Pollenizer management in a hazelnut orchard

S.A. Mehlenbacher and A.N. Miller – Oregon State University, 1988

HGA newsletter, Winter 2003

Three factors must be considered in choosing pollinizer cultivars: 1) the amount of viable pollen produced, 2) compatibility, and 3) time of pollen shed.

The amount of viable pollen produced by a hazelnut tree is largely a function of the number of catkins on the tree and the viability of the pollen produced. Some cultivars set pollen in abundance Others typically set very few catkins. Some cultivars drop their catkins prior to pollen shed. Since one good Daviana catkin is estimated to produce 4 million pollen grains, the amount of pollen produced by a single pollinizer tree is tremendous.

Incompatibility occurs when plants having functional pollen and functional female flowers are unable to set seed when self-pollinated or crossed with some of their relatives. Incompatibility is a physiological between pollination and fertilization. Pollinizers are required in hazelnut orchards because cultivars are self-incompatible. They are also cross-incompatible in certain combinations. Hazelnuts are similar to Brassica species (cabbage, broccoli, and their relatives) in that all have the sporophytic incompatibility system All pollen produced by a tree exhibits the same incompatibility reaction. The reaction is under simple genetic control. There is one locus (one gene), the S-locus (for self-incompatibility) with 22 known alleles. Don’t let this terminology scare you. In humans, there is a locus controlling eye color, an allele for brown eyes, and an allele for blue eyes. Humans have a locus controlling hair color, an allele for brown hair, and an allele for blond hair. In hazelnuts, there is one S-locus which controls the incompatibility reaction, and alleles S1, S2, S3,…..S22. Hazelnut trees, like humans, are diploids. Thus at every locus, they have two alleles. Brown eyed people may have two alleles for brown eyes or one allele for brown eyes and another for blue eyes. The allele for blue eyes is recessive. In hazelnuts, both alleles are always expressed in the female flowers. One or both may be expressed in the pollen. If a given allele expressed in the pollen is met with the same allele in the female flower, the cross is incompatible. In simpler terms, if like meets like, the reaction is incompatible. If a given allele expressed in the pollen is met with different alleles in the female flower, the cross is compatible. Because of the dominance relationships among S-alleles, some crosses are compatible yet the reciprocal cross is incompatible.

A third factor to consider in choosing pollinizer cultivars is time of pollen shed. It is essential that the pollinizer shed pollen when the female flowers of the main crop cultivar are receptive. Pollen which is shed before the female flowers emerge is wasted. Unpollinated female flowers remain receptive for up to 3 months, so a late pollinizer can be very effective. Although actual dates vary from year to year, relative dates of pollen shed are consistent. Female flowers and catkins respond differently to temperature. In colder winters, female flower development is accelerated relative to catkin elongation. Thus a pollinizer which sheds at the ideal time one year may shed too early or too late the next year. Some cultivars shed their pollen over a very short time while others shed over a much longer time. By planting 2 or 3 rather than a single pollinizer, growers increase the odds of having pollen shed at the optimum time.

Another dehiscence and pollen release from catkins requires lower relative humidity and warmer temperatures. If temperatures are too cold (320F or less) and humidity is too high (85%+) then pollen will not be shed.

Temperature will also have an effect on pollen viability. If temperatures exceed 730F then pollen viability decreases. Optimum pollen germination in artificial culture has been obtained at 68-720F, but will also germinate at 34-390F.

Once pollen is released from the anther, wind is required for its distribution. The absence of wind could result in pollen dropping to surfaces below the catkins. Theoretically, pollen could travel about 250ft. in 36 seconds and drop 3 feet. if a 15mi/hr wind was blowing. However, hazelnut pollen is moved by eddy diffusion where the air movement rolls and swirls. Pollen consentration drops very quickly to relatively low amounts around 46-72 ft. distance away from the orchard edge. At this time we do not know what the required amount of pollen is to guarantee optimum nut set and development. However, in 2 research articles Schuster (5,6) states, “In observation in the field is shows that trees planted 40-60 ft. from a pollinizer bear good crops though the ones at 60 ft. sometimes appear light.” and “In the field it has been noted that trees more than 50ft. from the pollinizer yield smaller crops than closer.”

