Brittleness Score

The Brittleness Score (1–10) is a context variable that calibrates what is possible, and how fast, for any given piece of land. It is not something management can change. It is a characteristic of the place, inherited from the climate.

The concept, in one paragraph

Brittleness describes how an environment responds to rest. In non-brittle environments (score 1–3), resting land recovers naturally through biological decomposition: fungi, soil fauna, mycorrhizal networks do the heavy lifting and the cover knits back together on its own. In brittle environments (score 7–10), rest alone is not enough: dead plant material oxidises and degrades on the stem rather than rotting back into the soil, the land needs periodic disturbance (such as grazing) to maintain health. Semi-brittle environments (4–6) sit in between.

Understanding brittleness is essential for choosing the right management tools, because the same low Land Health Score demands very different responses depending on where a site sits on this continuum.

The six brittleness classes, in full

The table below sets out the six classes EcoIntel uses. Each class has a characteristic rainfall pattern, a characteristic rate of decomposition in the field, and a characteristic set of management risks. The continuum runs from Atlantic Ireland to North African desert.

Class	Definition	Field validation	Management implication
1 Very Non-Brittle	Even rainfall, year-round growth, humid decomposition.	Dung & litter disappear in weeks; minimal bare soil.	Main risk = overgrazing / compaction. Focus on rest, diversity, avoiding disturbance.
2–3 Non-Brittle	Rainfall well spread, short dry season.	Dung / litter break down within 1–2 months.	Root-cause issues tend to be nutrient leaks and compaction, not water stress.
4–5 Semi-Brittle	Rainfall patchy, longer dry spells.	Dung / litter persist 2–6 months; some bare soil.	Both overgrazing and over-rest are possible; effective rainfall begins to matter.
6–7 Brittle	Long dry season; rainfall arrives in bursts.	Dung / litter persist >6 months; large patches of bare soil.	Water and mineral cycles fragile; animal impact required to restore cover and litter.
8–9 Very Brittle	Rainfall erratic; ecosystem inactive for long periods.	Dung / litter persist >1 year; moribund material accumulates.	Over-rest is the major problem; trampling and littering essential to kick-start processes.
10 Hyper-Brittle	Desert / near-desert; rainfall extremely irregular.	No decomposition; biological crusts; persistent bare ground.	Water design (keyline, catchments, ponds) essential; biology needs intensive re-boot.

Reference ecoregions on the brittleness curve

Brittleness is computed from the coefficient of variation of monthly rainfall, a measure of how unevenly precipitation is distributed across the year. This replaces the subjective estimation that has been standard practice since Savory introduced the concept.

The piecewise mapping from rainfall coefficient of variation (X-axis) to Brittleness Score (Y-axis), with reference dots from European and North African landscapes. The red dot is Wilder Wood Farm at CV 0.51, B 3.5.

The reference dots on the curve illustrate the full continuum:

• Ireland, Brittleness ~2 (Very Non-Brittle). Rainfall distributed across virtually every month, dung disappears in weeks, year-round biological decomposition. Rest produces recovery reliably. The principal management risk is over-stocking.
• UK Devon, Brittleness ~2.8 (Non-Brittle). Atlantic temperate, occasional short dry summers, rainfall well-spread. Rest works; the main issues are nutrient leaks and compaction rather than water stress. (Wilder Wood Farm sits just above this point at 3.5.)
• Germany interior, Brittleness ~4.2 (Semi-Brittle). Continental temperate, drier summers, intermediate behaviour. Both over-grazing and over-rest can degrade the system; effective rainfall begins to dominate.
• Hungary plains, Brittleness ~6.3 (Brittle). Pannonian basin, hot dry summers, long dormancy. Dung and litter persist months on the surface; rest alone produces moribund material. Animal impact in the right windows becomes essential.
• Spain Mediterranean, Brittleness ~7.7 (Very Brittle). Hot dry summers, rainfall in bursts, long inactive periods. Over-rest is the dominant failure mode; the land needs trampling and littering to restart biology.
• North Africa, Brittleness ~8.6 (Very Brittle / Hyper-Brittle edge). Desert margins, extremely irregular rainfall, persistent bare ground. Water design (keyline, micro-catchments, ponds) becomes essential alongside intensive biological re-boot.

