Model Range Related Changes Are Climate Related

The range-related pattern of change at this site matches expectations of animal populations shifting northward due to climatic warming. Figure 3.9 illustrates diagrammatically how northward shifts in geographic distributions of animals would result in the patterns we observed. Note both a population with a Gaussian distribution of abundance (bell-shaped curves) and a population with abrupt drops in populations at the edges of the range (steep-sided curves) can hypothetically show this pattern. This is significant because preliminary surveys (R. Sagarin, unpublished data) show

A. Original study B. After northward shift of species

A. Original study B. After northward shift of species

tions. As viewed from a single study site (indicated by star), the southern species abundances (solid lines) appear to increase, and the northern species abundances (dotted lines) appear to decrease as species move northward through time (e.g., from A to B).

that at least some of our species show extremely steep-sided abundance distributions.

It is difficult with only a single study site to separate actual range shifts from population changes within an existing range. Nevertheless, several of the intertidal species for which we have additional information have shown changes consistent with northern range shifts. For example, Serpulorbis squamigerus, a southern gastropod, which is now extremely abundant at HMS and in Monterey Bay (R. Sagarin, pers. obs.), was not recorded by Hewatt, and was recorded as rare or occurring as single, scattered individuals by investigators in 1966 and 1980 (Hadfield 1966, Morris et al. 1980). The rapid rise of Serpulorbis between 1980 and the first re-survey of Hewatt's transect in 1993, as well as abundance data I have collected for this species in the northern half of its range (R. Sagarin unpublished data) could indicate that Serpulorbis has shown a shift in range in a manner illustrated by the steep-sided distributions in Figure 3.8. Tetraclita rubescens, a southern barnacle that increased significantly in abundance has also shifted its range northward in recent decades (Connolly and Roughgarden 1998). Further observations of recent arrivals of southern species such as lobster (Panulirus interruptus), the Kellet's whelk, Kelletia kelletia (Herrlinger 1981, G. Villa and J. Watanabe, unpub. obs.) and the eel-grass limpet, Tectura depicta (Zimmerman et al. 1996) by other investigators lend additional support to the hypothesis of northward movement of animal populations. These observations, which cover either small spatial areas or limited time periods, are suggestive, but are too limited to confirm range shifts. Ideally, abundances at many sites (especially concentrated near the range edge) must be observed over several years to ensure that ranges have indeed shifted. The Partnership for Interdisciplinary Studies of Coastal Oceans (PISCO), a recent collaboration between intertidal ecologists at the Universities of California at Santa Barbara and Santa Cruz, Stanford University, and Oregon State may provide such a monitoring scheme.

An important consideration in interpreting our data from Hewatt's transect is that they are based on only two surveys separated by 60 years.They therefore can give no indication of the short-term variability in populations along the transect, suggesting the possibility that the pattern we observed was the product of a chance "snapshot" of widely fluctuating populations. To begin to address this scale of variation, we resampled 19 plots in 1996, 3 years after they were originally resampled. The comparison of these plots in 1996 versus 1993 reveals only very small changes in species' abundances (Fig. 3.10). Furthermore, no range-related pattern of change is evident.

The implication of this second, short-term study is that changes in populations are much different when viewed in short time frames (3 years) than in long time frames (60 years). Interestingly, the short-term changes, although inconsistent with the pattern of range-related changes seen over 60 years, did not erase the sharp range-related pattern seen in the long-term. This is illustrated when the 19 resampled plots are examined over the long term by comparing the 1996 survey to Hewatt's study (Fig. 3.11).Without additional intermediate data points, it is impossible to know how long the process of species composition shift takes, although the relatively small magnitude and lack of consistent direction of abun-

increase

increase

• Southern ■ Northern a Cosmopolitan decrease

Figure 3.10. Plot of density in 19 paired plots in 1996 vs. 1993. Symbols are as in Fig. 3.4. Filled symbols represent significant changes (P < 0.025). The lower alpha value represents a Dunn-Sidak correction on the paired t-test, used because the data were drawn from repeated measures of the same plot (Hewatt's study, 1993 and 1996). Adapted from Sagarin et al. 1999.

• Southern ■ Northern a Cosmopolitan decrease

Log(1993 Density +1)

Figure 3.10. Plot of density in 19 paired plots in 1996 vs. 1993. Symbols are as in Fig. 3.4. Filled symbols represent significant changes (P < 0.025). The lower alpha value represents a Dunn-Sidak correction on the paired t-test, used because the data were drawn from repeated measures of the same plot (Hewatt's study, 1993 and 1996). Adapted from Sagarin et al. 1999.

(g o increase increase

• Southern ■ Northern a Cosmopolitan

Log(Hewatt Density +1)

Figure 3.11. Plot of density in 19 paired plots (as in Fig. 3.9) between 1996 and Hewatt's study. Symbols are as in Fig. 3.3. Filled symbols represent significant changes (P ** 0.025). The lower alpha value represents a Dunn-Sidak correction on the paired t-test, used because the data were drawn from repeated measures of the same plot (Hewatt's study, 1993 and 1996). Adapted from Sagarin et al. 1999.

• Southern ■ Northern a Cosmopolitan

Log(Hewatt Density +1)

Figure 3.11. Plot of density in 19 paired plots (as in Fig. 3.9) between 1996 and Hewatt's study. Symbols are as in Fig. 3.3. Filled symbols represent significant changes (P ** 0.025). The lower alpha value represents a Dunn-Sidak correction on the paired t-test, used because the data were drawn from repeated measures of the same plot (Hewatt's study, 1993 and 1996). Adapted from Sagarin et al. 1999.

Ev dance changes in the short-term suggest that the range-related pattern may take many years to appear. It is likely that species respond individually to climate change depending on thermal sensitivities, reproductive strategies, mobility, life span, and interactions with other species (Graham and Grimm 1990).The pattern we see today thus integrates changes by many species over many years. This may help to explain the failure of other investigators to find such strong range-related changes in their own reexaminations of historical data sets that date back only 20 or 30 years. For example, John Pearse and collaborators, in reexamining extensive intertidal faunal surveys originally conducted from 1971 to 1973 from Santa Cruz, California, did not find a strong range-related pattern of change overall, but did find increases in several southern species (J. Pearse, unpublished data).

Nevertheless, an intertidal community with hundreds of invertebrate and algal species is an incredibly complex system character ized by a myriad of potential direct and indirect biotic interactions that may affect populations of any given species. Over the course of 60 years, factors other than warming sea temperatures have undoubtedly changed populations of intertidal invertebrates. We examined several alternative hypotheses to explain the changes we observed (Barry et al. 1995, Sagarin et al. 1999). Our findings are summarized in Table 3.1. While any of these alternative hypotheses might explain some of the faunal changes we observed, none of them can explain the strong range-related pattern of change in its entirety.

0 0

Post a comment