The above article has been abridged to conserve space in our news letter. Data tables showing compatible crosses, compatible cultivars, Cultivar pollen shed compatibility, Pollen concentrations at different distances, and charts showing pollinizer planting options within orchards have been left out. If a full copy of the article is wanted, please contact the Editor.

The hazelnut tree is a wonder

Jeff Olson, Extension agent, Oregon State University

HGA newsletter, Autumn 2003

Horticulturally speaking, the hazelnut tree is clearly out of the ordinary. It is more than just nutty. It is unique and wonderful. The way in which it achieves pollination in the winter and completion of nut set in the spring, is like no other horticultural crop that I have ever heard of. It is a “one of a kind”, just like some of the people in our industry!

Over the years, many researchers have investigated the growth and development of the hazelnut, in an attempt to unlock some of the secrets of this unusual plant. In fairly recent times, 1979, Dr. Maxine Thompson, of OSU, published a very informative article about the growth and development of the hazelnut flowers and nuts. It is one of those information-packed articles that is peppered with words like: megasporocytes, achesporial cells, funiculus of the anatropous ovule…you know what I mean, light reading.

Maxine followed the development of the flower and nut from the time of pollination to harvest. And she found some fascinating things. Pollination takes place in January and February, with the wind being the pollinator. There are a pair of styles that are joined at the base by a tiny ovarian meristem. When the pollen lands on the stigmatic surface of the style, hopefully it germinates and forms a pollen tube. The pollen tube grows to the base of the style in 4 to 7 days and then it rests there until May. Four to five months lapse between pollination and fertilization of the ovary.

This research project looked at how long the stigma would remain receptive to pollen. Did you know that the stigmatic surfaces of hazelnut female flowers remain receptive for up to 3 months? This study used a controlled hand pollination to test the length of receptivity at five different stages of flower development, and found them to be viable all the way up to March 13. Those old flowers were withering and necrotic, but the stigmas were still receptive.

The clusters that had not been pollinated successfully all dropped to the ground by the end of May. The flowers that were successfully pollinated continue with the development of the ovaries that become mature by the middle of June. Fertilization of the ovaries can then take place. The nut can now develop, and it reaches full size by the beginning of August. By July the shell begins to harden, and is hard by early August.

The cells that will form the following year’s crop also form in the spring and summer, while the present crop is being set. The male flower, the catkins, differentiates at the end of May and beginning of June. The female flowers for the next year differentiate between July and September. So, you can see that there is an awful lot going on inside that hazelnut tree in the spring and summer.

Way back in 1925 Joseph Newell did a study of the bearing shoots of the filbert. He found that the number of fruiting buds was directly proportional to shoot length, with the number of fruit buds increasing as the shoot length increased. The diameter of a shoot was a good index to its length. In most cases he found that there is very little degree of negative correlation between a shoot`s angle and the number of fruit buds. He noted that better light illumination results in better production in the upper third of the tree. Of course, this observation has been verified quantitatively by the more recent work of Dr. Anita Azarenko.

Newell found that in most cases the variations of illumination on different sides of trees was insufficient to cause any consistent variability in the number of fruit buds formed on the shoots. He observed that in Barcelona, a fair degree of correlation exists between the male catkins and the number of fruit buds per shoot.
In 1957 John Painter and Henry Hartman reported that the greater the length of the shoot: 1) The greater the number of female flower clusters; 2) the greater the percentage of female flowers that had nuts set in May; 3) the greater the number of nuts in a cluster; 4) the greater the number of nuts matured; 5) the larger the nuts. Overall evidence suggests twigs greater than 6.25 inches long bore the largest and greatest numbers of nuts. Twig length did not effect distribution of blank nuts. They noted that filberts can not produce both extensive twig growth and extensive nut production in the same year, hence the alternate year effect.

