2.1 Assessing changes in soil
2.2 Soil chemical status
2.3 Soil physical condition
2.4 Organic matter dynamics
2.5 Weed spectrum and intensity
Two important questions are:
Dyck et al. (1994) and Nambiar and Brown (1997) have recently summarised recent research. Another older but useful references is Chijioke (1980). However, it is important to be cautious: tree rotations are long, even in the tropics, compared with most research projects! Furthermore, most research has been on conifer plantations.
It is usually difficult to establish conclusively what soil changes result from forestry practices. Sound baseline data are uncommon and it is frequently difficult to prove a link to plantation silviculture. There are remarkably few examples of changes supposedly induced by growing trees that lead to less favourable conditions for that species. Equally, the irreversibility of changes has rarely been demonstrated, apart from obvious physical losses such as erosion of topsoil. A gradual trend, perhaps observed over several decades, can be quickly reversed as stand conditions change. As Nambiar (1996) points out “the most striking impacts on soils and hence productivity of successive crops occur in response to harvesting operations, site preparation, and early silviculture from planting to canopy closure.”
Most reports of site change in plantation forestry derive from matched plots. Increasingly today long-term experiments are being specifically designed to investigate change, e.g. CIFOR’s tropics-wide study (Tiarks et al. 1998), the network in USA (Powers et al. 1994), and those monitoring gross environmental change such as the Europe-wide extensive and intensive forest monitoring plots (level 1 and level 11). Modelling is widely used but has limited precision because of the assumptions made.
The older observational approach is often biased as the investigation follows signs of poor tree growth or health. It also suffers from soils being notoriously variable in space and time. Single plot or small sample comparisons can be wildly unreliable. Improving reliability requires intensive sampling within a plot at specific times and matching many, not just a few, pairs of plots.
Thus there is great danger of drawing conclusions from limited investigations. Short-term studies in plantations can be grossly misleading, especially when extrapolated over whole or successive rotations.
2.2.1 Soil as a mineral store
2.2.2 Nutrient removal
2.2.3 Litter and residues
2.2.4 Measured changes in soil chemistry
Plantations may have three impacts:
Soils vary enormously as a nutrient reservoir, particularly when comparing fertile arable soils and associated crops to many forest situations. In forestry with perennial, often deep-rooted trees soil reserves are often less important than nutrient supply dynamics at the soil surface. Indeed, forests are highly efficient re-cyclers of nutrients and almost ‘leak free’ if undisturbed. In the tropics, where recycling is usually efficient, the biomass and organic matter, not the mineral soil, is the major store of nutrients. The soil - in both temperate and tropical areas - often is relatively unimportant for nutrient supply in mature plantations. It is the surface organic zone and topsoil with its fine roots, which is important in concentrating energy flow from decomposing organic matter back into plants. The integrity of this layer and how it is affected by human activities, including silviculture, is critical to sustainability.
Nutrient removal in plantations occurs when any product is gathered or harvested such as leaves, fruits, litter, logs or whole trees. Many studies have been made; Goncalves et al. (1997) alone list 12 tropical examples. The ratio of nutrient export to the nutrient store is a key measure of long-term ecosystem stability, although it is not straightforward to measure the store. For example, Lundgren (1978) found that Pinus patula plantations in Tanzania had annual removals of 40 kg ha-1 of nitrogen (N), 4 kg ha-1 of phosphorus (K), 23 kg ha-1 of potassium, 25 kg ha-1 of calcium and 6 kg ha-1 of magnesium. These rates of removal are about one-third of those of maize (Sanchez 1976) and in the Tanzania study represented less than 10 per cent of soil store i.e. a stability ratio of <0.1. In contrast, Folster and Khanna (1997) report data for Eucalyptus urophylla x grandis hybrid stands with three very different site histories in terms of the previous plantation crop(s) at Jari in NE Amazonia. To quote “Twelve of the stands were in the second to fourth rotation, indicating that most of the previously grown Gmelina, Pinus or Eucalyptus had already extracted their share of base cations from the soil and left it greatly impoverished.” The stability ratio of >1 suggests non-sustainability. However, caution is needed. Others (e.g. Rennie 1955; Binns 1962; Johnson and Todd, 1990) have predicted from comparison of removals in harvested biomass with available quantities in soil that calcium nutrition will be a problem; yet trees continue to grow on soil where conventional soil analysis suggests there is virtually no calcium!
The impact of nutrient losses depends in part on what parts of the tree are actually removed - debarked log, log, whole tree including branches etc. - owing to the highly unequal concentration of nutrients in plant tissue. In general terms if the stability ratio is greater than 0.3 they may be serious stability questions in the longer term, and if it is above 0.5 in the immediate future.
Understanding these dynamics helps identify when, in the continuum of plantation productivities, sites, species and practices, the ratio becomes critical for long-term stability. There appear to be few examples of reaching such limits. It is worth remembering that nutrient removals by forest crops are typically only one-fifth to one-tenth that of arable farming (Miller 1995).
The influence of litter on soil chemical status may be important since leaves of different species decay at different rates. For example, in southern Africa substantial accumulations may develop under P. patula on certain sites (see Morris, 1993b) while this is unusual beneath the more lightly canopied P. elliottii. In broadleaved stands litter accumulation is uncommon though not unknown e.g. under some beech and oak stands on acid soils in Europe. Even under teak and Gmelina, which usually suppress all other vegetation, the large leaves readily decay. Similarly under the light crowns of eucalypts and ash (Fraxinus spp) and the nitrogen rich foliage of nitrogen-fixing trees such as Acacia, Leucaena and Prosopis spp, alders and casuarinas litter build up is rare.
