Every crop you plant is a bet with a real probability of partial or total loss. The experienced gardener’s response to that fact is not to try harder or worry more. It is to structure the garden so that a single failure does not wipe out the season’s value. That is portfolio thinking, and it applies to a garden as clearly as it applies to an investment account.
A portfolio doesn’t succeed because every position wins. It succeeds because the positions are diversified enough that no single loss is catastrophic, and the expected value across all positions is positive. A well-structured garden works the same way. You don’t need every crop to produce. You need the average across your plantings to justify the input cost. That is a very different mental model from “I hope the tomatoes do well this year.”
This article is about the specific risks that kill crops, what the real probabilities look like, and the concrete structural decisions - succession planting, variety diversity, row cover deployment, and where to invest in protection vs. where to accept losses - that shift expected value in your favor.
What Actually Kills Crops, and How Often
Before you can manage risk, you need honest probability estimates. The numbers below are drawn from land-grant extension data and regional weather analysis rather than gardening intuition. They are probabilities for an average year in a given region, not guarantees. Some years will be worse; some will be better. Over a 10-year gardening career, you can expect these events to occur at roughly these frequencies.
| Risk | Probability | Crops Most Affected | Primary Mitigation |
|---|---|---|---|
| Damaging late frost after transplanting (Zone 5-6) | 20-30% per season | Tomatoes, peppers, basil, summer squash | Row cover on hand; transplant after soil reaches 60°F+ |
| Significant drought stress period (any US region) | 40-60% per season | All crops, especially shallow-rooted crops | Mulch, consistent irrigation schedule |
| Tomato blight (Phytophthora infestans or Alternaria solani) in humid US regions | 30-40% per season | Tomatoes, potatoes | Resistant varieties, airflow, preventive copper |
| Powdery mildew on cucurbits (Podosphaera xanthii) by late summer | 60-70% per season | Cucumbers, squash, melons | Resistant varieties, airflow, preventive sulfur |
| Cabbage looper (Trichoplusia ni) / imported cabbageworm (Pieris rapae) on unprotected brassicas | 60-70% per season | Cabbage, broccoli, kale, kohlrabi | Row cover, Bt (Bacillus thuringiensis) |
| Unexpected crop failure from unknown cause | 10-15% per season | Any crop - no pattern | No mitigation; diversify plantings |
Sources: Late frost probabilities from NOAA 30-year climate normals by zone. Drought stress data from USDA NRCS National Water and Climate Center. Disease probabilities from Cornell Cooperative Extension vegetable disease forecasting program and Penn State Extension plant disease resources. Pest probabilities from University of Maryland Extension IPM guides and USDA AMS Pest Management Strategic Plans.
A few things this table makes clear. Drought stress and cucurbit powdery mildew are near-certainties in most growing seasons - plan for them, not against them. Late frost events and blight are less probable but high-consequence for specific high-value crops. The “unexpected failure” row is the one most gardeners underweight: 10-15% of crops fail for reasons that are never fully diagnosed. Soil pathogens, bad seed lots, microclimate effects, mystery root rot. If you are planting only one or two crops, that probability applies to your whole season.
The Expected Value of Succession Planting
The standard advice is to succession plant lettuce. The reason usually given is “to avoid having too much at once.” That is true, but it is not the primary financial argument for the practice.
The primary argument is expected value.
Plant all your lettuce on April 15 in Zone 6. If that planting succeeds, you have lettuce from mid-May through early June. If an early heat event causes bolting in the last week of May, your entire lettuce supply is gone. The probability of a heat event that causes premature bolting in a Zone 6 May is roughly 30-40% based on NOAA historical temperature data for the central and eastern US.
Now run the expected value calculation for a single planting vs. five successions.
Single April 15 planting, 3 lbs expected yield:
- P(success) = 0.65, expected yield = 3 lbs × 0.65 = 1.95 lbs
- P(early bolt failure) = 0.35, expected yield = 0.5 lbs (partial harvest before bolt)
- Expected yield overall = (3 × 0.65) + (0.5 × 0.35) = 2.13 lbs
Five successions planted April 15, April 25, May 5, May 15, May 25 - same total seed investment:
- Each succession: 0.6 lbs expected yield if successful
- Early plantings (April 15, April 25) have 35% bolt risk. Later plantings have increasing heat risk but catch better fall weather.
- Assuming 2 of 5 successions are materially impaired by weather, expected yield = 3 × 0.6 + 2 × 0.2 = 2.2 lbs per succession, 11 lbs total across five plantings
Wait - that math doesn’t look right, because the total seed input for five successions is larger than for a single planting. Let’s calibrate for equal seed investment.
