Studies on spiny amaranth (Amaranthus spinosus L.) interference with lettuce (Lactuca sativa L.) as influenced by phosph...

MISSING IMAGE

Material Information

Title:
Studies on spiny amaranth (Amaranthus spinosus L.) interference with lettuce (Lactuca sativa L.) as influenced by phosphorus fertility on histosols
Physical Description:
vi, 213 leaves : ill. ; 29 cm.
Language:
English
Creator:
Shrefler, James William, 1953-
Publication Date:

Subjects

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

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 205-212).
Statement of Responsibility:
by James William Shrefler.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001897882
oclc - 29809576
notis - AJX3179
System ID:
AA00003678:00001

Full Text










STUDIES ON SPINY AMARANTH AMARANTHUSS SPINOSUS L.)
INTERFERENCE WITH LETTUCE (LACTUCA SATIVA L.) AS
INFLUENCED BY PHOSPHORUS FERTILITY ON HISTOSOLS


















By

JAMES WILLIAM SHREFLER


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1993


































To My Parents












ACKNOWLEDGEMENTS


I wish to express my appreciation to my committee

members, Dr. Barry Brecke, Dr. Joan Dusky, Dr. Donn Shilling,

Dr. Bill Stall, and Dr. Daniel Colvin for their guidance and

support during the course of my studies at the University of

Florida. I also am greatly indebted to the faculty at the

Everglades Research and Education Center at Belle Glade for

valuable assistance and encouragement they provided me. The

companionship and assistance of the weed science graduate

students and staff and the Everglades Research and Education

Center staff is appreciated and will long be remembered. My

deepest appreciation goes to my wife, Carmen, for her support

and understanding.


iii













TABLE OF CONTENTS


ACKNOWLEDGEMENTS . . iii

ABSTRACT . . v

CHAPTERS

1 INTRODUCTION . . 1

2 LITERATURE REVIEW . . 4

Interference and Competition .... 4
Factors Influencing Competition ... 10
Factors for which Plants Compete . 12

3 SPINY AMARANTH INTERFERENCE IN CRISPHEAD LETTUCE AS
INFLUENCED BY PHOSPHORUS FERTILIZER APPLICATION
METHOD . . .. 17
Introduction. ... . .. 17
Materials and Methods . .. 24
Results and Discussion. ... .. 30

4 SPINY AMARANTH (Amaranthus spinosus L.) COMPETITION
WITH LETTUCE (Lactuca sativa L.) AS INFLUENCED BY
PHOSPHORUS FERTILITY OF MUCK SOIL .. 74
Introduction . . 74
Materials and Methods . .. 86
Results and Discussion . 95

5 SUMMARY AND CONCLUSIONS . .. 185

APPENDICES

A CHAPTER THREE MICRONUTRIENT DATA .. 191

B CHAPTER FOUR TABLES . .. 198

LITERATURE CITED . .. 205

BIOGRAPHICAL SKETCH. ... . .. 213












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

STUDIES ON SPINY AMARANTH AMARANTHUSS SPINOSUS L.)
INTERFERENCE WITH LETTUCE (LACTUCA SATIVA L.) AS
INFLUENCED BY PHOSPHORUS FERTILITY ON HISTOSOLS

By

James William Shrefler

May, 1993

Cochair: Dr. B.J. Brecke
Cochair: Dr. J.A. Dusky
Major Department: Agronomy Department

Studies were conducted to determine how spiny amaranth

(Amaranthus spinosus L.) interference with lettuce (Lactuca

sativa L.) grown on histosols would be influenced by soil

phosphorus (P) fertility and P application method. Additive

techniques were used in 3 field studies (spring 1991, fall

1991 and spring 1992) to test for the effects P application

method (banded versus broadcast), weed density and duration of

interference on lettuce yields. Phosphorus application by

either method increased lettuce yields ca. 100% over controls

not receiving P. Spiny amaranth biomass was less responsive

to P. Duration of interference reduced lettuce yields in a

quadratic fashion with broadcast P application in spring 1991

and 1992. With mid- to full-season interference, lettuce

yields were reduced more with 4 weeds per 2.3 m than 1.






Lettuce tissue P concentrations explained the yield response

to P fertilization; P being least with no added P,

intermediate with banded P and highest with broadcast P.

Lettuce yield loss due to weed interference was not explained

by nutritional status.

Replacement techniques were used in 3 greenhouse

experiments to study the effect of P fertility on competition

between lettuce and spiny amaranth during the first 4 weeks of

growth. Monocultures and 1:1 mixtures were used at total

densities ranging from 2 to 32 plants per 113 cm2.

Competition was assessed based on analyses of individual plant

weight, relative yields, relative crowding coefficients (RCC)

and relative mixture response. Competitive interactions

varied with experiments. When monoculture growth of the two

species was comparable, lettuce was more competitive than

spiny amaranth in mixture at all but the lowest P fertility

level. When spiny amaranth growth in monoculture was 2 to 4

times greater than that of lettuce, the weed was the dominant

competitor. However, relative yield analysis suggested that

the competitive ability of lettuce was improved by increased

P fertility. Competitive interactions were density dependent.

As measured by the RCC, the intensity of interspecific

competition increased with density; this density effect being

species dependent. However, there were no interactions

between density and P fertility in the analysis of

interspecific competition.












CHAPTER 1
INTRODUCTION


During the period between 1986 and 1991, lettuce

(Lactuca sativa L.) produced in Florida had an annual

average value of more than 38 million dollars (Freie and

Young, 1992). Greater than 90 percent of this production

occurred on Everglades histosols. Most of the remainder was

grown on histosols in the central region of the state.

Weed control is important in the production of quality

lettuce (Ryder, 1979). On Florida muck soils, available

weed control strategies are limited. For gramineous weeds,

postemergence herbicides are available (Stall and Dusky,

1992). Broadleaf weeds, however, are an important weed

problem on these soils (Dusky et al., 1988). Nonselective

weed control between the crop rows can be obtained with a

shielded application of paraquat (l,l'-dimethyl-4,4'-

bipyridinium). However, weeds in the crop row remain

uncontrolled with this measure. Soil applied herbicides are

available for broadleaf weed control in lettuce on mineral

soils; however, these are not recommended for use on muck

soils (Dusky and Stall, 1992). Thiobencarb (S-[(4-

chlorophenyl)methyl]diethylcarbamothioate) is one herbicide

that can be used on these soils for weed control in lettuce






2

(Dusky et al. 1988). Its herbicidal activity at a given use

rate was found to be less on histosols than on mineral soil

(Braverman, Locasio, Dusky and Hornsby, 1990). This was

attributed to greater soil adsorption and decreased mobility

of the herbicide in the organic soil compared to the mineral

soil (Braverman, Dusky, Locasio and Hornsby, 1990).

However, thiobencarb efficacy is dependent upon

environmental conditions (temperature and moisture) and

there is a relatively small safety margin for lettuce.

Due to the lack of options, weed control in lettuce

grown on muck soils is highly dependent on cultivation and

hoeing. Hoeing requires a high labor input. Hoeing also

results in damage to lettuce plants (Dusky and Stall, 1990).

Information concerning the effect weed infestations have on

a crop facilitates the decision making process of when to

apply control measures (Thill et al., 1991). Such

information might help to minimize labor costs and to reduce

crop damage that would result from excessive hoeing. Such

information for lettuce is scarce.

Crop mineral nutrition is important in the production

of quality lettuce (Ryder, 1979). Relatively large amounts

of phosphorus (P) fertilizer are required to obtain a

lettuce crop of good quality and yield on high organic

matter soils (Sanchez and Burdine, 1988). Current

environmental and economic issues, however, compel growers

to reduce fertilizer P use (Alvarez and Sanchez, 1991;








Sanchez et al., 1990). One promising approach to reduced

fertilizer P application for lettuce is the use of a banded

application (Sanchez et al., 1990).

One objective of this investigation was to characterize

the effect of interference by spiny amaranth (Amaranthus

spinosus L.), an important broadleaf weed species, on

lettuce. Spiny amaranth was chosen for these studies

because of its ubiquitous occurrence throughout the lettuce

growing areas of Florida and because of the difficulty of

controlling it in lettuce. A second objective was to

determine how P fertilization practices would influence the

effect of the weed on the crop. In following with these

objectives, two series of studies were conducted. These

studies are presented in chapters III and IV.

Chapter III of this investigation concerns field

studies which were conducted to determine how density and

duration of spiny amaranth interference affects lettuce

yields. These factors were examined at several fertilizer P

application options.

Chapter IV of this investigation presents studies on

competition between lettuce and spiny amaranth during early

growth. Competition is assessed in these studies under

varied levels of P fertilizer amendment of muck soil.

Several methods of analyzing the competitiveness of one

species upon another are discussed.












CHAPTER 2
LITERATURE REVIEW



Interference and Competition


It has been stated that weed control measures "focus

directly or indirectly on improving the competitive ability

of crops with regard to weeds" (Spitters and Van Den Berg,

1982, p. 137). This is an example of how the term

competition is commonly used to refer to the overall effect

that weeds have on crop growth due to interactions between

plants growing in close proximity to one another. Losses in

crop yield or quality from interactions among weeds and

crops form the basis of modern weed science (Radosevich,

1987). Radosevich and Holt (1984) discussed the existence

of ten possible interactions, previously published by

another author, that may occur between plants. These

interactions have to do with the influence a plant might

have on the environment of its neighbors and are

collectively termed interference. Three of the interaction

types, competition, amensualism allelopathyy) and parasitism

are of a negative nature. These are also of particular

importance to crop production. Competition refers to

mutually adverse effects of organisms utilizing a resource








in short supply. Other definitions for the terms

interference and competition are also found in the

literature (e.g. Hall, 1974a) and these terms have even been

used synonymously (Glauninger and Holzner, 1982).

In nonagricultural ecosystems plant competition is a

phenomena influencing both the immediate and the future

community. On a short term, it determines how individual

plants will fare as compared to neighbors of both like and

different species. On a longer term, competition is an

important force influencing the higher plant composition of

the ecosystem (Grime, 1979). In agricultural ecosystems,

plant competition is commonly viewed as a factor influencing

crop productivity (Radosevich and Roush, 1990), which

implies a shorter term focus than the study of competition

in nonagricultural ecosystems would take.

The study of plant competition in agriculture generally

concerns the optimization of crop yield. One application of

this concerns the determination of mixed cropping system

characteristics that enable maximum crop productivity

(Spitters, 1983). All components of the intended mixtures

are of economic value to the grower. The study of

competition, in this case, aims to achieve a optimal balance

between the desirable components. Another application of

the study of plant competition in agriculture concerns the

case in which crop plants grow in mixture with plants that

are not part of the intended crop. The study of competition








in this case would still concern the optimization of crop

productivity. In contrast to the mixed cropping scenario,

however, optimization is dependent on the productivity of

only the desirable components of the mixture. The study of

competition between weeds and agricultural crops is of this

nature.

Conceptualization of weed competition with crops is

linked to the methodology used in its study. General

treatment of the topic often begins with a discussion of

methodology (Radosevich and Roush, 1990; Spitters and van

den Berg, 1982). There are two basic methods which have

received considerable attention (Cousens, 1991; Radosevich,

1987). The replacement or substitutive series was advanced

by the work of de Wit (1960) although it was reportedly

established prior to this (Cousens, 1991). The additive

study is the second principle method (Cousens, 1991;

Radosevich and Roush, 1990). A third method, the area of

influence, has been considered a special case of the

additive study (Radosevich and Roush, 1990).

Each of these methods is useful for answering a unique

set of questions (Cousens, 1991). The additive design

addresses weed management concerns such as the establishment

of relationships between weed density and crop yield and the

comparison of weed species for their ability to suppress

crop yields. The replacement design is used for determining

which of two components of a binary mixture is the better








competitor. It is also utilized when study objectives

include the identification of the nature of the interaction

between the two components of the mixture (Cousens, 1991;

Jolliffe et al., 1984; Radosevich, 1987). The area of

influence approach is applicable for determining the effect

of an individual weed on crop plants growing at various

distances from the crop plant (Gunsolus and Coble, 1986).

Just as each of these basic methods has a specific

purpose, limitations are also associated with each (Cousens,

1991). It was argued that these methods have been

wrongfully criticized for these limitations. This is

because the criticisms have assumed that the methods should

provide more thorough analyses of competition than is

achieved with them. Thus, much of the criticism would be

better viewed as limitations to be recognized and accepted

(Cousens, 1991). In light of this situation, modifications

or extensions of these basic methods of analysis of plant

competition have been proposed.

Two concerns that have been raised regarding the

analysis of plant competition are that results are density

dependent and that intraspecific and interspecific

competition effects are not separable (Firbank and

Watkinson, 1985; Jolliffe et al., 1984; Spitters, 1983).

Variations on each of the replacement (Jolliffe et al.,

1984) and additive (Rejmanek et al., 1989; Spitters, 1983)








approaches have been developed to address these limitations

to the basic methods.

