Assessment of Macadamia Selections Based on Nut Quality



R.J. Nissen* & R.R. Williams**

Reprint from CMS 1980


Samples of Macadamia nuts were obtained from promising new selections in Queensland and northern New South Wales. These samples were assessed on a range of nut quality criteria. Of the 57 selections assessed, ‘Don’ and ‘Own Venture’ were considered to be exceptionally good whilst 12 others warrant further consideration. The two commercial cultivars, 246 and 508 were sub-standard.


Commercial Macadamia production in Australia is largely based on cultivars selected for their performance under Hawaiian growing conditions. However, the performance of these selections under Australian conditions, particularly with respect to yield, is significantly poorer than in Hawaii. In 1978, the Horticulture Branch of the Queensland Department of Primary Industries intensified its search for Macadamia selections, which perform exceptionally well under Australian conditions.

Macadamia nut growers were solicited to provide samples of nuts from individual seedling trees known to perform exceptionally well and having quality characteristics at least comparable with established cultivars growing in the same locality. Fifty-seven samples were received; predominantly Macadamia integrifolia from Queensland and M. tetraphylla from northern New South Wales.

Nut quality has been accepted as the first criterion for the selection of improved Macadamia cultivars. All the samples received were subject to a quality assessment procedure based on the characteristics and standards established by the Hawaiians and reviewed by Hamilton and Ito (1976) and Hobson (1976). Where a range of values have been suggested for minimum acceptable standards, we have adopted the lower values because these minimum limits, accompanied by a yield increase, would still represent a worthwhile improvement. In addition, we have included evaluations of shell characteristics not previously used.


Drying procedure

The nut samples forwarded by co-operators or collected by the department had been dehusked and air-dried to varying degrees.

These samples were spread on aluminum trays in a drying oven and dried under a draught of ambient air for one week, followed by forced air drying at 38 C for 2 days then 52 C for 5 days. This procedure results in kernel moisture content of approximately 1.2% and avoids kernel damage due to too rapid drying (Hobson 1976).

Shell assessment

After drying, the external appearance of the shell was assessed on the following characteristics.

The categories used were based on an overall comparison of the samples available and are therefore only relative to the material on hand.

Texture_ each sample was classified as either (a) rough, (b) slightly rough or (c) smooth.

Color_ all the samples were examined together and grouped according to the shade of color, i.e. (a) light brown, (b) brown or (c) dark brown.

Shape_ a representative shell from each sample was split in half and the perimeter of the split surface was traced. An acceptable shell is approximately spherical with no obvious angularity.

Flecking_ a relative classification into 3-category (a) slights (b) moderate, or (c) heavy flecking.

Micropyle — all nuts were examined; the occurrence of any open micropyles is unacceptable.

Any other defects or features were also noted. Since there are no established standards for the above characteristics, the sample was considered to be acceptable provided there was no open micropyle and the other characteristics were not extremely divergent from the average.

Nut weight

The number of dried nuts per 1000 g was determined by counting the number in a weighted sub-sample usually 100-500 gms. The acceptable range for this parameter was 121-187 nuts per kilogram (CSR recommendation) with approximately 133 nuts per kilogram producing kernels of the desired weight (Hobson 1976).

Nut size variation

A mechanical device involving a vibrating, inclined plane with diverging base plates was used to separate the nuts according to their diameter. The nuts were collected in 5 size groups; <2.2, <2.4, <2.6, < 2.8, 3.0 cm. The distribution of nuts within these groups was expressed as a percentage of the total number of nuts sampled.

Nut cracking characteristics

After the proceeding, whole-nut assessments, each nut was cracked by striking with a hammer along the suture adjacent to the micropyle. Spoiled kernels, along with their shells, were discarded from the sample and the total weight of their discarded material was recorded. The balance of the sample was assessed as follows.

Shell fracture_ the sample was classified as either (a) shatters or (b) splits, according to the performance of the majority of the nuts.

Sticking kernels_ the number nuts in which the kernel adhered to part of the shell after cracking was recorded and expressed as a percentage of the total number of acceptable nuts.

Kernel breakage_ the proportions of kernels, which remained as (a) whole kernels, (b) halves or (c) pieces were recorded and expressed as a percentage of total kernel weight. The combined proportion of the whole plus half kernels (i.e. useable kernel) was also calculated.

Average, individual kernel weight

The average kernel weight was obtained from the total weight and number of whole kernels. For even roasting characteristics, an average weight of 2-3 grams per kernel is acceptable (Hobson 1976).

Percentage total kernel recovery

This is the total weight of unspoiled kernels expressed as a percentage of their total nut (shell plus kernel) weight. The acceptable limit for new selections, based on the better Hawaiian selections, has been set at 38%. Recoveries in excess of 50% have been reported but these are usually associated with too thin shells.

Percentage No. 1 kernel

No. 1 kernels (including half kernels) are those, which float on tap water, i.e. SG 1.0, which corresponds to a minimum oil content of 72%. The weight of No. 1 kernels is expressed as a percentage of the total weight of whole plus half kennels. A minimum standard of 95% is used based on the Hawaiian experience.

Percentage recovery of No. 1 kernel

This is the weight of No. 1 kernel expressed as a percentage of the total (unspoiled) nut in shell. This figure is derived from the product of percentage total kernel recovery and percentage recovery of No. 1 kernel.

