Sauzet et al. (2024). The organic carbon‐to‐clay ratio as an indicator of soil structure vulnerability, a metric focused on the condition of soil structure

) investigated whether the soil organic car-

bon to clay ratio (SOC:clay) adequately characterizes the

structural condition of soils, as implied by the proposi-

tion of using a 1:13 value of the SOC:clay ratio as a met-

ric for healthy soils in the European Soil Monitoring Law

(European Commission,

). Poeplau and Don (

)

used German Agricultural Soil Inventory data from 2958

Received: 23 October 2023

|

Revised: 9 April 2024

|

Accepted: 19 April 2024

DOI: 10.1111/sum.13060

C O M M E N TA R Y

The organic carbon- to- clay ratio as an indicator of soil

structure vulnerability, a metric focused on the condition of

soil structure

Ophélie Sauzet

|

Alice Johannes

2

|

Cédric Deluz

|

Xavier Dupla

3

|

Adrien Matter

|

Philippe C. Baveye

4

|

Pascal Boivin

Creative Commons Attribution-NonCommercial

This is an open access article under the terms of the

License, which permits use, distribution and reproduction in any

medium, provided the original work is properly cited and is not used for commercial purposes.
© 2024 The Authors.

Soil Use and Management

published by John Wiley & Sons Ltd on behalf of British Society of Soil Science.

1

University of Applied Sciences and

Arts of Western Switzerland (HES- SO),

HEPIA inTNP, Soils and Substrates

Group, Geneva, Switzerland

2

Soil Quality and Soil Use Group,

Department of Natural Resources &

Agriculture, Swiss Federal Research

Station Agroscope, Zurich, Switzerland

3

Institute of Earth Surface Dynamics,

University of Lausanne, Lausanne,

Switzerland

4

Saint Loup Research Institute, Saint

Loup Lamairé, France

Correspondence

Ophélie Sauzet, University of

Applied Sciences and Arts of Western

Switzerland (HES- SO), HEPIA inTNP,

Soils and Substrates Group, Geneva,

Switzerland.

Email:

ophelie.sauzet@hesge.ch

Abstract

The soil organic carbon to clay ratio (SOC:clay) is a metric used in soil quality
management. In Switzerland and the United Kingdom, for example, threshold
values for SOC:clay ratios have been determined to indicate very good (

>

1:8) to

degraded (

<

1:13) soil structures. A recent article in Soil Use and Management by

Poeplau and Don, however, suggested that this metric is ‘strongly biased and mis-
leading’, based on their observation that German sandy soils and heavy clay soils
tend to show very high and very low SOC:clay ratios, respectively. An alternative
metric was proposed based on the ratio of actual SOC to expected SOC level for
a considered area. We offer a commentary on the proposal, arguing that because
soil structure quality is overlooked by the approach, it fails to provide appropriate
SOC levels for soil health and could lead to soils with highly depleted SOC being
classified ‘good’. The SOC:clay ratio, on the other hand, does address soil struc-
ture condition, providing a structure vulnerability index, a key function inde-
pendent of local soil management conditions. When soils are found to have high
structure vulnerability, as indicated by the SOC:clay ratio, the cropping practices
at the site should be investigated and ways to increase the SOC content consid-
ered. Structure condition threshold values may only need to be reassessed if it is
shown that the average structure quality observed is not in conformity with the
present thresholds, which would be expected for some soils, such as Andosols.

K E Y W O R D S

clay content, soil organic carbon, structure quality, structure vulnerability

2 of 7

|

SAUZET et al.

topsoil samples (0–30 cm) to evaluate the suitability of the

SOC:clay ratio, and concluded that it is not a suitable SOC

level metric since it is strongly biased, misleading and

partly insensitive to SOC changes.

Their statement was based on the observation that at

low and high clay content, soils tend to show large and

small SOC:clay, respectively, regardless of SOC content.

As a suggested unbiased alternative to the SOC:clay ratio,

Poeplau and Don (

) proposed a new metric that is

based on the ratio between actual and expected SOC

(SOC

exp

) levels, the latter being derived from a regression

between local data for SOC and clay content. In this con-

text, the purpose of this article was to address criticisms

of the SOC:clay ratio, and to provide a balanced perspec-

tive on metrics associated with the structural condition of

soils.

2

|

BACKGROUND

Some background information is useful to understand

the arguments presented by Poeplau and Don (

) later used the ratio to emphasize

).

