METHODS 



Two sets of information were compiled for each RNA 

 site before fieldwork. First, written information was 

 reviewed about the RNAs. This was in the RNA 

 establishment information as available from the U.S. 

 Fish and Wildlife Service. It had previously been 

 collected and entered by Montana Natural Heritage 

 Program in the Biological Conservation Database 

 (BCD) as representing recogniied natural areas and 

 public lands (Site Basic Record Database, and Managed 

 Areas Database, respectively.) 



Second, U.S.G.S. topographic maps (7.5') and available 

 aerial photos were assembled prior to or in conjunction 

 with fieldwork at each site, and RNA boundaries were 

 copied onto the maps. The photographs were used for 

 site stratification and plarming traverses across the 

 major features of the RNA. Mylar overlays were used to 

 map out areas having spectral signatures to consider for 

 ground-truthing, and as base maps for future map 

 production. Often photos were not readily available, so 

 that the topographic maps were used to guide the site 

 traverses, focusing mainly on unique combinations of 

 slope, aspect and elevation throughout the site. In 

 addition, comments were routinely requested from 

 U.S. Fish and Wildlife Service biologists familiar with 

 the RNAs for information on biological features and 

 management, and for information and clarification 

 about access and boundaries. Additional botany and 

 ecology resources were compiled in techiucal 

 preparation (described separately in methods.) 



Field investigations were restricted to established RNA 

 boundaries, with two additions. The Sandhills area of 

 the Medicine Lake NWR was included in surveys 

 because it shares some of the rare plant species features 

 as Big Island RNA, and has a special designation as 

 part of wilderness area. In addition, an area west of 

 Hell Creek State Park was identified by Bill Haglan 

 (Charles M. Russell NWR) as possessing features 

 potentially worthy of considering for RNA designation. 

 We refer to the set of seventeen study sites as including 

 the Medicine Lake Sandhills and Hell Creek areas 

 though they are not designated as RNAs. 



Ecological and botanical iriformation collected in the 

 field was used to expand the RNA establishment 

 information, fully described in this report and 

 summarized in BCD. In addition, the individual rare 

 plant records have been entered in BCD, and 



vegetation plot data is stored in vegetation databases 

 and draft classification documents. 



Ecological Methods 



Plant communities were identified and documented in 

 terms of their community composition, structure and 

 associated abiotic enviroiunental parameters by 

 establishing representative 1/10 acre plots (37.2 ft 

 radius). Data were recorded on a standardized 

 Community Survey Form as used by Montana Natural 

 Heritage Program consistent with ECADS vegetation 

 ordination analysis (Ecosystem Characterization and 

 Description System, USPS 1996; see Appendix A). 



At each RNA, plant associations were documented 

 that met one or more of the four following criteria: 



L Prevailing plant associatioris within designated 

 areas, i.e., the most extensive vegetation features 

 dominated by native plant species, 



2. Plant associations that were the basis of original 

 designation, e.g., the Douglas fir forest at the Two 

 Calf'Douglas-fir RNA, 



3. Well-developed plant associations that are 

 potentially rare statewide or rangewide, and 



4. Well-developed plant associations in outstanding 

 ecological condition regardless or rarity of extent 

 at the site. 



Vegetation sampling plots were placed within each 

 major natural vegetation type based on observed aerial 

 extent of the type. This approach provided 

 documentation for common vegetation types, but was 

 not intended for exhaustive sampling of localized or 

 atypical environments, large replications, or full 

 gradient representation. In some instances, a given 

 common community type may span a range of 

 environments, in which case the attempt was made to 

 sample the modal expression of a commuiuty's 

 environmental range. Sampling sites were chosen 

 "subjectively, but without preconceived bias" (Mueller- 

 Dombois and EUenberg 1974) to meet the criteria of 

 homogeneous vegetation composition, least 

 disturbance, and representative setting. Plot points are 

 mapped on U.S.G.S. topographic maps with 300 feet 

 precision. 



On the first Refuge visited, Medicine Lake, excellent 

 quality aerial photography was available at 8 inches / 

 mile that served as a base layer upon which vegetation 