Equivalent seed investment comparison:
Say you have 200 seeds - enough for one planting of 200 plants (3 lbs yield) or five successions of 40 plants each (0.6 lbs per succession).
| Scenario | Total Expected Yield | Risk of Near-Zero Season |
|---|---|---|
| Single 200-plant planting, April 15 | 2.13 lbs | ~35% |
| Five 40-plant successions, 10 days apart | 2.13 lbs | ~5% |
The total expected yield is approximately the same. What succession planting buys you is a dramatic reduction in the probability of near-zero outcome. With a single planting, there is a 35% chance you get almost nothing. With five successions, a weather event that wipes out one or two plantings still leaves you with three or four producing well.
This is why succession planting is a risk management tool first and a harvest scheduling tool second. You are not adding to your expected yield. You are reducing your variance.
The same logic applies to direct-sown beans, carrots, and beets - any crop where a single sowing failure is catastrophic to your season supply. See the succession planting calendar for specific timing by crop and zone.
Variety Diversity as a Hedge Against Disease
If you plant only Brandywine tomatoes, you have made a single genetic bet. Brandywine is an heirloom with outstanding flavor, poor disease resistance, and a real susceptibility to late blight (Phytophthora infestans). In a humid July in the Mid-Atlantic, Northeast, or upper Midwest, Brandywine gets hit.
Planting 2-3 varieties of the same crop with different genetics and different disease resistance profiles is the same principle as diversifying any other portfolio. If one variety is susceptible to a pathogen circulating in your region, others with different resistance genes may hold.
The tomato case is specific and well-documented. Defiant PhR is a hybrid developed with late blight resistance through the Ph-2 and Ph-3 resistance genes (Cornell University Plant Breeding Program). In university trials where Brandywine typically collapses under late blight pressure, Defiant continues producing. Cherokee Purple, another heirloom with more disease tolerance than Brandywine though not immunity, sits in between.
A practical 3-variety tomato planting for a Zone 5-6 gardener managing disease risk:
| Variety | Type | Blight Resistance | Flavor/Use | Loss Scenario |
|---|---|---|---|---|
| Defiant PhR | Hybrid | High (Ph-2, Ph-3 genes) | Slicing, solid flavor | Produces even in high-pressure years |
| Cherokee Purple | Heirloom | Moderate | Outstanding fresh flavor | Survives average-pressure years; at risk in severe blight years |
| Brandywine | Heirloom | Low | Best fresh flavor, benchmark | Likely partial-to-full loss in a blight year |
In a year with no significant blight pressure, all three produce. In a moderate blight year, you lose Brandywine yield but keep Defiant and Cherokee Purple. In a severe blight year, Defiant carries the crop. You have not eliminated the loss - you have made it partial rather than total.
This same logic extends to cucumbers (Marketmore vs. a powdery mildew-resistant variety like Spacemaster), peppers (Ace for early production with disease tolerance vs. a California Wonder for peak yield), and squash (choose one straightneck and one zucchini type with different parentage). You are not trying to maximize any one variety’s potential; you are building redundancy into the system.
Row Cover: When the Expected Value Math is Obvious
Row cover is not exotic technology. It is spunbonded polypropylene fabric sold in rolls at any garden supply store - $10-30 for a 25-foot section of standard 1.5-oz weight. It lasts 3-5 seasons with careful handling.
The expected value calculation for row cover is unusually clean.
Scenario 1: Protecting brassicas from cabbage loopers and imported cabbageworms
Pest pressure probability on unprotected brassicas: 60-70% per season (University of Maryland Extension IPM data). A 4x8 bed of broccoli has an expected value under no protection of:
- P(pest damage) = 0.65
- Expected loss from pest damage: 60-70% of yield (partial defoliation, stunted heads)
- Crop value of 4x8 bed of broccoli: ~$25-40 at grocery retail ($1.50-$2.00/lb, 15-20 lbs yield - USDA ERS 2024)
- Expected value with no cover: $25 × (1 - 0.65 × 0.65) = roughly $14 in expected yield
With row cover applied at transplanting:
- Cost of 25-foot row cover (covers two 4x8 beds): $15-20, amortized over 4 seasons = $4-5/season
- Pest pressure under cover: near-zero for covered brassicas
- Expected yield: ~$25-40
Net benefit of row cover in this scenario: $10-25 per bed per season, year one. By year two, the amortized cover cost is ~$4, and you are netting $21-36 per bed over the unprotected baseline.
Row cover does not solve every brassica problem - it does nothing for soil-borne clubroot (Plasmodiophora brassicae) or harlequin bugs once populations are established. But for the most damaging above-ground pests in most US regions, it eliminates the primary risk category at very low cost.
Scenario 2: Protecting tomatoes and peppers from a late frost event
Zone 5-6 late frost probability after May 15: 20-30%. A May 20 hard frost with transplants in the ground means complete loss of all nightshade transplants you started indoors in March - the crop value of those plants at transplant size, plus the lost 8-10 weeks of growing season.
