SEASONALITY FROM CHIONE

Venus clams (Chione undatella, C. fluctifraga, and C. californiensis) are probably the most abundantly represented genus in the shell middens of San Diego County. The availability of a method for determining the portions of the year during which this clam was harvested prehistorically would be important for reconstructing the seasonal subsistence round in the region and for assessing the seasonal duration of occupations at particular sites.

In southern California, seasonality studies based on mollusk growth increments were pioneered by Margaret Weide (1969), using Pismo clam (Tivela stultorum). The biological foundation for local studies of Chione undatella growth increments has been the work of Richard M. Barker (1970) at Cholla Bay in the Gulf of California. Barker distin­guished two key features on the external surfaces of Chione undatella valves: grooves, which were interpreted as marking annual winter hiatuses in growth; and ridge-and-trough sequences, which were interpreted as fortnightly growth increments. For the first through fifth years of life of the Cholla Bay clams, Barker reported averages of 16.9, 8.8, 4.5, 4.0, and 2.0 ridge-and-trough growth increments, respectively.

Several studies, many of them conducted in Orange County, have suggested the feasibility of using growth increment patterns on archaeologically recovered Chione undatella shells to determine seasonality. Other studies have called into question the validity of particular methods for calibrating growth increment patterns or have doubted the likelihood of any successful calibration method.

-- Christopher E. Drover (1974) proposed growth-increment methods for determining the seasonality of C. undatella and C. fluctifraga valves. For C. undatella, the Cholla Bay average number of ridges for each year of growth was divided by three (for spring, summer, and fall seasons) to give the number of ridges expected for each season in each growth year. On archaeological specimens, the number of grooves and the number of ridges since the final groove were then used to assign a season of death. (For C. fluctifraga, a linear measurement, in centimeters, of the amount of growth between grooves was substitut­ed for the ridge count.) Analyzing Chione spp. valves from ORA-119A, a Late Prehistoric site overlooking Newport Bay, Drover concluded that there was a strong predominance of winter Chione collection. Drover's method was also employed in several subsequent archaeological studies in Orange County (cf. Koerper 1980:137).

-- Christina Carter (1978) applied a modified form of Drover's method to C. undatella valves recovered from site LAN-702 in Long Beach. Rather than using Barker's averages for yearly ridge increments, Carter used average counts based on the valves from site LAN-702 itself. The site had two stratigraphic components: in the lower component, C. undatella growth increments indicated a predominantly fall collection season (44% of valves); in the upper, a predominantly winter collection (43%).

-- Michael E. Macko (1983) suggested another modification in Drover's method. Based on a New Zealand study by Peter J. F. Coutts and Charles Higham (1971), Macko proposed to partition Chione growth incre­ments unequally among the non-winter seasons, assigning 35% of growth to spring, 50% to summer, and 15% to fall.

-- Edward Lyons (1978) used Barker's data to propose a method of seasonality determination for C. undatella that differed from Drover's. Lyons fitted Barker's ridge increment averages to a logarithmic curve: Y = 12.13 ln (t) + 17.03, in which Y is the total number of ridges and t is the age of the clam in years. The total number of ridges on a valve was then used to infer the clam's age in years and fractions of a year. Lyons' model assumed that growth was continuous throughout the year and that spawning (and, implicitly, the formation of annual grooves) could be taken as having occurred on May 1. He divided the year into three four-month seasons (May-August, September-December, and January-April). Analyzing ridge counts on C. undatella valves from site ORA-82 in Huntington Beach, he found that all three seasons were substantially represented but that there was a statistically significant concentration of counts assigned to the January-April season. Subsequent studies by Lyons also applied the technique to other Orange County sites (cf. Koerper et al. 1984:70; Lyons 1984).

-- Clifford V. F. Taylor (1980) developed a variant of Lyons' method. Rather than using the logarithmic curve that Lyons had fitted to Barker's Cholla Bay ridge averages, Taylor proposed using site-specific logarithmic curves based on archaeological collections. Curves were fitted to data from sites SDM-W-192A (San Diego Bay), SDM-W-132A (Agua Hedionda Lagoon), and SDI-4281. Implicitly, Taylor also aligned the annual grooves with January 1 rather than May 1. He argued that stronger seasonal patterning was found when the site-specific curves were used in place of the Cholla Bay curve.

-- Henry C. Koerper (1980) challenged the validity of Drover's Chione seasonality method on several grounds. The sharply decreasing numbers of "fortnightly" ridges in successive years of C. undatella growth seemed to imply a lengthening of the "winter" inactive season, making assignments of most shellfish harvesting to a winter season misleading. Factors other than lowered winter temperatures, such as disturbance from storms and spawning, may produce pseudo-annual grooves. The annual growth averages observed by Barker at Cholla Bay in the Gulf of California may not be applicable to southern California. The rate of shell growth may be influenced by very localized environmental factors as well as by general temperature. Koerper, Richard Cerreto, and Karl P. Reitz (1984) also responded to Lyons' proposal, reiterating and elaborating their criticisms of the use of Chione seasonality methods in general and noting additional problems with Lyons' Chione growth model and its application. Lyons (1984) responded, reasserting the validity of the method and offering a computer program to assist with computations.

-- Richard Cerreto (1988, 1992) comprehensively criticized of the seasonality methods proposed by Drover, Lyons, and Macko. He based his rejection of their proposals on experimental studies of a modern population of C. undatella in Newport Bay. The individual shellfish were captured, marked, and released into special holding pens. After two years, they were recaptured, and their growth markers were evaluated against the expectations of the three seasonality models, none of which showed signficant agreement with the empirical observations.

-- Leslie Quintero (1987:157-161) discussed the problems in applying C. undatella seasonality methods to San Diego County sites. She raised the possibility of an approach substantially different from that of Drover, Lyons, and others, based on studies of other shellfish species by Cheryl Claassen (1986). This approach involved comparing the proportion of archaeological specimens having an outer annual groove with the percentage of individuals within a modern population which were in a growth arrest phase during a given month. Quintero did not apply the method to an archaeological collection, in part because of the absence of modern studies for comparison.

PROSPECTS

Future zoological and archaeological investigations may be able to clarify the nature of "annual grooves" and "fortnightly ridges," the amount of shell growth during successive years of an animal's life, the seasonal distribution of that growth, and the degree of variability in patterns of growth in different environmental settings.