Tải bản đầy đủ - 0trang
V. Improving the Efficiency of the Pea Fruit
C. L. HEDLEY AND
M. J . AMBROSE
Unlike other legumes, the illuminated pod of the lupin (Lupinus albus) can
make CO, gains from the atmosphere for all but the last 2 weeks of its life (Pate
et al., 1977). It is also suggested that the pod acts as a temporary reservoir and
agent for remobilization of the respiratory products of the seed.
The pea (Pisurn sarivurn) pod is committed to exporting assimilates, derived
from CO, fixation, to the developing seeds (Lovell and Lovell, 1970). This
carbon is mainly derived from fixation of COBrespired by the seeds into the pod
cavity (Flinn and Pate, 1970). Most of the carbon required for seed development
originates from the subtending leaf and stipule (Flinn and Pate, 1970), the photosynthetic activity of which is modulated by the growth rates of the pod and the
seed (Flinn, 1974). Although the pea pod is capable of a net uptake of CO, from
the atmosphere only during the very early stages of pod development (Harvey et
al., 1976), its role in refixing and recycling carbon to the seed is substantial and
accounts for up to 20% of the fruits assimilate requirement (Flinn et al., 1977).
Both the photosynthetic carbon fixation enzyme, ribulose- 1,5-bisphosphate
(RuBP) carboxylase and phosphoenolpyruvate (PEP) carboxylase, often associated with dark C 0 2 fixation mechanisms, have been demonstrated within the
pod tissue of peas (Hedley er al., 1975). The distribution of these enzymes
within the pod wall emphasizes the refixation role of the pod, the inner epidermis
being rich in both carboxylase enzymes (Atkins el al., 1977). At high light
intensities these enzymes are capable of fixing 60% of the CO, released by the
seeds, although this proportion is reduced at the light levels present in the canopy
(Atkins et al., 1977).
It is likely that the type of pod that will be most efficient in the “leafless”
canopy, with its increased light penetration and improved standing, will be very
different in structure and physiology from that more suited to the leafed pea crop.
A wide range of genetic variation is available for pod type in Pisum sativum.
Large differences exist between genotypes for pod size and pod growth rate
(Hedley and Ambrose, 1980), and also for pod wall thickness (Wellensiek,
1925b) and for the presence or absence of schlerenchyma layers (White, 1917).
Comparisons, using genotypes that differed for wall structure or chlorophyll
content, have been made to assess variation for carboxylase activity (Price and
Hedley, 1980). The activity of both carboxylase enzymes was shown to vary
with pod type and with pod age, the activity of the photosynthetic enzyme (RuBP
carboxylase) correlating with the chlorophyll concentration. Yellow-podded
types had lower activities of this enzyme but also had higher absolute levels of
the dark fixation enzyme (PEP carboxylase). The PEP-carboxylase system was
shown for all pods to comprise a far higher proportion of the total carboxylase
activity than that normally found in leaf tissue exhibiting the C, photosynthetic
system (Price and Hedley, 1980).
As yet there is little information about the significance of different pea pod
types for the seed yield of the plant. Even more important, there is no informa-
DESIGNING “LEAFLESS” PLANTS FOR DRIED PEA CROP
tion about the effect of pod phenotypes when incorporated into “leafless” plants.
There is a suggestion from studies using near-isogenic lines for the uf and st
genes, that the net CO, uptake of the pods from leafless plants is higher than that
for corresponding leafed plants (Harvey, 1978). No clear explanation, however,
has been put forward to explain this observation. There is also little information
for peas on the role of the pod as a temporary storage organ. Although some
studies suggest that pods are committed to the export of assimilate to seeds (Flinn
and Pate, 1970), there is no reason to suppose that variation does not exist for this
characteristic, especially among pods differing for wall thickness.
