Document Sample
TWENTY YEARS Powered By Docstoc
of tree breeding at the CSIR
                                                     Hectares in 000 ha.


                                                                                                                                                       First plantation


                                                                                                                                                       Large scale commercial plantations

                                                                                                                                                       Silvicultural research

                                                                                                                                                       First government sawmill

                                                                                                                                                       Timber investigations

                                                                                                                                                       First paper mill
                                                                                                                                                       Provenance research

                                                                                                                                                       Hydrological research
                                                                                                                                                       First integrated pulp and paper mill

                                                                                                                                                       Wattle breeding

                                                                                                                                                       Pine breeding
                                                                                                                                                       SAFRI Tree Improvement
                                                                                                                                                       Eucalypt breeding

                                                                                                                                                       Poplar breeding (Lion match)

                                                                                                                                                       Pine hybrids

                                                                                                                                                       Eucalypt hybrids

                                                          Approximate expansion of plantation forestry in South Africa
                                                                                                                                                                                   CSIR Tree Improvement

Adapted from van der Zel and Brink, 1980
                          Foreword by past leaders of the CSIR Tree Improvement Team......................................................................................................2

                           1            Finding suitable trees for South African conditions............................................................................................................4

                           2            Transforming the value of the plantation ............................................................................................................................6

                           3            Mitigating risks through genetics.........................................................................................................................................8

                           4            Trees for marginal areas ....................................................................................................................................................10

                           5            Overcoming hybrid hurdles ...............................................................................................................................................12

                           6            Producing early and abundant seed ................................................................................................................................14

                           7            Picking up the pace: accelerated breeding .....................................................................................................................16

                           8            The polyploid journey of discovery ..................................................................................................................................18

                           9            Clones versus seedlings – making informed choices for deployment ...........................................................................20

                          10            Tree improvement is data-intensive ...................................................................................................................................22
 Did you know?            11            Sharing knowledge ............................................................................................................................................................24

The total increase in     12            From the CSIR to the world................................................................................................................................................26
the forestry plantation
afforested area over      13            Social impact ......................................................................................................................................................................27
the last 20 years in
South Africa was only     14            Looking to the future...........................................................................................................................................................28
8.3%. Compare that
                          Development of our skills ..................................................................................................................................................................29
to an overall increase
of 69.5% in production    References..........................................................................................................................................................................................30
from these plantations
(South African Forestry   Our people: Past and present ..........................................................................................................................................................32
Industry Facts 2007/
2008).                    Concluding remarks............................................................................................................................................................................n

                                                                                                                                                               1        TWENTY YEARS OF TREE BREEDING AT THE CSIR
                                                by past leaders of the CSIR
                                                Tree Improvement Team

                                                Classical tree breeding laid the
                                                foundation for CSIR programme
                                                TREE BREEDING RESEARCH in South Africa started around 1947 with the
                                                establishment of the Wattle Research Institute (now the Institute for Commercial
                                                Forestry Research). Although Dr Hilmar Luckhoff planted a Pinus patula trial in
                                                the early 1950s, the government tree breeding research only formally started
                                                with the establishment of the D.R. de Wet Tree Breeding Station near Sabie in
                                                1958. This was soon followed with three more stations, namely Futululu,
                                                Saasveld and J.D.M. Keet. The early years saw selection of plus trees, seed or-
                                                chards and the planting of some progeny trials.

                                                I was fortunate to join tree breeding in 1970 at J.D.M. Keet. Dr Max Hodgson
                                                was in charge there and both he and Paul Lange from Saasveld encouraged me
                                                to study further in the United States. After completion of my studies, I was full of
                                                enthusiasm to “put South African tree breeding on the map”.

                                                Those were wonderful years as we could accelerate tree selection and plant
                                                many progeny tests to really implement true classical tree breeding on the
                                                principles of population/quantitative genetics. It all laid a wonderful foundation
                                                for the tree breeding programme of the CSIR, which started in 1990 with the
                                                transfer of the South African Forestry Research Institute to the CSIR.

                                                Prof Gerrit van Wyk
                                                Retired Professor of Forestry, University of Stellenbosch.
                                                CSIR, 1990 -1995.

                                                                                     What they do not tell you
Focus, commitment and dedication                                                     about tree breeding
leads to international recognition and                                               IT IS WELL-KNOWN that tree breeding requires long-term investment and
credibility                                                                          excellent technical skills. However, there is something extra, which is both rare
                                                                                     and difficult to foster and achieve, which is key to the success of a tree breeding
THE CSIR WAS in a very fortunate position when, in 1990, the forestry                programme and of a tree breeder.
research function and research assets of government (South African Forestry
Research Institute) were transferred to the CSIR, to form a new forestry division,   That “something extra” has to do with the personality of the breeders and the
Forestek. The classic tree breeding function had been expanded to include            supporting organisation. In order to achieve the long-term goals, the breeders
aspects of biotechnology, vegetative propagation and forest botany. During the       have to match breeding challenges with their own long-term dedication. This
early years at the CSIR, we still maintained the forestry herbarium and xylarium,    requires great personal sacrifice and therefore the need to be motivated to
and housed Dr Poynton’s programme, which was already developing the third            invest half a personal lifetime to such a cause – that is, to have PASSION for
volume in the Tree Planting in Southern Africa-series (finally printed in 2010).     what they achieve.
Excellent research, technical and field staff kept the momentum going in the
various breeding programmes, and we were able to build on the solid founda-          The CSIR tree improvement team has historically had a very low turnover of
tion laid by our predecessors.                                                       staff – many reaching in excess of 40 years excellent service. We owe each
                                                                                     one of them a big “thank you”. We also owe the organisation and industry a
Through focus, commitment and dedication, the CSIR’s tree breeding team has          big thank you for creating an environment which allowed the passion and pride
produced and continues to produce the goods and aspects of the research              to flourish, and recognised and rewarded the good work.
highlighted in this publication. There are many new eucalypt hybrids being
tested, evaluated and deployed nationally and internationally. Development of        We can reflect on various technical challenges and how we attempted to ad-
the low-splitting E. grandis population has significantly increased its value as a   dress them, and the successes and failures along the path. However, I would like
solidwood species. In addition, the scientific knowledge base has been in-           to suggest that we dedicate this booklet as a small and humble tribute to all the
creased and underpins tree breeding and the programme. The researchers,              good work, pride and passion that each member of the team, both past and
technologists and field staff enjoy international recognition and credibility.       present, has contributed towards tree breeding.

It was an absolute privilege to have been part of the team and, for a short time,    Thank you all, and well done.
to have led the team at the CSIR. May you go from strength to strength.
                                                                                     Dr Steve Verryn
Prof Colin Dyer                                                                      Former research group leader, Principal researcher and NRE Fellow,
Director: Institute for Commercial Forestry Research, CSIR, 1990 - 1998.             CSIR, 1990 - 2010.

                                                                                                                             3     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Finding suitable trees for South African conditions
Plantation forestry relies on exotics                                              Of the 23 eucalypt and 28 pine species that have been tested by SAFRI and the
                                                                                   CSIR, only eight eucalypt and eight pine species are now being managed as
South Africa’s forestry industry is almost exclusively based on three exotic       breeding populations by the CSIR.
species, namely, pines, eucalypts and wattle. The founder of the first European
settlement at the Cape of Good Hope in 1652, Jan van Riebeeck, was the
                                                                                                                COMMERCIAL GERMPLASM
first to realise that indigenous forests were not growing fast enough to meet
the demand for timber and he imported seed from Europe and India to plant
in South Africa. Since then, the quest was to find exotic trees that are adapted                             PRE-COMMERCIAL GERMPLASM
to local conditions and suitable for plantation forestry.
                                                                                                                 BREEDING POPULATION
In later years, foresters searched far and wide for suitable species. However,
not all species tested were suitable for South African conditions.
                                                                                                  Eucalypt hybrids                  Pine hybrids
Formal tree breeding in South Africa was initiated in the 1950s. In 1957, the                        E.grandis F 4                     P.patula F 3
South African Forestry Research Institute (SAFRI) started with the improvement                    E.camaldulensis F 3                 P.elliottii F 2
of pines. The programme is built on the large scale introduction of foreign                         E.urophylla F 2                   P.radiata F 2
plantation species in the 18th and 19th centuries and the provenance research                      E.tereticornis F 1                  P.taeda F 4
which had started in 1927.                                                                            E.saligna F 1                  P.caribaea F 1
                                                                                                     E.globulus F 1                 P.maximinoi F 1
The South African government continued with its tree improvement programmes                         E.resinifera F 1                  P.oocarpa F 1
until 1990 when SAFRI was transferred to the CSIR. The research and breeding                      E.longirostrata P 0              P.tecunumanii F 1
programmes of SAFRI – together with the research of the National Timber
Research Institute (NTRI) that was transferred to the CSIR in 1983 – were
joined to form a new CSIR operating unit, Forestek. Today the tree improvement
research group is situated within CSIR Natural Resources and the Environment,
with offices in Pretoria, Durban and Nelspruit.

