Techniques for determining the effects of tillage on the distribution of sicklepod (Cassia obtusifolia L.) seed in the soil


Material Information

Techniques for determining the effects of tillage on the distribution of sicklepod (Cassia obtusifolia L.) seed in the soil
Cassia obtusifolia
Physical Description:
vi, 98 leaves : ill. ; 28 cm.
Hacker, Larry Augustus, 1956-
Publication Date:


Subjects / Keywords:
Sicklepod   ( lcsh )
Tillage   ( lcsh )
Agronomy thesis Ph. D
Dissertations, Academic -- Agronomy -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1986.
Bibliography: leaves 90-96.
Statement of Responsibility:
by Larry Augustus Hacker.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000566648
oclc - 14253667
notis - ACZ3088
sobekcm - AA00004859_00001
System ID:

Full Text






May 1986


I would like to acknowledge my very true sincere

appreciation to Dr. Wayne Currey, Dr. David Hall, Dr. Jerald

Milanich, Dr. Sherlie West, and Dr. David Teem. I would

especially like to acknowledge Dr. David Teem and Dr. Wayne

Currey for their continuous support and guidance throughout

my graduate program. I thank them most for understanding

the problems encountered during the research and helping me

to overcome them.

Thanks are also extended to Dr. Larry Fitzgerald,

Dr. Charles Beatty, and Lester Burch for their help in the

preparation of the tools and loan of equipment necessary for

the completion of this work.

Finally, I would like to dedicate this work to my wife,

Barbara and my twin daughters Lauren and Lindsey for the

many sacrifices that they made during the course of this

study and preparation of this manuscript. Without their

support, love, and strength it could not have been achieved.



ACKNOWLEDGMENTS.... ....... .......................... ii

ABSTRACT .....................,............. .. .... ..... v


I LITERATURE REVIEW............................ 1

Techniques of Seed Recovery ................ 1

Estimation of Seed Number................... 5

Longevity......... .......................... 6

Number of Seeds Produced.................... 9

Cropping Systems Effects.................... 10

Effects of Chemical Weed Control on
Seedbanks .................................. 18

Distribution of Seedbank ................... 21

Predicting Weed Problems................... 25


Introduction................................ 28

Methods and Materials..................... 30

Results and Discussion...................... 40

DISTRIBUTION................................... 53

Introduction.................. ......... ... 53

Methods and Materials....................... 55


Results and Discussion.................... 59


Introduction................................ 71

Methods and Materials ..................... 72

Results and Discussion ................... 76

V SUMMARY AND CONCLUSIONS........................ 85

Techniques........................................ 85

Tillage Systems........... ................ 86

Predicting Infestations.................... 88

Future Studies ............................. 88

LITERATURE CITED..................................... 90

BIOGRAPHICAL SKETCH.................. ............... 97

Abstract of Dissertation Presented to the Graduate
School of the University of Florida in Partial
Fulfillment of the Requirements for the
Degree of Doctor of Philosophy

Techniques for Determining the Effects of Tillage on the
Distribution of Sicklepod (Cassia obtusifolia L.) Seed
in the Soil


Larry Augustus Hacker

May, 1986

Chairman: Dr. David H. Teem
Major Department: Agronomy

Field experiments were conducted from 1982 to 1984 at

the University of Florida to examine new techniques for

determining the distribution of sicklepod (Cassia

obtusifolia L.) seed in the soil following conventional,

minimum, and no-tillage operations.

Techniques involved were the development of a new

sampling device, synthetic seeds, and analysis by x-ray.

The sampling device and synthetic seeds were mainly

developed to assist in the use of x-ray for seed recovery.

Comparisons were made between the conventional and new

techniques to determine their effectiveness.

The results indicated that the new methods were

accurate tools for the study of seed distribution. The

synthetic seeds were effective in terms of providing

recapture with relatively no degradation. Two load

materials were used to load the synthetic seeds. Barium

sulfate was the best in that it allowed for better visual

separation on the radiographs. The x-ray technique was

highly effective when compared to an older wash method in

that it was faster and more accurate.

Conventional tillage was shown to result in a

distribution of approximately 75-80% of the recovered

synthetic seeds below the 10 cm level following a single

operation. However, following the second operation, the

distribution became more uniform throughout the working

depth. Significant differences were obtained between the

single and double tillage treatments with increases in the

top 6 cm and reductions in lower levels of the profile.

Minimum and no-tillage had similar distribution patterns for

all sample times with 80-90% of the synthetic seeds

recovered in the top 6 cm.

Infestation levels were found to vary between tillage

systems. Means of 10-14, 50-57, and 44-58 percent emergence

of the initial dispersed viable seeds were recorded for the

conventional, minimum, and no-tillage plots, respectively.

Thus a definite relationship exists between the resulting

distribution and later infestation levels. The conventional

tillage plots resulted in the lowest infestation numbers,

yet it was at the cost of an increase to the seedbank in the




Techniques of Seed Recovery

The study of weed seeds in the soil requires that

efficient techniques of seed recovery be developed. Most of

the early research was conducted in an attempt to examine

the different weed flora of various locations. Some early

work, reported to be the first, was that conducted by

Putensen (34) in which soil samples were removed at three

soil depths. The samples were spread in trays and placed in

the greenhouse, and the number of emerging seedlings was

recorded to determine the number of viable seeds. He also

collected the remaining seeds by wet-sieving the soil

material after germination had ceased. Snell (58) utilized

this technique in his study in which he sampled the soil

seedbank in 10 cm increments to a depth of 120 cm and

monitored the germination at each increment. Later,

Brenchley and Warington (3,4,5) conducted experiments

utilizing a sampler that removed a 3 x 4 x 6 in deep plug of

soil. They incorporated this sampling method in studies

examining weed seed populations in fallow systems,

cultivation programs, and re-establishment of weed species

following various programs. Many researchers have utilized

variations of both these sampling methods and analysis of

cores removed for viable seed estimations.

An early problem was noted in respect to soil sampling

for estimation of seedbank size in that statistical analysis

often varied between sampler size and number of samples

removed (6,37). Roberts (37) tested two different sampler

sizes in an attempt to better understand this problem. He

collected data from a small sampler (2.3 cm internal

diameter) and a large sampler (4.4 cm internal diameter).

Three samples were extracted from each plot with the larger

device and eight samples from each plot with the smaller

device. All samples were removed to a depth of 15 cm.

Results indicated that by utilizing the smaller sampler,

removing a greater number of samples increased precision.

This probably minimized the chance of variation resulting

from collecting larger samples in areas where seeds

aggregated. This effect was also reported by Champness (6).

Robinson (53) compared the effect of the number of samples

removed with the same device. In this study, two large

samples of 154 plugs each or four small samples of 19 plugs

each were taken with a sampler of approximately 1 in

internal diameter. The data collected showed that the

degree of varability was less with the smaller sample size.

This was supported by the fact that the smaller sample size

gave much higher estimations of seed number and therefore

could explain very high populations reported in past

studies. In order to better analyze the samples, Robinson

handled each method in a different manner. The large

samples were washed using sieves of various sizes. Then the

collected material was subjected to the germination test.

The smaller samples were handled by spreading the collected

soil in trays, placing them in favorable conditions for

germination of viable seeds, and counting emerging

seedlings. Although the methods were different, the

importance of sample number is clearly shown.

Most of the early work involved the estimation of weed

seed population entirely on the emergence test in which the

samples were placed in containers, subjected to a favorable

controlled environment and monitoring emergence. This

method allows for good estimations of viable weed seeds.

However, it typically requires large amounts of greenhouse

space and long time frames to allow total germination of

viable seeds to occur. This time period also involves

careful monitoring of emergence since some species die

rapidly following emergence. Kropac (24) discussed the work

of various researches and their many modifications of a new

flotation method of separation. It was felt that this new

method was somewhat costly, time consuming, and possibly

could be detrimental to seed viability if careful selection

of density altering chemicals was not utilized. Malone et

al. (27) utilized a flotation method of seed separation in

which seed viability was not altered. Separation was

obtained rapidly and effectively by mixing a solution of

sodium hexametaphosphate, sodium bicarbonate and magnesium

sulfate. The first two chemicals aided in the breakdown of

soil aggregates and enhanced the separation while the

magnesium sulfate allowed for the flotation of the seeds.

This material was then decanted off and wet-sieved after

which seeds were separated by hand with the aid of a


Following the development of synthetic screen materials

new methods were developed for seed extraction from the

soil. Fay and Olson (16) utilized mesh bags to remove seeds

from soil samples. Samples were placed in nylon bags which

were then placed in a washing device that gently agitated

the sample with the aid of an electric motor. The bags were

then removed and seeds obtained. Thorsen and Crabtree (64)

also used a mechanical washing device to wash away the soil

in polyester bags. Their device was motorized such that a

container tumbled and was flushed with a continuous water

flow. They also stated that the device could be modified

for unbagged samples by placing a mesh material on the

bottom of the container. Following the washing of free

samples, they suggested that the collected material could

then be separated by using sieves of various sizes.

The problem with the flotation method, or any method

involving wet sieving, is the subsequent separation of the

collected seeds. Several solutions have been offered in

this regard. The obvious is hand separation, which is

extremely laborious especially in situations with large

numbers of seeds represented by various species. Although

this method is slow, the early research on seed recovery

typically involved the germination test in which the

collected soil samples were spread out in greenhouses for

germination, identification, and counts. This method

required extended time frames, often greater than three

years, in a controlled environment. In order to decrease

the time requirements for such studies, other methods have

been developed. Standifer (60) utilized air velocity to

separate different species of seeds from collected residue

and found it to be an effective method; however, this method

is somewhat less precise than desired. Based on these

observations, it appears that researchers should focus on a

particular species to reduce the time requirements. Fay and

Olson (16) demonstrated that the time required could be

greatly reduced by focusing on one species. They examined

the distribution of wild oat (Avena fatua L.) in a study and

found that the subsequent separation was rather easy due to

the elimination of the other species in the samples


From this discussion it is evident that several

techniques have been utilized to determine the actual seed

number, and that the importance of new techniques is

important to future research on the study of soil seedbanks.

Estimation of Seed Number

All of the methods described above have been utilized

to estimate the number of seeds in various soils.

Researchers estimated seed numbers in the cultivated soil

layers to range from 10,000 to 100,000 seeds/m2 (2,7,

11,22,27,28,29). Many factors are interrelated in regard to

seed bank populations some of which include longevity,

number of seeds produced, cropping practices and herbicide



Many studies have examined the longevity of weed seeds

in the soil (14,15,23,26,67). Probably the best known study

was initiated by Beal who placed seeds of 20 different

species mixed with sand in pint bottles. The uncorked

bottles were then buried about 46 cm below the surface.

They were tilted to prevent accumulation of water in the

containers. Since the initiation of the study, bottles were

exhumed and the seeds subjected to germination tests.

Nearly half of the species had lost all viability by the

fifth year; however, eleven species were shown to be viable

after 20 years, two species were viable after 70 years, and

three species were found to be viable after 100 years

(14,23). A similar study was initiated by Duvel and

reported by Toole et al. (66) in which 107 species were

placed in sterilized soil buried in flower pots covered with

a porous saucer. These studies indicated the importance of

seed longevity in respect to potential problems from buried

weed seed. Apart from longevity of buried seeds in

undisturbed soil, these studies do not demonstrate the

effects of disturbances to the soil profile in respect to


Another aspect of seed longevity is that of dormancy

mechanisms. Roberts (36) suggested that weed seeds which

are buried in the soil are subjected to some form of

dormancy if they are to survive. Virtually all seeds are at

some time during their burial subjected to adequate moisture

conditions for germination. The degree in which soil

moisture affects seed longevity has been examined in terms

of degeneration of seeds (25,67). Villiers (67) suggested

that weed seeds often have a repair mechanism that allows a

reduction in the rate of degeneration, enabling the seeds to

persist for longer periods. However, as shown in earlier

research, once soil cores are subjected to optimum

germination conditions rapid emergence follows. Therefore,

the rapidity of this response indicates that before these

seeds were disturbed, they were held in check by "enforced"

dormancy mechanisms. Enforced dormancy as described by

Harper (19) is a condition when viable seeds do not

germinate because of some limitation in their environment.