B . THESEED
Ultimately it is the number and weight of individual seeds that survive through
to maturity that determines the yield of a crop. Although the ability of a genotype
to tolerate interplant competition is important in partitioning assimilate into economic yield, it is of equal importance for the seeds of such a genotype to tolerate
the resulting intraplant competition. As with the competition between plants, if
the competition between developing seeds for diminishing resources within the
plant is high, then the number of seeds that succeed in developing through to
maturity will be low. As discussed in Section IV,B, competition between seeds
developing at different nodes will be reduced if a genotype has a reproductive
indeterminate habit, and competition within reproductive nodes will be reduced
if only a single pod develops at each node. The number and weight of seeds that
develop within each pod, however, will be determined by the tolerance of each
seed for its neighbors. Consequently, if seed variants that are less demanding of
the plant’s resources can be found, then more seeds would be expected to develop successfully.
In general, seeds within a pod develop asynchronously with the largest individuals in the center, tapering to very immature seeds at both ends (Fig. 19). This
asynchronous growth is initiated very early in the development of the fruit. The
cause of the asynchrony is not known, but it does not appear to be due to a lack of
fertilization (Linck, 1961). As development continues there is a tendency for the
seeds at both ends of the pod to abort, and for only the central seeds to continue
development (Linck, 1961). When plants are grown in environments that induce
intensive intraplant competition, further abortion occurs and it is the smallest
seeds that appear to be most susceptible (personal observation). The interactions
between individual seeds within a pod are in many ways similar to those between
individual plants within a population. It is possible therefore to apply to the
behavior of seeds toward their neighbors reasoning that is similar to that applied
to explain the responses of individual plants within the crop (Section 111,C).
It can be assumed that there is a finite rate for the translocation of assimilate
C. L. HEDLEY A N D M. J . AMBROSE
Distribution of seed size within a pea pod
from the plant into the pod. This rate of assimilate input will therefore only
supply a finite sink demand from the seeds, the sink demand being equivalent to
the sum of the growth rates of all the developing seeds within the pod. Initially
the sum of the individual seed growth rates will not exceed the assimilate supply
and all of the seeds will begin to develop. As the seeds within the pod develop,
the sum of the growth rates may eventually exceed the rate of assimilate input
into the pod and competition between seeds will occur. As with plants within a
competitive sward, the seeds with the highest growth rate, situated in the center
of the pod, will continue to grow at the expense of the seeds with lower growth
rates, situated at the two ends of the pod. If this interseed competition occurs at a
critical stage in development, then the seeds with the lowest growth rates will
abort. In large-seeded types abortion occurs even among quite large developing
seeds and accounts for the low numbers of large seeds reaching maturity (personal observations).
It is apparent that the key to improved tolerance between seeds within the pod
is to ensure that the total seed sink demand throughout development does not
exceed the assimilate input into the pod. By definition this will entail maintaining
relatively low seed growth rates. This can be achieved by selecting seed types
that have a decreased relative growth rate (RGR) and that would therefore, for a
given seed size, have an increased duration of growth (Fig. 20a). The overall
effect of a reduced RGR will be to reduce the absolute growth rate and hence the
DESIGNING “LEAFLESS” PLANTS FOR DRIED PEA CROP
sink demand of the seed. Relative growth rate is determined by the interaction of
the rate of cell division and expansion with environment. Seed phenotypes with a
lower RGR would either have decreased rates of cell division and/or cell expansion, or more likely the proportion of seed growth attributed to cell division
would be reduced relative to that attributed to cell expansion. This will lower the
seed RGR because dividing cells have a higher RGR than expanding cells.
Another possibility for improving interseed tolerance, especially later in seed
FIG. 20. Theoretical variation for seed growth. (a) Seeds of similar size differing for relative
growth rate. (b) Seeds differing for size but with similar relative growth rate. (c) Seeds differing for
size and for relative growth rate. L, large seed; S , small seed.
