Landscape surrogates are easily measurable quantities of a landscape which are used to represent target responses (such as presence/absence of a particular species), which may be harder to measure.

The researchers examined the colinearity of the individual landscape surrogates and also the correlation between surrogates and direct evidence of the target responses for case studies in two different locations in which fragmentation of habitat had occurred. They identify fragmentation as a possible cause of loss of biodiversity.
 
    In the first case study, wilderness quality was measured by:

• Remoteness from settlement
• Remoteness from access
• Apparent naturalness (lack of permanent human structures)
• Biophysical naturalness (lack of biophysical disturbance caused by modern land use activities)
• Total wilderness quality (as a combination of the other four)

    Researchers found a strong colinearity among the surrogates, but no significant relationship between these surrogates and the target response, in this case presence and abundance of arboreal marsupials.
 
    In the second study, the surrogates were:

• Amount of habitat in a landscape
• Spatial sub-division of habitat in a landscape
• Patch isolation
• Patch shape and edge area
• Relative density of roads in surrounding area

    There was little colinearity among the surrogates, and they were compared to the target response by contrast of specific sets of surrogates. In this case there were significant relationships between the contrast of surrogates and measure of the direct response. However, the relationships were different for different faunal species, indicating that not all species respond identically to changes in habitat and fragmentation.
Using landscape surrogates is an important tool, because direct measurements of target responses are not always as feasible. However, as indicated by the first case study, not all surrogates are effective measures of their target responses. The authors presented a framework for evaluating possible surrogates:

• "The best attainable inference relating to the defined units for study must arise from direct measures of the target response on these units.
• "The validity of a surrogate measure should be judged by the probability that results based on the surrogates alone are 'concordant' with results that would be obtained if the direct measures were to be observed and analyzed."

Thus, the surrogates should be as close as possible to the actual target response. In addition, surrogates that have high colinearity should be avoided, because they just present the same information in slightly different form. Not all surrogates or sets of surrogates apply to every species. So if the flora group chooses to monitor amazonian flora by certain indicator species, we should make certain that the surrogates we use for a species are actually good indicators for that species, and we might also want to choose species which are applicable to different surrogates as to provide a broader base for study.