Is intensive management associated with low soil carbon in Irish farms? Implications for developing indicators of farm soil health and nature value

Abstract

Ecologically functioning soils are increasingly viewed as a key component of sustainable agriculture, a means of sequestering carbon and a major contributor to farmland biodiversity. It is generally considered that intensively managed agricultural systems are associated with reduced soil health and require high chemical inputs to maintain soil nutrients. It is important, therefore, to clarify which soil parameters exhibit clear differences between high- and low-intensity farm management and, subsequently, which parameters might then be considered as meaningful indicators of soil health. This study investigated 30 physico-chemical properties of soils collected from 31 Irish farms. Although several soil parameters showed clear differences between the two study areas (Counties Sligo and Wexford), a much smaller subset of properties showed meaningful differences among intensively, intermediately and extensively managed farms. The clearest patterns were shown by several, co-correlated, carbon-based properties, which were all lower in farms under intensive management. The average total carbon content of intensively managed farms was <5%, and we suggest that this might be used as an initial threshold indicator of a degraded and/or low-carbon soil for Irish farms. In general, our study adds additional support for the use of carbon-based parameters as indicators of farm soil health, but further work is needed to ascertain whether different calibrations are required in different regions and for different soil types. In a wider ecological capacity, soil carbon could be included in a suite of environmental and ecological indicators, such as habitat heterogeneity, floral richness and invertebrate diversity, already proposed as measures of farm nature value.

Main article text

Introduction

Within an agricultural context, well-maintained soils are generally considered essential for the efficient, and profitable, production of tillage crops and livestock pasture, and thus there is currently much interest in long-term, standardised monitoring of soil quality and health (e.g., FAO, 2020; European Commission, 2021; FAO/ITPS, 2021; Davis et al., 2023). Maintenance of healthy soil promotes several critical biological and ecological processes, such as nutrient cycling and the maintenance of a diverse soil fauna (Paterson et al., 2007). Additionally, agricultural soils are increasingly being viewed in terms of their potential for carbon (C) sequestration and using enhanced soil C both as a means for mitigating the environmental footprint of individual farms and for reducing atmospheric C in a global sense (Söderström et al., 2014; Lehman et al., 2020).

In order to describe any effects of farm management on soil quality and soil health, there is first a need to propose and evaluate indicators that may be more-or-less relevant to different stakeholders (Karlen et al., 2019; Lehman et al., 2020; Harris et al., 2022). Concepts of soil health often treat soil in terms of a dynamic, functioning ecosystem, which would ideally support diverse assemblages of macro- and micro-organisms, and could be described in terms of ecosystem processes such as trophic networks, nutrient cycling and stability (Cardoso et al., 2013; Karlen et al., 2019). For environmental scientists and ecologists, therefore, biological indicators related to ecosystem functioning, such as microbial diversity, invertebrate abundance and ecosystem resilience, offer ecologically meaningful proxy measures for soil health (O’Halloran et al., 1999; van Bruggen & Semenov, 2000; Stewart et al., 2018; Bhaduri et al., 2022). In contrast to soil health, descriptions of soil quality are dependent upon situation (e.g., natural vs. agricultural systems) and are only meaningful when placed into the context of what the target soil is required to achieve (Sojka & Upchurch, 1999). So, for farmers, the availability of macronutrients, such as nitrogen (N) and phosphorus (P), and micronutrients such as metal ions, would be considered informative indicators of soil quality as these measures are directly relevant to crop or pasture growth (Cardoso et al., 2013; Radujković et al., 2021; Irvine et al., 2023). When debating the many potential measures of soil health and/or soil quality, consideration must be given, therefore, to all the different stakeholders involved and decisions made regarding which proxy measure might be of value to each, or to all.

Ireland is dominated by highly modified landscapes: over 60% of land is classified as agricultural, and approximately 80% of this land is considered improved grasslands for livestock (Bottero et al., 2021). As only a minute area of Irish “countryside” represents true natural wilderness, biodiversity initiatives often focus on promoting “semi-natural” habitats found within agricultural landscapes (Larkin et al., 2019; Moran et al., 2021). The presence and quality of these semi-natural habitats, such as unimproved grasslands, ponds, hedgerows/treelines and patches of mixed woodland, are then each deemed to contribute to the overall farm “nature value” (Sheridan et al., 2017; Larkin et al., 2019). In terms of management, there is generally an inverse relationship with farming intensification and farm nature value (Sullivan et al., 2017; Maskell et al., 2019; Matin et al., 2020; Carlier et al., 2023). For example, intensively managed livestock farms, with high use of synthetic fertilisers and pesticides, and high animal stocking rates, tend to have low nature value because they lack unimproved grasslands, have uniform field boundaries and exhibit low overall habitat heterogeneity. Conversely, extensively managed farms are more frequently associated with a higher nature value, as they tend to have lower animal stocking rates, contain a greater diversity of habitats and involve lower chemical inputs (Boyle et al., 2015; Rotchés-Ribalta et al., 2021; Lomba et al., 2023).

From the above discussion, classification of farms in terms of their management intensity can, therefore, often provide a valuable proxy for the designation of farms in terms of their nature value. Recent research has supported this premise by describing how extensively managed Irish farms promote higher invertebrate diversity and abundance compared with intensive farming systems (Volpato et al., 2020; Ahmed et al., 2021). Additionally, in the context of farm management, highly intensive agricultural production can result in multiple aspects of soil degradation, such as erosion, contamination, increased salinity, compaction and depletion of organic matter (OM) (Lehman et al., 2020; Dupla et al., 2022; Khangura et al., 2023). Thus, as farm management practices can affect several physical and chemical soil properties, it can be hypothesised that the connections between farming intensification, nature value and biodiversity might easily be extended further to include additional soil-based indicators of environmental well-being and sustainable food production (Söderström et al., 2014; Lynch et al., 2019; Lehman et al., 2020).

