Tải bản đầy đủ - 0 (trang)
III. Importance of Variation in Time of Seedling Emergence to Crop Development

III. Importance of Variation in Time of Seedling Emergence to Crop Development

Tải bản đầy đủ - 0trang



two sowing dates. In mixed sowing date plots, the plants were sown on a

square grid with the early and late sowings alternating in a checkerboard

design. The difference in sowing time was either 2, 4, or 8 days and the

sowing positions were 1.5 cm apart. Despite sowing seeds carefully at

uniform depth in boxes of compost, there was a spread in emergence time

of 8 to 14 days for the early-sown plants and 5 to 10 days for the late-sown

plants. Thus, there was a wide range of preemergence growth rates, which

resulted in a wide overlap in the seedling emergence times from the different sowings. When the plants had a foliage height of 30 cm, the logarithm of

dry weight was negatively correlated with time of seedling emergence. The

novel point about this work was that the effect of time of sowing on the

regression of weight per plant at harvest on seedling emergence time was

examined and was found to have only a slight effect. Thus, time of seedling

emergence, and not any correlation with preemergence growth rate, was

responsible for the subsequent effects on plant weight.

In the following sections the effect of time of seedling emergence on the

growth, form, and composition of individuals in populations is examined.



At the point of seedlingemergence, plant weight is still minute compared

with the potential and probable weight that the plant will attain. Plant

growth at this time is nearly always exponential because there is virtually

no self-shading or competition for growth resources from neighbors. Consequently, a difference of a few days in seedling emergence time can result

in a manyfold difference in weight between plants. For example, Black and

Wilkinson (1963)found that a delay in emergence of 5 days brought about a

reduction in subterranean clover weight of about 50%, and a delay of 8 or 9

days produced a reduction in weight of at least 75%. Gray (1976) showed

that, in lettuce at 26 days after sowing, seedlings that had emerged at 7

days were five to six times larger than those that had emerged at 15 days.

When considering the importance of relative times of seedling emergence for the subsequent dry matter increment of individuals within a

population, a number of factors have to be taken into consideration. First,

the spread in time of seedling emergence in a population is important, but

most studies relied on the natural spread in time of seedling emergence

rather than attempting to modify it artificially. The time from seedling

emergence to harvest is also relevant. As this time interval increases, there

is a greater probability that individual plant growth will be influenced by

some extraneous factor, for example, herbivory. Finally, the “state” of

the population must be considered, such as the density of plants, or



whether the plants under consideration are a monoculture or part of a

mixed-species population. Table I presents a summary of these factors and

the percentage weight variation accounted for by spread in time of seedling

emergence in a number of species.

The percentage of weight variation associated with spread in time of

seedling emergence was usually greatest when fewer than 20 days have

elapsed between mean time of seedling emergence and harvest (Table I).

(Gray, 1976; Benjamin, 1987). The paper by Benjamin (1982) (Table I) best

illustrates how the percentage of weight variation associated with spread

in time of seedling emergence increases as spread in time of seedling

emergence increases. This trend is also seen in other work (Table I).

Benjamin (1982) reduced the spread in time of carrot seedling emergence

from 19 to 14 days by using a contact herbicide to kill the first seedlings to

emerge. This treatment had no detectable effect on the frequency distribution of storage roots in different diameter grades, suggesting that relatively

slight changes in spread of seedling emergence time have no detectable

influence on variation in weight. Controlling time of seedling emergence

might allow other sources of weight variability to be expressed more

strongly. However, the residual sum of squares of root weight (after taking

out the effect of emergence time) from plots that had small spreads in

emergence times was no greater than the residual sum of squares from

plots that had large spreads in emergence times (Benjamin, 1982). Therefore, there is no evidence that other sources of variation are able to express

themselves to a greater extent when time of seedling emergence is controlled.

The effect of density on the relationship between spread in time of

seedling emergence and plant weight is difficult to interpret from Table I

because the other factors were varied too. The very high percentage

variation in weight associated with spread in time of seedling emergence

reported by Ross and Harper (1972) is perhaps because they used extremely high densities (Table I). Benjamin (1982) showed that the increase

in coefficient of variation (CV) of carrot shoot and storage root weight at

wider spreads in seedling emergence time was greater at 400 than at 25

plants mP2.

