Figure 4 were conducted in the fall at temperatures of 4°-5°C 

 when the acclimation time would be slower, 2) no changes in 

 weight were noted after 18-20 h during repetitious weighings of a 

 few randomly selected bloodworms and sandworms collected peri- 

 odically during commercial sampling, and 3) commercial samples 

 collected on one day were never processed until at least 24 h later, it 

 is highly probable that all length and weight measurements were 

 made on commercial samples only after all worms had been fully 

 acclimated to standard high salinity conditions. 



The length measurement of a marine worm in its natural state is a 

 difficult if not impossible undertaking: the soft-bodied Annelid can 

 coil, undulate, expand, and contract. To avoid these problems, we 

 anesthetized the worms before measuring them. The relationships 

 of natural lengths to anesthetized lengths for bloodworms and sand- 

 worms collected from the Sheepscot River are shown in Figure 5. 

 These results demonstrate that the difference between anesthetized 

 length and natural length is greater for bloodworms than for sand- 

 worms; a bloodworm of 20 cm anesthetized length is equivalent to 

 approximately 13 cm natural length, whereas a sandworm of 20 cm 

 anesthetized length is equivalent to approximately 17 cm natural 

 length. 



Bloodworm samples collected during April and May were sexed 

 because in the region of Wiscasset, Maine, spawning occurs in 

 June (Creaser 1973). Sandworm samples were sexed during 

 August and September after spawning in April and May (Creaser 

 and Clifford footnote 1 1). 



Verification of Sampling Procedures and 

 Interview Responses 



Studies preformed to determine what effect the anesthetic might 

 have on the worm "s weight indicated that it had little effect. 



Studies performed to determine if length measurements upon 

 bloodworms and sandworms are true and reproducible indicated 

 that bloodworm lengths, over the range of sizes tested (15.7-36.6 

 cm), are reproducible within ±0.2 to + 1.0cm(at95% confidence 

 limits or 1.96 SE) and sandworm lengths, over the range of sizes 

 tested (12.1-64.3 cm), are reproducible within ±0.4 to ±2.4 cm 

 (at 95% confidence limits or 1.96 SE). 



Studies in which lengths were obtained on individual worms after 

 being relaxed in two different anesthetics (0.2% propylene 

 phenoxytol and 7.5% MgCl 2 ) demonstrate that when bloodworms 

 were first relaxed and measured in 0.2% propylene phenoxytol and 

 then relaxed and measured in 7.5% MgCU, the lengths recorded in 

 the MgCl, were usually smaller (23 out of 24 cases). The reduction 

 in size varied between 0.8 and 23.4%. When bloodworms were 

 first relaxed and measured in 7.5% MgCl, and then relaxed and 

 measured in 0.2% propylene phenoxytol, the lengths recorded in 

 the propylene phenoxytol were usually greater (16 out of 2 1 cases). 

 Increased lengths varied between 1 .0 and 12.0% and decreases var- 

 ied between 1.5 and 13.0%. These results suggest that caution 

 should be used when comparing the findings in this manuscript 

 (where 0.2% propylene phenoxytol was used as as anesthetic) with 

 the results in other publications (where other anesthetics were 

 used). 



More detailed information on the results of the studies above, 

 which were performed to verify various sampling procedures, is 

 reported in Creaser et al. 31 



3l Creaser, E. P.D. A. Clifford. M.J. Hogan, and D. B. Sampson. 1980. An anal- 

 ysis of the commercial baitworm fishery for sandworms Nereis virens Sars and 

 bloodworms Glycera dibranchiala Ehlers in Maine. Maine Dep. Mar Res. Lab. 

 Res. Ref. Doc. 80/1 8. 1 80 p. 



The results of studies to determine if the 25 worm samples were 

 truly representative of the entire contents of the bloodworm buckets 

 and sandworm hods are presented in Table 8. It is evident from 

 these results that on 10 out of 10 tries the range of bloodworm mean 

 lengths and weights ( ± 1.96 SE) overlapped the actual mean length 

 and weight of the entire "bucket" population. On 9 out of 10 tries 

 the range of sandworm mean lengths (±1.96 SE). and 8 out of 10 

 tries the range of sandworm mean weights ( + 1 .96 SE), overlapped 

 the actual mean length and weight of the entire "hod" population. 

 There were few problems inherent in our method of selecting 25 

 bloodworms for measurement and most of the time the same holds 

 true for sandworms. 



Few errors were observed when comparing the total landings we 

 recorded during the digger interview with the total the dealer 

 recorded and paid the digger for. In only a few instances during a 4- 

 yr period were intentional errors made by diggers. Occasionally, a 

 digger failed to report to the dealer that we had collected 25 of his 

 worms and his recorded landings with the dealer were therefore 25 

 worms short. 



The results of our efforts to check the accuracy of the diggers' 

 estimates of their digging time are shown in Table 9. This study was 

 necessary because certain industry factions shared the opinion that 

 diggers were reporting false information regarding their estimates 

 of beginning and ending time. The results in Table 9 demonstrate 

 that there is less than a 2 % discrepancy between the time estimates 

 of groups of diggers and their actual digging time recorded by 

 observation from concealed positions. However, when time esti- 

 mates for individual diggers are obtained through digger interviews 

 on the flat, these estimates are probably more accurate than the esti- 

 mates they would have made had they been interviewed at the 

 worm cellar some distance away. Because of manpower limitations 

 we were not able to follow individual diggers back to their respec- 

 tive cellars to obtain estimates of their digging time. We can only 

 state that had we been able to do this the discrepancy might have 

 been greater than 2 % , but probably still within very acceptable lim- 

 its. These data were analyzed to determine if the ratio of two varia- 

 bles (actual vs. reported time) was significantly different from a 1 : 1 

 ratio at 95% confidence limits (2 SE fi ). The results indicate that the 

 relationship between actual and reported time is not significantly 

 different from a 1:1 ratio ( 1 .01 764 ± 0.028 19 or 

 0.98945-1.04583). In other words, the mean estimate of digging 

 time, as reported to the sampling crew, is quite accurate. As far as 

 individual groups of diggers are concerned, some estimate a little 

 high, some estimate a little low, and some estimate precisely. Verifi- 

 cation of the accuracy of both reported landings and digging time 

 estimates enables us to conclude that the estimates of catch/hour, 

 one of the simplest indices of marine worm abundance, are proba- 

 bly quite accurate. 



Commercial Sampling for Length, 

 Weight, Sex, and Condition 



Table 10 shows that the 6-mo mean lengths ( ± 1 SE) for blood- 

 worms were 18.72±0.60 cm (1973), 19.84±0.38 cm (1974). 

 20.74 ± 0.59 cm (1975), and 20.83 ± 0.54 cm ( 1976). These means 

 are not significantly different from one another at 95 % confidence 

 limits ( ± 1 .96 SE). On the basis of this commercial sampling infor- 

 mation, no significant differences occurred in the size of blood- 

 worms harvested between 1973 and 1976. 



It is also apparent from Table 10 that during April and May poten- 

 tial spawners comprise between 7.33-13.58% and 0.50-1.63%, 

 respectively, of the commercial catch. Apparently, the diggers 



13 