The aim of this investigation was to screen multiple potential indicators of soil quality and health in Irish livestock farms and ascertain which of these indicators, if any, most clearly responded to farm management intensity in a logical and consistent manner. To do this, we quantified 30 soil properties in four general categories (physico-chemical, metal ions, macronutrients, C) from farms classified as being under intensive, extensive or intermediate farm management. To assess the generality of any observed patterns, soil sampling was carried out in two contrasting biogeographical areas of Ireland, namely County Wexford in the south-east and County Sligo in the north-west.

Material and methods

Study sites and sampling

Sampling was performed in two Irish counties: Sligo and Wexford. The Sligo sites were centred on the coordinates 54.243° N, 8.6065° W, and ran for approximately 13 km from Duneill in the west to Shemagh in the East, spreading over a total land area of approximately 45 km². The Wexford sites were in the north-west of the county, centred on coordinates 52.603° N, 6.681° W, and covered an area of 106 km² in an approximate triangle with Kiltealy in the west, Munfin Lower in the east and Kildavin in the north. Mean annual temperature is similar in both counties (Sligo 9.6°C; Wexford 9.8°C), whereas precipitation is considerably higher in Sligo (1260.1 mm) than in Wexford (840.2 mm) (www.met.ie/). The average altitude of the Sligo sites (∼50 m asl) was substantially lower than the average attitude of the Wexford sites (∼160 m asl).

The study sites were all primarily livestock farms and were classified into three categories based on their level of management intensification: intensive, extensive and intermediate. These classifications were based on an index presented by Boyle et al. (2015), which considers factors such as the area farmed, the livestock stocking rate, the proportion of grasslands that are improved, field size and the quality and quantity of linear habitats. The management classification cut-off points we used are based on the value of the summary index: extensive, index >5; intermediate, index 3.5–5 and intensive index <3.5. Based on the data provided by Boyle et al. (2015), a typical intensively managed farm could have a farmed area under 20 ha, with 65% of grasslands considered as improved, livestock stocking rates of around 2 units/ha and around 3 km of linear features. In comparison, a typical extensively managed farm could have a farmed area of approximately 50 ha, with only 25% of improved grasslands, a stocking rate of around 1 unit/ha and approximately 8 km of linear features (Boyle et al., 2015).

General soil properties for both sampling regions were obtained using online data sources (gis.epa.ie and gis.teagasc.ie) with additional information from Creamer & O’Sullivan (2018). The soils in both the Sligo (Water Framework Directive sub-catchment – Dunmoran_SC_010) and Wexford (Water Framework Directive sub-catchments – Slaney_SC_070 and Urrin_SC_010) sampling areas are generally described as “Brown Earths” and “Brown Podzolics”, which are mainly well-drained, but with some areas of poorly drained peats.

Soil samples were taken at a depth of 15 cm, with approximately 5 kg soil in total collected from each site. The 5 kg total sample was formed by pooling three sub-samples that were collected from a single field, approximately 50 m apart, by crossing the centre of the field diagonally. To avoid spurious nutrient concentrations due to livestock gatherings, no samples were collected within 25 m of a water trough, field boundary, fence line, hedgerow or farm gates (e.g., Ekanayake et al., 2017). Sampling was conducted in June and July 2018 for Sligo (15 farms) and August and September 2018 for Wexford (16 farms). In Wexford, two farms had areas under tillage and grazing land. For completeness, samples were taken from both grazing and tillage soils, and, for the purposes of this preliminary study, these spatially distant samples have been considered as independent replicates.

Soil analysis

Several physico-chemical properties were assessed to provide an initial description of each soil. Soil pH and electrical conductivity (EC) were measured using a 1:5 fresh soil:water ratio; soil moisture content was calculated after drying for 24 h at 105°C. Soil particle density was measured using a pycnometer, and soil porosity (SP) was then calculated as bulk density (BD) divided by soil particle density.

The OM content of each soil was calculated after weight loss on ignition at 450°C for 8 h. Total C (TotC), total organic C (TOC) (after acidification) and N were analysed by elemental analysis (LECO TruSpec CN analyser, LECO Corporation, UK) after air drying and fine grinding of samples. Plant available phosphorus (P) was extracted using methods described by Olsen et al. (1954) and Mehlich (1984), and measurements obtained via UV spectrophotometry. Plant available NO₂, NO₃ and NH₄ were extracted using a 1 M KCl solvent (McTaggart & Smith, 1993) and analysed using a Konelab auto analyser (ThermoFisherScientific, Ireland). Metal concentration was determined after acid digestion with aqua-regia and analysis via inductively coupled plasma atomic emission spectroscopy.

The labile and recalcitrant fractions of soil OM were separated by density fractionation conducted by shaking 15 g of air-dried soil in 50 mL of NaI solution (density >1.7 g·cm⁻³) for 30 min, and then centrifuged at 3,000 rpm for 10 min (Song et al., 2012). The light fraction (LF) was then Buchner-filtered on fibreglass. The remaining heavy fraction (HF) was shaken with 50 mL of deionized H₂O and centrifuged three more times to achieve full washing. Washing of the LF was carried out with CaCl₃. Both fractions were dried at 60°C for 48 h before being weighed and analysed for C and N.