Feller and Beare (

1997

) initially demonstrated that the

SOC:clay ratio influenced aggregate stability. Building on

this, Dexter et al. (

2008

the interplay between clay and carbon in controlling soil

structure stability, proposing a ‘complexed organic car-

bon’ theory. They concluded that an optimum SOC con-

tent was equal to 10% of the soil clay content. However,

Johannes, Weisskopf, et al. (

), reported that the rela-

tionship between structure quality and SOC content was

not linear. Working with 161 soil samples (5–10 cm depth)

obtained in Swiss cropland and whose structural quality

they evaluated visually based on the Visual Evaluation

of Soil Structure (VESS) method (Ball et al.,

), they

reported that the relationship followed a ‘broken stick’

pattern. A rapid drop in SOC content was observed in the

good to acceptable structure quality range (VESS scores

from 1 to 3), followed by a slower drop from acceptable

to degraded structure quality (VESS scores from 3 to 5).

Johannes, Matter, et al. (

) suggested that there is no

optimal SOC value corresponding to a fraction of the clay

content, contrary to Dexter et al. (

2008

), and observed that

the SOC:clay ratio decreased monotonically with struc-

tural quality.

On that basis, they determined empirically that thresh-

olds of the SOC:clay ratio of 1/8, 1/10 and 1/13 indicated

the boundaries, respectively, between ‘very good’, ‘good’,

‘moderate’ and ‘degraded’ levels of average structure qual-

ity across seasons and rotations. Prout et al. (

) sub-

sequently tested the SOC:clay index using data from the

initial sampling (1978–83) of the National Soil Inventory

of England and Wales, covering 3809 sites under arable

land, grassland and woodland, and found that the thresh-

old values identified by Johannes, Matter, et al. (

)

were applicable to this much larger data set. This led Prout

et al. (

) to suggest that, although by no means abso-

lute, these thresholds might also be applicable in soils in

similar climatic zones across Europe. More recent work

by Prout et al. (

) and Johannes et al. (

) confirmed

the practical usefulness of the thresholds of the SOC:clay

index, respectively, in England/Wales and in the Jura re-

gion of Switzerland, in soils with up to 87% and 52% clay

contents, respectively.

2.1

|

Testing the applicability of SOC/

clay thresholds is worthwhile

A first point concerning the recent article of Poeplau and

Don (

) is that any attempt to scrutinize the applica-

bility of the SOC:clay index and its various thresholds in

a different geographical context is entirely welcome. It is

customary to insist on the fact that conclusions are tied

to specific locations (Andrews et al.,

2002

; Bünemann

et al.,

2018

), and therefore, studies using local references

are required. Similarly, when considering the SOC:clay

ratio to characterize the structural condition of soils, the

range of applicability needs to be researched and the areas

identified where refinements are needed.

In sandy soils, as an example, the SOC:clay ratio con-

sistently tends to be very high and the structure is stable,

thereby placing the soil in the optimal class in Johannes,

Weisskopf, et al.'s (

) ranking. In the data set used by

Johannes, Matter, et al. (

), the minimum clay content

was 10%. In general, there is a need for a more sensitive

description of sandy soils, which remains to be developed.

This can be extended to all soils with low clay content, for

instance, peat soils and chalky soils for which the con-

cept of complexation of SOC on clay proposed by Dexter

et al. (

2008

), as well as the concept of structure quality,

may not apply. Indeed, these soils do not typically ex-

hibit classical structural behaviour observed in soils with

higher clay content. By studying the structural behaviour

of undisturbed soil samples on a large range of clay con-

tent, Boivin (

1990

) found that sandy soils with less than

12% clay did not demonstrate structural shrinkage and

behaved like particle beds. It is, therefore, likely that the

structure condition of such soils cannot be linked to the

SOC:clay ratio, though one could conclude that the vul-

nerability of their structure is low, which would be correct

but of little interest.

Likewise, clayey soils pose a particular challenge. As

pointed out by Johannes et al. (

), some clay soils have

SOC values above the ‘good’ or even ‘very good’ thresh-

olds of the SOC:clay ratio, though the SOC content has

14752743, 2024, 2, Downloaded from https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/sum.13060 by Hayley Peter-Contesse , Wiley Online Library on [18/10/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

|

3 of 7

SAUZET et al.

to increase proportionately much more than in other

soil types (Johannes, Matter, et al.,

). Johannes

et al. (

) showed that under similar cropping practices,

the structure quality and the SOC:clay ratio were indepen-

dent of clay content over a wide range (16%–52%) of clay

content, supporting the idea that the higher the clay con-

tent the easier it is to increase SOC. Again, some adjust-

ment to the approach might be in order.