Testing the suitability of trees                                                                                                                Did you know?
Since 1886, about 1 000 species were tested in species trials or arboreta es-         recommended that plant material should
                                                                                                                                              Did you know that
tablished throughout the country, assisting researchers to select the most appro-     be collected from the area in which the
                                                                                                                                              the first government
priate species for South African conditions. The species which performed well in      species is to be planted, as these prove-
                                                                                                                                              plantation was
one or more arboreta were tested on a variety of sites to allow for evaluation        nances generally perform best. The trial
                                                                                                                                              established in 1876
under different planting conditions. However, as many species have a wide nat-        at de Hoek in the Magoebaskloof area is
                                                                                                                                              in Worcester to supply
ural geographical distribution, different strains or races are adapted to different   still standing (on the land of Komatiland
                                                                                                                                              timber and fuel-wood
climatic conditions. For this reason, provenance trials have been established to      Forests) and is a potential source for both
                                                                                                                                              for the steam train
determine the best provenances within a species. Provenance refers to the origi-      conservation and selecting genetic
                                                                                                                                              running between that
nal geographic source of the seed.                                                    material.
                                                                                                                                              town and Cape Town?
                                                                                      The disc of an 1 800 year-old indigenous                With annual rainfall of
Historically, SAFRI had established a large number of species and provenance
                                                                                      yellowwood from Bluelilliesbush adorns the              only around 175 mm,
trials. These trials were also undertaken to identify trees suitable for specific
                                                                                      entrance to the head office of CSIR Natural             the eucalypt plantation
niche products or for specific environments. Examples of provenance trials for
                                                                                      Resources and the Environment in Pretoria.              had to rely on irriga-
niche products include a eucalypt furniture species/provenance trial established
                                                                                      And it is still admired by young tree improve-          tion. A growth rate of
by SAFRI in Frankfort and species and provenance trials of indigenous species
                                                                                      ment researchers like Rochelle Parsons and              5.7 m3 per hectare per
planted to evaluate wood for fuel-wood purposes. The CSIR continued to estab-
                                                                                      Nuveshen Naidoo.                                        annum was achieved.
lish new progeny and provenance trials.

Potential of indigenous species
Indigenous yellowwood (Podocarpus falcatus) is one of the few indigenous
species that shows potential for commercial planting. Although its growth
rate cannot compete with the exotic species, it is harvested for high-value
furniture timber. In 1982, the then Department of Water Affairs and
Forestry and SAFRI established a number of Podacarpus falcatus
provenance trials. In 1997 the CSIR assessed the trials and

                                                                                                                              5        TWENTY YEARS OF TREE BREEDING AT THE CSIR
Transforming the value of the plantation
It has become increasingly important to grow high-value crops, rather than
large quantities of low-value timber with a high degree of waste. The CSIR, in
                                                                                       Breeding high-value Eucalyptus hybrids
conjunction with a number of forestry sector partners, has been involved in a          for the saw-timber industry
number of projects aimed at increasing the value of the plantations in the last
two decades, in particular, for the pulp and paper and the saw-timber sectors.         In 2007, the CSIR and Merensky initiated a strategic cooperative research
                                                                                       project to accelerate the breeding of high value Eucalyptus hybrids for the
Improving solid wood properties:                                                       saw-timber industry using the best available germplasm and quantitative
                                                                                       genetic and selection methodologies.
the veneer and sawmill study
                                                                                       The vision was to develop new generation germplasm with highly-desirable
The juvenile-mature ‘Florida study’ investigated the importance of traits for the      growth, adaptability and saw-timber wood properties (particularly low splitting)
production of saw-timber and veneer in E. grandis, with CSIR researchers work-         to meet market demand for greater volumes of high quality products. At the
ing in partnership with Shell Forests (UK), Fletcher Challenge (New Zealand)           same time, the project would obtain even greater genetic improvements than
and the South African forestry companies Merensky, SAFCOL and Mondi.                   those achieved with pure species breeding. Researchers used advanced genera-
The mill study showed that log-end splitting, diameter at breast, height and brittle   tion elite selections of E. saligna and E. urophylla to serve as parents for the hy-
heart are the traits that have the greatest influence on the value of eucalypt solid   brids. The selections were based on growth and wood property assessments.
wood products.

Former research group
leader for the tree improve-
ment group at the CSIR,
Dr Steve Verryn, was
involved in addressing
the problem of log-end
splitting in the sawmill study.

                                                                                                                                              Did you know?

                                                                                                                          The total increase in the forestry planta-
                                                                                                                          tion afforested area over the last 20
                                                                                                                          years in South Africa was only 8.3%.
                                                                                                                          Compare that to an overall increase of
                                                                                                                          69.5% in production from these planta-
                                                                                                                          tions (South African Forestry Industry
                                                                                                                          Facts 2007/2008).

Low splitting E. grandis germplasm and                                               Improving pulping properties
seed orchard                                                                         After eight years of research, Project Pulp has produced the first eucalypt hybrid
                                                                                     clone in South Africa to be submitted for protection by Plant Breeders’ Rights.
Five E. grandis clones with low log-end splitting and low brittle heart were
                                                                                     The project started in 2002 as a partnership between the CSIR and NCT
developed as a high-value timber resource. The objective was to produce
                                                                                     Forestry Co-operative Limited. Since 2006, the Department of Science and
rapidly-growing E. grandis material suitable for good quality, economic, solid
                                                                                     Technology’s Innovation Fund has funded the project in order to develop novel
wood processing. The clones have rapid growth and more stable wood than
                                                                                     Eucalyptus hybrids with improved fibre quality and high pulp yield. This project
current eucalypt resources due to the very low log-end splitting. There is also
                                                                                     has added significant value to the resource planted in limited forestry land in
little or no loss of plantation wood due to brittle heart in the clones, producing
                                                                                     South Africa. To date, 18 trials with hybrids produced by the project have been
potentially even better value veneer products.
                                                                                     planted in Limpopo, Mpumalanga and KwaZulu-Natal.

                                                                                     A number of promising new clones have been identified, some of which will be
                                                                                     in a pre-commercialisation phase in the near future.

The CSIR’s tree breeder,                 Interlocking grain (one of the main                          One of 18 trials of
Dave Steyn (retired in 2003),            traits influencing the veneer value of                       Eucalyptus hybrids at
on some trees felled for the             E. grandis), is a good predictor for                         Nseleni established as
veneer and sawmill study                 brittle heart and log-end splitting.                         part of Project Pulp to
in a trial at Frankfort.                                                                              improve fibre quality
                                                                                                      and deliver higher pulp

                                                                                                                             7     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Mitigating risks through genetics
Tree breeding involves long-term research, but
risks present themselves rapidly, often within a                    TY PI CAL                  Fire: Out of the ashes
growth cycle. Newly-introduced or evolved pests                     THREAT S
                                                                                               In 2008, fierce fires raged through Mpumalanga, devastating the CSIR’s
and diseases pose a constant threat.                                  BIOTIC                   Mapiep breeding archive, with a severe setback to many projects. In the case of
                                                                       Pests                   reproductive biologist Amith Sivlal, two years of fieldwork for his MSc went up
In addition, the rapidly changing environment cre-                   Diseases                  in flames. With support from the CSIR, a massive salvage operation began; it
ates new threats and amplifies existing ones to                                                included conditioning of the orchard, recovery of eucalypts and regrafting of
the industry.                                                                                  pines. Turning disaster into opportunity, Amith changed the focus of his study,
                                                                                               and managed to make important findings in his research on improving the
Nature manages risk                                                   Drought
                                                                                               success rate of controlled pollinations. While Mapiep is now in a better condi-
                                                                                               tion than before the fire, the threat posed by fires has clearly shown the need for
through diversity, and                                                 Frost
                                                                                               not just preventative measures, but mitigation as well.
so do we
                       Species              Species that are currently not widely used still hold
                       conservation         genes that may be required in the future.

