Skip to content

The forest floor keeps its own weather

9 minJames

Revised 11 August 2026

Rain can be running in the wheel tracks and still leave the ground under an old spruce stand feeling held together rather than wet. A few metres away, in a stand of much the same age, the moss gives under the heel and dark water appears at the base of the needles. The rain was shared. The response was not.

It is tempting to call the difference “the soil”, as if the mineral ground begins at the boot sole. In barrskog it often does not. Between the weather and mineral soil sits a working layer: fresh needles, broken needles, twigs, fungal threads, humus, and often a moss carpet. It catches water, gives some of it back to the air, passes some downward, and changes the temperature and humidity at the surface. Its thickness matters, but thickness alone is a poor shorthand for what it does.

Rain is divided before it reaches the humus

The rain measured at an open gauge is not the rain received by the forest floor. Before a drop reaches needles on the ground it meets crown cover. SMHI describes interception as rain held on vegetation and evaporated without reaching the ground; in dense conifer forest, roughly a third, and in some circumstances as much as half, can be caught by trees. Short, repeated showers favour this loss more than the same total delivered in fewer long spells. That is the first division of a common rain event.

What does pass the canopy is throughfall, and it is not spread evenly. Needles, branches, gaps in crowns and trunks redistribute it. Work on conifer stands has found spatial variation not just in the amount of throughfall but in its intensity. The practical consequence is ordinary enough: a patch beneath a crown is not supplied in the same way as a gap between crowns, though both lie within the same stand.

Then comes the floor. A dry needle layer does not behave as bare mineral soil. It has empty spaces and dry surfaces that can take up water at first. The moss layer has its own structure and its own capacity to hold water. A light shower can therefore wet the forest floor conspicuously while adding little to the soil beneath. A longer rain, or a floor already close to wet, changes the sequence: storage fills and water can move down through the organic horizon and into mineral soil.

That sequence is why “it rained there” is not yet a description of moisture at the depth that matters to roots, mycelium or the lower humus. It says something about an input. It does not say how much passed the canopy, how much sat in the floor, how quickly it evaporated, or where it went laterally along a compacted or sloping layer.

The distinction is not pedantry. In a northern Swedish boreal landscape, Zignol and colleagues monitored surface soil moisture at 78 locations and found that topography, soil properties, vegetation and land cover all contributed to spatial variation. The relative importance also changed with conditions: topography explained more during wet periods, while soil and vegetation mattered more in dry periods. Weather had a lagged, cumulative effect rather than a clean same-day relationship. A rain total is therefore the beginning of an account, not its conclusion.

A thick floor is both store and valve

The useful image is not a sponge alone. A forest floor is a store with exits.

Needles and moss can retain water. That retention can soften a brief change in air humidity or a day of sun and wind at the soil surface. It can reduce direct exposure of the humus to radiation, and moss and organic matter also alter heat flow. Turetsky and co-authors’ synthesis of Alaskan boreal forest research treats mosses as regulators of soil climate and biogeochemical cycling; their review also notes that shifts away from Sphagnum can reduce moisture retention. That is a strong reason to regard the green and brown layer as part of the hydrology, not decoration on it.

But stored water is not necessarily water delivered to mineral soil. Forest-floor interception research makes the awkward point clearly. Litter can hold substantial water, affecting both evaporation and soil-moisture dynamics. Some of the water held near the surface returns to the atmosphere before it percolates. The floor can consequently delay recharge below it, especially after a dry spell and a small event. A thick dry layer may first make the soil beneath less responsive to rain, not more.

Later, the same layer can have the opposite practical appearance. Once wet, a deep organic horizon and moss cover can maintain a humid, shaded interface above mineral soil while bare or sparse ground follows hot, dry air more directly. This is buffering in the strict sense: it reduces and delays a fluctuation. It is not a permanent wetness guarantee.

The balance depends on event shape. Long gentle rain, hard rain, drizzle, wind after rain, solar exposure, and the starting water content of the layer all alter the result. SMHI’s account of evaporation is useful here because it names the drivers usually hidden by the word “weather”: energy from sunlight, and removal of water vapour by wind and turbulence. The same rain received by two forest floors can therefore be followed by quite different losses to the air.

That is also why the familiar gesture of lifting a moss edge after rain can mislead. A wet upper moss tuft establishes that the surface store is charged. It does not establish the water state underneath, and a dry-looking surface after wind does not establish a dry humus below. There are layers, and they can be out of step.

Two floors with the same depth are not the same floor

A ruler pushed through moss and needles produces a number. It does not produce a material description.

A floor made mostly of loose, recently fallen spruce needles has a different arrangement of voids from a compressed, partly decomposed humus layer. A moss carpet has yet another structure, and species composition changes its form. Deadwood, roots, stones and old machine disturbance break the layer into short distances with different paths for water. Even the apparently simple question of depth needs a location: beside a trunk, under the crown edge, in a moss hollow, or on the exposed shoulder of a rise.

