Articles
EFFECT OF SPACING AND ROOTSTOCK ON THE PERFORMANCE OF ‘COMICE’ PEAR IN NEW ZEALAND
Article number
596_105
Pages
609 – 614
Language
English
Abstract
Traditionally in New Zealand European pears have been planted at 5 x 3 m on quince BA29 or on Pyrus seedlings, and grown as a large centre leader tree.
In the spring of 1993 a spacing trial was planted at Nelson Research Centre with Doyenné du Comice on quince rootstocks BA29 and QC planted at three within-row spacings (1.5, 1.9 and 2.4 m) in rows 3.5 m apart, giving tree densities of 1905, 1504 and 1190 trees/ha.
The trees were well-branched at planting, being derived from bench grafted rootstocks in the spring of 1991. The trees were trained in a spindle tree form, although the height of the central leader was allowed to go up to 2.8 m, and an outrigger wire was used to assist tree training. Williams/BA29 was used as a pollinizer, at a ratio of 1 in 5. Vegetative growth was much reduced on QC compared to BA29; after seven years in the orchard, the mean TCA on BA29 was 89 cm2 compared to 44 cm2 on QC.
Along with this growth reduction, there was also a more pronounced overgrowth of the scion at the graft union on QC compared to BA29. Appreciable yields were harvested from the trees on both rootstocks in year two, reflecting the good tree quality at planting, but there was a significant effect of rootstock (P = <0.001) with a mean of 5.5 kg on QC and 2.5 kg on BA29. This initial advantage of a higher yield per tree with QC was not sustained to tree maturity; by years six and seven the mean yields per tree were 32 kg on QC and 38 kg on BA29. Mean fruit weight was significantly smaller with trees on QC compared to BA29 throughout the life of the trial; an annual mean of 274 g with trees on QC compared to 313 g on BA29. Mean number of fruit per unit TCA was always higher on QC compared to BA29. The effect of rootstock on fruit size appeared to be a direct effect of the increased fruit numbers per unit TCA with QC, but further analysis on the basis of fruit number per unit light interception revealed a significant smaller mean fruit size on QC compared to BA29 for the same level of crop load.
Significant effects of within-row spacing were slow to develop on this plot, partly due to soil variability.
There was no significant effect of spacing on yield per tree over the first five growing seasons.
Mean yields per tree for years six and seven, showed a significant interaction between rootstock and spacing (P = 0.012), with a much stronger decline in yield per tree with closer within-row spacings with trees on BA29 than on QC, suggesting greater tree to tree competition with the more vigorous rootstock.
This was borne out by the relationship between yield per hectare and light interception, where the slope of the relationship was smaller with BA29 than QC. This suggests that higher yields per hectare could be obtained with QC at higher tree densities than those used in this trial.
Mean yield per hectare in year six and seven at 3.5 x 1.5 m was 59 tonnes with QC and 63 tonnes with BA29.
In the spring of 1993 a spacing trial was planted at Nelson Research Centre with Doyenné du Comice on quince rootstocks BA29 and QC planted at three within-row spacings (1.5, 1.9 and 2.4 m) in rows 3.5 m apart, giving tree densities of 1905, 1504 and 1190 trees/ha.
The trees were well-branched at planting, being derived from bench grafted rootstocks in the spring of 1991. The trees were trained in a spindle tree form, although the height of the central leader was allowed to go up to 2.8 m, and an outrigger wire was used to assist tree training. Williams/BA29 was used as a pollinizer, at a ratio of 1 in 5. Vegetative growth was much reduced on QC compared to BA29; after seven years in the orchard, the mean TCA on BA29 was 89 cm2 compared to 44 cm2 on QC.
Along with this growth reduction, there was also a more pronounced overgrowth of the scion at the graft union on QC compared to BA29. Appreciable yields were harvested from the trees on both rootstocks in year two, reflecting the good tree quality at planting, but there was a significant effect of rootstock (P = <0.001) with a mean of 5.5 kg on QC and 2.5 kg on BA29. This initial advantage of a higher yield per tree with QC was not sustained to tree maturity; by years six and seven the mean yields per tree were 32 kg on QC and 38 kg on BA29. Mean fruit weight was significantly smaller with trees on QC compared to BA29 throughout the life of the trial; an annual mean of 274 g with trees on QC compared to 313 g on BA29. Mean number of fruit per unit TCA was always higher on QC compared to BA29. The effect of rootstock on fruit size appeared to be a direct effect of the increased fruit numbers per unit TCA with QC, but further analysis on the basis of fruit number per unit light interception revealed a significant smaller mean fruit size on QC compared to BA29 for the same level of crop load.
Significant effects of within-row spacing were slow to develop on this plot, partly due to soil variability.
There was no significant effect of spacing on yield per tree over the first five growing seasons.
Mean yields per tree for years six and seven, showed a significant interaction between rootstock and spacing (P = 0.012), with a much stronger decline in yield per tree with closer within-row spacings with trees on BA29 than on QC, suggesting greater tree to tree competition with the more vigorous rootstock.
This was borne out by the relationship between yield per hectare and light interception, where the slope of the relationship was smaller with BA29 than QC. This suggests that higher yields per hectare could be obtained with QC at higher tree densities than those used in this trial.
Mean yield per hectare in year six and seven at 3.5 x 1.5 m was 59 tonnes with QC and 63 tonnes with BA29.
Publication
Authors
J.W. Palmer
Keywords
HortResearch, Nelson Research Centre, Motueka, New Zealand
Online Articles (158)