Effects of two decades of organic and mineral fertilization of arable crops on earthworms and standardized litter decomposition

The experiment was established in 1991 at Ritzlhof near Linz in Upper Austria (N 48°11'18.42; E 14°15'15.12; altitude 280 m a.s.l.) as described in Ros et al. (2006a, b), Tatzber et al. (2015), and Lehtinen et al. (2017). The soil is classified as a loamy silt Cambisol with 13.6% sand, 69% silt, and 17.4% clay. On an average, soils on this site have a pH (CaCl₂) of 6.9, 1.28% of organic C, and 0.14 % of total N in 0–25 cm (Tatzber et al., 2015). The climate is temperate with mean long-term annual temperature of 8.5°C and a mean long-term annual precipitation of 753 mm. Field plots (each 6 m × 5 m) included 12 fertilization treatments with 4 replicates, a detailed description of the experiment is given in Lehtinen et al. (2017) and Tables 1 and 2. The plots were cultivated according to good farming practice using a traditional crop rotation (Table 2). The treatments consisted of a control without nitrogen fertilization (0 kg N), treatments fertilized with mineral N (40 kg N ha⁻¹ year⁻¹, 80 kg N ha⁻¹ year⁻¹, and 120 kg N ha⁻¹ year⁻¹ as calcium-ammonium-nitrate, CAN) and treatments with organic amendments—urban organic waste compost (OWC), green waste compost (GWC), cattle manure compost (MC), and municipality sewage sludge compost (SSC)—each treatment corresponding 175 kg N ha⁻¹ year⁻¹. Further variants consisted of the four compost amendments plus 80 kg mineral N (CAN) ha⁻¹ (Table 1). The average C inputs (1991–2011) with compost applications were 2,277 kg C ha⁻¹ year⁻¹ with OWC, 2,190 kg C ha⁻¹ year⁻¹ with GWC, 2,567 kg C ha⁻¹ year⁻¹ with MC, and 4,081 kg C ha⁻¹ year⁻¹ with SSC. Fertilization was applied annually with the exception of pea (no fertilization at all) and compost, no application took place for the cropping season 2004, 2008, 2010, 2012, and 2014 (Table 2). The study is of long term, and a set of parameters (abundance and activity of earthworms, litter decomposition and winter barley growth and yield) were obtained in 2014, while soil parameters were measured in 2012. In the cropping season 2014, all plots were ploughed (25 cm deep) with subsequent rotary harrow treatment 1 day before sowing. Winter barley (Hordeum vulgare L. variety Christelle) with a seeding rate of 270 kernels m⁻² was sown using a Reform seed drill with 2.5 m sowing wide on September 9, 2013. A uniform N fertilization of all long-term treatments was applied but compost was not applied (the last compost was given in October 2012 before sowing of winter wheat, see Table 2). The mineral N fertilization was carried out on February 25, 2014, in the developmental stage BBCH 25 (Zadoks et al., 1974) using Linzer plus Volldünger resulting in 40 kg N ha⁻¹ N, 16 kg P ha⁻¹, 16 kg K ha⁻¹, 6 kg MgO ha⁻¹, and 8 kg sulfur ha⁻¹. On April 9, 2014, at BBCH 35, the plots were treated with Linzer CAN 27% N at a rate of 49 kg N ha⁻¹.

The results of biological and yield parameters were complemented with earlier results of chemical soil parameters. Soil samples were taken in August 2012 in 0–25 cm soil depth. On each plot, 10 subsamples were taken with a single gouge auger (cores of 30 mm in diameter) that were mixed and stored in plastic bags. Before the analyses, the soil samples were air dried and sieved using <2 mm mesh. Soil pH (CaCl₂) was determined electrochemically (pH/mV Pocket Meter pH 340i, WTW, Weilheim, Germany) in 0.01 M CaCl₂ at a soil-to-solution ratio of 1:2.5 (ÖNORM L1083). Soil organic carbon was analyzed by dry combustion using a LECO RC-612 TruMac CN (LECO Corp., St. Joseph, MI, USA) at 650°C (ÖNORM L1080). Total nitrogen (N_t) was determined according to ÖNORM EN 16168 by elemental analysis using a CNS 2000 SGA-410-06 at 1,250°C. Nitrogen (N) mineralization potential on dried soils was measured by the anaerobic incubation method (Keeney, 1982), modified according to Kandeler (1993). Extractable P and K were determined with CAL (calcium acetate/lactate) according to Schüller (1969) and ÖNORM L1087 with spectral photometer (P, using molybdenum blue method) and flame photometer (K), using a Segmented flow Analyzer SAN (Skalar)). Magnesium (Mg) was analyzed by 0.0125 M CaCl₂ (method SCHACHTSCHABEL, ÖNORM L1093) using a flame atomic absorption spectrometer, Thermo Fisher iCE 3500.