C. L. HEDLEY AND M. J . AMBROSE
development, is to select small seeds, which by definition will mature while their
growth rates are relatively low (Fig. 20b). For a given ovule number per pod the
best seed phenotype will therefore be small-seeded with a low RGR (Fig. 20c).
In preliminary studies using six-leafed genotypes, we have found significant
differences in the RGR of seeds during the early part of development (Hedley and
Ambrose, 1980). We have not, as yet, studied the effect of this variation on
competition between seeds, or the effect of interplant competition on the different seed phenotypes.
Selection for a character such as RGR will be difficult because the seed is not a
genetically homogenous structure. The seed is composed of the embryo, the
testa, which is maternal, and the endosperm, which is triploid and composed of
two maternal and one paternal genome (Cooper, 1938). The seed phenotype is
determined by the development of these tissues and by the developmental interactions between them. The effect of the maternal influence on the development of
the seed can best be observed from the differences between reciprocal crosses of
genotypes that differ for seed size. In such crosses the resulting F, seeds usually
resemble the maternal parent in size (Davies, 1975). Such a maternal control may
act via the testa determining or controlling the transfer of nutrients to the embryo
(Murray, 1979, 1980).
In an attempt to understand the complex development of the seed, we have
studied the development of the component parts, in a range of genotypes (Hedley
and Ambrose, 1980). The physical relationship between the embryo and embryo
sac (the endosperm-filled vacuole formed within the developing testa) is primarily concerned with the rate at which the embryo and embryo sac volumes expand
relative to each other (Fig. 21). The initial increases in volume of the embryo and
embryo sac are exponential and therefore linear on a logarithmic scale. Variation
was found between the genotypes for the slopes of both lines and for the separation in time for the initiation of exponential embryo growth relative to that of the
embryo sac. The slope for increase in embryo volume was always considerably
greater than that for the volume of the embryo sac. The difference in time
between the initiation of the two exponentials, however, determined that initially
the absolute volume increase of the embryo sac was greater than that of the
embryo. The absolute difference between the two volumes therefore increases
initially and liquid endosperm accumulates. The difference in slope between the
embryo and embryo sac, however, determines that a point is eventually reached
where the absolute rates of volume increase are the same. This point corresponds
to the maximum volume of endosperm within the seed. After this point endosperm is absorbed, presumably by the developing embryo.
Variation between genotypes in the growth of the seed and in the final seed
size are determined by differences in the slopes of the exponentials and the
difference in time between their initiation. An understanding of the mechanisms
controlling these three variables will be necessary before seeds with specific
DESIGNING “LEAFLESS” PLANTS FOR DRIED PEA CROP
FIG. 21. Relationship between the expanding embryo sac (es) and embryo (e), showing point of
maximum (max. end.) and minimum (min. end.) endosperm volume. r is the difference in time
between the onset of embryo sac and embryo exponential growth.
growth characteristics can be selected. This forms the basis of our present research into seed development. As part of these investigations we have studied the
cellular composition of embryos from a range of genotypes. Differences have
been found between genotypes in the number of cells within a given embryo
volume, suggesting differences in the proportion of cells within the embryos that
are expanding. Such variation, as stated earlier, may affect the RGR of the
embryo. It is not known, however, if these observations are due to intrinsic
genetic differences between embryos or if they are the result of a maternal
influence. There is evidence from reciprocal crosses that the cell number of
embryos is greatly modified by the maternal parent (Davies, 1975).
As well as selecting seed phenotypes that are less demanding per unit time of
the assimilate input from the plant, it may also be possible to improve the
C . L. HEDLEY AND
M. J . AMBROSE
efficiency with which seeds utilize assimilate. As with the pod wall, both the
testa and embryo have significant levels of PEP-carboxylase activity as well as
much lower levels of RuBP-carboxylase activity. These enzymes may act to
reduce respiratory losses and recycle carbon within the seed (Hedley et ul.,
1975). More conclusive evidence of such a system has been found for the
developing seeds of lupin (Lupinus ulbus; Atkins and Flinn, 1978). The significance of this system for the developing embryo is not known, but it can be
suggested that the provision of C4 acids from such a recycling system will be
important in the synthesis of amino acids at a time when protein synthesis is high.
