Half the summer rain, the warmest summer since 1881, the soil dry down to two metres. What can still be saved in the coming weeks, why soil cover remains the most important lever — and where seeding from the air reaches its limits in dry conditions.
The figures for Baden-Württemberg have been official for a few days: summer 2026 was the warmest in the state since records began in 1881, with a new state record of 41.4 °C on 27 June in Waghäusel. Less than half the usual summer rain fell — 48.9 percent, to be exact. Lake Constance reports new record lows for the date day after day. And the farmers' association speaks of 30 to 40 percent of fodder missing on many forage farms, regionally half.
Anyone standing in the field does not need these figures. They see the maize ripening too early, the potatoes staying small, and the grassland that has not regrown after the second cut. This article aims to answer three questions that arise now: what can still be saved at all in the coming weeks? Which measure actually achieves something in dry conditions? And — because we offer seeding from the air and want to stay honest — where does drone seeding work in a drought, and where does it not?
The drought monitor of the Helmholtz Centre for Environmental Research (UFZ) distinguishes between the topsoil down to 25 centimetres and the total soil down to about two metres. The difference is decisive in practice.
The topsoil reacts quickly. One good steady rain, and the topsoil is moist again — that is the soil in which seed germinates and in which the cover crop spends its first weeks. This layer can still be influenced in the coming weeks.
The total soil is sluggish. After the record drought year 2025 and a March 2026 that the UFZ classifies as "extremely dry", it never properly refilled — spring brought only around 70 percent of the long-term precipitation nationwide, and according to the spring assessment Bavaria was the relatively driest region of Germany. Winter wheat and oilseed rape draw their water from this reserve next spring. The reserve will not fill up again this autumn, whatever is seeded now. It can only be conserved: by letting less of it evaporate unproductively.
That is the idea that carries everything else. In the coming weeks it is not about getting water into the soil — only the rain can do that. It is about keeping the water that is still there in the soil.
The most urgent question on many farms is not about next spring but about the winter. Anyone who has to buy in 30 percent of their basic fodder is looking at five-figure sums — and at prices that are rising right now because everyone is buying.
The chambers of agriculture in Austria, whose dry regions are very similar to our conditions this year, recommend for precisely this situation fast-growing cover crop mixtures: clover-grass, oat-pea-vetch mixtures, Landsberger mixture (vetch, ryegrass, crimson clover) or forage rye. Seeded in August to early September with sufficient moisture, they still deliver an autumn cut. The overwintering mixtures additionally bring a spring cut before the 2027 silage maize — given the current state of the silos, that may be the more important part.
Two things have to be said honestly here. First: the autumn cut depends on moisture. A stand that emerges in September and then sees no rain for four weeks brings no cut, but soil cover — which is valuable, but it fills no silo. Second: the Bavarian farmers' association reports from its current survey that, of all things, the cover crops seeded as fodder replacement have in many cases failed to emerge because of the drought. Anyone seeding now should choose the date by the weather forecast, not by the calendar.
The temptation is obvious in a drought year to skip the cover crop: it needs water, water is lacking, so better to grow nothing at all so that the soil does not lose even more. The figures say the opposite.
A cover crop needs around 120 millimetres of water between mid-August and mid-December. That sounds like a lot. But an uncovered soil loses up to 60 percent of its water in late summer through unproductive evaporation directly from the surface — water that benefits no plant. The cover crop consumes water, yes. The bare-fallow field next to it consumes it too, only without anything in return. Studies by the University of Natural Resources and Life Sciences Vienna (BOKU), summarised by the specialist platform Praxis-Agrar, therefore show: the covered soil loses considerably less water than the open area.
Then there is the rain that comes at some point. On a soil with 50 percent cover, around 80 percent less water runs off the surface than on open soil. The first autumn rain after a drought is exactly the one that rushes off a crusted topsoil instead of soaking in — the cover keeps it where it belongs.
After a cover crop, the pore volume of the soil is up to 15 percent higher than after bare fallow. These are the medium pores in which plant-available water is stored — precisely the water that decides the next dry June. And humus stores up to five times its own weight in water. Every cover crop that leaves root mass in the soil is a deposit into this account. The return does not come this year, but it comes.
What worries many farms about soil cover first is not the water but what happens on a brown field under the sun. An uncovered soil heats up at the surface to 50 degrees and more in late summer. Soil life retreats into the depths or dies off, and whatever organic matter lies in the topsoil is broken down quickly. The nitrogen released has no root to take it up: it lies in the soil as nitrate and is lost downwards with the first heavy rain — exactly the rain everyone is waiting for leaches it out. Farmers call this "burning off the nutrients", and the expression captures it better than any technical formula: what was built up over years is gone in one summer.
