Trimming aquarium plants

Every trimming question has the same answer

People ask trimming questions plant by plant. How do I trim Rotala? What about my Cryptocoryne? Should I cut my Anubias leaves? Asked that way there are seventy-five answers and no way to work out the seventy-sixth.

There is only one question, and it is this: where does this plant grow from?

Every plant has a meristem — a region of dividing cells that produces new tissue. Everything else on the plant is finished growing. Cut through a meristem and you have removed the plant’s ability to make that structure again; cut somewhere else and you have removed tissue the plant can replace. That is the whole of trimming.

Learn the five or six patterns and you can trim a plant you have never seen before, correctly, on the first attempt.

What a leaf node actually is

A node is the point on a stem where a leaf attaches. The bare length between two nodes is an internode. So far this is just vocabulary — and this is where most articles stop.

What matters is invisible. Sitting in the angle between the leaf and the stem, at every single node, is an axillary bud: a dormant meristem, fully formed, waiting. A stem plant is not one growing point. It is one active growing point at the tip and a dormant one at every node below, all the way down.

That is why the cut goes above a node. You are not cutting at a node because the tissue is tougher there or because it heals better. You are cutting above one because that leaves the bud attached to the piece that stays in the substrate. Cut halfway along an internode and the remaining stub has no bud in it; the plant must push new growth from the node below anyway, and the stub above it dies back.

Why cutting the top makes a plant bushy

Those buds stay dormant because the growing tip suppresses them. The tip produces the hormone auxin, which moves down the stem and holds the buds below in check — the phenomenon called apical dominance. It is why an untrimmed stem plant grows as a single leggy shoot with bare lower stem, racing for the surface instead of filling out.

Remove the tip and the source of that suppression goes with it. Buds below are released and break into side shoots, and one stem becomes two, three or more. The modern picture is more layered than a single hormone — auxin, cytokinin and strigolactone act as an integrated network rather than auxin alone¹ — but the practical consequence is unchanged and completely reliable: cut the top off and the plant branches below the cut.

This is the mechanism behind every “top it to thicken it up” instruction in the hobby. Cut low and you get branching low, which is what fills the bottom of a stem group. Cut high and you get branching high, which is why repeatedly trimming only the tips gives you a dense canopy above a bare, shaded, increasingly rotten base.

One honest caveat on the evidence: apical dominance is exhaustively demonstrated in terrestrial model plants such as Arabidopsis, pea and tomato. Submersed aquarium species are far less studied. The response is so consistent in practice that applying it is a sound inference rather than a leap — but it is an inference, not a result measured in Rotala.

Which half do you replant?

When you top a stem you now hold two pieces, and they are not interchangeable.

• The cut-off top is the cutting. It has the healthy apical meristem and the best leaves. Strip the lowest leaf or two, push it into the substrate, and it will root from the exposed nodes — the same nodes that carry buds also produce adventitious roots.
• The rooted base is what bushes out. It looks like the worse half — short, cut, sometimes leafless — but it has an established root system and freshly released buds, and it will usually out-grow the cutting.

The common mistake is binning the base because it looks finished. The second most common is replanting tops into a group whose bases have been removed, which throws away the established roots and resets the whole group.

Stem plants — the only group the classic rule fits

Rotala, Ludwigia, Bacopa, Hygrophila, Limnophila, Pogostemon, Alternanthera, Cabomba, Myriophyllum

Growing points at the tip and at every node. Cut a few millimetres above a node, replant the top, leave the base to branch. Trim before the plant reaches the surface rather than after — once the canopy shades its own lower stem, those lower leaves are shed. An exposed node may still wake up once light reaches it, but the bare internode between nodes never produces anything, so a badly shaded stem comes back patchy at best.

For a dense group, cut at staggered heights rather than levelling the whole thing, and cut lower than feels comfortable. Most people trim too high, too late, and then wonder why the bottom is bare.

Rosette and crown plants — never top these

Cryptocoryne, Echinodorus (swords), Vallisneria, Blyxa, Pogostemon helferi

Blyxa is the awkward one in this group. It forms a rosette and is trimmed like one, but it is stem-like enough that it readily throws side shoots and is usually propagated by separating those rather than by leaf removal alone.

These have one growing point, a crown at the base of the plant, and no axillary buds up a stem — because there is no stem. Everything the classic rule assumes is absent.

So there is nothing to release and nothing to gain by cutting the top: shortening a leaf does not make the plant branch, it just leaves you a cut leaf. Remove whole leaves at the base instead, following the leaf stalk down to the crown and cutting it there. Take the outer, oldest leaves — on a rosette those are the ones on the outside, and new growth emerges from the centre.

Several rosette plants also spread by runners — Vallisneria and dwarf sagittaria most obviously. That is propagation, not trimming: you thin the daughter plants to control spread, and manage each individual rosette by removing leaves at its own crown.

