The impact of repeated prescribed burning in semi-mature pine plantation forests of Mpumalanga on fuel loads, nutrient pools and stand productivity
[摘要] ENGLISH ABSTRACT: Wild fires in South Africa destroy vast areas of plantations annually and a growing need exists to reduce the fire risk effectively, economically and sustainably. Periodic prescribed burning is viewed by many researchers and managers as the only cost effective method to reduce fire damage and risk, and as such is being implemented on a large scale in SA. This thesis documents the effects of repeated prescribed burning operations on fuel load reduction, tree damage and stand growth rate, as well as nutrient dynamics in the system. Five field trials in semi-mature P. elliottii and P. patula plantations were laid out in Northern Mpumalanga during January 2014 to determine the effect of under-canopy burning on fuel loads, nutrient pools and stand productivity. Each trial consisted initially out of three control, three first burn and three second burnt plots.Fuel reduction. Under-canopy burning treatments were of low to moderate intensity with an average predicted FLI ranging between 44 and 602 kW m-1. With burning, the forest floor (FF) was significantly reduced in all trials (6.3 t ha-1 and 12.9 t ha-1 in P. elliottii and P. patula trials per burning event, respectively). Small quantities of additional litterfall (that would counter the initial goal of FF reduction) were noted for a single month in 3 out of 5 trials, but the short duration of this effect meant that the cumulative effect was insignificant. Average litterfall rates per species were 5.52 and 8.82 t ha-1 yr-1 for P. elliottii and P. patula, respectively. In general, the highest litterfall rate occurred during the winter months and the lowest during the summer months. The goal of FF reduction to values around 10-15 t ha-1 can be achieved with two prescribed burns in many P. elliottii stands (with smaller FF's), but it may require at least 4 burning events in some P. patula stands with FF loads of more than 50 t ha-1. A significant reduction in understorey vegetation was still evident 8 months after treatment implementation except for one trial but species abundance returned to pre-burn levels 14 months after treatment implementation with no changes in species composition.Effects on stand condition and growth. Crown damage was restricted to one trial whilst root combined with cambium damage occurred in two trials, a result attributed to the smouldering effect of the FF. Among burning treatments, only Non-significant dbh and BA increment differences were evident, 4 and 12 months after treatment implementation but a highly significant seasonal effect could be seen in all trials: dbh increment was 61% higher in summer than in winter.Nutrient dynamics. FF nutrient loss was directly related to FF consumption by fires, with N loss varying from 33 – 119 kg ha-1; P, K, Ca and Mg losses averaging 3.8, 12.6, 72.4 and 17.2 kg ha-1. An increase in foliar macro nutrient concentration of freshly fallen needles was observed, but only during the first month after treatment implementation. Most soil nutrients showed modest changes that were not significantly different from other burning treatments. Modest increases in soil pH was observed with one in 5 trials recording a significant increase in pH (KCl) after the second burn. Total N mostly decreased with increasing number of burning treatments, but this effect was only of a significant magnitude in one out of 5 trials. Exchangeable Mg showed a small increase in all trials after burning, but the magnitude of this effect was only significant in one out of 5 trials.In conclusion: The evidence collected across 5 replicated experiments showed that significant fuel reduction (in FF and understorey) is attainable with 2 to 4 low intensity, repeated prescribed burning events under pine tree canopies. There appears to be minimal damage to trees and no significant short term response in diameter increment after repeated prescribed burns. The effect on nutrient dynamics in the system is modest and non-significant in most cases.
[发布日期] [发布机构] Stellenbosch University
[效力级别] [学科分类]
[关键词] [时效性]