This type of dormancy is typically used to describe the

dormancy that is found on seeds buried in the soil, due to

the fact that germination occurs following extraction from

the soil and placement in conditions conducive to

germination. Although "enforced" dormancy is important, it

is not the only mechanism at work on buried seeds. Possibly

all three types of dormancy as described by Harper (19) are

found in some species. A better understanding of dormancy

in relation to seeds in the soil seedbank is needed.

One common dormancy mechanism in relation to weed seeds

is seedcoat hardness. Lueschen and Andersen (26) have an

ongoing study involving velvetleaf (Abutilon theophrasti

Medic.) in which they found seedcoat to be important in

terms of hardness. Along with data from other studies, much

work has shown the importance of seedcoat hardness as a

dormancy mechanism or survival mechanism. Stoller and Wax

(62) discussed the importance of seedcoat hardness of three

species: common ragweed (Ambrosia artimisiifolia L.),

ivyleaf morningglory [Ipomoea hederacea (L.) Jacq.], and

velvetleaf. They stated that this one characteristic was

possibly the most important in relation to survival of these

species in the soil. They also reported that depth of

burial was important from a standpoint of seed decay.

Results indicated that decay was faster at the 2.5 cm depth

than at the 10.2 cm depth.

Creel et al. (12) examined the seedcoat of sicklepod

(Cassia obtusifolia L.) and found it had a hard wax-like

covering that was approximately 0.1 mm thick. The embryo

itself was surrounded by a dense hard material. He also

subjected unscarified seeds to burial for 12 months at 23 C

in moist soil while germination was monitored. Results

demonstrated that only 15% of the buried seeds germinated,

and 5% of these did so after only two days. The other 10%

germinated over a thirty day period. Also working with

sicklepod and coffee senna (Cassia occidentalis L.), Teem et

al. (63) found that both species germinated under a wide

range of temperatures from 24 to 36 C with optimum seedling

growth from 30 to 36 C. They also found the seeds of both

species to respond to various types of scarification and

that both species can emerge from a depth of 12.7 cm. The

percent emergence was low being only 25% for sicklepod and

50% for coffee senna. Both species still demonstrated the

ability to emerge from these depths which greatly increases

their survival chances. In relation to the rate of

emergence, sicklepod was shown to obtain 63% emergence after

only three days from a depth of 2.5 cm. This rapid flush of

seedlings in a field situation would create a large problem

in terms of weed control and crop competition, which in turn

could possibly generate new seed sources for the soil


These studies indicate the importance of seed longevity

in terms of burial in the soil and also demonstrate the

potential problems that can occur in terms of weed

infestation. Yet, apart from the longevity of buried seeds

and some discussion of dormancy mechanisms which affect

survival, they do not offer explanations of longevity in

relation to soil disturbance and to the soil profile.

Number of Seeds Produced

In general, weeds are extremely dynamic in terms of

seed production. Weeds typically can produce vast numbers

of viable seeds per plant under most conditions. Stevens

(61) reported how prolific many species are in terms of

production with numbers ranging from approximately 2,000 to

over one million seeds produced by a single plant. Anderson

(1) reported that seed number is clearly related to seed

size and weight. Thus weeds with extremely small seeds tend

to have higher numbers produced per plant. Recent research

conducted by Jordan (21) offered new information about the

number of seeds produced per plant. In studies involving

sicklepod, numbers of seeds produced per plant were recorded

in a study comparing soybean [Glycine max (L.) Merr.] row

spacing and sicklepod density with yield. At a row spacing

of 75 cm and a sicklepod density of 0.5 plants/m2, a two

year average of 12.98 million seeds was produced per

hectare. Thus he obtained data showing plants in these

conditions produced approximately 2600 seeds/plant. The

greatest production reported was 168.15 million seeds/ha,

which occurred in the second year of the study at a density

of 5.0 plants/m2 and a soybean row spacing of 75 cm. This

represents approximately 3363 seeds/plant. Creel et

al. (12) stated that sicklepod seed yields of greater than

900 kg/ha have been calculated from cotton fields which were

heavily infested with sicklepod. These studies indicate the

importance of seed production in terms of replenishment of

the soil seed bank and survival of the species.

Cropping Systems Effects

Early studies of effects of cultivation on seed

longevity were typically conducted by artificial means.

Chepil (9) initiated studies in which metal trays in the

field acted as plots; each tray contained soil and seeds of

numerous species. Over a six year period, the trays were

subjected to shallow cultivation by hand. In a second

study, Chepil (10) again utilized small metal trays;

however, this time the focus was on five weed species and

evaluated over a shorter time frame. He subjected the trays

to various types of cultivation and concluded that the

number of viable seeds decreased more under cultivation

systems when compared to the undisturbed trays.

Roberts (41) initiated four experiments in which seeds

of eleven weed species were mixed in a 7.5 cm layer of soil

in pots which were buried in the ground. Three studies were

continued for approximately 5.5 years. On five or six

occasions during each year, the soil in the pots was

thoroughly mixed by hand. From these studies he concluded

that there was a progressive decline in the number of

seedlings emerging. Also, the rate of decline for each

species was found to be approximately constant, although

some species declined faster than others. From an earlier

study, Roberts (38) reported a decline of viable seeds of

approximately 50% per year in field situations involving

cultivation. Since in this study the rate of decline was

determined by emerging seedlings, he could only approximate

the decline of viable seeds in this later experiment. When

all species were combined however, the results were shown to

fit that of the earlier test in terms of rate of decline.

One additional observation was that after the five-year

period seeds were still emerging although quite few in

number, thus indicating the effectiveness of some species in

terms of survival in cultivated soils.

Similar studies were conducted by Roberts and Feast

(46,47) in which they subjected seeds buried in earthenware

cylinders to different cultivation regimes by hand

disturbance. In one study, the soil was either disturbed to

a depth of 15 cm three to five times a year, or it was left

undisturbed. Seedling emergence was monitored for six years

and each year a calculation was made of viable seeds

remaining. The mean numbers of viable seeds remaining after

completion of the sixth year was 5.9% of those initially

added in the disturbed cylinders, and 27.5% where the soil

was left undisturbed (47).

In an earlier study, Roberts and Feast (46)

incorporated depth into the test. Seeds were buried at

depths of 2.5, 7.5, and 15 cm. The depth of soil

disturbance was also tested at these various depths while

the checks were left undisturbed following burial. The data

collected from both the disturbed and undisturbed cylinders

demonstrated the effect of depth of seeds in the soil in

respect to depletion. After a five year period, the

percentages of viable seeds remaining at depths of 2.5, 7.5,

and 15 cm were 2.3, 4.0, and 7.7% respectively under the

disturbed conditions, and 6.8, 16.5, and 31.6 for the

undisturbed areas,respectively. All numbers were expressed

as percent of initial seeds buried. An interesting point

was that no two species responded in exactly the same way;

thus Roberts and Feast (46) stated that research should be

initiated on individual species in order to obtain a better

understanding of underlying differences in the mechanisms of

survival of buried seeds.

Although these studies involved artificial tillage

systems, and other conditions different from actual field

situations, they did demonstrate that seeds in the soil are

affected by tillage and that the existing seeds tend to

decrease exponentially with time in soils that are

undergoing cultivation.

Early work by Brenchley and Warington (2,3,4,5)

compared various types of cultural systems on existing weed

seed in the soil under natural conditions. They conducted

tests on areas consisting of both fallow and cultivation

systems, and later examined the re-establishment of weed

flora to these areas in relation to the soil seed bank. In

their earliest study involving different fallowing systems,

they found that by subjecting the plots to four year

fallowing periods the weed seed population could be greatly


Roberts conducted a series of tests involving vegetable

cropping systems (37,38,39,40,51) which resulted in five

papers. The results obtained in the second paper of the

series best summarize the effects of six years of cropping

on weeds in the soil (38). The field in which these studies

were conducted was shown to have a high initial population

of viable seeds. Samples taken in 1953 by Roberts (37)

determined which would be used in the later studies. During

the first of four years of cropping there was a progressive

fall in total weed seed population when compared to 1953

data. Also with the absence of seed replenishment, results

indicated that the number of viable seeds in any one year

was approximately half that of the previous year. During

the fifth year of the study, wet soil conditions at the time

of cultivation resulted in somewhat less effective control

than in the previous years resulting in extensive

replenishing of the soil seedbank. In the sixth year the

population was again reduced as in the previous years.

In the third paper (39), Roberts examined four

different primary cultivations on the weed seeds in the

soil. The comparisons consisted of deep plowing (36-41 cm),

shallow plowing (15-18 cm), shallow plowing (15-18 cm) plus

subsoiling (41-46 cm), and rotary cultivation (15-18 cm).

Over the six year period, the three plowing systems had

substantially the same effects on both the number of viable

seeds and on subsequent infestation levels. However,

following occasions in which seed production occurred, seed

number was shown to be greater in the top 15 cm on plots

receiving shallow plowings when compared to those deeply

worked. The rotary cultivation method, to the same depth as

the shallow plowing, was found to have greater numbers of

viable seeds of certain species when these two were

compared. Roberts also reported that both the deep plowing

and shallow plowing plus subsoiling methods decreased the

numbers above 15 cm while increasing numbers below 15 cm.

Uniform distribution patterns were also found in plots

receiving continuous deep plowing methods. In general, with

the absence of seed production, all methods of cultivation

reduced the population of viable seeds in the soil. Yet, if

plants were allowed to produce seeds the rotary cultivation

method allowed for build-up of seeds in the upper soil


Roberts (40) reported on cultivation methods with some

additional procedures added to the earlier study. Results

revealed that numbers of weed seeds and degree of

infestation was essentially the same for plowing (18-23 cm),

with and without subsoiling, deep plowing (30 cm) or hand

digging to 18-23 cm. Yet with shallow plowing (7.5-10 cm),

plots were extremely weedy which was thought to be due to

inadequate burial of seeds. As with the previous paper

(39), the plots with the highest seed number and largest

infestation levels were those receiving rotary cultivation

particular those with shallow rotary cultivation (7.5-10

cm). Roberts and Stokes (51) reported that the primary

cultivation methods utilized followed the results found in

the earlier years of the study in respect to seed

populations and infestation levels.

Roberts and Dawkins (45) conducted a six-year study on

the effects of cultivation on weed seeds in the soil.

During the course of the study, no crops were grown on the

test area. Plots were treated in the following methods:

dug to a depth of 23 cm each year in mid-March, mid-June,

mid-September, and mid-December; dug to a depth of 23 cm

each year in mid-March and mid-December; or left

undisturbed each year. Initial populations were determined

by sampling methods described earlier for determination of

population changes. Under all methods, the number of seeds

in the 23 cm profile decreased exponentially from year to

year; however, the degree of seed depletion occurred at

different rates for each cultivation system. On the

undisturbed soil, losses were found to be 22% per year, with

the plots disturbed four times a year and two times a year

resulting in depletion rates of 36 and 30%, respectively.

