• About this site|
• Site index|
• Contact us|
• Jobs

Quantification of the Flood and Erosion Reduction Benefits, and Costs, of Climate Change Mitigation Measures in New Zealand

Disclaimer

1 Introduction

2 Physical Effects of Afforestation and Reversion on Flood Occurrence and Other Hydrological Phenomena

3 Physical Effects of Afforestation and Reversion on Sediment Yield

4 Physical Effects of Afforestation and Reversion Effects on Erosion

5 Estimates of Monetary Benefits / Costs

6 Recommendations on Methodology for a National Perspective

7 Summary and Conclusions

8 Bibliography

9 Acknowledgements

4 Physical Effects of Afforestation and Reversion Effects on Erosion

This chapter deals specifically with effects on terrestrial erosion (ie, loss of soil from the land surface), as opposed to sediment yield. Terrestrial erosion includes removal of soil by mass movement processes, as well as by water or wind. There is a tendency across New Zealand agencies to regard sediment yield as a corollary of erosion rate; and to view reductions in sediment yield to rivers and estuaries, as being the principal benefit of erosion or flood control. This view stems from a preoccupation with water quality and aquatic habitat, and ignores the onsite effects of terrestrial erosion. It ignores the reality that much terrestrial erosion debris either does not enter a watercourse; or if it does, is swiftly deposited back on land. It also neglects to recognise that terrestrial erosion scars and debris have on-site effects, economic as well as environmental. Blaschke et al (2000) note a similar lack of interest in the on-site effects of erosion (especially mass movement erosion) in the international literature as well.

This chapter concentrates on measured evidence of reductions in terrestrial erosion from afforestation and reversion. The ‘bare ground’ percentages in the tables are from surveys that have measured fresh erosion of all kinds (mass movement, water or wind).

Older reviews of erosion control in New Zealand (eg, MacCaskill, 1974; Poole, 1981) describe techniques that were widely used from the 1940s through the 1970s, but cite few measured reductions in erosion. Benefits are anecdotally described as being greater farm production, revenue from timber, or reduced damage repair costs.

Field surveys or aerial photographic surveys were carried out by the Ministry of Works and Development (MWD) or catchment boards from the 1940s, to ascertain the extent of erosion in catchments prior to designing soil conservation works schemes. The early surveys entailed assigning visual rankings of erosion severity (usually on a scale of 1 to 5) – a good procedure to determine priorities, but uninformative as regards effectiveness of works once installed, or the amount of erosion.

Measurements of erosion did not start until the 1970s, when university and MWD researchers and a few catchment board staff started to investigate differences in erosion amongst planted trees, compared with pasture and native cover. Their results started to appear as published papers and reports from the 1970s. It was not until the late 1990s that enough measurements had accumulated from around the country, to be compared and compiled. Some initial evaluations, carried out by staff of DSIR Land Resources, exist as internal reports (eg, Harmsworth and Page, 1991) or as contract reports (eg, Blaschke et al, 1992; Clough and Hicks, 1992). Since then several compilations have been published (notably Crozier et al, 1993; Glade, 1996; Glade and Crozier, 1996).

What follows is a summary of key findings from the unpublished reports (that contain more extensive data lists) as well as the published reviews. As with earlier sections, subsequent discussion will focus onto MfE’s brief, ie, what do these summary data actually tell us about terrestrial erosion benefits of any afforestation or reversion carried out for the purpose of carbon sequestration?

Summary of New Zealand data

Table 8 summarises storm damage survey information about the spatial extent of soil erosion under pasture, planted forest, natural forest and scrub.

Points to note about Table 8 are:

• storm damage surveys since 1970, that compare pasture with bush or planted forest, show that the area of soil eroded by storms is consistently less (but not zero) where forest is planted or scrub is allowed to revert, or bush is retained

• reductions vary a great deal from survey to survey, but are mostly in the 50 to 90% range (when bush, scrub or forest data are re-computed as percentages of pasture data)

• publications give hectares eroded for each vegetation cover, and usually also express this figure as a percentage of surface area. Few give margins of error for the measurements.¹³

Table 8 confirms a widespread perception that less erosion occurs in forested or scrubby terrain, than on land in pasture. However bush and scrub do not provide total protection, especially in higher-intensity storms (a fact often overlooked by advocates of afforestation / reversion options).

Erosion data from state-of-environment surveys

A second source of information are regional point samples of erosion, collected by one of the authors (DLH) and colleagues since 1998 for seven of the country’s regional authorities. Relevant data from these surveys are summarised in Table 9.

Points to note about Table 9 are that:

• The first three surveys give percentage of sample points eroded. This is not the same as percentage area eroded (despite some optimism in the accompanying contract reports, that this would be the case). They could only be used to estimate reductions in erosion, if one assumes that average area eroded per sample point is the same for different land uses.

• More recent surveys give percentage area eroded for each land use (from measurement of 1-hectare areas around each sample point). So they can be used to estimate reductions in erosion, for the regions concerned.