Permanganate oxidisable C (POxC) was determined in triplicate using 0.02 M KMnO₄ on 2.5 g of air-dried soil and sieved to 2.0 mm (Weil et al., 2003). Absorbance at 550 nm was measured on a Shimadzu single cuvette reader. The evolution of CO₂ was measured after a 24 h incubation in which oven-dried soil was re-wetted prior to incubation (Franzluebbers et al., 1996; Haney & Haney, 2010; Haney et al., 2018). A 20 g soil sample was incubated in 50 mL head space vial in the dark at 22°C, and after 24 h the headspace was sampled using a syringe and analysed using gas chromatography.

Microbial biomass C (MBC) was determined by chloroform fumigation-extraction (Vance et al., 1987). First, field holding capacity of each soil was derived by saturating a 500 g sample of dried soil which was then allowed to drain overnight before being re-weighed. Microbial biomass C was determined using 250 g samples of dried soil which were adjusted to 65% field holding capacity and then incubated for 1 wk at 25°C. The samples were halved, and one half was extracted by chloroform fumigation for 24 h, with the remaining sub-sample used as a paired control. Samples were then extracted with K₂SO₄ and analysed by the potassium dichromate oxidation method.

Data analyses

Data analyses were performed using Community Analysis Package v4 (Pisces Conservation Ltd, UK) and Genstat v20 (VSNI Ltd, Hemel Hempstead, UK) software packages. Within each county, soil profiles from the intensive, extensive and intermediate farms were compared using analysis of similarity (ANOSIM) which provides an indication of statistical significance by comparing the relative within- and between-group similarity with that obtained by 1,000 random permutations of the raw data (Clarke, 1993; Henderson & Seaby, 2008). The ANOSIM procedure also provided a probability value to indicate the likely differences between management categories in a pairwise fashion. Principal component analysis (PCA) was used to visualise the separation of soil profiles in each county and identify which soil properties were most strongly influencing any separation.

Differences in individual soil parameters between the two counties and among the three farm management categories were assessed by unbalanced univariate analysis of variance (ANOVA). If the ANOVA identified any overall significant differences, pairwise comparisons of all six groups of samples were made using Fisher’s least significant difference (LSD, P < 0.05). Relationships among the soil parameters based on C and OM content were evaluated using Spearman’s rank correlation coefficient.

Results

Multivariate analysis of soil chemical and physical profiles

For both Sligo and Wexford, the first two PCA axes explained >40% of the variance in the respective multivariate data sets (Figure 1). In both counties, the soils from the intensively managed farms formed more-or-less distinct clusters based on their physico-chemical properties. Conversely, the soils from the intermediate and extensively managed farms were more dispersed across the PCA biplots with some samples positioned near to the cluster of soils from the intensive farms (Figure 1). In both counties, the PCA analysis indicated that the soils from the intensive farms were negatively associated with measures related to the C content (e.g., TOC, TotC, POxC, HFC). In the Sligo soils, the PCA suggested that intensively managed farm soils might have higher levels of Al and Mg, whereas in Wexford the intensive farms were associated with high levels of Morgan’s phosphorus (MrgP) (Figure 1).

Figure 1.

PCA biplots based on 30 physico-chemical properties of soils collected from farms under intensive, extensive or intermediate management systems in Counties Sligo and Wexford, Ireland. The % variation explained is given for the first and second PCA axes for each county. Arrows indicate the direction and magnitude of individual soil properties with the highest eigenvector scores. HFC = heavy fraction carbon; MrgP = Morgan’s phosphorus; PCA = principal component analysis; POxC = permanganate oxidisable carbon; TN = total nitrogen; TOC = total organic carbon; TotC = total carbon.

The ANOSIM performed on the Sligo soils revealed a significant difference among the soil profiles from farms under the different management regimes (P = 0.009), but this was primarily because of the difference in the soil profiles of the farms under intensive management compared with those under extensive management (P = 0.012; Figure 1). The analysis also indicated that there were no statistically significant differences between the soil profiles of the intensive and intermediate farms (P = 0.258) or the extensive and intermediate farms (P = 0.448).

The ANOSIM performed on the Wexford soils also revealed a significant difference among the overall soil profiles from farms under the different management regimes (P = 0.002). The soil profiles from the intensively managed farms were significantly different from those of the farms under extensive management (P = 0.002) and intermediate management (P = 0.016), but there was no statistically significant difference between the extensive and intermediate farms (P = 0.695; Figure 1).

Individual soil physico-chemical properties

Although several of the general soil properties differed between the two study areas, only one, soil pH, differed significantly among the three farm management categories (Table 1). In Sligo, the soils became more acidic from intensive, to intermediate, to extensive farm management. This same pattern was also present in the Wexford soils but was less distinct, and the three farm categories were not separated statistically (Table 1).

Table 1:

General soil properties and metal ion concentrations (mg/L) of farm soils under three different management systems in Sligo and Wexford