Ross and Harper (1972) noted that the last cocksfoot seedlings to emerge

in a population growing at 30,000 plants m-’ had a weight only a little

greater than that of seeds, even after 35 days of growth. The implication is

that late-emerging seedlings are denied resources for growth by the earlieremerging plants and these competitive interactions are more intense at

high densities.

More direct evidence that relative time of seedling emergence determines the relative amount of growth resources that can be captured by an

individual comes from a second experiment by Ross and Harper (1972) in

Table I

Percentage Shoot Weight Variation Associated with Spread in Time of seedling Emergence






Lactuca sativa

Spread in

time of






Mean days



to harvest


% variation

in weight






Ross and Harper
















Gray (1976)















24.48, or%

seedlings in

34 x 12cm


500- 1000









ca. 105


Naylor (1980)

Fecundity measured, plants grouped

into six emergence classes

Howell (1981)

Daucus carota

Lactuca sativa

Daucus carota

Allium porrurn





Beta vulgaris


Not stated.

25 and 400

25 and 400

25 and 400


ca. 120

ca. 120

ca. 120

ca. 33

ca. 117

ca. 27

ca. 75

ca. 140

ca. 259


ca. 50








ca. 0








1/23-cm pot

10 seeds in 10 x

10-cm tray


ca. 56

ca. 0










1/18-cm pot










ca. 0


6-1 1


ca. 10













Root weight measured

Root weight measured

Root weight measured

Individual plants harvested from

within crop when mature

Root weight measured

Root weight measured

Root weight measured

Root weight measured

Benjamin (1982)

Wurr and

Fellows (1983)

Benjamin (1984a)



Dolan (1984)

Fecundity measured, plants grouped

into two emergence classes

Stanton (1985)

Benjamin (1987)



which the growth of seedlings that had emerged at different times with and

without the presence of neighbors was compared. Where neighbors were

not present, then the difference in weights between plants that had

emerged at different times was explained by the differences in durations of

growth. However, where neighbors were present, then the late-emerging

seedlings were smaller than expected from just their reduced period from

emergence to harvest. Benjamin (1984~)grew carrots in either pure or

mixed sowing date combinations at 400 plants m-*. The sowing dates were

separated by a fortnight, thus ensuring different times of seedling emergence. At 35 days after the first sowing, the late-sown plants in mixed

sowing date treatments had ceased growing, whereas there was no evidence that the late-sown plants in pure stand had ceased growth at the end

of the experiment at 56 days. This experiment provides some of the

strongest evidence that the rank each individual occupies in the hierarchy

of sizes maintained by competition is largely determined by its time of

seedling emergence relative to that of its neighbors.

VanBaalen et al. (1984) set up a replacement series experiment with

Scrophularia nodosa and Digitalis purpurea, species that have similar

seed weights, but the former had a higher relative growth rate when grown

in monoculture. The replacement series was repeated with the D. purpurea sown either at the same time as, 7 days earlier, or 7 days later than

S . nodosa. They concluded that a delay of 1 day in the emergence of

D. purpurea in mixed stand resulted in a 28% reduction in yield per plant,

whereas a delay of 1 day in the emergence of S . nodosa gave a reduction in

yield per plant of 5%. Therefore, the importance of spread in time of

seedling emergence depends on the seedling relative growth rate,

and perhaps the effect of given delay in the time of emergence on the

growth of an individual should be “scaled” by the mean relative growth


The relative times of seedling emergence have been considered to be of

importance in studies of competition between species. Nieto et al. (1968)

first introduced the idea that there were critical periods in the development of crops during which the presence of weeds will suppress the

growth of the crop, these critical periods being defined in terms of time

after crop seedling emergence (Hewson and Roberts, 1971; Spitters and

vandenBergh, 1982). Elberse and deKruyf (1979) set up replacement series

experiments using Hordeum vulgare and Chenopodium album at a range

of densities and with sowing the Chenopodium at various times before the

Hordeum. They concluded that a difference of 15 days was the critical

period; Hordeum outcompeted Chenopodium if Chenopodium was sown

less than 15 days before Hordeum, but Chenopodium outcompeted



Hordeurn if sown move than 15 days before Hordeurn. The relative densities of the two species was of relatively minor importance to the outcome

of these competitive interactions. More recently, Cousens et al. (1987)

have developed a model to describe the relationship between crop yield

and the combined effect of weed density and its time of seedling emergence

relative to that of the crop.