To determine the barley biomass production of all 48 plots, 1 m² was harvested 5 cm above the soil surface by hand using a pruning clipper. The barley was harvested on 1 July at BBCH 89. The stem-to-ear ratio was assessed on 20 randomly chosen plants from every plot that were cut close to the ground at harvest time. From every plant, the length of the ear and the length of the stem were measured. After weighing the biomass, the ears (spikes) were cut off with a pair of scissors and threshed by a manual threshing machine. One sample was taken from every plot to calculate the thousand kernel weight (TKG) at a humidity rate of 14% (Contador Pfeuffer machine, Pfeuffer GmbH, Germany, Kitzingen).

A two-way analysis of variance (ANOVA) was used to assess the influence of the factor fertilization class at four levels: no fertilization/control (n = 4), mineral fertilization (n = 12), organic fertilization (n = 16), and organic-mineral fertilization (n = 16). The dependent variables were earthworm activity (cumulative surface cast weight, average cast weight, cumulative number of surface casts, average cast number), abundance and biomass of earthworms, litter decomposition (decomposition rate and stabilization factor after 28 and 56 days), and barley growth and yield (biomass of 1 m², TKG, length of stem and ear, and stem-toear ratio). Mean values were computed; Tukey’s post-hoc test was used for mean comparisons. Correlations between variables were calculated with the Spearman correlation coefficient. All statistical analyses were performed using SPSS (version 20, IBM SPSS Statistics, USA).

Activity, abundance, and biomass of earthworms are shown in Table 3. The cumulative surface cast weight was significantly affected by the fertilization class. The highest cumulative cast weight was found under long-term organic-mineral fertilization, with significant differences compared to the control and only mineral fertilization. The average cast weight per sampling was also highest under combined mineral and organic fertilization, revealing significant differences only compared to mineral fertilization. The cumulative number of surface casts showed similar results. The highest amount of cumulative cast numbers occurred under organic-mineral fertilization, significantly lower ones under mineral fertilization. The average cast numbers were statistically higher under both organic and organic-mineral fertilization, compared to only mineral fertilization.

Total number of earthworms was not affected by the fertilization class. Organic-mineral fertilization resulted in the highest number of earthworms, the lowest number of earthworms occurred in the control. Earthworm biomass was unaffected by fertilization: the control showed the lowest biomass and the organic-mineral fertilization showed the highest biomass. Average weight per earthworm was unaffected by treatments but tended to be highest under organic-mineral fertilization and lowest in the control.

Litter decomposition rate (k) and stabilization factor (S) after 28 and 56 days, respectively, are shown in Table 4. Litter decomposition rate (k) after 28 days was affected by the fertilization class. It was significantly higher in the control compared to the organic-mineral class and the mineral class. The decomposition rate in the organic fertilization class was not different from control treatments. Litter decomposition rate after 56 days was not affected by the fertilization class.

The stabilization factor (S) after 28 days was affected by the fertilization class. The stabilization factor in the control was significantly lower compared to the organically fertilized class and the organic-mineral fertilized treatment. The mineral fertilization class showed an intermediate stabilization factor. The stabilization factor (S) after 56 days did not reveal significant differences between the fertilization classes.

After long-term organic fertilization with composts and combined organic-mineral fertilization, the pH values, SOC, N_t contents, and the N mineralization potential, as well as plant available P-CAL and K-CAL, increased significantly compared with only mineral fertilization (Table 5). Earthworm activities (cumulative and average cast weights) showed highly significant (p<0.01) positive correlations with SOC, N_t, the N mineralization potential, and plant available P-CAL and K-CAL (Table 6). The cumulative and average cast numbers were positively influenced by soil pH, SOC, N_t, N mineralization potential, and P-CAL. The number of earthworms (living and total) was significantly positively affected by pH, SOC, N_t, and the P-CAL content. The same soil properties (except for N_t) had a stimulating influence on the total earthworm biomass. Furthermore, a positive correlation between SOC and P-CAL and the barley stem length was observed. Barley biomass increased with higher SOC and K-CAL and was positively influenced by N_t and the N mineralization potential. The litter decomposition rate (k) after 28 days correlated positively with pH and P-CAL and negatively with plant available soil Mg-Sch. The stabilization factor after 56 days correlated negatively with pH.

The results of barley growth and yields are shown in Table 7. Neither barley biomass yields nor the TKG showed significant differences between the fertilization classes. Only the stem length was significantly higher with organic-mineral fertilization compared to the control. No significant differences occurred for the ear length and the stem-to-ear ratio.