Conventional analysis of replacement series studies

uses expected yields as a reference from which inferences

about competition are made (Jolliffe et al., 1984; Rejmanek

et al., 1989). Analysis based on expected yields does not

permit the partitioning of interspecific and intraspecific

competition effects in the overall analysis. Jolliffe et

al. (1984) proposed the use of a "projected yield" as a

reference against which competitive effects can be measured

in a replacement series. This technique has been referred

to as a "synthetic no-interaction approach" (Roush et al.,

1989, p. 270). The projected yield is a hypothetical curve

of how the yield of a monoculture would respond to plant

density in the absence of intraspecific competition. Using

this approach, intraspecific competition is quantified by

comparison of actual monoculture yields with projected

yields. The response of the individual species to mixture

is measured by comparison of yield obtained in mixture with

those of the monoculture. In order to address the question

of the effect that density has on competitive relations, the

method can be further expanded by conducting the replacement

series at varied densities (Jolliffe et al., 1984; Paul and

Ayres, 1987; Rejmanek et al., 1989).

The additive study methodology can be expanded to

enhance competition assessment by a technique termed






9

"reciprocal yield analysis" (Rejmanek et al., 1989; Roush et

al., 1989). Application of this approach involves the

establishment of additive series for each of the mixture

components of interest, which is termed a "complete additive

design". These series can be established at each of several

levels for the set density component of a series

(Vleeshouwers et al., 1989). With this approach, mixtures

of two or more components can be studied (Rejmanek et al.,

1989). The effect of each component of the mixture on the

others) can be assessed.

The results obtained in competition studies can be

influenced by a number of factors (Aldrich, 1987) and basic

methods are found to be biased with certain mixtures

(Connolly, 1986). Studies have been conducted to compare

some of the alternative approaches to competition analysis

when applied to specific situations (Rejmanek et al., 1989;

Roush et al., 1989). The reciprocal yield analysis

technique was found in both studies to be useful for the

assessment of interspecific versus intraspecific competition

effects. The basic replacement series was considered in

both studies to be inferior in describing competitive

effects over several densities and in describing the effect

of species proportions. In the study by Roush et al. (1989)

the synthetic no-interaction approach was also evaluated.

It was found to be superior for analysis of the influence of

proportion on interspecific competition. The reciprocal








yield approach was preferred for the partitioning of

interspecific and intraspecific competition effects and the

measurement of interactions between density and proportion.


Factors Influencing Competition


Factors that have been found to influence plant

competition include environment, plant emergence

characteristics, growth rates and other components of plant

size and function (Radosevich, 1987). Proximity of

individual plants to others is directly influenced by

density, spatial arrangement and species proportion, which

are considered influential factors in the study of

interactions between plants. Planting pattern has been

found to influence competitiveness of crop plants with

square and equilateral triangle patterns making crops more

competitive than rectangular patterns (Aldrich, 1987).

Aldrich (1987) examined causes of variability found in

results of weed density/crop yield studies from apparently

similar situations. In agricultural crops, the amount of

plant growth that has accrued up to any moment is density

dependent at the start of the crop cycle and becomes density

independent as time passes (for closed canopy crops).

Biomass production rate is maximum once the transition has

been made to the density independent stage. Environmental

factors or factors related to weed presence may influence

the time required to reach the density independent stage,








which in turn influences the length of time over which the

maximum biomass production rate occurs.

The nature of acquisition of resources by plants can

influence the potential onset of competition for them

(Aldrich, 1987). Mobile resources in a given region of the

soil may be used simultaneously by crop and weed plants

whose root systems have not yet overlapped. On the other

hand, relatively immobile resources, such as phosphorus,

would only be taken up from soil in close proximity to roots

of the individual plants. Consequently, competition for

mobile resources could begin sooner than for less mobile

ones.

Crop emergence timing relative to that of weeds can

influence competitive interactions between the two. Elberse

and deKruyf (1979) studied the influence of emergence timing

on competition between lambsquarters (Chenopodium album L.)

and barley (Hordeum vulgare L.) using a replacement study

approach. Barley was sown at 7, 21 and 31 days after

lambsquarters and three harvests were made for each sowing.

With the earliest barley planting, the crop progressively

suppressed the weed. The 21 day planting resulted in weed

and crop coexisting with neither becoming dominant. For the

latest planting the crop replaced the weed. Aldrich (1987)

emphasized the importance of the effect of emergence timing

as a source of variation in weed interference studies,

pointing out that it can be difficult to control when








working with various species under field conditions.

Temperature may influence growth rates and competitive

ability of weed species differentially (Aldrich, 1987).

Studies are cited in which redroot pigweed (Amaranthus

retroflexus L.) grew faster and was more competitive than

lambsquarters with increased temperature.


Factors for which Plants Compete


Of the factors that influence plant growth some are of

a consumable nature (nutrients, water, CO2, 02 and light)

while others are conditions, e.g. temperature (Radosevich

and Holt, 1984). It is those of the first category for

which competition might occur (Radosevich and Holt, 1984;

Aldrich, 1987). An additional distinction is made that

light is available in a fixed amount while water and

nutrients, though available in limited amounts, can be

manipulated by the grower (Glauninger and Holzner, 1982).

Aldrich (1987) cites that environmental conditions would

preclude competition for CO2 and 02. DeWit (1960) discussed

space as a factor for which plants compete, contending that

it is not feasible to separate out individual components.

However, Glauninger and Holzner (1982) consider space to be

just the combined effects of competition for light,

nutrients and water. Nevertheless, competition for both

below and above ground factors appears to be important.

Wilson (1988) reviewed literature concerning studies in








which attempts were made to separate competitive

interactions occurring between shoots from those arising in

the root environment. Consideration of 47 cases led to the

conclusion that both types of competition can be important

although root competition is generally of greater

importance.

Light is a resource that is "continuously available in

a constant amount" (Glauninger and Holzner, 1982, p. 150).

In situations where water and nutrients are adequate, light

is the factor that will ultimately determine plant growth

(Gardner et al., 1985). Competition for light begins when

one plant begins to shade another (Glauninger and Holzner,

1982). Because light is nontransferable within a plant,

shading of a portion of one plant by another results in the

shaded region not receiving the resource (Aldrich, 1987).

Characteristics such as stature, growth habit and ability to

photosynthesize in less than full sunlight contribute to the

ability of plant species to compete effectively with other

species (Glauninger and Holzner, 1982). Well developed crop

stands may offer sufficient light interception to provide

significant reduction of weed development. Redroot pigweed

biomass was found to be less in soybeans planted on 25 cm

row spacings than when the crop was grown on 76 cm row

spacings (Legere and Schreiber, 1989). With more closely

spaced rows the canopy would close earlier. Weed species

vary in their susceptibility to growth reduction due to








shading. Common purslane (Portulaca oleracea L.) is

reportedly very susceptible while lambsquarters is much less

so (cited in Glauninger and Holzner, 1982). For crops,

early growth, height and density are important for

successful competition with weeds to occur (Glauninger and

Holzner, 1982).

Only a limited amount of information is available on

the effects of weed interference on lettuce. This

information has focused on temporal aspects of weed presence

in the crop. Field studies in England demonstrated that

when the crop was free of weeds at 3 weeks after emergence,

no yield loss resulted from subsequent weed infestation

(Roberts et al., 1977). In studies conducted in Florida,

the effects of duration of weed interference was assessed

for monospecies weed infestations of livid amaranth

(Amaranthus lividus L.) at densities of 120 plants per m2

and common purslane at 15 plants per m2 (Shrefler et al.,

1991). In the case of livid amaranth, if weeds were removed

at 19 days after planting of the crop no yield loss

occurred, whereas delaying weed removal an additional 15

days resulted in complete loss of marketable yield. In the

case of common purslane, delaying weed removal from 16 to 37

days after planting of the crop resulted in a 36% decrease

in marketable yield, which occurred in a linear fashion with

time.








On high organic matter soils in Florida, relatively

large amounts of phosphorus fertilizer are required for

lettuce production (Sanchez and Burdine, 1988). This

nutrient has been found to influence the outcome of plant

competition in some instances (Buckeridge and Norrington-

Davies, 1986; Kranz and Jacob, 1977; Weiner, 1980).

However, no information is available concerning the effect P

fertility would have on weed interference with lettuce.

The general objectives of the studies discussed herein,

and as outlined in chapter I, are to characterize the

effects of interference by spiny amaranth (Amaranthus

spinosus L.) on lettuce and determine how P fertilization

would, in turn, influence these effects. Studies in chapter

III concern the effects of weed interference on the crop

under field conditions as influenced by P fertilizer

application method. The additive method of the study of

weed interference was used in order to assess the impact of

the weed on the crop. The effect of the duration of weed

interference was assessed at 2 weed densities that fall

within the range of what might be encountered under

commercial production conditions.

Studies in chapter IV concern the determination of how

P fertility influences the growth of each of the weed and

the crop when grown in the presence of one another. A

replacement study design at a series of densities was used

for these studies. This design permits assessment of the






16

effects of competition on each species of the mixture. By

conducting studies at several densities, the effect of plant

density on competition, as assessed by the replacement

design, can also be evaluated. In addition, this approach

also permits analysis of competition using the synthetic no-

interaction approach (Roush et al., 1989). The methodology

used in this study, therefore, was chosen to enable some

comparison of methods for use in the assessment of

competition as influenced by P fertility.












CHAPTER 3
SPINY AMARANTH INTERFERENCE IN CRISPHEAD LETTUCE
AS INFLUENCED BY PHOSPHORUS FERTILIZER
APPLICATION METHOD


Introduction


Production of high yielding, quality lettuce (Lactuca

sativa L.) requires the intensive management of a number of

factors (Ryder, 1979), two important ones being weed control

and mineral nutrition of the crop. While weed control is

commonly employed in commercial lettuce production, detailed

documentation of the effects of uncontrolled weed growth on

lettuce is scarce. Some information concerning the effect

weeds have on lettuce is found in studies in which

herbicides and other control methods are evaluated (Dusky et

al., 1988; Giannopolitis et al., 1989). Such studies

indicate that substantial weed losses will occur if weeds

are not controlled.

Studies specifically concerned with how weeds affect

lettuce have also been reported. Field studies in England

showed that densities of 65 weeds per m2 resulted in

complete loss of marketable yield when weeds were allowed to

remain present in the crop (Roberts et al., 1977). Weed

populations in these studies were comprised of a mixture of

gramineous and dicotyledonous species. Elimination of weeds

17








by hand at 3 weeks after emergence of the crop, with no

additional weed removal, was sufficient to prevent

significant yield loss.

The effect of early season interference with crisphead

lettuce by specific weed species has also been assessed

(Shrefler et al., 1991). Livid amaranth (Amaranthus lividus

L.) at 120 plants per m2 did not affect marketable lettuce

yield when removed 19 days after planting of the crop.

Delaying weed removal an additional 15 days resulted in

complete loss of marketable yield. Common purslane

(Portulaca oleracea L.) at 15 plants per m2 resulted in a

reduction in the quantity of marketable lettuce heads as

weed removal was delayed from 16 to 37 days after planting

of the crop (Shrefler et al., 1991). Weed interference

during this 21 day period resulted in a 36% yield decrease,

which occurred in a linear fashion with time.

Weeds adversely affect crops due to competition for

light, water and nutrients (Patterson, 1985). When water

and nutrient availability are optimal, light is the factor

which ultimately limits crop growth (Gardner et al., 1985).

If sunlight interception by a plant is reduced due to

shading caused by a neighbor plant then interference for

light necessarily occurs (Aldrich, 1987). This is in

contrast to the below-ground factors, i.e. nutrients and

water, which may be available in excess of the consumptive

capacity of the plants. Because competition for light is








the result of shading, the onset of this interaction is

related to weed density. As weed density decreases,

individual weeds will have to achieve greater size in order

to cause comparable amounts of shading.

The sensitivity to shading which would occur in

competition for light could be expected to vary with plant

species. This is because plants differ in their capacity to

use light in photosynthesis (Gardner et al., 1985). Leaves

of plant species in which carbon is fixed via the C4 pathway

do not generally light saturate at full sunlight. Species

of the C3 type, on the other hand, commonly reach maximal

photosynthetic levels at less than full sunlight. This was

found to be the case for lettuce, a C3 plant, well into the

vegetative growth portion of the crop cycle (Sanchez et al.,

1989). As lettuce approached maturity, however,

photosynthetic photon fluxes equivalent to full sunlight

were required for maximum carbon exchange rates.

Competitive interactions between plants may also take

place through the root systems. As a means of assessing the

relative importance of plant competition due to below-ground

interactions versus those occurring above-ground, Wilson

(1988) summarized the results of a number of studies found

in the literature in which separation of root and shoot

competition was attempted. These studies used partitions to

isolate root systems or shoots between plants of mixtures in

which competition was being studied. Wilson (1988)








concluded that both root and shoot competition were

important although the former tended to be so more

frequently.

