raised by a major nurseryman were on dwarfing rootstocks and interstems compared to 60% in
1998.

EXPERIMENTALRESULTS
In 1991 there was very little experience in New Zealand with virus-free M.9, therefore it was
important that we gain experience of its vigor to know what tree densities were feasible.I was
fortunate to secure the cooperation of two Nelson growers who planted blocks of Royal Gala on
NZ9 in the spring of 1993.Both growers were keen for some trial work to be done on their
orchards and therefore a within-row spacing trial was established at each site.These trials then
supplemented within-row spacing trials of Royal Gala, Fuji and Braeburn on M.9 that were
planted at the same time on the Nelson Research Orchard.In the following year another spacing
trial of Royal Gala, Fiesta and Fuji was planted in the Marlborough region.The majority of the
results presented in this paper are from the two spacing trials on the Nelson grower orchards.
Some of the early results from this work were presented by Palmer and Adams (1997).

On both Nelson grower sites unfeathered whips of Royal Gala/M.9, clone Galaxy, were planted
in the spring of 1993 at five within-row spacings in a randomized block experiment with
5 blocks and 3 trees per plot separated by a single guard tree.On the first grower property, on a
deep clay soil, the trees were planted with a row spacing of 3.5 m (11.5 ft) and within-row
spacings of 0.75, 1.0, 1.25, 1.5 and 1.75 m (2.5, 3.3, 4.1, 4.9 and 5.7 ft), giving a range in tree
densities from 3810 to 1633 trees/ha (1542 to 661 trees/acre).On the second orchard, on a fertile
light silt soil, the trees were planted with a row spacing of 3.7m (12.1 ft) and within-row
spacings of 0.9, 1.15, 1.4, 1.65 and 1.9 m (3.0, 3.8, 4.6, 5.4 and 6.2 ft), giving a range in tree
densities from 3003 to 1422trees/ha (1215 to 575 trees/acre).Records included total number
and weight of fruit per plot and trunk cross-sectional area (TCA).The trees were pruned as
slender spindles by the researchers while the growers were responsible for the other cultural
operations.

Tree growth, as indicated by TCA, was largely unaffected by tree density after the first growing
season (Figure 1).Over the next 3 years tree density showed a more pronounced effect on TCA,
with the suppression of growth becoming more severe with time at the higher tree densities.
Yield per hectare showed a mirror image of the effects on growth (Figure 2), although curves
have been fitted through the data as yield would be zero at a tree density of zero.As the yields
were similar on both sites, the curves have been fitted through both data sets.There was no
effect of spacing on yield per tree in the first cropping year (second leaf), with an average yield
per tree of 3.5kg (7.7 lb), so yield per hectare was a linear function of tree density.This is a
frequent finding with trials of this type.The relationship between yield and tree density became
more curvilinear from year 3to year 5, with an increasing reduction of yield per tree at the
closer spacings compared to the wider spacings, so that by the fifth leaf yield per hectare showed
only a relatively small increase beyond 2500 trees/ha (1012trees/acre).Trees at 2200 trees/ha
(890 trees/acre) reached yields of 80t/ha (36 ton/acre) by year 5.These results illustrate how
rapidly new cultivars could be brought on stream using such planting systems.The average yield
per tree and mean fruit weight from all spacings for each year are given in Table 1.Fruit quality
generally was very good although no details of defects were collected.Mean fruit weight was
large, above 180 g in all cases except for Grower 2 in year 4 when there was a misunderstanding
over the hand thinning and the trees were thinned very late.Other than the effect of spacing on
growth and yield per tree, spacing effects to date have been small.Closer spacings resulted in a
significant decline in mean fruit weight in years 3 and 4 for Grower 1 and year 4 for Grower 2

but not in other years.There has been no significant effect of spacing on the fraction of fruit
removed in the first pick.

The use of well-feathered trees at planting would have further improved precocity and probably
saved almost a year in the cumulative cropping cycle.Work with unfeathered and feathered trees
on M.9 in New Zealand has shown that yields in the first 2 or 3 years can be increased by two to
threefold by using feathered trees (Palmer and Adams, 1997).

Royal Gala does not normally suffer serious problems of russet or sunburn and therefore is
perhaps an “easy”cultivar to try on M.9.In our other trials of Braeburn and Fuji on M.9 we
have seen serious downgrading of Fuji due to russet, particularly in the Nelson region.In drier
regions such as Marlborough, we have not seen such serious problems of russet on M.9.
Sunburn levels on Fuji and Braeburn on M.9 reached 12% and 9% in Nelson in the bad sunburn
year of 1997/98.This is somewhat higher than growers experienced on larger trees, although
this was counterbalanced by good color and high soluble solids.

Although results have been very encouraging from these trials on M.9, it is appropriate to
consider the issue of woolly apple aphid.At present we have no commercially available
dwarfing apple rootstocks resistant to woolly apple aphid.No serious woolly apple aphid
problems have been seen in these orchards to date, although none of these orchards are under
IFP. Certainly woolly apple aphid seems generally to be a more common pest under IFPthan
under conventionally sprayed systems but the parasiteAphelinus malican successfully control
this pest once established under IFP. Although M.9 interstem trees on MM.106 or M.793 offer
an insurance policy against woolly apple aphid root infestation, tree size to date from our
interstem trials is more in the M.26 size range or slightly larger.The economic advantages of the
intensive systems on M.9 rootstock may, however, outweigh the risk of serious woolly apple
aphid debilitation for growers wanting to move to intensive plantings in the immediate future.In
the longer term, dwarfing rootstocks with woolly apple aphid resistance will remove this
concern.

Further work is in progress to examine in detail fruit quality from the trials described here and
from an experiment comparing Royal Gala planted intensively on M.9 and at the more
conventional spacings on MM.106.

SUMMARY
Intensive apple orchards on M.9 have shown good precocity, high yields and excellent fruit size.
These systems offer the New Zealand pipfruit growers new management options for their future
orchards as they seek to respond to the ever-changing economic and market pressures on the
pipfruit business.In my 8 years in New Zealand there has been a positive shift in the attitude of
growers to smaller trees and undoubtedly this will accelerate as dwarfing rootstocks resistant to
woolly apple aphid become widely available.

ACKNOWLEDGEMENTS
These trials could not have been done without the helpful support of the growers concerned.It is
no mean task in a hectic picking season to ensure that the trial plots are looked after.Further
thanks are extended to ENZA for funding some of the work referred to in this article.

LITERATURECITED
Palmer, J.W. and H.M. Adams.1997.Early results with intensive systems of apples on virus-
free M.9 rootstock in New Zealand.Acta Hort. 451:487-493.
Tustin, D.S., P.M.Hirst, W.M. Cashmore, I.J. Warrington and C.J. Stanley.1990.The principles
and practices of training slender pyramid trees for high intensity production. Compact Fruit
Tree 23:83-92.
Warrington, I.J.1994.The ‘Granny Smith’apple.Fruit Var. J. 48:70-73.

Table 1.Mean yield per tree and average fruit weight of Royal Gala/M.9 over four seasons at
two grower orchards in Nelson, New Zealand.

Grower 1
____________________________

Grower 2
_____________________________

Year

Yield (kg)

Mean wt. (g)

Yield (kg)

Mean wt. (g)

2
3
4
5

4
13
21
34

198
191
185
180

3
16
26
38

208
180
147
189