  Conservation         Stratified           The CSIR implements breeding strategies that maintain
  of Diversity         breeding             diversity in the breeding populations while not compromising
                       strategy             on the genetic gains of the deployed improved trees.

                       Tree banks           In vivo conservation through maintaining trials.

  Creating             Hybrids              Hybrids allow for a rapid response to the sudden introduction
  Diversity                                 of risks (refer to hybrid breeding).

  Deployment           Cloning              Cloning allows highly tolerant individuals to be rapidly

Disease: dealing with Fusarium                                                                      0.25

Fusarium circinatum (pitch canker) has within the last two decades become the                       0.20

                                                                                    Volume (m 3 )
most important nursery disease affecting pines, particularly Pinus patula.
The industry has responded with a multi-pronged plan including:
• Increased nursery hygiene;                                                                        0.10
• Identification of Fusarium-tolerant genotypes and families; and
• The creation of hybrids to introduce Fusarium-tolerance.                                          0.05

Serendipity: The Pinus elliottii x                                                                  0.00
                                                                                                           P. ell x P. car          P. ell
P. caribaea case                                                                                                                (Commercial)
Pinus elliottii x P. caribaea hybrids were first deployed due to superior growth
over P. elliottii on P. elliottii sites. While the new hybrid performed well, the
extra expense of control pollinations prevented its deployment on a large scale.
However, more recently, the hybrid has become very popular because of its
improved disease-tolerance compared to certain other commercially-deployed
pines. Fortuitous results – though unintended – highlight not only the unpre-
dictability, but sometimes the helping hand of nature.

                                                                                                                                  9     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Trees for marginal areas
With less land available for plantation forestry, industry has been forced to      Tree breeders explored the potential of the production of interspecific hybrids
expand into dryer and colder sites.                                                for marginal sites. Interspecific hybrids reunite two genomes that have evolved
                                                                                   separately and therefore have accumulated a series of differences. The resulting
Backed by a solid research foundation, forestry can still provide commercial       heterozygosity offers many opportunities to the breeder. For example, the
returns on sites with low and unpredictable rainfall, poor soils and potentially   exploitation of hybrid vigour, complementarity of parental genes (e.g. combining
cooler temperatures, while the spectre of climate change will challenge the        the volume growth of E. grandis with the drought tolerance of E. camaldulensis)
suitability of current species for traditional plantation areas.                   and adaptability to a range of environmental conditions. Hybrids therefore play
                                                                                   an important role in addressing the need for germplasm that is able to adapt to
Most mainstream plantation species that are currently available are adapted to     the rapidly changing environments that result from climate change as well as
higher rainfall zones – for example, both Eucalyptus grandis and Pinus patula      creating a resource for marginal areas.The success of hybrids in marginal areas
prefer a mean annual precipitation of more than 1000 mm – making droughts          is illustrated by the performance of E. grandis x E. tereticornis hybrids in a trial
a significant threat to the industry. Tree breeding researchers are therefore      series in Swaziland. On a high rainfall site at Mdzimba in western Swaziland,
continuously searching for forestry trees adapted to lower rainfall.               E. grandis was the top performer, whereas on the dry site at Manyonyaneni
                                                                                   in eastern Swaziland, E. grandis did not survive. In the dry site the hybrid
Producing hybrids for marginal sites                                               outperformed both parents.

An international collaborative project funded by the Australian Centre for
International Agricultural Research focused on the development of hybrids for
marginal areas.

CSIR tree geneticist Cyndi
Snedden at a field visit to
an 18-year-old eucalypt
hybrid trial at Prinsloosrust,
a Merensky plantation
in Limpopo.
                                                                                                                                     Did you know?
                                                                                                                                   Did you know that
                                                                                                                                   there are more than
                                                                                                                                   700 eucalypt species?
                                                                                                                                   Only about 12 of these
                                                                                                                                   are of commercial
                                                                                                                                   importance. Eucalyptus
                                                                                                                                   grandis x E. urophylla
                                                                                                                                   and E. grandis x E.
                                                                                                                                   camaldulensis are con-
                                                                                                                                   sidered to be the most
                                                                                                                                   important commercial
                                                                                                                                   eucalypt hybrids.

                                                                  An illustration of the relative expression of traits for various eucalypt species.
                    54 months
             15                                                    SPECIES                Vigour      Drought        Cold      Disease      Rooting     Termite Snout Beetle
                                                                                         (Growth)    Tolerance    Tolerance   Resistance    Ability    Resistance Resistance

                                                                   E. camaldulensis         x           xxx          xx          xx           xxx           x           xx
Height (m)

                                                                   E. grandis              xxx           x            x           x           xx            x           xxx
                                         42 months
              5                                                    E. nitens               xxx           x          xxx          xx            x            x           xx

                                                      30 months    E. tereticornis          xx          xx            x          xxx          xx           xxx          xx

              0                                                    E. urophylla             xx          xx            x          xxx          xx           xx           xx
                      High             Intermediate      Low
                     rainfall             rainfall     rainfall
                    >1250mm              >750mm       >250mm

                                                                  Hybrids: a success story
                  E. grandis
                  E. grandis x E. tereticornis                    In a 19-year-old Eucalyptus camaldulensis provenance trial at a dry site at Salique in Limpopo,
                  E. tereticornis                                 with a mean annual precipitation of 760 mm, the trees showed poor vigour and poor stem form
                                                                  as opposed to the performance of a 10-year-old Eucalyptus camaldulensis hybrid trial at the same
                                                                  site. In contrast to the E. grandis control in the latter trial that did not survive a drought spell, the
                                                                  E. grandis x E. camaldulensis hybrids survived and thrived, offering a much-needed resource in
                                                                  this wood-denuded area.

                                                                                                                               11      TWENTY YEARS OF TREE BREEDING AT THE CSIR
Overcoming hybrid hurdles
Breeders are faced with many challenges in the quest to develop hybrid varieties as the union
of the parental genomes that have evolved separately within a hybrid poses many problems.                                                         Thesis
The differences between the genomes often severely affect the reproductive potential, while
incompatibility and low fertility hamper the propagation of hybrids. Only a small percentage of
viable hybrids will show hybrid vigour or favourable combination of characteristics. These few
                                                                                                        Segregation in F2 hybrids
individual trees have the potential to surpass the pure species and are therefore highly sought
after by breeders. It is, furthermore, difficult to predict the performance of hybrids as appropriate   Commercial deployment of the F2 Pinus elliottii
quantitative genetic models have not yet been developed for hybrids.                                    x Pinus caribaea hybrid could contribute substan-
                                                                                                        tially to cost-effectively meeting the operational
The CSIR’s tree improvement group has initiated a number of studies to investigate ways to              demand for the Pinus elliottii x Pinus caribaea
overcome the barriers to successful hybrid production.                                                  hybrid. The study investigated the segregation in
                                                                                                        a nursery trial of F2 Pinus elliottii x Pinus caribaea
                                                                                                        hybrid seed from elite F1 hybrid germplasm
                                                                                                        developed for South African conditions. Varia-
                                                                                                        bility within the F2 hybrid was compared to the
                                                                                                        F1 hybrid as well as Pinus caribaea and Pinus
                                                                                                        elliottii pure species controls. The study showed
                                                                                                        increased levels of segregation in the F2 genera-
CSIR tree geneticist                                                                                    tion to that of the F1, indicating that if operational
Rochelle Parsons                                                                                        deployment of the F2 seed is considered, then the
assessing F2 P. elliottii                                                                               uniform performance found in the F1 is not likely.
x P. caribaea seedlings
in a nursery trial.


