86
3. MEASURING GROWTH AND FORM
in Figs. 3.13 and 3-14 was < 1/27f;this was also observed in H. oculata. In contrast with the observations on H. oculata (Table 3.2) the degree of negative
substrate tropism does not seem to be affected by the change in exposure to
water movement in the two other species.
In Fig. 3-16 notable differences can be seen in the standard deviation
of br_spacing between Pocillopora damicornis and the two other species.
In P. damicornis the standard deviation was relatively low compared to the
two other species, fusion of branches never occurred, and there seemed to be
a mechanism which suppressed growth of branches in the immediate vicinity
of other branches. In the study by Rinkevich and Loya (1985a) it was proposed for the branching scleractinian Stylopora pistillata (see also Sect. 2.2.4)
that there is a chemical signal mechanism which suppresses the growth of
branches when a certain distance to another branch is reached. Their experiments indicated that a chemical signal was possibly secreted into the
water column and worked as a repellent, growth suppressing agent. A similar mechanism might be present in P. damicornis. In Fig. 3.16b it can be seen
that the average br_spacingvalues decreased at increasing exposures to water
movement. A possible explanation might be that the repelling agent was also
dispersed by the hydrodynamic action, which may have diluted the overall
concentration of the chemical signal. In Fig. 3.16ait can be seen that there was
a high standard deviation in branch spacing in Millepora alcicornis; in this
species there does not seem to be an active chemical agent controlling growth
of branches as these tend to grow in proximity to one another and fusion
of branches frequently occurs. This observation corresponds to the observations of Rinkevich and Loya (1985a) in experiments with the branching
hydrozoan Millepora dichotoma. In Fig. 3-16c it can be seen that the variance
in branch spacing and also the probability of fusion of branches increased
with increasing exposure to water movement in Haliclona oculata.
Morphological measurements on projected images are a substantial simplification of reality, and this method was used only for practical reasons.
However, several morphological properties of the organisms may be adequately represented by this approach. The branches, for example in Haliclona
oculata,are formed more or less in one plane. In more complex growth forms,
especially the more compact growth forms of Pocillopora damicornis where
branches often overlap, information is lost by projection. In Fig.3.11b, the
"worst case", it can be observed that due to occlusion effects the individual
branches cannot be distinguished in the inner part of the object. Valid measurements can be made only at the periphery. In Fig. 3.11b it can also be seen
that the morphological skeleton does not give meaningful results in the inner
part of the image; in this area artificial straight lines and junctions are formed
in the thinning algorithm. Fig. 3.11 shows that the method works only partially for objects with a complex three-dimensional geometry; the number
of meaningful results decreases with an increasing degree of compactness of
the objects. For these more complex forms a full three-dimensional analysis
of the growth form is ultimately the only solution. A three-dimensional analysis would also allow determination of the degree of anastomosis, which is
another relevant morphological property of these organisms.
3. MEASURING GROWTH AND FORM
in Figs. 3.13 and 3-14 was < 1/27f;this was also observed in H. oculata. In contrast with the observations on H. oculata (Table 3.2) the degree of negative
substrate tropism does not seem to be affected by the change in exposure to
water movement in the two other species.
In Fig. 3-16 notable differences can be seen in the standard deviation
of br_spacing between Pocillopora damicornis and the two other species.
In P. damicornis the standard deviation was relatively low compared to the
two other species, fusion of branches never occurred, and there seemed to be
a mechanism which suppressed growth of branches in the immediate vicinity
of other branches. In the study by Rinkevich and Loya (1985a) it was proposed for the branching scleractinian Stylopora pistillata (see also Sect. 2.2.4)
that there is a chemical signal mechanism which suppresses the growth of
branches when a certain distance to another branch is reached. Their experiments indicated that a chemical signal was possibly secreted into the
water column and worked as a repellent, growth suppressing agent. A similar mechanism might be present in P. damicornis. In Fig. 3.16b it can be seen
that the average br_spacingvalues decreased at increasing exposures to water
movement. A possible explanation might be that the repelling agent was also
dispersed by the hydrodynamic action, which may have diluted the overall
concentration of the chemical signal. In Fig. 3.16ait can be seen that there was
a high standard deviation in branch spacing in Millepora alcicornis; in this
species there does not seem to be an active chemical agent controlling growth
of branches as these tend to grow in proximity to one another and fusion
of branches frequently occurs. This observation corresponds to the observations of Rinkevich and Loya (1985a) in experiments with the branching
hydrozoan Millepora dichotoma. In Fig. 3-16c it can be seen that the variance
in branch spacing and also the probability of fusion of branches increased
with increasing exposure to water movement in Haliclona oculata.
Morphological measurements on projected images are a substantial simplification of reality, and this method was used only for practical reasons.
However, several morphological properties of the organisms may be adequately represented by this approach. The branches, for example in Haliclona
oculata,are formed more or less in one plane. In more complex growth forms,
especially the more compact growth forms of Pocillopora damicornis where
branches often overlap, information is lost by projection. In Fig.3.11b, the
"worst case", it can be observed that due to occlusion effects the individual
branches cannot be distinguished in the inner part of the object. Valid measurements can be made only at the periphery. In Fig. 3.11b it can also be seen
that the morphological skeleton does not give meaningful results in the inner
part of the image; in this area artificial straight lines and junctions are formed
in the thinning algorithm. Fig. 3.11 shows that the method works only partially for objects with a complex three-dimensional geometry; the number
of meaningful results decreases with an increasing degree of compactness of
the objects. For these more complex forms a full three-dimensional analysis
of the growth form is ultimately the only solution. A three-dimensional analysis would also allow determination of the degree of anastomosis, which is
another relevant morphological property of these organisms.