Of greater importance than such long-term processes is how the litter and organic matter layers are handled, especially during site preparation and harvesting operations (see section 2.4).
Most studies have either compared conditions before and after plantations were established or examined trends as a plantation develops. Few have examined changes over successive rotations and there are even fewer direct comparisons between plantations and farmland. Few consistent trends emerge.
In both temperate and tropical studies increases and decreases in carbon, nitrogen and macro-nutrients under plantations compared with natural forest or pre-existing conditions have been reported (Evans 1999b). Not surprisingly nitrogen accumulation is widely found under nitrogen fixing species.
Many studies in temperate plantations have focused on pH change, litter type, podzolization and so on. Recent investigations have concerned acid rain impacts, though distinguishing these from direct tree effects on soil acidity is difficult. On the whole, tree impacts are relatively small compared with the soil nutrient store (Evans 1999b).
2.3.1 Site preparation and planting
2.3.2 Impact of tree growth
2.3.3 Indirect impact of vegetation suppression
2.3.4 Harvesting damage
Plantation forestry may impact soil physical conditions, and hence sustainability through
Cultivation and drainage affect soil physics for many years and sometimes for more than one rotation. Obviously site preparation seeks to improve growing conditions for trees and not impair sustainability or productivity. Longer-term benefits include reduction in bulk density, increased infiltration capacity and aeration, improvement in moisture storage and enhanced mineralization rates of accumulated organic matter (Ross and Malcolm, 1982). Physical disruption of indurated layers and deep cultivation such as tining or ripping are actually designed to reverse ‘undesirable’ soil profile development. However, poor practices can be detrimental by topsoil removal or increasing erosion rates.
The general conclusion about water use by trees, compared with grassland and many crops, is that trees have higher evapo-transpiration rates and thus ‘use’ more water. This has actually been harnessed to lower water tables. Eucalypts and other trees are planted for this purpose to control salinity.
However, it is difficult to quantify this effect on the growth of later rotations. If a plantation loses more moisture than is received in precipitation, soil moisture will not recharge and reserves are depleted. In the US mid-West many plantations established in the early 1900s initially thrived but died once moisture reserves were used up and precipitation was inadequate to sustain growth (Kramer and Kozlowski 1979).
Roots of trees are known to strengthen soils and reduce soil mass movement on hillsides.
Plantations of teak and Gmelina in the tropics and many conifers, in both tropical and temperate conditions, may suppress all ground vegetation. Where this exposes soil, perhaps because litter is burnt or gathered, erosion rates increase. Under teak Bell (1973) found soil erosion 2½ to 9 times higher than under natural forest. The protective function of tree cover derives more from the layer of organic matter that accumulates on the soil surface than from interception by the canopy. In India, raindrop erosion was nine times higher under Shorea robusta plantations where litter had been lost through burning (Ghosh, 1978). Soil erosion beneath Paraserianthes falcataria plantations was recorded as 0.8 t ha-1y-1 where litter and undergrowth were kept intact but an astonishing 79.8 t ha-1y-1 where it had been removed (Ambar 1986). Wiersum (1983) found virtually no soil erosion under Acacia auriculiformis plantations with litter and undergrowth intact, but serious where local people gathered the litter. In Jamaica, Richardson (1982) reported that the dense needle mat under pine plantations was better than natural forest for minimising soil erosion.
Extracting trees from a site can cause soil compaction, scouring of soil surface and erosion, blocking of ditches and other drainage channels, and oil spillage. The method of extraction greatly influences the extent of damage with draft systems using mules, oxen etc being least harmful and skidding with tracked vehicles generally most damaging. Weather conditions and the type of soil also affect the severity of damage. Compaction often results from harvesting clay soils in wet conditions. There are many reports of impaired growth of planted trees on extraction routes and where soil has been compacted and suffered erosion (see Nambiar 1996).
The litter and organic matter layer at the soil surface is critical to sustainability for three reasons:
Activities that disturb these organic matter roles can have large effects. Perhaps most serious of all is regular and frequent litter raking or gathering. In commercial plantation forestry managing debris during site preparation or after harvesting is costly, but as Nambiar (1996) states ‘one shoddy operation can leave behind lasting problems’. The examples of yield decline discussed in section 3 usually include harmful practices to litter and organic matter.
Establishment and reestablishment of plantations greatly affects ground vegetation with many operations designed directly or indirectly to reduce weed competition. The objective of weed control is to ensure that the planted tree has sufficient access to site resources for adequate growth. Once canopy closure has occurred weed suppression is usually achieved for the rest of the rotation.
In subsequent rotations the weed spectrum often changes. Owing to past weed suppression, exposure of mineral soil in harvesting, and the accumulation of organic matter, conditions for weed species change. Birds and animals may introduce or spread new weed species, grass seed may be blown into plantations and accumulate over several years only to flourish when the canopy is removed. Roads and rides in plantations can become sources of weed seeds. Weed management must be an holistic operation. Often where yield declines have been reported in subsequent rotations the significance of changing weed competition has not been recognised.