These depletions were related in terms of infestation levels

with an annual decrease in infestation of 9, 7, and 3% from

the most to the least disturbed plots. In this study, as in

others, the decline of individual species was typically at

different rates, yet generally the decline was exponential.

Schwerzel and Thomas (56) also examined the effects of

cultivation frequency on certain weed species; however, they

introduced a species into an area that was essentially void

of the introduced species. This approach allowed them to

follow seed depletion without the need of initial population

estimation. Seeds of five species were mixed into the soil

to a depth of 22.5 cm in each plot. Plots were then

subjected to no cultivation, one cultivation, four

cultivations, or twelve cultivations. Cultivations were

done by hand-hoeing to a depth of 22.5 cm. They reported

that percent emergence of the original population over a ten

year period was as follows: no cultivation 20.0%, one

cultivation 34.2%, four cultivations 29.7%, and twelve

cultivations 41.8%. Although this does not agree with

reports by Roberts (38), it still demonstrates the

importance of soil disturbance in respect to depletion of

soil seed reserves.

An ongoing experiment involving seven cultural

practices on an area heavily infested with velvetleaf seed

is presently being conducted by Lueschen and Andersen (26).

They are attempting to determine the time period required

for total depletion of the soil seed bank. The first four

years of this study showed that the population of velvetleaf

seeds declined rapidly before leveling off, and this decline

was increased by tillage operations. The most intensive

tillage system, two plowings with continuous fallowing,

resulted in the fastest decline over this four-year period.

However, even though extreme care was taken to prevent

plants from producing new seeds, the population under this

intensive system was still estimated at 1300 viable

seeds/m2 to a depth of 23 cm. Although the initial

population was 130 million, this illustrates the difficulty

of totally eradicating such a vast number of viable seeds.

Pollard and Cussans (31,32) measured the effect of

different tillage systems on the weed flora by monitoring

emergence of seedlings under each tillage regime. Tillage

regimes were moldboard plowing, shallow or deep tine

cultivation, and direct drilling. They reported that small

seeded species were favored by reduced cultivation or direct

drilling, while larger seeded species were favored by

moldboard plowing and deep tine cultivation.

Froud-Williams et al. (17) also investigated the

effects of different cultivation regimes on weed floras of

arable soils. The cultivation treatments common to all five

locations were no cultivation and plowing to a depth of 20

cm. Typically they found that the total number of weeds

emerging in the count areas varied with cultivation

treatments depending on the dominant species at each

location. Total weed numbers were higher on plowed plots

than on the undisturbed plots at two sites where grasses

were not the predominant species. However, at the other

locations the reverse was true due to higher frequencies of

grass species. Their results indicate the importance of

understanding the effects of tillage on monocotyledonous and

dicotyledonous species and that different responses can be

related to seed size. Reports of reductions in the density

of dicotyledonous weeds and increases in the frequency of

annual grass weeds with the implementation of reduced

tillage operations were found by several researchers


Effects of Chemical Weed Control on Seedbanks

In an attempt to better understand how the addition of

chemical weed control affects the weed seed reserves in the

soil, Chancellor (8) examined the rate of increase of seeds

in the soil in the absence of chemical control. Plots were

either not sprayed or they were sprayed with a herbicide

mixture of bromoxynil (3,5-dibromo-3-sec-butyl-6-methylura-

cil), ioxinyl (4-hydroxy-3,5- diiodobenzonitrile), and

dichlorprop [2-(2,4-dichlorophenoxy)propionic acid]. After

a two year period, both treatments were shown to have 21

species making up the soil seedbank; thus the chemical

treatment did not change the seedbank composition. However,

as expected the number of seedlings emerging had changed in

regard to treatment. The total number of emerged seedlings

was 686 on the sprayed plots and 1402 on the untreated

plots. It would appear from these observations that the

unsprayed plots had doubled after a two year period; yet

when these figures were compared with those collected at the

time of initiation of the experiment, he reported that a 10%

decline occurred on the untreated area, and a 52% decline

occurred in the treated area. He suggested that some of

these variations were possibly due to differing temperatures

during seed set periods thus altering the seed rain to the

soil surface or the effect of cultivation timing in terms of

soil moisture and weather conditions following these


A twelve year study was conducted by Hurle (20) in

which the long-term effects of weed control on viable weed

seeds in the soil were examined. Seven treatments were

utilized and included the following: (1) no weed control,

(2) MCPA {[(4-chloro-o-tolyl)oxy]acetic acid) at 1.5 kg/ha,

(3) 2,4-D [(2,4-dichlorophenoxy)acetic acid] at 1.5

kg/ha,(4) DNOC (4,6-dinitro-o-cresol) at 4.0 kg/ha,

(5) Calcium cyanide at 143 kg/ha, (6) a rotation treatment

which consisted of MCPA, DNOC, Calcium cyanide, and

harrowing, and (7) a harrowing alone treatment. He found

that all weed control measures reduced the number of viable

seeds in the soil over that of no weed control. The DNOC

treatment resulted in the greatest reduction at 60% with the

other three herbicide programs offering approximately 36-44%

reduction. The rotation treatment and harrowing alone were

the least effective overall.

Roberts and Neilson (48) determined the effect of

long-term chemical control on four monoculture systems in

respect to soil seedbanks. They compared spring barley

treated with triallate [S-(2,3,3-trichloroallyl) diisopro-

pylthiocarbamate] at 1.68 kg/ha, spring barley treated with

MCPA at 1.68 kg/ha, maize treated with simazine [2-chloro-

4,6-bis(ethylamino)-s- triazine at 1.68 kg/ha, and carrots

treated with linuron [3-(3,4-dichlorophenyl)-1-methoxy-l-

methylurea] at 0.84-1.68 kg/ha either as a postemergent or

in split applications. The reduction in number of viable

seeds was expressed by the comparison of seeds recovered to

those numbers found in 1963. Initially, in the second year

of the study, they reported an appreciable reduction of

viable seeds in all four monoculture systems. This decline

also continued throughout the next five years of the

experiment. However, changes in actual composition of the

seedbank were directly related to the cropping system and

herbicide used. There were shifts in response to these

treatments, and although reductions occurred, little

qualitative effects on species composition were reported.

Thurston (65) reported on a seven year study where

continuous winter wheat was grown. Various compounds, no

weed control, and a continuous fallow situation were

compared. He found that the number of species had not

changed; yet the densities had changed greatly due to

treatments imposed. Basically the addition of herbicides to

the study of seedbank depletion has demonstrated the effect

of various compounds on existing weed seeds. One thing that

should always be kept in mind is the possibility of

population shifts due to poor weed control of a certain

species with a certain compound (65).

Distribution of Seedbank

In order to better understand the importance of seed

movement in regards to cultural practices, many researchers

have attempted to examine distribution in relation to these

operations (52,54,56,59). Robinson and Kust (52) analyzed

the soil profile to determine the distribution of witchweed

[Striga asiatica (L.) Kuntze] seeds in the soil. Cores,

each 3.8 cm in diameter, were taken to a depth of 152.5 cm

and divided into 15.25 cm sections. Each section was then

separated by a flotation technique and seeds recovered.

They reported that the vertical distribution was not uniform

and that although the greatest numbers were found near the

soil surface, the cultural practices that formed this

distribution were not known.

In an attempt to better understand these distribution

patterns in regard to tillage systems, Soriano et al. (59)

hand dispersed seeds of two crop species on the soil surface
at a density of 4000 seeds/100m2. The area was then plowed

to a depth of 20 cm and disked several times. Samples were

taken by randomly removing 30 blocks of soil within the

sample area. Each 30 X 10 X 20 cm sample was extracted with

the aid of a metal frame which was hammered in the ground,

soil was removed from around it, and the block was then

lifted out of the field. In order to examine vertical

distribution, the block was then sliced into 4 cm thick

layers. Each layer was then wet sieved with the aid of

running water. They reported that although the

characteristics of the two seeds in terms of geometry and

seedcoat were very different, the distribution was the same.

A very characteristic vertical distribution was found with

both species increasing in frequency from the surface to a

depth of 20 cm. Although crop seeds were used in this

study, they were selected due to similarities in regard to

weed seed characteristics of both size and seedcoat.

Kellman (22) made observations of both horizontal and

vertical distribution of seeds in two cropping situations.

In both a corn field and pasture, a 1 m baseline was

established and contiguous cores removed along this baseline

to a depth of 10 cm. Each core was then sliced into the

following sections: 0-1, 1-3, 3-6, and 6-10 cm. Sections

were then wet sieved using a 0.25 mm mesh and the sieved

material collected for germination test. Results indicated

that the pasture had many more species in the 0-3 cm layers

than the corn field; yet the corn field had a much greater

seed number in these surface layers. He proposed that the

vertical pattern in the corn field was probably a

combination of older, more evenly distributed seeds buried

by prolonged tillage, and very recent seeds produced by the

existing weed vegetation being concentrated at the surface.

Fay and Olson (16) monitored the distribution of wild

oat seeds in the soil profile under two tillage systems of

moldboard plowing or chisel plowing. Results showed that

seeds were more evenly distributed throughout the 18 cm soil

profile by annual moldboard plowing than by chisel plowing.

Over 60% of the wild oat seeds recovered occurred in the top

2.5 cm with none recovered below 10.2 cm in the soil

receiving the chisel plow treatment. Therefore, they

suggested that seeds would tend to concentrate near the soil

surface with continued chisel plowing as opposed to a more

uniform distribution with moldboard plowing.

Froud-Williams et al. (18) conducted a study in which

seed distribution as influenced by cultivation regime was

compared over a three year period. Using sampling

techniques discussed earlier and subsequent analysis of

cores with the emergence test, they effectively determined

vertical seed distribution. Distribution was determined in

5 cm layers to a depth of 20 cm. The two cultivation

regimes involved were direct drilling in uncultivated plots

or conventional seeding in plots plowed to 20 cm. They

found that the distribution of seeds in the soil profile was

not uniform, and fluctuated with each annual plowing of the

plots at all locations. Seed distribution on plowed plots

tends to reverse each year in terms of greatest number of

seeds. After the first years plowing, at two locations the

concentrated layers appeared at 10-15 cm. Following the

second year's plowing, distribution was relocated in terms

of major seed bands. It is clear from this work, that

plowing as a cultivation method results in severe relocation

of fallen seeds. In the no cultivation plots, very little

change in concentration occurred with the exception of one

site which possibly had undetected seed production. They

also stated that the rate of seed decline is directly

associated with soil disturbance.

A new method of examing seed distribution was

introduced by Rottele and Koch (54). Part of their study

involved the use of plastic pearls in an attempt to simulate

the seed of cleavers (Galium arparine L.). Both true seeds

and plastic pearls were distributed on an area 25 m2 and

replicated four times. Samples were taken with a soil probe

to a depth of 25 cm. All samples were sliced into 5 cm

sized pieces and stored in plastic bags until analysis.

Both seeds and plastic pearls were then extracted from the

soil by washing the sample sections in a sieve (1 mm) until

all the soil was removed. Both the seeds and pearls

remaining were counted and recorded. The results indicated

not only that the plastic pearls and seed demonstrated

similar distribution patterns, but also that the

distribution of both responded the same to each additional

tillage operation. After the first plowing (- 25 cm)

vertical distribution of both viable seeds and plastic

pearls were shown to have approximately 36% and 35% of each

recovered in the 15-20 cm zone. In the case of the plastic

pearls, 79% of the recovery was below the 10 cm depth with

85% of the viable seeds recovered below this depth.

