Soil and Water: The Engine Under the Grass · Lesson 10
Erosion, Wetlands and the Drainage Line
Settles how to triage erosion, which national datasets must never be added together, and what the law requires in a watercourse.
By the end of this lesson you can
- Triage the erosion on your own farm into a cheapest-first repair order
- Explain why the sheet-and-rill and gully datasets cannot be added together
- State the CARA duties that attach to a vlei, a drainage line and an eroded patch before you move any soil
#Two numbers, four orders of magnitude apart
Gauteng has 347 149 hectares classified at high sheet-and-rill erosion risk. Gauteng has 110 hectares of mapped gully.
Both figures come from the same national assessment (Le Roux & Smith, Grain SA, November 2014). They describe completely different things, and a farmer who reads the first as though it were the second will spend his whole erosion budget on the wrong problem.
Gabions, rock packs and check structures are what erosion looks like in the farming press, because dongas photograph well. They are the right answer in the Eastern Cape and Northern Cape. On the Highveld they are somebody else's answer to somebody else's problem, bought with your money.
#What 12.6 tonnes a hectare actually means
The headline national figure is a mean predicted water-erosion soil-loss rate of about 12.6 t/ha/yr, rising to roughly 13 t/ha/yr under annual cropland — about three times Australia's 4.1 t/ha/yr, against a soil-formation rate the same source puts at under 5 t/ha/yr.
Two honesty notes. The assessment is from 2014, a decade-old picture. And "under 5 t/ha/yr formation" is generous — much of the pedological literature puts actual formation an order of magnitude lower, which makes the deficit worse rather than better. No published South African soil-formation rate could be verified for this course, so read the comparison as "erosion vastly exceeds replacement", not as a precise ratio.
Work it on a 100 ha holding at the national mean:
- 100 ha × 12.6 t/ha/yr = 1 260 tonnes of soil per year leaving, on the modelled average
- less formation at the source's own generous 5 t/ha/yr: 100 × 5 = 500 t/yr replaced
- net deficit: about 760 tonnes a year
Seven hundred and sixty tonnes of topsoil off a smallholding in a year, with nobody watching it go and no line for it in the books. It is a modelled mean, not a measurement of your farm, so what it gives you is the order of magnitude — and the invisibility.
#Triage: three forms, three completely different bills
Sheet erosion removes a uniform skim across a whole slope. You cannot see it happening; you notice it when the A-horizon is gone and the veld no longer responds to a good season. Cover fixes it, and only cover.
Rill erosion is the first visible stage: shallow parallel channels a hand-span deep, usually below a track, a gateway or a bare patch. A rill under 30 cm deep is repaired with cover on the catchment above, a brush pack in the channel, and rest. Cheap.
Gully erosion (donga) is a capital project with a head-cut migrating uphill. It is the symptom, not the disease — the runoff that carved it was generated above it.
#The law arrives before the machine does
Farmers hear the first half of Regulation 7 and conclude the drainage line is a no-go zone. The regulation demands more: manage it, keep it working as a flood path, get written permission before draining or cultivating it.
#The vlei is the most valuable and most wrecked ground you own
Wetlands are the highest-value, most legally protected and most easily destroyed hectares on a South African livestock farm — and in the summer-rainfall interior they are the only green feed in late winter, which is exactly why they get destroyed.
Delineation follows the departmental field procedure using four indicators — terrain unit, soil form, soil wetness (gleying and mottling) and vegetation — with the outer boundary where hydromorphic indicators stop. A Katspruit or Kroonstad form, from the previous lesson, is a wetland flag before you look at a plant. (The guideline's current edition year is inconsistently cited; confirm it with the Department of Water and Sanitation or a registered wetland practitioner rather than quoting a year.)
Get it delineated once, properly, and mark it on the grazing plan.
#The highest-return hectare on the farm
One more reason the drainage line comes first. Invasive alien plants reduce national surface runoff by about 1 444 million m³ per year, equal to 2.9% of naturalised mean annual runoff, with the worst-affected catchments at 8.4% (Mossel Bay to Tsitsikamma) and 7.4% (Border, Eastern Cape) (Le Maitre et al. 2016). A figure near 2 444 million m³/yr also circulates; it is not a measured upper bound but a sensitivity analysis of what the estimate becomes if riparian invasion is as under-mapped as specialists believe.
The farm-scale point: riparian invasions cost disproportionately more water per hectare than dryland invasions. The wattle and gum in your drainage line are drinking water you paid for and destabilising banks you will otherwise pay to repair.
So before you price a gabion, do three things. Walk the drainage line and count active head-cuts, dated, so you can watch the number fall. Look hard at what your veld surface will be in September, because a bare winter surface meeting the first thunderstorm is where the real loss happens. And find out in writing which permissions you would need before moving a cubic metre of soil.
#Check yourself
4 questions — answers explained as you go
-
1A Gauteng farmer reads that his province has 347 149 ha of erosion and puts his whole erosion budget into gabions in his drainage line. What has gone wrong?
Why: The two datasets measure different things and differ by more than three thousand to one in Gauteng. The left-hand column is modelled risk, not observed loss, so a well-covered farm inside a high-risk polygon may not be eroding at all. Gabions have their place — in the provinces that actually have dongas, below a catchment already fixed. -
2Your farm has a 2 m deep gully with an actively retreating head-cut. Where does the first money go?
Why: A structure below an unfixed catchment gets paid for twice. Then armour the head-cut before anything downstream — working upstream from the mouth wastes an erosion budget. -
3Which statement about CARA Regulation 7 is correct?
Why: Farmers hear the prohibition and miss the duty. Regulation 7 is "don't degrade, and keep the flood path clear", and Regulation 13 adds a positive obligation to reclaim land already eroding. Doing nothing is not a safe harbour, and the National Water Act applies alongside CARA rather than instead of it. -
4You run 20% more stock than your CARA grazing-capacity figure and water them from a stream that bounds your farm. What is the water-law consequence?
Why: Schedule 1 has four conditions, and grazing within the capacity of the land is one of them, tying your water entitlement directly to CARA Regulations 10 and 11. There is no numeric volume threshold in Schedule 1 — the widely quoted 10 m³/day figure comes from registration practice, not the exemption. Module 8 takes this apart in full.
Sources for this lesson
- Le Roux & Smith, Grain SA, November 2014 — soil erosion in South Africa, its nature and distribution — The 12.6 t/ha/yr mean, the >100 000 mapped gullies, and the two separate provincial datasets
- Water SA — water erosion prediction at a national scale for South Africa — The underlying national erosion modelling method
- Conservation of Agricultural Resources Regulations, GN R1048 in GG 9238 of 25 May 1984 as amended — Regulation 7 on vleis and the 10 m strip, Regulation 8 on run-off diversion, Regulation 13 on reclaiming eroded land
- Conservation of Agricultural Resources Act 43 of 1983 — departmental text — The Act the regulations are made under
- National Water Act 36 of 1998, full text (DWS) — Section 21(c) and 21(i) water uses, section 22, and Schedule 1(1)(b)(iii) stock watering with its conditions
- SASRI Information Sheet 1.3 — understanding and managing wetlands — South African industry guidance that wetlands be managed rather than merely fenced off
- Milk SA / MPO — buffer and wetland management guideline for pasture dairy farms — South African industry buffer guidance for livestock operations
- Le Maitre et al. 2016, Water SA — impacts of invasive alien plants on water flows — Riparian invasion water cost and why the drainage line is the highest-return clearing
- SANBI / Working for Wetlands — South Africa's national wetland rehabilitation programme — A co-funding route for rehabilitation on private land