Runner and carpet plants — trim to force light down

Monte Carlo, Hemianthus callitrichoides, Glossostigma, Marsilea, Eleocharis, dwarf sagittaria, Lilaeopsis

Carpets spread by horizontal stems — stolons or runners — with a growing tip at the end of each. Left alone they do two things you do not want: they stack upward in layers, and the lowest layer, cut off from light, dies and rots until the whole mat lifts off the substrate in a sheet.

Trimming a carpet is not about neatness. It is about getting light back to the bottom layer before it dies. Cut the mat flat with scissors, take the clippings out with a net before they replant themselves across the tank, and do it while the carpet is still thin rather than after it has thickened.

The claim that trimming a carpet “encourages horizontal rather than vertical growth” is partly grounded. Plants detect crowding through the ratio of red to far-red light — leaves absorb red while transmitting and reflecting far-red, so a low red:far-red ratio signals shade and triggers elongation, the classic shade-avoidance response.² Thinning a mat raises that ratio and reduces the stretch signal. How much of the effect in a specific aquarium carpet is that, versus simply having removed the tall growth with scissors, is not something the hobby has measured.

Rhizome plants and epiphytes — leaves only, never the rhizome

Anubias, Bucephalandra, Java fern, Bolbitis

The thick horizontal stem these plants creep along is the rhizome, and it carries the plant’s growing points: an active one at its tip and dormant ones along its length. New leaves and new roots both emerge from it. It is a stem, not a root, and it must not be buried — smothered in substrate it rots, and rot spreads along the rhizome through exactly the buds the plant would otherwise grow from.

Trimming is therefore leaf by leaf: follow the stalk down to where it meets the rhizome and cut it there, leaving no stub to decay. Take old, holed, algae-covered leaves and leave the rest. There is no topping, no shaping, no cutting to height.

To make more plants you divide the rhizome rather than trim it: cut cleanly between leaf groups so that each piece keeps several leaves and its own roots. A bare rhizome fragment can shoot from a dormant bud, but it is far slower and far less certain than a piece that keeps working leaves.

Java fern adds a second route the others do not have: it grows small plantlets directly on its older leaf margins and tips. An old, tatty leaf covered in them is a nursery rather than a problem, and is worth leaving until the plantlets are large enough to detach.

Because these are slow growers with long-lived leaves, they accumulate algae — an Anubias leaf can be two years old, and that is two years of exposure. A leaf carrying heavy black beard algae is usually better removed than treated, since the plant will replace it and you have taken the algae out of the tank with it.

Mosses — the same problem as carpets

Java moss, Christmas moss, weeping moss, Riccia

Moss grows from many tips at once and simply keeps thickening. The failure is identical to a carpet’s: the lower layer is starved of light and flow, browns, dies, and eventually the whole pad detaches from its wood or stone and floats off in one piece.

Trim it hard and often, back to a thin layer, with scissors rather than by pulling. Do it before it browns underneath, not after — once the base has died there is nothing holding the green top on. Clippings will grow, so net them out unless you want moss everywhere.

Riccia is grouped here for practical purposes but is a liverwort, not a moss, and it is not attached in the same way — it is buoyant and tied down rather than rooted, which is why it escapes so readily when trimmed loose.

Floating plants — thinned, not trimmed

Amazon frogbit, water lettuce, Salvinia, red root floaters, duckweed

You never cut a floater. You remove whole plants by the handful, and the reason is light: an unchecked floating mat will take most of the light before it reaches anything below, and everything rooted underneath thins out.

Thin to keep a defined open area rather than a percentage of cover, and do it weekly — floaters double fast, so it is a small routine job or a large emergency one. Trimming the trailing roots is cosmetic and does the plant no favours; those roots are doing useful nutrient uptake.

Bulb and tuber plants — cut at the base and let them rest

Aponogeton, Madagascar lace, tiger lotus, banana plant

Leaves rise individually from a storage organ. Like a rosette, you remove whole leaves at the base and never cut them to length. The specific case is the lotus, where cutting surface lily pads at the base is how you keep the plant producing submersed leaves instead — take the pad before it opens flat and shades the tank.

Several Aponogeton species also go through a genuine dormancy in which they drop everything and appear dead. That is a rest period, not a failure, and the bulb should be left alone rather than dug up.

Marimo “moss balls” appear in plant lists but are a colonial green alga, not a plant at all. They are not trimmed — they are squeezed out in old tank water and rolled back into shape, and are divided by tearing rather than cutting.

When should a leaf come off?

This is the question nobody answers, and it has a real answer: it depends on whether the plant can get anything back out of the leaf.

