< Previous | Contents | Next >

Capercallie: A Review of Research Needs

3 RESEARCH PRIORITIES

3.1 A general approach

Clearly, targets set in the Capercaillie Biodiversity Action Plan (2.1.1) have not been met, or are unlikely to be met by the target dates if the current rate of decline is maintained. With only about 1000 birds left, there may be little time left to turn the population around, even if that can be achieved. The time factor is important because if the current rate of decline is maintained there will only be around 250 capercaillie by 2008/09. Moreover, there is a danger that any time lag associated with undertaking research, implementing the findings and of these having the desired effects on habitats will simply be too long. Therefore, in addition to some new research (below), a more pragmatic targeted approach to improving habitats is urgently required (see section 4 and below).

Most of the current research on capercaillie is being undertaken by the RSPB at Abernethy. This includes studies on (i) the impact of predator removal on capercaillie breeding success and numbers, (ii) the influence of forest structure on the ecology of blaeberry and (iii) field trials of capercaillie-friendly fences.

3.2 Research priorities

One problem in conducting research on capercaillie when numbers are low and declining, is the lack of birds to work with. This effectively rules out some lines of investigation, such as any study that involves using large numbers of radio-tagged birds. This has been one of the most important techniques available to scientists, and has been responsible for advancing our knowledge about little-known aspects of capercaillie ecology in Scotland and elsewhere. Thus, the main thrust of new research should be towards a better understanding of how habitats can be improved.

There is general agreement that low breeding success is the main cause of the current decline in capercaillie numbers, although increased mortality from fence collision may be a contributory factor. One concern is that climatic factors associated with poor breeding success may override any beneficial effects of improved management resulting from further research. However, there are clearly numerous habitat-related factors linked to poor breeding success and a better understanding of these should lead to improvements in how forests are managed for capercaillie.

In the box below, nine research topics are listed. The order of priority is somewhat flexible, but generally those at the top of the list should have greater priority to those at the bottom. Some of the proposals are surveys rather than research, but it is appropriate to include these in this section rather than in the next. A more detailed rationale for each topic is given below and summarised in Table 4.

3.2.1 Factors associated with high breeding success

(Importance 9/10: Likelihood of success 9/10: Tractability 7/10: Man-years 1 (plus T&S): Timescale <1 year)

Research has shown that breeding success in highly variable among years and forests, but that some woodlands consistently produce more chicks than other (Baines et al., Manuscript). This project would look in finer detail at those forests that consistently produce more chicks (top 50%) in the same sites used by Baines et al. (Manuscript), to see if there are any common factors that might explain higher breeding success. The islands of Loch Lomond should be an additional site. While this site has relatively high rainfall it still manages to maintain a small fluctuating population (Jones, 1990; Waltho, 1999) (2.2.3). Factors to be examined would include those relating to forest composition (tree species and age class distribution etc), the extent of brood habitats, predator abundance (red fox, pine marten, crow and buzzard) and the effectiveness of any predator control.

3.2.2 Inventory of deer fences in key sites

(Importance 8/10: Likelihood of success 10/10: Tractability 8/10: Man-years = 2 (plus T&S): Timescale 1 year)

More effort needs to be put into removing existing deer fences in key sites (Figure 2 & section 4). To facilitate this an inventory of existing fences is essential so that the scale of the problem can be properly assessed. This should include fences on FC and non-FC ground. The inventory should record the location of each fence and use a classification system that ranks fences according to their perceived danger to capercaillie. It should also record which fences have been erected over the last 5 years and the length, location and rank of fences removed over the same period. This will provide data about whether collision risk is increasing or decreasing.

3.2.3 Large-scale management package trial in key sites

(Importance = 8/10: Likelihood of success 7/10: Tractability = 6/10: Man-years = 5+ (plus T&S and materials): Timescale 3 years)

This is an essential project to assess the value of a proposed management package for capercaillie (4.3). This would be a forest-scale management trial rather than a strict experiment with controls. Even so, it is important to be able to demonstrate that the package has a beneficial impact on capercaillie. Therefore, it is proposed that three pairs of forests should be selected, with each pair similar in size and forest composition. One forest from each pair would be selected to trial the package. The six forests should lie within key areas and have capercaillie present, and at least two of the forest pairs should comprise largely managed forest (plantations of Scots pine and other conifers). The project is dependent on finding six forests with sympathetic owners. Part of the work would involve developing simple methods to monitor the impact of different elements of the package, both on the habitat and on capercaillie.

3.2.4 The role of brash in plantations

(Importance = 8/10: Likelihood of success 7/10: Tractability = 7/10: Man-years = 2.0 (plus T&S and materials): Timescale 5 years)

Observations indicate that capercaillie broods appear to select Scots pine plantations that have been recently thinned and even small clear-cuts where there is brash and blaeberry. This project will use a replicated experimental design in at least two widely separated forests (ideally with capercaillie present) in an attempt to discover why such areas are preferred. The experiment should use two treatments (thinned with brash left on the ground; thinned with brash removed) and a control (unthinned). Site factors to be assessed would include ground vegetation and brash, nutrient status of blaeberry in late April (important for maternal nutrition) , species richness and biomass of arthropods in the field layer from mid-May until the end of June (important for chick condition and survival) . Assessment should start in the spring before the treatment plots are thinned (year 1) and be repeated annually until year 5, after which the possible extension of the monitoring would be reviewed.

