Predicting the Distribution of Marine Benthic Biotopes
493
less than 5 m, code these areas with biotope type 1”. These codes were used to prepare
the maps (Figs. 3, 5 and 6) showing the distribution of the biotopes or biotope groups,
by assigning a shade or fill to each individual code.
The boat used for the survey of Loch Long was different to that used for the
rest of the survey due to a low bridge at the entrance of the Loch preventing access by
the main survey vessel. It is generally accepted that acoustic data obtained from
different boats under different environmental conditions may not be comparable,
therefore acoustic signatures were derived separately for Loch Long using only
biotopes found in that Loch, and acoustic signatures for Lochs Duich and Alsh using
only acoustic and ground validation data obtained in those lochs.
The range of data points within different signatures may overlap with each
other. When data are mapped, this may result in some areas with narrow data ranges
being obliterated by others with a very wide spread of data points within a signature.
Therefore manipulation of the signatures for each biotope was required, in particular
for those biotopes with only a relatively small number of associated data points, but a
very wide data range. This was an iterative process, of reviewing the data range for
each biotope, checking the ground validation data to see how well particular records
fitted with the relevant biotope description, possible re-allocation of ground validation
sites to another biotope, then re-calculating the statistics for the particular signature,
and reviewing again.
Some difficulty was encountered in this process, as we found that at this level
of survey, the acoustic data ranges for some biotopes were not distinguishable
separately. This was found where different biotopes have similar life forms (Bunker
and Foster-Smith, 1996), and occur on the same substratum types. An example of this
occurred on steep bedrock and boulders in the Lochs Duich, Alsh and Long pSAC.
Ground validation recorded at least two biotopes on steep rock below the kelp zones:
dead men’s fingers (Alcyonium digitatum) where the rock or boulders were wave
exposed or tideswept, and brachiopods (Neocrania anomala) and solitary sea squirts
(ascidians) where steep bedrock or boulders occurred in sheltered, still water
conditions. These two biotopes are very distinct, yet both consist of animals with
similar form and size, and occur on the same substratum type, albeit in differing
environmental conditions. Unfortunately, the range of acoustic data associated with
the buffers for ground validation sites allocated to these two biotope types was wide,
and overlapping, and we could not distinguish acoustically between the two on the
basis of the acoustic data. This problem was overcome by allocating one acoustic
signature (with a consequently wide data range) to several biotopes with similar life
form, in this case termed ‘deep rock and mixed substrata with faunal turf’.
Tables 1, 2 and 3 relate to the legend of the maps in Fig.s 3, 5 and 6, and
show the MNCR biotope codes (Connor et al., 1996) for each lifeform category used on
the maps. Biotope codes prefixed with DAL or PS are provisional new biotopes
identified during the Duich, Alsh and Long survey, or the Papa Stour survey
respectively. Phase II MNCR forms were completed for these biotopes during our
surveys, and the forms submitted to MNCR for verification.
493
less than 5 m, code these areas with biotope type 1”. These codes were used to prepare
the maps (Figs. 3, 5 and 6) showing the distribution of the biotopes or biotope groups,
by assigning a shade or fill to each individual code.
The boat used for the survey of Loch Long was different to that used for the
rest of the survey due to a low bridge at the entrance of the Loch preventing access by
the main survey vessel. It is generally accepted that acoustic data obtained from
different boats under different environmental conditions may not be comparable,
therefore acoustic signatures were derived separately for Loch Long using only
biotopes found in that Loch, and acoustic signatures for Lochs Duich and Alsh using
only acoustic and ground validation data obtained in those lochs.
The range of data points within different signatures may overlap with each
other. When data are mapped, this may result in some areas with narrow data ranges
being obliterated by others with a very wide spread of data points within a signature.
Therefore manipulation of the signatures for each biotope was required, in particular
for those biotopes with only a relatively small number of associated data points, but a
very wide data range. This was an iterative process, of reviewing the data range for
each biotope, checking the ground validation data to see how well particular records
fitted with the relevant biotope description, possible re-allocation of ground validation
sites to another biotope, then re-calculating the statistics for the particular signature,
and reviewing again.
Some difficulty was encountered in this process, as we found that at this level
of survey, the acoustic data ranges for some biotopes were not distinguishable
separately. This was found where different biotopes have similar life forms (Bunker
and Foster-Smith, 1996), and occur on the same substratum types. An example of this
occurred on steep bedrock and boulders in the Lochs Duich, Alsh and Long pSAC.
Ground validation recorded at least two biotopes on steep rock below the kelp zones:
dead men’s fingers (Alcyonium digitatum) where the rock or boulders were wave
exposed or tideswept, and brachiopods (Neocrania anomala) and solitary sea squirts
(ascidians) where steep bedrock or boulders occurred in sheltered, still water
conditions. These two biotopes are very distinct, yet both consist of animals with
similar form and size, and occur on the same substratum type, albeit in differing
environmental conditions. Unfortunately, the range of acoustic data associated with
the buffers for ground validation sites allocated to these two biotope types was wide,
and overlapping, and we could not distinguish acoustically between the two on the
basis of the acoustic data. This problem was overcome by allocating one acoustic
signature (with a consequently wide data range) to several biotopes with similar life
form, in this case termed ‘deep rock and mixed substrata with faunal turf’.
Tables 1, 2 and 3 relate to the legend of the maps in Fig.s 3, 5 and 6, and
show the MNCR biotope codes (Connor et al., 1996) for each lifeform category used on
the maps. Biotope codes prefixed with DAL or PS are provisional new biotopes
identified during the Duich, Alsh and Long survey, or the Papa Stour survey
respectively. Phase II MNCR forms were completed for these biotopes during our
surveys, and the forms submitted to MNCR for verification.