An accurate test of PVA predictions must incorporate a largenumber of data sets to obtain representative assessments for detailed models(Brook et al., 2000).
However, many field studies regarding wildlife populationsare of too short a duration to detect rare events, quantify the importance ofdensity dependence, or distinguish long-term trends (Beissinger & Westphal1998).The reliability of the predictions made using VORTEX is questionedin multiple studies. The predictionsbased on population viability analysis are generally considered overly optimisticunless all potential threats to an endangered species are included in the model(Lacy 1993)(Ralls and Taylor 1997). VORTEX software does not consider the possibilitythat even without any direct threat to the wildlife, population numbers can beas affected by habitat fragmentation just as much as they are by overall lossof habitat. The reduction in partridge spatial density caused by habitatfragmentation will be likely to be a cause of both the particular pattern offragmentation and the demography and behaviour of the partridges (Kendall, 1998)(Lacyet al.
, 1995). In addition, how threats such as hunting or feeding competitionwith other wildlife sharing the habitat, will affect population numbers andwill depend on the pattern of fragmentation and on the demography and behaviourof the partridges. For instance, the reserve size will affect poachingintensity because the edges of the reserves will determine how easilyaccessible the interior is to poachers and how often animals would roam out ofthe relative safety of the reserve (Akcakaya and Burgman, 1995).The threats facing endangered species influence population sizethrough their effects on demographic rates and the spatial structure of habitatpatches. Therefore, until population viability analysis model using VORTEX considereither the demography or spatial structure of the animals, it is likely thatthe results are going to be inaccurate.
In conclusion, the predictive accuracy of a PVA using VORTEXwill only be useful if the distributions of the population rates will not adjustin the future. As ecological systems are dynamic, regulating processes canchange, but it is usually impossible to predict how or when. Research isrequired to estimate how changes from one regulating factor to anotherinfluence the distribution of vital rates and population growth rates. Data onthe growth rate of populations should be collected following the initiation ofa management or conservation strategy, and the results and predictions of PVAsshould be reassessed, and if necessary strategies altered, following theaddition of these data.
A useful application of PVAs would be to compare theconsequences of different management or conservation strategies, and exploringtheoretically the implication of model assumptions on extinction probabilitiesand population dynamics (Possingham, 1993).