                                                                         Prediction of clonal performance
                                                                         in eucalypt hybrids
                                                                         The inheritance in hybrids is more complex than that of pure species and no
                                                                         suitable genetic models have been developed. It is therefore difficult to predict
                                                                         the ranking of progeny or clones. This is essential for the selection of parents;
Reproductive barriers in the                                             the production of seed; or the collection of material for cloning.
production of hybrid seed                                                CSIR tree geneticist Marianne Hettasch investigated the extent to which pre-
                                                                         dicted performance of eucalypt hybrid clones corresponds with actual realised
In recent years, there has been an unprecedented demand for
                                                                         performance in clonal trials in order to assess the suitability of the current
the P. elliottii x P. caribaea hybrid, due to its improved growth,
                                                                         selection methodology.
desirable wood properties and improved tolerance to Fusarium
                                                                         The results indicate that the current techniques are sufficiently reliable to predict
                                                                         performance, but that the predictions are more difficult to make the more distant
However, the F1 P. elliottii x P. caribaea hybrid is very difficult to
                                                                         the relationship is between the parents. The correlation between the predicted
produce as the cross typically has low seed set and poor germi-
                                                                         and observed performance decreases with genetic distance between parents
nation of full seed. The low seed set in the F1 hybrid thus makes
                                                                         and therefore the ability to correctly predict performance is diminished, irrespec-
the hybrid very costly and difficult to mass-produce for commer-
                                                                         tive of which technique is used.
cial deployment. In this study, various aspects around the repro-
ductive biology of the hybrid parents were investigated to obtain        Correlations of realised versus predicted clonal performances in E. grandis, E. grandis x
a better understanding of the mechanisms involved in successful          E. camaldulensis (GxC), E. grandis x E. urophylla (GxU) and E. grandis x E. saligna (GxS)
reproduction. Several aspects around the pollen biology of P.            for two different prediction methodologies are illustrated. One method consisted of Best
caribaea (male parent) were found to be key factors responsible          Linear Unbiased Predictions (BLUP) (broken line) and the second method (solid line)
for the low seed set in the creation of this hybrid. This includes       consisted of a series of selection indices with varying proportions of weight on individual
the poor viability and storability of P. caribaea pollen and micro-      versus family mean. Better correlations between predicted and realised performances were
                                                                         obtained for E. grandis (blue) than for the hybrids.
bial contamination of pollen grains.
                                                                           Correlation of realised vs. predicted clonal performance






                                                                                                                                            0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
                                                                                                                                                   Weight on individual tree value for stem form versus
                                                                                                                                                             family mean in a selection index

                                                                                                                                                      Indices for GxC                 Indices for GxS
                                                                                                                                                      BLUP for GxC                    BLUP for GxS
                                                                                                                                                      Indices for GxU                 Indices for E. grandis
                                                                                                                                                      BLUP for GxU                    BLUP for E. grandis

                                                                                                                                                                        13   TWENTY YEARS OF TREE BREEDING AT THE CSIR
Producing early and abundant seed
Tree breeders face several technical and biological challenges when producing
seeds for research or commercial purposes. A better understanding of the repro-
                                                                                       Understanding the pollen biology
ductive biology of forest species not only increases the body of knowledge of          of pines and eucalypts
these species but will also potentially make tree breeding efforts more efficient.
The reproductive biology group is therefore providing essential support to tree        Often the flowering times of trees do not overlap and pollen has to be collected,
improvement activities.                                                                processed and stored before controlled crosses can be conducted. Tree breed-
                                                                                       ers have to ensure that the pollen used in controlled crosses is of the highest
                                                                                       quality in order to obtain acceptable quantities of seed. Pollen of most tree
Improving controlled pollination                                                       species can drastically lose viability when handled and stored incorrectly.
                                                                                       Studies conducted by the CSIR have indicated that especially in eucalypts the
Controlled pollination involves multiple visits to trees as the flowers have to be     collection method plays a crucial role in maintaining the viability of pollen
isolated (bagged) to prevent contamination from unwanted pollen, then artifi-          during long-term storage. Recent studies on pine pollen have suggested that
cially pollinated by hand and, during the third visit, the isolation bags have to be   fungal attack on stored pollen can significantly reduce pollen viability, and
removed to allow normal development. In eucalypt seed production, the process          can be a major factor responsible for poor seed production in pine hybrids.
of artificial pollination is labour intensive and costly, especially when producing
commercial quantities of seed. In order to reduce the costs associated with
eucalypt seed production, several studies have focused on reducing the number
of visits through technical advancement of the controlled pollination technique.

CSIR reproductive biologist
Amith Sivlal is using microscopy
to investigate reproductive
incompatibilities in pine

                                                                                                     Ludwick Sekgale,
                                                                                                     technical field assistant
                                                                                                     based in Nelspruit,           Fungi on Pinus
                                                                                                     cleaning eucalypt pollen.     caribaea pollen.

Improving seed orchard practices                                                   Investigating reproductive barriers in
for increased seed production                                                      pine hybrid seed production
The age at which trees become reproductively active and first bear flowers is      Crosses between genetically distant species are in many instances unsuccessful
one of the challenges faced by tree breeders. Both eucalypt and pine trees take    due to incompatibility barriers. Incompatibilities between species normally
several years before flowers can be produced. This means that tree breeders        disrupt the process of pollen germination, fertilisation or embryo development,
have to wait excessively long periods to make genetic advancements. In order       resulting in trees aborting developing fruits or producing empty seeds.
to reduce the generation intervals in trees, chemical treatments which alter the
hormonal activities in trees, have been used to induce early flowering in seed     The CSIR is currently investigating incompatibility barriers in the production of
orchards. The CSIR has recently reviewed several studies on inducing early         a commercial important pine hybrid, Pinus elliottii x P. caribaea. The controlled
flowering in eucalypts, and will be adopting and potentially advancing several     cross in the creation of this hybrid has been well known to have a very low seed
of the reported techniques at the CSIR Mapiep breeding archive at Nelspruit.       set and researchers are using microscopy studies to determine at which point of
                                                                                   the developmental process the incompatibilities occur.
In addition, CSIR researchers are currently generating a wealth of knowledge
on the flowering times for different eucalypt species by frequently monitoring
flower bud development at the breeding archive at Mapiep. This flowering
information helps tree breeders to plan and execute controlled crosses at the
right times. The database also provides a wealth of information regarding the
flowering biology of these species.

                                       A eucalypt flower
Emasculated eucalypt                   is covered with a
flowers are prepared                   pollination bag after
for controlled pollination.            controlled pollination.

                                                                                                                         15     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Picking up the pace: accelerated breeding
Time is money – at least for tree breeders faced with long generation intervals.
Trees are long-lived crops and a typical pulp rotation in eucalypts, for example,
                                                                                    Assessing traits early:
is between seven and 12 years, depending on the site. For pines it is even          studying juvenile-mature correlations
longer. However, a breeder cannot wait this long.
                                                                                    Juvenile selection is practised in forestry based on the underlying assumption
Tree breeders have to investigate all possible avenues to shorten the breeding      that the performance of a trait in a young tree provides an indication of perform-
cycle – the time it takes to breed and select an improved tree – and to get these   ance at maturity or rotation age. The generation interval can thus be shortened
improved trees into production. They are also challenged by the need to             by selecting juveniles at a younger age, but the success of this technique
minimise response time to address urgent problems                                   depends on a high juvenile-mature correlation for the selection traits.
such as new diseases or changed market needs in a
crop with long generation intervals and breeding cycles.                            Selecting individuals at an early age and verifying the selections at a later stage
Tree breeding researchers have developed accelerated                                is more cost effective than delaying selection until advanced ages, depending
breeding technologies, backed by extensive research                                 on the reliability of the genotypic age-age correlations. The key to accelerated
to address these challenges.                                                                                      breeding is to determine the optimum age for selec-
                                                                                                                   tion. CSIR tree improvement group has undertaken a
                                                                                                        number of juvenile-mature or age-age correlation studies with
                                                                                                Eucalyptus grandis, Eucalyptus camaldulensis, Pinus patula and
                                                                                                    Pinus elliottii. In the E. grandis, researchers found high genetic
                                                                                                             correlations for growth assessments at five and 19 years
                                                                                                                     of age. Further research indicated that log-end
                                                                                                                           splitting can be assessed at three years of
                                                                                                                              age in high-splitting areas. However, if early
                                                                                                                                assessment is done at an age when one
                                                                                                                                cannot assess wood property, no progress
                                                                                                                               will be made.
CSIR forester Brian Pierce (right) with Australian client Rod Griffin at the site
of a 20-year-old E. grandis progeny trial where researchers investigated the
impact of 73 traits on saw-timber, veneer volume and value recovery. In
addition, studies of juvenile-mature correlations were initiated in these traits.

