Radiation interception measurement in poplar: sample size and comparison between tube solarimeters and quantum sensors

Monitoring radiation interception requires frequent measurements of incoming and transmitted radiation through the canopy. When planning and setting up such an experiment, the following must be taken into consideration: type of sensor to be used, how many measurements (in space and time) to use, and the statistical reliability of the collected data. To answer these and other questions, an experiment was carried out in Scotland using container-grown stands of three poplar clones. The transmittance data were not normally distributed. The mean and variance of hourly, daily and weekly values of transmittance, calculated using tube solarimeters or quantum sensors, were computed. The relationship between the mean and the variance, for both types of sensor, showed a good fit to the model S 2 = ax b (where S 2 is the variance, x is the mean, and a and b are equation parameters). The parameters of this equation have been used to determine the required sample size for the two types of sensor on the basis of the standard error of the mean. The reliability of the average transmittance does not change when considering the hourly, daily or weekly data. The standard error stabilises with 2-3 tube solarimeters and with 3-5 quantum sensors. There are high standard errors only with very high transmittance using the quantum sensors. The average seasonal ratio between the photosynthetic photon flux density (PPFD) and solar irradiance (SI) of the incoming radiation was 1.90 mmol J- l. With closed canopies (LAI > 2.5), the transmittance measured with tube solarimeters was as much as 4-8 times higher than that measured with quantum sensors. As a consequence, the fraction of intercepted solar irradiance was lower than the fractional PPFD interception. The seasonal average radiation conversion ratio based on PPFD and $I were 0.54 g mol-l and 1.21 g MJ-u, respectively