Some of the studies listed in Table 1 claimed to show that little of the

variation in plant weight was associated with time of seedling emergence.

How could this occur if the general tenet of this review is correct? Sometimes the lack of an association between plant weight and time of seedling

emergence was because of a very low spread in time of seedling emergence

(Dolan, 1984; Benjamin, 1987). The spread in time of seedling emergence

in carrots in the study by Benjamin (1984a) was 19 days, yet only 3%

weight variation at 33 days was associated with time of seedling emergence. However, in this study the vast majority of seedings (82%)emerged

over only a 5-day period. The studies of Howell (1981) and Stanton (1985)

were unable to show that spread in time of seedling emergence was important perhaps because they used fecundity rather than plant weight as a

measure of plant performance, and they also classified seeding emergence

date into a restricted number of categories.

Sangakkara and Roberts (1986), in a competition study between grass

species, argued that they had eliminated differences in time of seedling

emergence between ryegrass, prairie grass, and cocksfoot by using transplants. They concluded that “competitive relationships seem to be determined by initial seedling size and the capacity to accumulate greater

quantities of dry matter” and that “early emergence had no apparent

effect in determining the competitive hierarchy established between these

species.” However, they neglected the possibility that the different initial

seedling sizes at the time of transplanting were a direct consequence of

differences in time of seedling emergence.

The bulk of the literature indicates that spread in time of seedling

emergence is a major source variation in mature plant weight, probably

because differences in emergence time have a large effect on seedling size

at the point in time when plants start to compete with one another for

growth resources. Time of seedling emergence has its largest effect when

spread is large, when seedlings have a high relative growth rate (warm,

moist conditions and fast-growing species), when between-plant competition is intense (at high plant densities), and when harvests are made not

long after seedling establishment. In the following sections the effect that

seedling emergence time might have on the morphology, physiology, and

biochemistry of plants will be examined.





The previous sections have shown that the late-emerging seedlings become the smallest in the population and the ones most likely to be shaded

by their neighbors. The position occupied by the foliage influences the

temperature, humidity, and quantity and quality of light that shoots experience, which can have a large effect on the partitioning between organs. For

example, Acock et al. (1979) showed that with increasing radiation or

decreasing temperature, a greater percentage of dry matter is partitioned

to the stem at the expense of the roots in chrysanthemums. This study

shows that it is difficult to determine a general effect of late emergence on

partitioning of dry matter because late-emergers will tend to have foliage

both in the shade and at a lower temperature than early-emergers.

Studies that examine the competition between species are relevant because between-species differences in plant stature might be analogous to

the differences in stature within species caused by differences in time of

seedling emergence. These studies also show conflicting effects of stature

on partitioning of dry matter between organs. For example, Digitalis

purpurea (a rosette-forming plant) grew more slowly and had a lower

root : shoot ratio when grown in association with Scrophularia nodosa (a

stem-forming plant) than when grown in monoculture (VanBaalen et al.,

1984). In contrast, Hawthorn and Cavers (1982) reported that groups of

Plantago major and P. rugelii plants, suppressed by the presence of larger

neighbors, allocated a greater percentage of dry matter to the roots and

caudices, whereas the larger plants devoted a greater proportion of dry

matter to reproductive growth.

Barnes (1979) presented extensive data from field experiments with

carrots to show that at any given time there was a linear relationship

between the logarithm of shoot weight and the logarithm of storage root

weight, but the intercept of this relationship decreased with time. The

variation in plant weight at any one time was largely due to the different

imposed density treatments. The implication is that at any one instant, the

relative growth rate of the shoots is a fixed proportion of that of the storage

roots, but this proportion changes with time. Barnes reasoned that the shift

in the relationship between shoot and storage root relative growth might be

due to the lower maintenance respiration rate in storage organs. The

expectation would be that plants derived from seedlings that had emerged

at different times would exhibit different partitioning of photosynthates

between organs because the plants were effectively of different ages.

