الگوهای فضایی از حفاظت از تنوع زیستی در یک مدل تعادل عمومی چندمنطقه ای

Migration dynamics and local biodiversity are interrelated in a way that is likely to affect patterns of regional specialisation. We assess this relationship with a New Economic Geography model that has been extended with biodiversity. Biodiversity is heterogeneous, and responds to habitat availability. The results indicate that a symmetric pattern of regional specialisation is more likely, and that additional equilibria may emerge as the marginal utility of biodiversity increases. In the policy analysis we focus on the case where the overall social optimum is symmetric and show that it can be supported as a non-cooperative Nash equilibrium. However, multiple Nash equilibria may exist.

In this paper we propose and analyse a formal model to study the interaction between human population dynamics and biodiversity. The model integrates a number of feedbacks between the two phenomena. In this introduction we sketch the theoretical framework within which we wish to address this issue. The interactions between population and the environment are numerous and complex. In a special issue of the Population and Development Review, Lutz et al. (2002) advocate that population–environment (P–E) analysis is treated as a specific field of research. It has become increasingly clear that populations have impacts on the environment, e.g., air, water, climate, and biodiversity through the extraction and consumption of natural resources. Conversely, the environment may influence size, composition, location, etc. of human populations through pollution or by affecting fertility. Environmental disasters furthermore may lead people to move to safer places. The present paper tries to capture both sides of the P–E interaction. The emphasis will be on the dynamics of the population, in particular from the spatial perspective of location choice. It is a well-documented issue that individuals may migrate for environmental reasons. In their study on migration within the US, Mueser and Graves (1995) include weather variables as well as an indicator of recreational amenities of lakes. Hunter (2005) surveys the theoretical and empirical literature on migration caused by natural environmental hazards, such as earthquakes and hurricanes, as well as technological hazards, such as nuclear waste and chemical spills. Inspired by the issue of climate change, Reuveny (2005) investigates the relationship between environmental degradation and migration, and the possible conflicts arising from this migration. The relationship between population growth, migration and local air pollution is addressed by Cramer (2002). Marquette and Bilsborrow (1999) list numerous studies concerning the interaction between land use, natural resources and migration. A contribution by Hunter (2000) goes into the effect of population size and distribution on e.g. climate change, deforestation, and land use. Particularly interesting for our purposes is the work by Chu and Yu (2002), who deal with migration and biodiversity. They focus, however, on the impact of human populations on biodiversity, and not so much on the reciprocity that exists between the two. One of the aims of the present paper is to take into account the reciprocal relationship between local biodiversity, and particularly its amenity value, and human migration. An important issue in biodiversity conservation is the fragmentation of habitat reserve sites, which is generally perceived as a threat to long-term persistence of biodiversity (Barbier et al., 1995 and Armsworth et al., 2004). It prevents individuals of species’ populations to (re)colonise habitat areas, thus increasing the extinction risk of the overall population (Hanski and Gilpin, 1991). On the other hand, fragmentation may also have a positive effect on populations by spatially restricting the impact of stochastic disturbances, such as fire or disease (e.g., Fahrig, 2003). Fragmentation of habitat reserve sites, however, has largely been studied in isolation from spatial-economic considerations. Production and consumption levels, for instance, are dependent on the spatial distribution of economic activity as shown by the New Economic Geography model (Fujita et al., 2001; henceforth FKV). In this paper, we consider the problems of habitat fragmentation and biodiversity loss in connection to such a complete economic system. We first briefly survey the economic literature on the design of reserve sites, making a distinction between optimal design and the spatial structure as a result of decentralised economic behaviour. The optimality of the spatial design of reserves sites has traditionally been analysed by economists using one of two modelling approaches (see Eppink and van den Bergh, 2007). The first approach has been to include the use of land in models of renewable resource management. Sanchirio and Wilen (1999) present one of the first of such models that includes patchy habitats and analyse the optimal spatial distribution of harvesting effort (biodiversity is measured only in terms of biomass). Similar, more recent models focus on the impact of reserve sites on harvesting in catch-areas (e.g., Sanchirico and Wilen, 2001 and Armstrong and Skonhoft, 2006). The second approach poses the design of habitat reserve sites as a ‘maximum coverage problem’ (MCP). It finds the spatial pattern of reserve sites that maximises the number of species preserved given budget constraints. This type of modelling is starting to include economic costs in models for the planning of biodiversity reserve sites, as in, e.g., Ando et al. (1998), Polasky et al. (2005) and Naidoo and Adamowicz (2006). Both approaches to optimal spatial allocation of reserve sites remain characterised by an absence of feedback effects between production, consumption and levels of biodiversity that may arise at the macro level from the behaviour of individual economic agents. The economic literature has recently started to give attention to the use of space in conjunction with biodiversity conservation as an outcome of decentralised decision-making. Armsworth et al. (2006), for instance, have analysed land market outcomes when conservationists and land developers compete for the rights to undeveloped land. Their model shows that the participation of conservationists raises the market price of land, thus lowering the amount of land that is developed. The effect on biodiversity depends ultimately on the ‘biodiversity value’ of the land parcels to which development pressure is shifted. Eppink et al. (2004) simulate the effects of competitive bidding on biodiversity and find that alternative socio-economic conditions may have very different outcomes for the composition of land use and biodiversity levels. In general, the pressure on undeveloped land is partly determined by the relative economic competitiveness of regional economies. Biodiversity conservation in the context of regional economic specialisation and development has been studied by, e.g., Smulders et al. (2004) who consider the conservation of a species under threat of regional agricultural expansion in a general equilibrium model of international trade. Barbier and Schulz (1997) study habitat conversion and species richness for a single open economy in a model of dynamic consumption optimisation. In Polasky et al. (2004), species may inhabit either of two types of habitat that may be developed in a specific factors model. None of these models includes between-region mobility of firms, capital or consumers to study the impact of preferences for economic and environmental conditions on regional development. In the area of environmental economics Sandmo and Wildasin (1999), Wellisch, 1994 and Wellisch, 1995, Silva (1997), Hoel and Shapiro (2003) and Haavio (2005) belong to the few exceptions, but they all consider transboundary pollution. We take a local perspective, which seems more appropriate in the context of biodiversity loss and conservation, although in doing so existence values are not fully taken into account if some species only occur in certain regions. An appropriate vehicle to study mobility of production factors and consumers is the New Economic Geography (NEG) model, which is characterised by a general equilibrium approach with imperfect competition. The NEG model is nowadays frequently used in spatial economics to address the impact of regional pollution resulting from production on patterns of regional specialisation (e.g., Pflüger, 2001, Elbers and Withagen, 2004, Van Marrewijk, 2005 and Lange and Quaas, 2007). New Economic Geography provides a relatively comprehensive economic model, but many of the NEG models developed so far have disregarded or only marginally considered the influence that land use has on biodiversity. In this paper, we extend a basic NEG model as described in Chapter 4 of FKV with biodiversity conservation by including habitat loss and its effect on species. In our model, workers migrate from one region to another depending on the real wage differential between regions, as well as on the different levels of biodiversity. Obviously, we want two trading regions. The commodities traded are produced by labour in each of the two regions. We also take into account that the level of biodiversity in each of the regions is determined endogenously. Therefore, we will assume that industrial activities require land use, which through the so-called species-area curve goes at the cost of biodiversity. In this way we can also determine the fragmentation of habitats in a setting of decentralised trading as well as a coordinated social optimum. Rauscher and Barbier (2007) come rather close to our model, but they consider economies with one production sector (manufacturing) and one production factor, whereas we explicitly introduce land as a production factor for agriculture and biodiversity. Furthermore, in Rauscher and Barbier (2007) biodiversity is a global public good model instead of a local good as in our model, and the focus of their analysis is on the redundancy of species between regions and its effect on spatial equilibria. The following section describes the NEG model and our inclusion of biodiversity. We compare the basic results with those in FKV and provide in-depth analyses of parameter settings and policy measures in Section 3. Conclusions and implications for the debate on biodiversity reserve sites are presented in Section 4.

