Background & Summary

Almost 40% of the world’s terrestrial surface area is managed as agricultural land1. Those fields yield 95% of the global food either directly or indirectly, underlining the essential role of agricultural soils for human nutrition2. The exponentially growing world population and the increasing problems associated with climate change will put pressures on future global food sovereignty3,4. Moreover, biomass for energy and material use is also being increasingly produced on the limited resource soil. This calls for managing agricultural land in a sustainable manner to maintain soil fertility and feedstock supply for future generations.

In terrestrial ecosystems, a wide range of ecological processes and patterns extend over long periods of time and large spatial scales5. Through its multifactorial approach to agriculture6, long-term ecological research (LTER) can provide insights into those chemical, physical and biological processes that become apparent only after years or even decades. In contrast, short-term experiments offer insights into how a system is controlled at a specific time and place by a set of factors (for instance: initial limiting factors and their interactions). Importantly, however, agricultural systems are the summation of multiple components operating at various time scales. The initial response curves of the entire system or single components do not automatically show the direction of the long-term changes, for instance changes in soil organic carbon stocks5,7. Thus, LTER provides systemic insights into biological, chemical, hydrological and biophysical processes under temporal dynamics. The scientific knowledge gained is crucial, especially when LTER is accompanied by a factorial approach such as gradient studies or specific treatments that manipulate specific factors and measure key processes5,8.

Historically, tillage was used for weed control and to prepare the soil for subsequent planting9. The type of soil cultivation significantly influences the soil biosphere. In contrast to tillage, no- tillage (the practice of direct seeding) can reduce soil erosion10,11, improve nutrient cycling12, enhance the water infiltration capacity of the soil and, by providing an adequate cover crop, reduce evaporation in semi-arid and arid climates13,14,15. At the same time, no- tillage practices can result in lower crop yields than conventional ploughing16,17. The impact of no-tillage as a climate mitigation approach remains uncertain8,18. Namely, the soil can become more anaerobic under reduced tillage practices, which in turn can promote the production of N2O16,19. Conventional tillage damages the aggregates of the soil surface20, making the soil prone to soil erosion21. Furthermore, tillage enhances the mineralization of soil organic carbon (SOC), reducing the SOC stocks22, although the literature on subsoil effects is still scarce8.

This paper describes selected years (2018–2022) in an agricultural data set containing data on a) agricultural management, b) primary production (crop yield), c) soil parameters (organic carbon, nitrogen concentration) collected from the three LTER experimental sites. The three LTER sites are combined in a linked group, the LTER “Marchfeld” Cluster of the Pannonian region in Austria, with diverse crop rotations and different tillage systems. Two of the LTER study sites (Gross-Enzersdorf and Fuchsenbigl) are conventionally managed, while the third one (Rutzendorf; MUBIL) is organically operated. The tillage treatments consist of a reduced (soil depth 10–15 cm) and a mouldboard ploughing (25–30 cm) system at all sites. On the conventionally managed study sites, further gradations of tillage, namely no till and minimum tillage (5–8 cm), are implemented. The soil parameters are available for various soil depths, usually in 10 cm (MUBIL: 15 cm) increments up to 30, 50 or 100 cm for MUBIL, GE and FB, respectively.

Over five years (2018–2022), substantial records were stored and archived in the data set “Cluster Marchfeld”. The data set is available at Zenodo23 (https://doi.org/10.5281/zenodo.15212569).

Methods

Field study sites

The three long-term experimental field sites are located in the Marchfeld region, which is part of the Pannonian basin, Austria (Fig. 1). Overall, the Marchfeld is one of the most important production areas for arable farming, including vegetables, in Central Europe24. This area is among the driest in Central Europe25 and is characterized by heightened wind erosion26, elevated nitrate concentrations in the groundwater27 and a limited amount of landscape elements28. A detailed description of the region “Marchfeld” is given in Guarini29. The socio- ecological trajectories are available at district level and the data were collected from the district “Gänserndorf”, because the Marchfeld region is part of that district. Furthermore, the three LTER sites have the most common soil types in the district Gänserndorf, i.e. Chernozem 50% and Phaeozem 12%30.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Location of field study sites within the Marchfeld region in Austria.

Fuchsenbigl (FB)

The tillage experiment, in Fuchsenbigl (FB, Marchfeld, AUSTRIA), was instigated in 1988 to evaluate the impact of tillage on soil physicochemical and biological properties as well as crop yields (Table 1). The following treatments were tested: A) minimum tillage with a rotary driller to a soil depth of 5–8 cm; B) reduced tillage with a cultivator to a soil depth of 15–20 cm; and C) conventional tillage with mouldboard ploughing to a depth of 25–30 cm31. The experiment consisted of experimental units with a plot size of 720 m2 (l × w: 60 m × 12 m). The experimental set-up is designed in a completely randomized block design with three replicates. On this study site, an open crop rotation is cultivated with the following most common crops: winter wheat (Triticum aestivum L.), soybean (Glycine max L.), winter barley (Hordeum vulgare L.) and winter triticale (X Triticosecale Wittmack), millet (Sorghum bicolor L.). A detailed description of the experiment design can be found in31,32.

