Case Study: A Kyoto-Inspired Moss Tray, Two Years On

The brief and the constraints

The vessel is an unglazed stoneware tray, thirty-two by twenty-two centimetres, four centimetres deep, with no drainage holes. It sits on a north-facing sill in an unheated hallway that runs between eleven and nineteen degrees through the year. Light measures between 600 and 1,400 lux depending on season, with no direct sun at any point.

The brief was explicitly minimal: no lid, no wood, one stone, three species, and a surface that reads as a single landscape from above. Open builds are harder than closed ones because the humidity buffer disappears, but they also avoid the two most common closed-build failure modes — saturation and stagnant air.

The shallow depth was the main engineering constraint. Four centimetres does not allow a conventional drainage layer plus growing medium plus slope. The solution was to treat the whole tray as growing medium and manage water by weight instead of by reservoir.

Substrate and the weight method

The medium is a firm mix: five parts sieved coir, three parts akadama at two to three millimetres, two parts fine pumice. Akadama is the unusual ingredient and it does most of the work — it holds water within the granule while draining freely between granules, and it holds a shaped surface without slumping, which matters enormously in a tray with no walls to speak of.

The surface is not flat. It rises from six millimetres at one long edge to thirty-five at the diagonal corner, a gentle asymmetric swell rather than a slope. From above this reads as a low hill; from the side, which is how the piece is mostly seen, it gives the moss carpet a curve that catches light differently across its width.

Water is managed by weighing. The tray at correct moisture is 1,850 grams; dry it falls below 1,600. The builder weighs it on kitchen scales every Sunday and adds the difference as a slow drizzle of filtered water. This sounds fussy and takes forty seconds, and it has been more reliable than any moisture meter tried previously.

What was planted, what failed, what replaced it

The original planting was Leucobryum glaucum as three cushions, Hypnum cupressiforme as the connecting carpet, and Polytrichum commune as a taller vertical accent in the high corner. Two of those three worked.

Polytrichum failed within four months. It is a demanding species that wants high light, an acidic sandy medium and genuine air movement, and while the hallway provided the third it did not provide the first two. The shoots stretched, thinned and browned progressively from the base. It was replaced with Dicranum scoparium, which provides similar upright texture with far lower light requirements and has been stable since.

One Leucobryum cushion also failed, and the cause was diagnostic rather than species-related: it had been seated in the lowest corner of the tray where water collected. The remaining two cushions, both on the raised half, have expanded by roughly two centimetres in each direction. The lesson generalised into a rule the builder now applies to every tray — cushion mosses go high, sheet mosses go low, always.

How the surface changed over twenty-four months

Month one to three: visible gaps between fragments, distinct species boundaries, an obviously assembled look. Watering was twice weekly by mist, and the surface dried noticeably within a day.

Month four to twelve: the Hypnum carpet closed, the gaps disappeared, and the surface began holding moisture much longer as the mat itself became a water store. Weekly watering became sufficient. The stone, initially stark, acquired a thin algal film along its damp base and stopped looking like an object placed on soil.

Month thirteen to twenty-four: consolidation. The Hypnum began climbing the lower flank of the stone. The Dicranum thickened into a proper tuft with the characteristic swept appearance. Most significantly, the boundaries between species softened as the carpet grew over cushion edges, and the piece finally achieved the continuous-surface quality it was designed for. The builder now trims the Hypnum back from the cushion margins twice a year with curved scissors, which takes about ten minutes and is the only maintenance beyond watering.

What transfers to other builds

First, open trays are viable and underrated. They avoid the humidity management problem entirely at the cost of a weekly watering habit, and for temperate species in a cool room they arguably produce better colour and denser growth than a sealed vessel would.

Second, akadama or a similar structural granule solves the shallow-tray problem. Ordinary potting medium in four centimetres of depth becomes a slumping mud within months.

Third — and this is the transferable design lesson — the piece works because it is under-designed. One stone, three species, no props, no path, no forced perspective. At tray scale, viewed from above and close, the eye reads texture rather than landscape, and adding compositional devices intended for a vertical viewing angle simply clutters the surface. Matching the composition to the actual viewing geometry is a discipline that applies to every vessel, and it is where most ambitious builds go wrong.

Extended notes

• Akadama breaks down over roughly three to five years, softening into clay. In a permanent tray planting this is acceptable because the moss mat holds structure by then, but it is worth knowing when planning longevity.
• Unglazed stoneware trays sweat. This one sits on three cork pads, which keeps the sill dry and provides airflow underneath.
• The weekly weighing habit doubles as a health check. A tray that loses weight unusually fast is drying out somewhere it should not be, and that has caught a cracked base and a badly positioned radiator on separate occasions.

Frequently asked