	Sligo			Wexford			P County	P Man	P Int
	Int	Med	Ext	Int	Med	Ext
pH	6.66b,c	6.32b	5.63a	7.20c	7.10c	7.07c	<0.001	0.039	0.074
EC	66.3	53.4	67.5	71.2	77.2	75	0.150	0.808	0.594
MC	21.6a,b	12.6a	22.1a-c	26.3b,c	27.1b,c	31.7c	0.002	0.143	0.393
BD	0.70	0.84	0.58	0.77	0.87	0.83	0.137	0.210	0.421
SP	43.9a,b	41.3a,b	53.1b	39.9a,b	25.5a	28.4a	0.028	0.407	0.324
CO2	1,495a,b	2,134b	1,373a,b	1,023a	1,353a,b	1,326a,b	0.049	0.205	0.462
MBC	891a-c	1,403c	1,322b,c	560a	635a,b	748a,b	0.004	0.328	0.631
Al	12,940a,b	10,906a,b	8,942a	24,922c	17,950a-c	23,170b,c	0.002	0.605	0.734
Ca	2,979a,b	3,604a,b	4,432b	2,836a	2,905a,b	3,963a,b	0.285	0.041	0.860
Cu	18.9	19.3	14.9	20.1	17	24.4	0.303	0.926	0.295
Fe	14,882a,b	12,908a	14,429a,b	23,233c	22,559b,c	27,172c	<0.001	0.526	0.697
K	1,338a	1,247a	1,437a	1,361a	2,434b	2,498b	0.004	0.015	0.047
Mg	1,856a,b	1,626a,b	1,255a	2,293b	2,496b	2,599b	0.004	0.948	0.402
Mn	406a	395a	335a	976b	519a,b	750a,b	0.004	0.342	0.434
Mo	6.0a	4.6a	3.5a	5.2a	11.5b	12.3b	0.002	0.073	0.007
Na	696a	718a	649a	516b	509b	537b	<0.001	0.861	0.410
Zn	53.2a-c	48.5a,b	26.0a	86.6c	68.9b,c	83.2c	0.002	0.905	0.178

P-values obtained from unbalanced two-way ANOVA, with P-values highlighted in bold indicating statistical significance for differences between the two counties (P _County), among the three farm management intensities (P _Man) and the interaction term (P _Int). Values with the same superscript letter code are not significantly different based on Fisher’s LSD (P < 0.05).

ANOVA = analysis of variance; BD = bulk density (g); CO₂ = carbon dioxide (ppm); EC = electrical conductivity (μS/cm); LSD = least significant difference; MBC = microbial biomass carbon (mg/kg); MC = moisture content (% mass); SP = soil porosity (%).

In terms of soil metal concentrations, the Wexford soils were significantly higher in Al, Fe, Mo, Mg, Mn and Zn compared with the Sligo soils (Table 1). Two metal ions, Ca and K, displayed significant differences among the three farm management categories. Both of these metals were relatively low in intensively managed farms and relatively high in extensively managed farms, although the values within each county were not separated statistically (Table 1).

For parameters related to soil macronutrient levels, NH₄, Olsen’s P and MrgP were not different between study areas or among farm categories (Table 2). Nitrate was highest in the Wexford soils compared with Sligo soils, whereas heavy fraction N (HFN), light fraction N (LFN) and total N (TN) were higher in Sligo compared with Wexford. Only TN was identified as being significantly different among the farm management classes and, for each county, was lowest in the intensively managed farms (Table 2).

Table 2:

Parameters related to macronutrient levels, carbon and OM in soils from three farm management categories in Sligo and Wexford

	Sligo			Wexford			P County	P Man	P Int
	Int	Med	Ext	Int	Med	Ext
TN	0.44a	0.66b	0.68b	0.33a	0.42a	0.42a	<0.001	0.036	0.475
HFN	3.42a,b	6.36c	5.83b,c	2.76a	2.20a	4.32a-c	0.004	0.067	0.117
LFN	13.8a	28.3b	14.2a	12.0a	12.8a	13.0a	0.034	0.075	0.108
NO3	2.12a,b	1.55a	1.50a	2.05a,b	2.63b	2.73b	0.008	0.876	0.059
NH4	0.32	1.3	1.16	0.84	0.48	0.85	0.671	0.719	0.371
Ols P	6.8	7.8	5.8	5.2	6.7	8.2	0.728	0.252	0.100
MrgP	9.1	8.3	14.0	18.0	7.1	5.3	0.823	0.313	0.165
TotC	4.22a	8.57b,c	9.92c	3.62a	5.31a,b	5.20a,b	0.005	0.019	0.234
TOC	3.30a,b	7.05b,c	9.93c	2.85a	3.88a,b	4.06a,b	0.005	0.031	0.123
HFC	37.4a,b	74.6b,c	83.3c	29.6a	28.8a	50.7a-c	0.008	0.046	0.313
LFC	205.9a	244.2a,b	211.2a	272.4b	231.5a,b	242.8a,b	0.031	0.619	0.115
POxC	695b,c	851c,d	961d	500a	611a,b	662a-c	<0.001	0.015	0.743
OM	11.3a,b	19.9c	16.6b,c	9.3a	11.4a,b	13.9a-c	0.023	0.055	0.364

P-values obtained from unbalanced two-way ANOVA, with those highlighted in bold indicating statistical significance for differences between the two counties (P _County), among the three farm management intensities (P _Man) and the interaction term (P _Int). Values with the same superscript letter code are not significantly different based on Fisher’s LSD (P < 0.05).

ANOVA = analysis of variance; HFC = heavy fraction carbon (g/kg); HFN = heavy fraction nitrogen (g/kg); LFC = light fraction carbon (g/kg); LFN = light fraction nitrogen (g/kg); LSD = least significant difference; MrgP = Morgan’s phosphorus (mg/L); NH₄ = ammonium (mg/L); NO₃ = nitrate (mg/L); Ols P = Olsen’s phosphorus (mg/L); OM = organic matter (% dry weight); POxC = permanganate oxidisable carbon (per kg); TN = total nitrogen (% dry weight); TOC = total organic carbon (% dry weight); TotC = total carbon (% dry weight).

