The effect of grassland botanical composition and (co-) grazing cattle and sheep on herbage production and sward nutrient utilisation

Abstract

Plant diversity in grassland ecosystems has recently gained increasing interest, as plant diversity has often been found to be related to improved ecosystem functioning. So far, studies of diversity effects have mainly been performed in experimental grasslands with randomly assembled species mixtures varying in plant species number. However, there is little knowledge on diversity effects on ecosystem functioning in semi-natural and agriculturally used ecosystems. Grazing has been suggested to be a useful management tool for the conservation and recreation of diverse grasslands. The effect of co-grazing animals on vegetation composition has received little attention up to now. In this study, I investigated initial effects of cattle and sheep grazing alone or in combination (co-grazing) on vegetation development, herbage quality and productivity of grassland differing in initial plant diversity. Furthermore, I studied effects of fertilisation with nitrogen (N) and phosphorus (P) in amounts simulating input by excretion of animal urine and dung. The experiment was established in a semi-natural grassland of the plant-sociological community Lolio-Cynosuretum. This moderately species-rich sward (diverse sward) and its grass-dominated counterpart (grass sward), achieved by application of herbicides, is located in the Solling Uplands (Germany). Both sward types were rotationally grazed from spring to autumn by cattle and sheep grazing alone or together. Cattle-grazing significantly affected vegetation composition in grass and diverse swards. Sheep- and co-grazing had no effect on vegetation composition. The legume abundance was positively related to cattle-grazing, but was reduced in sheep paddocks. Cattle and sheep seem to have adapted their feeding behaviour to the sward composition; sheep were more selective than cattle in both sward types. Dicot abundance was increased by cattle- and co-grazing in the diverse swards and by cattle- and sheep-grazing in the grass swards. Co-grazing did not result in a complementary sward use or in higher plant diversity. Herbage productivity was similar between sward types and was unaffected by grazing treatment, plant species number and functional group proportions. The diverse swards produced forage of better quality. Sheep- compared to cattle-grazing significantly decreased post-grazing herbage quality. The accumulated nutrient uptake by the vegetation over three harvests was similar between sward types. Biomass production affected nutrient uptake. Fertilisation treatment did not affect nutrient uptake, except for an increased N content in the biomass in the high N treatment. In the diverse swards, the contribution of legumes to the N supply of the vegetation was generally small and only marginally affected by fertilisation. The fertilisation treatment had no effect on the abundance of AMF structures. There were hints to higher AMF functioning at higher plant species numbers and conditions of limited P. Our results indicate that vegetation development in grazed grasslands seems to be affected by initial vegetation composition and grazing animal species. Results of experimental grasslands seem difficult to confer to semi-natural grasslands in terms of productivity and nutrient use.