Woods Hole Research Center

As part of the throughfall exclusion experiment, Center scientists developed and deployed a system of automated soil respiration chamber measurements. Researchers have compared manual and automated measurement systems and have carefully analyzed sources of error and bias in both.

Methods and Measurements

Automated Soil Respiration System: The automated soil respiration measurement system consists of a Li-Cor 6252 IRGA, two sets of Campbell relay controllers, three sets of solenoid manifolds and a Campbell multiplexer, which are all controlled by a Campbell CR10X datalogger. There are two sets of controls for this system, those that control the movement of the chamber top (relay controller 2) and those that controlled the flow to and from the chambers (relay controller 1). Relay controller 2 controls the solenoid manifold, which is responsible for supplying the pistons with pressurized air to raise and lower the chamber tops. The other two sets of solenoids, connected to relay controller 1, controlled the flow from the chamber top to the IRGA, through the pump and flowmeter, and back to the chamber top. The flow rate through the system was 0.7 l min±1. Air temperature for each chamber is measured using a Type T thermocouple wired to the multiplexer. The collars are constructed from a 30.5 cm (0.07 m2 surface area) diameter schedule 80 PVC collar cut to 6.5 cm depth. A stainless steel tube (5 cm long, 2.16 mm ID) served as a chamber vent to equalize pressure between the inside and outside of the chamber. One end of the collar is bevelled so that it could be inserted into the ground approximately 2 cm. The chamber top is made of schedule 40 PVC and was 35.5 cm diameter and 15 cm tall. Automobile window weather stripping is applied to the top of the collar to form an O-ring, such that when the chamber top was lowered by applying 70 psi of pressure to the piston, the chamber top sealed against the weather stripping on the collar. Each automated chamber is activated for 5 min. The first 1 min flushes out the tubing with ambient air. The chamber top is then lowered onto the collar and the headspace concentration is logged for 4 mins. At the end of this 5 min period the chamber top is raised and the next chamber activated. One flux measurement from each of the six automated chambers is sampled each half hour over the course of a day. Changes in headspace concentration are converted to flux rates in the same manner as the manual flux values.

Automated soil respiration chambers in the up (left), and down (right) positions.

This automated system was used at the Harvard forest in 2002-2003 and is currently being used at Howland forest. A new autochamber system was built and deployed at the Harvard forest in 2007 and is currently running. The same wiring design was utilized however the material of the chamber design was altered. Stainless steel mixing bowls replaced the PVC caps as chamber tops, thin walled PVC pipe replaced the schedule 80 PVC collars. The overall design is the same as presented with some difference in material.

Video of chambers:

Related Documentation:

Chamber Flow Diagram
Chamber Part list
Chamber Wiring diagrams

• Automated hourly measurements provide high temporal resolution which can be used to determine effects of diel variation in temperature and immediate responses to wetting events, thus permitting >90% of the temporal variation in CO2 efflux rates to be accounted for by an empirical model of soil temperature and litter layer water content.

• Some of the remaining unexplained variation appears to be diel. Temperature functions underestimate the diel amplitude of CO2 efflux, which increases with increasing mean daily flux. Center researchers speculate that the unaccounted source of diel variation may be a diel pattern of substrate supply allocated to roots.

• Wet-up fluxes, although significant, account for only 6% of the annual flux at the Harvard Forest and are functions of precipitation amount and the change in the litter layer water content.

• Manual sampling once per week has a high probability (>90%) of providing estimates within 5% of the seasonal total from the half-hourly autochamber measurements.

• As previously established in the literature, chambers must have properly designed vents to avoid artifacts of pressure differentials that can cause significant biases.

Key Publications

Savage, K., E.A. Davidson, and A.D. Richardson. 2008. A conceptual and practical approach to data quality and analysis procedures for high frequency soil respiration measurements. Functional Ecology doi: 10.1111/j.1365-2435.2008.01414.x.

Savage, K.E., Davidson, E.A., Richardson, A. and Hollinger, D., Y. 2009. Three scales of temporal resolution from automated soil respiration. Agricultural and Forest Meteorology 149, 2012–2021.