Time is inside that material. Fresh litter is a transient cover; decomposition transforms it toward humus. Moss may grow upward while older material below is compacted. A stand of a stated age can therefore contain a floor shaped by former tree cover, thinning, windthrow, fire, drainage, browsing and the slower history of decomposition. Stand age is an indirect clue, not a measurement of the water buffer.

The canopy adds another mismatch. The study by Šrámek, Neudertová Hellebrandová and Fadrhonsová compared Norway spruce stands of different ages across contrasting seasons. Interception increased with stand age and dimensions, but differences in soil-water potential between stands did not simply track interception. The authors point to stand transpiration as an important driver of the water budget. That is a useful refusal of a neat story: more crown cover may alter water arriving from above, yet roots and foliage also alter water leaving the soil.

Nor is a same-aged pair necessarily alike in canopy density, basal area, tree species mix, ground vegetation or root distribution. One may be an old edge opened to wind and afternoon sun; the other may be enclosed, with a continuous canopy and deep moss. The first can lose surface water quickly even if its organic layer is deep. The second can receive less throughfall under a dense crown but retain a different surface climate. Neither observation settles the other.

Foragers often name this as “soft ground” or “a good moss bottom”. Those phrases contain an accumulated observation, not a universal mechanism. They may be noticing thickness. They may instead be noticing shade, a seep from upslope, a fine-textured mineral soil, an old ditch, or a change in canopy. The sensible response is to preserve the observation and keep the causes separate.

The weather above and moisture below keep different clocks

One reason two stands seem to contradict each other is that rain operates on several clocks. Crown interception responds during the event. The upper litter responds early. Movement to lower layers depends on prior wetness, soil structure and pathways. Evaporation begins again whenever energy and dry moving air are available. Plant water use continues on its own daily rhythm.

This makes the previous days more informative than a single cloudburst, but not sufficient. Zignol and colleagues found that soil-moisture variation in their Swedish study was better explained by hydrological and meteorological variables averaged across the preceding five to seven days than by immediate conditions. That result is useful for the idea of antecedent weather. It is not a licence to turn five or seven days into a Swedish rule. Their measurements were surface soil moisture in one northern landscape during one summer, not every humus profile in every län.

The same caution applies to regional evidence about fungal fruiting. Koelemeijer and colleagues excluded rainfall in a boreal forest landscape and reported reduced sporocarp production for both saprotrophic and ectomycorrhizal fungi. It is a valuable experimental result: withholding rain changed fungal output. It does not identify a particular litter depth, establish an amount of rain that switches fruiting on, or convert a wet forest floor into a promise about what will be present above it.

Older long records make the same larger point without resolving the hillside. Straatsma, Ayer and Egli found associations between precipitation and fungal productivity, and between July temperatures and fruit-body appearance, in a Swiss forest plot surveyed over many years. Kauserud and co-authors found changes in fruiting phenology in Norwegian herbarium records. Both works concern weather and timing at scales far broader than two moss patches. Neither measures the needle-and-moss buffer beneath a Swedish spruce stand. The gap is not a flaw in those studies; it is the gap that must remain visible.

A moisture reading, a rain gauge and a hand in the moss are thus different sorts of evidence. The first is a measurement at a depth and moment. The second is a weather input. The third is local memory and inspection. They can agree, but one cannot replace the others.

Thickness is evidence but not a verdict

There is settled ground here. Forest floors intercept and store water. Canopies alter what reaches them. Evaporation, transpiration, topography, soils and vegetation alter what remains. In Swedish boreal field data, those controls do not hold their ranks in every weather state. In a dry period, the properties of the floor and vegetation can become more consequential; in wet conditions, the route water takes through the terrain can dominate.

There is also a practical limit to maps. SLU’s Markfuktighetskarta is trained with Riksskogstaxeringen field data and combines topography, climate and soil information to represent persistent moisture conditions and their natural transitions. SLU is explicit that actual wetness at a particular time depends on the weather. The map does not measure the present thickness, composition or water content of the litter layer at a boot-sized patch. Riksskogstaxeringen is an annual sample of the country’s forest area, not a diary of one stand’s floor.

Unsettled: how much litter depth alone explains moisture beneath Nordic forest floors at the scale on which a person notices a patch. I could not find a published Swedish field study that repeatedly measures, in paired same-aged stands, litter and moss architecture, throughfall, lower organic-horizon moisture, mineral-soil moisture, evaporation and fungal fruiting together. There are excellent studies of pieces of that chain. The chain itself remains thinly measured.

That uncertainty is a reason to describe the ground carefully. A thick layer after rain may be a buffer, a temporary store, an evaporating surface, or all three in sequence. A neighbouring stand may differ because its layer is thinner, but just as plausibly because its crown, roots, slope or soil directs water differently. The rain was the shared event. The forest floor was part of the reason the two places did not live through it in the same way.

Sources