Location of Sampling Stations: Litterfall trap placement was determined by a stratified, random design. A line was drawn on a map of the two forks of Kings Creek representing the longest possible straight line through the forest. This line was scaled to the actual length of the forest, and divided into five equal sections. A random number generator was then used to select a point on each segment, and a line perpendicular to the main line was drawn. These perpendicular lines represented the five transect lines used to place litter traps in each fork of Kings Creek (PGLO1 Figure 1 available in Appendix M). The points where these lines emerged from the forest were located in the field, and a surveyor's transit and compass were used to measure the length of each transect and set stakes at 25 m along the transect lines. Once the total length of the transect lines was known for each forest, the number of traps allocated to each transect line was determined. The number of traps corresponds to the relationship: length of transect/total length of transect in forest x30. The only exception to this procedure was that each transect was assigned a minimum of two litter traps. A random number generator determined the location of each trap. Again, exceptions existed in site selection, and these were 1. Traps were to be separated by a minimum of 5 m, and 2. Traps were to be placed within 10 m of canopy foliage. This last rule was used in placing traps along transects N-1 and N-3.
Frequency of Sampling: Samples are collected monthly March to December, weather permitting. October and November samples are collected approximately every two weeks. No collections are made during January and February (little to no sample this time of year).
Variable Measured: Total deposition of litter (grams of dry weight/0.25 m2) per collection period, and subdivisions of wood and seeds (see sample data sheet). Please note that foliage is not measured directly; foliage mass may be determined by subtraction of woody and seeds from total. Insects, feathers and other animal debris and fecal droppings are removed prior to weighing.
Methods: Thirty litterfall traps, 50 x 50 cm (.25 m2) are located along the north fork of Kings Creek. Two are located on the south fork of Kings Creek. The north fork boxes are numbered 31 to 60 and the south fork boxes are numbered 1 and 2. Originally, the south fork also had boxes 3 to 30 but these samplers were terminated in 1993 due to repeated damage by bison. (Boxes 1 and 2 are located just outside the bison area). The design of the litterfall traps is similar to those used at Coweeta but have been modified overtime. Modifications include increasing the depth of the traps to 50 cm ('deep dish'), reinforcement of all corners and screens (to minimize the effect of rot) and metal legs. Most recent design attached in Appendix P.
Samples are collected by brushing all material to a corner. The sample is then brushed or scooped into pre-labeled paper bags. Any sample that gets caught on the upper rim of the trap (large branches, etc), will be marked and cut or broken apart to retain the portion that would be inside the sampling area (the rest is discarded). Samples are dried at 60oC for a minimum of 3 days and weights of woody material, seeds and total are measured.
Summary of All Changes:
FREQUENCY OF SAMPLING (Original set-up; 1981): Traps were placed in the field on September 4-6, 1981. Collections were made every two weeks during September, October and November and then monthly from December to September. Deviations: In 1983, collections were irregular. Beginning in 1984, the autumn collections were at monthly intervals. And 1987-1992, the monthly collections were from March through November.
VARIABLES MEARSURED (1981): Deviations: Foliage was not measured 1981-June 1992, and 1995-present (foliage mass may be determined by subtraction of wood and seeds from total). Foliage weights were measured July 1992-Dec 1994.
ORIGINAL METHODS (1981): Sixty litter fall traps, 50 x 50 cm (0.25 m2) were constructed during the summer of 1981. The design of these traps was similar to those used at Coweeta (e.g., Cromack, K. Jr.1973, Litter production and decomposition in a mixed hardwood watershed and white pine watershed at Coweeta Hydrologic Station, North Carolina. PhD dissertation. University of Georgia, Athens, GA.), except that the depth of these traps was 50 cm. This 'deep dish' design (available in Appendix P) was used to prevent losses due to high winds. The traps open at 1 m above the forest floor. Deposition of litter into these traps due to lateral movements by wind is believed minimal (e.g., virtually no foliage was found in these traps during January 1982, in spite of strong winds).
The allocation of 30 traps per forest floor resulted in a somewhat greater sampling intensity of the South Fork of Kings Creek. Transect lines measured 353.6 m resulting in one collector every 11.8 m. The total transect length for the North Fork was 543.3 m resulting in one trap placed every 18.1 m.
The transects are marked by a yellow-painted steel conduit stake at each end. The material is dried at 60oC and weights of woody material, seeds and total are measured. Foliage is assessing by subtracting the woody and seed weights from the total weight.
January 6, 1993: Collections from litter-fall traps within the bison area terminated. The sample ID's affected are 3 through 30, inclusive. This modification is due to repeated damage to litter fall traps by bison.
For additional metadata information see: http://lter.konza.ksu.edu/sites/default/files/DC.pdf
For additional methods information see: http://lter.konza.ksu.edu/sites/default/files/MM.pdf
