Although targeting macrohabitats, such as landscape features, for conservation actions and monitoring is an efficient approach to conservation (Noss 1987, Franklin 1993, Scott et al. 1993), obtaining a desired ecosystem condition may not result in the desired response from all species (Herkert 1994, Swengel and Swengel 1997) or even entire species assemblages (MacNally et al. 2002). Many species respond to very specific aspects of a landscape feature and are therefore not necessarily protected by feature-level actions. Because of these specific responses, species can be good indicators of specific changes within ecosystems that might otherwise be overlooked. Further, some species are not limited primarily by habitat and are threatened or otherwise influenced by forces independent of landscape feature conditions. Collecting, storing and assessing information on individual species will always be necessary to prevent extinction and to evaluate the condition of more-specific ecosystem components.

More rigorous monitoring of individual species should focus on species that are:

• highly imperiled;
  
  
• known indicators of ecological integrity;
  
  
• known to have specific requirements that are not assessed by landscape feature monitoring (e.g., area limited (size), resource limited (e.g., Karner blue butterfly and lupine), dispersal limited (distance between habitat patches), or process limited (fire regimes); or
  
  
• otherwise of considerable interest (e.g., socially or economically important species, keystone species).

General tracking of endangered, threatened and special concern species is already conducted by many entities who contribute data to MNFI's Biotics database. Identification of ways to collect, store and share information for SGCN that are not tracked in Biotics is also needed. By monitoring species assemblages (e.g., frog and toad survey, breeding bird survey, fish surveys), trend information on specific species can be assessed. In addition, broad tracking of species trends should be supplemented with specific, well-reasoned research efforts to validate trends and determine causes. Research to collect baseline data for poorly understood species should be a priority, as identified within the discussion of the 'Lack of scientific knowledge' threat in the Statewide Priority Threats.

Rigorous monitoring is not feasible for all SGCN across the entire State. Therefore conservation tools, including predictive models, can be useful for drawing broad-scale conclusions on distribution and abundance based on assessments made at smaller scales. These models can greatly increase the utility of information from monitoring programs.