover -0.057301876 -0.13347964  0.01887589 0.1475845
## rye-vetch-oat-clover  -0.029170174 -0.10534794  0.04700759 0.6650019
## rye-oat-vetch         -0.041792054 -0.11796982  0.03438571 0.3600720
## rye-clover-oat-vetch  -0.007791970 -0.08396974  0.06838580 0.9977700
## rye-vetch-oat-vetch    0.020339732 -0.05583804  0.09651750 0.8800008
## rye-clover-rye         0.034000084 -0.04217768  0.11017785 0.5376017
## rye-vetch-rye          0.062131786 -0.01404598  0.13830955 0.1112640
## rye-vetch-rye-clover   0.028131701 -0.04804607  0.10430947 0.6927701

Cover Crop Type (Pine)

Total Necromass

# perform anova for pine
anova_NECpinCCtype <- aov(MicrobialResidualC_mg_g ~ cc_type,  data = necPIN)
summary(anova_NECpinCCtype)

##             Df  Sum Sq Mean Sq F value Pr(>F)
## cc_type      3 0.04128 0.01376   0.862   0.53
## Residuals    4 0.06385 0.01596

The ANOVA revealed that there was not a statistically significant difference in total necromass between cover crop type (of the 4 types) for the pine site.

Fungal Necromass

# perform anova for pine
anova_NECFpinCCtype <- aov(FungalC_mg_g ~ cc_type,  data = necPIN)
summary(anova_NECFpinCCtype)

##             Df  Sum Sq  Mean Sq F value Pr(>F)
## cc_type      3 0.01781 0.005937   0.664  0.616
## Residuals    4 0.03575 0.008938

The ANOVA revealed that there was not a statistically significant difference in fungal necromass between cover crop type (of the 4 types) for the pine site.

Bacterial Necromass

# perform anova for pine
anova_NECBpinCCtype <- aov(BacterialC_mg_g ~ cc_type,  data = necPIN)
summary(anova_NECBpinCCtype)

##             Df   Sum Sq  Mean Sq F value Pr(>F)
## cc_type      3 0.005444 0.001814   1.466   0.35
## Residuals    4 0.004952 0.001238

The ANOVA revealed that there was not a statistically significant difference in bacterial necromass between cover crop type (of the 4 types) for the pine site.

Tillage Treatment

Total Necromass

# perform anova
anova_necTIL <- aov(MicrobialResidualC_mg_g ~ tillage,  data = necTIL)
summary(anova_necTIL)

##             Df  Sum Sq Mean Sq F value Pr(>F)
## tillage      1 0.00052 0.00052   0.038  0.848
## Residuals   18 0.24839 0.01380

The ANOVA revealed that there was not a statistically significant difference in total necromass between tillage treatments (minimum tillage vs. no-till).

Fungal Necromass

# perform anova
anova_necFTIL <- aov(FungalC_mg_g ~ tillage,  data = necTIL)
summary(anova_necFTIL)

##             Df  Sum Sq  Mean Sq F value Pr(>F)
## tillage      1 0.00033 0.000332   0.047  0.831
## Residuals   18 0.12766 0.007092

The ANOVA revealed that there was not a statistically significant difference in total necromass between tillage treatments (minimum tillage vs. no-till).

Bacterial Necromass

# perform anova
anova_necBTIL <- aov(BacterialC_mg_g ~ tillage,  data = necTIL)
summary(anova_necBTIL)

##             Df   Sum Sq  Mean Sq F value Pr(>F)
## tillage      1 0.000021 0.000021   0.014  0.906
## Residuals   18 0.026266 0.001459

The ANOVA revealed that there was not a statistically significant difference in bacterial necromass between tillage treatments (minimum tillage vs. no-till).

Two-Way ANOVAs

Total Necromass

# Performing two-way ANOVA for total necromass using Agroforestry Type & Tillage
anova_NECt_agtl <- aov(MicrobialResidualC_mg_g ~ agro_type * tillage, data = nec_nocntrl)
summary(anova_NECt_agtl)

##                   Df  Sum Sq  Mean Sq F value Pr(>F)
## agro_type          1 0.00434 0.004336   0.288  0.599
## tillage            1 0.00052 0.000520   0.034  0.855
## agro_type:tillage  1 0.00285 0.002851   0.189  0.669
## Residuals         16 0.24120 0.015075

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of agroforestry type and tillage.

# Performing two-way ANOVA for total necromass using Cover Crop Type & Tillage
anova_NECt_cctl <- aov(MicrobialResidualC_mg_g ~ cc_type * tillage, data = nec_nocntrl)
summary(anova_NECt_cctl)

##                 Df  Sum Sq  Mean Sq
## cc_type          9 0.16050 0.017833
## tillage          1 0.00052 0.000520
## cc_type:tillage  9 0.08788 0.009765

A two-way ANOVA was not able to be performed for cover crop type and tillage, likely due to a lack of replication / variability (only one season of data, so only 1 observation for each cover crop / tillage combination).