In a purely theoretical paper, Moms and Myerscough (1987) considered

the effect of partitioning of dry matter between shoots and roots on the

competition between plants in even-aged monocultures. They concluded



that the yield-density relationships could be influenced by shoot-root

partitioning of dry matter, which implies that there could be complex

feedback processes occurring in crops in which differences in time of

emergence lead to differences in amounts of resources available to each

plant, which in turn can lead to differences in shoot-root partitioning,

which in turn leads to differences in the ability of plants to compete

with their neighbors. Until data are presented to support this view,

then the more simple assumption should hold, that is, suppressed plants

adapt to allow them to scavenge resources not taken up by dominant plants

(Robinson, 1986), but this adaptation does not deny resources to the

dominant plants.

Thus, there appear to be no studies that have examined the partitioning

of dry matter between foliage and fibrous roots separately in plants derived

from early- and late-emerging seedlings, but the effects of different emergence times on shoot-root partitioning might depend on the light intensity

falling on the crop and the mineral status of its soil.


Iwaki (1959) made a large number of interesting observations on the

effect of mixed or pure stands and the relative times of sowing on the

canopy structure and weight of buckwheat (Fugopyrum esculentum) and

green grams (Phaseolus uiridissirnus). In pure stands, buckwheat increases in weight slightly faster, and grows much taller, than green grams.

In mixed stands there is a severe suppression of the growth of the green

grams. This suppression is lessened by sowing the green grams earlier than

the buckwheat, but a difference of 13 days was not sufficient to overcome

the dominance of the buckwheat. An interesting feature of this study was

that the canopy was harvested separately in 10-cm-deep strata, so that the

effects of mixed and pure stand on canopy architecture could be examined.

I have estimated the percentage of the photosynthetic system in each

10-cm band of the canopy from his Fig. 1 (Table 11). The plants growing in

pure stand deployed over half their photosynthetic dry matter in the top

10-cm layer of the canopy, whereas the plants growing with buckwheat

(which grew to a height of 120 cm) deployed their photosynthetic dry

matter more uniformly along the length of the plant. However, these

suppressed plants were nearly the same height as those in pure stand, but

their total weight was less than half that of the pure stand plants (Table 11).

Presumably similar morphological responses would occur where the cause

of the dominance was a difference in time of seedling emergence.

Perry (1985) reviewed the literature on competition in forest stands and



Table 11

Effect of Stand Type on the PhotosyntheticDry

Matter (Percentage of the Entire Plant) at

Various Heights above the Ground in a Crop of

Green Grams”

Stand type








Total (9)



with Buckwheat












From Iwaki (1959).

included the morphological differences between dominant and suppressed

trees. He concluded that shade leaves are larger and have more chlorophyll per unit area, and that sometimes the small trees have greater height

and radial growth rates than their larger neighbors because of a lower

respiration rate and higher specific leaf area. Dominant trees have larger

and broader crowns, which allows them to maintain high rates of photosynthesis per tree, but at the expense of high maintenance respiration rate.

Perry (1985) also pointed out that suppression of plant growth can influence the chemical composition of trees, with “reduced allocation of photosynthates to those chemical constituents that are energy-expensive.’’ To

give an idea of the effects he quoted Chung and Barnes (1977), who

determined the number of grams of glucose required to synthesize 1 g of

the following substances in Pinus taedu: lipids, 3.02 g; phenolics, 1.92 g;

lignin, 1.90 g; nitrogenous compounds, 1.58g; organic acids, 1.43 g; carbohydrates, 1.13 g. The more energy demanding compounds could be involved in deterring herbivory. Where insects and molluscs are the herbivores then the smallest plants in the community are most prone to

predation (Perry, 1985; Gange et af., 1989; Thomas and Weiner, 1989), but

where there are grazing mammals then the largest plants can be preferentially eaten (Weiner, 1988).