In one of the few research studies on the root systems of hazelnuts that I have found, scientists in Italy exposed two adjacent halves of two neighboring trees. The trees were closely spaced at 13′ x 13′, so the roots tended to extend diagonally to the corners of a square but laterally only as far as the roots of the next tree. The bulk of the roots occurred in the 8-31 inch depths of the soil. Many feeder roots occurred in the soil directly under the trunk. The study was conducted on deep fertile soil and that had a practice of cultivating to a 10-inch depth, which were the factors mainly responsible for this type of vertical root distribution.

So, you see that the quest for understanding the marvelous hazelnut tree have been going on for decades. Significant discoveries have been made, and certainly more await discovery.

The NZ hazelnut industry. Where to from here?

By Murray Redpath

HGA newsletter October 2010.
Originally presented at HGANZ AGM, 2010

NZ hazel industry 2010:

NZ has over 430 hectares planted, most since 2000. Almost 100% of this area is planted in Whiteheart plus pollinisers so we have an industry dependent on successfully growing and marketing Whiteheart.

There has been a change in the orchard size since 2000 – from mainly orchards of 200-1000 trees to mainly orchards of over 1000 trees (see graphs below).
Fig. 1: Number of hazel trees per grower, SNGA hazelnut seminar 1996 (McNeil, 1996)
Percentage of attendees
60 50 40 30 20 10 0

<200 200-399 400-999 1000->2000 1999
Orchard tree number


The Hazelnut Company is the largest hazelnut processor with a large factory in Canterbury; owned by about 100 grower shareholders.

There are five or six smaller processors with varying levels of sophistication. All sell into the NZ market. One small processor has a small export market.

Lessons from the past:

The NZ hazelnut industry has been developing for at least 30 years. Early research indicated the possibility of economical yields.

Example: Yields from the Gum Block, DSIR Research Orchard, Appleby, Nelson for the “Barcelona”* variety (multi-stemmed bush planted 1953) averaged over the years 1970 to 1975 = 10.3 kg/tree/year.

At 400 trees/ha. (5×5 spacing) = over 4 tonnes/ha/year.

*Note: the tree tested as “Barcelona” was renamed “Appleby” after Prof. Maxine Thompson established that it was not the true Barcelona during a visit in 1981. It appears to be identical to the Spanish cultivar Negret.

Appleby was the main variety planted in the early orchards. Crops well, kernels blanch well.

What happened to these early hazel orchards? A typical example is one in Maraekakaho, Hawke’s Bay.

Planted in 1984 at a 4 x 2 metre spacing. Mostly Appleby, later plantings included Whiteheart, Barcelona, Ennis, and Butler. 1989 photos show excellent yields 5 years after planting. The owner used a local hazelnut cracker copied from an American design to crack and sell the crop, successfully tested the sorting of kernel from shell at local seed dressing facilities, and tried vacuum packing nuts at local meat processing facilities.

The property was sold in 1990 just as the trees were starting to crop well. The new owner had no interest in hazels and did nothing to the trees. By 2000 the trees were over 5 metres high, a “forest” with a dense canopy, and the trees were riddled with big bud mite and lemon tree borer. Often the main tree had died and been replaced by vigorous suckers.

The property sold again in 2000, the new owner was told that hazels were of no use and should be pulled out. A chance visit revealed that he was interested in cropping the hazels if they could be rehabilitated. The orchard was trimmed to about 1.5 metres high and 1.5 m wide. Further trimming of excess suckers and branches occurred over the following 2 years.

The trees put on over a metre of healthy growth in the first year and the orchard was back to having a vigorous productive canopy carrying a heavy crop in 2004 when……..
The orchard was sold again (to an absentee landowner).