Water is one below-ground factor for which competition

between weeds and crops can occur (Aldrich, 1987; Griffin et

al., 1989). Competition for water would occur when soil

moisture is inadequate to provide the combined needs of weed

and crop plants (Aldrich 1987). Soil moisture levels can

also influence the nature of competition that occurs between

plant species. Griffin et al. (1989) studied competition

between soybean (Glycine max (L.) Merrill) and Florida

beggarweed (Desmodium tortuosum Sw. (DC) under different

soil moisture regimes. These experiments showed that at

optimum levels of available soil water soybean was more

competitive than the weed. Soybean, however, was more

sensitive to decreased soil water potential than the weed.

Consequently, at low levels of water availability the weed

became more competitive than the crop.

Below-ground competition between different species can

sometimes be attributed to the availability of mineral

nutrients and nitrogen. In the previously discussed work by

Wilson (1988), one category of study considered is where

competition is evaluated at several levels of nutrient

availability. Indices of competition that were calculated

for each study evaluated were highest at low levels of

nutrient availability as often as at high levels. No








generalization that competition was reduced by increased

nutrient levels was possible.

Alkamper (1976) reviewed literature on the combined

influence of weeds and fertilization on crop production.

Based on this, he concluded that fertilization without weed

control results in increased crop damage by weeds, except

possibly in the case of low weed densities.

Soil fertility requirements for lettuce production on

organic soils in Florida have been established (Sanchez,

1990). Phosphorus is applied based on quantitative analysis

of soil for water extractable P. Once this value is

obtained a corresponding fertilizer P rate can be determined

from a calibration curve (Sanchez and Burdine, 1988).

Inadequate soil P predictably results in sub-optimal lettuce

yields (Sanchez et al., 1988).

In lettuce production on south Florida histosols

fertilizer P is applied prior to planting the crop (Sanchez,

1990). Sanchez et al. (1990) demonstrated that fertilizer P

is utilized with greater efficiency when applied in a 'band'

instead of broadcast, the traditional application method.

With the band application, fertilizer P is placed in line

with the crop row, 5 cm below the soil surface. These

studies showed that equivalent yields can be obtained with

two thirds lower fertilizer P rates if the nutrient is

applied in a band rather than broadcast.






22

Fertilizer P placement close to lettuce seeds has been

shown to provide the additional benefit of an increase in

seedling growth that can not be achieved when the fertilizer

is mixed throughout the soil (Costigan, 1984). This

response was attributed to the slow early growth of lettuce

roots. Thus, close placement of fertilizer P resulted in a

growth response to the nutrient even though root system

development was still minimal.

Economic analysis of vegetable production in Florida

suggests that reduced use of fertilizer P could render

lettuce production unprofitable (Alvarez and Sanchez, 1991).

Environmental concerns, among others, call for reduced

fertilizer application due to its potential adverse effect

on wetland ecosystems which are closely linked to

agricultural lands. Band application of fertilizer P

appears to be one means by which nutrient use rates could be

reduced without reducing lettuce yields.

In anticipation of the adoption of a banded fertilizer

P application practice, studies were carried out to

determine how this might influence the impact of weed

interference on lettuce. The weed selected for use in these

studies was spiny amaranth (Amaranthus spinosus L.). Spiny

amaranth is the most ubiquitous weed in lettuce production

areas in Florida and difficult to control in the crop due to

a lack of registered herbicides (Dusky et al., 1988). Since

little detailed information on the effect of weeds on








lettuce is available, studies were designed to assess both

the effect of weed density and duration of weed presence on

the crop under several fertilizer P application regimes.

Several methods can be used to assess interference

between different plant species when grown in mixture

(Cousens, 1991). There are two basic methods which have

received considerable attention (Cousens, 1991; Radosevich,

1987). The replacement or substitutive series was advanced

by the work of de Wit (1960) although it was reportedly

established prior to this (Cousens, 1991). The additive

study is the second principle method (Cousens, 1991;

Radosevich and Roush, 1990). Each of these methods is

useful for answering a unique set of questions (Cousens,

1991). The additive design addresses weed management

concerns such as the establishment of relationships between

weed density and crop yield and the comparison of weed

species for their ability to suppress crop yields. The

replacement design is used for determining which of two

components of a binary mixture is the better competitor. It

is also utilized when study objectives include the

identification of the nature of the interaction between the

two components of the mixture (Cousens, 1991; Jolliffe et

al., 1984; Radosevich 1987). An additive design was chosen

for these studies since the principle objective was to

determine the effect weeds have on the crop under the

different fertilizer P application regimes.









Materials and Methods


Three field experiments were conducted during 1991 and

1992 at the Everglades Research and Education Center at

Belle Glade, Florida as shown in Table 3.1. Soil at the

study site is classified as a Pahokee series muck [Euic,

hyperthermic Lithic Medisaprists (Soil Survey Staff, 1978)].

Prior to selecting the field location for each experiment,

soil was sampled and analyzed for nutrients as described by

Sanchez (1990). Soil chemical characteristics of the fields

in which the experiments were conducted, which were selected

for low P fertility, are given in Table 3.2. Fertilizer

application of all nutrients except P was made to the entire

study area as recommended based on soil test results.

Phosphorus application, as triple superphosphate, was also

made based on soil-test recommendation. Instead of being

applied to the entire study area, fertilizer P was applied

as an experimental treatment. Specific treatments used

were; 1) no added P, 2) broadcast applied P and 3) band

applied P. Broadcast applied P was distributed evenly on

the soil surface and mixed in the soil with a disk harrow.

Where P was applied as a band treatment in the 1991

experiments, one third the amount of P was used, on a per

plot basis, as was used in the broadcast treatments. In

1992 the P fertilizer rate was increased to one half of the








Table 3.1 Schedule of procedures performed in the spring
1991, fall 1991 and spring 1992 experiments.


Experiment

Procedure Spr. 1991 Fall 1991 Spr. 1992
date -
Apply broadcast
fertilizers 2/27 10/10 2/19

Apply banded P
fertilizer 3/1 10/28 2/21

Plant 3/1 10/29 2/21

Thin crop stand 3/22 11/26 3/17

Establish weed
densities 3/22 11/28 3/23

Harvest crop 5/8 1/8 4/27


aPotassium and micronutrients applied to the entire study
area. Phosphorus applied broadcast as an experimental
treatment.








Table 3.2 Soil chemical characteristics of the study sites
and fertilizer phosphorus application rates in the spring
1991, fall 1991 and spring 1992 experiments.


Phosphorus application

Experiment pH Pwa broadcast band
Kg ha' -

spring 1991 6.2 3 1320 440

fall 1991 6.6 3 1320 440

spring 1992 6.1 3 1320 660

aWater extractable phosphorus soil test index (Sanchez,
1990).


bPhosphorus applied as triple superphosphate.






27

amount of the broadcast application. A pressed bed planting

system was used (Lucas, 1982). Fertilizer P for the banded

treatments was applied during the bed formation process.

This was done in such a way that a fertilizer band about 8

cm wide was placed 5 cm below the soil surface, in line

with, and centered on, the crop row. Beds were constructed

on 91 cm centers and had a width of 48 cm at the top.

Crisphead lettuce 'Southbay' (Guzman, 1984) was direct

seeded two rows to a bed. Seeding dates are given in Table

3.1. Rows were spaced 30.5 cm apart. Pelleted lettuce seed

was sown in groups of three closely spaced seed every 30 cm

of row. Plants were thinned to 1 every 30 cm at the times

indicated in Table 3.1.

Spiny amaranth densities of 1 and 4 plants per 2.3 m of

bed were established by selective removal of unwanted weeds

from the naturally arising seedling population. The dates

at which selective weed removal was performed are given in

Table 3.1 and will be referred to as 'plot establishment'.

The weeds kept in the plots were located on top of the bed

and in the region in between the two crop rows. Once

established, plots were maintained free of unwanted weeds by

hand weeding or hoeing.

Following plot establishment, duration of weed

interference treatments were achieved by removal of weeds

from plots at various intervals. Weed removal timings for

the spring 1991, fall 1991 and spring 1992 studies were 7,






28

21, 35 and 49 days, 7, 17, 28 and 36 days and 6, 16, 27 and

36 days, respectively. Following removal of spiny amaranth,

these plots were maintained free of all weeds. Spiny

amaranth plants removed from plots were dried at 60 C. Dry

weights were obtained as a measure of biomass. Nutrient

composition analysis was conducted on spiny amaranth of the

different removal times.

A randomized complete block experimental design with

four replications and with a split plot treatment

arrangement was used for each experiment. Phosphorus

application method comprised the main plots. A complete

factorial arrangement of the weed densities and removal

times constituted the subplots. There was a weed-free check

plot for each P application regime. Individual plots

consisted of a 9.1 m long section of raised bed. Subplots

were separated by a weed-free bed which was planted to

lettuce.

Lettuce harvest was timed so that the most mature

lettuce plants would be harvested as late as possible

without becoming excessively hard or cracked due to over

maturity. Harvest dates are given in Table 3.1. Fifteen to

20 lettuce heads were harvested from each plot. In the

spring of 1991, 10 heads were taken from each side of the

bed. Lettuce heads were cut such that several of the

outermost leaves were discarded. In the fall 1991 and

spring 1992 studies lettuce was harvested from the western








side of the bed and with all intact leaves attached.

Lettuce heads were counted and weighed. After the initial

weighing of lettuce in the fall 1991 and spring 1992, a

second weight was obtained following removal of the wrapper

leaves (outer leaves which fall free from the head).

Lettuce yield data were analyzed using analysis of variance

techniques (Steele and Torrie, 1980) and SAS General Linear

Models software (SAS Institute Inc., 1987). Lettuce yield

data for the weed-free subplots were used as a control for

each of the two weed density treatments. Duration of weed

interference effects were analyzed with regression

techniques (Freund et al., 1986). Regression analyses were

performed on data means. Interactions between main effects

were assessed using least squares means analysis and paired

means comparisons (SAS Institute Inc., 1987).

Spiny amaranth biomass data were analyzed using

techniques similar to those used for lettuce. Prior to

analysis, amaranth data were log transformed after adding 1

to the data values. This was done because correlation of

means and residuals was suggested by plotting residual

values (Freund et al., 1986). Logarithmic transformation is

appropriate when data for which values differ by a large

magnitude are to be analyzed with analysis of variance

(Steele and Torrie, 1980). This transformation resulted in

residual plots which appeared to be of random distribution.








Samples of 4 to 6 lettuce plants were kept from each

plot for nutrient composition analysis. In the spring of

1991 and 1992 heads were kept as harvested. In the fall of

1991 trimmed heads were used. Lettuce heads of these

samples were cut up, dried and stored at 60 C.

Weed and lettuce samples to be used for nutrient

analysis were ground in a stainless steel outfitted

laboratory mill to pass through a 1 mm mesh sieve. Samples

were thoroughly mixed following grinding and a subsample of

the ground material was used for analysis. Subsamples were

wet ashed as described by Wolf (1982). Nitrogen was

determined by a micro-Kjeldahl method (Bremner and Mulvaney,

1982), P was determined colormetrically and K, Ca, Mg, Zn,

Fe, Mn and Cu by atomic absorption spectrophotometry.

Nutrient concentration data in lettuce and spiny amaranth

were analyzed using the techniques discussed for lettuce

yield and weed biomass data. Micronutrient data for lettuce

and spiny amaranth are in Appendix A.


Results and Discussion


Spinv Amaranth Growth


The effects of P application (PA), weed density (WD)

and duration of weed interference (DWI) on spiny amaranth

biomass (on a per plant basis) for the spring 1991, fall

1991 and spring 1992 experiments are shown in Table 3.3.

Substantially greater biomass was achieved in the spring








Table 3.3 The effect of P application, weed density,
duration of weed interference and their interactions on
spiny amaranth biomass in the spring 1991, fall 1991 and
spring 1992 experiments.


Experiment
Main
Effectab Spring 1991 Fall 1991 Spring 1992
biomass (g plant) -
PA

None 49.9ac 5.9 55.8
Band 60.7b 6.0 69.0
Broadcast 57.4b 4.9 67.2

Signif.d ** ns ns

WD

Low 65.0 6.8 70.7
High 47.0 4.5 57.3

Signif. ns *** ***

DWI

1 0.1 0.1 1.2
2 3.7 1.2 8.0
3 55.9 7.4 55.5
4 169.2 13.8 191.3

Signif. *** ** ***


Interactions
level of significance -
PA x WD ns ns ns
PA x DWI ** ns ***
WD x DWI ns ns ns
PA x WD
x DWI ns ns ns


aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI). Low and high weed
densities are 1 and 4 spiny amaranth plants per 2.3 m of
bed. Durations of weed interference, in days after plot
establishment, are (7, 21, 35 and 49), (7, 17, 28 and 36)
and (6, 16, 27 and 36) for experiments 1, 2 and 3,
respectively.






32

Table 3.3--continued.

bAnalysis performed on log transformed data. Log
transformation was made after addition of 1 to data. Non-
transformed means are presented.
cPhosphorus application means within a column followed by a
common letter are not significantly different at alpha=0.05
based on Duncan's Multiple Range Test.
significance for the main effect.







