Wildlife conservation evaluation (Smith & Theberge 1986; Usher 1986a; Wathern et al. 1986) is used to assess the conservation value of different sites, some of which may be protected later. It may also be used to evaluate different types of management within or among sites. To rank or group sites, clearly defined evaluation criteria should be used. One of these, naturalness, is frequently used (Margules & Usher 1981; Smith & Theberge 1986; Usher 1986b), but has been difficult to define, and it has been used in different ways by different conservationists (see Margules & Usher 1981; Middleton & Merriam 1985; Rackham 1991; Smith & Theberge 1986; Taylor 1990). Recently, Anderson (1991) proposed a method for assessing naturalness. He applied the method to the Greater Yellowstone Ecosystem but suggested it could be used more generally in conservation evaluation. Below, I argue that Anderson's proposal is inadequate in many situations, and I discuss aspects of the criterion of naturalness that need to be considered. Natural usually describes a process, situation, or system more or less free of human influence (Anderson 1991). In its most common usage, naturalness reflects degree of human influence-the lower the influence, the more natural the site. Implicit in Anderson's proposal is the assumption that human influence is negatively correlated with value for biological conservation. But exactly what is being valued is not clearly stated by Anderson. He seems to focus on ecosystems and communities, but he also mentions the need for component populations and discusses species composition. Ideally, every evaluation criterion should be considered at all levels of biodiversity (landscapes, ecosystems, communities, species-especially endangered ones, populations, genes), although all authors to some extent are influenced by their own experiences. Human influence often decreases the value of sites, but in some important cases it is necessary to maintain biodiversity. A good example is ancient types of farmland, which are increasingly considered for protection in Sweden (Gotmark & Nilsson 1992; Nilsson & Gotmark 1992). As in many other countries (Fry 1991), farmland in Sweden has changed rapidly during the last decades. Wetlands have been drained, streams and ditches have been fed into underground pipes, smaller groves in fields have been removed, and small farms have either disappeared or merged into bigger ones, creating larger fields with less biological variation. Also, fields have been sprayed with pesticides, and fertilizers have been applied to fields and grasslands. There still remains, however, smaller heterogeneous areas of farmland managed by older methods (see Robertson et al. 1990). These areas are extremely rich biologically. For instance, in permanent old pastures managed without fertilizers there may be forty-five species of vascular plants per square meter, compared to twenty-seven in abandoned old pastures (no longer grazed) and fourteen in fertilized old pastures (Ekstam et al. 1988). In temporary sown pastures, species richness is even lower. Diversity and density of terrestrial birds are considerably higher in older than in modern agricultural landscapes (Robertson et al. 1990). Wet meadows grazed by livestock along the coast and at lakes support many more species of shorebirds than ungrazed, abandoned meadows (Larsson 1969; Soikkeli & Salo 1979). Also, old farmland is important to reptiles, amphibians, and insects that prefer open, heterogeneous habitats (C. Andren & B. Ehnstrom, personal communication). Thus, species richness is extremely high in old farmland in Sweden and elsewhere in Europe (see Green 1989). Sixty-eight percent of the 419 threatened species
Read full abstract
Feb 28, 2021
Birger Hjelm 
Open Access