When last visited it was being leased by an organic grower who was using the hazel orchard to run free range hens. At least the orchard is still there – most of the ones that I visited in the 1980’s were pulled out in the 1990’s.

It is difficult to grow an industry and sustain processing plants and supply to markets when orchards disappear just as they are starting to produce well. At present I am hearing rumours of some growers talking about pulling their trees out – are we going to see history repeat itself?

Why do we all grow Whiteheart?

Whiteheart was selected from a group of old varieties and seedling selections by Prof Maxine Thompson during her 1981 visit and became widely recommended after it performed well in the Lincoln hazelnut trial. It has been promoted because it is seen as offering a quality advantage (a round clean kernel that blanches 100%) over Turkish or Italian kernel varieties.

It does have some problems such as vigorous suckering, late flowering, being susceptible to big bud mite and bacterial blight, and it does not yield well in many orchards.

Research Projects

MAF Sustainable Farming Fund L08-088:

Helping growers to record flowering information and write a web-based guide to hazel varieties. Due for completion in July 2010.

MAF Sustainable Farming Fund L10-148:

A 3 year project building on the last project by collecting and analysing flowering and yield data from growers. The aim is to identify possible causes of poor performance in some orchards (pollination problems, location, frost damage, variety selection?).

Others – Agmardt?

An application may be made for funding from Agmardt for a literature search and analysis into the potential for hazel growing as an economical alternative land use in water or environmentally sensitive areas.

Lessons from the Jeff Olsen road trip.

Prof. Olsen considered that the growth rate and yields of many of the Whiteheart orchards in Canterbury were much lower than Oregon orchards of a similar age.
We visited an interesting 11 year old orchard on the down lands west of Oamaru. Planted mainly in seedlings, it had 2 rows of Whiteheart that were yielding very well on heavy soils with irregular irrigation. This orchard showed the necessity for shelter – the Whiteheart were in the lee of the slope and well sheltered. Trees on the windward slope were struggling.

Prof. Olsen felt that the orchards in Nelson and Marlborough were more similar in growth to Oregon hazel orchards. He commented that the climate had a similar feel to it.

Site selection:

Prof. Olsen felt that more consideration needs to be given to selecting the correct variety for each site. Whiteheart is a low vigour variety and should only be grown on the best sites. On more difficult sites, a vigorous variety should be considered where the vigour will compensate for the slower growth rates.
Some interesting statistics from Prof. Olsen’s presentations at the April meeting:
• About 75% of the Oregon crop is sold in the shell.

• About 60% of the crop is exported.

Southwest Pacific In-shell Export Totals (Tons) August 2009 report

• Australia = 11.02

• China = 631.06

• Hong Kong = 13820.22

• Vietnam = 4282.79

• New Zealand = 1.5

• Total Tons = 18776.59

• This represents 85% of the total US Exports, up from 71% in 2007.

Costs of establishment for Oregon orchards: the break even point, when returns cover annual cash costs, is year 7 at standard densities (6 metre by 6 metre) and year 4 for double density orchards.

All previous cash costs of establishment are paid back by year 11 in standard density and year 9 for double density.

South east injury”:

Prof. Olsen recognized the damage to young trees in some orchards (Shore’s; McCracken’s) as damage by bacterial blight (Xanthamonas). It is common in Oregon on varieties susceptible to blight. It is caused when blight bacteria enter through bark damaged by rapid thawing on the sunny side of the trunks after severe winter frosts. In Oregon this is on the southeast side of the plant. This is the reason why they recommend painting the trunks over the first year.

Grafting onto existing trees to establish new varieties.

Oregon State University has successfully grafted new varieties onto existing trees in the orchard. Both top working trees and grafting onto suckers worked, grafting onto suckers was more successful (grafts at ground level were more sheltered than those in the canopy).

For grafting onto suckers, the tree was chopped down during the summer before grafting and a few suckers allowed to grow. Scion wood was collected in winter. Grafting took place in summer using a standard whip and tongue graft. It was critical to have daytime temperatures consistently above 20° C and to have high light levels around the tree to be grafted.

