band in cross-section was identical to that previous- 

 ly described and illustrated by photograph (Peter- 

 son et al. 1983) of clams harvested from the Middle 

 Marsh locality. 



The pattern of band deposition relative to times 

 of initial planting, first measurement (12 mo), and 

 collection (24 mo) was also extremely consistent 

 across all data sets. Initial planting in 1980 occurred 

 during the period of annual band deposition for 70 

 of the 72 clams. (In one clam, the 1980 annual band 

 was just completed and in another the 1980 annual 

 band was just about to begin at the time of initial 

 planting.) The disturbance check caused by the 

 12-mo measurement fell near the end of the growth 

 band for 70 clams and just after the band for the 

 two others. The time of collection in 1982 fell dur- 

 ing or just immediately after the deposition of the 

 1982 annual band for all clams except those from 

 local origin at North River. Of the 16 cut clams in 

 that data set with sufficient growth for band resolu- 

 tion, 12 were just beginning to deposit their 1982 

 band at the time of collection (2 August, 3-4 wk 

 earlier than the other sites— Ikble 2). Two of the 16 

 lacked the terminal band, whereas the remaining 

 two had already deposited a substantial amount of 

 the 1982 band. This North River local data set was 

 the only one that contained any clams (only three) 

 which had bands sufficiently faint to cause any doubt 

 about recording them. 



By counting all presumptive annual bands over the 

 complete growth record of each clam, we also 

 estimated the age of each of the 89 M. mercenaria 

 used in this experiment. The estimated age at the 

 experiment's initiation for the 17 clams excluded 

 from our 2-yr tests ranged from 6 to 29 yr and 

 averaged 15.5 yr (±1.7 SE). For the 72 clams that 

 grew sufficiently and included a sufficient shell 

 marker at initiation to be used in our 2-yr tests, age 

 at experimental initiation ranged from to 17 yr and 

 averaged 3.9 yr (±0.3 SE). Thus, the average age of 

 the clams that could not be used for our tests was 

 significantly (P < 0.01 in a t-test) higher than that 

 of the 72 clams that were used. Most (16 of 17) of 

 the excluded clams lacked both sufficient growth and 

 an obvious disturbance check at initiation. Only one 

 clam was excluded with sufficient growth but with- 

 out an adequate disturbance check. Although 16 

 clams lacked sufficient growth to determine ac- 

 curately the shell size at the initiation of the experi- 

 ment and were therefore excluded from our tests, 

 all of these clams possessed discrete bands in their 

 shells that could be counted separately. They were, 

 however, close together at the terminal margin of 

 the shell where separating them was not always 



possible and caused some uncertainty in their age 

 estimates. 



Discussion 



Our banding data from recovery of marked and 

 measured M. jnercenaria after virtually 24 mo of 

 terminal growth provide a compelling case for the 

 reliability across different habitats of using major 

 growth bands in sectioned shells to age hard clams 

 in the Cape Lookout region of North Carolina. Our 

 previous test of the annual periodicity of banding 

 in North Carolina's M. mercenaria (Peterson et al. 

 1983) was carried out in only single locality, a Back 

 Sound seagrass bed, characterized by almost equal 

 proportions by weight of sands and muds in its sur- 

 face sediments (Tkble 1). Through this study, we ex- 

 tend our tests of the reliability of annual band deposi- 

 tion in M. mercenaria to several additional sites, 

 located in different bodies of water and characterized 

 by much sandier sediments (Ikble 1). Of the six sites 

 that we used for these tests, three were vegetated 

 by seagrasses and three lacked macrophytic cover 

 (Tkble 1). (Although our initial experiment in Mid- 

 dle Marsh was situated inside a seagrass bed, sea- 

 grasses were removed from the experimental plots 

 during each sampling.) Despite these differences in 

 local geographic location (and probably salinity), 

 sediment grade, and seagrass presence, banding pat- 

 terns were consistent and bands were deposited 

 annually. 



By using relatively high densities of 80 m"^ (over 

 10 times the average natural density found in a 

 Bogue Sound seagrass bed by Peterson 1982, in 

 North River, Bogue, Back, and Core Sounds by Beal 

 1983, and in Back Sound seagrass beds and sand 

 flats by Peterson et al. 1984), we ran the risk of 

 causing inhibition of growth. In fact, we were unable 

 to analyze growth band deposition in 17 of our 89 

 clams largely because of insufficient growth in the 

 terminal 2-yr increment. This problem may have 

 been induced by our choice of relatively high densi- 

 ty in these experiments, but it does have a natural 

 analog. In areas with relatively slow growth and in 

 older age classes where growth rate slows, aging 

 North Carolina's M. mercenaria by counting annual 

 bands in shell cross-section may be more difficult and 

 lead to greater error than the consistency of band- 

 ing results on our other 72 clams implies (Tkble 3). 

 Nevertheless, banding even in these generally older 

 clams that were excluded from our analyses was 

 discrete and sufficient to permit us to estimate their 

 ages. Aging does not appear to imply cessation of 

 annual band deposition but only an increased dif- 



675 



