Divisi

More than 90% of the seed planted across sub-Saharan Africa comes from farmers' own fields and neighbours, not agro-dealers - which means how well you choose and store it matters more than most people realise.

Not every crop is worth saving seed from. Open-pollinated maize, beans, cowpea, groundnut, sorghum, millet and rice breed true, so the seed grows into a plant just like the parent. Hybrid (F1) maize does not - in Kenya, recycled hybrid seed lost 16-20% of its yield in the first replant and 32-46% by the third. Keep buying that fresh, or switch to a good open-pollinated variety you can save for free.

Choose your seed before harvest, not from the grain heap at the end. Walk the field and flag 30-50 of your healthiest, best-yielding plants of one variety, harvest them separately, and keep only full, clean grain. Dry it to about 8% moisture - a maize kernel should crack when you bite it, not dent - and store it airtight, cool and dark; seed left in an open tin can lose half its germination within six months.

Before you plant, test it: roll 50 seeds in a damp cloth, count the sprouts after 5-7 days and double the number for your germination percentage. Below about 70%, don't risk it.

Send this to whoever's saving seed off the heap this year.

A banana mat is a perennial business - one well-managed clump can keep earning for years across the East and Central African banana belt, from Uganda and Kagera to Rwanda, Burundi, eastern DRC and the plantain zones of Cameroon. A stormy El Niño season threatens that in three separate ways, and each one is cheap to head off now.

Wind topples heavy bunches because the roots are shallow under all that weight - improved matooke hybrids can carry bunches up to 51 kg. Prop every bunching plant with a forked pole, and dig a test hole a hand deep after the next downpour; if water is still sitting in it a day later, that plot needs drainage before the next storm, not just more mulch.

Keep only three stems per mat - a fruiting mother, a half-grown daughter, a young follower - and cut everything else at ground level. A crowded mat is a windy, disease-friendly one.

Wilt has no chemical cure, so tool hygiene is the whole defence. Banana Xanthomonas Wilt once infected a third of Uganda's farms and cut yields by up to half, but a simple routine - snapping off the male bud by hand, cutting out only the diseased stem, and flaming your panga for 20-30 seconds between cuts - dropped incidence below 10% within a month in one pilot.

Send this to whoever's mat is carrying a heavy bunch right now.

Ghana and Côte d'Ivoire grow about 70% of the world's cocoa, and in the last El Niño, global output fell almost 11% while Ghana's crop hit a 14-year low. That swing isn't just weather - it's decided farm by farm, by shade, sanitation and drying.

Cocoa wants 1,500-2,500 mm of rain a year and a dry season under three months. Push past that and disease climbs fast: one Cameroon study found 65-70% of pods hit by rot once rainfall passed 2,200 mm. Keep your shade at 30-40% - roughly 16-18 tall trees per hectare, spaced 24 m apart - and you cool the canopy without keeping it so damp that black pod takes hold.

Black pod itself is beaten at the source, not with a spray. Harvest ripe pods every week and bury every infected one - a Cameroon trial that did this cut the black pod rate by 22-31% in its first season. Add just three well-timed copper sprays instead of the usual six, and yield rose 10.9-51.8% while paying back nearly twice its cost.

None of it matters if the drying goes wrong at the end. Ferment 6-8 days, turning every two days, then dry slowly to 7-8% moisture - a Ghana trial that did this hit 98% bean purity, against 18% mouldy beans from a rushed five-day ferment.

Send this to whoever's canopy has gone too thick this year.

The cheapest fertiliser on your farm doesn't come in a bag - you grow it between seasons for the price of seed.

A legume cover crop like mucuna, lablab or cowpea pulls nitrogen out of the air and stores it in its leaves and roots - the same nitrogen you'd otherwise buy. A strong stand of mucuna or sunn h**p can hold 190-320 kg of nitrogen per hectare, and a well-run mucuna relay lifted the following maize crop by up to 1.9 tonnes per hectare in East Uganda trials - around 38 extra bags of 50 kg.

Match the legume to your rainfall: lablab, cowpea and pigeonpea for short, uncertain rains; mucuna, jack bean or sunn h**p where the wet season runs long. Sow it into your maize at knee-height, 6-8 weeks after planting, and slash it as it starts to flower - that's when the leaves carry the most nitrogen, no tool needed to tell.

It also does double duty in a heavy-rain year: the living cover breaks the force of the rain and cuts how much topsoil you lose.

Send this to whoever's field sits bare between seasons.

Bees don't need your water, your fertiliser or space you could plant something else in. They need a shady corner and something in flower nearby - and in return, one improved hive gives you around 20-30 kg of honey a year, close to a full debe, while pollinating the beans and pumpkins growing near it for free.

Start with a Kenya Top-Bar Hive - it's the cheapest to build, needs no extractor, and lets you lift and check one comb at a time. An improved hive like this yields more than double what a traditional log hive gives (26.5 kg against 10.4 kg on average), and you don't need to buy bees: rub the inside with beeswax, set it near flowering trees and water, and a wild swarm will usually move in within a few weeks.