Getting it back is expensive. Anyone who wants to bring the humus and the nutrients back into the field needs organic fertilization, lime, several years of cover crops and root mass — and money for each of them. A cover crop now is the cheaper option: it shades the soil and keeps it noticeably cooler, it keeps soil life active, and it takes up the nitrogen being released and binds it in its biomass. The LTZ Augustenberg puts this binding, depending on mixture and growth, at several dozen kilograms of nitrogen per hectare, released again in spring when the following crop needs it. Soil cover is thus three things at once: water store, sun protection and nutrient account.
This is the point where many have doubts, and the data here are unusually clear. In multi-year field trials in typical dry regions with around 594 millimetres of annual precipitation — that is, less than falls at Lake Constance in a normal year — the cover crop showed no yield effect on the following crop in 73.5 percent of plots for spring barley, in 80.7 percent for sugar beet and in 85.1 percent for maize. And in the designated drought years of the trial there was no negative yield response — the sugar beet yield after a cover crop actually tended to be higher.
When choosing species, it is worth looking at water efficiency: mustard forms around four grams of dry matter per litre of water consumed, about 40 percent more than oilseed rape. Crucifers as a whole have the fastest early development and close the soil earliest — in dry conditions that is the most important move.
This is where the drought year parts ways with the normal year. Normally a cover crop means: work the stubble, prepare the seedbed, drill. This autumn many farms are facing a topsoil into which the cultivator barely penetrates — and where it does, it brings up clods that lie there for three weeks.
The Hessian agricultural office sums it up in one sentence in its assessment of drone seeding: every tillage operation promotes mineralisation and reduces soil moisture. In a year in which moisture is the scarcest resource, that is an argument against every operation that can be avoided.
The alternative is seeding without tillage — and the earliest form of this is seeding into the still-standing cereal crop, seven to ten days before threshing. The seed then lies in the shade of the stalks on a soil that still has its residual moisture, because nobody has torn it open. If rain falls, it germinates where moisture lasts longest. And it germinates two to three weeks earlier than seed that only gets into the soil after harvest, stubble cultivation and seedbed preparation. In a late summer in which every week of growing season counts, that is the real advantage — not the pass saved.
The practice is simple: apply shortly before forecast rainfall. The combines then run as normal, the straw settles over the seed and holds the moisture. If the stand has not emerged by threshing because the rain failed to come, a shallow pass with the straw harrow or disc harrow, at most three centimetres deep, is enough to make up the seed-to-soil contact — without tearing open the topsoil.
We offer this seeding, and precisely for that reason what does not work belongs here. A guide that conceals it is not a guide.
Seeding from the air has no seed-to-soil contact. The grain lies on the surface, and what it needs to germinate has to come from above. In extreme drought with no rainfall in sight, drone seeding is risky — no riskier than drilling into a bone-dry seedbed, but no safer either. Anyone with no rain in the five-day forecast is better off waiting. The drone's advantage is the timing, and that is no use if the timing is wrong.
In 2024 the Nutrinet practice research network tested drone seeding on ten farms in Baden-Württemberg — a mixture of clover, niger, sudangrass and Phacelia, applied around two weeks before the wheat harvest. The result was sobering: on eight of ten farms the mixture did not establish optimally; drilling delivered the visually better stands. The report names as causes cool sites where the heat-loving species emerged weakly, and calibration problems that had to be improvised in the field.
We mention this not to talk the method down, but because it shows what it depends on: the choice of species for the site and the clean setting of the application rate. Both can be solved. But both have to be solved before the drone flies — not afterwards.
From the trials of the agricultural offices and our own experience, four points emerge. The seed rate has to be increased by at least a factor of 1.5 compared with drilling, because some of the grains land on straw or clods. Only small to medium seeds are suitable — mustard, oil radish, Phacelia, clover-grass, niger; field bean or large lupin belong in the drill. The straw has to be distributed evenly; thick straw mats suppress emergence just as much as they hold moisture. And fields with heavy pressure from grass weeds or thistles are unsuitable, because drone seeding omits the operation that would deal with the weeds at the same time.
