The research was carried out at Experimental Farm of Agronomic Institute (IAC) located at geographic coordinate of 21˚12'S e 47˚51'N and 645 m above sea level. The annual average rainfall is 1.425 mm, concentrated especially from October to March with maximum average temperature of 25˚C and minimum average temperature of 19˚C. According USDA soil taxonomy soil is classified as red Oxisol (Ferralsol—World Reference Base for Soil Resources) with clay texture (sand = 276 g kg of soil⁻¹, silt = 332 g kg of soil⁻¹ and clay = 530 g kg of soil⁻¹) and following chemical fertility: pH = 5.5, Ca = 39.5 mmolc dm⁻³, Mg = 15.1 mmolc dm⁻³, K = 4,2 mmolc dm⁻³, S-SO₄ = 10.5 mg dm⁻³, P = 132.6 mg dm⁻³, organic matter = 24.1 g dm⁻³, cationic exchange capacity (CEC) = 85.1 mmolc dm⁻³. The experimental site has 2.000 m² and it is under conservation agriculture since 2008.

The experimental treatments of the research were two tillage method, conventional, with subsoiling and harrowing (occasional tillage—just one tillage along of 16 years under no-tillage) compared with no-tillage (established along 16 years) interacting with two cover crops, millet (Pennisetum glaucum) and white lupin (Lupinus albus). Millet is a grass that has fast growth with high production of aboveground biomass besides has a fasciculate root which improves soil structure; white lupin is a leguminous that in addition to incorporating atmospheric nitrogen by biological fixation, it solubilizes phosphorus in the soil and also produce great aboveground biomass to lay down on soil surface. Both species are used as cover crops in the tropical Conservation Agriculture. Occasional tillage was characterized by conventional tillage performed before tomato transplanting and was done with subsoiling at 0.40 m of soil depth following by harrowing at 0.20 m of soil depth. For no-tillage the site was kept under conservation agriculture.

The experimental was a randomized block with split-plot design with four replicates, accounting 16 experimental unit with 4.0 m wide and 20.0 m long each one (80 m²).

Cover crop was seeded in 2023 March with no-till seeder (Vence Tudo enterprise, AS-7300 model) at 0.25 m row spacing. For millet it was used 20 kg ha⁻¹ of seed and 60 kg ha⁻¹ for white lupin. The cover crops grew freely without any agricultural input and irrigation. Along the cover crops grown the average of minimum and maximum temperatures was 14.9˚C and 28.3˚C respectively and the total precipitation was 217 mm. This weather condition was enough to good growth of cover crops and straw production on soil surface. The cover crops were killed with a pass of roller crimper in July 2023 (110 days after seeding).

Liming was done after killing cover crops with application of 1 t ha⁻¹ on the soil surface, based on the technical recommendation (Trani et al. 2018) [12] . Dolomitic lime (54% CaCO₃ and 45 MgCO₃) was incorporated into the soil at the conventional tillage site while in the no-tillage lime was kept on the surface, leaving the lime displace naturally in the soil profile. After killing cover crops, it was performed conventional tillage in the plots of the occasional tillage treatment.

After killing cover crops, tomato seedling (Hybrid with determinate growth from HM-Clause—HM2798) was transplanted in both, conventional tillage and no-tillage on the straw of millet and white lupin. For tomato transplanting the rows were open with the aid of the same no-till seeder (Vence Tudo enterprise, AS-7300 model) used for seeding cover crops, in both conventional and no-tillage treatments. The tomato spacing was 1.25 m between row and 0.5 m between plants.

The base fertilization of tomato was done with 40 kg ha⁻¹ of N, 450 kg ha⁻¹ of P₂O₅ and 150 kg ha⁻¹ of K₂O, and the tomato topdressing fertilization was done with 300 kg ha⁻¹ of N, 100 kg ha⁻¹ of P₂O₅ ha⁻¹ and 300 kg ha⁻¹ of K₂O, according Trani et al. (2018) [12] technical recommendation.

The tomato cropping was dripping localized irrigation and the moment and quantity of irrigation was determined by tensiometer and crop evapotranspiration, respectively (Marouelli et al., 1994) [13] . Biological control was performed for tomato cropping protection under integrated management of pests and diseases. Along the tomato cropping the average of minimum and maximum temperatures was 17.5˚C and 32.2˚C, respectively and the total precipitation was 162 mm, that was adequately for good performance of tomato cropping.

The experimental traits were evaluated by cover crop dry biomass yield, soil water storing, soil fertility from 0.0 to 0.20 m of depth, soil resistance to penetration and agronomic tomato performance.

The cover crops dry biomass production was evaluated before roller crimped pass to kill cover crops. In this way, plants of 1 m² of each experimental unit was sampled and drying in dry forced air oven at 110˚C by 72 hours until getting constant weigh. After, dry biomass production was estimated for Mg ha⁻¹.

