the bacterial density of the immediately fixed sample with that of the incubated one.
Bacterial density in the CF was determined using the acridine organe epifluorescence
direct counting (AODC) technique (Hobbie et al. 1977). The number of bacteria and wet
biomass were calculated using the expressions of Linley et al. (1981). Conversions to dry
biomass and carbon equivalent of the wet biomass were obtained using the coefficients 0.2
(Troitsky and Sorokin, 1967) and 0.1 (Luria, 1960), respectively.
RESULTS
Bacterial density of the CF of anthozoans obtained from in situ investigations
Bacterial densities in the CF of L. pruvoti and P. axinellae varied considerably during the
investigation period. Rod-shaped bacterial density ranges between 0.5 to 7.37 X 10 5 cells
ml1 CF (x = 3.09 X 10
5 , S.D. = 2.08 X 10 5 , n = 22) in L. pruvoti, while coelenteric cocci
vary between 2.74 x 10
5 cells m1 CF and 40.19 X 10 5 cells ml- 1 (x = 15.35 X 10 5 , S.D. =
10.3 X 10 5 , n = 22).
In P. axinellae values of rods range between less than 10 3 and 6.1 x 10 5 cells ml-1 (x = 2.13
x 10
5 cells ml -1 , S.D. = 1.36 x 10 5 , n= 22) and of cocci from less than 10 3 to 54.94 x 10 5
cells'ml"
1 (x= 11.7 x 10 5 cells ml- 1 , S.D. = 15.7 x 10 5 , n=22). Mean rod-shaped bacterial
density in ambient water is 1.41 x 10 5 cells ml- 1 (S.D. = 0.4 x 10 s , n= 22) and the mean
density of coccoid bacteria is 2.23 x 10 5 cells ml -1 (S.D. = 0.7 x 10 5 , n = 21). Therefore,
total bacterial density is approximately 6 times higher in L. pruvoti and 4 times higher in
P. axinellae than in ambient waters. Although large deviations of bacterial densities occur
in L. pruvoti and P. axinellae, no distinct diurnal cycles could be observed. Within the
coelenteron of the symbiotic sea anemone S. giganteum, bacterial densities are generally
above those of ambient waters. Mean bacterial density of the CF is 2.49 X 10 6 cells ml* 1
(S.D. = 1.07 X 10
6 ), i.e. 2.6 times higher than in ambient waters (Herndl et ai, in press).
Rods contribute 90% to the number of bacteria and 97 % of bacterial biomass. Mean
rod-shaped bacterial density is 2.3 x 10 6 cells ml -1 (S.D. = 1.1 X 10 6 ) and therefore 14
times higher than the density of rod-shaped bacteria in the adjacent water. The biovolume
of coelenteric bacteria did not differ significantly from that of ambient waters. Mean
volume obtained for rods is 0.6 µm 3 (S.D. = 0.14, n = 100) and for cocci 0.13 µm 3 (S.D. =
0.03, n = 30).
The mean bacterial density in the coelenteron of the symbiotic sea anemone A. sulcata is
2.33 X 10 6 cells ml' 1 (S.D. = 2.2 X 10 6 , n = 7), consisting of up to 90% of cocci. The
volumes of both rod-shaped and coccoid cells of the CF were significantly higher
(ANOVA, P < 0.001) than those of ambient waters.
Bacterial fluctuations in the CF
Figure 1 demonstrates that in P. axinellae the regulative function of the CF against
bacteria depends upon the initial bacterial density in the coelenteron. C. cespitosa
demonstrates very low generation times of coelenteric bacteria at low rod-shaped bacterial densities below 0.1 X 10 5 cells ml -1 CF, but at 0.2 X 10 5 cells ml -1 clearance activity of
the CF already begins (Fig. 2). Coelenteric coccoid bacterial density shows no such
marked fluctuations in turnover rates as do rod-shaped bacteria. During the clearance
phase, however, rods and cocci show quite similar slopes, although clearance of rods
starts at lower bacterial densities than cocci clearance.