Site it at least 100 metres from the house, footpaths and tethered animals, with a hedge in front so the bees rise above head height as they leave - the African honey bee defends its home hard and can chase an intruder over 200 metres. Only harvest comb that's at least three-quarters capped with white wax; that's your sign the honey is dry enough to keep for years without spoiling.

It fills while you're doing everything else on the farm. Send this to anyone with a quiet corner going spare.

The onion you dig up today is worth the least it will ever be — because every other farmer nearby is digging up theirs at the same time.

Cure it properly and it holds 3-5 months, long enough to sell once the market runs dry and the same bulb is worth far more.

Start with a short-day variety — Red Creole, Bombay Red or Mang'ola Red. Near the equator, long-day types just grow leaves and never bulb, which is a costly mistake if you plant the wrong seed. Transplant seedlings at 45-55 days with 3-4 true leaves (never past 60, or yield collapses), spaced 20cm x 10cm for the biggest marketable bulbs.

Scout weekly from the 4-5 leaf stage: open the leaf neck and count thrips. At 1-3 larvae per leaf, act — thrips-damaged plants showed 100% infection with purple blotch versus 30-40% on clean plants, and that fungus alone can cost 85% of the crop unchecked.

Harvest when about 80% of the tops have flopped over and the necks have gone soft. Cure the bulbs in shade with airflow for 14-20 days, until the neck is dry and shrunken and the skins rustle, then store in a raised, ventilated crib — never sealed plastic, it sweats and rots the crop.

Send this to whoever in your group sells the same week everyone else does.

Late blight doesn't wait for you to notice it — the leaf lesions can spread within a day, and by then it may already be in the tubers.

In a wet El Nino year, it can destroy 60-100% of an unsprayed potato crop, as recorded in Uganda. The plants that come through it are the ones sprayed BEFORE the first spot, not after.

Start with certified, blight-tolerant seed — 30-80g tubers spaced 75cm between rows and 30cm in the row. Ridge up at the 4-8 leaf stage and again 2-3 weeks later, along the contour so water drains off the slope instead of through your crop.

Spray a protectant fungicide (mancozeb or chlorothalonil) every 7-10 days from the moment plants emerge, tightening to every 4-5 days if the leaves stay wet past 8-9am for three days running. Once you spot blight nearby, add a systemic fungicide tank-mixed with the protectant, every 5-7 days. Stop spraying about 14 days before you dig, to clear the label's waiting period.

Cure the tubers for a week or two in shade with airflow, then store ware potato dark at 6-8C and seed potato in diffused light — dark stops greening, light keeps seed sprouts short and sturdy.

Send this to whoever's about to plant potato this wet season.

Last season, Madagascar's Grand South went 200-300mm short of normal rain between October and December — while the east coast took three cyclones in three weeks, each dropping up to 300mm in a single week. Same window, two completely different problems, and your plan depends on where your plot sits.

In the deep south — Androy, Anosy, Atsimo-Andrefana — cassava is your anchor: its roots hold in the ground for 8 to 12 months until you need them, and even simple methods lift 5-12 tonnes per hectare. Add orange-fleshed sweet potato for the vitamin A, and never take a cutting from a plant with yellow-mottled, twisted leaves — that's cassava mosaic disease, and there's no spray for it, only clean cuttings.

On the east coast, get ready before November: seal your grain store, clear drainage so a flooded field empties within 24 hours, and keep spare seed aside to replant fast once the water drops. Track Meteo Madagascar's alert colours — red means impact within 12 hours.

In the highlands, a short-cycle, cold-tolerant rice nursery, ready to transplant the day water arrives, keeps erratic rain from deciding your season for you.

Send this to a farming friend on a different part of the island — their plan probably isn't the same as yours.

See the week's Daily Quiz farming topics, participation and lessons shared by the Divisi community.

The maize seed at the agro-dealer doesn't wait for you to find the money. Miss the planting window by two weeks and you're already behind for the season.

A village savings group — a chama, stokvel, susu or tontine — is one way farmers get around that without a bank. 15-25 people you trust save a small amount every week into a shared pot, and members can borrow up to three times what they've put in, repaid within three months at a monthly charge the group sets itself (commonly around 10%, though groups range from 5% to 20% — read the total cost across the whole loan, not just the monthly number).

According to a Mastercard Foundation review, members have typically seen their savings grow 30-60% over a year, because the interest paid by borrowers stays inside the group instead of going to an outside lender. Time the annual share-out for when the rains are due and you can walk into the agro-dealer with cash already in hand.

If you've already got a harvest to store, warrantage works differently — you borrow against the stored grain itself (always for less than it's worth) and sell later, when a study of East African markets found maize prices rising 25-40% by the lean season.

Borrow only for seed and fertiliser, and never for more than one normal harvest can repay.

Send this to whoever in your group is deciding what to save toward this season.