Feasibility of strip-tillage for field grown vegetables

Eignung des Strip-Tillage Verfahrens für den Feldgemüsebau

Material and Methods

Fig. 1. Monthly average temperature and precipitation in 2011 [a] and 2012 [b] at Ihinger Hof

Monatliche mittlere Temperaturen und Niederschläge 2011 [a] und 2012 [b] am Ihinger Hof

White cabbage was transplanted with a total-control transplanter from Checchi & Magli (Type TRIUM, Budrio, Italy). The machine was modified by installing row-cleaners in front of special blades for mulch-planting systems. The row-cleaners clear the planting rows from straw residue and large soil clods to achieve an exact planting result.

Both the strip-tiller and planting machine were equipped with a device for placed nitrogen fertilizer application during strip-tillage and transplanting. The fertilizer granules were placed 5 cm deep exactly in the planting row. Furthermore, for the strip-tillage and the transplant processes, an RTK-GPS auto-guidance system was used with the precision of 2.5 cm to make sure that plantlets and fertilizer was placed in the targeted position.

Spring time nitrogen fertilization was based upon SMN target values for lettuce and white cabbage. For lettuce 150 kg N ha^–1 and for white cabbage 270 kg N ha^–1 were applied. The 2012 nitrogen fertilizer rate was split for placed nitrogen application (ST_Int_pN). The first half was applied with the second strip-tillage pass in spring and the second nitrogen rate was applied while transplanting white cabbage. For MP and the non-intensive ST treatment, the nitrogen fertilizer was applied broadcast one day prior to transplanting. For cabbage, potassium and magnesium sulfate were applied via broadcast if necessary, in accordance with official fertilizer recommendations.

An Eijkelkamp Penetrologger was used to measure penetration resistance down to a depth of 50 cm on 15 May 2011 in all MP_lettuce and ST_lettuce plots with ten measurements taken per plot. For ST_lettuce, measurement samples were taken again both in the planting rows (IR) and between the planting rows (BR).

Disturbed soil samples were taken on five dates between June and August 2011 during white cabbage cultivation to determine the gravimetric water content. All MP and ST plots were sampled with a soil core sampler (1 cm inner diameter) at 0–10 cm, 10–30 cm and 30–40 cm depths. Strip-tillage sampling positions were in row (ST_IR) and between rows (ST_BR). Positions for sampling within each plot corresponded to the hill slope, with the components of the back slope, and the foot slope. In each plot, two samples were taken per position and depth. Soil samples were dried at 105°C for 24 h and gravimetric water content was calculated by the mass difference of wet soil and dry soil weight.

Plant-non-available water content was measured, corresponding to the bulk density, in one plot per treatment and at 10 cm, 30 cm and 40 cm depths (data not shown). The plant available water content [L m^–2] was calculated by multiplying the gravimetric water content by the soil horizon thickness and the bulk density. Plant-non-available water content was subtracted. Subsequently, the sum was calculated for all values from each depth within each treatment.

In 2011, lettuce and white cabbage were harvested after 55 days and 118 days, respectively. In 2012, white cabbage was harvested 98 days after transplanting.

Three field transects were harvested with 18 plants per plot. In total, 288 white cabbage plants were cut per year. Whole cabbage plants were harvested and weighed. Subsequently, all cover leaves were removed and the head fresh weight was determined.

T + PC + PC·T + R: P + Pos,

where fixed effects are given before and random effects are given after the colon, and interactions by a dot between the corresponding main effects. In 2012 different previous crops (PC) were replaced by different N fertilizer application systems (N). For soil mineral nitrogen, gravimetric water content and penetration resistance at different soil depths (D) and repeated core sampling at the same position (rPos) were assumed. The model is given by:

R + T + D + T·D: P·D + P·Pos·D + P·Pos·rPos·D,

again with fixed effects given before and random effects given after the colon, and interactions by a dot between the corresponding main effects. For bulk density, the model syntax was adjusted due to lack of field replicates.

For residual error effects of depth (D) a joint variance structure was assumed because of the existing autocorrelation effect of the different soil depths.

Different sampling dates of soil mineral nitrogen and gravimetric water content were separately analyzed from each other. For letter description, a multiple t-test was used only after finding significant differences via an F-test.