Both Johannes, Matter, et al. (

); Johannes

et al. (

) and Prout et al. (

open for such developments, as neither claimed that

their conclusions were universal or absolute. On the

contrary, both groups of authors pointed out that fur-

ther research was needed to determine to what extent

the SOC:clay index and its various thresholds could be

applied in other areas than those in which the original

investigations took place. For instance, Andosols are

well known to show very large SOC content, and to be

poorly dispersible (see Feller et al.,

2001

), thus leading

to much higher SOC:clay ratios than those reported by

Johannes, Matter, et al. (

), Prout et al. (

) and

Poeplau and Don (

).

Poeplau and Don (

) also mentioned Vertisols and

Chernozems. Vertisols, characterized as heavy clay soils

showing uniform profiles with low SOC content to large

depths because of their vertic properties, would likely

show very low SOC:clay ratios in the topsoil. Generally, it

can be expected that soils with large SOC protection, such

as Andosols, and soils with high active carbonate fraction

(Rabot et al.,

2024

) or Chernozems, will not present the

same SOC:clay thresholds for their structure condition as

the soils assessed in Johannes, Matter, et al. (

Prout et al. (

2.2

|

Testing under identical conditions

For a broader assessment of the SOC:clay index to be

meaningful, one could reasonably argue that it should be

carried out under the same conditions as those used in the

original articles that first described the approach. Whereas

Johannes, Weisskopf, et al.'s (

) and Johannes et al.'s

(

) soil samples were obtained from 5 to 10 cm, with

structure quality assessed on the same samples, and Prout

et al. (

,

) used the top 15 cm of the soil profile,

Poeplau and Don's (

) data set consists of SOC and clay

values from 0 to 30 cm layer of soils. Part of the rationale

behind the Johannes, Weisskopf, et al. (

); Johannes

et al. (

) and Prout et al. (

) choice of soil

samples is that this depth involves the portion of agricul-

tural fields that tends to be homogenized by tillage, whose

structure is affected by traffic and tillage activity, and is

typically enriched in organic matter. The significantly

deeper samples used in the data set by Poeplau and

Don (

) would mean that they included soil material

that would tend to contain less organic matter and, pos-

sibly, more clay if clay illuviation took place in situ to any

significant extent. Therefore, the data set used by these au-

thors may not have been adequate to assess the robustness

of the SOC:clay index approach, as originally devised by

Johannes, Weisskopf, et al. (

Poeplau and Don (

) did not use soil structure qual-

ity but bulk density (BD) values in the second part of their

studies, with no information on the depth at which undis-

turbed samples were collected. For SOC and clay contents,

information from the 0 to 10 and 10 to 30 cm depth layers

was aggregated, but how the BD values corresponding to

the 0–30 cm layer were derived was not described.

2.3

|

Centrality of soil

structural condition

In another more fundamental respect, Poeplau and

Don (

) misrepresent, the essence of the SOC:clay

index approach, when they wrote that ‘although initially

developed in the context of soil structure, the rationale be-

hind the index is to provide a basis for comparing actual

SOC levels of differently textured soils’. If the goal was in-

deed to compare actual SOC levels of differently textured

soils, then a different route would certainly have been

taken by Johannes, Matter, et al. (

) than to rely on

the SOC:clay ratio. Indeed, these authors clearly showed

that physical parameters including the soil BD are far bet-

ter correlated with the SOC content directly than with the

SOC:clay ratio. A similar observation has been made by

Poeplau and Don (

) and was indeed reported long

ago by many authors, as reviewed in Johannes, Matter,

et al. (

).

However, the objective pursued explicitly by Johannes

et al. (

,

), contrary to what Poeplau and

Don (

) wrote was to find how the level of SOC affects

the soil

structural

condition. Consequently, they do not

deal with a SOC content that could be achieved in rela-

tion to observed averages, as in Poeplau and Don (

),

but with a soil condition that should be achieved, referring

to the functions associated with the structure and inde-

pendently of the observed SOC distribution.

While many of the soil physical properties linearly cor-

relate to SOC content, none alone successfully define the

soil structure quality, as reviewed in Johannes et al. (

2019

As explained in detail in Baveye et al. (

,

) and

Vogel et al. (

), this question has huge practical con-

sequences, since climatic change threatens the ability of

soils to fulfil a number of essential services, for which

soils having a resilient structure—now often referred

14752743, 2024, 2, Downloaded from https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/sum.13060 by Hayley Peter-Contesse , Wiley Online Library on [18/10/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

4 of 7

|

SAUZET et al.

to as their ‘architecture’—and hence, an adequate SOC

content, is a key requirement. In that context, Johannes,

Matter, et al. (

); Johannes et al. (

) focused on pur-

pose not only on soil physical parameters in general but

also specifically on the architecture of soils, about which

the BD, considered by Poeplau and Don (

), is a poor

descriptor.