The soil moisture content was measured by gravimetric method. Thus, the procedure was done by soil sampled in the steel ring with volume of 100 cm³ at 0.0 to 0.10 m and 0.10 to 0.20 m of soil profile at two points of each experimental unit and between row of tomato cropping, after a period of 15 days with no precipitation which provides adequate condition to evaluate soil moisture in the experimental treatments. Then, soil samples were weight and after dried in a dry forced air oven at 105˚C by 84 hours until getting constant weigh. Furthermore, it was calculated the percentage (%) of the soil moisture content by the following equation: SM = (Wu − Wd)/Wd*100, where SM = soil moisture content; Wu(g) = soil weight with moisture at moment of sampling; Wd(g) = soil weight after drying.

Soil fertility was evaluated in two soil profile extracts at the samples taken from 0.0 to 0.10 m and 0.10 to 0.20 m depth, in two points of each experimental unit at the end of tomato cropping. The samples were sent to laboratory for soil chemical analysis of Calcium (Ca), Magnesium (Mg), Potassium (K), Sulphur (S-SO₄), Phosphorus (P), organic matter (OM), pH, cationic exchange capacity (CEC), bases saturation (V%). Furthermore, it was analyzed the enzymes β Glucosidase and Arylsulfatase that are important indicators of soil biological fertility. β Glucosidase act on the organic matter decomposition and soil carbon cycle and Arylsulfatase act on the Sulphur cycle in the soil making available S inorganic to the plants. The enzymatic analysis was procedure as https://www.embrapa.br/en/busca-de-publicacoes/-/publicacao/1133109/tecnologia-bioas-uma-maneira-simples-e-eficiente-de-avaliar-a-saude-do-solo .

Soil resistance to penetration was measured with aid of resistance electronic meter (PLG1020—Falker enterprise) with measurement capacity until 0.60 m of soil depth, recording values in MPa at each 0.10 m. The measurements were taken in two points of each experimental unit in the row of tomato and at the end of tomato cropping.

The tomato growth was measured at 55 days after transplanting (DAT) by accounting the bunches quantity, fruit quantity, fresh fruit biomass and dry biomass of plant aerial parts (leaf, stems and fruit). For this, two plants from each experimental unit were sampled and evaluated.

The tomato root growth was measured by “mini-rhizotron” method at 55 DAT. Firstly, it was insert into the soil acrylic tubes at 0.40 m depth in one point of each experimental unit between two tomato plants. The dimension of the tubes is 60 mm diameter and 400 mm long. The root images capture was done with the Root Scanner CI-600 model from CID-Bioscience in two soil profile, 0.0 m to 0.20 and 0.20 to 0.40 m depth. Afterwards, the images were analyzed in the Winrhizotron® software to quantify total root number, root length (cm), root volume (cm³), root surface area (cm²).

The tomato yield was measured by accounting marketable fruits, biomass of marketable fresh fruit and estimative of yield in Mg ha⁻¹. For this, fruits from six plants of each experimental unit were harvested, classified and recorded for tomato yield accounting.

The statistical analysis was performed by PROC MIXED of the SAS® program (Littell et al., 2006) [14] . Before the data were subjected to homoscedasticity, normality and outlier tests and then was running the variance analysis (ANOVA) followed by treatments average comparison by PDIFF test at 5% of probability.

At the end of cover crops growth, millet yielded 4.8 Mg ha⁻¹ of dry biomass while white lupin yielded 6.6 Mg ha⁻¹ of dry biomass. This yield of dry biomass from both cover crops was adequate to performance no-tillage. Although millet had yielded less quantity of straw, the great carbon/nitrogen relation (C/N) of this specie could contributed to keep straw on soil surface for a long time, due to the slow straw decomposition. On the other hand, white lupin provided nitrogen into the soil by biological nitrogen fixation throughout bacterium symbiotic association with rhizosphere.

The results from the soil chemical analysis at 0.0 to 0.10 m depth at the end of tomato harvest are shown in Table 1 . For the calcium (Ca) content and the cationic exchange capacity (CEC) had significant interaction between experimental factors, because the soil Ca content and CEC increased under white lupin cover crop and conventional tillage ( Figure 1(a) and Figure 1(b) ).

*Significant differences by the variance analysis (ANOVA) at 5% of probability.

The soil magnesium (Mg) concentration at 0.0 to 0.10 m depth had no difference between cover crop, millet and white lupin. However, the soil Mg concentration in the conventional tillage (27.6 mmol dm⁻³) was larger than no-tillage (23.6 mmol dm⁻³).