The brittleness curve is steep at low CV values and flattens at high CV. Small differences in rainfall variability between Ireland and Devon produce only ~1 point of brittleness change; small differences between Hungary and Spain produce closer to 1.5 points; differences beyond Spain push toward the hyper-brittle edge quickly.

How EcoIntel computes it

EcoIntel computes brittleness from reanalysis climate data. Precipitation CV (coefficient of variation) measures month-to-month rainfall variability across all assessment years. A low CV means rainfall is evenly distributed; a high CV means it arrives in bursts separated by dry periods.

The precipitation CV is converted to a raw brittleness score using the piecewise mapping above, then adjusted for ecoregion: the zone baseline reflects the regional climate norm. The final score blends the precipitation-derived value (70%) with the zone baseline (30%).

The brittleness score and precipitation CV are related but distinct measures. The CV is a pure statistical measure of rainfall variability. The brittleness score incorporates the CV but also adjusts for regional context: a CV of 0.5 in Atlantic Britain (where the baseline expectation is reliable rainfall) has different ecological implications than the same CV in the Mediterranean (where seasonal drought is the norm). The score reflects the ecological meaning of the variability, not just the variability itself.

For general illustration, this chart shows how daily rainfall patterns differ across brittleness levels. Non-brittle climates (green) have consistent small rainfall events throughout the year. Semi-brittle climates (gold) show distinct wet and dry seasons. Brittle climates (orange) have long dry periods punctuated by intense downpours. The ecological consequences cascade from there: where rain is consistent, biology decomposes litter quickly; where it is bursty, litter accumulates and management has to do biology’s job.

The underlying rainfall data for Wilder Wood Farm. Each cell is one month coloured by precipitation volume; the right-hand column shows the annual coefficient of variation that feeds the brittleness calculation.

For each parcel, EcoIntel pulls the multi-year ERA5-Land precipitation record at that location, computes the coefficient of variation of monthly rainfall (pooled across all assessment years), maps the result onto the piecewise curve, and blends with the ecoregional baseline. The output is one numeric Brittleness Score per site, recorded against the assessment for traceability.

Brittleness is not an opinion. It is a number derived from observed climate. And because it is derived from a multi-year record, it changes only with genuine climate shift, not with last week’s weather.

Mapping the CV bands

The piecewise mapping in compact form, for the technically inclined:

Precipitation CV	Brittleness	Category	Characteristics
< 0.30	1.0–3.0	Non-brittle	Reliable moisture, biological decay dominant
0.30–0.55	3.0–5.0	Semi-brittle (low)	Mostly reliable moisture, occasional gaps
0.55–0.75	5.0–6.0	Semi-brittle	Seasonal moisture, mixed decay processes
0.75–1.00	7.0–8.0	Brittle	Unreliable moisture, physical weathering dominant
> 1.00	8.0–10.0	Very brittle	Arid / semi-arid, extreme unpredictability

Why brittleness reorders the management cascade

Two farms with the same Land Health Score can need very different first moves depending on their brittleness:

• A non-brittle site at Land Health Score 5 (Transitional) responds to rest plus species enrichment. Rest gives the existing biology time to do its work; diversity unlocks the next ceiling. Buying seed is the last resort.
• A brittle site at the same Land Health Score 5 responds to planned high-density grazing pulses plus water design. Rest alone allows litter to stand un-decomposed, capping the surface. Animal impact in narrow windows trampling that litter is the lever; building infiltration and water capture is the foundation everything else needs.

EcoIntel uses brittleness as a context variable behind every diagnosis. Recommendations adjust to it. Rest periods lengthen or shorten. Intervention timing shifts. The BROWN → BLUE → GREEN → BLACK cascade reorders. Calibrating expectations to brittleness prevents both over-optimism (in brittle systems) and unfair disappointment (in non-brittle ones).