Poeplau and Don (

) appear to have overlooked

this property of the SOC:clay ratio, which is not provided

by the SOC content metric they proposed. The structure

quality follows a broken stick pattern with respect to SOC

(Johannes, Matter, et al.,

), as well as the structural

porosity volume, which is a key parameter of soil structure

quality. Indeed, in the general context of the influence

of SOC on the structural quality of soils, the results of

Johannes, Weisskopf, et al. (

) and Prout et al. (

)

suggest a method to assess the structure

vulnerability

of

cropped soils.

Their data were collected across a wide range of crop-

ping practices, soil use and seasons. The observed aver-

age structure quality was, therefore, the result of their

resistance and resilience under multiple stresses, which

are key components of soil quality (Seybold et al.,

1999

).

According to Kay (

1998

), structure vulnerability ‘reflects

the combined characteristics of resiliency and stability’,

and stability is the resistance of the structure to stresses.

Various authors, including Seybold et al. (

1999

), have

pointed out that vulnerability and structural condition

evaluations are two critical components of soil structure

assessment. In that respect, one could consider that the

different average structural states of soils, delineated by

the SOC:clay ratio thresholds identified by Johannes,

Weisskopf, et al. (

), reflect the combination of the

resistance and resilience properties of soils (Kay,

;

From this perspective, Fell et al. (

2018

), Dupla

et al. (

) and Johannes et al. (

) proposed that the

SOC:clay ratio should be seen as a ‘structure vulnerabil-

ity indicator’ (SVI). Hu et al. (

) recently pointed out

that soil structural vulnerability assessments should focus

on key soil structural indicators, such as pore network

morphology or hydraulic properties. Bulk density is not

informative in that context, in spite of what Poeplau and

Don (

) suggest. Moreover, Hu et al. (

) pointed out

a lack of studies demonstrating linkage between soil struc-

tural vulnerability and loss of soil functions or ecosystem

services, which is consistent with the non- linear relation-

ship between SOC or physical properties and structure

quality. The SOC:clay ratio, if used as an SVI, may help to

quantify such a link.

As Poeplau and Don (

The indicator proposed by Poeplau and Don (

) acknowledge, the European

Union is currently developing a Soil Health Law as a legal

framework to achieve the objectives of the European Soil

Strategy. In that respect, it is of key importance to in-

troduce soil health metrics based on soil quality rather

than on objectives that merely try to keep up the current

situation.

2.4

|

Observed SOC content is misleading

) does

not refer to soil functions but qualifies the SOC level from

‘very good’ to ‘damaged’ according to the current situa-

tion that could very well already be SOC depleted. In this

respect, we believe that this indicator can be useful in ena-

bling farmers to exchange information on the SOC con-

tent of their soils in relation to regional averages. But this

should be done with the knowledge that in a context of

SOC loss such as observed during the twentieth century,

a field can show a locally ‘good’ SOC content with respect

to this index (i.e. with respect to the regional average) and

have a SOC content considered too low in terms of the

soil's functions.

There is, therefore, disadvantage with this strategy in

allowing or hiding a poor and decreasing structure con-

dition. For instance, Dupla et al. (

) showed that the

SOC:clay ratio in the Geneva canton is far below the ‘de-

graded’ structure vulnerability threshold for most fields.

As an illustration,

Figure

presents the analyses of the

fields from arable land of Geneva canton used in Dupla

et al. (

), with the structure vulnerability thresholds

and the quality classes based on the metric introduced by

Poeplau and Don (

). It can be seen that these soils

show a very low SOC content on the full clay content range,

with an average SOC:clay of 0.06. However, there are ex-

amples of fields reaching the 1:8 SOC:clay over the full

clay content range. The overall relationship between SOC

and clay in this region is SOC

=

0.0282.clay

+

0.8018 and

shows a

R

2

of .2252, thus presenting a better determined

relationship but a very similar slope than illustrated in

Poeplau and Don (

) (0.0282 and 0.0288, respectively).