Following the second plowing, both the seeds and plastic

pearls were returned to the upper layers. Recovery numbers

for seeds and pearls from 0-15 cm were 71% and 82%,

respectively. In terms of an indicator, the plastic pearls

were shown to mimic the actual seed well, and this allowed

for analysis of seed response to tillage.

Predicting Weed Problems

According to Radosevich and Holt (35), the importance

of research involving the determination of soil seed banks

and early observations of seedling establishment from these

reserves offers the ability to predict potential weed

infestations. By inputing the parameters of seed production

and dispersal, soil seed bank size, rates of seedling

recruitment, and expected mortality into a demographic

model, it should be possible to determine both the dominant

species and its predicted density at a particular site.

Sagar and Mortimer (55) have developed a "working

model" which can be used to study population dynamics of

weeds. The model uses a known population of plants and

predicts the expected infestation levels for the next year.

Mortimer (30) examined four weed species with the aid of

similar models although they were simplified to examine the

effects of various levels of disturbance and the exclusion

of animals from the study area. Attempts were made to

compare the treatments in regard to seeds becoming buried,

eaten by predators, germinating, and forming of healthy


Roberts and Potter (49) studied the emergence patterns

of weed species in relation to cultivation and rainfall.

The study was conducted over a four year period, with a

cultivation every two weeks. The number of seedlings

emerging was counted weekly and showed the effects of

cultivation and rainfall on seedling emergence. They found

a flush in the spring of each year, and that subsequent

flushes on both disturbed and undisturbed soil were directly

related to rainfall. They estimated 90% of the seedling

emergence following disturbance occurred within ten weeks

following the early spring cultivation, and three weeks

after cultivation in the summer. A better understanding of

the influence of cultivation on germination and the

relationship to the soil seedbank may allow predictions of

weed emergence. Roberts and Ricketts (50) conducted a study

in which this association was examined. By comparing


seedling emergence in respect to cultivation with existing

viable seeds in the soil, they examined the possibilities of

a quantitative prediction of the weed flora. They found

that when soil moisture was adequate for seed germination,

only 3-6% of the seeds in the soil to a depth of 10 cm

germinated. They felt that this type of data between actual

and potential weed flora can offer predictive values with

additional study in regard to cultural practices and time.



Weed seeds in the soil have been studied in an attempt

to better understand their behavior and to obtain an increase

in their control. Initially researchers studied the ability

of certain species to survive for long periods of burial in

the soil (14,23,66). Longevity is possibly the most

important aspect involved in the typical formation of large

reserves in the soil, called soil seedbanks. The importance

of these studies is that they were the first to demonstrate

the ability of weed seeds to remain viable in the soil for

extended periods.

In order to study the soil seedbanks in their natural

state, methods of seed recovery are of importance.

Researchers have utilized various soil core sampling tools

in studies to recover weed seeds in the soil (4,10,42).

Studies have also been conducted on the size and number of

cores to be removed for the least variation in population

estimations (6,37,53). The development of sampling methods

lead to the study of separation techniques which allowed for

seeds to be removed from the soil more efficiently and in

less time. This showed that the seedling emergence method

was both time consuming and required a controlled

environment to be most effective.

The next step was to study these buried weed seeds in

relation to actual cropping systems. Roberts and co-workers

(41,43,44,45,46,47,48) conducted an array of studies

offering information about the decline of populations under

normal cropping practices and the effects of these practices

on the many aspects of seeds in the soil. In order to more

clearly understand the effects of tillage on weed seeds in

the soil, Rottele et al. (54) introduced a new method to

examine seed distribution with the aid of plastic pearls.

He utilized the plastic pearls to simulate the seed of

cleavers and subjected them both to cultural practices. By

sampling the area receiving the artificial seeds he studied

the changes in distribution with each additional tillage

operation. This novel approach with modification allows for

closer examination of weed seeds in the soil and of those

factors which influence their survival or demise.

The objective of this study was to examine these

techniques more closely, and possibly develop additional

techniques for the study of weed seeds in the soil. Primary

interest was the effects of tillage operations on the

distribution of sicklepod seeds in the soil, and new methods

of determination of the soil seedbank distribution.

Methods and Materials


During the fall of 1982, an area of land with a seven

year history of pasture grass breeding was located. This

area was chosen due to its history of low broadleaf weed

infestation. The area was randomly sampled in twenty

places. Each sample consisted of taking approximately one

cubic foot of soil by shovel. The samples were then taken

to the greenhouse, spread out in trays and watered.

Seedling emergence was monitored, with the main goal to

determine presence or absence of sicklepod. This method

indicated that sicklepod seeds were not present; however,

some of the samples were also screened through sieves to

collect existing seeds in the soil. No sicklepod was found

in the screenings. Since no sicklepod was present, any

sicklepod emerging would be from those dispersed at

initiation of the study.

In the winter of 1982, a study was established at the

University of Florida Agronomy Farm on an Arredondo fine

sand soil (siliceous, well drained hypothermic, Grossarenic

Paleudult). A split-plot experiment with a randomized

complete block design and six replications was used. The

main plots were tillage systems. The subplots were areas

receiving year one viable sicklepod seed, year two viable

sicklepod seed, year one and year two synthetic seed, and an

unseeded check. The main plots measured 8.5 X 12.2 m, and

the subplots measured 4.25 X 6.1 m with four subplots per

main plot. All tillage treatments were maintained in the

same plots for two years. Both tillage systems and seed

dispersal in subplots will be discussed individually.

Viable Seed Collection

In the fall of 1982 and 1983, sicklepod seeds were

collected from a local field on the University of Florida

Agronomy Farm in which high infestation levels were found.

Seeds were collected by removing mature dry plants from the

field and hand threshing them over a 1.8 m2 cloth. Seeds

were then gathered and hand cleaned with the aid of various

sieves and compressed air. Care was taken throughout the

collection and cleaning to reduce the chances of seedcoat

damage. After cleaning, the seeds were stored at room

temperature in cloth bags.

Production of Synthetic Seeds

In order to develop a synthetic seed a random 100 g

sample was taken from the viable seeds collected in 1982.

This sample was used to determine the average diameter,

length, and density.

Synthetic seeds were produced with help from the

Material Science Department at the University of Florida.

Seeds were produced for both years, 1982 and 1983, from a

branched form of the polymer, polyethelene. The form used

had a low density of approximately 0.93 g/cm3. Due to the

low density of the polymer, a load material was required

to increase the density to the 1.25 g/cm3 of sicklepod.

Two separate load materials were used for 1982 and 1983

seeds to allow for identification and separation following

analysis. The two load materials utilized were iron powder

and barium sulphate for 1982 and 1983 respectively. The

materials for each of the two years were blended with the

polymer and extruded from an Extruder (Mini Max) at 175 C to

produce a rod with an average diameter of 3.0 mm. Following

production of the rod material, they were then chopped into

average lengths of 4.5 mm with the aid of a Pelletizer

(Custom Scientific Ins.). All of the procedures described

above resulted in the production of synthetic seeds that

best mimicked the true sicklepod seed collected locally in

terms of size, shape, and density. Both year one and year

two synthetic seeds were subjected to density checks and

were shown to be approximately 1.25 g/cm3

Counting Method

Both the viable seeds and the synthetic seeds were

counted with an Electronic Counter into lots of 2200

seeds/bag. This number was based on previous research

conducted at the University of Florida. Jordan (21)

reported the number of sicklepod seeds produced per plant at

a density of 0.5 plants/m2 and a soybean row spacing of 75

cm to be approximately 2100 in 1981 and 3100 in 1982. From

these data and personal discussion, a figure of 2200

seeds/plant was utilized. Although the subplot areas were

to receive 28,600 seeds each, observations were made during

the counting procedure that demonstrated less error in

smaller lot sizes; thus lots of 2200 seeds/bag were used.

Seed Dispersal

The test area was established in the winter of 1982,

and the dispersal of synthetic or viable seeds on the soil

surface of the subplot areas was accomplished by hand

spreading. Distribution was carefully executed to produce

as uniform seed distribution as possible. All areas in

which seed were dispersed were allowed 28,600 seeds which

resulted in an initial population of 1100 seeds/m2

Immediately after dispersal the entire area was disked twice

in the same direction with a tandem disk, and wheat

(Triticum aestivum L.) 'Florida 301' was planted. Second

year seeds of both types were handled in the same manner;

however, winter wheat was not sown after the two diskings,

and replication five received no seeds of either type. This

area was to be excavated later to examine the distribution

of year one synthetic seeds.

Tillage Operations

In the spring of 1983, three tillage systems were

randomly applied to each replication following the removal

of the wheat to a 15 cm stubble. The three systems were

(1) Conventional Tillage--moldboard plowing (24 cm),

followed by two diskings (7.5 cm); (2) Minimum Tillage--two

diskings (7.5 cm); and (3) No-tillage--direct seeding into

stubble. Following the completion of tillage operations,

"Bragg" soybeans were planted with a row width of 75 cm.

Early in the summer of 1984, all procedures were repeated as

in 1983; however, no soybeans were planted in the second

year and replication five was left undisturbed for

excavation and analysis of year one tillage effects. In

both years, shortly after the completion of the tillage

operations, oryzalin (3,5-dinitro-N,N-dipropylsulfanilamide)

was applied at 0.68 kg/ha over the entire test area for

annual grass control.

Seed Sampling Technique

Soil samples were extracted with the aid of a new

device developed for this and additional studies (Figure 1).

The soil sampling device was composed of three components.

The first of these was the sampler which was constructed

from 0.32 cm steel sheet in a manner that allowed the second

component to be inserted. On the bottom of the sampler were

two tooled blades approximately 5 cm high which aided in the

penetration of the sampler into the soil profile (Figure 2).

The second components, soil collection trays, were

constructed from 0.32 cm sheet aluminum to specifications

that allowed insertion into the sampler (Figure 2).

Originally the collection trays were constructed from

polyvinal cloride; however, it proved to be unstable and

trays were not strong enough to allow repeated use. These

were replaced with the trays constructed from aluminum. The

collection trays served as the container for each sample

following extraction. A sliding impact hammer was utilized

to drive the sampling device into the soil (Figure 2).

Figure 1. Soil sampling device utilized for studying the
distribution of synthetic seeds in the soil
profile. Designed to remove a vertical slice
with dimensions 39.25 X 3 X 24 cm. Top:
Insertion of collection tray into sampler
Below: Collection tray with intact soil slice

V *,;


ILi ~

(B) Impact

- s35

(A) Sampler / '

(0 252

-i7 ( 744*)

39 37cm (1549 )I

^o Tray
On'), (C) Collection

Figure 2. Schematic drawing of soil sampler illustrated
by components. Main component diagram: (A)the
sampler. Additional components: (B)collection
trays and (C)impact hammer.

Once positioned in the sampler, the inside of the tray

was flush with the bottom opening which allowed the sampler

to be impacted into the ground and the sampler filled with

undisturbed soil. Following removal of the sampler from the

soil, it was positioned such that the lid of the aluminum

tray was on top. The lid was then removed to provide more

freedom for the tray to be removed. Before the tray was

taken out a stainless steel blade was inserted in the gap

between the lip and the bottom of the tray. This separated

the sample from fill soil contained in the 5 cm extension

below the tray bottom. The tray with an intact vertical

slice of soil was then easily removed.

Following removal of the trays, samples were then

sealed by placing the lid back on the tray and taping the

bottom opening with duct tape. Two large rubber bands were

used to help hold the tray closed during transport, although

they were not necessary.

