Embedded Files
The data for this project were collected by Dr. Anne McIntosh and her team in 2013 in Dried Mixed Grassland sites of Southern Alberta near Brooks and Taber, AB (Figure 1). The analysis will focus on the grasslands data, specifically on lichen, percent cover of vascular plants, and soil parameters data in individual quadrats, with permission from Dr. Anne McIntosh and under the supervision of Dr. Diane Haughland.
Figure 1. Map of all the sample locations near Brooks and Taber, AB.
18 paired wellpad and reference sites were originally sampled. Many of the wellpad sites were not active, having been drilled, abandoned, and plugged. There is a gap for some sites between abandonment and certification; therefore, the ages used to define the sites are the years they were reclaimed. The sites were grouped into three different age classes of 10, 20, and 30 years by rounding the year reclaimed to the closest age class. All sampling was conducted according to the Alberta Biodiversity Monitoring Institute’s Field Data Collection Protocols for Native Grasslands (2014) as shown in Figure 2.
Figure 2 from ABMI Field Data Collection Protocols for Native Grasslands (2014). Sampling layout for field data collection on native grassland wellpads and reference sites using nested square plots, where the wellsite is in the center, and reference plots are adjacent to the wellsites. The wellsite plot was divided into 4 quadrats. 4 points directly 35 m from the wellsite centre were identified, and around these points were placed pigtails for a 5x5 (green), 10x10 m (pink), and 25x25 m (blue) plots. 4 Reference quadrats were delineated 35 m from the edge of the wellsite, with a similar nested structure.
A community survey was conducted for lichen in the wellpad and reference sites (Table 1). For the vascular plants, percent cover was estimated in all of the systematically located quadrats of each site (Table 2). Percent cover for vascular plants was thus recorded at five wellpads and four reference quadrats per site. Soil parameter data includes measurements on pH, electrical conductivity, carbon and nitrogen content, and bulk density at different soil depths in all of the systematically located quadrats of each site (Tables 3 and 4).
Table 1. Snapshot of the original lichen dataset. The first two columns have to do with the date the sample was collected and by whom. The following columns have to do with the species found and any ID notes. The following columns are the Site ID number, Age, and Sample Location of the site where that species was found.
Table 2. Snapshot of the original vascular plant percent cover dataset. The first column has the sample number, following the Site ID number in the next column. Species_ID includes the abbreviated/short form name of the species found in that Site. The following columns describe the percent cover of each species in 4 parts of each quadrat, and the strata it was found on, where different letters represent different kinds of substrate. Quadrants A-E were on wellsites, while F-I were on reference sites.
Table 3. Snapshot of soil parameter data. Soil parameters were measured for depths of 0-15 cm (0), 15-30 cm (15), 30-60 cm (30), and 60-100 cm (60) for each quadrant. The Sample ID was formed as follows: W=Wellsite/R=Reference - Quadrant (A-I) - Soil depth (0, 15, 30, or 60). Site_ID, Depth, and Wellpad or Reference Site are also explicitly stated in the following 3 columns. Measurements for pH, electrical conductivity (microS/cm), total nitrogen (%), and total organic carbon (%) in each quadrat are listed in the rest of the columns.
Table 4. Snapshot of soil bulk density data. Bulk density was measured for depths of 0-15 cm (0) and 15-30 cm (15) for multiple sub-samples within each quadrat. The Sample_ID is formed as follows: W=Wellsite/R=Reference - Quadrant (A-I) - Location within quadrant ( 1-5) - Soil depth (0 or 15). StudyArea_ID is DMG for Dried Mixed Grasslands. ID is the sample number. Site_ID is for the site number. Bulk density, depth (cm), and whether it was a Wellpad or Reference site follows these.
The data analysis will use Cluster Analysis, PCoA, NMDS, perMANOVA, and other univariate methods on R 4.5.2 GUI 1.82 High Sierra build (8556). A PCA and Cluster analysis with a scaled version of my data (Hellinger Transformation) combined into one spreadsheet will allow me to look at general recovery trajectories and relationships amongst all my data points. PCoA and Jaccard Distances, focusing solely on my lichen species matrix, will allow me to analyze the differences among age classes and sample locations for lichen communities. NMDS and Bray-Curtis distances with the vegan package will help me to look at the relationships between age classes and sample locations for graminoid and forb plant cover. A perMANOVA and pairwise perMANOVA tests using the vegan package will help me to test for the statistical differences between sample locations and age classes, and the lichen, graminoid/forb, and soil data. Finally, a univariate ANOVA, Linear Model, and Mixed Linear Model will quantify these relationships, allowing me to test the correlations between different species of graminoids/forbs and lichen; between these communities’ recovery and different age classes or sites; and between the different clusters of my cluster analysis.
Page updated
Report abuse