VI. A PLANT IDEOTYPE FOR IMPROVING YIELDS OF
The preceding sections have described some of the problems associated with
the “leafless” phenotype, the main problem being the necessity to grow leafless
plants at high planting densities in order to attain a sufficiently high biological
yield per unit area. We have also suggested how the efficiency of the plants
within the dried pea crop can be improved. Many of these suggestions will apply
irrespective of whether or not the crop plant has a leafless phenotype.
If we persist with the leafless phenotype (ufafstst), it is evident that the
ideotype must be tolerant of high planting density and therefore by definition
must be a relatively uncompetitive plant. The relationship between a plant’s
competitiveness and its growth rate determines that a plant with a reduced growth
rate is required. Unless variants can be found that grow at a reduced relative
growth rate, which seems unlikely, plant growth rate can only be reduced by
selecting plants that have developed from a small embryonic axis. This will
determine that the ideotype will have small seeds, unless the relationship between the size of the embryonic axis and seed size can be broken.
A modification to the leafless model, and one that we are now considering, is
to incorporate the gene for normal stipule size (St) while maintaining the gene for
converting leaflets into tendrils (uf).There is some evidence that this phenotype
(ufafStSt) has a higher growth rate than leafless plants of comparable seed size
(Snoad, 1981), but it will hopefully maintain some of the improved canopy
characteristics of the leafless phenotype. It is possible that such a modification
will overcome the absolute requirement for high planting densities and many of
the problems, both economic and physiological, that accompany such densities.
Even if such a modified model proves successful, we have no reason to suggest
DESIGNING “LEAFLESS” PLANTS FOR DRIED PEA CROP
that other features of our crop ideotype, incorporated to improve plant tolerance
to the crop environment, will change.
Ideally the plant should have a genetically determined nonbranching habit.
Branches do not add significantly to economic yield at high planting densities,
although they do utilize environmental resources and increase biological yield
per unit area. If such a plant type cannot be identified, then plants that branch late
in development should be selected, since such plants will be inhibited physiologically from branching by competition within the sward.
The ideotype will be relatively early-flowering so that partitioning of assimilate into reproductive structures is initiated when competition between plants is
low. Early flowering should be coupled with a reproductive indeterminate habit.
This will allow the ceiling biological yield per unit area to be attained and will
increase the duration of assimilate partitioning into economic yield. An increased
duration of partitioning will reduce the competition between yield components
Each reproductive node will contain a single pod selected for maximum efficiency of carbon refixation and recycling. The seeds will have a low demand per
unit time of available resources, and therefore will have a reduced relative
growth rate and will mature when small. In addition, seeds that have an improved
efficiency for recycling carbon will be incorporated into the ideotype, if such a
system can be shown to have a beneficial effect on yield.
Any incompatibility between the seed’s dual role as the main reproductive sink
and as an embryo plant must be taken into account, and may result in an appropriate compromise.
B . BREEDING
The overriding effect of interplant competition on plant characters, such as the
number of reproductive nodes and overall plant size, determines that the assessment of an individual’s suitability as a crop plant must be made in an environment akin to that encountered by the plant within the crop. If, however, a
segregating population of individuals that differ for competitiveness is grown as a
microplot at commercial planting densities, then strongly competitive individuals
will thrive relative to weak competitors. Therefore in such an environment the
weak competitors would perform much worse and the strong competitors much
better than if each were grown in a crop of genetically similar individuals. This
presents the breeder with a problem, since for reasons discussed earlier, those
individuals within a segregating population that are most likely to make successful crop plants will be weak competitors. In order that such plants not be discarded early in the breeding program, it is essential that selection be delayed until