In A. sulcata, cocci dominate the bacterial population in the coelenteron during the
course of incubation experiments ; this corresponds with the values of the in situ investi409
Bacterial density in the CF was determined using the acridine organe epifluorescence
direct counting (AODC) technique (Hobbie et al. 1977). The number of bacteria and wet
biomass were calculated using the expressions of Linley et al. (1981). Conversions to dry
biomass and carbon equivalent of the wet biomass were obtained using the coefficients 0.2
(Troitsky and Sorokin, 1967) and 0.1 (Luria, 1960), respectively.
RESULTS
Bacterial density of the CF of anthozoans obtained from in situ investigations
Bacterial densities in the CF of L. pruvoti and P. axinellae varied considerably during the
investigation period. Rod-shaped bacterial density ranges between 0.5 to 7.37 X 10 5 cells
ml1 CF (x = 3.09 X 10
5 , S.D. = 2.08 X 10 5 , n = 22) in L. pruvoti, while coelenteric cocci
vary between 2.74 x 10
5 cells m1 CF and 40.19 X 10 5 cells ml- 1 (x = 15.35 X 10 5 , S.D. =
10.3 X 10 5 , n = 22).
In P. axinellae values of rods range between less than 10 3 and 6.1 x 10 5 cells ml-1 (x = 2.13
x 10
5 cells ml -1 , S.D. = 1.36 x 10 5 , n= 22) and of cocci from less than 10 3 to 54.94 x 10 5
cells'ml"
1 (x= 11.7 x 10 5 cells ml- 1 , S.D. = 15.7 x 10 5 , n=22). Mean rod-shaped bacterial
density in ambient water is 1.41 x 10 5 cells ml- 1 (S.D. = 0.4 x 10 s , n= 22) and the mean
density of coccoid bacteria is 2.23 x 10 5 cells ml -1 (S.D. = 0.7 x 10 5 , n = 21). Therefore,
total bacterial density is approximately 6 times higher in L. pruvoti and 4 times higher in
P. axinellae than in ambient waters. Although large deviations of bacterial densities occur
in L. pruvoti and P. axinellae, no distinct diurnal cycles could be observed. Within the
coelenteron of the symbiotic sea anemone S. giganteum, bacterial densities are generally
above those of ambient waters. Mean bacterial density of the CF is 2.49 X 10 6 cells ml* 1
(S.D. = 1.07 X 10
6 ), i.e. 2.6 times higher than in ambient waters (Herndl et ai, in press).
Rods contribute 90% to the number of bacteria and 97 % of bacterial biomass. Mean
rod-shaped bacterial density is 2.3 x 10 6 cells ml -1 (S.D. = 1.1 X 10 6 ) and therefore 14
times higher than the density of rod-shaped bacteria in the adjacent water. The biovolume
of coelenteric bacteria did not differ significantly from that of ambient waters. Mean
volume obtained for rods is 0.6 µm 3 (S.D. = 0.14, n = 100) and for cocci 0.13 µm 3 (S.D. =
0.03, n = 30).
The mean bacterial density in the coelenteron of the symbiotic sea anemone A. sulcata is
2.33 X 10 6 cells ml' 1 (S.D. = 2.2 X 10 6 , n = 7), consisting of up to 90% of cocci. The
volumes of both rod-shaped and coccoid cells of the CF were significantly higher
(ANOVA, P < 0.001) than those of ambient waters.
Bacterial fluctuations in the CF
Figure 1 demonstrates that in P. axinellae the regulative function of the CF against
bacteria depends upon the initial bacterial density in the coelenteron. C. cespitosa
demonstrates very low generation times of coelenteric bacteria at low rod-shaped bacterial densities below 0.1 X 10 5 cells ml -1 CF, but at 0.2 X 10 5 cells ml -1 clearance activity of
the CF already begins (Fig. 2). Coelenteric coccoid bacterial density shows no such
marked fluctuations in turnover rates as do rod-shaped bacteria. During the clearance
phase, however, rods and cocci show quite similar slopes, although clearance of rods
starts at lower bacterial densities than cocci clearance.
In A. sulcata, cocci dominate the bacterial population in the coelenteron during the
course of incubation experiments ; this corresponds with the values of the in situ investi409