LAKE DWELLING TRICLADS
11
correlation positively significant at the P = 0-02 level. The correlation
between winter and eit,her spring or summer was l ~ s i t i v e but not significant. The species composition of each triclad fauna was identical in
each season. Lakes in Snowdonia have also been sampled several times
and given comparable estimates of population size and the species composition has always been the same.
A further potential drawback with a single sample is possible variation in the size of the population over the year. This was minimized by
restricting routine sampling to the summer months and in any case
seemed to be an unlikely source of large error because study of the population dynamics of the planariid species (Reynoldson, 1980, 1961a;
Taylor and Reynoldson, 1962) has shown that annual changes from
births and deaths were around 50%; they were larger in Dendrocoelum
lacteum however, up to 500%. This argument was supported by the data
from Anglesey lakes which showed only small variation in amplitude
from season to season. It seems reasonable to conclude that this simple
method of estimating populatioii size and species composition in triclads was adequate for the purpose of examining trends in distribution
and abundance. Clearly, it is insufficiently sensitive to compare individual lakes, especially those close together on the chemical spectrum
and it has not been used for this purpose.
During the search for triclads, other organisms found on the stones
were noted or collected for identification. Samples of water for chemical
analysis were taken at the same site lts the triclad sample. Again, only
single samples were taken and what has been said about the reliability
of the triclad method applies here but there is no doubt that the chemical samples have a narrower range of variability as the Anglesey data
showed (Reynoldson, 1958a). It has also been possible to compare our
measurements of various ions with those of other workers (e.g. Gorham,
1957, on the Shropshire meres) and there was reasonable agreement.
Further, the data of Tucker (1958) and Olsen (1955) indicated that the
range of values over a season would not substantially alter the ranking
of lakes within a season. Of the several analyses, calcium and total dissolved matter have been most illuminating. In considering anomalies it
has sometimes been useful to look a t other chemical features such as for
example, magnesium, chloride, oxygen uptake from permanganate. An
indication of the range in chemical qualities of the lakes is given in
Table I.
Under favourable conditions five or six lakes could be sampled in one
day and still leave time for examination of material in the evening; triclads were generally identi6ed on the day of collection. Dugesia lugubris,
Dendrocoelum lacteum and Phagocuta vitta are identifiable by eye or lens,
but with Polycelis species the large specimens had to be squashed and
11
correlation positively significant at the P = 0-02 level. The correlation
between winter and eit,her spring or summer was l ~ s i t i v e but not significant. The species composition of each triclad fauna was identical in
each season. Lakes in Snowdonia have also been sampled several times
and given comparable estimates of population size and the species composition has always been the same.
A further potential drawback with a single sample is possible variation in the size of the population over the year. This was minimized by
restricting routine sampling to the summer months and in any case
seemed to be an unlikely source of large error because study of the population dynamics of the planariid species (Reynoldson, 1980, 1961a;
Taylor and Reynoldson, 1962) has shown that annual changes from
births and deaths were around 50%; they were larger in Dendrocoelum
lacteum however, up to 500%. This argument was supported by the data
from Anglesey lakes which showed only small variation in amplitude
from season to season. It seems reasonable to conclude that this simple
method of estimating populatioii size and species composition in triclads was adequate for the purpose of examining trends in distribution
and abundance. Clearly, it is insufficiently sensitive to compare individual lakes, especially those close together on the chemical spectrum
and it has not been used for this purpose.
During the search for triclads, other organisms found on the stones
were noted or collected for identification. Samples of water for chemical
analysis were taken at the same site lts the triclad sample. Again, only
single samples were taken and what has been said about the reliability
of the triclad method applies here but there is no doubt that the chemical samples have a narrower range of variability as the Anglesey data
showed (Reynoldson, 1958a). It has also been possible to compare our
measurements of various ions with those of other workers (e.g. Gorham,
1957, on the Shropshire meres) and there was reasonable agreement.
Further, the data of Tucker (1958) and Olsen (1955) indicated that the
range of values over a season would not substantially alter the ranking
of lakes within a season. Of the several analyses, calcium and total dissolved matter have been most illuminating. In considering anomalies it
has sometimes been useful to look a t other chemical features such as for
example, magnesium, chloride, oxygen uptake from permanganate. An
indication of the range in chemical qualities of the lakes is given in
Table I.
Under favourable conditions five or six lakes could be sampled in one
day and still leave time for examination of material in the evening; triclads were generally identi6ed on the day of collection. Dugesia lugubris,
Dendrocoelum lacteum and Phagocuta vitta are identifiable by eye or lens,
but with Polycelis species the large specimens had to be squashed and