33

1991 and 1992 studies than in the fall of 1991. This can be

attributed to the late time of establishment of the amaranth

relative to the crop in the fall of 1991 (Table 3.1). The

amaranth seedling emergence was very sparse in the fall 1991

experiment until two to three weeks after the crop was

planted. This may have been due to moisture at the soil

surface being inadequate for weed seed germination to occur.

There was less than 1.2 cm of rainfall during the first 14

days following planting in the fall 1991 experiment while in

the 1991 and 1992 spring experiments there was 2.6 and 3 cm,

respectively, during the same period. In the spring 1992

study, plots were also established later than in the spring

1991 study (Table 3.1). In this case, however, amaranth per

plant biomass at the first weed removal time was about 10

times greater than in the other studies (1.2 g per plant in

1992 versus 0.1 g per plant in the 1991 studies) (Table

3.3). In the fall 1991 experiment, only the main effects

DWI and WD affected amaranth biomass (Table 3.3). Effects

of DWI reflect the influence of plant age on biomass. Weed

biomass per plant was 50% greater (6.8 versus 4.5 g per

plant) under low than high weed density (Table 3.3). A

density affect also occurred in the spring 1992 experiment

such that amaranth biomass was 20% greater under low than

high weed density. These reductions in biomass at the high

weed density can likely be attributed to increased

intraspecific competition. Classic plant population biology






34

studies show an inverse relationship between plant density

and yield of individual plants in a population (Firbank and

Watkinson, 1990).

In the spring 1991 and 1992 experiments there were

significant interactions between PA and DWI on weed biomass.

The effect of PA and DWI in the spring 1991 experiment is

given in Table 3.4. Biomass of amaranth removed from

lettuce at 21 days after plot establishment was greater

where P was applied banded or broadcast than where no P

applied. The effect of PA and DWI in the spring 1992

experiment is given in Table 3.5. Phosphorus application

was found to affect amaranth biomass only in the case of the

final DWI. When weeds were present until time of harvest of

the crop, weed biomass under the banded P application was

greater than where no P was added.

In general, the principle factors affecting spiny

amaranth biomass were the length of time it was grown (DWI)

and the density at which it was grown. While there is some

evidence of a positive response to P application, this was

not the case in the fall 1991 experiment (Table 3.3) or at

the majority of the durations of weed interference in the

spring studies (Tables 3.4 and 3.5). Thus, under the low P

fertility conditions of these experiments, spiny amaranth

growth, as reflected in per plant biomass, was similar with

or without the addition of fertilizer P. This is in

contrast to another weedy amaranthus species, redroot








Table 3.4 The effect of fertilizer phosphorus application
on spiny amaranth biomass in the spring 1991 experiment.


Phosphorus Duration of weed interference
Application 7 21 35 49

biomass (g plant") -

None 0.07a 1.9 a 46.8a 165a

Banded 0.1 a 3.97b 64.6a 174a

Broadcast 0.14a 5.2 b 56.4a 168a

aDuration of weed interference in lettuce in days after plot
establishment.

bValues within a column followed by the same letter are not
different based on paired means comparison at alpha = 0.05.

cAnalysis performed on log transformed data. Log
transformation was made after adding 1 to data values. Non-
transformed means are presented in the table.








Table 3.5 The effect of fertilizer phosphorus application
on spiny amaranth biomass in the spring 1992 experiment.


Phosphorus Duration of weed interference
Application 6 16 27 35

biomass (g plant') -

None 1.50a 8.55a 55.0a 158a

Banded 1.05a 8.34a 59.8a 215b

Broadcast 1.08a 7.20a 51.5a 201ab


aDuration of weed interference in lettuce in days after plot
establishment.

bValues within a column followed by the same letter are not
different based on paired means comparison at alpha = 0.05.

cAnalysis performed on log transformed data. Log
transformation was made after adding 1 to data values. Non-
transformed means are presented in the table.









pigweed (Amaranthus retroflexus L.), which was found by

Hoveland et al. (1976) to be one of the more P responsive

weeds of several that were studied. Phosphorus uptake

variability is known to occur between plants of different

species (Chapin and Bieleski, 1982; Hoveland et al., 1976).

The effect of P fertility on biomass production by 17 plant

species occurring as weeds in southern United States was

studied (Hoveland et al. 1976). Considerable variation was

found among the weed species. Those species exhibiting

greatest response to P also showed the most severe

deficiency under low P conditions. Variation in P uptake

capacity by plants has been attributed to root size

differences (Lindgren et al., 1977) and the presence of root

hairs (Itoh and Barber, 1983). Spiny amaranth apparently

has a root system which is well adapted to low P conditions.


Lettuce Yields


The effects on lettuce yields of phosphorus

application, weed density and duration of weed interference

and their interactions are shown in Table 3.6. Lettuce

yields are presented on a weight per plant basis. In the

fall 1991 experiment no interactions were found between the

main effects (Table 3.6). In this experiment, application

of P and duration of weed interference affected lettuce

yields. Yields obtained with band and broadcast application

were not found to differ. Regardless of the method used, P








Table 3.6 The effect of P application, weed density,
duration of weed interference and their interactions on
lettuce yields in the spring 1991, fall 1991 and spring 1992
experiments.


Experiment
spr. 1991 fall 1991b spr. 1992
Main -trim +trim -trim +trim
Effect
Lettuce Yield (kg plant) -
PA

None 0.34ac 0.42a 0.28a 0.57a 0.43a
Band 0.65b 0.89b 0.67b 1.01b 0.78b
Broadcast 0.70b 0.93b 0.68b 1.07b 0.83c

Signif.d *** *** *** *** ***

WD

Low 0.58 0.74 0.54 0.91 0.70
High 0.53 0.75 0.54 0.85 0.66

Signif. *** ns ns *** ***

DWI

1 0.60 0.76 0.55 0.92 0.71
2 0.60 0.75 0.55 0.94 0.73
3 0.55 0.77 0.55 0.92 0.71
4 0.54 0.76 0.54 0.84 0.65
5 0.51 0.69 0.50 0.78 0.60

Signif. ** ns *** ***


Interactions
level of significance -
PA x WD ns ns ns ns ns
PA x DWI ns ns ns *
WD x DWI *** ns ns *** ***
PA x WD
x DWI ns ns ns ns ns

aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI) following plot
establishment. Low and high weed densities are 1 and 4
plants per 2.3 m of bed, respectively. Durations of weed
interference 1, 2, 3, 4 and 5 correspond to numbers of days
after plot establishment of (0, 7, 21, 35 and 49), (0, 7,






39

Table 3.6--continued.

17, 28 and 36) and (0, 6, 16, 27 and 36) for the spring
1991, fall 1991 and spring 1992 experiments, respectively.

yields for untrimmed (-trim) and trimmed (+trim) lettuce.

cPhosphorus application means within a column followed by a
common letter are not significantly different at alpha=0.05
based on Duncan's Multiple Range Test.
dSignificance of the main effect.






40

application resulted in 2.2 and 2.4 times greater yields of

untrimmed and trimmed lettuce, respectively, than when no P

was applied. Duration of weed interference also had a

significant effect on yield of untrimmed lettuce (Table

3.6). No meaningful trend over time was found however.

Yields were consistent for the durations of weed

interference of 0 through 28 days and appeared to decrease

only when amaranth was present through harvest of the crop

(36 days after plot establishment).

In the spring 1991 and 1992 experiments, P application

resulted in significant lettuce yield increases (Table 3.6).

In the 1992 experiment no differences were found between the

band and broadcast application methods for untrimmed

lettuce. Untrimmed lettuce yields were an average of 82%

greater when P was applied, regardless of the application

method. For spring 1991 yield, and trimmed yield in 1992,

interactions occurred between PA and DWI (Table 3.6). The

affect of PA and DWI in the spring of 1991 is shown in

Figure 3.1. Phosphorus application by either method

resulted in greater lettuce yields at all durations of weed

interference. Broadcast application resulted in greater

yields than band application in weed-free lettuce but not

where weeds were present for any duration. Duration of weed

interference had a significant effect on lettuce where P was

applied broadcast but not where it was applied in a band or

not applied at all.




















0.8 -
a
a .
a

S a a --. a
0.6 b a
a
Os a



0.4 b b b




0.2
0 7 21 35 49
Duration of Weed Interference (days)

-no P, predicted X no P, actual -band, predicted
band, actual --broadcast, predicted broadcast, actual









Figure 3.1 The effect of phosphorus fertilizer application
and duration of weed interference on lettuce yields in the
spring 1991 experiment. Actual means and regressions of
yield on durations of weed interference are shown.
Regression equations of yield versus time for no added P,
band applied P and broadcast applied P are Y=0.364-
0.003X+0.000058X2 (R2=0.81, not significant),
Y=0.657+0.0031X-0.0001X2 (R2=0.86, not significant) and
Y=0.789-0.00057X+0.00004X (R2=0.99, sig. at p=0.001),
respectively. Actual values associated with a common letter
at a given duration are not different at alpha=0.05 based on
paired means comparisons.








The interaction of PA and DWI for trimmed lettuce

yields in the spring 1992 study is shown in Figure 3.2.

Phosphorus application by each of the methods resulted in

increased yields at all durations of weed interference.

Broadcast application resulted in greater yields than band

application in weed-free (duration of weed interference

equal to 0) lettuce but not where weeds were present for any

duration. Duration of interference had a significant effect

on lettuce yields both in the case of no added P and in the

case of the broadcast application.

Lettuce responded to P application as would be expected

based on the results of Sanchez et al. (1990). Phosphorus

application by either the band or broadcast method resulted

in substantial yield increases over those obtained where no

P was applied. Analysis of the spring 1991 yield data

suggested that slightly greater yields were obtained with

the broadcast than the band application. For the spring

1992 study the P fertilizer rate in the band application was

increased from 0.33 to 0.5 of that used in the broadcast

application (Table 3.2). In spite of this modification,

yields were still slightly greater for the broadcast

application in the weed-free lettuce. It is not clear why

this yield difference occurred in the weed-free plots.

In the fall 1991 experiment lettuce yields were not

found to be differentially affected by weed density (Table

3.6). Lettuce yields in spring 1991 and 1992 were less



















1
a a a

.. .a
0.8-. ~- -- *. a
o aa

Sa
L 0.6
c b b

0.4



0.2
0 6 16 27 36
Duration of Weed Interference (days)

-no P, predicted X no P, actual -band, predicted
band, actual -broadcast, predicted broadcast, actual







Figure 3.2 The effect of phosphorus fertilizer application
and duration of weed interference on trimmed lettuce yields
in the spring 1992 experiment. Actual means and regressions
of yield on durations of weed interference are shown.
Regression equations of yield versus time for no added P,
band applied P and broadcast applied P are Y=0.454+0.0043X-
0.00021X2 (R2=0.98, significant at p=0.05), Y=0.825-0.00245X
(RO=0.65, not significant) and Y=0.891+0.00056X-0.000149X2
(R2=0.99, sig. at p=0.05), respectively. Within given
durations, actual values associated with common letters are
not different at alpha=0.05 based on paired means
comparisons.








under high weed density than low weed density. There were

also interactions between weed density and duration of weed

interference in each of the spring experiments. The effect

of WD and DWI on yields in the spring 1991 experiment is

shown in Figure 3.3. Under high weed density, lettuce

yields decreased in a linear fashion with increasing

duration of weed interference. Under low weed density there

was no significant effect of duration of weed interference

on lettuce yields.

For the spring 1992 experiment the effects of WD and

DWI are shown in Figure 3.4 for untrimmed yields and in

Figure 3.5 for trimmed yields. Under each of the weed

densities, untrimmed and trimmed lettuce yields decreased in

quadratic fashions with increased duration of weed

interference. The effect of high weed density was more

pronounced than that of low weed density for each of

untrimmed and trimmed lettuce yields.

In the fall 1991 experiment lettuce yields were

essentially not affected by spiny amaranth (Table 3.6). In

this experiment, the emergence of spiny amaranth relative to

that of the crop occurred later than in the spring 1991

experiment. In the spring 1991 study, plots were

established at 3 weeks after planting of the crop (Table

3.1). In the fall, weeds were not large enough for plots to

be established until 4 weeks after the crop was planted.

The final spiny amaranth biomass obtained in the fall 1991


















0.8


Sa a a
0.6 ) K -- X5-
a -
I a
-0.4 b
Sb


0.2



0
0 7 21 35 49
Duration of Weed Interference (days)

low den predicted X low den actual
--high den predicted high den actual












Figure 3.3 The effect of weed density and duration of weed
interference after plot establishment on lettuce yields in
the spring 1991 experiment. Actual means and regressions of
yield on duration of weed interference are shown.
Regression equations of yield versus time for low and high
weed densities are Y=0.586+0.0003X (R2=0.27, not
significant) and Y=0.624-0.0041X (R2=0.95, sig. at p=0.01),
respectively. Within given durations, actual values
associated with common letters are not different at
alpha=0.05 based on paired means comparisons.