Several parameters pertaining to soil C and OM responded significantly to farm management (Table 2). For example, TotC, TOC, heavy fraction C (HFC) and POxC were all highest in the farms under extensive management and lowest in the farms under intensive management (Table 2). There was also moderate evidence (P = 0.055) that OM was also highest in the extensively managed farms compared with intensively managed farms (Table 2). Although these trends were seen in both counties, the patterns were generally more pronounced, and separated statistically, in the Sligo farms compared with the Wexford farms. For example, in Sligo, the TOC in soils from extensively managed farms was over 3-fold that found in soils from intensively managed farms, compared to <1.5-fold in Wexford. On the other hand, soil OM from extensive farms was around 50% higher than that found in intensive farms in both Sligo and Wexford (Table 2).

Evaluation of the relationships among the C-based soil properties indicated that TotC, TOC, POxC and HFC were all closely related (r _S > 0.7 in all cases). Additionally, soil OM generally had a weaker, but still statistically significant, relationship with most of the other C-based variables (Table 3). The spurious C measure was light fraction C (LFC), which showed a weak negative relationship with all of the other five C-based soil properties (none of which were statistically significant; r _S < |0.16| in all cases; Table 3).

Table 3:

Relationships between properties of Irish soils pertaining to soil carbon and OM content

	OM	TotC	TOC	POxC	HFC	LFC
OM		0.001	0.001	0.003	0.001	0.686
TotC	0.688		0.001	0.001	0.001	0.753
TOC	0.716	0.922		0.001	0.001	0.856
POxC	0.498	0.771	0.767		0.001	0.387
HFC	0.589	0.791	0.717	0.808		0.905
LFC	0.073	−0.057	−0.033	−0.156	−0.022

Lower half of the table shows Spearman’s rank correlation coefficients; upper half of the table shows P-value based on 33 soil samples.

HFC = heavy fraction carbon (g/kg); LFC = light fraction carbon (g/kg); OM = organic matter (% dry weight); POxC = permanganate oxidisable carbon (per kg); TOC = total organic carbon (% dry weight); TotC = total carbon (% dry weight).

Discussion

The primary aim of the study was to identify soil parameters that exhibited meaningful and logical trends across different levels of farm management intensification. Several soil parameters were identified as being significantly different between Sligo and Wexford, which was not unexpected given the geographic separation of the study areas and differences in environmental factors such as annual rainfall and altitude. Considerably fewer soil properties showed consistent, statistically significant and meaningful trends with farm management intensification, although the multivariate statistical analyses based on overall physico-chemical profiles clearly separated soils from farms under intensive and extensive management systems in both counties.

In general, the soil parameters which separated management groups were related to soil C and OM. Total organic C, TotC, HFC and POxC, were found to be significantly separated among farm management groups by univariate statistical analysis and were also identified by the PCAs as tending to be lower in soils from intensively managed farms. Soil OM and soil organic C are central to many physical and chemical processes within soils and play critical roles in maintaining soil structure, water holding capacity and nutrient retention (Lal, 2014, 2016). Several parameters based on soil C are already accepted as being meaningful indicators of both soil quality and soil health, and/or have been demonstrated to respond to land management (Chantigny, 2003; Harrison et al., 2011; Bongiorno et al., 2019; Ramesha et al., 2019; Lehman et al., 2020). Additionally, C-based soil measures are often positively correlated with other ecological indicators, such as soil microbial diversity and invertebrate abundance, and so provide a potential soil-based link with the wider nature value of the farm (e.g., Hurisso et al., 2016; Morrow et al., 2016; Prout et al., 2021).

With the exception of LFC, there were strong positive correlations between the different C-based measures. For simplicity, and parsimony, this result offers the opportunity for developing a single C-based indicator in lieu of measuring a whole suite of parameters. POxC is a measure of active or “labile” C that is relatively easy to quantify, and previous studies have suggested that POxC is sensitive to farm management practices (Tan et al., 2007; Mirsky et al., 2008; Ibrahim et al., 2015). Soil OM is also a relatively simple and cost-effective parameter to measure, and was positively correlated with other carbon measures, such as TOC, TotC, POxC and HFC. In our study, in both Sligo and Wexford, the intensively managed farms had an average TotC of <5%, whereas the farms under intermediate and extensive management had an average TotC of >5%. Thus, based on these results, a TotC of <5% might represent a sensible threshold value for use as an initial indication of low soil health or soil degradation due to intensive farm management.

We concede that our results are based on only a single sampling event from a relatively small number of farms in each management category, and we have not accounted for the considerable variation in soil properties that can occur within a single farm, or even within a single field (Ekanayake et al., 2017). Lehmann et al. (2020) previously suggested that many soil health indices either require regional calibration or are inappropriate for direct comparisons of different soil types, so there could clearly be some issues with proposing a single threshold carbon content for all Irish regions. Also, whether a soil C content is designated as high or low can also depend on the soil clay content, and so presenting a combined measure, such as a C:clay ratio, could be considered a more meaningful indicator of soil health (Prout et al., 2021).

Although the soils under intensive management had similar levels of C-based measures in both counties, the magnitude of the positive responses to extensive management in terms of increased C was much more profound in Sligo than in Wexford. In Ireland, the use of carbon as a measure of both soil health and soil quality is further complicated because peat-dominated soils have very high C and OM content but are often considered low-quality soils in terms of productivity. Clearly, our suggestion to use a TotC value of 5% as a threshold is based only on these preliminary data, and more extensive sampling, across many regions, is required to substantiate this proposed threshold and decide whether more geographically specific carbon values are required at a province or county level.

