Van de wal in de sloot : een typologisch onderzoek aan makrofaunaprocessen

Ditches are an important structural feature of the landscape in the holocene part of the Netherlands. Ditches and other water-courses have been dug to improve drainage and make the land suitable for agriculture.
Since ditches are two to six metres wide and are generally less than a metre deep and seldom shaded, there is enough light to stimulate an abundant growth of water-plants. This growth of water-plants increases the hydraulic resistance in the water-courses. The transport function of the water-courses can be restored by removing these waterplants by manual or mechanical equipment.
The cleaning of ditches generally takes place in autumn, but some water-courses that are particularly important for drainage are cleaned several times between May and October. Ditches are therefore strongly influenced by man and consequently the flora and fauna in these ditches are affected too.
Although Heimans & Thijsse described the richness of their flora and fauna as early as 1896, little research has been done since that time. The research described in this thesis, was undertaken largely because so little was known about the natural flora and fauna in ditches.
The project started in 1975 and was finished in 1981. The principal aims of the research were:
-To investigate whether the macrofauna of ditches shows a certain constancy in time and location, or whether it consists of a variable random collection of individuals or species.
- To find out whether it is possible to construct a typology for ditches with the help of macrofauna.
- To see whether there are correlations between the composition of the macrofauna coenoses and the physical and chemical composition of the water and other factors, such as dimension, flow, the growth of water- and bankplants.
- To find practical applications of the research.

The research was carried out in two areas: The Uithof, which can be regarded as a representative of an area with ditches in river-clay sedimentations and the Oostelijke Binnenpolder of Tienhoven, which is a representative of a peat polder.
The macrofauna was sampled six times a year: a standard-net was used over a total distance of 5 metres for each sample. The net was moved through and samples were taken from five distinct habitats and as many microhabitats as possible. Parallel with the sampling of the macrofauna analyses were made of the vegetation at each sampling station, water samples were taken for physical and chemical analysis and a great number of abiotic measurements were made.
As about 250 taxa with a presence from one to several thousands of individuals had to be processed in each area, we decided to use the computer program BIOPAT (Hogeweg & Hesper 1972, Hogeweg 1976).
It is possible to determine the degree of (dis)similarity of two up to n-samples with the help of the Ward criterion (1963), after projecting n-taxa into a n-dimensional space. The clusters of samples found in this way, can be incorporated in a dendrogram in which the fork-length is a measure of the degree of (dis)similarity (see Fig. 1).

In this processing it is possible to take into account the presence or absence of the species (binary) and their number of individuals.
The numerical analysis can be worked out according to a normalised or a logarithmized scale. With the help of the CLASER part of BIOPAT the minimum number of differential taxa was deduced for a number of the clusters presented in the dendrogram.
The characteristical and accompanying taxa were deduced with the CLUSTA part of BIOPAT, which is used for working out statistical data.
Since little is yet known about the ecology of macrofauna species in general and especially in ditches, we decided to use the same type of processing for all taxa (non-supervised) to form the dendrogram and not to attach more weight to certain species. Other parts of the BIOPAT program were supervised, because they start from the clusters of the dendrogram.

In both areas the clusters of samples can be described in terms of environmental features. The features are as follows:
- Water-courses with a width of more than 10 metres and a water-depth of 0.5-3 metres.
- Organically polluted water-courses.
- Ditches with a standard profile: a talud of 2:3, a width of 4 metres and a waterdepth of less than one metre.
- Old ditches without a standard profile, with a width of 4 metres and a waterdepth of less than one metre.
- Trench-ditches; not always containing water, water-content dependent on the season and water conditions.
After this clustering, the year was divided into seasons and we investigated whether the found division into water types could also be recognised in the various seasons. The divisions into water types, as described above, could be maintained for nearly all seasons, although the difference between the various ditches was not so clear in winter and in early spring.
Since the total macrofauna could give the most information about the "constancy" of the sampling stations per cluster, the separate taxa were lumped together and the individuals of each taxon were added up. This clustering gave almost the same results.
Each taxon was labelled with a seasonal mark in order to cheek whether the division into clusters could be determined by a certain season, or whether the composition of the clusters could be caused by a difference in the development of the macrofauna. After the analysis of this data-set too, it appeared that the same cluster preference was maintained for the sampling locations and that this division showed a similar pattern in all seasons.
With these four ways of analysing, it could be clearly demonstrated that the composition of the macrofauna showed a considerable constancy in time and space. During the seasons and the years, water types could be distinguished on the basis of the composition of the macrofauna. Starting from the clusters of water types, taxa were selected, which typified the water types. These taxa were said to be characteristical or accompanying taxa and were used in the composition of the preliminary typology.

The datasets from the Oostelijke Binnenpolder of Tienhoven were also first analysed separately. In this area it appeared that a group of ditches subject to seepage was permanently separated from the other sampling stations. The rest of the ditches showed only a weak division into two clusters. These two groups of ditches are situated in different parts of the polder and
have a different form of ditch management. One group of ditches is cleaned by a mowing-basket, the other by a ditch-scoop.
A division into seasons was determined in this area too. The data-set for each season showed a division between ditches highly influenced by seepage and the other ditches in de polder. In winter and early spring these other ditches could hardly be differentiated on the basis of the composition of the macrofauna.
The analysis of the data added up for each sampling station and for each year showed a similar pattern. In the analysis with the taxa labelled by season the only interpretable difference was found between the "seepage"-ditches and the other ditches; the weak division in these other ditches almost disappeared.
The characteristical and accompanying taxa were determined for the division on the basis of the year summation. These taxa were used for the composition of the preliminary typology. The preliminary typology of the clay- and peatditches occured to be different.
Since the two typologies might only have a very local validity, their applicability was analyzed with data-sets from the same polders for other periods and with data-sets from other polders.
These analyses showed that the water types distinguished in the research area were generally applicable to river-clay areas and peat areas.

Summing up it can be concluded that macrofauna in ditches has its own peculiar composition and is constant in time and space. For water-courses in river-clay areas and peat areas it is possible to compose a typology with the help of the macrofauna. The water types can also be identified on the basis of abiotic factors.
Some of the results obtained in this research, e.g. those for the way of sampling, the effects of ditch-management and other external influences, might be useful to local authorities and other bodies concerned with ditchmanagement.