Table 1 Overview of the three long-term experimental field sites.
Full size table

Organic farming Trial (“MUBIL”)

With its conversion to organic farming, the long-term field monitoring “MUBIL” (“Monitoring der Auswirkungen einer Umstellung auf den Biologischen Landbau”) was founded in 2003 (Table 1). The MUBIL trial is located in Rutzendorf (Marchfeld, AUSTRIA) and managed by the Institute of Organic Farming, University of Natural Resources and Life Sciences, Vienna (BOKU University). The soil is classified as a Calcaric Phaeozem33 with a soil pHCaCl2 of 7.634 Table 1.

In 2003, an eight-year crop rotation was introduced at the MUBIL trial with the following sequence: 1st Year: Lucerne (Medicago sativa L.) 2nd Year: Lucerne; 3rd Year: Winter wheat (Triticum aestivum L.) + catch crop; 4th Year: Grain maize (Zea mays L.) 5th Year: Spring barley (Hordeum vulgare L.) + catch crop; 6th Year: Grain pea (Pisum sativum L.) + catch crop; 7th Year: Winter wheat (Triticum aestivum L.) + catch crop; 8th Year: Winter rye (Secale cereale L.). The field plot trail was set up in a two-factorial, completely randomized block design with four replicates. The experiment consisted of experimental units with a plot size of 270 m² (15 m × 18 m). Four organic fertilization systems have been tested in the trial since 2003. In 2016, a new soil tillage trial was instigated at the MUBIL site in Rutzendorf. Prior to starting the experiment, the homogeneity of the soil from the whole site was examined and a medium-quality soil was selected for the experiment.

Two tillage treatments are tested in one organic fertilization system: A) one is ploughed with a mouldboard plough to a soil depth of 25–30 cm, while B) the other half is managed through reduced tillage by using a cultivator to a soil depth of 10–15 cm. All plots are fertilized by mulching Lucerne and/or catch crops.

Experimental farm gross- enzersdorf (GE)

The third long-term experimental field trail is located in Gross-Enzersdorf (Marchfeld, AUSTRIA, Table 1). In 1996, a soil tillage trial with the following treatments was instigated: A) no- tillage treatment: direct seeding in un-tilled soil with a disc drill without removing the previous crop residues. B) minimal tillage with a wing share cultivator to a soil depth of 8–10 cm; C) reduced tillage with a wing share cultivator (soil depth: 20–25 cm) and every four years the soil was tilled with a subsoiler to a depth of 35 cm; thus the crop residues remain only partly on the soil surface; and D) mouldboard ploughing to a depth of 25–30 cm, which implies incorporating the residues into the soil35.

The experiment was set up in a split plot design with four replicates. Thereby, the factor tillage was attributed to main plots (48 m × 40 m), whereas the factor crop rotation was assigned to subplots (24 × 40 m). The second factor, crop rotation, consisted of two levels: Treatment A) a four-year crop rotation with maize (Zea mays L.), winter wheat (Triticum aestivum L.), sugar beet (Beta vulgaris L.) and winter wheat (Triticum aestivum L.). And B) a four-year crop rotation with winter wheat (Triticum aestivum L.), soybean (Glycine max Merr.), winter wheat (Triticum aestivum L.), and oilseed rape (Brassica napus L.). The non-harvested crop residues remain on the field. In order to meet the nutritional requirements of main crops, the experimental site was fertilized according to good agricultural practices as indicated in the Austrian Guidelines. The whole experimental design is discussed in the following publications35,36,37:

Agricultural management

All applied agricultural management practices were documented for 5 years, from 2018 to 2022. Mandatory data on management events were sowing (either main crop or cover crops), fertilization (type and quantity), harvest with crop name, tillage, mowing, integrated plant protection (type and quantity) and irrigation (if applied). Each activity and its associated device were described in detail. One aspect of it, the crop sequence of the three LTER sites for the years 2018–2022 is given in Table 2. The detailed list of agricultural management is given in the repository23 (https://doi.org/10.5281/zenodo.15212569).

Table 2 Plan of crop cultivation (2018–2022).
Full size table

Sampling

Grain yield and straw of the main crops were harvested and the dry matter content of the sampled materials was analysed. Every year, soil samples were taken annually in different soil depths. Soil samples were taken at the beginning of vegetation in spring (FB: end of February/beginning of March; MUBIL/GE: March/ April) and plant/crop samples were taken at the time of harvest.

