Fig 1.
Pocket gopher mounds in a longleaf pine forest at the Ordway-Swisher biological station.
(a) Pocket gopher mounds covering longleaf pine and turkey oak litter near the edge of a recent burn. In the burned section, a large portion of the litter was consumed by the fire, but litter under pocket gopher mounds was protected from consumption. (b) Pocket gopher mounds have a clumped distribution, often resulting in locally high densities.
Fig 2.
Annual fine litterfall flux derived from canopy foliage in longleaf pine forest at the Ordway-Swisher biological station.
Values are mean g m-2 yr-1 ± 1 SE from 27 litterfall traps; a) mass, b) nitrogen, and c) phosphorus.
Fig 3.
Decomposition and nitrogen and phosphorous dynamics of longleaf pine needle litter, turkey oak leaf litter, and mixed pine and oak litter in litterbags on the surface of the forest floor and buried beneath pocket gopher mounds over a four year period; a) percent of the initial mass remaining, b) percent of the initial nitrogen content remaining, and c) percent of the initial phosphorous content remaining. Values are means ± 1 SE. Litter types and location that are significantly different (p < 0.05) have different letters. Significance levels for percent mass remaining were calculated based on models for decomposition coefficients, k, calculated for sets of litterbags in each 0.5 ha plot. Complete statistics are in Table 2.
Fig 4.
The relationship between cumulative mass loss and nitrogen concentration in litter in litterbags on the surface of the forest floor and buried beneath pocket gopher mounds over a four year period; a) longleaf pine litter, b) turkey oak litter, c) mixed pine and oak litter. Complete statistics for the regression lines for surface (solid lines) and buried (dotted lines) litterbags are in S3 Table.
Table 1.
Initial carbon, nitrogen and phosphorus content of longleaf pine and turkey oak litter in pine, oak and mixed litterbags.
Values are means ±1 SE; n = 15 for C and N contents, n = 10 for P content.
Table 2.
Results of analyses with linear mixed models for the decomposition parameter (k), nitrogen mass remaining at 24 and 48 months, and phosphorus mass remaining at 24 and 48 months in pine, oak, and mixed litter on the forest floor and buried beneath pocket gopher mounds.
Satterthwaite approximation was used to calculate degrees of freedom, df.
Fig 5.
Fine root ingrowth into longleaf pine and turkey oak litter in litterbags on the surface of the forest floor and buried beneath pocket gopher mounds over a four year period.
Values are mean g fine roots / litterbag ± 1 SE.
Fig 6.
Model simulations of litter mass, nitrogen, and phosphorus content on the forest floor of a longleaf pine forest.
(a-c) Litter layer dynamics of an individual pocket gopher mound over a 10–year period predicted by the single mound simulation; (a) mass of pine needle and oak leaf litter, (b) nitrogen in pine needle and oak leaf litter, and (c) phosphorus in pine needle and oak leaf litter. Burial occurred at year zero, and the appropriate litter decomposition rates and average litterfall values were used in simulations. (d-f) Simulated litter layer dynamics of the forest floor as a function of new mound formation at 0%, 1%, 2.3%, 5% and 10% of the forest floor covered per year over an eight–year period predicted by the mound density simulations; (d) mass of pine needle and oak leaf litter, (e) nitrogen in pine needle and oak leaf litter, and (f) phosphorus in pine needle and oak leaf litter. Vertical lines above each bar indicate variation (± 1 SE) in the amount of litterfall mass, nitrogen and phosphorus. (g-i) Simulated litter layer consumption, nitrogen volatilization, and phosphorus pyro-mineralization during low-intensity fires occurring at a five-year return interval predicted by the mound density and fire simulations; (g) consumption of pine needle and oak leaf litter, (h) nitrogen volatilization from pine needle and oak leaf litter, and (i) phosphorus pyro-mineralization from pine needle and oak leaf litter. Simulated rates of new mound formation were 0%, 1%, 2.3%, 5% and 10% of the forest floor covered per year. Vertical lines above each bar indicate variation (± 1 SE) in the amount of litterfall mass, nitrogen or phosphorus.
Fig 7.
Simulated mass consumption, nitrogen volatilization, and phosphorus pyro-mineralization of the litter layer at three fire return intervals and five rates of pocket gopher mound formation.
Simulated fire return intervals are 3, 5 and 10 years, and rates of new mound formation are 0%, 1%, 2.3%, 5% and 10% of the forest floor covered per year. Values are percent reduction of (a) mass consumption, (b) N volatilization, and (c) P pyro-mineralization predicted by the mound density and fire simulations compared to prescribed fire only simulations.