GrowingHQ

Diagnose a plant problem

Describe what you can actually see. GrowingHQ ranks the possible causes, shows what fits and what does not, and tells you the one check that separates the leading candidates — rather than naming a single cause and hoping.

What this covers, stated plainly

23 symptoms and 542 diagnostic rules. That is not every plant problem and this is not a plant doctor — a combination GrowingHQ has no rules for gets a refusal rather than generic advice that would fit anything.

Some rules are specific to houseplants, and 211 plants carry rules of their own, because one diagnostic tree does not fit every plant. Root rot from a full saucer is a houseplant’s commonest killer and a rare diagnosis for a border shrub.

Which plants have rules of their own

Agapanthus, Aglaonema, Alocasia, Aloe vera, Amelanchier, Anthurium, Apple, Apricot, Asparagus, Aubergine, Basil, Bay, Bearded iris, Beetroot, Berberis, Bird of paradise, Blackberry, Blackcurrant, Blue fescue, Blueberry, Borage, Boston fern, Broad bean, Broccoli, Brunnera, Brussels sprouts, Buddleja, Cabbage, Calathea, Californian lilac, Camellia, Carrot, Cauliflower, Celeriac, Celery, Chamomile, Chard, Cherry, Chervil, Chicory, Chilli pepper, Chinese money plant, Chives, Christmas cactus, Clematis, Climbing hydrangea, Coriander, Courgette, Crab apple, Crocosmia, Crocus, Cucumber, Curry leaf, Daffodil, Dahlia, Deutzia, Dieffenbachia, Dill, Dragon tree, Echeveria, Echinacea, Endive, Epimedium, Escallonia, Evergreen climbing hydrangea, Fatsia, Feather reed grass, Fennel, Fig, Firethorn, Fittonia, Forsythia, Foxglove, French bean, Fuchsia, Garlic, Garlic chives, Gladiolus, Globe artichoke, Golden-scaled male fern, Gooseberry, Grape, Grape hyacinth, Hard shield fern, Hardy cyclamen, Hardy geranium, Hart's tongue fern, Haworthia, Hawthorn, Heartleaf philodendron, Hebe, Hellebore, Heuchera, Hippeastrum, Holly, Honeysuckle, Hosta, Hoya, Hyacinth, Hydrangea, Hyssop, Jade plant, Japanese forest grass, Japanese maple, Jerusalem artichoke, Kale, Kentia palm, Kohlrabi, Lady fern, Lamb's lettuce, Land cress, Lavender, Leek, Lemon balm, Lemon thyme, Lemon verbena, Lemongrass, Lettuce, Lily, Loganberry, Lovage, Magnolia, Mahonia, Maidenhair spleenwort, Male fern, Marjoram, Melon, Mexican feather grass, Mexican orange blossom, Mint, Miscanthus, Mock orange, Monstera, Moth orchid, Nerine, Onion, Oregano, Ornamental allium, Ostrich fern, Pak choi, Parlour palm, Parsley, Parsnip, Passionflower, Pea, Peace lily, Peach, Pear, Peony, Peperomia, Pieris, Plum, Potato, Pothos, Pumpkin, Quince, Radish, Raspberry, Redcurrant, Rhododendron, Rhubarb, Rocket, Rose, Rosemary, Rowan, Royal fern, Rubber plant, Rudbeckia, Runner bean, Sage, Salad burnet, Salvia, Savory, Sedge, Shallot, Shiso, Silver birch, Skimmia, Snake plant, Snowdrop, Soft shield fern, Sorrel, Spider plant, Spinach, Spring onion, Star jasmine, Strawberry, String of hearts, String of pearls, Swede, Sweet pepper, Sweetcorn, Swiss cheese vine, Syngonium, Tarragon, Tayberry, Thai basil, Thyme, Tomato, Tuberous begonia, Tulip, Turnip, Viburnum, Vietnamese coriander, Weigela, Whitecurrant, Winter savory, Winter squash, Wisteria, Yew, ZZ plant.

