The first source of variation is the uncertainty we have in our population estimates. Two sources of variation must be assessed before we are confident of our inference from these estimates. However, if the 1995 estimate was 40 pairs, we might still be concerned, but would be less confident that immediate action is required. For example, if the population size of Mexican spotted owls in Mesa Verde National Park is determined as 50 pairs in 1990, and as only 10 pairs in 1995, we would be concerned that a significant negative trend in the population exists during this time period, and that action must be taken to alleviate the trend. Methods to separate process variation from sampling variation will be presented.ĭetection of a trend in a population's size requires at least two abundance estimates. The latter two sources often are referred to as process variation, i.e., variation in the population dynamics process associated with environmental variation (such as rainfall, temperature, community succession, fires, or elevation).

Three sources of variation must be considered: sampling variation, temporal variation in the population dynamics process, and spatial variation in the dynamics of the population across space. In this section, we discuss sources of variation that must be considered to make inferences from data when trying to detect trends. Charles Gowan, in Monitoring Vertebrate Populations, 1998 5.2 VARIANCE COMPONENTS

Calculate standard error