Numerous experiments, including Park Grass, have demonstrated that low levels of soil fertility are associated with high species co-existence on a wide range of semi-natural lowland grasslands (Janssens et al, 1998).

One of the most important constraints to restoration is therefore the high residual soil fertility associated with arable conversion and improved grassland.

Experiments at Trawsgoed and Pwllpeiran have demonstrated that grassland restoration may take place more rapidly on shallow free-draining upland soils, where there is a relatively high rate of leaching (Hayes et al., 2000; Hayes and Sackville Hamilton, 2001). The cessation of fertilisation, and reinstatement of hay cutting and grazing may therefore be sufficient to gradually reinstate grasslands in upland sites.

Experiments at Colt Park suggest that the restoration of upland hay meadows using such extensive management alone may take over 20 years (Smith et al., 2002). The rate of reversion may be accelerated by the removal of two hay crops, which helps to accelerate reductions in residual soil fertility. Subsequent aftermath grazing has also been shown at Trawsgoed to open up the sward, creating germination gaps for for seed from the hay crop (Smith et al., 2000).

By contrast, hay cutting without grazing has been shown to favour coarse grasses, whereas grazing alone encourages the establishment of undesirable weed species (Hayes and Sackville Hamilton, 2001).

Where the effects of residual fertilisers on competitive interactions are more pronounced, and there is a relatively impoverished or isolated seed bank, the rate of diversification and grassland reversion may be significantly limited. It has been estimated that it would take 70-90 years for the fertilised plots at Park Grass to revert to their previous composition (Dodd et al., 1994).

Successful restoration must therefore consider the soil nutrient and pH levels at the site to be restored, and the extent to which these differ from semi-natural reference levels, such as those of local semi-natural grassland or BAP priority grassland (Critchley et al., 2002). These suggest that semi-natural levels of P and exchangeable K are in the range 4-11 mg/l and 76-210 mg/l, respectively. pH levels range from 4.9-6.1 on acid grasslands, 6-6.4 on neutral grasslands and 6.8-7.9 on calcareous grasslands.

Experiments on brown soils in west Wales have shown that where soil extractable P are less than 10 mg/l above desired NVC reference levels, hay cutting and extensive grazing may be sufficient to restore grasslands in less than a decade. Where P levels are >10 mg/l higher than NVC reference levels, more interventionist techniques, such as deep cultivation may be necessary (Hayes et al., 2000; Hayes and Sackville Hamilton, 2001). Further work, however, is needed to understand the applicability of these levels to other areas, given the high variability of soil types and phosphate levels in the soil.

Reducing fertility


The relationships between plant communities and the species that feed on them or on their associated herbivores, is highly complex.

Experiments can provide a microcosm for study, giving an insight into invertebrate life cycles, and how they respond to changes in their hosts and habitats.

SOIL pHRESTORATION_plceholder_2.html

Monitoring of lowland heathland at Thursley Common, following the cessation of nitrogen fertilisation experiments, demonstrates that vegetation growth, soil chemistry and soil microbial communities showed persistent effects of eutrophication.

Although habitat management techniques, including mowing and high intensity burning, reduced nitrogen in the plant biomass, effect on below ground nitrogen stores were limited. The residual effects of nitrogen on heathland may therefore persist for several decades (Power et al., 2006).


Experiments at Tealham and Tadham Moor, on the Somerset Levels, have also demonstrated the difficulties of restoring wet grasslands following nitrogen addition.

15 years after the cessation of experimental treatments,  differences in soil chemistry and vegetation species composition were still detectable in plots where treatments exceeded 50 kg/ha.

Where nitrogen deposition on heathland has resulted in a shift in community composition to favour grasses, the higher nitrogen cycling rates associated with grass litter may be self-reinforcing and prevent natural recovery (Terry et al., 2004; Power et al., 2006)

In such cases, heathland restoration may require active removal of  organic layers by sod-cutting (van den Berg et al., 2003; Dorland et al., 2004).



SOIL pHRESTORATION_plceholder_2.html


Experiments comparing the effectiveness of grazing, cutting and fertilisation have demonstrated that it is necessary to cease fertilisation coupled with grazing to increase the biodiversity of such grasslands over the long term (Willems 1983; Bakker 1987; 1989).



Arable cropping has been used to accelerate the decline in soil fertility by accelerating the removal of nutrients, most notable phosphorus, which naturally leach relatively slowly from the soil.

Experiments at Trawsgoed, however, demonstrate that increased take off of nutrients within the crop may be compensated for by an increase in the mineralisation of organic phosphorus in the soil humus (McCrea et al., 2001).

Cropping alone may therefore be ineffective in significantly reducing soil phosphorus (Marrs, 1993; Marrs et al., 1998).




The adsorption of phosphate by oxides of iron and aluminium is recognised as a method for reducing soil phosphates (Wild 1988). Aluminium sulphate has been demonstrated to be effective on improved pasture in Wales (Adams et al., 1999), and reduce soil phosphorus on a sandy soil in East Anglia.

However, due to the potential for toxic side effects, chemical treatments are not deemed suitable for sensitive sites.

JOIN USRESTORATION_fertility_files/ECT%20flyer.pdf