Laserfiche WebLink
V.E. Experimental Design Strategy Across Multiple Scales <br />The study design for each monitoring and research component of the IMRP has been integrated <br />into one overarching statistical survey design. Through the use of a common design, the <br />monitoring and research activities will collectively determine the biological response to Program <br />management actions. This integrated survey design is intended to maximize the use of <br />monitoring and research resources by enabling the shared use of data for multiple studies, and <br />ensure unbiased estimates with inference to the intended resource. <br />The cornerstone of the IMRP is a systematic sample of survey units throughout the length of the <br />Platte River reach (Figure 13). This systematic sample will allow unbiased estimation of <br />monitoring and research parameters at the system, Program, or project level scale. Utilizing an <br />equal probability based sample of units will enable post-stratification and will provide pre- <br />treatment data for all areas in the system (specifically useful for Program land purchases at <br />locations unknown at the start of the Program). For any given stratification scheme, the sample <br />for each stratum will contain population units in proportion to their presence in the landscape. <br />Monitoring activities are intended for trend estimation and the examination of the influence of <br />other variables on the estimated trend. Hypotheses developed from monitoring (i.e. factor A was <br />the cause of a trend) will be evaluated with experimental research. Research will investigate the <br />hypothesized cause and effect relationship and may involve the application of treatments such as <br />habitat manipulation. <br />V.E.1. Monitoring Design <br />Monitoring revisit design - Survey designs for environmental monitoring are greatly enhanced <br />by the use of panels to identify which sample units are surveyed on each visit through time. A <br />panel is a collection of sample units that are always sampled at the same time (Fuller 1999). The <br />frequency and pattern at which panels are visited through time is the revisit design (McDonald <br />2003). <br />The revisit design of a monitoring program reflects the relative importance of each monitoring <br />objective. Visiting a set of sample units every year (pure panel) ensures low variance for trend <br />estimates but the sites tend to wear out and obtain biases through conditioning, particularly when <br />destructive sampling is used (Fuller 1999, McDonald 2003). Visiting a set of sites in alternating <br />years (rotating panel) allows for the inclusion of more sites in the sample (increasing the chance <br />of observing rare elements) and results in low variance for the estimation of inean levels (status) <br />within a year (Fuller 1999, McDonald 2003). Urquhart and Kincaid (1999) found the pure panel <br />to be the best for detecting linear trends through time and revisiting new sample units each time <br />to be the best for estimating status. The revisit design for biological monitoring under the IMRP <br />will balance the objectives for status and trend estimation equally as suggested by McDonald <br />(2003), Fuller (1999), Breidt and Fuller (1999), and Urquhart et al. (1998). <br />The IMRP revisit design involves a split panel; a panel (group of sample units) that is visited <br />every year and several panels that are visited in rotating years. Using the notation specified by <br />McDonald (2003), the revisit design for biological monitoring will be [1-0, 1-3]; one panel will <br />be surveyed each visit (the 1 pure panel visited each time is indicated by the 1-0) and four panels <br />will be surveyed once every fourth visit (the 4 rotating panels each surveyed 1 time and then not <br />September 1, 2006 Adaptive Management Plan 33