Results

Fig. 2. Cumulative soil loss [g m^–2] during 20 min of simulated rainfall (2 L min^–1) in [a] 2011 and [b] 2012. MP_lettuce: moldboard plowing with lettuce as previous crop, ST_lettuce: strip-tillage with lettuce as previous crop, MP: moldboard plowing, ST: strip-tillage, ST_Int_bN: intensive strip-tillage with broad­cast nitrogen fertilization. Identical letters show no significant differences, P < 0.05. Error bars indicate the standard errors of means.

Kumulativer Bodenabtrag [g m^–2] während 20minütigem simulierten Regen (2 L min^–1) 2011 [a] und 2012 [b]. MP_lettuce: Pflug, Vorfrucht Kopfsalat, ST_lettuce: Strip-Tillage, Vorfrucht Kopfsalat, MP: Pflug, ST: Strip-Tillage, ST_Int_bN: Intensives Strip-Tillage mit breit gestreuter N-Düngung. Gleiche Buchstaben zeigen nicht signifikante Unterschiede, P < 0,05. Fehlerbalken sind Standardfehler des Mit­telwerts.

The penetration resistance was proportional to bulk density. In the tilled soil layer, the highest penetration resistance was measured at a depth of 7 cm in ST_lettuce_BR (1.76 MPa; Fig. 3). The penetration resistance values in ST_lettuce_IR were lower from 0 cm to a depth of 14 cm than under MP_lettuce and ST_lettuce_BR. At 7 cm ST_lettuce_IR and ST_lettuce_BR differed significantly. Penetration resistance decreased in MP_lettuce and ST_lettuce_BR from 7 cm to 17 cm respectively, in ST_lettuce_IR from 7 cm to 14 cm followed by increased values down to 30 cm depth. From 16 cm to 50 cm, penetration resistance was not affected by tillage treatment, previous crop or sampling position.

Fig. 3. Penetration resistance under different tillage practices on 15 May 2011. MP_lettuce: moldboard plowing with previous crop lettuce, ST_lettuce_IR: strip-tillage with previous crop lettuce, sampling position in tilled rows, ST_lettuce_BR: strip-tillage with previous crop lettuce, sampling position between planting rows in untilled area. Test of significant differences were performed in 7, 16, 27, 33 and 45 cm soil depth. Data with the same letter are not significantly different, P < 0.05; no interactions of treatment and depth were detected. Horizontal bars indicate standard error of means.

Eindringwiderstand bei unterschiedlicher Bodenbearbeitung am 15. Mai 2011. MP_lettuce: Pflug, Vor­frucht Kopfsalat, ST_lettuce_IR: Strip-Tillage, Vorfrucht Kopfsalat, Probenahme in der Pflanzreihe, ST_lettuce_BR: Strip-Tillage, Vorfrucht Kopfsalat, Probenahme zwischen den Pflanzreihen. Mittelwert­vergleiche in 7, 16, 27, 33 und 45 cm Tiefe. Gleiche Buchstaben zeigen nicht signifikante Unterschiede, P < 0,05; es bestand keine Wechselwirkung Bodenbearbeitung × Tiefe. Fehlerbalken sind Standardfehler des Mittelwerts.

For the first three sampling dates there were no signifi­cant differences in soil water contents between the treatments at 30 cm depth. For the fourth and fifth sampling date, significantly higher gravimetric water content was measured in MP compared to ST_IR. Treatments and sampling position did not differ significantly at 40 cm depth, except on 25 July 2011. In MP, soil water content was significantly higher at 30 cm depth compared to 10 cm and 40 cm with the exception of 25 July 2011. In ST_IR and ST_BR soil water content increased from 10 cm to a depth of 30 cm and decreased from 30 cm to a depth of 40 cm. In deeper soil horizons (10–30 cm and 30–40 cm) differences in gravimetric water content were less noticeable between treatments.

Results of plant available water (Fig. 4) showed significantly lower values in ST_IR compared to MP and ST_BR. MP and ST_BR were not significantly different to each other. Plant available water content ranged between 11.9 L m^–2 and 43.1 L m^–2 during the cultivation period. Most consistent plant available water values were detected in ST_BR over the sampling period. Strong variations were measured in MP and ST_IR with differences in plant available water content of 16 L m^–2 in MP and 18 L m^–2 in ST_IR.

