| dc.description | Soil respiration is the largest natural flux of carbon dioxide (CO2) to the atmosphere,
often accounting for as much as 50 percent of global gross productivity. As tropical and
subtropical forests sustain the highest rates of soil respiration globally, a mechanistic
understanding of the temporal variation of soil respiration in these forests is essential for
accurately projecting the effects of climate change on the global carbon (C) cycle. Thus,
my thesis describes the temporal variation of soil respiration at diel and seasonal timescales
in a novel subtropical moist forest, and relates this variation to relevant climatic,
edaphic, and biotic factors. I used 3-years of soil respiration measurements taken at a fine
temporal scale (i.e., hourly) to (1) determine the annual magnitude of soil respiration and
(2) identify the main environmental factors (air/soil temperature, moisture, rainfall,
relative humidity, PAR, litterfall, and nutrients) that may be driving seasonal and diel
variation of soil respiration in a novel subtropical moist forest in Arecibo, Puerto Rico.
Mean annual emissions of C as CO2 were 2,684 g C m-2 (or 28.5 Mg C ha-1 yr-1), showing
significant seasonal variation at daily, monthly, and annual time-scales. Even with a low
monthly range (4 °C), air temperature explained 70% of the monthly variation in soil
respiration, which ranged from 4.0 to 11.4 μmol CO2 m-2 s-1. I found a significant
parabolic relationship between volumetric soil moisture and soil respiration, with an
optimal moisture value of ~ 0.350 m-3 m-3, at both daily and monthly time-scales (R2 =
0.41 and 0.50, respectively). Overall, the seasonal pattern of soil respiration was expected
considering changes in air and soil temperature throughout the year, as well as the
seasonality of rainfall, but the large range in monthly mean soil respiration was surprising
and suggests changes in forest productivity across seasons. Throughout our study period,I observed significant variation in hourly soil respiration (df = 23, 4254, F = 66.06, p <
0.0001), with the diel patterns of soil respiration changing considerably in magnitude and
shape from month-to-month. Overall, soil respiration showed a bimodal diel pattern,
peaking in the morning and late afternoon with a midday drop. Soil respiration and soil
temperature were significantly, positively related (linear and exponential regressions) in
some months (R2 = 0.21 to 0.74 and p = 0.02 to 0.002), but not in others (March, July,
and August). However, across all months hourly soil respiration and soil temperature
demonstrated hysteresis (out of phase with each other), consisting of one or two
hysteresis loops, suggesting that other biophysical factors apart from soil temperature are
driving the diel pattern of soil respiration in this forest. | |
