Details of micropropagation technologies have been well reviewed in recent years (Durzan 1988, Durzan and Gupta 1988, Wann 1988, Thorpe 1988, Thorpe et al. 1991, Le Roux & van Staden 1991, Tartorius et al. 1991). The alternative approaches have been grouped into three categories (Lutz et al. 1985, Thorpe et al. 1991), namely:
Multiplication of plants through the sequential subculture of axillary bud explants has been achieved for a large number of plant species, and is the basis of most of the commercial systems. Included in this category is the eucalypt axillary budding system, although some of the multiple shoots arising from the culture of eucalypt axillary buds may be adventitious in origin (Le Roux & van Staden 1991). Although best results for most tree species concern young seedling material, the method has been applied successfully to older trees of certain species, through the culture of adult material in some cases, and through the propagation of coppice material for those species for which such can be obtained. Compared to other micropropagation approaches, multiplication rates are low - five to ten propagules per culture cycle in many of the commercial operations (Lutz et al. 1985). Nevertheless, even a multiplication factor of this order can amount to rates of millions per year for many species (Wang & Charles 1991), and rates for eucalypts are frequently much higher (Le Roux & van Staden 1991). Favourable responses for pines are generally limited to young trees, and multiplication rates are lower (Thorpe et al. 1991).
This includes systems involving direct induction on explants and those involving adventitious budding in callus cultures (Lutz et al. 1985). Well described protocols are the meristic nodule systems for poplar (McCown et al. 1988) and radiata pine (Aitken-Christie et al. 1988). These involve the formation, multiplication and ultimately regeneration of plants from spherical cell clusters which show tissue differentiation and some vascularization (McCown et al. 1988). These are similar to systems reported for a number of herbaceous species (Aitken-Christie et al. 1988). In general, juvenile tissues are much more responsive than adult, and embryos have been the favoured explant, particularly for conifers (Thorpe et al. 1991). For responsive species, multiplication rates through adventitious budding are commonly substantially higher than for the axillary budding route (Wang & Charles 1991). In radiata pine, for example, it was estimated that an embryo of one of the more responsive clones could yield 260 000 plants (ready for the field) in 2.5 years (Aitken-Christie et al. 1988).
In this process, analogous to zygotic embryogenesis, a single cell or small group of vegetative cells is induced to undergo differentiation to form a somatic embryo. This embryo must be “matured” and then germinated to form a plant (Tartorius et al. 1991). Although less commonly achieved than the other technologies, regeneration through embryogenesis has been reported for over 50 woody species, encompassing over 20 angiosperm families, and at least a dozen conifer species (Wann 1988, Tartorius et al. 1991, Attree & Fowke 1991). Included are species of Picea, Pinus, Larix, Abies, Pseudotsuga and Sequoia (Attree & Fowke 1991), and some eucalypts (Watt et al. 1991). Species of Pinus have shown some recalcitrance compared to Picea (Becwar et al. 1988). For most species, in particular the conifers, favourable responses can be elicited from explants from embryonic or young seedling material only (Tartorius et al. 1991, Thorpe et al. 1991, Attree and Fowke 1991). Potential multiplication rates, particularly from cell suspension cultures, are very high - e.g. about 100 embryos per ml of culture and a doubling time of about 48 hours for Picea (Becwar et al. 1988). The rate of conversion to plantlets, however, remains very low - e.g. highest mean efficiency of 4% in the Picea example, and realized multiplication rates are far below potential. Commercial development of somatic embryogenic systems is yet to be reported.
To briefly summarize the alternatives comparatively, the axillary budding, adventitious budding and somatic embryogenesis approaches are characterized by progressively increasing potential multiplication rates, progressively lower numbers of species for which success has been reported, and increasing apparent restriction to juvenile material. Over 1 000 plant species can now be micropropagated (Bajaj 1991), including many forest tree species. Thorpe et al. (1991) listed over 70 angiosperm and 30 gymnosperm tree species for which successful methods for the production of plantlets have been reported, and Le Roux and van Staden (1991) listed over 25 species of Eucalyptus alone. These lists include most of the major plantation species. It is reasonable to conclude that, with sufficient research effort, successful protocols could be developed for most forest tree species. In general, a research project of this type might be described as short to medium term with a moderately high expectation of success.
