**3. Statistical analysis**

The resulting datasets from FAMEs analysis on the microbial groups were analyzed using generalized linear mixed models procedure (GLIMMIX Procedure) in SAS® 9.4 software package and means separated at p < 0.05. This analysis involved testing sampling methods (ECa-directed vs. transect-based), the effect of sampling date/time as well as any interactions between sampling method and time. To further investigate any impacts of plant (brome grass monoculture) as well as the effect of soil physical and chemical parameters on microbial diversity and abundance, principal component analysis (PCA) was conducted. The PCA was performed to determine the contribution of soil characteristics (e.g., ECa, pH) in the variation and separation of both temporal and sampling method. Soil chemical, physical and biological attributes were used in the PCA to elucidate the effects of

**177**

*Spatio-Temporal Dynamics of Soil Microbial Communities in a Pasture: A Case Study…*

sampling technique and timing of sampling. This was conducted using R programming software version 3.4.1, utilizing ade4 package [44] and factoextra package for

In addition, the relationship between soil microbes and soil physicochemical characteristics was visualized using heatmap. The heatmap was generated using a combination of packages and their functions in the R programming language. The hclust function and scale both available in stats package were utilized, respectively, for hierarchical cluster analysis after applying the scale function to centralize the various data about the mean and to generate z-scores around each variable's standard deviation about the mean. Cluster analysis was performed using the default 'complete' agglomeration method. The melt function in the reshape2 package was used to organize the dataset before plotting the heatmap using ggplot function acquired from the ggplot2 package. The z-score legend was generated using scale\_ fill\_gradient2 function in ggplot2 package and color breaks represented z-scores

The physical and chemical characteristics of soil in the brome grass pasture are summarized in **Table 1**. Soil organic matter contents were relatively high, averaging 4.14%. High organic matter content in the top 20 cm of the pasture soil can be attributed to dense rhizomatous roots of brome grass root biomass and shoots [32]. Soil nutrient availability to crops is influenced by soil pH. The pasture site exhibited

(Nitrates-N ppm) of soils sampled from the pasture site averaged 4.42, which can be attributed to the cow dung manure, N fertilizer, as well as the high biomass from the brome grass shoots and roots. 1:1 soil soluble salts (mmho cm−1) was strongly and

uted significantly to the total sum of cations me 100 g−1 (CEC) of the pasture site. Values for ECa ranged from 21 to about 44 mS m−1 (**Figure 1**), which indicated a low to moderate level of spatial site heterogeneity. The mean ECa value was 32 mS m−1 with ECa of 30.1–35 and 25.1–30 mS m−1 being more common (the two combined covered nearly the entire the pasture). Regions with lowest and highest ECa values mainly constituted small pockets that were randomly distributed across the pasture with no clear pattern that could be discerned (**Figure 1**). As a result, soil in this pasture was considered to be less variable and the ECa values fell within normal range of 0–150 mS m−1 for grass pasture [35] but lower than in a fertilized maize

General composition of microbial communities, namely, total microbial biomass,

diversity, and composition of soil microbes detected using FAMEs assay in soils sample collected over two seasons and methods is presented in **Figures 2** and **3**. Each sector of the pie chart (**Figure 3**) represents individual microbial composition as a percentage of total recovered microbial fatty acid. Microbial biomass was dominated by bacteria (55–65%), followed in declining order by arbuscular mycorrhizae (15–25%) saprophytic fungi (8–9%) actinomycetes (8%) micro-eukaryotes (4%).

0.89, p < 0.05) with

0.79, p < 0.05). Available

) cations contrib-

0.95, p < 0.05). Available N

a slightly acidic soil pH (5.98) and was positively correlated (R2

positively correlated to measured available soil nitrates (R2

P in the pasture site was measured at 18.07 ppm P. Hydrogen (H+

calcium but negatively correlated to percent H+ (R2

field [46] within Eastern Nebraska.

**4.2 Soil microbial community**

*DOI: http://dx.doi.org/10.5772/intechopen.93548*

ranging from −3 to 3 over the entire dataset.

visualization purposes [45].

**4. Results and discussion**

**4.1 Site characteristics**

*Spatio-Temporal Dynamics of Soil Microbial Communities in a Pasture: A Case Study… DOI: http://dx.doi.org/10.5772/intechopen.93548*

sampling technique and timing of sampling. This was conducted using R programming software version 3.4.1, utilizing ade4 package [44] and factoextra package for visualization purposes [45].

In addition, the relationship between soil microbes and soil physicochemical characteristics was visualized using heatmap. The heatmap was generated using a combination of packages and their functions in the R programming language. The hclust function and scale both available in stats package were utilized, respectively, for hierarchical cluster analysis after applying the scale function to centralize the various data about the mean and to generate z-scores around each variable's standard deviation about the mean. Cluster analysis was performed using the default 'complete' agglomeration method. The melt function in the reshape2 package was used to organize the dataset before plotting the heatmap using ggplot function acquired from the ggplot2 package. The z-score legend was generated using scale\_ fill\_gradient2 function in ggplot2 package and color breaks represented z-scores ranging from −3 to 3 over the entire dataset.