“I don’t know” is a real answer and costs nothing. An observation you have not made counts neither for nor against any cause — it simply does not enter the reasoning.

What the questions are asking for

Three of the observations decide the answer more than everything else combined, and each is easier to recognise drawn than described.

The four leaf damage patterns, and what each points toFour leaf outlines showing distinct damage patterns. First, uniform yellowing across the whole leaf including the veins, which points to nitrogen shortage or normal ageing. Second, yellowing between the veins with the veins themselves staying distinctly green, which points to magnesium or iron shortage. Third, browning starting at the leaf edges and working inwards, which points to drought, scorch or salt damage. Fourth, browning confined to the leaf tip, which points to dry air or hard water. The distinction between the first two is whether the veins keep their colour.Uniform

Nitrogen shortage, or normal ageing

Between the veins

Magnesium or iron — veins stay green

Edges first

Drought, scorch, or salt

Tips first

Dry air, or hard water

These four are the options the diagnosis form offers, and the difference between the first two decides the answer more often than anything else you can report.
Why it matters which leaves are affectedTwo plants side by side. On the left, the lower and older leaves are yellow while the growing tip stays green, which indicates a shortage of a mobile nutrient such as nitrogen, magnesium or potassium: the plant withdraws these from old leaves to feed new growth, so shortage appears at the bottom first. On the right, the newest growth at the top is yellow while the older leaves below stay green, which indicates a shortage of an immobile nutrient such as iron or calcium, which the plant cannot move once placed, so shortage appears at the top first. A label notes that this asymmetry is why the question is asked.Oldest leaves firstoldest at the bottom

A MOBILE nutrient — nitrogen, magnesium, potassium. Moved up out of old leaves to feed new growth, so the bottom suffers first.

Newest growth firstoldest at the bottom

An IMMOBILE nutrient — iron, calcium. Cannot be moved once placed, so only the newest growth goes short.

Nitrogen and magnesium move within the plant; iron and calcium do not. That asymmetry is why 'which leaves' separates causes that otherwise look identical.
Whole plant, a few leaves, or one sideThree plant outlines showing how damage is distributed. The first has every leaf affected, which points to a root, water or temperature cause acting on the whole plant. The second has only one or two of the lowest leaves affected while the rest is healthy, which is usually normal leaf ageing rather than a problem. The third has damage confined to one side of the plant, which points to a directional cause such as light, a draught or sun scorch.Whole plant

Roots, water, or temperature

One or two leaves

Often normal ageing

One side onlywindow

Light, draught, or scorch

A plant losing its two oldest leaves is behaving normally. A plant losing them everywhere is not — and the difference is what lets the engine rule 'nothing is wrong' in or out.

How it works

Yellow leaves has at least eight common causes. What separates them is not the symptom — it is the other observations: which leaves, in what pattern, how wet the compost is, what changed recently. Most tools map the symptom straight to a cause and are wrong often enough to be worse than useless.

This one ranks the candidates and shows the reasoning both ways, including the causes your description rules out. Where two causes fit equally well it says so rather than picking one, and leads with the check that separates them.

Where too little has been described it refuses and asks for more, because ranking on general frequency alone is exactly the mistake it exists to avoid.

How the diagnosis engine reaches an answerA flow diagram. A symptom and a set of observations feed into two kinds of test: supporting predicates, which raise a candidate cause, and contradicting predicates, which lower it or eliminate it entirely. These produce a ranked list with the reasons for and against each candidate. Below, four refusal outcomes are shown as equally valid results: too few observations, in which case the engine asks for more; no rules for that plant and symptom, in which case it says so; every candidate eliminated, in which case it says so; and two candidates that fit equally well, in which case it shows the check that separates them rather than picking one.Symptomwhat you can seeObservationspattern, moisture, changesSupportingraises a candidateContradictinglowers, or eliminatesRanked listwith reasons both waysOr it declines, which is also an answer:Too few observations → asks for moreNo rules for this plant → says soEverything eliminated → says soTwo equally good → shows the check
The four outcomes along the bottom are results, not errors. A tool that always produces a diagnosis is producing some diagnoses it should not.