Micropropagation is now the basis of a large commercial plant propagation industry involving hundreds of laboratories around the world. For example, 65–70 laboratories produced over 53 million plants in 1988 in Holland alone (Wang & Charles 1991). While herbaceous ornamentals remain the major emphasis of the industry, the list of commercially propagated plants does include some fruit trees (e.g. apple rootstocks) and woody ornamentals (Zimmerman 1985). In these species, the technology is being applied largely to the multiplication of highly desirable individual clones. Micropropagation is used because it permits rapid multiplication to meet market demand, and in some cases to overcome difficulties in alternative methods such as propagation by cuttings. In pineapple, for example, release of new varieties typically took 25 years with traditional multiplication procedures, but can be accomplished in one year using micropropagation (Drew 1993).
Although research results indicate that many forest tree species, including the major plantation species, can be micropropagated, widespread application of the techniques in commercial plantation establishment is not yet evident, and there is widespread scepticism among forest managers and tree improvers that such techniques will ever be applied. Possibly the best example of commercial application remains that of Tasman Forestry in New Zealand (Gleed 1992). Most forest plantation programmes are based on seedlings, and a few on cuttings. It is useful then to examine the characteristics of micropropagation systems in relation to the requirements of propagation technologies in commercial forest plantation programmes:
In current value, estimated costs of plants resulting from the axillary and adventitious budding systems vary from about $US 0.20 to $1.50 (Smith 1986, Mascarenhas et al. 1988, Beversdorf 1990, Aitken-Christie 1991), with the perhaps more widely tested costs of the ornamentals tending to lie in the upper half of this range. Forest tree seedling costs vary from about $0.05 to $0.20, and cuttings up to perhaps twice that. The margin between the cost of a micropropagule and those of seedlings or cuttings could be about $US 1.00, using existing commercial technologies. Micropropagules are not expected to be inherently better than cuttings or seedlings of the same genetic quality. To be worth applying then, micropropagation must have the capacity to deliver, to the commercial planting program at a particular time, genotypes which are genetically superior to those which could be delivered as seedlings or cuttings. It is instructive then to calculate the extent of the genetic superiority which would be required in order to produce a gain of $1.00 per plant at establishment. In Tables 4 to 6, these calculations are presented respectively for short (7 years), medium (25 years) and long (60 years) rotation species, under various levels of stocking, productivity, value of wood, and interest rate.
Table 4. Percentage gain to return $ 1.00 per plant at establishment (7 year rotation)
|Stocking (per hectare)|
Table 5. Percentage gain to return $ 1.00 per plant at establishment (25 year rotation)
|Stocking (per hectare)|
Table 6. Percentage gain to return $ 1.00 per plant at establishment (60 year rotation)
|Stocking (per hectare)|
Superiority, in terms of the genetic quality of planting stock available for commercial deployment at a particular time, of micropropagation over alternative propagation methods could result from either a capacity to propagate genotypes (families or clones) which cannot be propagated by other means, or to deliver them more quickly. This principle is illustrated in Figure 11.1, which is a quantitative comparison of some propagation options for the hybrid between slash and Caribbean pines. The solid line in Figure 11.1 represents the progression in genetic quality of the “breeding population” (the progeny trials being established) with time - a stepwise increase each generation resulting from a new round of recombination and selection. The dashed lines represent the genetic quality of stock being planted commercially. These production population lines are derived from the breeding population line by quantifying:
Selection in the breeding population, of the small number of elite families or clones to be deployed commercially. This includes selection for economically important traits such as vigour and straightness, and negative selection for ability to propagate the selections. Selection results in a net genetic superiority of the selected material over the breeding population, and location of the production population line above the breeding population line. It must be noted that the intensity of selection which can be applied is limited by the size of the testing program which can be afforded (the number of families or clones tested), and the by the need to maintain some diversity in the plantations (setting a minimum number of families or clones to be selected and deployed at a particular time).
The time between establishment of trials and the availability, in numbers sufficient for the commercial planting program, of planting stock derived from selections made in those trials. This includes the time required to test and identify the superior families or individuals (the selection interval), and the time required for appropriate multiplication by seed and/or cuttings (the multiplication intervals). With some propagation strategies, the selection and multiplication intervals can overlap. This time factor results in an offsetting of the production population line to the right.