46












a a


0.9
a


00.8
Sb \
Sb
0.7 \"
\


0.6
0 6 16 27 36
Duration of Weed Interference (days)

low den predicted X low den actual
-- high den predicted high den actual









Figure 3.4 The effect of weed density and duration of weed
interference on untrimmed lettuce yields in the spring 1992
experiment. Actual means and regressions of yield on
durations of weed interference are shown. Regression
equations of yield versus time for low and high weed
densities are Y=0.925+0.00207X-0.000118X2 (R2=0.96,
significant at p=0.05 and Y=0.927+0.00218X-0.00024X2
(R =0.86, significant at p=0.05), respectively. Within
given durations of weed interference, actual values
associated with common letters are not different at
alpha=0.05 based on paired means comparisons.



















0.8

a
a
0.7
Sa a a



0.6 N
b b
\ b


0.5
0 6 16 27 36
Duration of Weed Interference (days)

low den predicted X low den actual
--high den predicted high den actual










Figure 3.5 The effect of weed density and duration of weed
interference after plot establishment on trimmed lettuce
yields in the spring 1992 experiment. Actual means and
regressions of yield on durations of weed interference are
shown. Regression equations of yield versus time for low
and high weed densities are Y=0.72+0.00086X-0.000076X2
(R2=0.95, significant at p=0.05) and Y=0.72+0.0018X-0.0002X2
(R2=0.96, significant at p=0.05), respectively. Actual
values associated with a common letter at a given duration
are not different at alpha=0.05 based on paired means
comparisons.








experiment was only 8% and 7%, respectively, of that

achieved in the spring 1991 and spring 1992 studies (Table

3.3).

Crop plants are often found to have critical weed-free

periods (Zimdahl, 1980). If the crop is maintained free of

weeds during such critical periods, yield losses to weeds do

not occur. For lettuce, Roberts et al. (1977) found that if

the crop was free of weeds at 3 weeks after emergence of the

crop, yields would not be affected by subsequent weed

emergence. It appears that in the fall 1991 experiment weed

establishment did not take place within the critical period

for crisphead lettuce.

In the spring experiments lettuce yields were reduced

most when spiny amaranth was present at high densities

throughout the season until the time of harvest of the crop.

Under such conditions, yields were reduced by 30% in the

spring of 1991 (Figure 3.3) and by 20% in the spring of 1992

(Figures 3.4 and 3.5). The apparently greater yield

reduction in 1991 than 1992 may be related to the manner in

which yield reductions occurred. Yields decreased in a

linear fashion with duration of weed interference in the

spring of 1991, suggesting that weeds began to affect yields

beginning with the earliest durations of weed interference.

In 1992, yield reduction occurred in a curvilinear fashion

(Figures 3.4 and 3.5), which suggests that amaranth did not








begin to affect lettuce yields as early as in the spring

1991 study.

The onset of crop loss to weed competition can be

attributable to variation in prevailing environmental

conditions and resource availability (Zimdahl, 1980). For

example, competition has reportedly been shown to limit crop

yield earlier on when moisture availability is the limiting

growth factor rather than light. Timing of the onset of

competition has also been attributed to nutrient

availability, with crop yield loss occurring earlier under

high levels of nutrient availability. It may be that a

factor such as these may have resulted in spiny amaranth

competition beginning to have an affect on lettuce yield

earlier in the spring of 1991 than in 1992.

Under low weed density lettuce yields were affected

only in the 1992 experiment even though weed biomass

achieved in the two spring studies was comparable (Table

3.3). It is not clear why yields were reduced by the low

weed density in the spring of 1992 but not in 1991. One

possibility is that lettuce growing under the conditions of

the 1992 study was more susceptible to the effects of

interference by spiny amaranth. Lettuce in the spring 1991

experiment was only partially trimmed in contrast to the

trimmed lettuce of the 1992 study. Direct comparison of

yields for the two spring studies cannot be made due to

differences in the degree of trimming used at harvest in the








two experiments. Comparison of spring 1991 yields with

those of the more thoroughly trimmed lettuce of spring 1992

does suggest, however, that the yields of the latter were

greater (Table 3.6). It may be that where yield was

potentially less (i.e. in the spring of 1991) it was also

less susceptible to reduction by interference from the low

density of spiny amaranth. For example, if lettuce growth

was limited by light availability, then nutrient uptake by

the low weed density may not have been sufficient to result

in mineral nutrient availability becoming more limiting to

growth than light availability. During the final five weeks

prior to harvest of the spring 1991 experiment there was

considerable rainfall and substantially reduced solar

radiation during two periods of several days. Under field

conditions, lettuce yield has been shown to be sensitive to

light reductions of as little as 27% of prevailing solar

radiation when such shading occurs during head formation

(Sanchez et al., 1989). Based on these findings, it was

suggested that cloudy weather may be a limiting factor for

lettuce yields in south Florida. In the 1992 study, where

greater yields were obtained, a resource other than light

may have been the yield limiting factor. In this case,

spiny amaranth at the low density of the study may have

effectively reduced the availability of the factor

ultimately limiting lettuce yield, thus resulting in the

yield reductions that occurred.








Lettuce Nutrient Analysis


Concentrations of macronutrients in lettuce in the

spring 1991, fall 1991 and spring 1992 studies are given in

Tables 3.7, 3.8 and 3.9, respectively. Nitrogen

concentrations ranged from 1.21% in the 1992 study to 3.5%

in the fall of 1991. Greatest yields were obtained in the

spring of 1992 (Table 3.6), the study in which lettuce N was

lowest (Table 3.9). In the spring 1992 study, fertilizer P

application by either of the two methods resulted in a 15%

decrease in lettuce N concentrations.

Concentrations of P in lettuce ranged from 0.4 to 1.2

in the studies (Tables 3.7, 3.8 and 3.9). In each of the

spring studies (Tables 3.7 and 3.8), P concentrations were

greatest with broadcast P application, intermediate with

banded P application and lowest when no P fertilizer was

applied. In the spring 1991 study there was an interaction

between WD and DWI. This interaction is presented in Table

3.10. For the duration of weed interference of 35 days, P

concentrations in lettuce were lower under high than low

weed density. This suggests that spiny amaranth may have

been competing with lettuce for P; the more intense

competition under the higher spiny amaranth density

resulting in lowered P concentrations in lettuce.

In the spring 1992 study PA interacted with weed

density (Table 3.9). The interaction is explored in Table








Table 3.7 The effect of P application, weed density,
duration of weed interference and their interactions on
macronutrient concentrations in lettuce in the spring 1991
experiment.


Main
Effecta
PA
None
Band
Broadcast

Signif.b

WD
Low
High

Signif.

DWI
0
7
21
35
49

Signif.


N

2.38
2.02
2.08


2.18
2.16

ns


2.27
2.16
2.26
2.12
1.99

ns


Nutrient
P K


- concentration (% of
0.54ac 6.18
0.79b 5.73
1.14c 6.33


0.83
0.79

ns


0.84
0.83
0.78
0.80
0.80

ns


6.34
5.80

ns


5.61
6.39
6.52
6.47
5.46

ns


Ca
dry wt.)
1.89
1.78
1.65


1.87
1.69


1.73
1.62
1.80
1.97
1.83


Interactions level of significance -

PA x WD ns ns ns ns *
PA x DWI ns ns ns ns ns
WD x DWI ns ns ns ns
PA x WD
x DWI ns ns ns ns

aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI). Low and high weed
densities are 1 and 4 plants per 2.3 m of bed, respectively.
Durations of weed interference in days after plots
established.

bSignificance of the main effect.

cPhosphorus application means within a column followed by a
common letter are not significantly different at alpha=0.05
based on Duncan's Multiple Range Test.


Mg

0.45a
0.41b
0.39b


**


0.43
0.40


0.41
0.41
0.42
0.44
0.41








Table 3.8 The effect of P application, weed density,
duration of weed interference and their interactions on
macronutrient concentrations in lettuce in the fall 1991
experiment.


Main Nutrient
Effect N P K Ca Mg
PA concentration (% of dry wt.) -
None 3.31 1.12 8.78 1.44 0.36
Band 3.31 1.14 8.34 1.49 0.36
Broadcast 3.50 1.22 8.49 1.48 0.36

Signif.b ns ns ns ns ns

WD
Low 3.39 1.16 8.56 1.49 0.36
High 3.36 1.16 8.51 1.45 0.36

Signif. ns ns ns ns ns

DWI
0 3.24 1.17 8.65 1.49 0.35
7 3.48 1.13 8.53 1.44 0.36
17 3.32 1.17 8.52 1.46 0.36
28 3.39 1.20 8.37 1.45 0.36
36 3.44 1.13 8.61 1.5 0.36

Signif. ** ns ns ns ns

Interactions level of significance -

PA x WD ns ns ** ns ns
PA x DWI ns ns ns ns ns
WD x DWI ns ns ns ns ns
PA x WD
x DWI ns ns ns ns ns

aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI). Low and high weed
densities are 1 and 4 plants per 2.3 m of bed, respectively.
Durations of weed interference in days after plots
established.


bSignificance of the main effect.








Table 3.9 The effect of P application, weed density,
duration of weed interference and their interactions on
macronutrient concentrations in lettuce in the spring 1992
experiment.


Main Nutrient
Effect N P K Ca Mg
PA concentration (% of dry wt.) -

None 1.44ac 0.45a 5.57 2.27a 0.58
Band 1.24b 0.67b 5.50 2.50b 0.61
Broadcast 1.21b 0.80c 5.16 2.59b 0.62

Signif.b *** ns ns

WD

Low 1.29 0.65 5.33 2.47 0.61
High 1.30 0.64 5.49 2.43 0.60

Signif. ns ns ns ns ns

DWI

0 1.37 0.65 5.32 2.39 0.60
6 1.25 0.67 5.65 2.50 0.62
16 1.28 0.61 5.13 2.40 0.59
27 1.34 0.63 5.47 2.48 0.60
36 1.24 0.65 5.49 2.49 0.61

Signif. ns ns ns ns ns

Interactions
level of significance -
PAx WD ns ns ns ns
PA x DWI ns ns ns ns ns
WD x DWI ns ns ns ns ns
PA x WD
x DWI ns ns ns ns ns


aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI). Low and high weed
densities are 1 and 4 plants per 2.3 m of bed, respectively.
Durations of weed interference in days after plots
established.

bSignificance of the main effect.

cPhosphorus application means within a column followed by a







55

Table 9--continued.

common letter are not significantly different at alpha=0.05
based on Duncan's Multiple Range Test.








Table 3.10 The interaction between duration of weed
interference and spiny amaranth density for phosphorus
concentration in lettuce in the spring 1991 experiment.


Spiny amaranth density
Duration of
weed interferencea Low High

P conc. (% of dry wt.) -

0 0.84ac 0.84a

7 0.83a 0.81a

21 0.77a 0.85a

35 0.94a 0.72b

49 0.83a 0.78a

aDuration of weed interference in days after plot
establishment.

Low and high densities are 1 and 4 plants per 2.3 m of bed,
respectively.

cValues within a row followed by the same letter are not
different based on paired means comparisons at alpha = 0.05.








Table 3.11 The interaction between phosphorus fertilizer
application and spiny amaranth density for phosphorus
concentration in lettuce in the spring 1992 experiment.


Spiny amaranth density
Phosphorus Within row
application Low High comparisons

P cone. (% of dry wt.) -

None 0.46ac 0.44a ns

Band 0.65b 0.69b ns

Broadcast 0.82c 0.77c ns

aLow and high densities are 1 and 4 plants per 2.3 m of bed,
respectively.

bDifferences are significant based on paired means
comparisons.
cValues within a column followed by the same letter are not
different based on paired means comparisons at alpha = 0.05.








3.11. Phosphorus concentrations in lettuce under each of

low and high weed densities were lowest where no P was

applied, intermediate with banded P and greatest with

broadcast P. The source of the interaction, however, was

not detected. There were not, therefore, any consistent

interactions between the P fertilizer treatments and the

other experimental factors for P concentrations in lettuce

in the spring 1991 and 1992 experiments, the two in which

yields were effected in a similar manner in the two

experiments.

In the fall experiment, the high P concentrations in

lettuce for which no fertilizer P was added were unexpected

(Table 3.8). Even though lettuce P concentration at

maturity was not found to be affected by P application,

lettuce yields did show a response to applied P fertilizer

(Table 3.6), as was previously discussed. Lettuce leaf

samples were collected at midseason from plants in the weed-

free plots of this experiment. Phosphorus concentration

data for these leaf samples are given in Table 3.12. These

data indicate that P concentration in lettuce leaves did

show a response to fertilizer P application. Sanchez et al.

(1990) found that, on occasion, good lettuce yields are

obtained when soil P status is below that which would

normally be required. They attributed this to a combination

of weather conditions being optimal for efficient P use by

the plant and to movement of P into the root as soil water








Table 3.12 Phosphorus concentrations in outermost intact
lettuce leaves of weed-free lettuce at mid-season in the
fall 1991 experiment.



Phosphorus Phosphorus
Application Concentration

% of dry weight -

None 0.21aa

Banded 0.39b

Broadcast 0.49c


aValues followed by the same letter are not different based
on Duncan's multiple range test at alpha=0.05.








moved upwards under conditions of high evapotranspiration.

