Bush and Shrub Encroachment

#The block that looks fine and carries a quarter

Walk a Highveld camp in February with knee-high green grass between the shrubs and you will conclude the veld is holding. Count the shrubs and the arithmetic changes. In Mesic Highveld Grassland, dense stands of Seriphium plumosum at 10 000 or more individuals per hectare cut grass growth and primary production by up to 75% (Graham, Barrett & Brown 2020). That is not a camp that needs a better rotation. That is a camp that has quietly become a different piece of land.

Seriphium plumosum — formerly Stoebe vulgaris — does three things at once. It is allelopathic, chemically inhibiting the germination of other species, so the gaps between plants do not fill with grass. Mature plants are hard and scratchy and are avoided by ungulates, so no grazing animal will fix this for you. And the grass it displaces is the grass your stocking rate was built on.

#The map will not tell you

The DALRRD long-term grazing capacity map is the legal reference for what your land can carry, but the department states two limits in its own background document that matter here: tree crown cover was not incorporated, and browsing capacity is not included at all (DALRRD). On encroached veld the map overstates what you can run. The polygon says 6 ha/LSU because the vegetation type says 6 ha/LSU; the shrubs are not in the model.

Water goes the same way, and further than most farmers expect. In a southern African semi-arid savanna, interception loss roughly doubled with about a thirteen-fold increase in woody cover (PMC10017313). Rain that never reaches the soil is rain you did not receive, whatever the gauge says.

#Count it, then cost it

The department's biome means give you the conversion you need: Grassland 6 ha/LSU, Savanna 12 ha/LSU, Nama-Karoo 25 ha/LSU (DALRRD). Work the loss on your own camps.

Take a 120 ha Highveld block at the Grassland mean of 6 ha/LSU. On paper it carries 20 LSU. Now suppose 45 ha of it is standing at 10 000 or more Seriphium individuals per hectare.

• Those 45 ha should carry 45 ÷ 6 = 7.5 LSU.
• At up to 75% production loss they carry roughly a quarter of that — about 1.9 LSU.
• Effective capacity on the encroached portion is therefore about 24 ha/LSU.

Read that last number again. A Highveld camp at 24 ha/LSU is grazing like the Nama-Karoo, which the department puts at 25 ha/LSU. You have not moved and the rain has not changed; the vegetation has.

In grazing days — the currency that plans a year — the clean block supplies 120 ÷ 6 × 365 = 7 300 LSU-grazing-days. Encroached, it supplies (75 ÷ 6 × 365) + (45 ÷ 24 × 365) ≈ 4 563 + 684 = 5 247. You have lost about 2 050 LSU-grazing-days off 37% of the area — roughly 28% of the farm's year, on a block that still looks green in February.

Two honesty notes on that arithmetic. The 75% is an upper bound at high density; if your stand is thinner and the real loss is half that, halve the answer. And national extent and cost figures for encroachment circulate widely — secondary sources put South African encroachment in the order of 10–20 million hectares, with an accompanying national annual loss estimate in the tens of millions of US dollars, undated in the sources traced here (Ecosystem Services). Neither figure could be confirmed against a primary departmental or peer-reviewed South African source, so treat both as order-of-magnitude and do not build a budget on them. Your own camp count is worth more than any national number.

#Zero is not the target

This is the finding that should change how you think about the job. The same study measured Seriphium stands across a density gradient and found that plant diversity peaked in the middle, not at either end.

So the ecological answer is not eradication. On the measured gradient, driving a camp from 1 500 down to 24 individuals per hectare buys you grass and costs you diversity. Driving it from 9 500 down to something intermediate buys you both. The study reports the pattern rather than the mechanism, so do not over-read it — but it is enough to say that zero shrubs is not the ecological target, even where it is the tidy one.

That is a genuinely useful management target, because it is also the cheap one. Chasing the last shrub is where control budgets go to die, and it is the part of the job with the weakest ecological justification.

#What actually controls it

Three levers, strongest first: fire at adequate fuel load; browsers; mechanical or chemical control.

Fire is lever one, and the previous lesson has already set its terms. The shrub job means a burn before the first spring rains, hot and dry, and it needs 3–4 t DM/ha of flammable grass to carry (SA Grain). That is the trap encroachment sets: the more advanced the encroachment, the less grass there is to burn with, and the less able fire is to fix it. Fire is a lever you lose as the problem gets worse — which is the argument for acting at intermediate density rather than waiting.

Browsers are lever two. Goats and browsing game species eat what cattle will not, and the current South African work is explicit that fire and herbivory together control shrub encroachment and restore herbaceous vegetation in mesic grassland, where either alone is weaker (Plant Ecology & Diversity, 2025; Frontiers in Ecology and Evolution, 2023). The goat and game modules take the animal side of this further — the stocking arithmetic for a browsing enterprise runs on separate units from your grazing arithmetic, because the departmental map covers grazing capacity only.

Mechanical and chemical control is lever three, and it is the expensive one. Regrowth is the norm without follow-up, so a single clearing operation with no follow-up plan is money spent on a delay. If you use a herbicide it must be one registered for that plant under the Fertilizers, Farm Feeds, Agricultural Remedies and Stock Remedies Act 36 of 1947 and used strictly according to its label. This course gives no product, no rate and no mixing instruction, and neither should anyone else who is not the label.

#It is not optional

Note what Regulation 16 lists among its remedies: reducing livestock numbers. The regulator's view is that encroachment and stocking are the same file. And note the interaction with the previous lesson — if fire is your chosen method, the Regulation 16 duty does not release you from the Regulation 12 permission, the firebreak duty or the liability position. The duty to act and the permission to act are separate instruments.

#Do not let a carbon argument stop you

#What to do on Monday

Pick your worst camp, mark three plots, count the shrubs, and convert the count into lost LSU and lost grazing days using the arithmetic above. Then decide whether that camp still has enough grass to burn with — because if it does, you have a lever this year that you may not have in three.