These production population lines are stepped in the same manner as the breeding population line - increases in the genetic quality of material being propagated under a particular strategy will be attained each generation as an improved breeding population in which to select becomes available. Both the selection and time factors obviously influence the net vertical positioning of a production population line relative to the breeding population line, and hence the relative genetic quality of planting stock being deployed commercially at a particular time. Propagation options plotted in Figure 11.1 are: the production of full-sib families by mass pollination with caribaea pollen in monoclonal slash pine orchards; the production of full-sib families by controlled pollination in clone banks followed by multiplication by cuttings; and propagation of superior individual clones, multiplied sequentially by cuttings simultaneously with a clonal test. The first two are proven technologies. The third remains to be proven, although results of experiments examining the extent to which juvenility can be maintained are very promising.
The results of micropropagation experiments conducted with P. caribaea suggest that the proportion of genotypes displaying a satisfactory propagation response is considerably lower than that for cuttings. Micropropagation is therefore not likely to permit the propagation of genotypes which cannot be propagated by cuttings, and the only advantage to be gained will thus be a time advantage. It is evident in Figure 11.1 that, on average, the advantage (in terms of the genetic quality of planting stock being deployed commercially) of a one year shortening of the multiplication interval is numerically equivalent to the average annual rate of improvement in the breeding population - a little over 1% in the pine example used here. For the Figure 11.1 example, the use of rapid micropropagation method might allow the deployment of planting stock of families a year earlier than would be achievable by cuttings, equating to an advantage of about 1% on average. This analysis is somewhat conservative, since it is based only on volume, and does not take into account genetic gains in quality factors, which are likely to be just as important. In Queensland, the slash × Caribbean pine hybrid is planted at about 800 per hectare, and grown on a rotation of about 28 years for sawn timber, with Mean Annual Increment (MAI) between 10 and 15 and stumpage currently about $US 50 per cubic metre. The interest rate generally applied in the Government plantation venture is 4%. Consideration of Table 5 suggests that the genetic advantage of a one year reduction in the multiplication interval will not be sufficient to pay the $1.00 margin at establishment discussed above. Other factors being equal (in particular the maintenance of juvenility through the sequential multiplication procedure), micropropagation also offers little or no time advantage in the propagation of tested clones of this taxon - also insufficient to pay the $1.00 margin. If, however, propagation of superior clones by cuttings were not possible, and micropropagation were, then the margin of micropropagated clones over best full-sib families would probably pay $1.00 per plant at least at establishment.
With many influential factors, it is difficult to generalize such conclusions to other species, in particular because coefficients of variation will vary substantially. For other medium rotation species, a $1.00 margin per plant at establishment is likely to be affordable where stumpages are high (sawn timber product), MAIs are at least 20–30, rapid multiplication by cuttings is not possible, and preferably where interest rates to be applied are low. This might apply to some tropical hardwoods and conifers, grown for sawlogs in tropical areas, and which are difficult to propagate by cuttings. For short rotation pulpwood species, with stumpage say $10 to $20 and higher rates of interest frequently applied, payment of a margin of $1.00 for a propagule to be used in plantation establishment is not likely to be attractive, even where MAIs are very high, and even where other methods of propagation are difficult, e.g. Eucalyptus nitens and E. globulus. For long rotation species, where MAIs would normally be less than 10, payment of a margin of $1.00 per plant at establishment is likely to be marginal at best for general sawn timber production. Table 6 suggests, however, that this should be quite affordable for very high value species.
Analysis in terms of reductions in rotation length will make payment of $1.00 per plant at establishment more affordable, particularly for long rotations. For many purposes, though, wood quality considerations limit the extent to which rotations can be shortened, particularly with softwoods. A general conclusion therefore is that the application of commercial micropropagation, as it exists in the horticulture industry, to the direct production of commerical planting stock for forest plantation species would be quite limited. There are two approaches to this planting stock cost problem:
Cost dilution, through the use of micropropagated plants as stool plants from which cuttings are harvested.
Automation of the axillary or adventitious budding systems. Recent developments in reducing the high labour component in micropropagation by automation have been well reviewed (Aitken-Christie 1991; Tisserat 1991). Although several systems have been developed, commercial use to date is limited and capital costs are high. For pines, the meristematic nodule system is particularly amenable to automation (Aitken-Christie 1991).
