2011-08-10_APPLICATION CORRESPONDENCE - C2010088

Home Browse Search

Corey Heaps CAM Colorado LLC August 10, 2011 Page S started out sequentially but became selective after #24, yet, in the predisturbance areas, transects were sampled for cover and shrub density sequentially, 1 through 25. The Division does accept "both random and systematic (selective) sampling designs" as per Rule 4.15.11(1), but also requires "consistency in sampling." The explanation given in the last adequacy response does not clarify the issue as to whether the transect selection was unbiased. Please provide an explanation within the appropriate section of Exhibit 5, explaining why cover data was collected sequentially in the predisturbance area and sequentially /selectively in the reference area to meet sample adequacy. CAM - Response: The Division's comments regarding the perceived potential bias of the vegetation sampling program are confusing and difficult to understand. Part of the problem appears to be that the DRMS appears to have made assumptions on how the data were collected that are inconsistent with the sampling approach used. For example, they state that "the sampling sequence in the pre - disturbance area is sequential" while the sampling on the Greasewood Reference Area involved a combination of sequential and selective sampling. These comments are not correct. All of the sampling on the site, baseline and reference area was performed using a "random" sampling approach. None of the sampling was performed in a "sequential" manner as suggested by the DRMS. Perhaps in the theoretical world of textbook ecology, this is how it is done, but in the field no one takes one sample, then travels miles away to the next sample and then several miles back and so forth. In performing field work, time is not wasted traveling back and forth across a mine site or project areal several miles in width, to ensure that sampling is done in a "sequential" order. What the DRMS has seemingly ignored is the excellent discussion found in Part 3, entitled "Vegetation Sampling Methods, Considerations and Design Approaches" found in Division's Bond Release Guideline. Therein it states a total of at least five times on pages 15 and 16 that all sampling must be performed in an "efficient" manner. The concept of "efficiency" is totally ignored in the DRMS comment which concentrates totally on the "theoretical" aspects of "sequential" sampling and specifically concludes that unless "sequential" sampling is performed the data a somehow biased. The Bond Release Guideline which states on page 1 that this document takes "Precedence" of this document over the Vegetation Guideline, and further states on page 15 that "vegetation sampling design approaches acceptable to the Division include simple random sampling, and stratified random sampling design modifications including proportional allocation and area weighted sampling. Although not addressed in detail in this document, multi - stage, or cluster sampling and systematic designs as discussed below may be adaptable to certain reclamation situations, and may meet regulatory requirements when design assumptions are met and specified procedures are followed." This document further states that in instances where there is "difficulty of attaining sample adequacy... operators are advised to give careful consideration to sampling approaches which minimize variation among sample observations. Such approaches might include with -in parcel stratification and acreage weighting. Multistage or cluster sampling has been suggested as an approach which might be of use in this respect."