New varieties:

By “new varieties” I am referring to varieties that are already in NZ (such as Ennis) as well as imported varieties and new selections. There is renewed interest in varieties other than Whiteheart. Lansing and Ennis, two varieties grown for the in-shell market, are two varieties that are being planted. They are known to crop heavily and give good quality nuts in some locations (eg Marlborough) but we need more information about their performance in all growing areas before recommending large plantings. In Australia, Ennis often had a lot of shriveled kernels (Baldwin, 2007) and my experience supports this.

New Zealand has a useful number of the main varieties used internationally (eg Barcelona, Ennis, Tonda Romana, and Tonda di Giffoni) and perhaps we should examine their performance in various growing areas.

I have imported several new varieties from the Oregon State University programme and these are now out of quarantine. The main commercial variety is Jefferson. This has large nuts similar in size to Barcelona nuts but with fewer nut defects. It is a smaller tree and very productive. Jefferson flowers late and Eta and Theta were selected to pollinate it. It is susceptible to bacterial blight (similar level to Barcelona) so it may not be suited to harsher sites where blight has been a problem in Whiteheart.

The other 4 varieties imported are late pollinisers that are compatible with Whiteheart. In Oregon, Epsilon, Zeta, and Eta all drop pollen about a week after Merveille de Bollwiller and Theta about 2 weeks after M. de Bollwiller. They all have nuts about the same size as Whiteheart with kernels suited to the kernel market. Epsilon does not blanch but the other 3 varieties all blanch well (80-90% pellicle removal). Epsilon was imported because it is very resistant to big bud mites and may be useful where BBM is a problem.

I have used Merveille de Bollwiller as a standard reference and graphed the expected pollen release times below using data from Oregon and Tim Medlicott’s orchard near Dunedin. Theta may be a useful replacement for Alexandra in Whiteheart orchards.

We now have a wide range of hazel selections that drop pollen over August. There are selections from old Canterbury plantings and my breeding programme that cover this period and have kernels that blanch well. These may be useful as future pollinisers but we have not yet tested them for compatibility with Whiteheart whereas we know that the imported varieties are compatible.

The NZ hazelnut industry: At a crossroads?

We now seem to have a crop that is approaching the size of the local NZ market and we know that we will exceed the present NZ market for hazelnuts if all planted trees yield to expectations.

Do we want to keep planting and growing our crop?

Can we expand the local market or should we be looking at exports?
If we export, where are our markets likely to be and what varieties do we need for those markets?

Flowering and pollination

By Murray Redpath

HGA newsletter September 2011

The warm autumn experienced this year has highlighted the importance of current research into flowering and pollination behaviour of different cultivars at different locations around the country.

  • In South Canterbury (Geraldine), Lansing was in flower in mid-July but of the others only Butler had any flowers. Merveille, Alex and Whiteheart were shedding pollen from early July at least. By late August, there was still some flower on Butler, Ennis and MT18/114 but Lansing flowering was over. Merveille pollen was finished but Alexandra and Butler were still shedding.
  • In North Canterbury, Alexandra had an unusually long lead-in to pollen shed and peak dates seem about 2 weeks earlier than last year; MdB peaked 2 weeks later than Alex and had more prolonged shedding than previous years. Whiteheart flowering is later than previous years and looks likely to miss most of the pollen altogether.

Information on flowering and pollination in your orchard is vital to building up a complete picture of the behaviour of different cultivars throughout the country.
Flowering will be over soon so please email the results through as soon as possible so details can be analysed before Dr Grau’s visit.

Email: wairataforestfarm@farmside.co.nz
Submit online via http://www.hazelnut­growers.org.nz.
Post: Bag 1070, Opotiki 3162

A recording sheet and information about how to record your flowering information are available in the members’-only section of http://www.hazelnut-growers.org.nz.