Following the tillage operation and after the soil had

been weathered to some degree, all subplots receiving the

synthetic seeds were sampled as described above. Subplots

were sampled 10 times in a 1 X 3 m area within the subplot.

Samples were then sealed, labeled, and stored for analysis

at a later time. Methods were repeated following year two

tillage treatment. No samples were removed from

replications five and six.

Excavation Methods

Efficiency and accuracy of the sampling device was

accomplished by excavation of replications five and six at

randomly designated sites within the subplot area.

Replication five received only first year synthetic seeds

and only one tillage operation on the main plot areas.

However, replication six received both first and second year

synthetic seeds and two tillage operations. These were

excavated by removing a randomly selected block of soil

measuring 60 X 80 cm at 2 cm depth intervals. Each layer

was carefully measured, the soil hand removed with the aid

of a small garden spade, and samples placed in labeled bags.

The conventional tillage subplots were excavated to a depth

of 24 cm; while both the minimum tillage and no-tillage

subplots were sampled to a depth of 16 cm. The two reduced

tillage systems were never found to have seeds below 14 cm,

thus a depth of 16 cm proved to be adequate. These bagged

layers of soil were then analyzed by the wash method

utilized for all other soil samples. Seeds were recorded,

and the data collected were utilized for comparisons with

other recovery methods.

Analysis of Soil Samples

All samples collected were subjected to a washing

technique which involved a sieve apparatus constructed of

two 39.25 X 25 cm perforated steel plates containing 5.5

perforations/cm2 with a diameter of 2.5 mm (Figure 3).

These two plates were cut such that when they were placed in





10 25'81)

Figure 3. Drawing of washing device utilized to remove soil
from samples for determination of synthetic seed
recovery. Designed to allow for high pressure
wash with reduced seed loss.

the metal box frame they were slightly offset. This allowed

for the rapid removal of soil while decreasing the chance of

seed loss. The metal box frame was 40 X 25 cm with sides 15

cm high and was placed in a stand made from angle iron.

Each sample was unsealed and marked into 2 cm sections from

the top (0 cm) to the bottom (24 cm). Each section was then

sliced, placed in the washing apparatus, flushed with

running water until all soil was removed, and the remaining

synthetic seeds were recorded. Care was taken to check the

organic matter in each sample to assure accurate seed

recovery. This technique allowed for the comparison of seed

distribution by depth and year in relation to the three

tillage systems implemented.

Results and Discussion

Comparison of Wash and Excavation

New techniques utilized

This study was initiated to evaluate the effectiveness

of two new techniques for studying the distribution of seeds

under three separate tillage regimes. The first of these, a

new sampling device, was tested for both effectiveness in

the removal of undisturbed slices of the soil profile and

accuracy in the determination of the seed distribution

within these slices of the profile. The second technique of

this study examined was the use of synthetic seeds as

mimickers of the actual weed seed. In order to determine

the degree of accuracy of seed recovery of the sampling

device, an excavation treatment was implemented as a control

for examing seed distribution.

After the tillage treatments the test area received

both the sampling and later parts were excavated. A

comparison of these two methods of seed recovery was made

first. Overall, there was little difference between the two

methods. Also there appeared to be no compaction with the

sampling device. There were some differences in the percent

seed recovered. These differences typically occurred in the

reduced tillage systems in the upper soil layers. Possibly

a large portion of these differences was due to the

difficulty in excavating the upper layers of soil. When the

excavation was initiated, the soil surface was generally

uneven and therefore made the initial layers difficult to

remove exactly 2 cm at a time. Yet, even with these

differences, the two methods were very much the same in

terms of predicting the actual distribution of seeds under

the three tillage systems.

Conventional tillage

The wash and excavation techniques resulted in

essentially the same seed distribution pattern (Figure 4).

These data show the means for the three sample times at each

depth interval. The same data also show that the wash

method is not significantly different from the excavation

method, representing the control. Results indicate that the

conventional tillage caused the movement of the largest

portion of seeds below the 10 cm depth interval following




Figure 4.

Mean distribution of synthetic seeds following
conventional tillage with comparisons between
wash and excavation methods. *=excavationD=
wash HSD value determined by Tukey's at Alpha=
.05 level.

20 -

1 -

4-6 8-10 12-14 16-18 20-22

+ + + +
+ +
+ +




one tillage operation. Approximately 78% and 83% of the

seeds were recovered below this depth during the wash and

excavation methods respectively. This trend was shown to be

consistent in all cases after the first conventional tillage


Minimum tillage

Data from the minimum tillage plots shows that

distribution was less uniform throughout the working depth

when compared to the conventional tillage treatment (Figure

5). In the reduced tillage system, approximately 88 and 87%

of the seeds recovered were in the top 6 cm for both the

wash and excavation methods, respectively. The maximum

recovery depth throughout the study was typically at 12-14

cm. There was a greater chance for error between the two

sampling methods due to the increase in number of seeds in

the upper layers.

In the reduced tillage systems accurate recovery was

more difficult to achieve with the excavation method;

however only one significant difference exists between the

two methods at the 2-4 cm depth interval (Figure 5).

Distribution patterns were generally the same for both

methods with a decrease in seeds as depth increased and no

seeds recovered below the 12-14 cm interval. This same

pattern was found in all cases including the sequential

treatment in year two.


The no-tillage system like the minimum tillage

treatment showed a simple decrease in seeds as depth


8-10 10-12 12-14



Figure 5. Mean distribution of seeds following minimum
tillage with comparisons between wash and
excavation methods. *=excavation -=wash
HSD values determined by Tukey's at Alpha=.05

0 -



2-4 4-6 6-8

+ + 4





I .


increased (Figure 6). Mean seed recoveries in the top 6 cm

for the sampling and excavation methods were only 85 and

78%, respectively. Although lower than minimum tillage,

this is basically the same trend in terms of distribution.

Comparison between the two methods showed several

significant differences; however, much of this is believed

to be human error due to the excavation technique. This

technique was difficult to perform accurately at 2 cm

intervals because the surface of the excavation area was

uneven and irregular.

Comparisons of Sequential Tillage Operations

Distributions following tillage treatments

This portion of the research was designed to allow a

detailed examination of the seed distribution for each year

as influenced by tillage. In the first year of the study,

seeds were dispersed and plots were subjected to their

respective tillage operations. Thus sampling after this

period was year one of the study with one tillage operation.

In the second year of the study, additional seeds were

dispersed and plots were again subjected to their respective

tillage operations. However, sampling after this date gave

data on year one seeds after two tillage operations, and

year two seeds after one tillage operation. These sample

times allowed comparisons of the effects of one tillage

operation on distribution of seeds and, also, how this

pattern changes with the introduction of a second tillage

operation. The graphs (Figures 7, 8, and 9) show the means

2-4 4-6


2-4 4-6 6-8

6-8 8-10 10-12 12-14



10-12 12-14


Figure 6.

Mean distribution of seeds following no-till
tillage with comparisons between wash and
excavation methods. *=excavationO=wash
HSD values determined by Tukey's at Alpha=
.05 level.

0 -

of the wash and excavation methods by sample time, depth,

and tillage. The means of the two recovery methods were

utilized since there were insignificant differences in early


Conventional tillage

Examination of the effects of conventional tillage

shows a similar pattern to those discussed (Figure 7). The

key differences are how the pattern changes with the

addition of the second tillage and how the single tillage in

the second year resulted in a slight shift. As with the

data discussed earlier, the conventional tillage resulted in

a bulk of the seeds being distributed below 10 cm in depth.

In both the first and second year when seeds received a

single tillage operation, 84 and 89% of the seeds recovered

were below 10 cm. The bulk of this recovery was in the zone

12-18 cm deep. Statistical comparisons between the one

tillage and two tillage operations revealed several

significant differences in terms of seed distributions; when

a second tillage was added seeds were moved in two

directions, deeper and towards the surface. This resulted

in significant differences in the numbers of synthetic seeds

recovered in the 0-2, 2-4, and 4-6 cm layers with an

increase in seeds recovered following the second tillage

operation. Significant differences also were shown at 12-14

and 14-16 and 18-20 cm and a significant increase occurred

in the 18-20 cm layer. When the movement of seeds in the

plots receiving two tillages was compared to the second year

single tillage plots, essentially the same results were

noted with the exception being the 18-20 cm depth layer in

which no significant difference occurred. The first and

second year tillage plots offered interesting results.

Between the two years, even though both involved dispersing

the seeds followed by a single tillage operation there was a

shift in synthetic seed recovery. It is believed that since

this area was carefully selected to be free of sicklepod,

and since its history involves the last seven years in

pasture, that some of this shift could be explained in terms

of the condition of the soil before tillage. In year one

the soil was tightly packed, hard and less friable. In year

two of the study it had been worked once the year before.

This may explain the significance occurring in depths of 8-

10, 12-14, and 14-16 cm in which an increase of recovered

seeds occurred in the second year of the study at the 14-16

cm depth while the number decreased significantly in the

8-10 and 12-14 cm depth. The complete downward shift is

readily noticed on the graph shown (Figure 7).

Minimum tillage

Data collected under the minimum tillage system offered

few differences from those discussed earlier. The patterns

were essentially the same for all sample times; however, a

significant change occurred in the 2-4 and 4-6 cm depths

with the addition of the second tillage operation (Figure

8). The change showed a movement of seeds from the 2-4 cm

layer into the 4-6 cm layer, demonstrating that each






w 14

o 12

E 10

W 8




0-2 4-6 8-10 12-14 16-18 20-22



w +
D 6+ + +
a 4-
a + +
2 + +

0-2 4-6 8-10 12-14 16-18 20-22


Figure 7. Distribution of synthetic seeds as influenced by
one or two conventional tillage operations. 0=
batch 1 with 1 tillage, *=batch 1 with 2 tillage
operations, and A=batch 2 with 1 tillage. HSD
values determined by Tukey's at Alpha=.05 level.

additional tillage possibly would more uniformly distribute

seeds in the working depth of 6-8 cm. As with the

conventional tillage system, there were significant

differences in seeds recovered between the two single

tillage operations. More seeds remained in the 0-2 cm layer

in the second year tillage plots. Again, this could be

related to the condition of the soil in terms of friability

between the two years.


The no-tillage system followed the established pattern

showing a decrease in seeds as depth increased (Figure 9).

The pattern shifts were much the same as those observed in

the minimum tillage plots. A significant decrease again was

noted in the 0-2 cm layer after the second tillage

demonstrating that even though less tillage occurred the

trend towards uniform distribution in the worked depth of

0-8 cm possibly would occur with a continuous no-tillage

system. However, this difference was only statistically

significant when compared to the second run of the single

tillage in year two. Again as in the minimum tillage

treatment, an increase was observed in the 0-2 cm layer

which tends to support the conclusion of soil friability

following the first years tillage. It is possible that due

to less clod formation and less compaction, the soil tends

to have fewer cracks and crevices into which the seeds may

fall during the second years tillage.




Figure 8. Distribution of synthetic seeds as influenced by
one or two minimum tillage operations. O=batch 1
with 1 tillage operation, *=batch 1 with 2 till-
age operations, and A=batch 2 with 1 tillage
operation. HSD values determined by Tukey's at
Alpha=.05 level.





2-4 4-6 6-8 8-10

10-12 12-14










8-10 10-12 12-14





Figure 9.

Distribution of synthetic seeds-as influenced by
one or two no-tillage operations. O=batch 1
with 1 tillage operation, *=batch 1 with 2
tillage operations, and A=batch 2 with 1 tillage
operation. HSD values determined by Tukey's at
Alpha=.05 level.