Genetic markers – the quest to find                                                     Researchers developed a highly-significant model for the prediction of lignin
                                                                                        content based on brightness, surface growth strain (assessed using a surface
genetic markers for wood splitting                                                      strain gauge) and tree volume assessments in Eucalyptus grandis.

Log-end splitting in eucalypts is one of the most important traits that affect veneer
and sawmill value. In the past, splitting was only assessed at about eight years
                                                                                        Strategies to accelerate breeding
of age and could only be assessed on sites that promoted splitting. Research
                                                                                        The choice of an appropriate and sound breeding strategy is the key to achiev-
focus was then shifted to developing molecular markers (DNA sequences linked
                                                                                        ing sustained genetic gains and delivering an improved tree to the grower.
to the gene/s coding for the trait of interest) for splitting using bulked segregant
                                                                                        The CSIR has and continues to research strategies that will accelerate breeding.
                                                                                        For example, research on E. grandis showed that the cloned breeding popula-
                                                                                        tion produced far higher gains than non-cloned seedlings from the same
Although potential markers were identified, the applicability of the markers to
                                                                                        number of families and individuals per family.
screen the available eucalypt populations proved to be too low. CSIR Bio-
Sciences has subsequently developed a microarray-based DNA amplified
                                                                                        By comparing predicted gains for the production population, deploying tested
fragment polymorphisms in combination with cDNA clones from seven indivi-
                                                                                        clones into production only two years (possibly even less) after selection in the
dual Eucalyptus grandis clones (DNA sequences arrayed onto silanised slides)
                                                                                        cloned breeding population demonstrated potential benefit. Predicted gains
thus continuing the quest to find a genetic way of identifying Eucalyptus trees
                                                                                        for the conversion of the cloned breeding population into a seed orchard
with preferred wood quality traits.
                                                                                        were only slightly lower than the predicted gains for a forward selected clonal
                                                                                        orchard, but the time saved (and therefore, increased gain per year) makes this
Developing models for indirect selection                                                an attractive option for the production of improved seed.

It may be difficult to physically measure a particular trait, for example the lignin
content of the growing tree. Lignin is a complex chemical compound found in
wood and forms an integral part of the secondary cell walls of trees and plants.                                                                             Thesis
Too much lignin is bad news for paper manufacturers as it affects the quality
of the paper being produced. Lignin may be costly and difficult to assess.
In such cases, a genetically-correlated trait may be used for indirect selection.
                                                                                                          Understanding the stability
                                                                                                          of elite clones
                                                                                                          To understand how clones respond to the environment
                                                                                                          helps tree breeders to breed stable clones that will
                                                                                                          perform well on different sites, thereby increasing the
                                                                                                          efficiency of the breeding programme.

                                                                                                          A genotype-by-environment interaction study by forester
                                                                                                          Brian Pierce looked at 27 E. grandis clones planted on
                                                                                                          31 traditional E. grandis sites in South Africa.

                                                                                                          The study showed that the ranking of the clones remain
                                                                                                          the same despite changes in the productivity of the
                                                                                                          E.grandis site. The result from this study were used to
                                                                                                          inform the testing strategy for pre-commercial E. grandis
                                                                                                          and accelerate the delivery of clones for growers.

                                                                                                                               17     TWENTY YEARS OF TREE BREEDING AT THE CSIR
The polyploid journey of discovery
                                                 In the late 1990s, the CSIR’s tree geneticists took the first steps to attempt
                                                 the production of polyploid eucalypts. While many breeders in the horticul-
                                                 tural and agricultural sector have manipulated the ploidy level (chromo-
                                                 some number) of their crop to create varieties with novel traits such as
                                                 increased plant biomass, increased size of fruit and flowers (gigas form)
                                                 and seedlessness, this was a relatively new territory for tree breeders.

                                                 Although there has been interest in creating and exploiting polyploids
                                                 in forestry, specifically for reducing fertility and increasing biomass, the
                                                 successful induction has been limited to few species. There is also little infor-
                                                 mation available on the mature traits of the polyploids, such as their wood

                                                            Tracy Maritz completed her
                                                            MSc on the induction of
                                                            polyploids in eucalypts and      Morphology of leaves of treated
                                                            eucalypt hybrids in 2008.        plants.


          Producing polyploidy eucalypts                                       Identifying polyploids
          A number of methods were compared to identify the most               The second step on the journey was to assess different methods to distinguish
          successful techniques in forestry; these included exposing           between diploids, tetraploids and mixoploids. Eucalypt chromosomes are small
          seeds and buds to a chemical mutagen named colchicine.               and direct counting of chromosomes is difficult. Therefore indirect techniques for
                                                                               polyploidy identification were investigated. Assessment of stomata size proved
          The treatment of eucalypt seed with colchicine was successful:       to be a useful pre-screening tool for polyploid seedlings and bud sports. The size
          one E. grandis tetraploid (four sets of chromosomes), 11             of the stomata is larger in tetraploids than in diploids. The frequency of the stom-
          E. grandis mixoploids (mixture of diploid and tetraploid             ata also decreases with an increase in ploidy level. Stomatal chloroplasts are
          sections), six E. urophylla tetraploids and 11 mixoploids were       also a useful pre-screening tool as the chloroplasts are larger in tetraploids and
          produced. However, bud induction was less successful and             show a dispersed distribution as opposed to the distal distribution in diploids.
          only sectional or mericlinal chimeras were produced.
                                                                               The ploidy of the putative polyploids identified using stomatal size was then
                                                                               confirmed using flow cytometry. Flow cytometry is one of the most reliable tech-
                                                                               niques for the determination of ploidy level.

                                                                               Today the journey still continues, with field trials and crossings between diploids
                                                                               and tetraploids.

                                                          Leaf of a mixoploid with diploid-sized stomata
Stomatal guard cell          Stomatal guard cell          in the top left corner and tetraploid-sized            Chloroplast arrangements in stomatal guard cells in a
length of a diploid.         length of a tetraploid.      stomata in the bottom right hand corner.               diploid (a) and a tetraploid (b).

                                                                                                                   A                          B

                                                                                                                           19     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Clones versus seedlings –
making informed choices for deployment
Making choices between the use of vegetative propagation (cloning) and             • The genetic gains achieved with clonal deployment are higher than for
seed for the commercial production of improved germplasm is a key challenge          seedlings because the entire genotype is captured; and
for tree breeders. The breeder must make an informed choice based on the           • The success of cloning depends on the rooting ability of the tree, which in
genetics of the species, market, economic and time constraints, available infra-     turn depends on its genetic makeup (genotype).
structure, environment and available resources.
                                                                                   One of the fundamental questions affecting the choice between seed and clonal
Clones and seedlings, as a method of deploying improved germplasm in the           deployment is to what degree target characteristics are inherited non-additively
plantation, have both advantages and disadvantages. Examples of these are:         (as opposed to additively). If a trait
• Clones provide greater uniformity than seed. Uniformity has advantages           is largely additively-inherited, then the     100%
   for the harvesting and processing of timber, as well as recovery;               genetic makeup of the progeny                  90%
• Seed is cheaper to mass-produce than clones;                                     (offspring) is largely determined by
• Clones may take longer than seed to develop and test before deployment;          the genes it inherits from the parents.