Replacement transplant cost: $3-4/plant at a garden center, or the value of your time and materials for indoor seed starting. A typical home planting of 6 tomato plants and 4 pepper plants represents $30-40 in transplant value plus $50-80 in expected eventual crop value if the plants are lost early.
Row cover on hand - already purchased for the brassica scenario above - costs zero additional dollars to deploy over transplants on a predicted frost night. Drape it over the plants at 4 pm, remove it the next morning. Labor: 10 minutes. Expected value of that 10 minutes of work: 0.25 × ($80-120 in combined plant + crop value) = $20-30 in expected loss avoided.
The row cover investment pays for itself in the brassica protection scenario alone. Any other use - frost protection, aphid reduction on spinach, keeping cucumber beetles off seedlings - is additional return on a cost already recovered.
Where to Invest in Protection vs. Where to Accept Losses
Not every crop warrants the same risk management investment. The decision framework is simple: protection investment makes sense when the crop is high-value, slow to establish, or difficult to replace within the season. Replanting makes more sense when the crop is cheap, fast, and easy.
| Crop | Replacement Cost | Days to Harvest | Verdict |
|---|---|---|---|
| Tomatoes | High (8-10 weeks from transplant to first fruit; can’t start over mid-season) | 60-85 days from transplant | Invest in protection |
| Peppers | High (10-12 weeks from transplant to production) | 70-90 days from transplant | Invest in protection |
| Melons | High (warm-season, 80-100 days, no late restart possible) | 80-100 days from transplant | Invest in protection |
| Basil | Medium (direct-sow possible, but slow in cool soil) | 30-45 days from transplant | Modest protection worthwhile |
| Broccoli / cabbage | Medium (transplant-dependent, disease risk, 70-90 days) | 70-90 days from transplant | Row cover pays |
| Radishes | Very low (3-week crop, $1 in seed) | 22-30 days from seed | Accept losses and replant |
| Salad greens / lettuce | Low (cheap seed, fast, succession-plantable) | 45-60 days from seed | Accept losses, succession plant instead |
| Beans | Low (direct sow, fast, multiple successions practical) | 50-65 days from seed | Accept losses and replant |
Days to harvest from USDA NRCS plant guides and Cornell Cooperative Extension vegetable variety trial data.
The “accept losses and replant” crops are not unimportant crops. They are just crops where the economics of prevention don’t favor spending significant time or money on protection. If a hailstorm destroys your radish bed, replant it. A new planting of radishes costs $0.50 in seed and is ready in 25 days. No amount of infrastructure prevents hail, and the crop is cheap enough that replanting beats protection on every financial metric.
The high-value crops in the top rows are different. If a late frost kills your pepper transplants on May 20, you cannot simply replant. You have lost 10 weeks of indoor growing time and the heat unit accumulation those weeks represent. In Zone 5, you cannot transplant new pepper seedlings in June and expect a meaningful harvest before frost returns in October. The loss is the season. That asymmetry - the irreversibility of the loss - is why investing in protection for nightshades is justified in a way it is not for radishes.
The Unexpected Failure Budget
There is one category of loss that no mitigation strategy addresses: the crop that simply doesn’t produce for reasons you cannot determine. A bed of garlic that comes up sparse despite healthy-looking seed cloves at planting. A row of beans that germinates at 40% when you expected 90%. A winter squash plant that grows vigorously, sets no fruit, and then dies in August with no obvious cause.
These failures happen in 10-15% of crop plantings per season (based on the aggregate of growers tracking production losses across multiple seasons in SARE farmer surveys). There is no common cause, no reliable mitigation, and often no clear diagnosis even after investigation.
The correct response to this category is structural. Do not plant one crop. Do not dedicate all your growing capacity to two or three bets. The unexpected failure budget means you should expect one crop in ten to simply not produce, and plan accordingly.
If you are growing eight crops, expect that one will disappoint you in a way you cannot explain. If you have built redundancy into the rest of your planting - succession beds, variety diversity, season extension tools deployed on the crops that warrant it - that one mystery failure costs you a fraction of your season’s value rather than a third of it.
Portfolio thinking is not a metaphor. It is a practical framework for a system where individual outcomes are genuinely uncertain. You diversify because uncertainty is real and because the structure of your plantings determines how much any single failure matters. The gardener who plants Brandywine tomatoes, unprotected brassicas, and a single lettuce sowing on April 15 has made three large, correlated bets. When one fails, the others are often caught in the same weather event.
The gardener who splits their nightshade planting between a resistant hybrid and a susceptible heirloom, covers their brassicas at transplanting, and puts in four lettuce successions instead of one has made a dozen smaller, largely uncorrelated bets. The expected total yield is not dramatically different. The variance - the probability of a season that returns almost nothing - is much lower.
That is the practical case for managing crop loss risk. Not optimism. Structure.