Undersowing is good erosion control and good soil cover — in normal years. In dry years the calculation tips. Trials in Bavaria show that water competition between the main crop and the undersown crop was considerably stronger in the dry summer of 2020 than in the rainy 2021, with noticeable yield losses in silage maize. Anyone who already has too little maize in the silo this year should not give it another competitor for the last of the water.
Anyone who already has the undersowing in the crop now has covered soil and less evaporation after harvest in return — that is an advantage this autumn. Establishing it new belongs in spring, after the winter rain, not in the drought. What we offer in this respect is on the page Undersowing by drone; for this year we tend to advise against it.
Anyone who has to decide in the next two weeks whether and how to still seed will get a long way with four questions.
Is rain forecast in the next five days? If not: wait. No method turns dust into a seedbed. If yes: act now, and before the rain, not after — after the rain the soil is not trafficable for a day or two, and the head start is gone.
Is the soil trafficable and workable? If the topsoil is so hard that the cultivator brings up clods, every tillage operation costs moisture and produces no seedbed. Then seeding without tillage is the better choice — whether by drone or direct drilling.
Is the crop still standing? If the cereal has not yet been threshed, seeding into the standing crop is the earliest option and the one that conserves the most moisture. After threshing, what remains is seeding onto the stubble, which also works without tillage — the head start is then smaller, but it is there.
Do you need fodder or soil protection? For fodder: fast-growing, mowable mixtures and a date as early as possible. For soil protection and next spring: crucifers with fast cover, ideally as a mixture with Phacelia. Both together is possible — but anyone who wants both gets a little less of each.
This autumn will not refill the soil water reserve. But it decides how much of what is still there remains until spring. Soil cover is the strongest lever a farm has in its hands for this — it keeps the water in the soil, protects it from the sun and binds the nutrients that would otherwise be lost on a brown field; the figures from dry regions are unambiguous, and the myth that a cover crop costs yield in a drought year does not stand up to the trials. Anyone seeding should set the date by the rain, not by the calendar, and where the topsoil is too hard, forgo any tillage. Seeding from the air is a tool for this — with a clear advantage in timing and a clear limit when the rain fails to come. What we specifically offer is on the page Seeding cover crops by drone. And if you are unsure whether it is worthwhile on your field, tell us — we also advise against it when advising against it is the honest answer.
Yes, but only with a rain window in sight. In the trials from dry regions, the cover crop showed no negative effect on the following crop even in designated drought years. What it needs is moisture to emerge — without forecast rain the seed lies there and waits, whether drilled or broadcast.
The term is farmers' language, the process is real. An uncovered soil heats up strongly at the surface, soil life declines, organic matter is broken down, and the nitrogen released is leached with the next heavy rain because no root takes it up. A cover crop keeps the soil cooler, keeps soil life active and binds the nitrogen until the following crop needs it. Rebuilding it afterwards costs years and money.
Clover-grass, oat-pea-vetch mixtures, Landsberger mixture (vetch, ryegrass, crimson clover) or forage rye. Seeded by early September with sufficient moisture, an autumn cut is possible; the overwintering mixtures additionally deliver in spring. The later, the tighter the autumn cut.
It works if rain comes, and fails if it does not — like any seeding without seed-to-soil contact. Its advantage in a drought: it seeds into the standing crop, without tillage that would cost moisture. If rain then falls, the head start is two to three weeks.
Then seeding without tillage is the only route that does not cost moisture. If the stand has not emerged after threshing, a shallow pass with the straw harrow, at most three centimetres deep, helps.
Better not to establish it new. Water competition was considerably stronger in trials in the dry year, with yield losses in silage maize. Anyone who already has it benefits from the cover now. Establishing it belongs in spring.
UFZ Drought Monitor Germany · UFZ spring assessment 2026 · DWD summer 2026 assessment for Baden-Württemberg · Praxis-Agrar: cover cropping in dry conditions · LK Austria: fodder shortage 2026 · Bavarian Farmers' Association: drought · LLH Hesse: drone seeding · Nutrinet: drone seeding trial Baden-Württemberg · LfL Bavaria: erosion control in maize
The four decision questions, the choice of species by site and goal, the seed rates for seeding without tillage and the dates for autumn and spring fodder — on two pages, for the edge of the field. The printable drought guide and our newsletter are currently available in German. Ask us at info@feldfalke.com or request a quote directly.
Draw your field or upload a field file, see a guide price, send your enquiry — in two minutes, no obligation. Or call if you first want to know whether it is worthwhile on your field.
Send a no-obligation enquiryStuck in the portal? Just give us a call: +49 173 3706561 · WhatsApp
