Data

Data was compiled into a standard data table, a simplified version is shown in table 1.

In this data table, rows serve as separate sampling units. Response variables are soil temperature and soil moisture, both continuous data. Soil temperature and moisture values were averaged to get a mean value of all values in each month where data was available. Predictor variables are treatment type and statistical block, both categorical. Temperature and moisture measurements from one datalogger at differing points in time could not be considered separate sampling units without violating the assumption of independence. There was a sample size of 79 (thinned and unthinned) although July soil moisture and temperature measurements were missing from one unthinned sampling unit. Because of the missing data, the sampling design was slightly unbalanced.

There is substantial variation in soil temperature values observed throughout the season and between thinned and unthinned plots (Figure 1). The spread of values appears to be larger during warm periods in the data and smaller during cold periods.

There appears to be substantial variation in  soil moisture values and substantial overlap in values for thinned and unthinned plots (figure 2). The values for May 20, 2023 appear to be showing a dry period and there has likely been a rain event starting on about May 21, 2023. The spread in values appears to be consistent throughout the season, with the possible exception of just after a rain event.