The soil potassium (K) concentration increased in white lupin crop rotation with 8.1 mmol_c dm⁻³ regarding to soil under millet cover crop, that’s recorded 6.1 mmol dm⁻³. There is no difference between tillage methods for soil K concentration at 0.0 to 0.10 m depth.

The soil sulfur (S) concentration had no difference between experimental treatments of cover crops and soil tillage methods. The soil phosphorus (P) concentration at 0.0 to 0.10 m depth there is no difference between cover crops, but in the conventional tillage the soil P concentration (120 g dm⁻³) increased compared to no-tillage (101 g∙dm⁻³).

The experimental treatments had no significant effect on the soil organic matter at 0.0 to 0.10 m depth. Regarding soil pH, conventional tillage increased pH compared to no-tillage.

The results of the soil chemical analysis from 0.10 m to 0.20 m depth are shown in Table 2 . For the soil chemical traits evaluated in the research there was no interaction between experimental factors, cover crops and soil tillage methods.

*Significant differences by the variance analysis (ANOVA) at 5% of probability.

The Ca, Mg, K, P soil concentration and soil CEC increased under the conventional tillage regarding no-tillage. However, there is no difference for these soil nutrients between white lupin and millet at 0.10 to 0.20 m depth. White lupin increased soil organic matter regarding to millet as well as conventional tillage compared to no-tillage. The soil pH at 0.10 to 0.20 m depth there is no difference between treatments.

The results of soil enzymatic activity are shown in Table 3 . The experimental treatments of cover crop, white lupin and millet did not differ for soil activity of β Glucosidase enzyme. Meanwhile, conventional tillage increased the activity of β Glucosidase enzyme regarding to no-tillage, recording values of 102.6 and 62.2 ug p-nitrophenol/g/h, respectively.

The experimental treatments of cover crops, white lupin and millet and tillage methods, conventional and no-tillage did not differ for Arylsulfatase soil activity, which had average value of 57.0 ug p-nitrophenol/g/h.

*Significant differences by the variance analysis (ANOVA) at 5% of probability.

Regarding to soil moisture content at 0.0 to 0.10 m depth, cover crops did not differ between them, which had recorded average value of 20.5 g H₂O 100 g solo⁻¹, but soil tillage methods differed between them that’s no-tillage stored more water with 22.1 g H₂O 100g soil⁻¹ against 18.7 g H₂O 100 g soil⁻¹ from conventional tillage, which means 18.2% more water in soil under no-tillage at top soil 0.0 to 0.10 m depth ( Table 4 ).

*Significant differences by the variance analysis (ANOVA) at 5% of probability.

At 0.10 to 0.20 m depth, it was not recorded no difference between cover crops for soil moisture content which average value was 20.1 g H₂O 100 g soil⁻¹ for both, white lupin and millet. Meanwhile, under no-tillage the soil moisture content was 13.3% larger than conventional tillage which was recorded 22.2 g H₂O 100g soil¹ and 19.6 g H₂O 100 g soil⁻¹, respectively.

Soil resistance to penetration did not differ for cover crop treatments, but regarding to tillage methods conventional tillage decrease soil resistance for penetration especially up to 0.00 to 0.20 m depth compared to no-tillage. For soil depths below 0.20 m there was not significative difference between conventional and no-tillage ( Figure 2 ).

The cover crops treatments were significant for quantity of tomato brunch accounted at 55 days after transplanting (DAT), where tomato grown successively to white lupin produced 16 fruit bunches against 11 produced in succession of millet ( Table 5 ).

The same result was recorded for the quantity of tomato fruits and to the tomato fruit fresh biomass, because in succession to white lupin, tomato produced 35 fruit per plant against 21 fruits per plant from the succession of millet and produced 1535.7 g plant⁻¹ of fruit fresh biomass in successive grown to white lupin against 998.8 g plant⁻¹ at millet succession.

*Significant differences by the variance analysis (ANOVA) at 5% of probability.

The experimental treatments of cover crops and tillage methods had not significative effect on the tomato aboveground dry biomass evaluated at 55 DAT.

There was no significative difference among experimental treatments of cover crops and tillage methods for root quantity, root length, volume and area of soil explored by tomato roots measured at 55 days after transplanting at 0.00 to 0.20 m of soil depth ( Figure 3 ). The same results of tomato root growth parameters were recorded at 0.20 to 0.40 m measurement that’s any experimental treatments were not efficient to differ on root growth parameters ( Figure 4 ).

The cover crops treatments did not differ between them for tomato yield, but for tillage methods, conventional tillage with 70.4 Mg∙ha⁻¹ of tomato yield was greater than no-tillage which had 56.5 Mg∙ha⁻¹ ( Table 6 ).

*Significant differences by the variance analysis (ANOVA) at 5% of probability.