                                                                                                                                % OF TOTAL GENETIC VARIANCE
• Clones are genetically-uniform and therefore there is a higher risk associated                                                                              70%
   with their deployment as they are not as genetically-buffered to changing       Conversely, high non-additive genetic                                      60%
   conditions as seedlings;                                                        variation is a result of dominance and
                                                                                   epistatic effects of the specific genotype                                 50%

                                                                                   (combination of genes in the individual).                                  40%
                                                                                   It has, for instance, been shown by the
                                                                                   CSIR that the percentage non-additive
                                                                                   (dominant) inheritance is increasing                                       20%
                                                                                   with advancement of the Eucalyptus                                         10%
                                                                                   grandis breeding populations, and this
                                                                                   has important consequences for the                                         0%
                                                                                                                                                                      38     66
                                                                                   plant deployment strategies of com-
                                                                                                                                                                     AGE (MONTHS)
                                                                                   panies.                                                                          %NA         %A


                               Understanding inheritance in Eucalyptus grandis
Non-additive and additive genetic variances as well as broad and narrow           however, an expensive strategy and the increase in cost and time, may not
sense heritabilities were estimated using 38- and 66-month data from a            be necessary if other measures are taken to reduce experimental error and
cloned Eucalyptus breeding population of families derived from open-polli-        increase the heritability of selection traits.
nated selections. The study provided information to assess the appropriate-
ness of strategies. Growth traits were found to be under predominantly-           The choice of the strategy will also be influenced by the economic impor-
additive genetic control.                                                         tance of the traits and the benefit in terms of genetic gain (realising the gain
                                                                                  faster in the plantations) and cost afforded by the reduction in the time to
The current selection strategy, which is based on general combining ability,      deployment of select clones from the cloned-breeding population.
is appropriate in circumstances where the selection traits are under strong
additive control. Cloning and the resultant increase in cost and time may         The benefits afforded by the cloned population are mainly two-fold namely,
not be necessary if other measures are taken to reduce the experimental           an increase in the accuracy of selection; and the time-saving and increase in
error (e.g. reduce weed competition, accurate trial layout in nursery and         gains through the selection of tested clones for deployment at the same time
field and clearly-labelled plots in the field).                                   as selection of the next generation. Deploying clones exploits all the genetic
                                                                                  variation (additive and non-additive).
Disease-tolerance and stem form were, however, found to be under rela-
tively strong non-additive control and a different selection strategy may be      The trend in the distribution of genetic variance in the F1 and F2 families
required.                                                                         indicates a higher proportion of non-additive variance in the F2 families for
                                                                                  all traits except disease which had not been selected for in previous gene-
One of the main advantages of the cloned-breeding strategy is that                rations. This may indicate that, with advanced generations of breeding in
cloning facilitates the more precise assessment of the genotypic differences      this population of E. grandis, gains achieved through selection for additive
between individuals within families, as ramets of clones do not have the          variance will decline, compared with the one achieved in previous genera-
genetic variation that exists among seedlings.                                    tions. The lower narrow sense heritabilities in the F2 in this study support
                                                                                  this indication. A strategy for future generations that exploits the non-additive
Clone means are available to assess the genotypic differences between             variance may be appropriate.
individuals within a family. The cloned-breeding population strategy is,

           Did you know?
                                                     To what degree is rooting
          The CSIR’s tree improve-                   ability genetically controlled?
          ment research group
          developed a determinis-                    Eucalyptus grandis x Eucalyptus longirostrata is a promising cross. However, as with most hybrids, their
          tic gains modelling tool,                  production through controlled pollinations is expensive. While deployment of hybrids as rooted cuttings
          G-assist? The group is                     allows the capture of additive and non-additive genetic variation of superior genotypes, poor rooting
          encouraged by the                          can invalidate the perceived cost-competitiveness of clones. In a study on this new hybrid, the CSIR and
          relative accuracy of                       industry partner NCT, attempted to quantify the influences of genetic and environmental effects on the
          G-assist in predicting                     rooting success of adventitious cuttings. The study showed that while the inclusion of rooting as a selec-
          genetic gains.                             tion trait can improve productivity, greater gains can be realised through optimising the production

                                                                                                                          21     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Tree improvement is data-intensive
Managing the data- and knowledge-                                                    • Increase efficiencies by integrating tree breeding and data management
intensive nature of tree improvement
research                                                                             Harness the power of information
The science of tree improvement is by nature data-intensive. Statistical synthesis
                                                                                     technology (IT) for tree improvement
of data provides insights into the genetics of trees. The data of both current and
                                                                                     A process of ongoing IT development has been integral to the team’s research
previous generations of breeding are invaluable to an array of studies.
                                                                                     • In the early 1990s, a genetic ranking tool was developed to harness com-
Appropriate software is needed to:
                                                                                        puter power in order to handle very large data sets. It was called Matgen –
• Effectively manage large quantities of data which are
                                                                                        short for matrix generator.
  - multi-year in nature
                                                                                     • With the emergence of new web infrastructure and open-source software,
  - collected from many sites globally
                                                                                        TreeBase was initiated in 2007. TreeBase enables researchers to manage
  - interrelated in a complex manner (e.g. complex pedigrees);
                                                                                        information regarding reproductive biology, trial information, data, intellec-
• Manage the intellectual property of genotypes;
                                                                                        tual property, literature and pedigrees, and is accessible to clients and
• Provide real-time access to information for companies collaborating with the
                                                                                        partners globally.
                                                                                     • A new generation of integrated open-source analytical tree improvement
• Effectively manage knowledge over long periods of time given the long-term
                                                                                        tools is in an advanced stage of development. The Mthiwolwazi package
  nature of tree breeding; and
                                                                                        (Xhosa-name for ‘tree of knowledge’) will further improve the efficiency of
                                                                                        one of the core processes in tree improvement research.
     Did you know?

    You can access over
    600 CSIR tree breeding
    reports and publica-
    tions from the last 20
    years, searchable on
    contents, in TreeBase?

                                                                                            Did you know?
First empirical confirmation of negative
impact of collinearity                                                                     In the early 1990s, the CSIR’s mainframe would
                                                                                           run overnight on tree breeding data, and then abort
CSIR tree breeders provided the first empirical confirmation of the potential negative     due to lack of memory? Consequently, a memory-
impact of collinearity in tree breeding, confirming earlier simulation studies of tree     friendly technology – subdividing data into smaller
breeding data and models.                                                                  matrices to do selection analysis (Best Linear Predic-
                                                                                           tion) – was developed, spawning Matgen (matrix
Tree breeders mainly use the selection method called Best Linear Unbiased Predic-          generator).
tion (BLUP). However, problems may be encountered when using data sets and
models with a high degree of collinearity which may adversely affect selection
index predictions.

Simulation studies using Best Linear Prediction (BLP) highlighted that instability aris-
ing from collinearity caused lower genetic gains than alternative techniques in 80%
of the cases.

CSIR researchers used experimental data from a large number of multigenerational
E. grandis and P. patula progeny trials to investigate the occurrence of instability or
collinearity in BLUP and possible methods to deal with it. A series of BLUP predic-
tions (rankings) using three selection traits and 10 economic weighting sets, were

Backward and forward prediction models with different matrix inversion techniques
were used in calculating BLUP indices and two different numerical precisions were
used in the study. Results indicated that it may not always be optimal to use a higher
numerical precision programme when there is collinearity in the data and instability
in the matrix calculations. These simulation studies revealed the need to investigate
the effects of instability in experimental data and possible mitigation methods.

                                                                                                           23     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Sharing knowledge
The tree improvement group has developed five courses. The first course was
presented in 1993. Participants have come from 33 countries world-wide.
                                                                                   Practical data analysis tools for tree
The tree breeding short course
                                                                                   The data analysis course has a duration of one week and enables participants
The two-week tree breeding short course covers all aspects of breeding of tree     to analyse data that is typically encountered in a tree improvement programme,
species, focusing on Eucalyptus and Pinus species. Every year, the course brings   such as elementary trial design, analysis of variance, heritability estimates,
together tree breeders, geneticists, managers and foresters from around the        genetic correlations, repeatabilities, selection using Best Linear (Unbiased)
world; these are people who are interested in sharing and expanding their          Prediction indices, genetic gains prediction and interpretation of results. The
knowledge of tree breeding.                                                        course was developed on request by Stora Enso Celbi and was held for the first
                                                                                   time in 2003 in Óbidos, Portugal. It has also been offered in Durban, South
The course combines theoretical with practical aspects and covers a wide           Africa, in 2004 and 2007.
array of topics such as trial design, trial establishment and maintenance,
vegetative propagation, breeding strategies, quantitative genetics, wood
properties, conservation and biotechnology.

                                                 In 2001 the students visited
                                                 a 100-year-old Eucalyptus in
                                                 the Tzaneen area. The tree
                                                 was approximately 96.7 m
                                                 high and had a diameter of
                                                 1.23 m.

