Historical Science Defined

In Francis and Hare's (1994) definition, historical-descriptive science begins with observations of a system. At this stage, the obser vations may be anecdotal (e.g., fishermen noticing unusual species in their catches during warm-weather periods) or tangential to a main line of research (e.g., an investigator noticing changes in species composition over many years at a study site). Whatever the source, these observations are intriguing enough that an investigator uses them as a launching point for a more in-depth study of the system.

Simple models, or narratives, are then developed that suggest hypotheses of how the various observations fit together. For studies of how climate changes will affect biotic systems, many of these models have already been suggested in the form of predictions of population or community changes under expected warming climate. The most commonly proposed model is that species' geographic ranges will shift poleward in order to maintain the same position relative to climatic conditions. This hypothesis has been forwarded for a wide variety of biotic systems including terrestrial plants (Davis 1989, Woodward 1992), infectious disease (Shope 1991, Patz et al. 1996), invertebrates (Bhaud et al. 1995), birds (Root 1993,Taper et al. 1995), freshwater fishes (Scott and Poynter 1991, Murawski 1993), algae (Breeman 1990), marine fishes (Frank et al. 1990), and coastal marine communities (Fields et al. 1993, Lubchenco et al. 1993). Other types of narratives that have been developed for species under warming climate and that might be tested using historical studies include (but are not limited to) changes in timing of reproductive or other life-history stages (Bree-man 1990, Bhaud et al. 1995, Sparks and Carey 1995), increased stratification of oceans affecting nutrient upwelling (Fields et al. 1993, Roemmich and McGowan 1995), and alterations of species interactions (Murawski 1993, Davis et al. 1998).

In the final step, historical observations from a variety of sources are examined for patterns that support or don't support the model. The physical data used typically include long-term time-series data for one or more climatic variables, such as daily temperature records or annual tallies of frost-free days. Historical data on biotic systems might include presence/absence data, population counts, or measures of community composition and diversity through time. These data may take the form of "snapshot" data taken at the same location at separate points in time, or preferably, as time series data collected for many years, such as catch records for fisheries or Christmas bird counts. The analysis of these data may be as simple as noting a gradual increase in temperature alongside gradual changes in species' abundances (Southward et al. 1995), or more rigorous attempts to correlate physical and biological variables using time-series analysis (Francis and Hare 1994, McGowan et al. 1996).

0 0

Post a comment