Analytical measurements

Twenty-three different analytical measurements were performed on all three LTER sites (Table 3). Some of those are project-specific (indicated by their year of analysis in the Table 3), but most of them have been applied on LTER sites since their year of origin. For our data set, we selected those years in which the chemical parameters were examined with the same analysis at the same accredited National reference laboratory in Austria (AGES). Thereby, the “total” and “systematic” measurement error variances are reduced. This is important, since both the analysis and performance of the laboratory could have an influence on the result itself38. Considering that, we have selected total organic carbon (TOC) using the dry combustion method39, total nitrogen using the40 and the C/N ratio of those analysis. Those soil parameters are available in various soil depths, usually in 10 cm (or 15 cm MUBIL) increments up to 30, 50 or 100 cm for MUBIL, GE and FB, respectively. The complete data set is available at Zenodo23 (https://doi.org/10.5281/zenodo.15212569).

Table 3 All analyzed soil parameters.
Full size table

Data Records

The data set of the “ClusterMarchfeld” is online available in csv (Comma Separated Values) via Zenodo23 (https://doi.org/10.5281/zenodo.15212569). The data set has been created using Microsoft Access 201941. In total, the data set consists of 1153 records archived long- term data from 3 LTER sites. The data set consists of four csv Files: a) DataDescription (explains the headings of the csv files); b) Agricultural Management (offers Metadata to the experimental sites); c) Data Crop yield (provides yield data); d) Data Soil parameters (provides qualitative data on selected soil parameters).

The different sites are distinguished by their name (column: “Site_Name”: Fuchsenbigl; Gross-Enzersdorf; Rutzendorf), their farm management (column: “Farm_Category” either “organic” or “conventional” farming) as well as their respective crop rotation is shown (column: “Crop_Name”). The treatment tillage is given by the column “Tillage_Treatment” with the following gradations: “No till”; “Minimum tillage”; “Reduced tillage” or “Conventional tillage”. Soil parameters are available for various soil depths and it’s assigned soil layer (a range in cm; Column “Soil_Horizon”). The analyzed parameters are provided as “Crop_Yield_Dry”, “Total_Nitrogen”, “Total_Organic_carbon” or “C_N” ratio. Their corresponding units as well as their method of analysis are given through column “.._Unit” or “.._Method”, respectively (Table 4).

Table 4 Description of the provided data files at Zenodo23 (https://doi.org/10.5281/zenodo.15212569), including column name, explanation as well as its range or group.
Full size table

The data of all applied agricultural management practices follows the following structure: column “Site_Name”, “Farm_Category” and “Management_Type” (Divided into the type of agricultural management such as Tillage, Seeding; Harvest, Plant Protection, Irrigation or Fertilization). General agricultural practices are applied to the entire trial, while the specific factors namely tillage are only applied to subplots. This is indicated with the column “Experimental_Unit” (either whole or subplot). Tillage practices are further divided into their assigned factor by the column “Tillage_Treatment”. All applied agricultural management practices are presented by the date (column “Management_Date”) and their description (column “Management_Description”). This data set is available at Zenodo23 (https://doi.org/10.5281/zenodo.15212569).

Technical Validation

The quality of the data is ensured on four levels:

  1. A).

    LTER study site. Prior to starting the experiment, the homogeneity of fields was examined by measuring physical and chemical soil parameters. Compliance with the trial design, sample collection and all agricultural management measures were monitored and organized by the respective trial site managers. The collection of soil samples followed the guidance for sampling and storage42. The sampling campaign was conducted by a regularly trained sampling team.

  2. B).

    Laboratory. The efficiency and effectiveness of the laboratory is regularly ensured by measuring reference materials, standard solutions, laboratory replicates, and by participating in interlaboratory comparisons. The maximum allowed relative standard deviation between replicates was set to 5%.

  3. C).

    Data collection. In custom-made data templates, the data have been collected in an iterative manner. In those data templates, experiment names, treatment names, replicate number, observed year, measured value, units and methods are predefined to reduce the susceptibility to errors during data entry. Furthermore, each individual value was plotted to detect possible errors in the data input as well as to identify and correct deficiencies of the data input. An ANOVA was used to compare all measured parameters (crop yield as well as soil physical and chemical parameters) of each treatment. A pairwise comparison of treatments was performed with Tukey’s post-hoc tests (statistical significance set at p < 0.05). The completeness and quality of data was examined through data transfer templates, data verification checks, quality flagging and quality assessment exercises of the data set.

  4. D).

    Data set. The presented data set has been publicly made available at Zenodo23 (https://doi.org/10.5281/zenodo.15212569).