BLUP course                                                               B REEDING PO PU L ATI ON

Best Linear Unbiased Prediction (BLUP) is a statistical procedure
used to optimally predict and rank the genetic value of indivi-
duals. The one-week course covers the advantages and disadvan-
tages of BLP, BLUP and other selection index techniques. Theore-
tical principles are combined with the practical. Participants are                                                     recurrent selection

empowered to apply their knowledge to breeding programmes.
                                                                                                                        seed from females

Tree improvement course for                                                site 1      site 2            site 3                                          site 1     site 2           site 3
The three-day course is specifically designed for managers with                                 elite trees (grafts)                                                         elite trees (grafts)
an interest in tree improvement programmes. The course enables
managers to constructively engage in strategic discussions relating      PRO DU CTI O N
to the opportunities and risks that tree improvement presents for        POPU L AT IO N
forestry companies.

Optimising tree breeding
This one-week course gives a comprehensive review of traditional
tree breeding strategies. The primary objective of the course is to                   seed orchards                                                               seed orchards
empower participants to make informed decisions around the
development, management and implementation of breeding,
production and conservation strategies for their particular environ-
ment. The course has been presented in Portugal and South
                                                                                                         backward selection                  backward selection
Africa.                                                                  PRO GENY
                                                                         TES T S
CSIR tree breeder Dave Steyn, retired in 2003, explains the safe use
of a tree bicycle. Today, tree bicycles are still used to climb trees.


                                                                                                open pollination                             selection

                                                                           Example of a conventional breeding strategy, one of the eight discussed at the course.

                                                                                                                                       25           TWENTY YEARS OF TREE BREEDING AT THE CSIR
From the CSIR to the world                                                                      Built on a solid foundation of 50 years of research and tree breeding
                                                                                                in South Africa, the CSIR produces genetically-improved Pinus and
                                                                                                Eucalyptus species and hybrids with superior qualities. The improved
                                                                                                germplasm is mainly developed for use by the pulp and saw-timber
                                                                                                industry for their nursery production facilities. Germplasm is also made
                                                                                                available to commercial nurseries and researchers from all over South
                                                                                                Africa, Africa, and the world.

Germplasm from
the CSIR is on demand
from countries all over
the world where it is
used for further research
or commercialised.

                                                                                                                    Did you know?
In 2010, the CSIR appointed Vuyisile Hobololo to assist in the commercialisation of its germplasm.                • The Rooikoppies
Here Vuyisile (right) visits the CSIR research nursery in Nelspruit together with clients from York Timbers.        low-split seed orchard
                                                                                                                    and five low-splitting
                                                                                                                    eucalypt clones were
                                                                                                                    registered as invention
                                                                                                                    disclosures by the CSIR
                                                                                                                    in 2008.
                                                                                                                  • The CSIR was the first
                                                                                                                    organisation in South
                                                                                                                    Africa to apply for
                                                                                                                    Plant Breeders’ Rights
                                                                                                                    for a forest tree.
                                                                                                                    An application for the
                                                                                                                    new pulp hybrid CSIR
                                                                                                                    GxN2107 was submit-
                                                                                                                    ted in 2010.

Social impact
Supporting transformation with                                                    The nursery propagates and sells superior eucalypt and pine material to
                                                                                  industry as well as indigenous species for government urban renewal projects.
technology and products                                                           After the CSIR trained these women in cloning techniques, they have possibly
                                                                                  produced the first known clonally produced Acacia karoo trees.
In 2002, the tree improvement team initiated the ’Waste-to-Wood-to-Energy’
project, resulting in the establishment of a small nursery company – Siyatyala    The tree improvement group is also involved with other upliftment projects such
– an isiXhosa name that means ‘we are planting’. The company is owned and         as ‘Seeds for a Better Future’, Kilmo Trust (in East Africa) and the former Depart-
managed by two women and only employs previously-disadvantaged women              ment of Water Affairs and Forestry’s fuelwood project. Another major project
from the impoverished town of Alice and its surrounding villages in the Eastern   is a feasibility study in the Richards Bay area for potential rehabilitation of sand
Cape.                                                                             dunes following mining activities and use by local communities.

    Did you know?
  In 2002, the CSIR donated 900
  eucalypt seedlings to establish
  Siyatyala – a small nursery
  company in the Eastern Cape.
  Five years later they were
  contracted to supply over
  720 000 trees to government
  and industry.

                                                                                                                         27      TWENTY YEARS OF TREE BREEDING AT THE CSIR
Looking to the future
The trees we breed today are targeting the needs and markets
of the distant future. Our challenge is to anticipate and manage                  200
these future requirements. Some potential new demands on                          190
forestry are:                                                                     180
• Trees as a source of bio-energy;                                                170
• Trees for carbon sequestration;
• Climate change and the impact on forests;
• A new social dynamic with respect to land ownership and

                                                                      1980 =100
• Africa as the new frontier; and                                                 130

• The impact of new technologies on both the use of wood and                      120
    breeding.                                                                     110
Increased wood density is anticipated to be important for the                      90
future, as it is positively correlated with higher production of                    1980   1984       1988        1992         1996          2000        2004         2008
fibre per hectare. It is also positively correlated with calorific
                                                                                                          YEARS             Plantation Production         Plantation Area
value, as well as the strength and quality of the wood.
                                                                     This graph shows the need to grow higher quality wood on the increasingly scarce available land
Lastly, the ability to rapidly produce new genotypes which           (courtesy Foresty South Africa).
are significantly different to those traditionally available has
motivated the CSIR’s tree improvement research group to
develop significant hybrid technologies.

                                                                                                                         Former CSIR tree breeder Dave Steyn
                                                                                                                         inspecting a first generation selfed tree
                                                                                                                         (50% homozygous). Tree improvement now
                                                                                                                         has 75% homozygous trees as part of an
                                                                                                                         initiative to explore the potential of heterosis
                                                                                                                         or hybrid vigour through line breeding.

of our skills
The CSIR encourages and supports the development of the skills and formal
qualifications of staff, in line with the vision and research strategy of the tree
improvement research group and the mandate of the CSIR. Over the years we
have produced a number of PhD- and MSc-studies. Many current staff are also
conductiong research towards post-graduate qualifications.

Steve Verryn                                         Tracy Maritz
PhD-thesis: Improving on Best Linear Prediction      MSc-thesis: Induction of polyploidy in Eucalyptus
for tree breeding (1994), University of Pretoria.    species and interspecific hybrids (2008),
                                                     University of KwaZulu-Natal.
Brian Pierce
MPhil-thesis: The influence of the environment       Christopher Komakech
on the volume growth, stem form and disease          MSc-thesis: Assessment of growth potential of
tolerance of Eucalyptus grandis clones in the        alternative Eucalyptus species for mid and high
summer rainfall areas of South Africa (2000),        altitude sites in the summer rainfall growing
University of Stellenbosch.                          regions in South Africa (2008),
                                                     University of KwaZulu-Natal.
Andrea Louw
MSc-thesis: Investigation of the application of      Karen Eatwell
Best Linear Prediction for breeding and clonal       PhD-thesis: Remediation of instability in Best
production purposes in a Eucalyptus grandis          Linear Unbiased Prediction (BLUP) in tree
population (2006), University of KwaZulu-Natal.      breeding populations (In progress), University
                                                     of Pretoria.
Marianne Hettasch
MSc-thesis: Applicability of best linear unbiased    Nuveshen Naidoo
prediction (BLUP) for the selection of ortets        MSc-thesis: Factors affecting rooting of
in Eucalyptus hybrid populations (2009),             adventitious cuttings in Eucalyptus grandis x
University of Pretoria.                              E. longirostrata hybrids (In progress),
                                                     University of KwaZulu-Natal.
Roy Parfitt
MSc-thesis: Stem breakage in Colombia and            Rochelle Parsons
South Africa of Pinus tecunumanii from high-         MSc-thesis: Phenotypic segregation in F2 Pinus
elevation sources (1996), University of Pretoria.    elliottii x Pinus caribaea hybrid seed and
                                                     seedlings (In progress), University of Pretoria.
Cyndi Snedden
MSc-thesis: Broad- and narrow sense heritabili-      Amith Sivlal
ties in a cloned open pollinated Eucalyptus          MSc-thesis: Study of the reproductive barriers
grandis breeding population (2001),                  in production of Pinus elliottii x Pinus caribaea
University of Pretoria.                              hybrid seed (In progress), University of

                                                    29   TWENTY YEARS OF TREE BREEDING AT THE CSIR
                                                 Barros, E, van Staden, CA and Lezar, S. 2009. A microarray-based method
                                                 for the parallel analysis of genotypes and expression profiles of wood-forming
                                                 tissues in Eucalyptus grandis. BMC Plant Biology (9):51.