The model explored in this paper analytically and numerically studies the interaction between population dynamics and local biodiversity conservation. We have adapted the spatial Fujita–Krugman–Venables (FKV) model by making land use by manufacturing firms explicit and considering its effects on biodiversity. The model is characterised by labour mobility between two regions, imperfect competition, general equilibrium, and decentralised decision-making. Being firmly grounded in economic theory, the model can provide some new insights into the optimal spatial distribution of biodiversity conservation. The results show that a positive marginal utility of biodiversity indeed affects the laissez-faire solutions to the FKV model. First, the equilibria that indicate regional specialisation become less likely as the utility from biodiversity increases. The marginal effect of local biodiversity can be made large enough to always yield symmetric outcomes. Second, for a range of transportation costs, the bifurcation points of the model can be shifted by increasing the marginal utility from biodiversity, so that additional equilibria emerge. The new stable equilibria are located just off the boundary solutions. The distance between the boundary solutions and these new equilibria can be made to vary by changing the parameter settings for biodiversity heterogeneity. We have also considered environmental policies and compared cooperative and strategic behaviour of the regional governments. When governments cooperate to maximise total welfare and the utility of biodiversity is sufficiently high, the optimal degree of regional specialisation is symmetric for a large range of transportation costs. With decreasing transportation costs, the likelihood of symmetric regional levels of specialisation decreases, and the distribution of firms over regions switches to solutions that are located just off the boundaries. In a Nash game of non-cooperating governments, the symmetric second-best solution can be supported. This result, however, depends on initial conditions, as an asymmetric initial distribution of firms will yield a Nash equilibrium that is asymmetric as well.