Problems in the Field

How to prevent animals and insects from falling into the traps.
A number of investigators have reported that small mammals, insects and leaves fell into the trap during the course of the year. Most of this material presents no difficulty (apart from the aesthetic problem of handling the small mammals) providing that preservative has been added at the outset, but where this has not been added there is a problem of decaying animal remains, and the attendant presence of detritus feeding beetles as well as fungal growth.

One way to exclude small animals is to place a net either over the top of the trap or inside the trap just below the opening. The problem with a net is the danger that leaves will fall onto it, get stuck and thus block the trap. This is particularly a problem in mixed deciduous woodland with large-leafed trees such as Acer and Castanea. Wires stretched just above the trap may also help. Preliminary results of a monitoring experiment (Pardoe, in N. Wales, pers. comm.) in which there are parallel traps, one set covered with a coarse plastic net and one without, suggest that the presence of nets - at least in areas with rather open vegetation and no large-leafed trees - does not affect the composition of the pollen assemblage.

The addition of both thymol and formalin, as recommended in the original guidelines is obviously essential (NB however, the addition of too much concentrated formalin will 'fix' the pollen, thus preventing the removal of pollen grain contents during laboratory treatment and making identification difficult).

Several types of the insects, which may get caught in traps (especially bees and butterflies), bring with them non wind-pollinated pollen. If such insects are present, then the count should be evaluated critically and any large quantities of obviously entomophilous pollen deducted from the total. In these cases reference to earlier years' catches gives an indication of the 'normally' expected occurrence of such insect pollinated taxa. It is feasible that colouring the top of the trap blue might deter insects although the effectiveness of such treatment has not been tested.

What to do with traps which are full and apparently overflowing by the end of the field season.
Some investigators have recorded that certain of their traps have been so full of water that they have overflowed. There was a feeling that many of these traps had filled with surface water during times of high water table and this conclusion was supported by the presence of lower water volumes in adjacent traps. Care should, therefore, be taken to avoid flood-prone locations. If the overflowing is not due to flooding, then simply using a trap with a larger volume may solve the problem. If possible traps should be visited during the year to check on their water content. Some water can be drawn off mid-year, stored until the time of normal trap collection, and then added again before the material is processed. Alternatively, the trap can be emptied and replaced and a separate preparation and count made and the results amalgamated at the end of the year. This is time consuming but may be of value if there is interest in studying pollen variation during the year.

If Lycopodium spore tablets are added when the trap is first placed in the field, theoretically it should be possible to correct influx figures for any loss brought about by flooding. There have been no experiments, however, to see how the Lycopodium spores behave during the course of the year. It is possible that they may deteriorate or conversely that they become waterlogged and therefore may be better prepared for laboratory treatment. It has been demonstrated that Lycopodium spores may be lost during the laboratory preparation of samples if dry Lycopodium tablets are added directly to the pollen samples without prior soaking. Until the results of current experiments are available, it is recommended that Lycopodium tablets should be added to the traps at the start of the field season only where duplicate traps are available.

Disturbance of Traps
Some investigators have reported the theft or vandalism of pollen traps with the inevitable loss of a whole year of data from that site, leading to disappointment and frustration and a gap in their data series. It is suggested that, where possible, remote or protected sites should be chosen for the location of pollen traps and that in national parks etc. the co-operation of park wardens should be enlisted. Where this is not possible the provision of duplicate traps could be considered in the hope that one will survive. The recommendation for fencing and labelling the traps, which is given in the original guidelines, is by no means obligatory and each researcher should consider the relevance of this in his/her area. While in the wilds of Lapland a fence and notice may be the only means of relocating a trap, in the mountain areas of the Alps and the Mediterranean it may act as a 'flag' enticing both local inhabitants and tourists to interfere.

Location of Pollen Traps in Relation to the Surrounding Vegetation
As emphasised in the original guidelines; 'The minimum requirement for participation in the programme is a transect of three pollen traps placed:

However, as mentioned on p.10 (op. cit.), 'Traps in the forested part of each transect should be placed in an opening within the vegetation being monitored' and 'Ideally the opening should be large enough so that predominantly regional rather than local pollen is trapped.' This means that the 'in the forest' trap is not intended to be placed actually beneath the trees forming the forest canopy. The rational behind this perhaps needs further explanation.