                                                 Barros, E, Verryn, SD and Hettasch, MH. 2002. Identification of PCR-based
                                                 markers linked to wood splitting in Eucalyptus grandis. Annals of Forest Science

                                                 Dyer, C. 2000. Forestry research in South Africa. In: Owen et al. (eds.) South
                                                 African Forestry Handbook, South African Institute of Forestry, Pretoria: pp.

                                                 Geldenhuys, CJ and von dem Bussche, GH. 1997. Performance of Podocarpus
                                                 falcatus provenances in South Africa. Southern African Forestry Journal

                                                 Gerischer, GFR. 1983. Pulp and paper manufacture in South Africa. In: Oden-
                                                 daal,PB. et al. (eds). South African Forestry Handbook. South African Institute
                                                 of Forestry. Pretoria: pp. 219-227.

                                                 Hettasch, MH. 2009. Applicability of best linear unbiased prediction (BLUP)
                                                 for the selection of ortets in Eucalyptus hybrid populations, University of Pretoria.
                                                 Available at:

                                                 Hohls, DR and Verryn, SD. 2008. Developing a National and International
                                                 Research Community in Tree Breeding through a Web Based Information. CSIR
                                                 Conference, Science Real and Relevant, 2008.

Malan, FS. 1994. The wood properties and quality of three eucalypt hybrids.            Van Wyk, G. and Verryn, S. 2000. The basic principles of tree breeding in
South African Journal of Forestry (167):35-44.                                         South Africa. In: Owen, D.L. (ed.) Forestry Handbook Volume 1, Southern
                                                                                       African Institute of Forestry, Pretoria: pp. 61-68.
Poynton, RJ. 1977. Tree Plantings in Southern Africa. Vol. 1 The Pines. Depart-
ment of Forestry, Republic of South Africa.                                            Verryn, S and Roux, CZ. 1996. A memory-friendly algorithmic strategy for
                                                                                       Best Linear Prediction (BLP). In: Tree improvement for sustainable tropical
Shelbourne, CJA, Hong, SO, McConnochie, R and Pierce, BT. 1998. Early                  forestry. Edited by MJ Dieters, AC Matheson, DG Nikles, CE Harwood and SM
results from trials of interspecific hybrids of Eucalyptus grandis with E. nitens in   Walker Caloundra, Australia:Queensland Forestry Research Institute: pp. 217.
New Zealand. Forestry Science 29 (2):251-262.
                                                                                       Verryn, SD. 2000. Eucalypt hybrid breeding in South Africa. In: Dungey,HS,
Snedden, CL. 2001. Broad and narrow sense heritabilities in a cloned open              Dieters, MJ, Nikles, DG. (eds). Hybrid Breeding and Genetics of Forest Trees.
pollinated Eucalyptus grandis breeding population. University of Pretoria.             Proceedings of a QFRI/CRC-SPF Symposium, 9-14 April 2000, Noosa, Queens-
Available at:             land, Australia. Department of Primary Industries, Brisbane: pp. 191-199.

Snedden, CL and Verryn, SD. 2004. A comparative study of predicted gains               Verryn, SD and Turner, P. 2000. The prediction and selection of E. grandis solid
for selection from a cloned breeding population and the implications for deploy-       wood: Phase one. CSIR Report ENV-P-C 2000-19.
ment. In: Proceedings IUFRO Conference: Eucalypts in a changing world.
Edited by NMG Borralho, JS Pereira, C Marque, M Madeira and M Tome.                    Verryn, SD, Snedden, CL and Parfitt, RC. 2000. Program for the deterministic
RAIZ Institutto de Investigacao da Floresta e Papel, Portugal: pp. 137-144.            modelling of genetic gains of tree breeding and seed and clone production
                                                                                       strategies. Southern African Forestry Journal (189):3-9.
Van der Zel, DW and Brink, AJ. 1980. Die Geskiedenis van Bosbou in Suider-
Afrika. Deel II: Plantasiebosbou. South African Forestry Journal (115):17-27.          Verryn, SD and Snedden, CL. 2000. Optimising the expected genetic gains of
                                                                                       various breeding and selection strategies. In: Proceedings: Forest Genetics for
Van Wyk, G. 1983. The basic principles of tree breeding in South Africa. In:           the next millennium. IUFRO working party 2.08.01. Species Breeding and
Odenaal, PB et al. (eds). South African Forestry Handbook. Gutenberg                   Genetic Resources, Durban, South Africa: pp. 240-243.
Bookprinters, Pretoria: pp. 28-33.
                                                                                       Verryn, SD, Snedden, CL, and Malan, GJD. 2000. The Early Expression of
Van Wyk, G and Roeder, KR. 1977. The status of tree breeding in South Africa.          Splitting in E. grandis. CSIR Report ENV-P-C 2000-085.
South African Forestry Journal (107):54-59.

                                                                                                                             31     TWENTY YEARS OF TREE BREEDING AT THE CSIR
Our people: Past and present
Althea Adey                      Emily Kgosana          Samson Mkhonto     Brian Pierce          Dave Steyn
Letta Badimo                     Christopher Komakech   Simon Mng'omba     David Rabapapi        Louis van der Merwe
Flic Blakeway                    At Kruger              Jack Mokhitli      Melrose Ramokonyane   Gerrit van Wyk
Pat Brown                        Frans Letsoalo         Marelize Moolman   Karien Ras            Lenie Venter
Vourinus Coetzee                 Petrus Letsoalo        Temba Msweli       Wigram Segage         Anzell Venter
Colin Dyer                       Andrea Louw            Joseph Mtembu      Ludwick Sekgala       Di Verhaeghe
Karen Eatwell                    Piet Louwrens          Nuveshen Naidoo    Isaac Shabangu        Steve Verryn
Ruan Espach                      Gert Malan             Andre Nel          Bongani Shozi         Lynnette Wolter
Annabel Fossey                   Tracy Maritz           Jantjie Nolonolo   Amith Sivlal
Peter Geerthsen                  John Mather            Joseph Nyundu      Gerhard Smith
Karin Harding                    Jacob Matroos          Renusha Pandoy     Cyndi Snedden
Annalie Havenga                  Sipho Mdletsha         Roy Parfitt
Paul Hendriks                    Jeffery Melema         Rochelle Parsons
Adam Hendriks                    Johannes Mhlaba
Marianne Hettasch                Matthew Miles
Vuyisile Hobololo
Derek Hohls

Closing comments
                                                                                                            TECHNOLOGY AND HUMAN RESOURCES
                                                                                                            FOR INDUSTRY PROGRAMME
This publication was conceived and designed with the intention
to provide a compilation of impact-driven tree improvement
research at the CSIR over the past 20 years. Commitment and
passion of visionary and dedicated people has made it possible.

The tree improvement publication project team would like to
acknowledge the input and ideas of colleagues in partner orga-
nisations in the forestry sector, without whose cooperation,
commitment and knowledge-sharing, the research may not have
had an impact.

The entire tree improvement team has contributed to the contents
of the publication. In particular, Marianne Hettasch, Rochelle
Parsons, Cyndi Snedden and Steve Verryn have generously given
it their time. The enabling environment created over the past
four years by Dr Pat Manders of the Natural Resources and the
Environment unit at the CSIR has made much of the content come
to fruition. He has also ensured the publication of this document.
In addition, the support of CSIR executive and the financial sup-
port of THRIP are gratefully acknowledged.

The tree improvement team – current and future, and the
research partners and stakeholders with whom it interacts –
are well-resourced to contribute to addressing the challenges
facing the forestry and forest products sector in South Africa
and Africa. From 1990 to 2010, the teams’ research has had
some excellent impact and the years 2010 to 2030 promise to
be as productive, relevant and impact-filled.

Flic Blakeway
Manager: Forestry Competence Area
July 2010

Design and print production: Loretta Steyn Graphic Design Studio
Photography: Loretta Steyn, Dewald Reiners,, i-line films, Monsoon Photography and CSIR
Printed on Symbol Freelife Ivory environmentally friendly paper. The product is manufactured using 50%
Elemental Chlorine Free (ECF), 40% selected recycled post consumer fibre and 10% de-inked recycled fibre.
of tree breeding at the CSIR

Shared By: