Linking Predation and Temperature

This sensitivity of Pisaster predation to small changes in water temperature is particularly surprising because it occurs over the range of 9 to 13°C, in the middle of this sea star's thermal tolerance range. Pisaster ochraceus is found from at least Punta Baja, Baja California, to Prince William Sound, Alaska, and populations living near these

Alternating

Figure 4.7. Results of laboratory predation experiment. Sea stars were maintained under three temperature treatments: constant 12°C, constant 9°C, and alternating (periods 1-3 = 12°C, 9°C, 12°C, respectively). Bars are mussels consumed per sea star per day (+ SEM) in treatments (n = 4 tanks/treatment) during three consecutive 14-day periods. Data were analyzed in a repeated measures analysis of variance. "Treatment" (F2 9 = 59.81, p < 0.001), "time" (F218 = 36.59, p < 0.001), and "time x treatment" (F4 18 = 5.70,p = 0.004) were all significant. Within each time period, shared lines below bars indicate groups whose means did not differ (Tukey-Kramer, p > 0.05). Results are presented for the first three periods of the experiment. Thereafter, sea stars apparently became temporarily satiated on the ad libitum diet, and feeding rates declined in all treatments (Sanford 1999b).

Period

Figure 4.7. Results of laboratory predation experiment. Sea stars were maintained under three temperature treatments: constant 12°C, constant 9°C, and alternating (periods 1-3 = 12°C, 9°C, 12°C, respectively). Bars are mussels consumed per sea star per day (+ SEM) in treatments (n = 4 tanks/treatment) during three consecutive 14-day periods. Data were analyzed in a repeated measures analysis of variance. "Treatment" (F2 9 = 59.81, p < 0.001), "time" (F218 = 36.59, p < 0.001), and "time x treatment" (F4 18 = 5.70,p = 0.004) were all significant. Within each time period, shared lines below bars indicate groups whose means did not differ (Tukey-Kramer, p > 0.05). Results are presented for the first three periods of the experiment. Thereafter, sea stars apparently became temporarily satiated on the ad libitum diet, and feeding rates declined in all treatments (Sanford 1999b).

range limits experience seasonal water temperatures > 20°C and < 4°C, respectively.

Although the ecological effects of small changes in water temperature have generally been overlooked in marine systems, some decrease in predation rate with declining temperature is expected from a physiological perspective. The direct effects of colder temperatures on rates of movement, metabolism, and digestion should slow feeding rates (Clarke 1987, Cossins and Bowler 1987, Birkeland and Lucas 1990). Pisaster shows a lowered metabolic rate at lower temperatures (Paine, pers. comm. in Mauzey 1966) and feeding has been observed to decline or stop completely at low temper-

atures (< 5°C) in several Atlantic sea star species (Feder and Christensen 1966).

In the field, Pisaster predation was sharply reduced during upwelling (Fig. 4.5B), not only because each sea star consumed less, but also because densities of foraging sea stars were locally reduced (Fig. 4.5C). This twofold effect may be a common response of ectothermic consumers to temperature change. Not only should colder temperatures slow feeding rates, but a greater proportion of individuals should become inactive and seek shelter as temperature declines to some species-specific level (e.g., Aleksiuk 1976, Frazer and Gilbert 1976, Hunt 1977,Whicker and Tracy 1987). For example, the rate of predation of ladybird beetles on aphids was found to decrease exponentially as temperature decreased, since both mean walking speed and the proportion of ladybirds that were actively foraging decreased linearly with temperature (Frazer and Gilbert 1976, Kingsolver 1989).

My study indicates a direct link between water temperature and the strength of keystone predation in these rocky intertidal communities. These data provide mechanistic insight into how temperature change impacts an important species interaction. Such information lies at the heart of individual-based and physiologically structured models (Dunham 1993, Kingsolver et al. 1993, Murdoch 1993). These models seek to predict how the influence of environmental change on the physiology of individual organisms translates into population- or community-level effects. Knowledge of how temperature changes influence the Pisaster-Mytilus interaction should thus facilitate specific predictions about how anticipated climatic changes would impact these rocky intertidal communities.

There is however a major obstacle to this and similar global change studies; physical models used to predict climatic changes are resolved to a scale far broader than the relatively local scale of ecological studies (Root and Schneider 1993). General circulation models (GCMs) represent the globe as a grid of boxes and use laws of atmospheric physics to predict average conditions in each box. Although improvements have been made, computational limits still restrict GCMs to using boxes that are many kilometers on each side. This scale is too coarse to accurately predict changes in regional climate (Root and Schneider 1993).

Even more uncertain are changes in the frequency, timing, and intensity of seasonal or episodic events such as coastal upwelling, storms, frosts, heat waves, or fire (Root and Schneider 1993). Many researchers have suggested that it is changes in extreme or seasonal events, rather than changes in mean conditions, that may have the greatest effects on natural communities (Dobson et al. 1989, King-solver et al. 1993, Root and Schneider 1993, Bhaud et al. 1995). It is possible to generate some estimates of climatic variability from physical models, yet these data are rarely considered by climatic modelers (Root and Schneider 1993). There is thus little information available regarding the sort of variability that may be central to predicting the responses of ecosystems to climate change. For example, in the Pacific Northwest, the density and impact of Pisaster in the rocky intertidal zone is highest during the summer (the upwelling season). Thus it is changes in the frequency and intensity of upwelling, rather than changes in annual sea surface temperature, that should play the greater role in regulating sea star predation.

0 0

Post a comment