It may be that such a situation occurred during the latter

stages of crop development, resulting in the high P

concentrations found at maturity. Yield differences for the

P application treatments would therefore have occurred as a

result of the nutritional status that prevailed early in the

season, when a response to P was evident. The finding that

70% of the total nutrient uptake by a crisphead lettuce crop

can occur during the last three weeks prior to harvest (Zink

and Yamaguchi, 1962) favors the plausibility of this

explanation.

Potassium concentrations across experiments ranged from

5% in the spring 1992 study to greater than 8% in the fall

of 1991 (Tables 3.7, 3.8 and 3.9). In the spring 1991 and

1992 experiments, K was not found to be influenced by the

experimental factors (Tables 3.7 and 3.9). The interaction

between PA and WD in the fall 1991 experiment (Table 3.8)

did not appear to be meaningful.

Calcium concentrations in lettuce ranged from as low as

1.44% in the fall 1991 study (Table 3.8) to as high as 2.5%

in the spring 1992 study (Table 3.9). There did not appear

to be any meaningful trends for Ca concentration response to

P application.

Magnesium concentrations ranged from 0.36% in the fall

study (Table 3.8) to 0.6% in the spring 1992 experiment

(Table 3.9). There did not appear to be any meaningful








interactions between the experimental factors for Mg

concentrations.

Nutrient concentrations in lettuce were influenced more

often by fertilizer P application than by weed density or

duration of weed interference. For each of N, P and Ca, P

application influenced the concentrations of the nutrient in

lettuce in at least one experiment. The most pronounced

effect of P application was on lettuce P concentrations.

This P concentration response to fertilizer P application by

lettuce somewhat parallelled the response of lettuce yields

to P application, as previously discussed, for the two

spring studies. There was, however, some difference in how

yields and lettuce P concentrations responded to the

application method. For the broadcast application of

fertilizer P, lettuce P concentrations were greater than in

the case of the band application for each of the spring

studies. In each of these experiments, yields were

essentially not found to differ for the band and broadcast

fertilizer P application (Figures 3.1 and 3.2). This

suggests that with broadcast fertilizer application lettuce

may have been taking up more P than was required for the

yields obtained in the experiments.

Lettuce yields in the spring studies were influenced by

interactions between WD and DWI (Table 3.6). There was

little indication, however, of interactions between WD and

DWI for P concentrations in lettuce. The only reduction in






62

P concentration due to increased weed density was for the 35

day duration of weed interference in the spring 1991

experiment. It does not appear that interactions between

these experimental factors were of underlying importance to

the P nutrition of lettuce as measured by nutrient

concentrations in the crop at maturity.

Although the crop was affected by spiny amaranth

interference, the immediate cause of yield loss to the weed

does not appear to be due to interference with the P

nutrition of the crop. Phosphorus nutrition has been

implicated as a factor which can influence the outcome of

intraspecific competition in plants (Buckeridge and

Norrington-Davies, 1986; Chapin and Bieleski, 1982). In

these studies with lettuce and spiny amaranth, however,

essentially only the growth of lettuce was found to be

influenced by P fertility. Competition between lettuce and

spiny amaranth was influenced by P nutrition only in the

sense that it was a determinate factor for lettuce growth.

Although spiny amaranth interference reduced lettuce

yields in each of the spring studies, these yield reductions

do not appear to have been due to the reduction of P

availability to lettuce, as measured by P concentrations at

maturity. However, the fact that some reduction in lettuce

P concentrations did occur due to increased weed density

suggests that there may have been some interaction between

spiny amaranth and the P nutrition of lettuce. The








nutritional data was obtained for lettuce in these studies

at maturity. This data, therefore, does not provide direct

indication of the nutritional status of the crop during the

entire period during which competition between it and the

weed occurred. Nutritional status while competition is

occurring, rather than after it has occurred, can be of

greater value in assessing the importance of mineral

nutrition on competition (Glauninger and Holzner, 1982).


Spiny Amaranth Nutrient Analysis


Macronutrient concentrations in spiny amaranth for the

spring 1991, fall 1991 and spring 1992 experiments are given

in Tables 3.13, 3.14 and 3.15, respectively. In the spring

of 1992, N concentrations were influenced by duration of

weed interference (Table 3.16), which is a measure of plant

age. In the fall of 1991, duration of weed interference

affected N concentrations in amaranth (Table 3.14). The

interactions between the experimental factors for N

concentrations did not appear to be meaningful.

Phosphorus concentrations in the fall 1991 study were

influenced by duration of weed interference but not the

other experimental factors (Table 3.14). In the spring 1991

study there was an interaction between density and duration

of weed interference (Table 3.13), which is explored in

Table 3.16. At the final duration of weed interference, P

concentration was lower for the high than the low density.








Table 3.13 The effect of P application, weed density,
duration of weed interference and their interactions on
macronutrient concentrations in spiny amaranth in the spring
1991 experiment.


Main Nutrient
Effecta N P K Ca Mq
PA -- concentration (% of dry wt.) -
None 3.51 0.60ac 7.03 2.81a 0.80a
Band 3.26 0.69b 6.80 2.65b 0.73b
Broadcast 3.23 0.80c 7.00 2.68b 0.70b

Signif.b ns ** ns *

WD
Low 3.38 0.69 6.85 2.72 0.75
High 3.29 0.71 7.04 2.70 0.74

Signif. ns ns ns ns ns

DWI
7 4.64 0.88 5.04 3.32 0.99
35 3.10 0.80 9.47 2.98 0.78
49 2.26 0.43 6.43 1.84 0.46

Signif. *** *** ** *** ***

Interactions level of significance -

PAx WD ** ns ns ns ns
PA x DWI *** ns *** ns ns
WD x DWI ns ** ns
PA x WD
x DWI ns ns ** ns *


aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI). Low and high weed
densities are 1 and 4 plants per 2.3 m of bed, respectively.
Durations of weed interference in days after plots
established.

bSignificance of the main effect.

cPhosphorus application means within a column followed by a
common letter are not significantly different at alpha=0.05
based on Duncan's Multiple Range Test.








Table 3.14 The effect of P application, weed density,
duration of weed interference and their interactions on
macronutrient concentrations in spiny amaranth in the fall
1991 experiment.


Main Nutrient
Effect N P K Ca Mq
PA concentration (% of dry wt.) -
None 2.24ac 0.96 6.45 3.15 0.83
Band 2.32a 0.97 6.00 3.27 0.79
Broadcast 2.23a 1.07 6.20 3.27 0.75

Signif.b ns ns ns ns

WD
Low 2.18 0.98 6.29 3.19 0.78
High 2.34 1.02 6.16 3.28 0.80

Signif. ns ns ns ns ns

DWI
7 4.02 0.91 4.90 2.89 0.85
17 1.92 0.86 6.04 3.75 0.83
28 1.69 1.11 6.65 3.86 0.74
36 1.71 1.12 7.09 2.38 0.75

Signif. *** *** *** *** ***

Interactions level of significance -

PA x WD ns ns ns ns ns
PA x DWI ** ns ns ns ns
WD x DWI ns ns ns ns ns
PA x WD
x DWI ns ns ns ns ns

aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI). Low and high weed
densities are 1 and 4 plants per 2.3 m of bed, respectively.
Durations of weed interference in days after plots
established.

bSignificance of the main effect.

cPhosphorus application means within a column followed by a
common letter are not significantly different at alpha=0.05
based on Duncan's Multiple Range Test.








Table 3.15 The effect of P application, weed density,
duration of weed interference and their interactions on
macronutrient concentrations in spiny amaranth in the spring
1992 experiment.


Main Nutrient

Effect N P K Ca Mq
concentration (% of dry wt.) -
PA
None 2.64 0.65 5.83 3.92 0.82
Band 2.86 0.65 5.66 3.87 0.82
Broadcast 2.92 0.68 5.33 3.86 0.84

Signif.b ns ns ns ns ns

WD
Low 2.83 0.67 5.66 3.93 0.83
High 2.79 0.66 5.55 3.84 0.83

Signif. ns ns ns ns ns

DWI
6 4.77 0.87 4.56 4.89 0.86
16 1.47 0.78 5.70 3.61 0.96
27 3.20 0.57 6.17 3.72 0.81
36 1.79 0.42 5.98 3.32 0.69

Signif. *** *** *** *** ***

Interactions level of significance -

PAx WD ns ns ns ns ns
PA x DWI ns *** ** ns **
WD x DWI ns ns ns ns ns
PA x WD
x DWI ns ns ns ns ns


aMain effects are P application (PA), weed density (WD) and
duration of weed interference (DWI). Low and high weed
densities are 1 and 4 plants per 2.3 m of bed, respectively.
Durations of weed interference in days after plots
established.


bSignificance of the main effect.








Table 3.16 The interaction between duration of weed
interference and spiny amaranth density for phosphorus
concentration in spiny amaranth in the spring 1991
experiment.


Spiny amaranth density
Duration of
weed interference Low High

P conc. (% of dry wt.) -

7 0.86ac 0.90a
35 0.76a 0.83a
49 0.45a 0.40b

aDuration of weed interference in days after plot
establishment.

bLow and high densities are 1 and 4 plants per 2.3 m of bed,
respectively.

cValues within a row followed by the same letter are not
different based on paired means comparisons at alpha = 0.05.








In this experiment there was also a response to P

application (Table 3.13). Fertilizer P application resulted

in higher spiny amaranth P concentrations, which were also

greater with the broadcast than the band application. Spiny

amaranth biomass showed a positive response to P application

(Table 3.3), however, biomass was not differentially

influenced by the P application method. It thus appears

that, in the case of the broadcast treatment, amaranth took

up more P than was needed for the biomass accumulation that

occurred in these studies. In the spring 1992 experiment

there was an interaction between P application and duration

of weed interference (Table 3.15), which is explored in

Table 3.17. For the duration of weed interference of 6

days, P concentrations were found to be greater for the

broadcast applied P than for the other treatments. Spiny

amaranth biomass, however, did not show a differential

response to the P application treatments for this duration

of weed interference (Table 3.5). Thus, P concentration in

spiny amaranth does not appear to have been a limiting

factor to growth of the plant, as reflected by biomass, in

these studies.

In the fall 1991 experiment, K concentrations in spiny

amaranth were influenced by duration of weed interference

but not the other experimental factors (Table 3.14). The

interactions between the experimental factors for K

concentration did not appear to be meaningful.








Table 3.17 The interaction between duration of weed
interference and phosphorus application for phosphorus
concentration in spiny amaranth in the spring 1992
experiment.


Duration of Phosphorus application
weed interference None Band Broadcast

P conc. (% of dry wt.)- -

6 0.74ab 0.86a 1.03b

16 0.76a 0.75a 0.84a

27 0.64a 0.57a 0.52a

36 0.48a 0.40a 0.39a

aDuration of weed interference in days after plot
establishment.

values within a row followed by the same letter are not
different based on paired means comparisons at alpha = 0.05.






70

In the fall of 1991 (Table 3.14) and the spring of 1992

(Table 3.15) Ca concentrations were only influenced by

duration of weed interference. The interaction in the

spring 1991 study did not appear to be meaningful. In the

spring 1991 experiment Ca concentrations showed a negative

response to P application but did not differ for the two P

application methods (Table 3.13).

In the fall 1991 experiment Mg concentrations showed a

response to duration of weed interference but not to the

other experimental factors (Table 3.14). Interactions

between the experimental factors for Mg concentrations did

not appear to be meaningful.

The predominate factor which influenced nutrient

concentrations in spiny amaranth was duration of weed

interference. Mineral nutrient concentrations in plants are

controlled by genetic uptake potential of the plant,

nutrient availability in the growing medium and plant age

(Mengel and Kirkby, 1987). Plant age was increasing with

DWI and therefore may have been a factor influencing the

nutrient concentrations found in amaranth. Nutrient

availability of muck soils varies with field conditions

(Lucas, 1982). Rainfall, for example, can leach nitrate N

out of the root zone. The effects of time, therefore, may

be due to a combination of several factors.

Spiny amaranth biomass was influenced by density.

However, nutrient concentrations do not appear to have been








the underlying cause of this affect. Foliage of adjacent

plants at low density never overlapped but foliage did

overlap extensively for plants at the high density. This

suggests that interference for light between plants may have

been a factor involved in the biomass responses to density.

The effect of spiny amaranth on lettuce yields was

dependent on growth of the weed. Lettuce yields were not

affected by the relatively low weed biomass of the fall 1991

experiment. In the spring experiments, where substantially

greater weed biomass was achieved, lettuce yields were

reduced as the duration of weed interference was extended

(Figures 3.3, 3.4 and 3.5).

Lettuce yields were also influenced by fertilization

such that application by either method resulted in yield

increases of approximately 100 percent. Spiny amaranth

biomass, on the other hand, was only marginally influenced

by P fertilization. There was also some indication that

broadcast P application resulted in lettuce being more

susceptible to weed interference than did band application

(Figures 3.1 and 3.2). Lettuce yields in the absence of

weed interference were slightly greater when P was applied

broadcast than when banded. The more pronounced effect of

weed interference on lettuce where P was applied broadcast

suggests that lettuce susceptibility to weed interference is

greatest where yield is potentially greatest. This effect

did not seem to be due to the nutritional status of the






72

crop. Nutrient concentrations in lettuce were not found to

be differentially influenced by weed interference in any

meaningful fashion.

