A simulation model of the effect of mowing of agricultural grassland on the breeding success of the corncrake (Crex crex)

Abstract

The corncrake has been declining in numbers and distribution in Britain for more than 100 years. Meadows mowed for hay or silage are the most frequent habitat and nests and broods placed there are often destroyed during mowing. Observations of radio‐tagged females and watches of meadows during mowing are used to describe the timing of breeding, the duration of various stages of the breeding cycle and mortality rates of clutches and broods. It is shown that females frequently produce two clutches per season and that clutch and brood survival are high in the absence of mowing. The majority of nests are placed in meadows liable to be mowed during the breeding season and this proportion increases during the season. Mowing from the outside of the meadow towards the centre causes greater mortality of corncrake chicks than mowing from the centre outwards. A simulation model is developed to permit productivity (chicks reared to independence per female) to be estimated for a range of average mowing dates and for mowing from the outside inwards and from the centre outwards. The model indicates a large effect of the average date of mowing on productivity and only a small effect of the spread of mowing dates. Mowing from the centre of the field outwards is estimated to result in substantial increases in productivity via improved chick survival, especially when the average mowing date is early. Differences in productivity, estimated from average mowing dates, between regions of Britain and Ireland with stable or slowly declining corncrake populations and those with rapidly declining populations, are approximately large enough for differences in mowing practice to be the principal cause of the observed difference in population trend. Current conservation programmes for corncrakes in Britain and Ireland, which include payments to farmers to delay mowing and to mow from the centre of the field outwards, are evaluated using the simulation model. The improvement in productivity caused by the conservation measures should be sufficient to halt the long‐term population decline provided that they can be implemented widely enough.