Fungal Necromass

# Performing two-way ANOVA for fungal necromass using Agroforestry Type & Tillage
anova_NECf_agtl <- aov(FungalC_mg_g ~ agro_type * tillage, data = nec_nocntrl)
summary(anova_NECf_agtl)

##                   Df  Sum Sq  Mean Sq F value Pr(>F)
## agro_type          1 0.00001 0.000011   0.001  0.971
## tillage            1 0.00033 0.000332   0.042  0.841
## agro_type:tillage  1 0.00033 0.000329   0.041  0.841
## Residuals         16 0.12732 0.007957

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of agroforestry type and tillage.

# Performing two-way ANOVA for fungal necromass using Cover Crop Type & Tillage
anova_NECf_cctl <- aov(FungalC_mg_g ~ cc_type * tillage, data = nec_nocntrl)
summary(anova_NECf_cctl)

##                 Df  Sum Sq  Mean Sq
## cc_type          9 0.07818 0.008686
## tillage          1 0.00033 0.000332
## cc_type:tillage  9 0.04948 0.005498

Bacterial Necromass

# Performing two-way ANOVA for bacterial necromass using Agroforestry Type & Tillage
anova_NECb_agtl <- aov(BacterialC_mg_g ~ agro_type * tillage, data = nec_nocntrl)
summary(anova_NECb_agtl)

##                   Df   Sum Sq  Mean Sq F value Pr(>F)
## agro_type          1 0.003907 0.003907   2.961  0.105
## tillage            1 0.000021 0.000021   0.016  0.901
## agro_type:tillage  1 0.001243 0.001243   0.942  0.346
## Residuals         16 0.021116 0.001320

A two-way ANOVA showed that there was not a statistically significant interaction between the effects of agroforestry type and tillage.

# Performing two-way ANOVA for bacterial necromass using Cover Crop Type & Tillage
anova_NECb_cctl <- aov(BacterialC_mg_g ~ cc_type * tillage, data = nec_nocntrl)
summary(anova_NECb_cctl)

##                 Df   Sum Sq   Mean Sq
## cc_type          9 0.019137 0.0021263
## tillage          1 0.000021 0.0000210
## cc_type:tillage  9 0.007129 0.0007922

Summary Statistics

Corrplot

This first correlation matrix using corrplot( ) uses data from only the third season (because we only have necromass data for the third season, so otherwise it would have to be removed):

This next correlation matrix using corrplot( ) uses data from all seasons (so necromass has been removed):

Scatter Plots

This first scatter plot uses the function chart.Correlation( ) with data from all 3 seasons, without necromass being removed (this means that there are many NAs in the data from the first 2 seasons, but the function still seems to work):

This next scatter plot uses the function chart.Correlation( ) of all parameters with NAs removed, so effectively just for the third season (necromass would show up as NA before that):

This final scatter plot uses the function chart.Correlation( ) with data from only the parameters that we have data all three seasons for (so necromass has been removed):

Thesis Data: Agroforestry v3.0

Murray Sternberg

2024-03-07

Microbial Biomass C

Agroforestry Type

Cover Crop Type (Combined)

Cover Crop Type (Pecan)

Cover Crop Type (Pine)

Tillage Treatment

Two-Way ANOVAs

Microbial Biomass N

Agroforestry Type

Cover Crop Type (Combined)

Cover Crop Type (Pecan)

Cover Crop Type (Pine)

Tillage Treatment

Two-Way ANOVAs

24-Hour Microbial Respiration

Agroforestry Type

Cover Crop Type (Combined)

Cover Crop Type (Pecan)

Cover Crop Type (Pine)

Tillage Treatment

Two-Way ANOVAs

Enzyme Activities

Agroforestry Type

Cover Crop Type (Combined)

Cover Crop Type (Pecan)

Cover Crop Type (Pine)

Tillage Treatment

Two-Way ANOVAs

Necromass

Agroforestry Type

Total Necromass

Fungal Necromass

Bacterial Necromass

Cover Crop Type (Combined)

Total Necromass

Fungal Necromass

Bacterial Necromass

Cover Crop Type (Pecan)

Total Necromass

Fungal Necromass

Bacterial Necromass

Cover Crop Type (Pine)

Total Necromass

Fungal Necromass

Bacterial Necromass

Tillage Treatment

Total Necromass

Fungal Necromass

Bacterial Necromass

Two-Way ANOVAs

Total Necromass

Fungal Necromass

Bacterial Necromass

Summary Statistics

Corrplot

Scatter Plots