When considering overall soil properties, in both Sligo and Wexford, the samples from the intensively managed farms exhibited tight clustering across the PCA axis, indicating that these soils had relatively consistent profiles. Conversely, the samples from the intermediate and extensively managed farms were more dispersed across the PCA biplots, indicating that these soil profiles were comparatively disparate and, in some cases, actually had similar profiles to soils from intensive farms. This interspersion of soil samples from the extensive and intermediate farms on the PCA plots, and the lack of statistical separation of these profiles by the ANOSIM procedure, is at least partly the result of difficulties encountered in classifying some farms in terms of management regime. Placing of an individual farm into a management class was based on a summary index and did not necessarily mean that this farm met all the stipulated thresholds for all the individual parameters that were considered (e.g., stocking rate, field size, linear features, etc.). Similarly, we did not consider how long the farm had been managed in a particular way and if this time was adequate for effects on soil profile to become apparent (see Söderström et al., 2014).

In a wider sense, implementing a management approach that enhances soil C is often associated with secondary ecological benefits, such as increased biodiversity and maintenance of ecosystem functions, and farmers are now being encouraged to consider the natural capital of their farms and how adopting management systems with an underlying ethos of “carbon farming” can increase this capital. Soil C and OM therefore appear to be highly suitable, logical indicators of soil health and valuable tools for quantifying temporal changes in soil composition after farmers have adopted more sustainable management systems. Additionally, it seems logical that some measures of soil carbon could be included in a suite of environmental and ecological indicators, such as habitat heterogeneity, the quantity and quality of linear features and botanical and invertebrate diversity, that are already proposed as measures of farm nature value.

Conclusions

From the suite of soil properties we examined, our results clearly indicated that several measures related to soil C were lowest in farms under intensive management. Specifically, we feel that soil OM and C content have good potential to be developed further as bioindicators of farm soil health in Ireland and provide a means of directly monitoring the progress of farm management practices, such as organic and regenerative agriculture, that deliberately attempt to enhance soil C and heighten soil biodiversity.

References

1. Ahmed KSD, Volpato A, Day MF, Mulkeen CJ, O’Hanlon A, Carey J, Williams C, Ruas S, Moran J, Rotchés-Ribalta R, Ó hUallacháin D, Stout JC, Hodge S, White B, Gormally MJ. 2021. Linear habitats across a range of farming intensities contribute differently to dipteran abundance and diversity. Insect Conservation and Diversity. Vol. 14:335–347. [Cross Ref]

2. Bhaduri D, Sihi D, Bhowmik A, Verma BC, Munda S, Dari B. 2022. A review on effective soil health bio-indicators for ecosystem restoration and sustainability. Frontiers in Microbiology. Vol. 13:938481. [Cross Ref]

3. Bongiorno G, Bünemann EK, Oguejiofor CU, Meier J, Gort G, Comans R, Mäder P, Brussaard L, de Goede R. 2019. Sensitivity of labile carbon fractions to tillage and organic matter management and their potential as comprehensive soil quality indicators across pedoclimatic conditions in Europe. Ecological Indicators. Vol. 99:38–50. [Cross Ref]

4. Bottero I, Hodge S, Stout J. 2021. Taxon-specific temporal shifts in pollinating insects in mass-flowering crops and field margins in Ireland. Journal of Pollination Ecology. Vol. 28:90–107. [Cross Ref]

5. Boyle O, Hayes M, Gormally M, Sullivan C, Moran J. 2015. Development of a nature value index for pastoral farmland – a rapid farm-level assessment. Ecological Indicators. Vol. 56:31–40. [Cross Ref]

6. Cardoso EJBN, Vasconcellos RLF, Bini D, Miyauchi MYH, dos Santos CA, Alves PRL, de Paula AM, Nakatani AS, de Moraes Pereira J, Nogueira MA. 2013. Soil health: looking for suitable indicators. What should be considered to assess the effects of use and management on soil health? Scientia Agricola. Vol. 70:274–289. [Cross Ref]

7. Carlier J, Doyle M, Finn JA, Ó hUallacháin D, Ruas S, Moran J. 2023. The development and potential application of a land use monitoring programme for high nature value farmland and forest quality and quantity in the Republic of Ireland. Environmental Science and Policy. Vol. 146:1–12. [Cross Ref]

8. Chantigny MH. 2003. Dissolved and water-extractable organic matter in soils: a review on the influence of land use and management practices. Geoderma. Vol. 113:357–380. [Cross Ref]

9. Clarke KR. 1993. Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology. Vol. 18:117–143. [Cross Ref]

10. Creamer R, O’Sullivan L. 2018. The Soils of Ireland. Springer Cham. Germany: 300 pages

11. Davis AG, Huggins DR, Reganold JP. 2023. Linking soil health and ecological resilience to achieve agricultural sustainability. Frontiers in Ecology and the Environment. Vol. 21:131–139. [Cross Ref]

12. Dupla X, Lemaître T, Grand S, Gondret K, Charles R, Verrecchia E, Boivin P. 2022. On-farm relationships between agricultural practices and annual changes in organic carbon content at a regional scale. Frontiers in Environmental Science. Vol. 10:834055. [Cross Ref]

13. Ekanayake D, Owens J, Hodge S, Trethewey J, Roten R, Westerschulte M, Belin S, Werner A, Cameron K. 2017. Soil inorganic nitrogen in spatially distinct areas within a commercial dairy farm in Canterbury, New Zealand. Journal of New Zealand Grasslands. Vol. 79:83–88. [Cross Ref]

14. European Commission. 2021. EU Soil Strategy for 2030: Reaping the Benefits of Healthy Soils for People, Food, Nature and Climate. Brussels:

15. FAO. 2020. A Protocol for Measurement, Monitoring, Reporting and Verification of Soil Organic Carbon in Agricultural Landscapes – GSOC-MRV Protocol. FAO. Rome: 140 pages[Cross Ref]

