Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Individual study: Application of nitrogen fertilizer to make grasslands more attractive as feeding areas for pink-footed geese Anser brachyrhynchus at Loch of Strathbeg RSPB Reserve, Aberdeenshire, Scotland

Published source details

Patterson I.J. & Fuchs R.M.E. (2001) The use of nitrogen fertilizer on alternative grassland feeding refuges for pink-footed geese in spring. Journal of Applied Ecology, 38, 637-646

Summary

In the UK, increasing populations of wintering geese Anser spp. may cause agricultural damage by grazing on pastures and winter-sown crops. In some areas alternative feeding refuges have been established which, if made as attractive as possible to geese, are effective at reducing damage to nearby agricultural fields. An experiment was undertaken on a Royal Society for the Protection of Birds (RSPB) Reserve in north-east Scotland to investigate the effectiveness of different application rates of nitrogen fertilizer in the spring on the amount of pink-footed goose Anser brachyrhynchus grazing.

Study site: The study was undertaken during spring 1990 and 1991 at the Loch of Strathbeg RSPB resrve. Within the reserve are 18 or so grass fields (dominated by perennial rye-grass Lolium perenne) used as roost and grazing areas by around 30,000 pink-footed geese in the spring. These fields were sheep and cattle grazed in summer, when they also recieved applications of inorganic fertilizer.

Spring fertilizer application: Application was undertaken in one field in 1990 and an adjacent one in 1991. In 1990, the field used had a drier upper area and lower wet area. It was marked out in a randomized block design with three blocks comprising five, 12 m wide by 100 m long strips, representing four treatments and a control. The three blocks were replicated on the upper and lower field sections (six replicates).

In 1991, the experiment was conducted on a grass sward subject to a mowing regime in the summer of 1990 with grass cut when it reached 12 cm and 30 cm, with cuttings left on the field. The experiment was confined to the upper part of the field as the lower section was waterlogged.

Nitrogen, phosphate and potash are normally applied during the growing season but in this experiment only nitrogen was used. This was because a direct growing response to phosphate is only expected in areas where the soil has low levels of this nutrient (not the case in the fields used), and use of potash was avoided as this could have led to hypermagnesemia problems in cattle.

Nitrogen (N) was applied as ammonium nitrate (34.5% w/w N), this form being commonly used on grasslands in spring to encourage early leaf growth. The fertilizer in prill form (spherical granules) was spread using a tractor-mounted pneumatic spreader with a 12 m boom.

On 5 March 1990, fertilizer was applied at four rates: 0 (control), 40, 80, 120 and 160 kg N/ha. In 1991, despite double rear wheels on the tractor to reduce ground pressure to minimize damage, wet weather delayed application until 18 March. Due to erroneous application, the rates applied were substantially increased; 72, 144, 216 and 288 kg N/ha. In north-east Scotland such high applications would be considered excessive, with usual rates being about 50 kg N/ha.

Goose dropping densities: Dropping density was used to estimate goose grazing intensity. Five permanent 5 m² quadrats were established in each strip. Dropping counts were made immediately prior to the experiment to test for variation between areas selected for the treatments, and subsequently at about 2-week intervals from time of fertilizer application to the end of the dropping count census. Each quadrat or sample area was a circle with 1.26 m radius; all droppings were counted and removed. Cumulative dropping totals were calculated.

Grass production: Grass production was measured by erecting exclosures made from rabbit-proof wire mesh 0.5 m high x 4 m long, folded to produce a 1 x 1 m square exclosure with a mesh lid. In 1990 three, and in 1991 four, exclosures were erected in each strip immediately after fertilizer application and left until the grass was harvested. The grass was mown using a hedge trimmer. At the end of April, each exclosure was removed and the first peripheral 10 cm of grass was cut and removed to avoid any edge effect. The trimmer was then pushed through the grass to achieve a cut of about 2 cm high, similar to the sward height in the goose-grazed areas. The grass samples were dried; the weight of each was then measured and the dry matter (DM) or grass production under each exclosure calculated as grammes DM/m².

Stock were excluded from the field during the course of the experiment and the only other large grazing animals present were rabbits Oryctolagus cuniculus.

Dropping densities: In 1990, droppings in the quadrats were counted and cleared on the day that fertilizer was applied (5 March). Most quadrats contained no droppings. Over the study period in 1990, there was significant variation in cumulative dropping density with fertilizer treatments in both the upper and lower field areas (Table 1, attached). Interestingly, at the 80 kg N/ha rate compared with the higher rates there was no significant increase in dropping density.

In 1991, there was no significant variation in dropping density between the experimental strips prior to the start of the fertilizer treatment. Cumulative dropping density at the end of April varied significantly with increases detected in the control and all N levels, but within fertilizer treatments only between 72 and 288 kg N/ha.

Cost effectiveness: Taking the data from both years, the cost effectiveness of each application rate was estimated from the increase in dropping density relative to the zero application in the control, per kg N/ha applied. The lowest application rates gave the greatest percentage increase in goose grazing per kg N/ha

Grass production: In both 1990 and 1991 grass production broadly mirrored that of cumulative goose dropping densities, with dry weight production of grass varying significantly with fertilizer level. The results are summarised in Table 1 (attached).

In 1990, on the upper field area, there were significant increases in production between the control and the three highest levels of fertilizer treatment but no differences between the different fertilizer treatments. On the lower area again there were significant increases in production between the control and the three highest levels of fertilizer, and also between the 40 kg N/ha and the three higher treatments. The apparent decrease in grass production at 160 kg N/ha compared with the 120 kg N/ha rate was not significant. In 1991, significant increases were detected between the control and the three highest fertilizer levels but no differences between the fertilizer levels.

In both 1990 and 1991, goose dropping density increased significantly with increased levels of grass production, and in both years the swards were heavily grazed to a height of around 20 mm.

Conclusions: In this study, the amount of goose grazing (as measured by the cumulative dropping density) increased with fertilizer application rate up to around 80 kg N/ha and mirrored a similar increase in grass production. Above this application level there was little further increase and cost-effectiveness tailed off rapidly. Thus application of N fertilizer up to around 80 kg/ha at this site appears to be an effective way to greatly increase spring goose-grazing on grassland refuges through increasing the amount of grass available to the geese. This in turn may be a means to reduce damage to cereal crops and other farmland crops in the vicinity by encouraging the geese to remain within these refuge areas.


Note: If using or referring to this published study, please read and quote the original paper.