Fig. 4. Plant available water in the soil [L m^–2] grown with white cabbage and daily precipitation [mm] across 5 sampling dates in June and July 2011. Sampling was conducted down to 40 cm depth. MP: moldboard plowing, ST_IR: strip-tillage, sampling position in tilled rows, ST_BR: strip-tillage, sampling position between rows in untilled area. For individual sampling dates, data with same letter are not significantly different, P<0.05. Vertical bars indicate standard error of means.

Pflanzenverfügbares Bodenwasser [L m^–2] unter Weißkohl sowie tägliche Niederschläge [mm] zu 5 Pro­benahmeterminen im Juni und Juli 2011. Probenahme bis 40 cm Bodentiefe. MP: Pflug, ST_IR: Strip-Tillage, Probenahme in der Pflanzreihe, ST_BR: Strip-Tillage, Probenahme zwischen den Pflanzreihen. Gleiche Buchstaben zeigen nicht signifikante Unterschiede innerhalb jedes Probenahmetermines, P < 0,05. Fehlerbalken sind Standardfehler des Mittelwerts.

For white cabbage, in 2011 significant differences in average head weight were detected between MP_lettuce (1311 g) and MP (1476 g). There were no significant differences between ST, ST_lettuce and MP (Fig. 5a). In 2012, a significantly lower head weight was detected in MP (1622 g) than in ST (1850 g). Head weight of ST_Int_bN (1723 g) and ST_Int_pN (1684 g) did not differ significantly to MP and ST (Fig. 5b).

Fig. 5. Average head weight of white cabbage in 2011 [a] and 2012 [b] under different tillage systems, previous crop [a] and fertilizer application systems [b]. MP: moldboard plowing, MP_lettuce: moldboard plowing with previous crop lettuce, ST: strip tillage, ST_lettuce: strip-tillage with previous crop lettuce, ST_Int_bN: intensive strip-tillage with broadcast nitrogen fertilization, ST_Int_pN: intensive strip-tillage with placed nitrogen application. Values with the same letter are not significantly different, P < 0.05. Ver­tical bars indicate standard error of means.

Mittlere Kopfgewichte von Weißkohl 2011 [a] und 2012 [b] in Abhängigkeit vom Bodenbearbeitungs­verfahren und Vorfrucht [a] bzw. Düngerausbringung [b]. MP: Pflug, MP_lettuce: Pflug, Vorfrucht Kopfsalat, ST: Strip-Tillage, ST_lettuce: Strip-Tillage, Vorfrucht Kopfsalat, ST_Int_bN: Intensives Strip-Tillage mit breit gestreuter N-Düngung, ST_Int_pN: Intensives Strip-Tillage mit platzierter N-Dün­gung. Gleiche Buchstaben zeigen nicht signifikante Unterschiede, P < 0,05. Fehlerbalken sind Standard­fehler des Mittelwerts.

The artificial rainfall simulation showed a high erosion protective potential of strip-tillage in both the lettuce and white cabbage cultivation.

The amount of soil loss in the current study was similar to investigations of different tillage systems in Saxony (Germany), which were conducted with the same rainfall simulator and the same level of rainfall intensity. In these investigations, cumulative soil losses in sugar beet, barley and winter wheat after 20 minutes of irrigation were up to 270 g m^–2 in the conventional tillage treatment compared to 100 g m^–2 in the conservation tillage plots (Nitzsche and Zimmerling, 2004). In general, rainfall simulators have very small working areas. Approximately 50% of the 229 simulators described by Cerdà (1999) have an irrigation area of less than 1.5 m². Such simulators are not suitable to reproduce soil erosion processes which are scale dependent, for example overland flow or rill erosion, as they required larger areas (Gómez and Nearing, 2005). To conclude, small scale rainfall simulators are appropriate to establish the effect of soil properties, splash erosion, or the erosion potential of different tillage practices, as in this study with strip-tillage and moldboard plowing. The applied amount of 40 liters per hour reflects a rainfall event which occurs at 20 to 50year intervals (Schmidt et al., 1996).

