Influence of stocking, clone, fertil ization, and weed control on above-ground biomass and soil CO₂ efflux in a Pinus radiata D.Don silvicultural trial, Canterbury, New Zealand.

Abstract

Pinus radiata D. Don is a widely planted exotic tree species in New Zealand as it is a major source of carbon (C) sequestration and industrial timber. Developing precise biomass models is the most essential step in assessing carbon sequestration potential of the forests. Common silvicultural practices comprise site preparation, weed control and fertilization, with clonal forestry playing an increasing role in improving stand productivity and wood quality. These management practices, along with environmental variables, are known to influence above- and below-ground carbon dynamics.

The experimental site was located just south of Rolleston within the Canterbury region of New Zealand. The experiment consisted of 48 permanent plots with a randomized complete block split-split design (Mason, 2008), with an arrangement of factors within four complete blocks. The main plots consisted of three levels of stocking. A first split consisted of four levels of fertilization and follow-up weed control treatment. A second split consisted of five different embryogenic clones randomly allocated to all plots. Three studies were carried out: (a) to find the best models to predict above-ground biomass for Pinus radiata; (b) to assess the effects of silvicultural treatments along with environmental variables on soil CO₂ efflux (Fs); and (c) to examine the linkage between above-ground biomass and Fs across silvicultural treatments.

In a first study, two broad procedures were implemented for biomass modelling: (a) independent, and (b) additive. In the independent procedure, linear ordinary least-squares regression with scaled power transformations and y-intercepts produced more precise models than nonlinear biomass estimation methods using power equations and no y-intercepts. In the additive procedure, models fitted in a joint generalized linear least-squares regression, also called seemingly unrelated regression (SUR), provided better goodness-of-fit statistics, standard errors of estimates, residual plots, and histograms of residuals. Compared with independent and additive procedures, additive equations fitted in SUR recorded unbiased estimates of biomass in contrast to linear ordinary least-squares regressions. SUR produced the best goodness-of-fit statistics with unbiased estimates in seven out of ten biomass components. Separate allometric equations were developed to predict biomass for six components, three subtotals of two or more components and total above-ground biomass for Pinus radiata.

In a second study, the effects of silvicultural treatments on Fs, soil temperature (Ts), and volumetric water content ( v) for the whole period of the experiment, as well as separately for each season, were evaluated using mixed-effect models. The relationships among Fs, Ts, and v were investigated by linear and nonlinear regressions. Season, stocking, and clone had a significant influence on Fs. Estimated mean Fs rate was 22.71 tonnes CO₂ ha⁻¹ yr⁻¹ ( 6.2 tonne C ha 1 yr 1). No significant effects of fertilization and follow-up weed control on Fs were observed. Autumn (27.76 tonne CO₂ ha⁻¹ yr⁻¹) and winter (15.64 tonne CO₂ ha⁻¹ yr⁻¹) exhibited the greatest and smallest rate of Fs, respectively. Greatest Fs rates were observed at 1,250 stems ha 1, without weed control and for clone 3. A soil moisture threshold was determined (i.e. 14.3%) to separate whether Fs was limited by Ts or v. Above this threshold, a clear exponential relationship between Fs and Ts was observed. The values of Ts and v jointly explained relatively high variability (27.90 48.94%) in Fs compared to simply Ts (26.63 47.82%), based on modelling across all silvicultural treatments. Seasonal changes in Ts and v influenced Fs.

In a third study, effects of silvicultural treatments on below-ground soil respiration (BSR), above-ground biomass production (AGB), the ratio (BSR/AGB), tree diameter (DBH), height (H), basal area (G), and leaf area index (LAI) were examined. Mixed-effects analysis of variance was carried out. Stocking, follow-up weed control, and clone significantly influenced above-ground production and below-ground carbon partitioning. Increased above-ground biomass production with stand density was primarily determined by the better use of site resources. Decreased BSR/AGB with stand density was mostly associated to greater resource limitation due to competition. AGB and G increased while DBH and H decreased as stand density increased. Follow-up weed control enhanced above-ground growth by reducing BSR suggesting weed control would decrease competition for below-ground resources. Clones with poorer growth above ground partitioned proportionally more carbon below ground, and vice versa. In conclusion, certain clones were more productive above-ground at the expense of less carbon partitioning below-ground, stocking controlled Fs, Ts and v, and follow-up weed control increased above-ground growth by reducing BSR compared to the treatment without follow-up weed control, which may suggest that weed control reduced competition for below- ground resources.