16. FAO/ITPS. 2021. Recarbonizing Global Soils – A Technical Manual of Recommended Management Practices. Vol. Volume 1. FAO. Rome: 52 pages[Cross Ref]

17. Franzluebbers AJ, Haney RL, Hons FM, Zuberer DA. 1996. Determination of microbial biomass and nitrogen mineralization following rewetting of dried soil. Soil Science Society of America Journal. Vol. 60:1133–1139. [Cross Ref]

18. Haney RL, Haney EB. 2010. Simple and rapid laboratory method for rewetting dry soil for incubations. Communications in Soil Science and Plant Analysis. Vol. 41:1493–1501. [Cross Ref]

19. Haney RL, Haney EB, Smith DR, Harmel D, White MJ. 2018. The soil health tool – theory and initial broad-scale application. Applied Soil Ecology. Vol. 125:162–168. [Cross Ref]

20. Harris JA, Evans DL, Mooney SJ. 2022. A new theory for soil health. European Journal of Soil Science. Vol. 73:e13292. [Cross Ref]

21. Harrison RB, Footen PW, Strahm BD. 2011. Deep soil horizons: contribution and importance to soil carbon pools and in assessing whole-ecosystem response to management and global change. Forest Science. Vol. 57:67–76. [Cross Ref]

22. Henderson P, Seaby R. 2008. A Practical Handbook for Multivariate Methods. Pisces Conservation Ltd. Great Britain:

23. Hurisso TT, Culman SW, Horwath WR, Wade J, Cass D, Beniston JW, Bowles TM, Grandy AS, Franzluebbers AJ, Schipanski ME, Lucas ST, Ugarte CM. 2016. Comparison of permanganate-oxidizable carbon and mineralizable carbon for assessment of organic matter stabilization and mineralization. Soil Science Society of America Journal. Vol. 80:1352–1364. [Cross Ref]

24. Ibrahim M, Cao C-G, Zhan M, Li C-F, Iqbal J. 2015. Changes of CO₂ emission and labile organic carbon as influenced by rice straw and different water regimes. International Journal of Environmental Science and Technology. Vol. 12:263–274. [Cross Ref]

25. Irvine R, Houser M, Marquart-Pyatt ST, Bogar G, Bolin LG, Browning EG, Evans SE, Howard MM, Lau JA, Lennon JT. 2023. Soil health through farmers’ eyes: toward a better understanding of how farmers view, value, and manage for healthier soils. Journal of Soil and Water Conservation. Vol. 78:82–92. [Cross Ref]

26. Karlen DL, Veum KS, Sudduth KA, Obrycki JF, Nunes MR. 2019. Soil health assessment: past accomplishments, current activities, and future opportunities. Soil and Tillage Research. Vol. 195:104365. [Cross Ref]

27. Khangura R, Ferris D, Wagg C, Bowyer J. 2023. Regenerative agriculture – a literature review on the practices and mechanisms used to improve soil health. Sustainability. Vol. 15:2338. [Cross Ref]

28. Lal R. 2014. Societal value of soil carbon. Journal of Soil and Water Conservation. Vol. 69:186A–192A. [Cross Ref]

29. Lal R. 2016. Soil health and carbon management. Food and Energy Security. Vol. 5:212–222. [Cross Ref]

30. Larkin J, Sheridan H, Finn JA, Denniston H, Ó hUallacháin D. 2019. Semi-natural habitats and ecological focus areas on cereal, beef and dairy farms in Ireland. Land Use Policy. Vol. 88:104096. [Cross Ref]

31. Lehmann J, Bossio DA, Kögel-Knabner I, Rillig MC. 2020. The concept and future prospects of soil health. Nature Reviews Earth & Environment. Vol. 1:544–553. [Cross Ref]

32. Lomba A, McCracken D, Herzon I. 2023. Editorial: high nature value farming systems in Europe. Ecology and Society. Vol. 28:20. [Cross Ref]

33. Lynch J, Donnellan T, Finn JA, Dillon E, Ryan M. 2019. Potential development of Irish agricultural sustainability indicators for current and future policy evaluation needs. Journal of Environmental Management. Vol. 230:434–445. [Cross Ref]

34. Maskell LC, Botham M, Henry P, Jarvis S, Maxwell D, Robinson DA, Rowland CS, Siriwardena G, Smart S, Skates J, Tebbs EJ, Tordoff GM, Emmett BA. 2019. Exploring relationships between land use intensity, habitat heterogeneity and biodiversity to identify and monitor areas of high nature value farming. Biological Conservation. Vol. 231:30–38. [Cross Ref]

35. Matin S, Sullivan CA, Finn JA, Ó hUallacháin D, Green S, Meredith D, Moran J. 2020. Assessing the distribution and extent of high nature value farmland in the Republic of Ireland. Ecological Indicators. Vol. 108:105700. [Cross Ref]

36. McTaggart IP, Smith KA. 1993. Estimation of potentially mineralisable nitrogen in soil by KCl extraction – II. Comparison with soil N uptake in the field. Plant and Soil. Vol. 157:175–184. [Cross Ref]

37. Mehlich A. 1984. Mehlich 3 soil test extractant: a modification of Mehlich 2 extractant. Communications in Soil Science and Plant Analysis. Vol. 15:1409–1416. [Cross Ref]

38. Mirsky SB, Lanyon LE, Needelman BA. 2008. Evaluating soil management using particulate and chemically labile soil organic matter fractions. Soil Science Society of America Journal. Vol. 72:180–185. [Cross Ref]