A leaf is not only a solar panel, it is a store. As a leaf senesces the plant dismantles it and moves the contents to new growth. That process is efficient for some nutrients and barely happens for others. Measurements in Arabidopsis found nitrogen and potassium remobilised by up to about 80% from a senescing leaf and phosphorus by more than 40%, while calcium, manganese and several others moved by under 20%.³

Those are figures from one terrestrial species, so treat the exact percentages as illustrative rather than as numbers that transfer to a submersed plant. The pattern — some nutrients are worth reclaiming and some are not — is general, and it is the pattern the rule rests on:

• Old leaf yellowing from the bottom of the plant? That is a mobile-nutrient story — nitrogen, phosphorus, potassium, magnesium — and the plant is deliberately emptying that leaf to build new growth. Cut it off early and you throw the contents away. Let it finish. Take it when it is spent, translucent or starting to break down and foul the water.
• New leaf pale, distorted or holed at the growing tip? That is an immobile-nutrient story — iron, calcium, boron — and it will not recover, because the plant cannot move those nutrients into the damaged tissue. There is nothing to wait for. Remove it whenever it looks bad, and go and fix the cause.

Which end of the plant the damage appears at tells you which of those you are looking at — that is exactly what our guide to nutrient mobility is for, and the same signal is the backbone of diagnosing deficiencies.

Three cases override all of that and mean take the leaf now, whatever its age: it is badly algae-covered, since the leaf is now a reservoir; it is actively decaying and adding ammonia; or it is melting after planting, which is a normal transition rather than a health problem and is covered in our note on plant melt.

Every plant we cover, by growth form

Find your plant, read its form, then read that form’s section above. This table is generated from our own plant profiles, so it stays current as more are added.

Stem · 33

Top it above a node; replant the top, let the base branch

Alternanthera reineckii · Ammannia gracilis · Anacharis · Asian ambulia · Bacopa · Bacopa monnieri · Brazilian pennywort · Cabomba caroliniana · Creeping Jenny · Dwarf hygro · Guppy grass · Hornwort · Hottonia palustris · Hydrocotyle 'Japan' · Hygrophila corymbosa · Hygrophila pinnatifida · Limnophila aromatica · Lobelia cardinalis · Ludwigia arcuata · Ludwigia repens · Ludwigia super red · Micranthemum umbrosum · Myriophyllum mattogrossense · Pearlweed · Pogostemon erectus · Pogostemon stellatus · Rotala 'H'ra' · Rotala macrandra · Rotala rotundifolia · Rotala wallichii · Star grass · Staurogyne repens · Water wisteria

Rosette / crown · 10

Never top. Remove whole outer leaves at the crown

Amazon Sword · Blyxa japonica · Cryptocoryne 'Flamingo' · Cryptocoryne balansae · Cryptocoryne parva · Cryptocoryne undulata · Cryptocoryne wendtii · Downoi · Vallisneria spiralis · Water sprite

Runner / carpet · 8

Trim the mat flat to get light to the bottom layer

Dwarf Hairgrass · Dwarf chain sword · Dwarf four-leaf clover · Dwarf sagittaria · Glossostigma · HC Cuba · Micro sword · Monte Carlo

Rhizome / epiphyte · 10

Leaves at the stalk only. Never cut or bury the rhizome

African water fern · Anubias 'Mini Coin' · Anubias 'Petite' · Anubias Nana · Anubias barteri · Bucephalandra · Java Fern · Java fern 'Narrow' · Java fern 'Trident' · Java fern 'Windelov'

Moss · 4

Trim hard and often before the base browns

Christmas moss · Crystalwort (Riccia) · Java Moss · Weeping moss

Floating · 5

Thin whole plants by hand; never cut

Amazon frogbit · Duckweed · Floating fern (Salvinia) · Red root floaters · Water lettuce

Bulb / tuber · 4

Whole leaves at the base; expect dormancy

Aponogeton crispus · Banana plant · Madagascar lace · Tiger lotus

The short version

Before you cut anything, find the growing point. If it is at the tip and repeated at every node, you have a stem plant and you can cut almost anywhere — above a node, and the plant will branch below the cut. If there is only one growing point, whether that is a crown at the base, a rhizome along a piece of wood or a bulb in the substrate, then you are only ever removing whole leaves, and the plant decides its own shape.

Everything else — carpets, mosses, floaters — is a light problem wearing a trimming costume. You are cutting to get light to the layer underneath before it dies, and the timing matters more than the technique.

References

1. Domagalska, M. A. & Leyser, O. (2011). Signal integration in the control of shoot branching. Nature Reviews Molecular Cell Biology 12, 211–221. doi:10.1038/nrm3088
2. Franklin, K. A. (2008). Shade avoidance. New Phytologist 179(4), 930–944. doi:10.1111/j.1469-8137.2008.02507.x
3. Himelblau, E. & Amasino, R. M. (2001). Nutrients mobilized from leaves of Arabidopsis thaliana during leaf senescence. Journal of Plant Physiology 158(10), 1317–1323. doi:10.1078/0176-1617-00608