In general, lower soil erosion risk under conservation tillage compared to conventional tillage was observed in several studies (Blevins and Frye, 1993; Jin et al., 2008; Delaune and Sij, 2012). The key factor of erosion control in conservation tillage systems is the surface covering by straw or mulch (in the current study, 60% of the soil surface was covered), which reduces water velocity and rain drop impact and results in reduced runoff.

Corresponding to the results of the current study, in other investigations higher bulk densities were observed between rows than within rows in strip-tillage treatment with vegetable rotations (Overstreet and Hoyt, 2008). In short-term field experiments, such as the current study, bulk density in conservation tillage systems is often higher compared to conventional tillage (Al-Kaisi et al., 2005; Puget and Lal, 2005). In contrast, in long-term reduced tillage experiments, the bulk density was similar or lower than in conventional tillage fields (Tebrügge and Düring, 1999; Dolan et al., 2006). Lower bulk density is often caused by higher soil organic matter contents because the particle density of soil organic matter is lower than that of mineral soil. Soil organic matter increased in conservation tillage systems, which was due to the crop residues on the soil surface being turned over by micro-organisms into organic matter over time, together with soil particles forming stable aggregates. These factors in conservation tillage systems contribute towards preventing soil losses (Fawcett and Caruana, 2001) and improving water infiltration (Jabro et al., 2011). Evidence of higher water infiltration rates and greater water-holding capacity in this current study helps to explain the later start of soil loss in the strip-tillage plots when compared to the moldboard plowed plots.

Penetration resistance is the main decisive factor controlling root growth and it is a factor for determining the structure and quality of a soil (Tebrügge and Düring, 1999). Similar to the results of bulk density, top soil penetration resistance is higher under conservation tillage than under conventional tillage. This is consistent with most other studies (Vetsch and Randall, 2002; Licht and Al-Kaisi, 2005). In general, penetration resistance increases with depth, whereas the tillage treatment is less influential as depth increases (Erbach et al., 1992). A threshold for critical penetration resistance values for impeded root growth and reduced yields is given between 2.5–3.0 MPa (Taylor and Gardner, 1963). In summary, for our study penetration resistance did not exceed this critical value.

In assuming that strip-tillage between rows (ST_BR) can be compared to no-till, the results of higher moisture content and higher plant available water content between the rows in the current study are consistent with studies which detected higher top soil moisture contents in no-till treatments when compared to conventional treatments (Franzluebbers et al., 1995; Rasmussen, 1999). A possible reason could be a reduced evaporation rate and again an increased infiltration due to the soil being covered with straw residues under conservation tillage techniques (Smika and Unger, 1986; Jones et al., 1994; Licht and Al-Kaisi, 2005).

In the strip-tillage treatment, the low available water content in the tilled zone of ST_IR could be dependent on a variable pore size distribution in the tilled zone. A higher macropore volume was observed by Hussain et al. (1998) in the tilled area within the strip-tillage treatment in the top soil, but these were generally not well connected with subsoil macropores. It might be that in ST_BR, comparable to no-till treatments, the macropores are fewer than in moldboard plowed treatments but they are more homogenously distributed across the top soil and subsoil layers. Consequently, conservation tillage treatments have a larger volume of storage pores that lead to higher water infiltration and plant available water content in no-till treatments compared to conventional tillage treatments (Shukla et al., 2003).

The cabbage yield under strip-tillage of the current study was found to be equal or even higher. This is in contrast to a study examining strip-tillage treatments with different mulches which showed that cabbage yields in strip-tillage treatments were lower than in the conventional tillage treatment using the moldboard plow and disk management (Hoyt, 1999). Decreased yields are often associated with increased weed population, lower soil temperature in spring and no uniform seedbed preparation for guaranteed crop establishment, and in some cases slower nitrogen mineralization (Tiarks, 1977; Triplett Jr. and Dick, 2008). It is possible that the highly modernized techniques and the technical modifications of strip-tillage and the planting equipment used within the present study were responsible for the high yield potential. In another study, sugar beet yield in strip-tillage plots was similar to conventional treatments after 5 years of development and modification of machines and techniques (Evans et al., 2010).