SEASONAL PARTITIONING OF EVAPORATION BETWEEN TREES AND UNDERSTOREY INA WIDELY SPACED PINUS-RADIATA STAND

1. Tree transpiration, E(t), and understorey evaporation, E(u), were m easured hourly on 2-3 fine days during 12 periods throughout a year in a Pinus radiata plantation. There were 450 stems ha-1 and 60% of the herbaceous understorey vegetation and forest floor was covered by dead stems, branches and foliage from earlier thinning and pruning operati ons. Annual rainfall was 1154 mm and the trees and understorey vegetat ion were supplied adequately with water throughout the year. 2. When s tomatal conductance, g(s), was normalized to a maximum value, g(max), for each shoot, 87% of the variability was attributable to air saturat ion deficit above the forest canopy, D, and incident quantum flux dens ity, Q. g(s) increased rapidly in response to increasing Q and, when D less-than-or-equal-to 0.4 kPa, g(max) was obtained when Q < 200 mumol m-2 s-1. g(s) decreased linearly in response to increasing D. For fol iage expanding in the current year, g(max) increased by 90% between ea rly and late summer and this was accompanied by a 51% increase in the sensitivity of g(s) to D. In winter, night frost depressed g(s) by 33% at midday and throughout the following day. 3. The dynamics of foliag e development and loss were used to scale the seasonal change in tree canopy leaf area index. Foliage growth began in spring and 97% was com pleted after 5 months, although growth continued for 7 months in all. A second, smaller flush of foliage growth began in mid-summer and cont inued for 4 months. Loss of older foliage began in mid-summer. At the start and the end of the year, leaf area index was 3.2 and 5-2 (all-su rfaces basis), respectively, and maximum leaf area index occurred in s ummer, which coincided with maximum g(s). Canopy conductance and thus E(t) were very sensitive to increases in leaf area index. 4. The avera ge daily transmittance of shortwave radiation through the tree canopy on a clear day in late summer was 0.45. This increased to 0.57 in clou dy conditions. Similarly, transmittance through the tree crowns plus t he dead trees and branches was 0.11 and 0.16 for clear and cloudy cond itions, respectively. E(u) was closely related to the available energy and exceeded the equilibrium rate, except during mid-summer when the soil near the surface was dry. 5. E(t) varied from 0.5 mm day-1 in win ter to a maximum of 1.3 mm day-1 in late summer and E(u) varied betwee n 0.3 mm day-1 in winter and 1-6 mm day-1 in early summer. Estimates o f E(t) + E(u) were between 4 and 27% of independent measurements of to tal forest evaporation made using an energy balance/eddy covariance te chnique (average difference = 0.2 mm day-1). 6. The combination of wea ther variables, changes in g(max), the response of g(s) to D increasin g leaf area index and the relationship between E(u) and available ener gy led to the fraction E(t)/(E(t) + E(u)) remaining roughly constant t hroughout the year with an average of 0.52.