2-4 4-6 6-8

+ +






An increase in interest for analyzing the distribution

of weed seeds in the soil produced an increase in the

development of new sampling and separation methods

(16,24,27,38). Most of the research conducted in these

areas was initiated to develop more effective methods to

separate seeds from the soil samples collected in order to

better understand the seedbank distribution. Originally,

distribution was determined by utilizing the emergence test

in which the soil samples were placed in a controlled

environment to allow viable seeds to emerge (2,3,4,9,10).

This method is not only very laborious, but also requires

long periods of time to assure accurate counts. It also

makes no estimation of nonviable seeds in the soil seedbank.

Putensen (34) realized that the nonviable seeds might

possibly offer some information so he introduced wet-sieving

to collect these following the completion of the emergence


Following the development of new mesh materials,

separation methods in which washing the soil through various

sieve devices was typically utilized (27,53,60,64). The

main problem found with the sieve methods was that there was

still a time requirement for the separation of the sieved

material especially in studies of soil seedbanks containing

a large number of species with various seed sizes. Fay and

Olson (16) showed that the time required for the separation

of the material collected by sieve techniques was greatly

decreased when a single species was used. Mechanical

devices were constructed by several researchers to lower the

time requirement for the separation procedure (16,53,64).

Some researchers studied different flotation methods but

there was some question about damaging the viability of the

seeds collected (27,57).

The main purpose of the study reported here was to

develop and evaluate a new technique for the determination

of seed distribution under different tillage systems. The

method developed involves the use of synthetic seeds made

from polyethelene with parameters of size, shape, and

density made as close as possible to the seed being

mimicked. Due to the utilization of the synthetic seed, new

methods of separation from or detection in the soil samples

could be examined.

The introduction of the x-ray technique was

incorporated into this study to examine its accuracy.

Comparison of this new method against the older more widely

used method of washing would offer data which hopefully

would demonstrate its effectiveness. The sampling device

described in Chapter II was shown to be an effective device

for soil sampling and, it is hoped these samples obtained

would lend themselves to rapid analysis via x-ray. A

comparison was made between the two methods, and the

efficiency of the new method when compared to the older

washing technique was examined.

Materials and Methods


Methodology of the field experiment was discussed in

Chapter II. The experimental design of the study was a

split-plot experiment with a randomized complete block

design. Main plots were tillage systems, and the subplots

were treatments utilizing dispersal. The main focus of the

study was on the subplots receiving synthetic sicklepod

seeds; therefore, referal to the methods involving

synthetic seed production, dispersal, recovery, and analysis

may be helpful. For simplicity, "Conventional tillage"

refers to moldboard plowing to a depth of 24 cm, followed by

two killings to a depth of 7.5 cm with a tandem disk;

"Minimum tillage" refers to two tillings to a depth of 7.5

cm with a tandem disk; and "No-tillage" refers to direct

seeding with no soil disturbance.

Soil samples were collected using the device described

in Chapter II, with 10 samples being extracted per subplot

from an area 1 X 3 m centrally located within the subplot.

Analysis of Soil Samples

Wash method

Synthetic seeds were recovered by slicing each sample

into 12 sections each 2 cm in height. Each section was

washed individually to remove soil, leaving only small

debris and the synthetic seeds. The depth of the seeds in

the profile was noted and recorded (Refer to Chapter II).

X-ray method

One half of the samples collected from each subplot was

taken to Shands Teaching Hospital, University of Florida,

for analysis by x-ray. Due to the load materials utilized

in the production of synthetic seeds (Chapter II), both 1982

and 1983 synthetic seeds could be detected and

differentiated by this analytical technique.

Initial studies in developing the technique utilized

soil samples spiked with synthetic seed. These were

examined under a fluoroscope which allowed manipulation of

settings while viewing the samples on a monitor, thus

verifying that the seeds could be located. Next, the spiked

soil samples were subjected to various exposure levels and

durations of x-ray, from which a working technique was

developed (Table 1).

Analysis of radiographs was accomplished by using a

modified view screen that was divided into 2 cm sections

with 1.5 mm graphing tape. The radiographs of each sample

were then viewed. The synthetic seeds were counted and

recorded in each 2 cm band (Figure 10). Comparison of the

data collected from both recovery methods will be discussed

in terms of seed distribution, effectiveness of seed

recovery, and time requirements for each analysis method.

Figure 10. Radiograph shown on view screen illustrating the
location of synthetic seeds which allowed the
determination of actual numbers at each depth.
Circled areas show examples of synthetic seeds.


Samples were shot using: a

1). Philips Super 100 Control and Generator

2). Table Top setting at 100 cm from tube

3). Small Focus

4). 55 KVP

5). 130 mAs

6). Film type: Lanex medium OH Kodak

7). Film size: 16 X 14"

a Note: settings may vary between equipment
KVP=kilovolt potential mAs=milliamps/sec.

Results and Conclusions

Comparison of Wash and X-ray Techniques

Conventional tillage

General trends for each of the tillage operations will

be examined using plots constructed from the means of each

recovery method under each tillage system. Examination of

the conventional tillage system shows very similar patterns

of distribution (Figure 11).

The results show that the wash and x-ray methods are

very similar in terms of seed recovery per depth interval.

One significant increase in seeds recovered by the x-ray

method at the 8-10 cm depth interval was found (Figure 11).

Other than this difference, the two recovery curves were

similar. As in the first study, seed distribution was very

similar in terms of bulk recovery. The largest portion of

the seeds recovered was found below the 10 cm depth

interval with approximately 81 and 74% of the seeds

recovered in the wash and x-ray methods, respectively.

These curves were similar to those reported earlier and also

to those reported by other researchers following a single

tillage operation.

Minimum tillage and no-tillage

The comparison of both the minimum tillage and

no-tillage systems was virtually the same. There were few





w 7
a 6
W 5

0-2 4-6 8-10 12-14 16-18 20-22

x-ray. HSD values determined by Tukey's at
7 .
V( 5

4 + + +

23 +


0-2 4-6 8-10 12-14 16-18 20-22

Figure 11. Mean distribution of synthetic seeds recovered
following conventional tillage with comparisons
between wash and x-ray techniques. O=wash e=
x-ray. HSD values determined by Tukey's at
Alpha= .05 level.

statistically significant differences between the two

methods. Both tillage systems resulted in a similar

patterns of a steady decline in seed recovery down through

the working depth. In both tillage systems, the bulk of the

seeds recovered were in the top 6 cm of the soil surface

which would be expected with a 7.5 cm average working depth

(Figures 12 and 13). In the minimum tillage treatments,

approximately 88% of the seeds recovered were found in the

top 6 cm with both recovery methods; while the no-tillage

treatments resulted in approximately 90 and 80% of the seeds

recovered in the top 6 cm by the wash and x-ray methods,

respectively. The only significant differences occurred in

the no-tillage system with an increase of seeds recovered in

the wash method over that of the x-ray method at 0-2 cm and

a decrease at 8-10 cm. However, when the curves are

examined, it is shown that some of these differences are due

to the actual handling of the sample necessary for the wash

method. This is concluded from the x-ray data. It appears

that seed shift in the physical slicing of the soil into 2

cm layers offers some displacement into adjacent layers,

whereas the x-ray method reduces this error.

Comparisons of Sequential Tillage Operations

Conventional tillage

This portion of the study compares distribution

patterns with additional tillage operations. As in Chapter

II, the plots represented are the various sample times which

actually represent the number of tillage operations and year









10-12 12-14

8-10 10-12



Figure 12. Mean distribution of synthetic seeds recovered
following minimum tillage with comparisons
between wash and x-ray techniques. O=wash O
=x-ray HSD values determined by Tukey's at
Alpha=.05 level.

2-4 4-6 6-8 8-10

+ + 4

.4 p

10-12 12-14




Figure 13. Mean distribution of synthetic seeds recovered
following no-tillage with comparisons between
wash and x-ray techniques. O=wash e=x-ray HSD
values determined by Tukey's at Alpha=.05 level.






2-4 4-6 6-8 8-10


+ _+





of study. The graphs were constructed from the means of the

wash and x-ray data because very little differences between

the two methods were found in the initial study. The first

tillage method to be discussed is the conventional tillage

treatment (Figure 14). As with the initial study, there

were interesting changes in seed distribution. Shifts

between the two single tillage operations and the double

tillage operation altered the pattern in comparison to both

single treatments. The two curves representing single

tillage operations in both year one and year two of the

study were very similar in terms of pattern; however the

shift in the second year is quite noticeable. Although

there was no statistically significant difference between

the two single operation curves, this shift raises questions

in terms of soil compaction changes between year one and

year two. The bulk of the seeds recovered following the one

conventional tillage operation were found below the 10 cm

depth interval, with the main concentration for the first

year being from 10-20 cm. While in the second year,

possibly due to previous tillage the concentrated area

shifted down to the 14-23 cm range. For both of the single

conventional tillage operations, approximately 87 and 88% of

the seeds were recovered below the 10 cm depth.

Results obtained by comparing the seeds subjected to

the two conventional tillage operations show that repeated

conventional tillage results in a uniform distribution

throughout the working depth. Following this second tillage




It 16
W 14

o 12
W 10









12-14 16-18 20-22

+ +
+ +

4-6 8-10 12-14 16-18 20-22

Figure 14.

Distribution of synthetic seed as influenced by
one or two conventional tillage treatments.
Percent represent the means of the x-ray and
wash methods. O=batch 1 with 1 tillage
operation, *=batch 1 with 2 tillage operations,
and A=batch 2 with 1 tillage operation. HSD
values determined by Tukey's at Alpha=.05 level.


0 -


operation, the distribution is such that this effect can be

easily seen. The high levels found below 10 cm from the two

single operations are redistributed both deeper and

shallower following the second tillage operation. This

redistribution generates a new pattern that becomes more

linear. The second tillage operation generates several

statistically significant increases in the upper layers

while the areas that had high numbers of seeds with one pass

decrease significantly. The 0-2, 2-4, and 4-6 cm layers all

increased significantly from both single conventional

passes. The two highest recovery zones from each single

pass were also significantly decreased. These zones were

the 16-18 cm level in year one and the 18-20 cm level in

year two. Other areas demonstrated decreases, yet they were

not found to be statistically significant.

Minimum tillage

The minimum tillage operation was very similar in

regard to all sample times and number of tillage operations

(Figure 15). There were no statistically significant

differences between any of the times of recovery.

Approximately 88% of the recovered seeds were found in the

upper 6 cm for all three sample times. The only visual

difference or trend was during the first year of the study

when one tillage pass was made. In the first year plot 74%

of the seeds were found in the top 4 cm while during the

second year, this recovery decreased to approximately 65%









8-10 10-12 12-14

8-10 10-12


Figure 15. Distribution of synthetic seed as influenced by
one or two minimum tillage operations. Percent
represent means of the x-ray and wash methods.
O=batch 1 with 1 tillage operation, *=batch 1
with 2 tillage operations, and A=batch 2 with 1
tillage operation. HSD values determined by
Tukey's at Alpha=.05 level.

2-4 4-6 6-8

4. .9.




with the additional tillage. Again, this is believed to be

due to the original compaction of the soil.


The no-tillage system resulted in several statistically

significant differences between sample times and number and

time of tillage (Figure 16). It is important to note that

although this system involves no tillage at the time of

soybean planting, there still were two passes prior to the

winter wheat planting which contributed to the similarity to

the minimum tillage regime. The first year curve actually

represents distribution following two passes with a disk at

7.5 cm; the second year curve represents distribution

following four passes with a disc at 7.5 cm. Although this

area was treated as a no-tillage soybean program, the seeds

were subjected to some tillage following dispersal in land

preparation for winter wheat.