39. Moran J, Byrne D, Carlier J, Dunford B, Finn JA, Ó hUallacháin D, Sullivan CA. 2021. Management of high nature value farmland in the Republic of Ireland: 25 years evolving toward locally adapted results-orientated solutions and payments. Ecology and Society. Vol. 26:20. [Cross Ref]

40. Morrow JG, Huggins DR, Carpenter-Boggs LA, Reganold JP. 2016. Evaluating measures to assess soil health in long-term agroecosystem trials. Soil Science Society of America Journal. Vol. 80:450–462. [Cross Ref]

41. O’Halloran K, Booth LH, Hodge S, Thomsen S, Wratten SD. 1999. Biomarker responses of the earthworm Aporrectodea caliginosa to organophosphates: hierarchical tests. Pedobiologia. Vol. 43:646–651. [Cross Ref]

42. Olsen SR, Watanabe FS, Cosper HR, Larson WE, Nelson LB. 1954. Residual phosphorus availability in long-time rotations on calcareous soils. Soil Science. Vol. 78:141–152. [Cross Ref]

43. Paterson E, Gebbing T, Abel C, Sim A, Telfer G. 2007. Rhizodeposition shapes rhizosphere microbial community structure in organic soil. New Phytologist. Vol. 173:600–610. [Cross Ref]

44. Prout JM, Shepherd KD, McGrath SP, Kirk GJD, Haefele SM. 2021. What is a good level of soil organic matter? An index based on organic carbon to clay ratio. European Journal of Soil Science. Vol. 72:2493–2503. [Cross Ref]

45. Radujković D, Verbruggen E, Seabloom EW, Bahn M, Biederman LA, Borer ET, Boughton EH, Catford JA, Campioli M, Donohue I, Ebeling A, Eskelinen A, Fay PA, Hansart A, Knops JMH, MacDougall AS, Ohlert T, Olde Venterink H, Raynaud X, Risch AC, Roscher C, Schütz M, Silveira ML, Stevens CJ, Van Sundert K, Virtanen R, Wardle GM, Wragg PD, Vicca S. 2021. Soil properties as key predictors of global grassland production: have we overlooked micronutrients? Ecology Letters. Vol. 24:2713–2725

46. Ramesha T, Bolan NS, Kirkham MB, Wijesekara H, Kanchikerimath M, Rao CS, Sandeep S, Rinklebe J, Ok Y, Choudhury B, Wang H, Tang C, Wang X, Song Z, Freeman OW II. 2019. Soil organic carbon dynamics: impact of land use changes and management practices: a review. Advances in Agronomy. Vol. 156:1–107. [Cross Ref]

47. Rotchés-Ribalta R, Ruas S, Ahmed K, Gormally M, Moran J, Stout J, White B, Ó hUallacháin D. 2021. Assessment of semi natural habitats and landscape features on Irish farmland – New insights to inform EU Common Agricultural Policy implementation. Ambio. Vol. 50:346–359. [Cross Ref]

48. Sheridan H, Keogh B, Anderson A, Carnus T, McMahon BJ, Green S, Purvis G. 2017. Farmland habitat diversity in Ireland. Land Use Policy. Vol. 63:206–213. [Cross Ref]

49. Söderström B, Hedlund K, Jackson LE, Kätterer T, Lugato E, Thomsen IK, Jørgensen HB. 2014. What are the effects of agricultural management on soil organic carbon (SOC) stocks? Environmental Evidence. Vol. 3:2. [Cross Ref]

50. Sojka RE, Upchurch DR. 1999. Reservations regarding the soil quality concept. Soil Science Society of America Journal. Vol. 63:1039–1054. [Cross Ref]

51. Song B, Niu S, Zhang Z, Yang H, Li L, Wan S. 2012. Light and heavy fractions of soil organic matter in response to climate warming and increased precipitation in a temperate steppe. PLoS One. Vol. 7:e33217. [Cross Ref]

52. Stewart RD, Jian J, Gyawali AJ, Thomason WE, Badgley BD, Reiter MS, Strickland MS. 2018. What we talk about when we talk about soil health. Agricultural & Environmental Letters. Vol. 3:180033. [Cross Ref]

53. Sullivan CA, Finn JA, Ó hUallacháin D, Green S, Matin S, Meredith D, Clifford B, Moran J. 2017. The development of a national typology for high nature value farmland in Ireland based on farm-scale characteristics. Land Use Policy. Vol. 67:401–414. [Cross Ref]

54. Tan Z, Lal R, Owens L, Izaurralde RC. 2007. Distribution of light and heavy fractions of soil organic carbon as related to land use and tillage practice. Soil and Tillage Research. Vol. 92:53–59. [Cross Ref]

55. van Bruggen AHC, Semenov AM. 2000. In search of biological indicators for soil health and disease suppression. Applied Soil Ecology. Vol. 15:13–24. [Cross Ref]

56. Vance ED, Brookes PC, Jenkinson DS. 1987. An extraction method for measuring soil microbial biomass C. Soil Biology and Biochemistry. Vol. 19:703–707. [Cross Ref]

57. Volpato A, Ahmed KS, Williams CD, Day MF, O’Hanlon A, Ruas S, Rotchés-Ribalta R, Mulkeen C, Ó hUallacháin D, Gormally MJ. 2020. Using Malaise traps to assess aculeate Hymenoptera associated with farmland linear habitats across a range of farming intensities. Insect Conservation and Diversity. Vol. 13:229–238. [Cross Ref]

58. Weil R, Islam K, Stine M, Gruver J, Samson-Liebig S. 2003. Estimating active carbon for soil quality assessment: a simplified method for laboratory and field use. American Journal of Alternative Agriculture. Vol. 18:3–17. [Cross Ref]