I began my data collection by measuring the soil moisture in randomly selected locations. I measured twenty replicates and recorded whether the soil was dry or damp and the location of the sample. The random placement of the samples was determined using a random number generator to dictate the direction and number of paces taken to the next sample. This method was easy to implement but the data collection process was not enjoyable because it took place on a very hot and windy afternoon. Measuring the soil moisture was also straightforward because it’s assignment of dry or damp was based entirely on touch. However, this type of measurement makes the predictor variable categorical, and therefore the effects of a continuous soil moisture gradient will not be determined.
Using these twenty samples, I noticed that dry soil was primarily found on the hilltop and hillside, whereas damp soil was primarily found near the creek. I used this observation to stratify the field according to soil moisture into the damp, creekside and dry, hill stratifications.
The next day I recorded the presence/absence of C. fulgida in twenty quadrats randomly placed in the creekside stratification. This method was easy to implement but observation of C. fulgida was difficult in areas with long grass. Only one quadrat was observed to have C. fulgida present. My next step is to collect twenty replicates in the hill stratification and analyze whether the amount of quadrats with C. fulgida present differs between stratifications, and thus differs according to soil moisture.