All surveys give standard errors for sample means; an indicator of reliability when extrapolating sample means to areas from which each sample was drawn (the regions).

Table 8: Vegetation effects on terrestrial erosion (from storm damage surveys)

View vegetation effects on terrestrial erosion (from storm damage surveys) (large table).

Table 9: Vegetation effects on terrestrial erosion (from point sample surveys)

View vegetation effects on terrestrial erosion (from point sample surveys) (large table).

Table 10: Effects on terrestrial erosion (from soil conservation effectiveness surveys)

View effects on terrestrial erosion (from soil conservation effectiveness surveys) (large table)

Data in Table 9 confirm that eroded areas are reduced where land has been afforested, or allowed to revert or remain in bush. Given the large sample sizes (region-wide) and the small error margins, there is some certainty about these reductions.

The reason why reductions vary so greatly from one region to the next, is that in some, storms or wet winters preceded the date of survey. The surveys are literally ‘moment in time’ snapshots with an aerial survey camera.

Erosion data from soil conservation effectiveness surveys

A third source of information are some surveys of the effectiveness of tree planting as a means of soil conservation. These were undertaken intermittently at the request of a few regional councils wishing to find out how their tree plantings had performed in the wake of storms or wet winters. Table 10 summarises key data.

Points to note about Table 10 are:

• all surveys give percentages of soil surface area, bared by fresh erosion of whatever form. The percentages include deep-seated mass movements (slumps and earthflows) as well as shallow mass movements (soil slips and debris avalanches); also fresh erosion by running water (sheetwash, rills, gullies and streambank collapses)

• standard errors for sample means in Table 10 are somewhat large. If a sample mean of say 2% were extrapolated, the true value typically would be 1 to 3%. So percentage reductions based on these datasets are less certain, than ones based on regional datasets (Table 9).

Table 10 shows that eroded areas reduce, where tree plantings are ‘sufficient’ (closed canopy established over most or all unstable parts on a hillside). Where this is so, substantial reductions have been measured for spaced trees in pasture (whether planted or natural), for close-canopy afforestation, and for close-canopy scrub reversion / bush retention. Reductions in erosion are minimal where tree plantings are ‘insufficient’ (established on just some unstable parts in the case of spaced plantings, or not yet closed canopy in the case of afforestation).

The surveys in Table 10 do not differentiate whether trees are more effective at stabilising shallow mass movements than deep-seated ones. Nor do they differentiate tree-planting’s effectiveness for control of gully and streambank erosion, from its effectiveness for control of mass movement. Their aim was to assess soil conservation trees’ effectiveness ‘as planted’ irrespective of where.

Considerable scientific information about the effectiveness of tree planting in different situations is already available. For useful summaries, refer Thompson and Luckman (1993) and Phillips et al (2000).

Implications for forestry or reversion established primarily for carbon sequestration

Benefits of reduced erosion occur on-site and are proportionate to the extra area of soil protected from erosion. They can become significant and take the form of:

• saved pasture growth, stock shelter and stock fodder, where trees are space-planted on farmland

• timber yield, on afforested land (minus the costs of loss of pasture production from the same land)

• saved costs through not having to apply fertiliser, spray weeds or scrub, or incur other grazing costs (such as fencing and stock control) on marginal or reverting land.

The question arises, how can base figures from Tables 8 to 10 be used to estimate nationwide reductions in terrestrial erosion? It is near-impossible to apply the base figures. A brief explanation of why this is so, follows.

Some geological structures, and some soils, are more susceptible to erosion than others. Clearly these effects are present in surveys that have been carried out in different parts of the country. It would be tempting to say that Gisborne-East Coast data, for instance, apply to other terrain with shattered and crushed marine sedimentary rocks; or that Waikato-Bay of Plenty data represent typical reductions for terrains mantled by airfall tephra. However the second source of variation – storm rainfall – precludes making extrapolations along these lines.

Mass movement erosion, in particular, does not commence until storm rainfall exceeds a threshold value (Crozier and Eyles, 1980), then increases in proportion to rainfall, but tails off to an asymptotic maximum where rainfall is very high (Omura and Hicks, 1992). The publications cited in Table 8 confirm this, reporting very little erosion where storm rainfall was less than 100 mm, and antecedent conditions were dry. Where more than 100 mm of continuous rain fell, eroded area increased in proportion (though not indefinitely), as average area eroded did not exceed c.20%. Where antecedent conditions were wet, mass movements were triggered by quite small rainfalls, sometimes only 5 to 10 mm, falling as intense bursts; but again, their accumulated areas did not exceed c.20%.

Each eroded area listed in Tables 8 to 10 pertains to a particular combination of storm rainfall and antecedent soil moisture. Unfortunately there are not enough of them to plot the relationships as graphical curves for the most common vegetation cover (pasture), let alone the rest. In the absence of such curves, it would be foolhardy to extrapolate specific percentages out of the tables onto similar terrain elsewhere. To do so, would be to say that the same levels of erosion will always be attained whenever there is a storm.

