Introduced vertebrate predators are one of the most important threats to endemic species throughout a range of ecosystems, in particular on islands in biodiversity hot spots. Consequently, the reduction of predator numbers is considered a key conservation action in the management of many native vertebrates vulnerable to predators. It is now established that control attempts may affect non-target species through trophic interactions, but little is known concerning their consequences on competitive relationships. We study a mathematical model mimicking the effects of controlling introduced species in the presence of their competitors.We used two competing rodents to illustrate our study: black rats, Rattus rattus, and mice, Mus musculus. Analyses of the model show that control of only one introduced species logically results in the dramatic increase of the overlooked competitor. We present empirical data that confirm our theoretical predictions. Less intuitively, this process, which we term 'the competitor release effect', may also occur when both introduced competitors are simultaneously controlled. In our setting, controlling both predators can promote their coexistence. This occurs as soon as the inferior competitor benefits from the differential effect of the simultaneous control of both competitors, that is, when the indirect positive effect of control (the removal of their competitors) exceeds its direct negative effect (their own removal). Both control levels and target specificity have a direct influence on the extent of this process: counter-intuitively, the stronger and more specific the control, the greater the effect. The theoretical validation of the competitor release effect has important implications in conservation, especially for control management.

Abstract. The accurate estimation of wildlife population density is difficult and requires considerable investment of resources and time. Population indices are easier to obtain but are influenced by many unknowns and the relationships to actual population densities are usually unclear. Wildlife biologists, whether in the public or private sector, often find themselves in difficult situations where a resource manager or landowner wants good information, quickly, at low cost, and without clear objectives. In many situations, in addition to establishing clear objectives, a budget and timeframe, a biologist must understand and deal with the reality of many logistical concerns that will make the achievement of the objectives difficult or impossible. The situation is often complicated because the biology and ecology of the species of interest may be poorly understood in the specific setting and the species may be very rare or strongly influenced by current or past human activities. Methods to monitor a wildlife population may need to be tested or validated, extending the time and resources needed to complete the assigned task. In this paper, I discuss many of the challenges faced and the decisions to be made when a biologist is requested to provide useful, timely information on the status of a wildlife population. IntroductionThere are many reasons why natural resource managers need to monitor wildlife populations. A large array of methods has been developed and used to that purpose (Seber 1982(Seber , 1986(Seber , 1992Lancia et al. 1994;Thompson et al. 1998;Schwartz and Seber 1999). The biologist assigned the task, however, should be aware of the many factors and difficulties that can hinder the successful outcome of a monitoring effort. Because the determination of wildlife population abundance or density can be very difficult and expensive, one should have a clear set of objectives and adequate resources available for the task. Additionally, one must carefully select one or more field methods to apply to the population of interest. Many considerations can influence the method(s) selected and the value and accuracy of the data that result. Finally, implementing the monitoring strategy can be fraught with difficulties, especially when applied in remote or restricted areas or in lesser-developed countries. Careful planning and the anticipation of problems are required to improve the chances of success given the many things that can go wrong.In this paper, I discuss the practical side of important factors (objectives, method selection and implementation) involved in monitoring wildlife populations. My intention is to aid future practitioners from revisiting many of the difficulties that have been encountered in the past. I do not discuss models, statistical formulations and assumptions, or the evaluation ofdatasets in the estimation of wildlife population abundance or density. These topics have been well covered in several books and reviews (e.g. Seber 1982Seber ,1986

No abstract