Coupling of somatic embryo systems to artificial seed technology. Bioreactors are being developed for a number of propagation systems (Styer 1985, Takayama 1991, Preil 1991, Tisserat 1991), but obviously have particular application to the induction and maturation of somatic embryos. Obstacles to the actual engineering of these bioreactor systems do not appear to be high. Major problems remaining to be overcome are asynchronous development and a high percentage of abnormal embryos (Ammirato 1987, Beversdorf 1990, Watt et al. 1991, Redenbaugh et al. 1991). Artificial seeds have been produced through the sodium alginate and polymer encapsulation of somatic embryos of several crops - celery, Brassica, carrot, cotton, lettuce, alfalfa, rice and corn (Redenbaugh et al. 1991). A unit cost of 0.03 cents (US) per seed reported for alfalfa is very attractive, but low storage life (significant reduction in viability after 7 days) remains a problem.
Successful application of the methods discussed here rests heavily on the assumption that the genotypes propagated are identical to those selected. While axillary budding systems involve methods quite analogous to propagation by cuttings and provide genetic stability, the principle that genetic aberrations can occur with systems involving adventitious budding from callus has been well established (Wang & Charles 1991, Bajaj 1991). Of considerable interest is the extent to which genetic stability is retained during somatic embryogenesis. While the high proportion of morphologically and histologically abnormal embryos remains a major obstacle to the commercialization of somatic embryogenesis systems, there is little evidence to date of genetic aberrations (Thorpe 1988, Tartorius et al. 1991). No isozyme pattern variation was detected among over 1 500 somatic embryos of a single line of Picea glauca-engelmannii (Eastman et al. 1991).
Successful application of micropropagation technologies also rests heavily on the assumption that field performance of the propagules will be acceptable. While plants derived from axillary and adventitious budding in hardwoods have generally performed satisfactorily, the situation with conifers is less clear (Thorpe et al. 1991). Quite detailed studies with douglas fir and loblolly pine have suggested that plantlets display a lag phase with respect to both height and diameter growth, and display some more mature morphological characteristics (Ritchie & Long 1986, Amerson et al. 1985). Plantlets derived from embryos showed signs of advanced physiological age also for radiata pine (Menzies et al. 1991). Plantlets of loblolly pine displayed a lower incidence of infection with fusiform rust (Amerson et al. 1985). Grafting studies for loblolly pine have shown that these morphological effects are features of the shoots themselves, and not indirect effects of differing root systems (Anderson et al. 1992). Data on the field performance of plants arising from somatic embryogenesis are more limited. Planting artificial seeds directly into soil offers the attraction of removal of the acclimatization steps, potentially a major source of physiological or anatomical abnormalities (Tartorius et al. 1991, Thorpe et al. 1991). For Picea, plants arising from somatic embryogenesis have been similar to seedlings with respect to bud phenology (Becwar et al. 1988), growth rate, shoot and root morphology and frost hardiness (Webster and Attree 1990). More limited testing for loblolly pine and douglas fir has suggested that plants are at least normal in appearance (Durzan 1988). Recommendations for the application of a new propagation technology to any particular species should be preceded by specific testing - to demonstrate the satisfactory general performance of the propagule type, and the absence of genotype × propagule type interactions, for propagation strategies involving the deployment of propagules differing from those on which selection was based. As pointed out by Libby (1988), field testing to a level of reliability that will support responsible deployment is not quickly done.
The number of clones safely included in an annual planting program of a commercial plantation species may be as low as 7 to 25, although legislation in some countries dictates that higher numbers be used - 30–120 in Sweden and 20–500 in Germany, with the lower numbers restricted either to special sites or to clones subjected to intensive and lengthy testing (Muhs 1988). It might be reasonable to conclude therefore that the numbers required per clone in a particular planting program would be in the order of hundreds of thousands to low numbers of millions. As discussed above, the higher multiplication rates through micropropagation offer the potential to achieve these numbers more rapidly than is possible through propagation by cuttings. For many of the conifers, the required numbers could be achieved more rapidly with somatic embryogenesis, but the time advantage over a good meristematic nodule system will not be great. For eucalypts, axillary/adventitious budding systems in general give quite adequate multiplication rates. Multiplication rate is therefore unlikely to be an important reason for changing from a successful axillary or adventitious budding system to a somatic embryogenesis system.