This study corroborates the findings of Sanchez et al.

(1990) that band application of P is a viable alternative to

broadcast application for lettuce production on south

Florida histosols. However, the yield differences obtained

in weed-free lettuce for band versus broadcast application

suggests that the latter method was slightly superior in the

current study. These differences occurred regardless of

whether one half or one third of the broadcast rate was used

in the band application and, therefore, may be due to more

than optimization of rate. These results suggest that

further study of fertilizer P placement is needed.

Spiny amaranth interference with lettuce was not found

to be dependent on P fertilization to any great extent.

While lettuce growth was dependent on P fertilization, this

was not found to be the case for spiny amaranth. Spiny

amaranth generally grew as well with or without the addition

of P fertilizer. Thus, any changes in the competitive

interactions between lettuce and spiny amaranth due to P

fertilization would be due to responses by the crop, but not

the weed, to P fertility. Because spiny amaranth grows well

at low P, it could be expected to be a good competitor under

conditions of low P. In light of the findings of these

studies, in order to avoid crop loss to competition by spiny






73

amaranth, growers will need to continue to exercise the same

weed control practices when using a banded P application

system as with the currently used broadcast technique.












CHAPTER 4
SPINY AMARANTH (Amaranthus spinosus L.) COMPETITION
WITH LETTUCE (Lactuca sativa L.) AS INFLUENCED BY
PHOSPHORUS FERTILITY OF MUCK SOIL


Introduction




Weed management in lettuce (Lactuca sativa L.) produced

on histosols in Florida is highly dependent on control

measures applied after weeds emerge, consequently resulting

in intermittent weed presence in the crop. This is

especially so during the first several weeks of crop growth

since weed control measures are generally not initiated

until 3 weeks after planting, the time at which the crop is

thinned to the desired stand and weeds are removed. Studies

concerned with how the duration of weed presence affects

lettuce have been reported. Field studies in England showed

that weed densities of 65 per m2 resulted in complete loss

of marketable yield when weeds were allowed to remain

present in the crop (Roberts et al., 1977). Weed

populations in these studies were comprised of a mixture of

gramineous and dicotyledonous species. These studies showed

that elimination of weeds by hand at 3 weeks after emergence

of the crop, with no additional weed removal, was sufficient

to prevent yield loss. The effect of early season

74







75

interference with crisphead lettuce by specific weed species

has also been assessed (Shrefler et al., 1991). Livid

amaranth (Amaranthus lividus L.) at 120 plants per m2 did

not affect marketable lettuce yield when removed at 19 days

after planting of the crop. Delaying weed removal an

additional 15 days resulted in the complete loss of

marketable yield. Common purslane (Portulaca oleracea L.)

at 15 plants per m2 resulted in a reduction in the quantity

of marketable lettuce heads as weed removal was delayed from

16 to 37 days after planting of the crop (Shrefler et al.,

1991). Weed interference during this 21 day period resulted

in a 36% yield decrease, which occurred in a linear fashion

with time.

Lettuce production on histosols also requires careful

management of soil fertility (Sanchez, 1990; Lucas, 1982).

The crop responds to each of P and K amendment in a

predictable manner on these soils (Sanchez and Burdine,

1988). Critical concentrations in lettuce tissue have been

established for these nutrients (Sanchez et al., 1988).

While K fertilizer can be applied as needed after the crop

is planted, the entire amount of P to be used for a crop is

applied before planting. This P application has

traditionally been made by broadcasting the fertilizer

uniformly over the field and incorporating it into the soil

prior to planting. It has been shown, however, that when

fertilizer P is applied in a band below the crop row,








optimum yields can be obtained with substantially reduced

rates of P fertilizer on a field basis (Sanchez et al.,

1990). It was also found that concentrations of available P

in the crop root zone at 30 days after application were

greater in the case of the band method than the broadcast

method. Weed and crop interactions during early growth

would therefore be occurring under different P availability

conditions for these two fertilizer application methods.

Lettuce is particularly sensitive to phosphorus supply

during early growth. This was demonstrated in greenhouse

experiments where P was applied simultaneously by 2 methods

(Costigan, 1984). One method was to thoroughly mix dry,

granulated triple superphosphate with the soil. The highest

rates used were adequate to give optimal yields when this

was the only source of P fertilization. Pregerminated

lettuce seed were sown into, and covered with, this soil.

The second source of P was a solution of NH4H2PO4 which was

applied to the seeding zone of the soil. By 21 days after

planting, lettuce dry weight had responded positively to the

increased rate of P supplied as a solution but not to P

applied dry to the soil.

Phosphorus status has also been shown to influence the

capacity of lettuce seedlings to respond to other nutrients

(Costigan and Heaviside, 1988). In these studies, starter

solutions of varied nutrient concentrations were compared

for use with lettuce transplants. As P concentration in








lettuce tissue increased from 0.3 to 0.6 %, the growth

response of lettuce to the various starter solutions

increased in a linear fashion. Thus, P availability was the

ultimate factor limiting plant growth.

In another study the effect of interruptions in the

supplies of N, P and K to lettuce during early growth

following transplanting was assessed (Burns, 1987). Plants

were grown in sand culture to which nutrient solutions were

supplied. The effects of deprivation of each of N, P and K

were tested by withholding the individual nutrients,

beginning at the time of transplant, for several durations.

Shoot growth rates decreased within 2 and 6 days after

withholding the nutrients N and K, respectively, but not

until 9 days in the case of P. Restoration of the supply of

each of the nutrients resulted in a rapid increase in plant

growth rate. Their data also suggest that growth rate

sensitivity to withholding N and K was not as great as for P

deprivation, which resulted in abrupt cessation of growth

once the nutrient was withheld for 9 days.

Interactions between weeds and crop plants are found to

be mediated by the availability of resources consumed during

plant growth (e.g. water and nutrients) (Aldrich, 1987;

Radosevich and Holt, 1984), non-consumable conditions that

influence growth (e.g. temperature) and allelopathic

responses (Radosevich and Holt, 1984). Competition refers

to situations in which the growth of plants interacting with








one another is limited due to consumable resources being

quantitatively inadequate (Goldberg, 1990; Radosevich and

Holt, 1984). In the case of nutrients as the resource for

which competition occurs, there are two characteristics

which may influence the competitive ability of plants

(Goldberg, 1990). One of these is the capability of rapid

depletion of the resource in question. The other is the

ability to grow at depleted resource levels.

Plant competition studies provide a means of evaluating

the effects that certain management practices have on weed

and crop interactions (Conolly, 1988). A technique for the

study of nutrients as limiting factors of plant growth under

interspecific competition was established by Hall (1974a)

and has been applied for several nutrients and several plant

mixtures (Bhaskar and Vyas, 1988; Hall, 1974a; Hall, 1974b).

The approach utilizes the replacement series method.

Monocultures of each of two species are established along

with mixtures of the two species at varied proportions.

Overall density is held constant for the series. While the

replacement series technique is generally applied to biomass

or yield, Hall's approach is to also look at nutrient

content in plant tissue. By including varied fertility as a

factor, the results give an indication of how component

plants of the mixture fare in extracting the nutrient in the

presence of the other species. Using this approach, Hall

(1974b) was able to conclude that the grass component of a








grass-legume mixture competed more successfully than the

legume for K. Bhaskar and Vyas (1986) applied the technique

to a wheat (Triticum aestivum L.) and common lambsquarters

(Chenopodium album L.) mixture. They found that competitive

interference occurred for P (with common lambsquarters being

more aggressive), and to a lesser extent for N (in which

case wheat was more aggressive).

Wilson (1988) tabulated findings of studies testing the

effects of nutrient addition to plant mixtures in which

plant interaction was limited to that which would occur

below ground. Several indices of competition were utilized,

depending on the nature of the original study, to assess the

effect of nutrient addition on competition. Competition was

found to increase as often as it decreased, leading Wilson

to conclude that there is "no case for using the concept of

reduced competition effects at higher resource levels in the

interpretation of ecological processes or experimental

results". It appears that the influence of nutrient

addition on plant competition must be considered on a case

by case basis rather than broad generalization.

Nevertheless, it has been demonstrated that nutrients

can be important in influencing the outcome of interspecific

plant competition in agricultural settings (Hall, 1974b;

Siddiqi et al., 1985; Shribbs, 1986; Weiner, 1980). In

these studies, competition is found to be dependent on

amounts of nutrients supplied to plants grown in mixture.








The influence of one nutrient on competition may also be

dependent on the status of other nutrients. In studies with

red clover (Trifolium incarnatum L.) and Italian ryegrass

(Lolium multiflorum Lam.) the effect of combined P and K

addition under low N fertility was examined using

replacement methods (Weiner, 1980). Red clover was more

sensitive to intraspecific competition than it was to

interspecific competition with ryegrass. These competitive

interactions were differentially influenced by soil

fertility in several ways. In portions of the study

conducted at low N, ryegrass negatively affected red clover

at low P and K fertility but not when P and K levels were

increased. When N was added along with P and K, ryegrass

became dominant over the legume.

Glauninger and Holzner (1982) discussed why competition

for nutrients is not estimable by simply looking at nutrient

levels in plants. Recognizing that crops can suffer

severely due to nutrient uptake by weeds, these authors are

of the opinion that more important than absolute uptake of

nutrients by competitors are the relations between nutrient

availability and the needs of the crop. Interpretation of

altered nutrient status of crops due to weed competition is

probably dependent on the nature of competition. In studies

with bell peppers, increased weed "cover" resulted in

decreased pepper foliage Fe while B, Cu, P and K were

increased (Frank et al., 1988). These increases may be the








result of reduced plant growth due to others factors

becoming limiting rather than an increase in nutrient

uptake. Reduced growth has a concentrating effect on

nutrients (Mengel and Kirkby, 1987).

Additional reasoning for nutrient levels in plants

being of limited value as an indicator of competition for

nutrients is that the roots of weeds and crops in mixture

may vary in the region of the soil profile they inhabit, and

thus not be drawing totally on the same nutrient pool. A

study by Chambers and Holm (1965) using P32 to study the

site of P uptake from soil by green foxtail (Setaria viridis

(L.) Beauv.), redroot pigweed (Amaranthus retroflexus L.)

and common bean (Phaseolus vulgarus L.) demonstrated this

effect. Common bean received 60% of its P from within a 7.5

cm radius around itself, to a depth of 7.5 cm, while redroot

pigweed absorbed 60% of its P at the 15 cm depth of a 15 cm

radius. Localized nutrient supply can have a strong

stimulatory effect on root proliferation in the vicinity of

the nutrient (Anghinoni and Barber, 1980; Drew, 1978; Mengel

and Kirkby, 1987). The possibility of such a response

should be kept in mind in studying fertility effects on

plants growing in mixture when nutrient additions are made

to the rooting media.

Studies on nutrients as limiting factors for which

plants compete have most often considered N although any

nutrient could conceivably be important in this context








(Zimdahl, 1980). Several studies provide evidence for the

occurrence of competition for nutrients other than N between

roots (Caldwell et al., 1987; Kranz and Jacob, 1977).

Caldwell et al. (1987) demonstrated that plants whose roots

shared a mutual soil location were drawing on a common pool

of P. This was shown by measuring uptake of dual isotopes

of P. Defoliation of one of the species of the mixture,

which would reduce its nutrient uptake, resulted in an

immediate increase in P uptake by the other species.

Competition for P was studied in mixtures of flax

(Linum usitatissimum L.) and large seeded falseflax

(Camelina sativa (L.) Crantz) by monitoring the uptake of

P32 over time by both species (Kranz and Jacob, 1977). When

grown in mixture, uptake of P by flax was slower than when

grown in monoculture. Falseflax uptake of P, conversely,

was greater in mixture than in monoculture. Falseflax thus

competed more strongly for P than flax, both

interspecifically and intraspecifically.

The occurrence of competition between roots for

resources is dependent on the overlap of soil depletion

shells of adjacent roots (Fusseder et al., 1988). A model

for the determination of root overlap and the effect of

nutrient uptake on soil P status in the root zone was

applied to field studies with maize. These studies

demonstrated that competition for P was nearly non existent

in this case. This was because soil P in the root






83

microenvironment was not altered by P uptake for a distance

greater than 1 mm away from the root.

Changes in competitive interactions between plants as

nutrient availability is modified could conceivably be

dependent on the relative responsiveness of the components

of the mixture (Hoveland et al., 1976; Weiner, 1980) as well

as on the existence of an actual nutrient competition

phenomena (Hall, 1974b; Weiner, 1980). The fact that a

mixture responds to nutrient availability, however, does not

indicate that the nature of competition within the mixture

will necessarily be influenced by alteration of nutrient

levels (Buckeridge and Norrington-Davies, 1986).