2.5

|

From SOC:clay ratio to soil

vulnerability assessment

In the arable land of Geneva canton, 100%, 97%, 85% and

46%, of soil falling in the ‘Degraded’, ‘Moderate’, ‘Good’

and ‘Very Good’ SOC level classes proposed by Poeplau

and Don (

), respectively, are below the 1:13 SOC:clay

ratio threshold, and are therefore considered to have vul-

nerable structure. For the 1:10 threshold, these categories

are 100%, 100%, 99% and 83%, respectively. As shown in

Dupla et al. (

), in the neighbouring canton of Vaud,

with similar soils but a different agriculture management

14752743, 2024, 2, Downloaded from https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/sum.13060 by Hayley Peter-Contesse , Wiley Online Library on [18/10/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

|

5 of 7

SAUZET et al.

history, the SOC:clay ratios are much higher, being typi-

cally above 1:13. Using Poeplau and Don's (

) metric

in Geneva state would significantly mislead farmers about

their SOC level and structure vulnerability, as many of

the ‘good’ and ‘moderate’ categories would actually cor-

respond to very high vulnerability, requiring correction by

increasing the SOC content.

Introduction of the SVI has made farmers aware of the

dramatic condition of their soils. Many stated, however,

that even reaching the moderate threshold was not fea-

sible, or that such thresholds were out of reach in clayey

soils, which was challenged by the field results presented

in Dupla et al. (

) and Johannes et al. (

) in their

regional surveys, with farmers managing to reach ‘very

good’ SVI in all cases including clayey soils. In fact, Dupla

et al. (

) have shown that SOC levels are affected by

agricultural practices and are now increasing rapidly

throughout the region, making the 1:10 SOC:clay goal

realistic for farmers and agriculture management. The

practices identified by Dupla et al. (

) that could sig-

nificantly impact SOC change at the regional scale were,

unsurprisingly, (i) soil tillage intensity, (ii) cover crops in

terms of both duration and diversity and (iii) organic mat-

ter application. This is in line with conservation agricul-

ture principles and can be used to assess cropping systems

with respect to SOC management.

2.6

|

Practical implications

Therefore, the observation that soils from some regions

show low SOC:clay ratios in no way means that the thresh-

old values are ‘strongly biased or misleading’. From an

agronomic point of view, this should lead us to question

current farming practices, identify pioneering farmers

who achieve a good SVI and understand how their prac-

tices can be applied more broadly on a larger scale.

Interestingly, Poeplau and Don (

) presented fig-

ures that, though as a minority, show some large SOC:clay

even in the large clay content range. The same observation

could also be made in the Swiss cropland reported above,

and investigating the corresponding farms may facilitate

the identification of cropping practices that improve soils

(Dupla et al.,

It is clear that certain soil and climate conditions are

not suitable for significant increases in SOC content, for

example, non- irrigated arid regions. However, we ob-

serve a strong tendency to underestimate the achievable

SOC level when soils are depleted. If SOC content targets

should be adapted, it would require us to carefully address

this issue, by investigating the cropping systems with re-

spect to the SOC management principles, and identifying

the best practices. This agri- environmental management

standpoint should not be confused with the intrinsic re-

sponse of soil health to the SOC:clay ratio. From a pedo-

logical and soil health point of view, observing low SOC

values at the regional scale should lead us to assess the

structure condition of the corresponding soil groups

and reassess the structure vulnerability thresholds if it is

shown that the observed structure quality does not match

the thresholds published in Johannes, Matter, et al. (

From this point of view, the observations on the SOC:clay

ratio reported by Poeplau and Don (

) calls for both

agronomical and pedological investigations that have not

been carried out so far, before being able to conclude on

the usefulness of the SOC:clay ratio.

This leads us to comment on the European

Commission's proposal to use the SOC/clay ratio of 1:13

as a threshold for healthy soils. In should be noted that

in this proposal, the depth of soil sampled is not defined.

FIGURE 1

Observed soil organic

carbon (SOC) and clay content in %

w/w in the fields from Geneva canton

arable land. Dots represent the quality

based on the SOC metric from Poeplau

& Don,

. Bold lines represent

the structure vulnerability thresholds

as introduced in Johannes, Matter,

et al.,

.

14752743, 2024, 2, Downloaded from https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/sum.13060 by Hayley Peter-Contesse , Wiley Online Library on [18/10/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

6 of 7

|

SAUZET et al.

If the soil has been ploughed and sampled within the

ploughed layer, the sample depth probably makes no dif-

ference as the SOC profile would be fairly homogeneous.

However, under reduced tillage and soil regeneration

practices, SOC can accumulate nearer the surface. In

such cases, the sampling depth is significant, with a larger

0–30 cm layer arguably masking SOC increases of the top

5–10 cm. This presents a dilemma in agri- environmental

management regarding the appropriate depth for the 1:13

SOC:clay ratio threshold to be applied to achieve accept-

able soil health while avoiding demotivating farmers.

In short, it is important that pedological, agronomic

and agri- environmental considerations not be confused

when using the SVI index, such as in the articles by

Poeplau and Don (