For both days, snail size was divided into categories and graphed (fig 2.1, 2.2, 2.3, 3.1, 3.2, 3.3). Each graph showed different peaks for different snail sizes. When these size categorizes were broadened to encompass one third of all shell heights, patterns of snail height distribution emerged. In September, the abundance of the three size groups were very similar in the low and middle tidal zones, and the graphs showed overlapping slopes (fig 1.1). This is supported by the t-test, which showed that these samples are not significantly different (p=.24, table 1.2). The abundance of snail height that fell into the three size categories in the high zone differed from the low and middle zones (fig 1.1), which is also supported by the t-test results, because both p-values were low, indicating a significant difference between the low and high distributions and the middle and high distributions (p= .017 and p=.023, table 1.2). In November, a similar trend was noted, but was not as extreme. The low and middle zones had similar distributions of the three size groups, and the distributions differed slightly in the high zone (fig 2.2). The t-test showed that the distribution of snail height in the low and middle zones was not significantly different (p=.061, table 1.2). However, a t-test also showed that the distribution of snail height in the low and high zones was not significantly different (p=.30, table 1.2). I found this to be quite surprising because of the different selection pressures in the two zones. At high intertidal heights, the physical pressures are usually greater, and at low intertidal heights biological pressures are usually greater (Bertness 2007). I had hypothesized these different pressures would select against different snail sizes. The sample from September showed this but the November sample did not, perhaps because the sample days were taken during two different months.
Several single factor ANOVA tests were used to analyze the distribution categories of snail height. Tests were conducted for both months, on each month separately, using both the three broad group distribution sizes and the narrow groups distribution sizes (1mm groups). When both months were assessed together there was a significant difference for the narrow groups but not the broad groups (p = .0001, table 1.4 and p= .1605, table 1.3). When the samples were treated separately this trend repeated; the difference being significant when snail size was assessed by small categories (p= .000659, table 2.1 and p= .073514, table 3.1) and non significant when larger categories were used (p=.134, table 2.2 and p=.4532, table 3.2).
Both the low and middle zones were more populated in September than in November, while the opposite was true for the high zones. Differences between these two samples could be due to seasonal variability in abundance of L. littorina (Saier, 2000). Snail abundance peaks in October because of an influx of juvenile snails (Saier, 2000). Because the first sample was taken in late September (the 28th), this could have increased not only the number of snails, but also the abundance of snails falling into the smaller size groups in the low and middle zones. A t-test found that the difference between the two days was significant (p= 0.044, table 1.2), thus it is easier to understand the patterns found because the days are considered separate samples, not replicates.