Response of Eucalyptus grandis in Colombia to mid-rotation fertilization is dependent on site and rate but not frequency of application

Abstract A nutrient dose and application frequency study was installed in Eucalyptus grandis stands at six sites in the Colombian Andes to examine three hypotheses: (1) individual sites have different treatment responses (there is a significant site effect); (2) the relationship between volume growth response and applied nitrogen is not linear (there is an optimal amount of applied nitrogen beyond which growth improvements would be small); and (3) application frequency does not affect response if the cumulative dose applied is equivalent (there is no frequency effect). Nitrogen, phosphorus and boron were applied at a 1.0:0.1:0.005 ratio, where nitrogen rates ranged from 0 to 250 kg ha −1 and application frequency was 6, 12, 24 or 36 months. Fertilization began when trees were 11–24 months old and we examined volume growth response three years after study initiation when the maximum cumulative nitrogen application reached 720 kg ha −1 . There was a significant site effect: two sites were responsive and four were non-responsive to treatment. At the two responsive sites, the relationship between volume growth response and applied nitrogen was not linear. Three years after treatment initiation, the maximum absolute response was 142 m 3  ha −1 (91% increase) and 116 m 3  ha −1 (56% increase) for Sites 1 and 4, respectively. Sites 1 and 4 reached maximum volume response at cumulative nitrogen doses of 360 and 480 kg ha −1 , respectively. Tests that compared different nitrogen application rates and frequencies to achieve the same cumulative dose were not significantly different except for one test at Site 4 where the volume growth response to three applications of 180 kg nitrogen ha −1 was 58% greater than the response to six applications of 90 kg ha −1 .