The results indicate (Figure 16) that the major

differences were those between the first year single tillage

and the second year single and double tillage operations.

The first year data showed that recovery was largely

obtained in the 0-2 cm layer which was approximately 53%.

Both the second year operations resulted in statistically

significant decreases to 31 and 35%, respectively. This

result again is most likely due to the soil conditions prior

to the start of the tillage operation. Both the 4-6 cm and

6-8 cm depths were found to be statistically significantly

with higher numbers of seeds recovered in the samples














10-12 12-14

8-10 10-12 12-14


Figure 16. Distribution of synthetic seed as influenced by
one or two no-tillage operations. Percent
represent means of x-ray and wash methods. 0=
batch 1 with 1 tillage operation, *=batch 1 with
2 tillage operations, and A=batch 2 with 1
tillage operation. HSD values determined by
Tukey's at Alpha=.05 level.

2-4 4-6 6-8 8-10

. w g


removed during the second year when compared to year one.

There was no comparable difference between the two

recoveries in the second year.



In order to better utilize the information obtained

from studies examining weed seed populations in the soil,

research has been directed towards prediction models.

Radosevich and Holt (35) examined the potential of weed

infestations by studying the soil seedbank. Some researchers

have developed working demographic models to aid in the

study of weed population dynamics (30,55).

Research has also been conducted on various forms of

cultivation and tillage regimes in respect to seedling

emergence. Roberts and Potter (49) studied the relationship

between cultivation and rainfall in terms of seedling

emergence. The importance of becoming more familiar with

weed seed emergence in regard to cultivation will aid in the

ability to make predictions of future infestations.

Roberts and Ricketts (50) investigated the relationship

between cultivation and existing weed seeds in the soil.

They reported that such studies could possibly offer

information which would allow for predictions to be made on

weed infestations. They demonstrated how predictive values

can be obtained from actual and potential weed flora data

with additional work in regard to cultural practices.

This research was directed towards weed emergence and

distribution patterns. Attempts were made to predict weed

emergence with the aid of a model. Data collected on the

population in the soil seedbank was utilized through an

examination of distribution patterns under different tillage

systems, and actual emergence numbers collected from the

same study areas. The comparison between the actual and

predicted emergence was analyzed to answer questions

relating seed populations and their distribution in the


Methods and Materials


The establishment of the field experiment in terms of

the methodology applied was discussed in Chapter II. The

experimental design of the study was a split-plot experiment

with a randomized complete block design. Main plots were

tillage systems, and the subplots were treatments of seed

dispersal. The focus of this comparison will be on the

correlation between the distribution of the synthetic seeds

due to the tillage systems implemented and the effects of

this distribution on seedling emergence. A review of

methods involving viable seed collection, synthetic seed

production, seed dispersal, seed recovery, and analysis may

be helpful (Chapter II). For simplicity, "Conventional

tillage" refers to moldboard plowing to a depth of 24 cm,

followed by two tillings to a depth of 7.5 cm with a tandem

disk; "Minimum tillage" refers to two tillings to a depth

of 7.5 cm with a tandem disk; and "No-tillage" refers to

direct seeding with no soil disturbance.

Soil samples were collected using the device described

in Chapter II, with 10 samples being extracted per subplot

from an area 1 X 3 m centrally located within the subplot.

Both analysis methods discussed in Chapter II and Chapter

III were utilized in terms of synthetic seed recovery.

Seedling Emergence Counts

The population of sicklepod seedlings emerging

following the application of tillage treatments was obtained

by periodic counting of seedlings in a 1 X 3 m permanent

area within each subplot. All subplots receiving the viable

sicklepod seed dispersal treatments, which included either

1982 or 1983 viable seeds, were treated in this manner.

Counts were made until emergence had ceased in the subplots.

After counts were made, the seedlings were removed by hand

pulling; however, since they were typically in the

cotyledon to three leaf stage, no major soil disturbance

occurred. Emergence numbers were recorded for analysis and

compared to distribution in the soil profile.

Calculating Predicted Emergences

The calculations utilized to generate the predicted

emergence values were constructed from data collected in a

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15


Figure 17. Curve fit to points generated from emergence
study corrected to initial germination of 55%
to find correction equation (63).

study by Teem et al. (63). This study involved the

capabilities of sicklepod to emerge from various depths.

The seeds were placed at depths from 2.5 to 15.2 cm, and

emergence was determined as a percentage of seeds planted.

The values were first transposed to an initial emergence of

55% which was a base starting point. This point is believed

to be closer than that of the acid scarified seeds in the

work by Teem et al. (63) to the unscarified seeds used in

this study. A multiple regression on values of emergence at

the various depths was run. This regression fit a curve so

that an equation could be found to allow the calculation of

predicted emergence percent at any depth within the range of

0 to 15.2 cm (Figure 17). From this equation the recovered

synthetic seeds within each depth interval could be

corrected to a value representing the predicted number

emerging from that depth.

Y = 36.89 + 5.92 X DEPTH 0.543 X DEPTH2

When the depths from 0 to 14 cm were handled in this

manner and the total recovery corrected from the equation,

the predicted values were then compared with the actual

values collected on the test areas. The depth of 14 cm was

utilized as the bottom depth since this is below the maximum

depth of reported sicklepod emergence.

Results and Discussion

Actual versus Predicted


In order to compare the actual emergence numbers with

the predicted values, it was important to first examine the

main differences observed. The most obvious differences in

the actual emergence were shown in the conventional tillage

treatment in comparison to either the minimum or no-tillage

system. The reduction in the emergence following the one

tillage operation is approximately four fold lower than

either of the reduced tillage systems (Figure 18). The

conventional tillage resulted in approximately 14% of the

seeds dispersed emerging following a single tillage

application. This is a definite response of the

distribution pattern observed where approximately 85% of the

seeds were found below the 10 cm depth. Both the minimum

and no-tillage systems were found to have approximately 60%

emergence in comparison to the conventional tillage


Seed Batch

Figure 18 shows the grouping patterns of each tillage

with the actual values plotted against the predicted values.

However, the predictions appear to be higher than expected

in some cases, reflecting higher recoveries. Since the

predicted emergence values are calculated from the recovery

values by depth, the relationship between seed batch and

recovery becomes more important than in the previous











0.0 0.2 0.4 0.6 0.8 1.0
(Thousands) 2

1.2 1.4 1.6

Figure 18. Actual vs. predicted emergence by tillage.
O=conventional, *=minimum, and A=no-till.






0-- I I 1 I I I I -

studies. An analysis by seed batch was run and linear

models fitted to both batch 1 and 2 (Figure 19). The R

values were better when the points were separated by seed

batches than when the linear model was fit to all points

(Figure 18).

Recovery Values

These questions brought about the need for closer

examination of the recoveries of seeds by the various

methods utilized. Analysis of the recovery data from both

the x-ray versus wash, and wash versus excavation studies

revealed similar findings. Comparisons of the wash and

excavation methods (Figure 20) demonstrated that the means

were higher than the predicted recovery, however they were

not statistically significantly different among the tillage

systems. Yet, when the seed batches were compared overall

by the means, the recoveries of the second seed batch was

significantly higher than either recovery of batch one from

year one or two (Figure 21).

Similar results were noted with the wash versus x-ray

recovery such that there was no difference in the mean

recoveries by tillage systems (Figure 22), yet again the

second seed batch was recovered in significantly higher

numbers over batch one (Figure 23).

These observations help to explain why the predictions

in some cases are higher than the expected numbers by at

least two fold. It appears also that the load material and

color are also important especially in methods involving



6.6- n

w *
0 6.2- 0 o
S 6.0-

2 5.8-
w 0
j 5.6-


< 5.2-
0 5.0-
( 4.8- *


4.2- *

4.0- -\-i\- i -r- i
4.4 4.8 5.2 5.6 6.0 6.4 6.8 7.2


Figure 19. Log plot of actual vs. predicted emergence of
sicklepod by seed batch. O=batch 1 *=batch 2






o0 1.6-







Figure 20. Seed recovery means for wash and excavation by
tillage. CON=conventional, MIN=minimum, and NOT
=no-tillage. Predicted recovery shown by
horizontal line. HSD=500 at Alpha = .05






BT 12 B

W 1.8-
Oc 1.6

0 1.2-


Figure 21. Seed recovery means for wash and excavation by
synthetic seed batch. B1T1=batch 1 tillage 1,
B1T2=batch 1 tillage 2, and B2T1=batch 2 tillage
1. Predicted recovery shown by horizontal line.
HSD=340 at Alpha = .05




B1T1 B1T2 B2T1


Figure 21. Seed recovery means for wash and excavation by
synthetic seed batch. BITl=batch 1 tillage 1,
B1T2=batch 1 tillage 2, and B2Tl=batch 2 tillage
1. Predicted recovery shown by horizontal line.
HSD=340 at Alpha = .05


Figure 22.

Seed recovery means for wash and x-ray by tillage.
CON=conventional, MIN=minimum, and NOT=no-tillage.
Predicted recovery shown by horizontal line.
HSD = 520 at Alpha = .05




















0.0 1





M 1.8-

oc 1.4

0.2- B
111 B1T2 B2T1-
r 0.8




81Ti B1T2 B2T1


Figure 23. Seed recovery means for wash and x-ray by synthe-
tic seed batch. B1T1=batch 1 tillage 1, B1T2=batch
1 tillage 2, and B2T1=batch 2 tillage 1. Predicted
recovery shown by horizontal line. HSD=322 at
Alpha = .05


washing where visual separation from darker soil particles

and litter are involved. Although some of the predicted

values are high, it is clear that with an understanding of

distribution and population of seeds in the soil as well as

knowledge of the seeds capabilities and limitations,

acceptable prediction models can be developed to estimate

future weed infestation levels.



The techniques developed and utilized in these studies

were found to be efficient in terms of seed recovery and

determination of distribution patterns. The soil sampling

device was found to be effective in removing an intact

vertical slice of soil which was well suited to analysis by

both wash and x-ray. When compared to the excavation

method, or control, the data collected demonstrated that the

soil sampling device collected samples that were not

compacted eliminating shifts in seed location; the device

samples also allowed for excellent distribution analysis.

The synthetic seeds were found to be excellent in terms of

their ability for recapture following dispersal; they

should remain buried for extended periods of time without

degradation. The source of load material and color of the

synthetic seeds was found to be extremely important. The

barium sulfate load was better in terms of the utilization

of x-ray for analysis because of its higher resolution in

comparison to the iron powder. Barium sulphate also

resulted in a brilliant white color that increased the ease

of separation from soil and litter in the wash method.

One of the greatest values of the techniques developed

was that they were found to be much faster than the other

methods. The wash method has been used for many years in

various ways, yet it still involves a large degree of hand

separation which is time consuming and laborious. The use

of x-ray was found to be at least five times as fast and

required no disturbance to the soil sample which means a

lower chance of seed movement during handling. Although the

cost of x-ray equipment is high, it was found to be readily

available for research purposes at a local hospital. The

major cost involved was the purchase of film for the actual

radiographs which was compensated for by the amount of time

saved. The samples are collected, exposed, and developed;

the sample requires no opening, slicing, or hand separation.

A larger number of soil samples could be handled at a much

faster rate with this method than with the wash and

germination method.