A combined long-term application of composts and mineral N fertilizers resulted in higher earthworm activity compared to solely mineral fertilization. In contrast, abundance and biomass of earthworms were unaffected by fertilization classes. However, in tendency, the highest values always occurred in the treatments with organic-mineral fertilization. This is in accordance with Estevez et al. (1996) who claimed that solid cattle manure as an available food source can improve abundance of earthworms in soil. These authors also found that the effects of mineral fertilization on earthworms can vary. For example, Edwards (1983) observed that the application of mineral fertilizer on clay soils had no significant effect on the earthworm population. Edwards and Lofty (1982) reported that treatments receiving both inorganic and organic N had the largest populations of earthworms. The tendency of the increase in the abundance of earthworms in the treatments receiving organic amendments can be linked with the higher SOC contents in these treatments (Lehtinen et al., 2017). An addition of organic matter is believed to be one of the major management variables affecting abundance of earthworms (Leroy et al., 2008; Amossé et al., 2013). Our current results are in line with results of other long-term field experiments, showing that organic amendments increase both earthworm number and biomass compared to the sole use of mineral fertilizers (Jouquet and Doan, 2014; D´Hose et al., 2018). In another field experiment with spring wheat, which received no farmyard manure, earthworms produced 20% less surface casts than earthworms in treatments with amendments of farmyard manure for more than 9 years (Zaller and Köpke, 2004). In a long-term study from 1994 to 2004, Riley et al. (2008) stated that the incorporation of large amounts of organic matter and a longer ley period may explain the high density of earthworm channels in these systems. The improved development of earthworms with compost was also confirmed by the increased weight of earthworm casts in a study by Doan et al. (2013). In our study, compost plus inorganic fertilization had a significantly higher effect on earthworm activities than mineral N fertilization alone. Therefore, the addition of organic matter appears advisable in order to obtain maximum benefits from mineral fertilizers (Tiwari, 1993). The effects of compost application were still detectable in our study 18 months after the last application.

We studied decomposition and stabilization rate by the TBI method (Keuskamp et al., 2013) and found the highest decomposition after 28 days in the control and a significant lower decomposition under solely mineral and organic-mineral fertilization. This is in contrast to other findings showing increased decomposition in organic amendments with manure (Zaller and Köpke, 2004). Nitrogen fertilization could have reduced soil microbes (Thirukkumaran and Parkinson, 2000) by soil acidification and plant community changes (Zeng et al., 2016). In contrast, organic fertilization enhanced the microbial activity in the soil (Pokorná-Kozová and Novák, 1975). In the data that obtained after 56 days, no statistical differences were observed, which is partly in agreement with previous pot experiments (Zaller et al., 2016; Van Hoesel et al., 2017) in which litter decomposition in different agricultural management practices were compared. In a longterm field experiment, like in our case, the microorganisms might be well adapted to the different fertilization classes and thus decompose litter in a similar manner; however, we would have expected differences between the treatments.

The increase in pH, SOC, and plant available nutrients such as P-CAL and K-CAL resulting in a better fertility status of compost fertilized soils was described in Lehtinen et al. (2017). Furthermore, the authors have already addressed the danger of applying P and K in excess with long-term compost amendments at allowed N rates (corresponding to 175 kg ha⁻¹ year⁻¹) and the necessity of monitoring plant nutrients, which was visible in the actual evaluations as well. However, the increase in pH, SOC, and soil nutrients had, at least partly, a stimulating effect on activity, abundance, and biomass of earthworms. Our results confirm the improved prediction of biomass yields with the easily measurable biological soil indicator “N mineralization potential” (Dersch et al., 2003). Decomposition rates after 28 days were positively correlated with soil pH and plant available P-CAL, indicating better living conditions for soil microbes under these conditions. The role of soil pH is well known in the literature (e.g., Leifeld et al., 2008). The TBI method excludes the influence of soil meso- and macrofauna on decomposition and only allows soil microorganisms to enter the 0.25-mm mesh (Keuskamp et al., 2013). The stabilization factor after 56 days was negatively correlated with soil pH, also indicating the importance of soil pH in carbon cycling in soils (Leifeld et al., 2008).

The significantly higher barley stem length under longterm organic-mineral fertilization reflected the effect of nitrogen as the most important nutrient for plants production. The reason why only significant differences in the stem length but not in the other investigated parameters (biomass yield, ear length, stem to ear ratio) were observed was probably the uniform cultivation in the experimental year 2014. Lehtinen et al. (2017) reported that yields of winter barley with solely compost application were comparable to 40 kg of mineral N fertilization and increased significantly with additional mineral fertilization. Sacco et al. (2015) argued that yield reductions after organic fertilization are related to lower soil nutrient availability, because nutrients are provided mainly after mineralization. According to Bedada et al. (2014), the compost and mineral fertilizer treatments exceeded all the other treatments (compost, mineral fertilizer, control). The addition of either compost alone or in combination with NP fertilizers improved soil properties and crop productivity compared with the control and the treatments with mineral fertilizers. Therefore, compost addition can serve as a supplement to mineral fertilizer use and reduce the dependency on those fertilizers.