Sistema contable peruano

For both Sedge and Reed Warbler, CES data and NRS data provide informa- tion on productivity, although the nature of this information is quite different. Whereas the indices of productivity,Pt, estimated from CES count data, provide a measure of breeding performance across the entire breeding season, integrat- ing fecundity, multiple breeding attempts, egg and chick loss, and immediate post-fledging mortality (Freeman et al., 2001; Robinson et al., 2007), the NRS gathers information on the components of productivity: clutch size, brood size, and daily nest failure rates during the egg and nestling stages (Crick et al., 2003; Baillie et al., 2007). Measures of productivity across the entire breeding season are calculable from NRS data (see Peach et al., 1999; Siriwardena et al., 2001; Freeman and Crick, 2003), but, in general, constructing an informative mea- sure is problematical. In part this is due to the unknown number of breeding attempts per pair (i.e. multiple broods or replacement nests) as a result of in- sufficient NRS sampling late in the breeding season (Bibby, 1978; Baillie et al., 1986), and as individual breeding birds are not monitored for the duration of the breeding season (Crick et al., 2003). This inability of the NRS to provide

information on the annual variation in the number of breeding attempts can prevent reliable estimation of true (i.e. total) seasonal productivity. For exam- ple, in a scenario described by Freeman et al. (2001), changes in agricultural practices, such as earlier harvesting, could cause a decline in farmland bird pro- ductivity via a reduction in the annual number of possible breeding attempts. This goes undetected in the NRS data. For species such as Reed Warbler which are typically double brooded (Catchpole, 1974), or species like Sedge Warbler for which the degree of double brooding is unknown (Alker and Redfern, 1996), CES count data provides a much better means of monitoring the true seasonal productivity over time.

Secondly, immediate post-fledging mortality is automatically incorporated into the CES index of productivity, Pt, for juveniles are defined to be fully fledged birds. As NRS data collection ends once the offspring have left the nest, it pro- vides no information on post-fledging mortality. Consequently this important, and often appreciable (see Robinson et al., 2004), component of productivity is frequently over-looked in NRS reports on breeding performance (Freeman et al., 2001; Miles, 2005). Additional nestling ringing programmes may provide this information, but for many species, such as Reed Warbler, nests are not readily accessible rendering it impossible to ring the sufficiently large number of nestlings needed for accurate estimation (Freeman et al., 2001).

Although the CES scheme provides the only practical method of indexing annual productivity over the entire breeding season (Robinson et al., 2007), rather than on a per attempt basis as NRS data do, the worth of NRS data is not in doubt. They can yield information from habitats not amenable to CES monitoring, and have proved particularly valuable in highlighting the effects of climate change (Crick et al., 1997; Crick and Sparks, 1999). Along with joint interpretation with CES productivity indices, NRS can also, uniquely, provide an insight into which individual components of productivity are driving the observed changes in annual productivity.

2.4

Estimating Adult Survival

The CES live-recapture data consist of live resighting histories for individual birds over the study period and fall into the category of mark-recapture models (Lebreton et al., 1992; Williams et al., 2002). For example, the live-recapture data for an individual bird with ring K710666 are:

K710666 SEDWA AD 154 1997 0 0 1 0 1 0 1 1 0 0 0 0 K710666 SEDWA AD 154 1998 0 0 0 0 1 1 1 0 0 0 0 0 K710666 SEDWA AD 154 1999 0 1 0 0 0 0 1 0 0 0 0 0 K710666 SEDWA AD 154 2000 1 1 1 0 0 0 0 0 0 0 0 0

indicating that this bird, an adult (AD) Sedge Warbler (SEDWA), was initially caught and ringed at site 154 in 1997, and subsequently recaptured at this site in years 1998, 1999, and 2000. Furthermore, bird K710666 was caught at multiple visits within each of these years, e.g. in 1997 this bird was encountered at visits 3, 5, 7, and 8.

The annual recaptures of individual birds, under the CES scheme, provide valu- able information on their survival. Adult Sedge and Reed Warblers are markedly site-faithful (Wernham et al., 2002), generally returning to the same breeding grounds annually, whereas juveniles disperse from their natal areas (Catchpole, 1974; Peach, 2002). On account of the high dispersal of juvenile birds, and con- sequently the low number of juveniles that are recaptured in years subsequent to ringing (see Table 1.3), we restrict this part of the analysis to the capture histories of adult birds.

An asset of the “constant effort” design is that probabilities of recapture (at a site) can be considered constant over years, simplifying the model structure (Peach et al., 1990; Pratt and Peach, 1991; Peach, 1993; DeSante et al., 1999). However, it is imperative that captured birds are representative of the pop- ulation under interest, in this case resident breeders at the CES sites (Hines et al., 2003). The temporary presence of transient birds, i.e. birds not breeding in the vicinity of the site but migrating through it, needs to be accounted for by explicit probabilities within the analysis (Peach, 1993; Thaxter et al., 2006; Cave et al., 2009b; Freeman, 2008). As transient birds do not return to the site of their inaugural capture and ringing, if not adequately modelled, their presence in the CES live-recapture data produces negative bias in the estimates of survival (Peach et al., 1990; Pradel et al., 1997). Operationally, transient

birds can be thought of as having a zero probability of “survival” after their initial capture as they permanently leave the study area (Pradel et al., 1997). The occurrence of transients is a common problem in the analysis of live- recapture data for birds and small mammals (Pradel et al., 1997). In this section we describe, compare, and contrast several alternative models appro- priate for CES live-recapture data known to contain a mixture of resident and transient birds. We develop, and describe in detail, a new, modified approach to the standard Pradel et al. (1997) transient model that more fully utilises the information pertaining to residency provided by the CES scheme, and advocate its benefits over the alternative models.

We note that estimates of adult survival derived from CES data should be considered those of “apparent” adult survival (denoted bySa), which in addition to mortality, also incorporates a component due to the permanent emigration of birds away from their site of ringing (Pratt and Peach, 1991; Peach et al., 1995b; Thaxter et al., 2006). Assuming the rates of permanent emigration are constant over time, annual estimates of apparent adult survival will enable temporal changes in “true” adult survival (denoted by φa) to be detected (Peach, 1993). We address the conversion of apparent survival rates to true survival rates via the integrated population model in Chapter 5.