Thus, the relative times of seedling emergence are likely to have a large

effect on the shape and composition of plant organs. This is likely to be

important in terms of the adaptation of plants to competition for growth

resources. However, there are also commercial implications because there



is an increasing emphasis on the quality aspects of harvested produce,

which includes its shape, size, and composition. For example, Wurr and

Fellows (1983) showed that in studies with two cultivars of lettuce sown at

three occasions, the heads of plants derived from late-emerging seedlings

matured later than those derived from early-emerging seedlings.

The results of these studies that follow the development of individual

plants contrast strongly with those of Symonides (1978), who observed the

year-to-year difference in the rate of development of five species contrasting in their rate and duration of development. Symonides (1978) found that

the rate of development in all five species was faster in those years in which

the date time of seedling emergence had been delayed because of climatic

conditions. In this latter study, the development of the entire population

rather than that of individual plants had been observed. Clearly, within

populations, the tendency for accelerated development with late seedling

emergence is overridden by the effects of competition with larger plants.


There have been a number of studies that have demonstrated that the

plants derived from late-emerging seedlings are more likely to die. For

example, Stanton (1985) split a wild radish population into two emergence

time cohorts. The early-emerging cohort consisted of larger plants and the

probability of mortality before flowering was 0.21, but the corresponding

value was 0.55 for the late cohort. vanderToorn and Pons (1988) studied

the emergence patterns and survival of two Plantago species in plots in

which other species (Poa triuialis, Phleum pratense, Alopecurus prarensis, Trifolium repens, and Ranunculus repens) had either been removed,

clipped to a foliage height of 5 cm or 15 cm, or left unclipped. They

reported a spread in seedling emergence times of 10 and 12 weeks for the

two species. Seedlings that emerged earlier had higher survival rates, and

there were greater survival rates in more open environments, with survival

being positively correlated with the transmission of light in the different

clipping treatments. Shaw and Antonovics (1987) made a population census of Slauia lyrata plants growing for nearly 3 years in pots. They reported that the plants that founded the population had the greatest chance

of survival and produced the greatest number of seeds.

Westoby and Howell (1986) approached the relationship between emergence time and mortality in a different way. They investigated whether the

population structure of a community influenced the course of densitydependent self-thinning in radish. This course was described by graphs of

logarithm of biomass per unit area plotted against logarithm of numbers of



surviving plants. When plants are self-thinning as a result of between-plant

competition, the path of points on the graphs runs from bottom right (high

density, low biomass) to top left (low density, high biomass). The slope of

this path is claimed to be -4 (see White, 1980; Antonovics and Levin,

1980; Westoby, 1984 for reviews of self-thinning). Different population

structures were achieved by making either a single sowing or two sowings

separated by between 4 and 23 days in a pot. They concluded that the

course of self-thinning in those pots that had a staggered sowing time was

not consistently different from the course in the pots that had a single

sowing. The implication is that plants derived from late-emerging seedlings

are no more likely to die from competition with their neighbors than those

derived from early-emerging seedlings. However, the relationship between mean weight of survivors and plant density varied greatly between

the single sowing date pots in different experiments. This implies that this

relationship is not sufficiently consistent to be used to distinguish between

large differences in the probability of mortality of plants derived from

early- and late-emerging seedlings. Indeed, Weller (1987) has now shown

that many of the sets of data that were thought to support the -4 power law

in fact do not have statistically significant correlations between logarithm

of weight and logarithm of density.


Clearly, there can be a wide range of seedling emergence times, even in

synchronously sown commercial crops. The causes of different times of

seedling emergence are numerous, but the limited evidence suggests that

in monocultures these preemergence processes are important only because they influence time of seedling emergence.

This is an important hypothesis that deserves more attention and further

experimental verification. If true, it replaces the plethora of possible interactions between individuals in monocultures by a simple measurable statistic, the relative times of seedling emergence. This statistic can be used

as a predictor of future plant growth.

If time of seedling emergence is to be used as a predictor of plant

performance, there is a need for more consistent intervals of time into

which emergence is classified. Where spread of emergence time is very

wide, then seedlings that have emerged over several days often have been

grouped, but more usually emergence is recorded daily. Consequently,

comparisons between experiments have been made difficult because the

time intervals into which seedlings were grouped have been inconsistent.

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

III. Importance of Variation in Time of Seedling Emergence to Crop Development

Tải bản đầy đủ ngay(0 tr)