The analysis in the maturation state section suggests that, where a reasonable clonal testing program is feasible and affordable, propagation of entirely juvenile material can permit the capture of quite good gains. The frequently observed recalcitrance of mature material to micropropagation technologies, in particular to somatic embryogenesis, is therefore no major impediment to successful application of these technologies in clonal forestry. This is dependent though on successful storage of juvenile material for the duration of the clonal test.
As noted elsewhere in this report, embryogenic cell lines are more amenable to cryopreservation than are cultures involving more organized tissues. This attribute, together with the potential compatibility with artificial seed technology, is likely to be the greatest advantage of somatic embryogenesis as a propagation technology for clonal forestry.
Genetic variation in in vitro responses will result in the imposition of selection for multiplication rate. It appears that such selection could be intense for somatic embryogenesis systems in some species. Although as many as 75% of zygotic embryos of Picea have yielded embryogenic cultures (Becwar et al. 1988), initiation frequencies for Pseudotsuga (Durzan & Gupta 1987) and Pinus (Becwar et al. 1988) have been much lower - less than 3%. As discussed by Haines & Woolaston (1991), some selection for reproductive response can be tolerated without major losses in genetic gain. Reasonably heavy selection can be tolerated where such is applied as a screening prior to clonal testing (Borralho 1992). Demonstration of the absence of adverse genetic correlations between embryogenic potential and subsequent field performance, however, will be important. In a study with open-pollinated families of Picea abies, embryogenic capacity was not correlated with a number of traits associated with vegetative phenology in the field (Ekberg et al. 1992). In a few taxa, micropropagation may involve milder selection than for propagation by cuttings, e.g. for some poplar hybrids (Cooper 1992).
Successful protocols for axillary budding (particularly for hardwoods) and adventitious budding (particularly for conifers) exist for a large number of forest tree species, and the number of species for which somatic embryogenesis has been reported is increasing, particularly for conifers. Compared to cuttings, the higher multiplication rates available through micropropagation offer advantages with respect to the capture of genetic gains through clonal forestry. One major factor impeding early application in many industrial plantation programmes is that breeding is not sufficiently advanced for clonal forestry to be contemplated.
Current high costs will be an impediment to the direct use of micropropagules as planting stock in many programmes. Technologies resembling those used commercially in horticulture are most likely to be affordable for very high value species such as Juglans and Prunus, and perhaps also some long rotation tropical hardwoods, particularly those for which propagation by cuttings is difficult. These technologies are not likely to be affordable in pulpwood programmes, and will be marginal for medium rotation sawnwood species except where MAIs are very high. Major fluctuations in real stumpage rates, however, may change the above outlook.
Micropropagation clearly has a role in an integrated clonal propagation system - where micropropagation is used for the rapid multiplication of stool plants of the selected clones, but cuttings form the final planting stock. This approach is only of value, however, in very advanced breeding programmes which incorporate the generation and identification of outstanding clones.
The potential direct use of micropropagules as planting stock in industrial plantation forestry will broaden dramatically when costs are reduced. The most promising research areas in this context are those concerned with somatic embryogenesis and the development of artificial seeds. These are important strategic research areas underlying future applications, and should be pursued with model species such as Pinus taeda.
Uncertainty regarding field performance is another major impediment to early broader use of micropropagules. It is important that much more extensive field trials be established - in particular trials which will permit detection of genotype × propagule type interactions and genetic correlations between economic and reproductive traits.
Where clonal testing on a reasonable scale is possible and affordable, the current applicability of protocols mainly to juvenile material is not necessarily an impediment to the capture of good gains through clonal forestry. This conclusion, however, is dependent on the ability to store juvenile material for the period of a clonal test, and progress made with the cryopreservation of embryogenic cultures is highly relevant.
Genetic variation in response, often substantial, is not likely to be a major problem where clonal testing can be preceded by screening for responsive genotypes.
Micropropagation is unlikely to be used for the production of planting stock of non-industrial species, but may have a role in the multiplication of selected varieties prior to release. The availability of micropropagation technologies will also be essential in any future genetic engineering applications, and the development of simple protocols for those species for which such are not already available will be a useful research objective, although of lower priority.
Figure 11.1 Comparison of propagation options for Slash × Caribbean Pine