Increases in nutrient availability in competitive

situations can be more detrimental than advantageous to the

crop in some instances. In response to a proposal that weed

competition for nutrients could be compensated for by

increased fertilization, Alkamper (1976) reviewed literature

on the influence that fertilization has on crops and weeds

and the control of weed infestations. It was concluded that

fertilizing to reduce crop loss to weeds is not feasible,

except in the case of low weed densities. The reasons for

this are that weeds may grow faster than crops and absorb

nutrients more rapidly, that weeds are more responsive to

nutrients than crops in some cases and that fertilization of

weeds may only hasten the onset of crop loss due to weeds.

The study also suggested that proper timing of








fertilization, in conjunction with weed control, can be

valuable in enhancing the ability of the crop to suppress

weeds that survive control measures.

Weedy species vary in response to differences in soil

fertility. Some grow well at low levels of available

nutrients while phenotypic plasticity of others allows them

to take advantage of high fertility by growing luxuriously

(Glauninger and Holzner, 1982). Plants in the family

Chenopodiaceae respond so to high nitrate availability.

Redroot pigweed has been found to respond to P more so than

several other warm season weeds (Hoveland et al., 1976), and

exhibit P accumulation seven times that of beans (Zimdahl,

1980). Hoveland et al. (1976) cite that species of the

genera Amaranthus, Chenopodium and Portulaca are

particularly efficient in K uptake and that the presence and

growth of lambsquarters has been considered an indicator of

P deficiency in soil.

Nutrient addition can indeed modify the competitive

interactions in plant mixtures (Hall, 1974b; Siddiqi et al.,

1985; Shribbs, 1986; Weiner, 1980). Hall (1974b) studied

the influence of added K on competition between Nandi

setaria (Setaria anceps cv. Nandi) and Greenleaf desmodium

(Desmodium intortum cv. Greenleaf). At low K, desmodium

growth was greatly suppressed by setaria when the latter

comprised only 25% of the mixture. With added K, however,








desmodium growth was only marginally suppressed by setaria.

Setaria growth was not greatly influenced by added K.

The potential for nutrient addition to influence

competitive relations between plants can be not only species

dependent but also, for crops, cultivar dependent (Siddiqi

et al., 1985). The competitive ability between barley

(Hordeum vugare L.) cultivars and wild oat (Avena fatua L.)

at varied K fertility was studied. Some cultivars were

competitive at high and low K, others at high K only and

still others were only weakly competitive at even high K.

The influence of ground cover species on apple (Malus

pumila Miller) seedlings was studied by Shribbs et al.

(1986). Leaves of apples grown with groundcover were found

to have reduced N levels. Addition of N overcame the effect

only partially.

The overall objective of studies reported here was to

assess the competitive interactions under varied P regimes

between lettuce and spiny amaranth (Amaranthus spinosus L.)

during early growth. The replacement series design is

useful for exploring the way in which two species interact

(Cousens, 1991). The outcome of studies utilizing this

approach is found to be influenced by the density at which

experiments are conducted (Rejmanek et al., 1989). As

density is increased, indices that quantify competition in

replacement series tend to become less density dependent

(Cousens, 1991). Replacement studies should therefore be








conducted so that constant final yields are achieved

(Radosevich, 1987). This can be accomplished by the

selection of a suitable combination of density and

experiment duration.

In order to obtain constant final yield of lettuce

during early growth, densities substantially higher than

what would be practical in culture of the crop would be

required (suggested by data of Paul and Ayres, 1987). A

modified application of the replacement series which enables

assessment of competition over a range of densities was

proposed by Jolliffe et al. (1984). It was applied to

studies on competition between groundsel (Senecio vulgaris

L.) and lettuce (Paul and Ayres, 1987). Using this

technique, intraspecific and interspecific competition

effects can be quantified (Jolliffe et al., 1984; Roush et

al., 1989). Experiments in this study were therefore

designed so that competition between lettuce and spiny

amaranth grown under varied P regimes could be assessed

using both replacement series analysis techniques (Conolly,

1986; DeWit and Van Den Bergh, 1965; Rejmanek et al., 1989)

and the techniques of Jolliffe et al. (1984).


Materials and Methods


Preliminary Studies


Studies were conducted at Gainesville, Florida to

determine how the growth of lettuce 'Southbay' and spiny






87

amaranth would respond to amendment of soil with phosphorus

fertilizer when grown in monoculture under greenhouse

conditions. The purpose of these studies was to establish P

fertility rates to be used in the competition studies. Soil

used was a Pahokee muck {Euic hyperthermic Lithic

Medisaprists (Soil Survey Staff, 1978)} which was obtained

at the Everglades Research and Education Center (EREC) at

Belle Glade, Florida. Soil used in these studies was of low

P status (soil test value of 1-3 using the water extractable

P analysis) based on EREC soil test procedures (Sanchez,

1990) and had pH values which ranged from 6 to 6.6. Three

experiments were conducted in which Ca(H2PO4)2 (CP) was

applied to soils at the rates indicated in Table 4.1. Prior

to CP addition, soil was fumigated in steel drums with

methyl bromide, allowed to aerate for at least 3 weeks, and

then sieved through a 1 cm mesh galvanized screen. Soil was

amended by measuring the amounts of soil and CP to be used,

mixing the CP thoroughly with approximately 300 ml of the

soil, and then mixing the initial mixture with the remainder

of the soil. Pots in which plants were grown were plastic

containers, measuring 12 cm in diameter and 21 cm deep and

having 4 holes of 0.5 cm diameter in the bottom for

drainage. A 1.5 cm layer of Perlite was placed in the

bottom of the pots. All of the pots of a given CP rate (for

a given experiment) were placed side by side and filled

simultaneously with soil to insure homogeneity. After the








Table 4.1 Calcium phosphate soil amendment rates, planting
densities and species used in the preliminary experiments.


Calcium Species'
Phosphate Densities
Experiment Rate" (plants pot"') Lettuce Amaranth



Preliminary 1 0, 2, 4 4, 8, 16 + +
and 8 and 32

Preliminary 2 0, 4, 8, 4 + +
12 and 16

Preliminary 3 0, 8, 12, 4 +
16, 20
and 24


aCalcium phosphate rates are equivalent to 0.18 g calcium
phosphate per L soil for a rate of 1.

bEach species planted separately. Plus (+) sign indicates
that the species was used.






89

pots were filled a straight edge was used to level the soil

surface. Pots were then firmly tapped on the floor 4 times

to settle the soil. Using the same procedure as in the

initial filling, additional soil was added to again fill the

pots. The levelling and tapping procedures were then

repeated. In final preparation for planting, 200-300 ml of

deionized water was slowly added to moisten the soil. In

experiment 1 plantings were established at the densities

given in Table 4.1. Preliminaries 2 and 3 were limited to a

single density (Table 4.1). Preliminaries 1 and 2 consisted

of separate sets of treatments for each of lettuce and spiny

amaranth while only lettuce was used in preliminary three

(Table 4.1). Seed of lettuce and spiny amaranth (collected

at the EREC) were sown so that an even spacing between seeds

was maintained. Seed placement was confined to the region

between 1 and 4 cm away from the pot wall. To insure a

complete stand, seed were sown in excess and thinned after

emergence. Seed were then covered with a 1 cm layer of

vermiculite. Deionized water was applied to the surface as

needed in order to keep the vermiculite moist until seedling

emergence was complete. At 5 to 10 days after emergence,

plants were thinned to achieve the final density. Once

densities were established, soil was kept moist by adding

nutrient solution containing the micro and macro nutrients

needed for plant growth except P (Ross, 1974). These were

applied 1 to 3 times weekly. The concentration of Mn was








doubled to insure against deficiency (C. A. Sanchez,

personal communication). As plant growth progressed and pot

weight differences became evident, container weight

differences due to differential moisture loss were

compensated for by adding deionized water.

Once densities were established, and before foliage

extended beyond the confines of the pot walls, screen

cylinders were placed around the pots to confine the

foliage. Cylinders used in Preliminary 1 were constructed

of window screen. Otherwise cylinders were constructed of

hexagonal wire mesh with 2.5 cm openings. These extended to

30 cm above the pot surface.

Preliminary 1 was conducted in a fiberglass greenhouse

during February and March of 1991. Preliminaries 2 and 3

were conducted in an air-conditioned glasshouse during June

and July of 1991. Preliminaries 1, 2 and 3 were harvested

25, 24 and 27 days, respectively, after 50% seedling

emergence had occurred. Shoots were excised at the soil

surface, counted, dried at 60 C and weighed.


Competition studies.


Three experiments were conducted to determine the

effect of P fertility on competition between lettuce and

spiny amaranth during the first 4 weeks of growth following

seedling emergence. The studies were conducted in air-

conditioned glasshouses, the first during August and








September of 1991 at Gainesville, Florida, the second in

March of 1992 (spring) at Belle Glade, Florida and the third

in September of 1992 (fall) at Belle Glade. Plant culture

methods were essentially the same as those described for the

preliminary studies. The density, composition and fertility

treatments used in these experiments are given in Table 4.2.

In the Gainesville experiment, in addition to the treatments

listed in Table 4.2, there were also monocultures of 4

plants per pot for which no CP was added to the soil. This

treatment was included for each species in order to show the

overall response to CP under the conditions of the study.

The Belle Glade experiments differed from the Gainesville

one in that a density of 32 plants was included and that the

CP rates were modified (Table 4.2). For the spring Belle

Glade experiment, in addition to the CP rates listed in

Table 4.2, there were additional CP rates of 0, 10.5, 20.5

and 25.5. The only plant densities and compositions grown

at these additional CP rates were monocultures of 4 plants

per pot. These treatments were included for each species in

order to show the overall response to CP under the

conditions of the study. Each experiment was conducted as a

randomized complete block design with 3 replications.

Harvest involved excising shoots just above the soil

surface. Excised shoots were counted and then dried and

stored at 60 C. Dry weights were obtained and entire shoots

were analyzed for mineral nutrients and N. Weed and lettuce








Table 4.2 Calcium phosphate soil amendment rates and
planting densities used in the competition experiments.


Calcium Densities
Phosphate
Experiment Ratea monocultureb mixture

plants pot -
Gainesville 1.5, 7.5 1, 2, 4, 8 2, 4, 8
and 13.5 and 16 and 16

Belle Glade 0.5, 5.5 1, 2, 4, 8, 2, 4, 8,
- spring and 15.5 16 and 32 16 and 32

Belle Glade 0.5, 5.5 1, 2, 4, 8, 2, 4, 8,
- fall and 15.5 16 and 32 16 and 32


aCalcium phosphate rates are equivalent to 0.18 g calcium
phosphate per L soil for a rate of 1.

bMonocultures established for each of spiny amaranth and
lettuce.

cAll mixtures at a 1:1 ratio of spiny amaranth and lettuce.








samples to be used for nutrient analysis were ground in a

stainless steel outfitted laboratory mill to pass through a

1 mm mesh sieve. Samples were thoroughly mixed following

grinding and a subsample of the ground material was used for

analysis. Subsamples were wet ashed as described by Wolf

(1982). Nitrogen was determined by a micro-Kjeldahl method

(Bremner and Mulvaney, 1982), P was determined

colormetrically and K, Ca, Mg, Zn, Fe, Mn and Cu by atomic

absorption spectrophotometry.


Analysis of competition.


Statistical analysis were performed using General

Linear Models and Regression procedures (Freund et al.,

1986). Per plant dry weight data were analyzed using

analysis of variance with a complete factorial treatment

arrangement. Main factors were CP rate, total density,

composition (mixture or monoculture) and species. Data for

monocultures of a density of one plant per pot were excluded

from this analysis.

Relative yield of a species grown in a 1:1 mixture is

defined as the yield of the species in mixture divided by

the yield obtained in monoculture of the same species (DeWit

and Van Den Bergh, 1965). Relative yields were calculated

for each species, P rate and density combination. These

data were analyzed as a complete factorial arrangement with

the main factors being P rate, density and species.








A relative crowding coefficient was determined using

the techniques of Rejmanek et al. (1989) and was used as an

indicator of the relative aggressiveness of one species

versus another. It is defined as:

RCC12 = Wlt/W2t/W1P/W2P

The two component species of the mixture are identified as 1

and 2. Mean yields per plant for the plants grown in

mixture are Wit and W2t. Mean yields per plant for

monocultures are given by WI, and W2p.

Techniques of Jolliffe et al. (1984) were used to

calculate relative monoculture responses (RMR) and relative

mixture responses (RXR) at each of the CP rates for the

Gainesville and the spring Belle Glade experiments.

Relative monoculture responses were calculated from the

equation:

RMR = (Yp Y,) / Yp

The yield obtained in monoculture is represented by Y,. The

Yp is a hypothetical predicted yield which would result if

no competition occurred (Jolliffe et al., 1984). It is

derived from the initial slope of the relationship between

yield and density. In this study Yp was calculated from the

relationship given by Roush et al. (1989):

Yp = (Ymx / Kn) (N)
Where Ymx is the constant final yield, Kn is the density at

which one half of Ymx is achieved and N is the plant density