As the ultimate goal of the EPMP is to provide a more detailed and accurate interpretation of the vegetation represented by fossil pollen records, it is important that modern pollen traps are located in areas which collect pollen in a manner similar to the mires (and lakes) which existed in the past. This means that locating traps in openings in the forest canopy provides more appropriate information than if traps are located beneath a closed forest canopy. Different investigators will have varying priorities in terms of the size of clearings they choose for their traps depending on the type of fossil pollen sites they are currently studying. The size of this opening is important, as it will determine the source area of the arboreal pollen. Traps located in larger openings (30m or more in diameter) are more likely to collect the regional pollen rain and therefore provide an analogue for small mires that might have existed in the past. The choice should always be determined by the particular fossil situations the investigator wishes to interpret.

Various experiments to determine the degree to which the position of the trap affects the pollen deposition values can be envisaged. One sampling strategy could be to locate a series of traps at regular intervals from the edge of the forest to the centre of a large opening in the canopy. Another possible model is to locate a series of traps in canopy openings of different sizes.

Vegetation Mapping
As recommended earlier (Hicks et al. 1996) for the modern pollen results to be of any use it is essential that both the local vegetation in the immediate vicinity of the trap, and the vegetation in the wider region surrounding the pollen traps is numerically recorded. Only in this way can pollen influx values be related to the relative source area (Sugita, 1998) and ultimate catchment area of the traps. However, the relative source area for each trap will differ and there is, as yet, no reliable way of calculating what it is (Bradshaw and Jacobson, 1981, Prentice, 1985, 1986, 1988, Sugita, 1993, 1994, 1998, Calcote, 1995). It is not possible to stipulate the size or shape (see e.g. Hicks et al. 1997) of the surrounding area for which vegetation analyses are to be made. Initially on-site vegetation analyses, using a routine field method such as the Braun-Blanquet or Domin system, can be used to provide a representative view of the local vegetation. These can then be supplemented with vegetation classifications using remote sensing techniques for analysing digitised air photos and satellite images (Pellikka, 1997). Experiments with mathematical models for estimating the size of the catchment area (Prentice, 1986, Sugita, 1998) can then be applied to the digitised material. Preliminary trials (Valta pers. com.) indicate that although this may be achieved in reasonably flat terrain with homogenous vegetation it is not possible in mountainous areas or across ecotones. Ideally one would want to relate pollen influx of a taxon to the density of the relevant species in the vicinity. This figure would provide a strong tool in interpreting species migration and the position of tree-lines from fossil data.

Collection of Moss Polsters as Additional Surface Samples to Complement the Trap Results and Allow Comparisons With Other Projects (Hicks et al., 1996 p.7)
The original recommendation was that the moss polster should be collected when the trap is established. If this was not done, then sampling annually when the trap is emptied, or at the end of the trapping programme, is acceptable. Two complementary strategies are recommended.

  1. A single sample of 5 cm diameter (i.e. the same size as the trap opening) collected from adjacent to the trap.
  2. A composite sample consisting of a minimum of ten small, equal-sized sub-samples of moss (approximately a handful when combined).

The sub-samples should be collected ideally from within a 5m radius of the pollen trap. Care should be taken to avoid moss covered with soil. The moss samples should consist of living and dead plant material only; soil should be removed, preferably in the field, using scissors. Where possible, the situation of the moss should be comparable to the trap opening, i.e. it should be growing on the ground surface, rather than on vertical trunks or rocks. Plant litter or lichen can be collected if no moss is available. The sub-samples should reflect the range of species and life-forms growing in the area around the trap. The sub-samples should be mixed thoroughly prior to laboratory preparation. The moss polsters can be stored in sealed plastic bags, then frozen or air-dried if it is necessary to postpone laboratory preparation, but care should be taken to prevent contamination of the samples by airborne pollen. See also Hicks et al. (1998)

Top of page..
Back to Table of Contents