Tillage Systems

The results of these studies on three tillage systems

indicated that the conventional tillage system was the

system affecting distribution to the greatest degree. The

seeds were found to move deep into the soil profile with

approximately 80-85% found below the 10 cm depth. The

minimum and no-tillage systems resulted in little change

with all seeds being distributed in the top 7.5 cm.

Approximately 75-80% were found in the top 6 cm.

Another important factor was the number of tillage

operations, especially with the conventional tillage

treatment. The seeds were moved to different locations with

each tillage application, reflecting shifts in

concentrations. Following one tillage operation the bulk of

the seeds were found below 10 cm with high concentration

bands around 13-17 cm. Yet following the second tillage

operation, the distribution became more uniform throughout

the working depth. Even the seeds distributed in the second

year which received a single tillage operation resulted in a

somewhat different pattern. Although the second year seeds

were very similar in terms of actual patterns when compared

to the first year single tillage operation there was a key

difference. The concentrated bands shifted to deeper depths

in the profile. The most plausable explanation for this

was the difference in friability of the soil between the two

dispersals and tillage passes. Prior to the initiation of

the study, the field had been in grassland for approximately

seven years, thus the soil was tightly packed. In the

second year of the study it had been worked the previous

year and was in a more friable state.

Although few differences were noted between the two

reduced tillage systems, there was some statistically

significant difference in the upper 6 cm with additional

tillage applications.

Predicting Infestations

Understanding the distribution patterns and population

shifts among the seeds in the soil as influenced by tillage

are all important concerns when attempting to predict

emergence levels. However, of equal importance is a

complete understanding of the seed itself and its

capabilities and limitations in regard to both germination

and emergence. If these factors are known and if a good

method for analyzing the distribution and population within

the soil profile is available, then accurate predictions can

be made.

A better understanding of this area of weed science is

extremely important for achieving weed management. With a

complete knowledge of an adversary the battle can become

much easier.

Future Studies

With the introduction of the new techniques developed

in this work, there are many aspects of the soil seedbank

that could possibly be examined. The use of x-ray and

synthetic seeds offer two additional tools for a closer

examination of various soil seedbanks under many varying

conditions. The soil sampling device developed for these

studies was also found to be highly functional in relation

to both the wash method and the new x-ray techniques. The

advantages of a vertical slice as apposed to a core sample

are clearly seen when working with both of these methods.

The sampling device has immediate use in work involving soil

sampling for collection of data, because the area recovered

is greater and it lends itself to analysis better than a

core sampler. The study of seeds in the soil seedbank

becomes less of a problem when data can be collected in a

manner that requires less time and labor. The synthetic

seeds offer a new approach to questions of soil seedbank

structure. By loading synthetic seed with materials having

different atomic numbers various species could be examined

under different cropping practices and situations with the

ability to separate their individual responses. It is the

hope of the author, that this information and technology be

utilized in future studies. With these new tools, it is

hoped some new questions can be asked, and their answers found.


1. Anderson, W. P. 1977. Weed science: principals. West
Publishing Company, New York. 598 pp.

2. Brenchley, W. E. and K. Warington. 1930. The weed seed
population of arable soil. I. Numerical estimation
of viable seeds and observations on their natural
dormancy. J. Ecol. 18:235-272.

3. Brenchley, W. E. and K. Warington. 1933. The weed seed
population of arable soil. II. Influence of crop,
soil and methods of cultivation upon the relative
abundance of viable seeds. J. Ecol. 21:103-127.

4. Brenchley, W. E. and K. Warington. 1936. The weed seed
population of arable soil. III. The re-establish-
ment of weed species after reduction by fallowing.
J. Ecol. 24:479-501.

5. Brenchley, W. E. and K. Warington. 1945. The influence
of periodic fallowing on the prevalence of viable
weed seeds in arable soil. Ann. Appl. Biol.

6. Champness, S. S. 1949. Note on the technique of sampl-
ing soil to determine the content of buried viable
seeds. J. Brit. Grassl. Soc. 4:115-118.

7. Champness, S. S. and K. Morris. 1948. The population
of buried viable seeds in relation to contrasting
pasture and soil types. J. Ecol. 36:149-173.

8. Chancellor, R. J. 1964. Emergence of weed seedlings in
the field and the effects of different frequencies
of cultivation. Proc. 7th Br. Weed Control Conf.,
pp. 599-606.

9. Chepil, W. S. 1946. Germination of weed seeds: I. Lon-
gevity, periodicity of germination, and vitality of
seeds in cultivated soil. Sci. Agric. 26:307-346.

10. Chepil, W. S. 1946. Germination of weed seeds: II. The
influence of tilling treatments on germination.
Sci. Agric. 26:347-357.

11. Chippindale, H. G. and W. E. J. Milton. 1934. On the
viable seeds present in the soil beneath pastures.
J. Ecol. 22:508-531.

12. Creel, J. M. jr., C. S. Hoveland, and G. A. Buchanan.
1968. Germination, growth, and ecology of sickle-
pod. Weeds 16:396-400.

13. Cussans, G. W., S. R. Moss, F. Pollard, and B. J. Will-
son. 1979. Studies on the effects of tillage on
annual weed populations. The influence of differ-
ent factors on the development and control of
weeds, pp. 115-122. Proc. European Weed Res. Soc.
Sympos. Mainz.

14. Darlington, H. T. 1951. The seventy-year period of Dr.
Beal's seed viability experiment. Amer. J. Bot.

15. Dawson, J. H. and V. F. Burns. 1975. Longevity of
barnyardgrass, green foxtail, and yellow foxtail
seeds in the soil. Weed Sci. 23:437-440.

16. Fay, P. K. and W. A. Olson. 1978. Technique for sepa-
rating weed seed from soil. Weed Sci. 26:530-533.

17. Froud-Williams, R. J., R. J. Chancellor, and D. S. H.
Drennan. 1983. Influence of cultivation regime
upon buried weed seeds in arable cropping systems.
J. Appl. Ecol. 20:199-208.

18. Froud-Williams, R. J., D. S. H. Drennan, and R. J.
Chancellor. 1983. Influence of cultivation
regime on weed floras of arable cropping systems.
J. Appl. Ecol. 20:187-197.

19. Harper, J. L. 1957. The ecological significance of
dormancy and its importance in weed control.
Proc. 4th Int. Congr. Crop Protection, Hamburg
1957, 1:415-420. Braunschweig, Hamburg.

20. Hurle, K. 1974. Effect of long-term weed control mea-
sures on viable weed seeds in the soil. Proc. 12th
Br. Weed Con. Conf. pp.1145-1152.

21. Jordan, J. H. jr. 1983. Sicklepod (Cassia obtusifolia
L.) competition with soybeans as influenced by row
spacing, density, planting date, and herbicides.
Ph.D. dissertation., Univ. of Fl., Gainesville,Fl,
106 pp.

22. Kellman, M. 1978. Microdistribution of viable weed
seed in two tropical soils. J. Biogeography, 5:

23. Kivilaan, A. and R. S. Bandurski. 1981. The one hun-
dred-year period for Dr. Beal's seed viability ex-
periment. Amer. J. Bot. 68:1290-1292.

24. Kropac, Z. 1966. Estimation of weed seeds in arable
soil. Pedobiologia, 6:105-128.

25. Lewis, J. 1961. The influence of water level, soil
depth and type on the survival of crop and weed
seeds. Proc. Int. Seed Test. Assoc. 21:68-85.

26. Lueschen, W. E. and R. N. Andersen. 1980. Longevity of
velvetleaf (Abutilon theophrasti) seeds in soil un-
der agricultural practices. Weed Sci. 28:341-346.

27. Major, J. and W. T. Pyott. Buried, viable seeds in two
California bunchgrass sites and their bearing on the
definition of a flora. Vegetatio Acta Geobotanica

28. Malone, C. R. 1967. A rapid method for enumeration of
viable seeds in soil. Weeds, 15:381-382.

29. Moore, J. M. and R. W. Wein. 1977. Viable seed popula-
tions by depth and potential site recolonization
after disturbance. Can. J. Bot. 55:2408-2414.

30. Mortimer, A. M. 1976. Aspects of the seed population
dynamics of Dactylis glomerata L., Holcus lanatus
L., Plantago lanceolata L., and Poa annua L..
Proc. 1976 Br. Crop Protection Conf. Weeds, pp.

31. Pollard, F. and G. W. Cussans. 1976. The influence of
tillage on the weed flora of four sites sown to suc-
cessive crops of spring barley. Proc. 1976 Br. Crop
Protection Conf. Weeds, pp.1019-1028.

32. Pollard, F. and G. W. Cussans. 1981. The influence of
tillage on the flora in a succession of winter cere-
al crops on a sandy loam soil. Weed Res. 21:

33. Pollard, F., S. R. Moss, G. W. Cussans, and R. J. Froud-
Williams. 1982. The influence of tillage on the
weed flora in a succession of winter wheat crops on
a clay loam soil and silt loam soil. Weed Res. 22:

34. Putensen, H. 1882. Untersuchungen uber die im acker-
boden enthaltenen samereien. Hanoversches landund
forstwirtch. vereinsblatt. Hildesheim. 21:512-

35. Radosevich, S.R. and J.S. Holt. 1984. Weed ecology:
implication for vegetation management. pp. 86-91.
John Wiley and Sons Inc. New York, New York.

36. Roberts, E. H. 1972. Dormancy: a factor affecting seed
survival in the soil. pp. 321-359 in E. H. Roberts,
ed. Viability of seeds. Syracuse University Press,
Syracuse, N. Y.

37. Roberts, H. A. 1958. Studies on the weeds of vegetable
crops. I. initial effects of cropping on the weed
seeds in the soil. J. Ecol. 46:759-768.

38. Roberts, H. A. 1962. Studies on the weeds of vegetable
crops. II. Effect of six years of cropping on the
weed seeds in the soil. J. Ecol. 50:803-813.

39. Roberts, H. A. 1963. Studies on the weeds of vegetable
crops. III. Effect of different primary cultiva-
tions on the weed seeds in the soil. J. Ecol.

40. Roberts, H. A. 1963. Studies on the weeds of vegetable
crops. IV. Further observations on the effects of
different primary cultivations. J. Ecol. 51:323-

41. Roberts, H. A. 1964. Emergence and longevity in culti-
vated soil of seeds of some annual weeds. Weed
Res. 4:296-307.

42. Roberts, H. A. 1968. The changing population of viable
weed seeds in an arable soil. Weed Res. 8:253-256.

43. Roberts, H. A. 1970. Viable weed seeds in cultivated
soils. Rep. Nat. Veg. Res. Stn. for 1969.
Wellesbourne, U. K., pp. 25-38.

44. Roberts, H. A. 1981. Seed banks in soils. Advances in
Applied Biology, 6:1-55.

45. Roberts, H. A. and P. A. Dawkins. 1967. Effect of cul-
tivation on the numbers of viable weed seeds in the
soil. Weed Res. 7:290-301.

46. Roberts, H. A. and P. M. Feast. 1972. Fate of seeds of
some annual weeds in different depths of cultiva-
ted and undisturbed soil. Weed Res. 12:316-324.

47. Roberts, H. A. and P. M. Feast. 1973. Emergence and
longevity of seeds of annual weeds in cultivated
and undisturbed soils. J. Appl. Ecol. 10:133-143.

48. Roberts, H. A. and J. E. Neilson. 1981. Changes in the
soil seed bank of four long-term crop/herbicide
experiments. J. Appl. Ecol. 18:661-668.

49. Roberts, H. A. and M. E. Potter. 1980. Emergence
patterns of weed seedlings in relation to cultiva-
tion and rainfall. Weed Res. 20:377-386.