An alternative is to re-compile data from Tables 8 to 10, expressing eroded areas amongst tree cover, as percentages of eroded area in open pasture for each survey. This approach does not provide exact estimates, but at least supplies scientifically defensible minima and maxima that can be applied to terrain elsewhere:

• spaced trees in pasture (whether planted or natural), reduce terrestrial erosion by 21 to 100%

• close-canopy afforestation reduces terrestrial erosion by 10 to 100%

• scrub reversion / bush retention reduces terrestrial erosion by 18 to 100%.

Three caveats must be attached to these ranges. Firstly, Tables 8 to 10 contain four instances (out of 28 surveys with comparable data) where erosion has increased rather than reduced. Second, these exceptions, plus the enormous range in reductions from the other 24, show that how much erosion is controlled – within the minima and maxima – depends on factors such as standard of tree planting and condition of reverting scrub. A more lengthy discussion of these factors is given by DL Hicks (1995).

Effect of afforestation/reversion patterns

The discussion in Chapter 2 cast doubt on a significant flood reduction effect from the afforestation / reversion. Chapter 3 also noted that sediment reduction can be a significant off-site effect, but is strongly site-dependent.

Terrestrial erosion differs from both flooding and sediment yield in that not just the effect, but also the benefits, accrue on-site. Consequently these benefits are ubiquitous wherever afforestation and reversion are allowed to happen on private land, even if their pattern is diffuse. Although unlikely to expand across more than 10 to 20% of catchments’ area in future years (see Chapter 2 discussion), these low percentages will translate to a large number of hectares where on-site reductions in erosion can be expected.

Attention is also drawn to two features of the soil conservation effectiveness surveys (Table 10), that seem to have a bearing on carbon sequestration proposals:

• Erosion can be reduced by space-planted trees in grazed pasture; though not as much as by close afforestation.

• Whether there are substantial reductions, depends on sufficiency of tree planting ie, it needs to extend over most of the unstable area.

If government is looking for forms of carbon sequestration that have erosion reduction as a significant additional benefit, then climate change mitigation incentives should include space-planted trees in pasture. The amount and permanence of carbon sequestration, associated with such plantings, perhaps also deserves greater investigation than it has received until now.

Conclusions: terrestrial erosion

Storm damage surveys, state of environment surveys, and soil conservation effectiveness surveys, provide enough data for us to conclude that large areas of soil can be protected from erosion by:

• spaced planting of trees in pasture

• close-canopy afforestation

• scrub reversion

• bush retention

in the following circumstances:

• where land is erodible

• where sufficient trees are planted (on most or all of the unstable area)

• and tree or scrub cover is maintained.

These circumstances can be achieved by the diffuse patterns of afforestation and reversion on private land, because their effects and benefits occur on-site.

However there is a substantial technical problem with estimating their magnitude. It would be scientifically invalid to extrapolate absolute areas eroded (whether expressed as hectares or percentages) from one catchment to another, in the absence of information about storm rainfall characteristics and antecedent soil moisture conditions.

An alternative approach is to apply an ‘envelope’ of reductions from Tables 8 to 10 – for instance the least and largest measured reductions in area eroded amongst reverting scrub relative to open pasture – to some area where scrub reversion is proposed. This does not provide an exact forecast of what will happen, but it does supply a scientifically defensible estimate of the minimum and maximum effects of scrub reversion on terrestrial erosion. This possibility is explored in the discussion on methods for national estimation (Chapter 7).

Summary of impact of afforestation on erosion under different scenarios

Diffuse reversion / afforestation (defined in Chapter 1) will only reduce terrestrial erosion if located on erosion-prone land. This scenario would ensure tree cover on some but not all such land in a catchment. Nevertheless it could have large benefits (saved soil, land retained in production, reduced land management costs), because the benefits occur on-site, in direct proportion to area of erodible land covered.

Widespread reversion / afforestation (defined in Chapter 1) will also only reduce terrestrial erosion if located on erosion-prone land; but if targeted, would ensure tree cover on most such land in a catchment. Whether the scenario’s benefits will be small or large, depends on how much of the catchment is erodible.

Whole-catchment reversion / afforestation (defined in Chapter 1) will reduce terrestrial erosion, because it would ensure tree cover on all erosion-prone land in a catchment. Its benefits will be large, provided a substantial part of the catchment is erodible. However this scenario also entails tree cover on other parts where land is not subject to erosion (and where no benefits from erosion control could be expected).

For areas already in forest, further effects on the incidence of erosion would require major changes in the condition of canopy and understorey vegetation, through sustained control of the animal pests that affect quantity or quality of vegetation (eg, deer, thar, goats and possums). Establishing the linkages between animal densities, vegetation response and on-site erosion incidence has been very elusive in the past.

[ Previous|Next