Kernel characteristics

Only kernels from nuts, which were acceptable in the entire preceding test, were subject to this assessment. The characteristics were considered under two-category (a) appearance, (b) quality. Acceptable kernels must conform to the following descriptions (Hobson 1976).

Appearance_ kernels should be almost spherical without any obvious flatness or protuberances. The color should be clear white to creamy with little or no discoloration (grayness) at the apical end. This was assessed by a general observation of the sample and a brief description recorded.

Quality_ fresh dried kernels should be odorless. The flavor should be delicate, mild and uniform, the texture crisp but tender. This was assessed by a tasting panel and was recorded by selecting the appropriate descriptive terms from the following:

Texture: Hard, tender, crisp, mealy, sloppy.

Flavor: Tasteless, bland, sweet, bitter, oily, mild, nutty.

Odor: odorless, slightly-, mildly-, highly aromatic.


Of the 57 samples assessed, 14 were considered to be worthy of further consideration. Their ratings are summarized in table 1.

When determining the overall acceptability of the standard of not quality for a particular sample, a sample, which received ratings above the minimum required standard for all the criteria was considered to be very good. Other samples, which were sub-standard in one or two criteria but otherwise rated well, were considered to be acceptable at this stage of selection. The commercial cultivar 246 failed the assessment but the results have been included for comparison.

Summary of Quality Assesment of Nuts Samples--1978

Table 1

Growers Name &Variety Name




Number of

NutsPer Kilo.


Weight of

Individual Kernels

%of Kernel


% of

#1 Kernels

% Recovery

of useable










CSR + 136 3.1* 39.1 98.6 38.5 ++ +  

Almo Plantation Seedling

+ 165 2.7 46.0 96.0 44.0 -* + Flavor Suspect
Wagner + 250* 1.7* 41.5 98.8 41.0 + +  

NR Greber Heilscher

+ 140 2.8 42.5 98..7 41.9 -* + Flavor Suspect
NR Greber NRGx4 + 136 3.2* 44.8 96.8 43.3 -* + Flavor Suspect

NR Greber GreberHybrid

+ 196* 2.0 40.0 94.2* 38.0 + +  
NR Greber Own Venture + 122 2.7 42.1 100.0 42.1 ++ ++  
NR Greber Armanasco + 148 2.9 37.6* 100.0 37.6 + +  
NR Greber PinkHybrid + 198 2,5 53.3 99.6 53.0 + +  
Ken Field + 125 3.3* 43.0 100.0 43.0 + +  
DPI Trail Nutty Glen + 138 3.2* 44.0 97.0 42.6 + +  

DPI Trail Don

+ 176 2.3 41.4 95.9 39.7 + +  
DPI Trail Rickard + 154 2.5 38.0 94.9 36.0 + ++  
DPI Trail 508 + 196* 2.2 40.5 100.0 40.5 + +  

DPI Trail 246

+ 124 2.8 36.5* 94.4* 34.4* + - Failed

++ very good -unacceptable
+ acceptable * unacceptable factors


Although the trees assessed in this exercise were selected on the basis of their superior performance compared to other trees in their locality, the majority of the nut samples did not have sufficiently high standards of quality to justify consideration of their potential. Two selections, ‘Don’ and ‘Own Venture’ were exceptionally good.

It is worthy of note that the Hawaiian cultivar 246 was not accepted because the recovery rates were sub-standard (Table 1) but c.v. 508 rates highly apart from the high number of nuts per kilogram. When the new selections ‘Don’ and ‘Own Venture’ are compared with c.v. 508, they are only marginally better in quality. Therefore they would need to have substantially greater total yields or some other meritorious characteristics to justify their use to replace c.v. 508.

When considering the outcome of this exercise, two important facts must be borne in mind. First, this assessment is based on single trees sampled at one time. Nut quality can vary considerably from year to year and between locations and growing conditions therefore the order of quality ratings presented in table 1 could be quite different for similar trees under different circumstances.

Secondly, the criteria and standards we have used to assess nut quality are based largely on the experience of the Hawaiian industry. The limits of acceptability are based on what represents a worthwhile improvement over the existing cultivars in Hawaii and they need to be revised in relation to the performance of the best of the current Australian selections. Furthermore, the relative importance of the various criteria needs to be determined bearing in mind the requirements of the producer, the processor and the consumer, both now and in the future. This latter point is important in light of the time required to find and establish productive plantings of new selections

The next step in this selection program is to re-assess the quality of the 14 promising selections in subsequent years and to obtain additional information on yield and tree characteristics, which will determine the final acceptability of the selections. Only trees which, continue to perform well would warrant the investment of resources necessary in an extensive field-testing program.


* Experimentalist, Dept. Primary Industries, Nambour Qld. Australia.

** Horticulturist, Dept. Primary Industries, Nambour Qld. Australia.




Hamilton, R.A. and Ito, P.J. (1976). Development of Macadamia nut cultivars in Hawaii. Calif. Macadamia Society Yrbk. 22:94-100.

Hobson, L. (1976). Macadamia quality — some questions answered. Part II Hobson’s Choice. Citrus and Sub-trop. Fruit J. No. 517: 14-22.