Methods for estimating density are very similar to those used with point counts (see Chapter 1 for details) and various approaches have been advocated (J¨arvinen & V¨ais¨anen 1975; Burnham et al. 1980; Bibby et al. 2000; Emlen 1977; Buckland et al. 2001). Simple indices of the number of birds recorded per unit length of transect obtained by counting birds, either up to an unlimited distance, or to a single fixed distance on either side of the transect, are unlikely to be reliable because birds
are inevitably missed. Increasingly sophisticated methods under the banner of distance sampling are now commonly applied to transect data for birds and the necessary software is available freely (see ‘Correcting for differences in detection probabilities’).
In practice it may be sensible to map all bird records onto a schematic representation of the transect in your notebook, or onto a recording sheet. It might also help if distance bands were drawn onto this, too. Recording the birds in this manner means that a variety of different techniques can ultimately be used to analyse the data. If birds are recorded in separate distance bands, check that the distances can be reliably estimated. Try to standardise the rate of movement along the transect route; walking too fast misses birds, but walking too slowly may result in double counting. A walking rate of 2 km per hour is reasonable in open habitats, though 1 km per hour would be more realistic in forest.
Methods for estimating density are very similar to those used with point counts (see Chapter 1 for details) and various approaches have been advocated (J¨arvinen & V¨ais¨anen 1975; Burnham et al. 1980; Bibby et al. 2000; Emlen 1977; Buckland et al. 2001). Simple indices of the number of birds recorded per unit length of transect obtained by counting birds, either up to an unlimited distance, or to a single fixed distance on either side of the transect, are unlikely to be reliable because birds
are inevitably missed. Increasingly sophisticated methods under the banner of distance sampling are now commonly applied to transect data for birds and the necessary software is available freely (see ‘Correcting for differences in detection probabilities’).
In practice it may be sensible to map all bird records onto a schematic representation of the transect in your notebook, or onto a recording sheet. It might also help if distance bands were drawn onto this, too. Recording the birds in this manner means that a variety of different techniques can ultimately be used to analyse the data. If birds are recorded in separate distance bands, check that the distances can be reliably estimated. Try to standardise the rate of movement along the transect route; walking too fast misses birds, but walking too slowly may result in double counting. A walking rate of 2 km per hour is reasonable in open habitats, though 1 km per hour would be more realistic in forest.
การแปล กรุณารอสักครู่..