Nahant: Analysis and Discussion

Coralline algae, red algae, and tide pool cover were all significant predictors of limpet abundance during diurnal low tide (Table P).  Canopy algae cover was not significant, likely because canopy algae species have different relationships with limpets and the algal species were not quantified separately.  Coralline algae cover had a slightly negative relationship with limpet abundance, where the abundance of limpets decreased while cover of coralline algae increased (r² = 0.002).  Both red algae cover and tide pool cover had more dramatic negative relationships (r²_redalgae = 0.091; r²_tidepool = 0.314).   None of the limpets exhibited foraging behavior during observation but rather remained flush with the substrate.

The negative relationship between coralline algae and limpet abundance signifies a new aspect to the algal association.  These negative relationships together suggest that A. testudinalis preferentially inhabits bare rock during diurnal emersion, at which time it stops foraging and selects a resting site.  Smooth rock may be a more protective substrate for a home site than C. circumscriptum.  Future experiments could investigate the benefits of selecting resting sites on bare rock versus sites on coralline algae, including the possibility of differential survival. 

There was no significant relationship between the amount of C. circumscriptum and the number of A. testudinalis within 0.25m² (p_coralline = 0.42; p_{>50%coralline} = 0.06) at diurnal low tide.  Results are contingent on the assumption that 0.25m² is a biologically meaningful scale for observing the hypothesized association.  Steneck (1982) found a significant relationship using 2.5x2.5cm (0.0625m²) quadrats.  There is likely a distance threshold, shorter than 0.5m, after which there is no association between C. circumscriptum and A. testudinalis.  If A. testudinalis preferentially selects resting sites with no algal cover but also preferentially forages on C. circumscriptum, than the distance between resting sites and coralline algae is likely short enough to allow the limpets easy access for nocturnal foraging.

Algae cover measurements are a potential source of methodological error.  C. circumscriptum does not occur in distinct sections but rather has a broken and patchy distribution (see Background Figure 3), often following cracks in rocks, making percent cover difficult to estimate.  Figure Y demonstrates these confounding effects by illustrating two operational definitions of coralline percent cover in relation to distance trends between C. circumscriptum and A. testudinalis.  The bottom histogram counted cells with any amount of coralline algae and has a distinct positive relationship trend, with limpet abundance decreasing with increasing distance from coralline algae.  The top histogram counted cells with >50% coralline algae and shows the opposite trend, with limpet abundance increasing with distance from coralline algae. 

These results suggest that limpets select rock resting sites within a short distance of any amount of C. circumscriptum.  Cells with much coralline algae and little bare rock are unlikely to contain resting limpets, which could explain the relationship in the top histogram of Figure 2.  These results also emphasize the need to select biologically meaningful scales, such as the average distance A. testudinalis moves to forage.  Future designs could investigate the effect of scale on this relationship by employing quadrats of varying sizes.  A more effective cover estimate measurement should be used at each scale, such as overlaying grids onto the quadrats and identifying the substrate beneath each gridline intersection.