Regulation of soil-surface respiration in a grazed pasture in New Zealand

Abstract

The work in this thesis investigated the regulation of soil-surface respiration (Rs) in a grazed pasture, located near Oxford, Canterbury, New Zealand. An environmentally controlled laboratory exercise was conducted to investigate how soil temperature (Ts) and root-zone volumetric water content (θ) interacted to regulate Rs. These data were used to develop a model that described Rs as a function of Ts and normalised θ (θn). Chamber measurements of Rs, carried out in the field approximately twice a month throughout the 2005 - 2006 growing season, were used to validate the Rs model.

A closed gas exchange system was used to measure the individual exchange rates of ecosystem respiration (Rₑ) and Rs , in order to gain knowledge of the contribution of Rs to Rₑ in this system. This knowledge was used to partition eddy covariance nighttime measurements of Rₑ into Rs and above-ground autotrophic respiration (Rₐₐ). Eddy covariance soil-surface respiration (RsE) measurements were compared to modelled estimates of Rs. Finally, the model of Rs and continuous field measurements of Ts and θ, were used to estimate total growing season Rs at the field site.

The laboratory exercise revealed that the influence of Ts on Rs was best described using an Arrhenius-type function, while the influence of θn on Rs was expressed with a linear function. The model determined that when θn> 0.90, Rs remained constant. In the field, Rs measured with a respiration chamber, was strongly influenced by Ts, which ranged from a low of 12.2 °C to a high 20.1 °C throughout the measurement period. Values of θn at the field site, ranged from 1.00 to 0.59, but were generally > 0.90, and as such, rarely limited Rs. Maximum and minimum rates of Rs in the field were 11 and 6 µmol CO₂ m⁻²s⁻¹ respectively. The field soil-surface respiration rate, normalised to 10°C (R₁₀) was 4.7 µmol m⁻²s⁻¹. The model developed in the laboratory exercise was able to explain half (r²=0.52, P<0.05) of the variability observed in the field Rs measurements.

The partitioning exercise revealed that Rₑ was comprised of 84% Rs and 16% Rₐₐ. There was a significant relationship between the contribution of Rs to Rₑ and leaf area index (L).

There were 62 nights of valid eddy covariance Rₑ data from the field site. Maximum and minimum nightly averaged Rₑ were 12 and 2 µmol m⁻² s⁻¹, repetitively. Eddy covariance soil-surface respiration (RsE) was determined by multiplying nocturnal eddy covariance Rₑ measurements by the fraction 0.84, as determined in the partitioning exercise. The eddy covariance soil-surface respiration measurements were generally lower than modelled Rs estimates. For nights of valid eddy covariance data, the model total Rs (44 mol CO₂ m⁻²), was 57% greater than the total RsE (28 mol m⁻²). Total growing season Rs, as estimated by the chamber based model was 162 mol m⁻² (1.94 kg C m⁻²).

This study has shown that Rs in this temperate grazed pasture is a function of both Ts and θn, however, during the measurement period, θn rarely reached low enough levels to limit Rs.