|
|
|
| Authors: | D. Almaliotis, S. Bladenopoulou, CH. Chatzissavvidis, N. Karagiannidis |
| Keywords: | vegetables, soil fertility, minerals, fertilization, multiple regressions |
Abstract:
Twenty-four fields were cultivated with broccoli (cv. ‘Marathon’) during the period late summer - early winter of 2003 in Northern Greece.
Broccoli heads were harvested, weighed separately for each field and processed as frozen product.
Nitrogen was applied at a rate of 250 kg/ha and additional nutrients were applied as required by soil analysis.
Yield ranged from 3.15 to 14.42 t/ha, mean yield was 8.12 t/ha with a standard deviation of 3.11 t/ha.
The soils were characterized as medium to moderate heavy in texture with its basic parameters being as follows: a) Physicochemical properties: pH 6.82-7.96, electrical conductivity 0.32-1.69 mmhos/cm, organic matter content 0.35-2.51%, CaCO3 content 0.50-10.10% and b) Nutrient elements (mg/kg): P 9.26-81.44, K 110.0-470.0, Ca 136.0-1100.0, Mg 22.8-186.0, B 0.22-0.77, Mn 3.80-24.90, Zn 0.61-6.73, Fe 4.36-28.00 and Cu 0.39-4.00. Multiple regression analysis revealed that yield was significantly related with the studied soil parameters (r2 = 0.942). Furthermore, stepwise variable selection program showed that the parameters pH, electrical conductivity and Cu were the most important variables for yield prediction (r2 = 0.923). Especially, pH and Cu had a positive effect on broccoli yield, while electrical conductivity, a negative one.
|
Download Adobe Acrobat Reader (free software to read PDF files) |
|