3.2.5 Developing capercaillie-friendly fences

(Importance = 7/10: Likelihood of success 7/10: Tractability = 8/10: Man-years = 1 (plus materials): Timesale 1 year)

It is unlikely that any improved fence design will stop bird collision, and so the emphasis should correctly be on limiting deer numbers by culling. However, there will be situations where fences are going to be required because there is no other satisfactory way of establishing trees. Marking deer fences with orange roadside barrier netting has been shown to reduce grouse collisions, but it does not stop them. One problem with designing capercaillie-friendly fences is that fence visibility is judged from a human visual perspective. Capercaillie may see quite differently, as vision in many bird species extend into the ultraviolet spectrum (c.f Siitari et al., 1999). Therefore, it is recommended that research is commissioned into (i) the visual ability of capercaillie, including their use of UV wavelengths and whether this can be used to design fences that can be seen more easily, and (ii) reviewing the plethora of novel fence designs currently under trial, mainly by the RSPB (Goddard et al., No date; Dugan et al., 1999), and the use of electric fencing. The contractor should also look at how fences are currently monitored for bird collisions and produce a standard methodology.

3.2.6 Maternal nutrition

(Importance = 8/10: Likelihood of success 9/10: Tractability = 8/10: Man-years = 1 (plus T&S and materials): Timescale 3 years)

The quality of food available to females prior to laying may be important in determining breeding success (Moss et al., Submitted). This project would examine the diet of females in April by faecal analysis (Picozzi et al., 1995) so that management could be directed towards encouraging the most important plant/tree foods. The project might be more effective with a small sample of radio-tagged females rather than relying on flushing birds at random to obtain droppings, but could be carried out using either technique. The main advantage of using radio-tagged birds is that breeding success could be directly related to diet prior to laying. This would entail catching a sample of chicks in July/August of the year prior to work commencing and re-catching them once fully-grown to fit longer-life radios.

3.2.7 Managing ground vegetation in semi-natural pinewoods

(Importance = 8/10: Likelihood of success 8/10: Tractability = 7/10: Man-years = 4 (plus T&S and materials): Timescale 4 years)

The grazing impact of red deer on dwarf shrub vegetation is difficult to manage. In woodlands with heavy grazing pressure there is often much blaeberry, but in common with the rest of the vegetation it is short and offers little shelter or food for broods. In contrast, in some pinewoods the vegetation is becoming increasingly long and dominated by heather, the result of either excluding deer by fencing or by reducing densities through culling. Capercaillie appear to prefer a mosaic of long and short vegetation, with the short vegetation providing food and access for chicks and the longer vegetation cover. There are three methods that could help to create the small-scale mosaics that capercaillie prefer; burning, swiping and winter grazing with cattle (2.5.3). This work should be based on an experimental design with replicated control and treatment plots monitored before and for four years after treatment. Indirect methods, similar to those in 3.2.4, would be used to assess the suitability of each treatment for capercaillie. However, plot sizes need to be considerably larger due to the cattle treatment, which may need to be fenced.

3.2.8 Monitoring population size and range

(Importance 9/10: Likelihood of success 9/10: Tractability 9/10: Man-years 0.1: Timescale <0.1)

A review of population monitoring is required because the sampling method used in the 1992-94 and 1998-99 surveys may not be appropriate for a repeat survey if the population continues to decline. With the population so low, it appears important for monitoring to continue at around 5-year intervals. The sampling procedure selected should provide data that are comparable with previous surveys (Catt et al., 1998; Wilkinson et al., 1999), but more sensitive to lower numbers. Therefore, it is likely to be based on the existing distance sampling method (Buckland et al., 1993), which involves counting birds along transects undertaken during November-March, when most birds are feeding in trees. Stratification for the previous two surveys was based on geographical area and abundance class. Data from these surveys should be used to design a new sampling protocol.

3.2.9 Composition of forests in the core and extension areas

(Importance 6/10: Likelihood of success 9/10: Tractability 9/10: Man-years 0.1: Timescale <0.1)

An analysis of forest composition (tree species and age class distribution) in the core and extension areas would provide a useful overview on the extent of suitable capercaillie habitat and complement the work of Summers et al. (1999) . This analysis could be further split by the four zones shown in Figure 2 or by any other zonation. The National Inventory of Woodland and Trees is now complete for Scotland. This uses the LCS88 data updated to 31 March 1994 for woodlands over 2 ha. From this database predictions could be run to show future scenarios at any time interval. It may also be possible to look at how forest composition has changed since the last survey of woodland in 1979-82 (2.2.2).

< Previous | Contents | Next >