Study Sites and Methods:

Site Selection in Missouri

The Daniel Boone Conservation Area (DBCA) in Warren County, Missouri, was selected as an appropriate site to conduct the LEAP study. This 3500-acre Conservation Area is owned and managed by the Missouri Department of Conservation. It is dominated by oak-hickory forest and contains approximately 35 ponds, most of which were created decades ago as enhancements for wildlife. A variety of amphibian species occur in DBCA, including two species of concern, wood frogs (Rana sylvatica) and ringed salamanders (Ambystoma annulatum).

We selected specific study sites within DBCA in March and April of 2003. Considerations in the site selection process included:

• Pond hydroperiod and presence of fish
• Proximity to roads, parking lots, or forest openings (e.g., food plots)
• Forest composition
• Future management activities, including existing timber sales
• Areas proposed for natural area designation and off-limits to disturbance
• Topography and proximity to permanent streams

click map for printable PDF

We eliminated about five potential ponds after consulting with Department of Conservation personnel and learning of past fish stocking. We also used GIS software (ArcView) and readily available data (e.g., digital orthophotos and topo quads) to aid in site selection by eliminating sites with undesirable features in close proximity. For example, we eliminated about 13 ponds that were < 175 m from paved roads and/or large food plots. Another five ponds had to be eliminated because they were in or near proposed natural areas within the larger conservation area. Five more ponds were located in steep terrain and were largely inaccessible, leaving six potential sites. We visited these sites and found that one lacked water (too ephemeral). The last five were judged to be suitable for our study and were all situated within continuous, mature hardwood stands. Unfortunately, at two of these ponds there was a conflict with an existing timber sale to the north and our experimental arrays would have overlapped substantially with these soon-to-be logged areas. Thus, to ensure a fourth replicate array, we decided to create half an array at each of these two ponds, which are located close together (approximately 300 m apart; Figure 1).

Experimental Forest Management Treatments in Missouri:
During the period of May-December 2004, all experimental forest cuts were finished at our four experimental arrays. Our final treatments are detailed below (see figure 2):

1. Control treatment - maintained in the original condition of the forest (average ~ 80 years old, oak-hickory) with only minor disturbance from skidders

2. Clear-cut with CWD retained treatment - all trees greater than 10 inches dbh were removed by loggers, the remaining trees were all cut by hand, tops and limbs were cut to 6 feet or less to facilitate contact with the ground.

3. Clear-cut with CWD removal treatment - all trees greater than 10 inches dbh were removed by loggers, the remaining trees were left standing and girdled by hand, tops and limbs were cut to 6 feet or less to facilitate contact with the ground, and quadrants will be burned in January 2007 to remove woody debris.

4. Partial cut treatment - we "thinned" our quadrant to 50% stocking density by girdling unwanted species and defective trees, each array was inventoried so that the target stocking density was achieved accurately.

Drift Fence and Pitfall Trap Construction in Missouri (figure 3):
Over the fall and winter of 2004, four woodland ponds in Daniel Boone Conservation Area were completely encircled with aluminum flashing (39 cm high) buried 5-10 cm in the ground and pitfall traps were installed. All ponds, parallel, and wing fences have been finished and are functional. Each pond has 24 parallel fences each 50 ft long, 6 per quadrant at 50, 100, and 150 m from the ponds edge (figure 3). Additionally, a set of 3 wing fences has been placed on each edge of the control quadrant to monitor lateral movement from clear cuts into unaltered control forest. Each length of fencing has 3 pitfall traps per side; two on each end and a pair in the middle (figure 4). Buckets are in and operational bordering all five ponds (opened February 8, 2005). Additionally, buckets are in and operational for all wing fences (opened February 27, 2005). All buckets in parallel fences are in and opened at ponds 1 and 2 (opened February 27, 2005). All of the buckets are in at pond 5 in the parallel fences and all of the buckets in the wing fences are in (opened March 6, 2005). Pitfall traps are currently being constructed in the remaining experimental array and will be opened upon completion (mid-March 2005). 

Figure 3 :  A schematic of one replicate experimental array. Each of the four treatments has 6 fences, set at 50, 100, and 150 m respectively, to monitor amphibian movement in the terrestrial habitat. Additionally, a set of 3 fences have been added to the edges of the control treatment to monitor lateral movement into and out of the unaltered forest control quadrant.

Figure 4:  A schematic of a 50 ft drift fence with pitfall traps.  Each length of fence has 3 pairs of buckets per side for amphibian capture.

Experimental Terrestrial Pens:
A total of 64 pens will be constructed within the LEAP arrays and will be used to assess metamorph survival under each of the four forestry treatments. Four 3x3 m pens will be built in each quadrant of all experimental arrays. This spring 32 pens will be stocked with Rana sylvatica and 32 with Bufo americanus metamorphs. Pens will be stocked with Ambystoma maculatum metamorphs the following spring. Pen construction began in November 2004 and is scheduled to be finished by mid-March 2005.