IMPROVING DUCK MARSHES

■•^WEED CQNIRQL

Marine Biological Laboratory

LIBRARY

^P^:n 1958

WOODS HOLE, MASS.

CIRCULAR 19 -REVISED

FISH AND WILDLIFE SERVICE
UNITED STATES DEPARTMENT OF THE INTERIOR

IMPROVING DUCK MARSHES

■•^WEED CONTROL

By ALEX C. AAARTIN, RAY C. ERICKSON,
and JOHN H. STEENIS, V/ildlife Biologists
Bureau of Sport Fisheries and Wildlife

CIRCULAR 19 -REVISED

FISH AND WILDLIFE SERVICE, Arnie J. Suomela, Commissioner

UNITED STATES DEPARTMENT OF THE INTERIOR

Fred A. Seaton, Secretary

UNITED STATES GOVERNMENT PRINTING OFFICE • WASHINGTON : 1957

For sale by the Superintendent of Documents, United States Government Printing Office
Washington 25, D. C. : Price 25 cents

CONTENTS

Page

Introduction 1

Marsh weeds defined 1

Acknowledgments 2

General suggestions 3

Preventive planning 3

Prompt action 3

Selecting the best method 4

Best season for treatment 4

Replacement by useful plants 5

Methods... 6

Control by water-level management 6

Herbicidal control 7

Dangers 7

2,4-D and 2,4,5-T 8

Other herbicides 8

When and how to spray 9

Spraying equipment 10

Judging results 10

Calculations 11

Mechanical control 14

Burning 15

Biological control.. -- 16

Marsh weeds 17

AUigatorweed (Allernanthera phUoxeroides) 17

Arrow-arum {Peltandra virginica) 20

Cattails {Typha). 21

Cocklebur (Xanthium) 24

Giant cutgrass {Zizaniopsis miliacea) 25

Hibiscus (Hibiscus) 27

Lotus {Nelumho lutea)... 29

Maidencane {Panicum hemitomon) 31

Needlerush {Juncus roemerianus) 32

Phragmites {Phragmites communis) 35

Pickerelweed {Pontederia cordata) 39

Spatterdock {{Nuphar luteum) 40

Waterchestnut [Trapa nutans) 42

Waterhyacinth (Eichhornia crassipes) 46

Waterlettuce {Pislia stratiotes) 49

Waterlihes (Nymphaea) 50

Waterprimroses {Jussiaea) 51

Woody weeds 52

Alders {Alnm) 54

Buttonbush {Cephalanihus occidenlalis) 55

Saltcedar {Tamarix pentandra) 56

Willows (Salix) 58

Other woody weeds 60

in

FOREWORD

Killing the goose that laid the golden eggs has its modern
counterpart. Today we are doing it a different way. As
advancing civilization "reclaims" marshes and swamps
throughout the country, habitat vital to the existence of wild
ducks and geese is being destroyed.

During the past decade, efforts have been concentrated on
accurate inventories of the continental waterfowl popula-
tion. The time has come to put equal emphasis on habitat
needed by the birds. Only by a coordinated, nationwide pro-
gram of habitat management and improvement, including
control of unwanted plants, can there be a continuing supply
of waterfowl. And only thus can our waterfowl heritage
persist as an important recreational resource for millions of
Americans.

IMPROVING DUCK MARSHES

''^WEED CONTROL

Two factors explain why there is growing interest in the control
of marsh weeds. One is the advent of new and better herbicides
for the purpose. The other is the increasingly critical situation
facing the nation's duck-hunting resource — on which 21/^ million
Americans spend over a hundred million dollars annually. Already
more than half of the country's original 125 million acres of wet-
lands have been spoiled for waterfowl use. As national census
figures continue climbing toward the predicted 200-million mark,
more and more of the places where ducks feed, breed, and are
hunted will have to be converted to the needs of advancing civiliza-
tion. In other words, prospects are for fewer hunting places and
lesser numbers of ducks for larger populations of hunters.

To help offset this trend, it is important to make the best use of
what is left. Thousands of poor or fair areas in the United States
can be made more useful for ducks and duck hunters by replacing
marsh weeds with plants that furnish food or cover for waterfowl.
Commonly, the costs involved are not prohibitive.

In Florida recently, 20,000 acres of waterhyacinths and other
pest plants were cleared away from ponds and lakes by the Game
and Fresh Water Fish Commission. Numerous other programs,
both large and small, are being waged against phragmites, cattails,
waterchestnut, and other marsh weeds in various States. On the
Federal refuge system, more than 2,000 acres of marshlands are
treated with herbicides annually to make them more productive for
waterfowl. Yet all this is small compared with what can be done.

MARSH WEEDS DEFINED

Marsh plants are either assets or liabilities. Those that do not
benefit waterfowl do the opposite by competing with desirable vege-
tation and lowering usefulness of the habitat. Growths that do
not justify the space they occupy in terms of food or cover for
waterfowl are regarded as weeds.

Species designated as weeds are not necessarily such in all places
and times. Usefulness or lack of usefulness of a particular kind

2 INTRODUCTION

of plant varies considerably, depending on local circumstances, i
Bulrushes, though generally desirable, become so dense and domi-
nant in some localities that they need to be reduced. In the South
and some other regions, cattails compete too successfully with desir-
able plants in duck marshes, yet in the Northwest cattail marshes
furnish useful nesting cover for redheads, ruddies, and other
waterfowl. Similarly, though phragmites and needlerush have no
food value for waterfowl they provide useful protective cover in
some flooded feeding sites. In the latter type of situation, weed
control can consist of opening up dense stands of the plants instead
of trying to eliminate them.

Some weedy species are practically nationwide in distribution,
whereas others have more restricted range. Cattails, phragmites,
spatterdock, and willows are examples of wide-ranging marsh
weeds. Waterchestnut is at present a problem only in the North-
east; saltcedar is confined mainly to the Southwest; and several
troublesome plants such as alligatorweed, giant cutgrass, maiden-
cane, needlerush, sawgrass, waterhyacinth, and waterlettuce are
limited largely or entirely to the South.

The control of floating or floating-leaf plants such as water-
hyacinth, waterlilies, and spatterdock is discussed in this circular,
but submerged aquatics are excluded because their suppression is
primarily the concern of fishpond management and a fishery leaflet^
on the subject is available. Objectionable trees and shrubs of
waterfowl habitat are discussed in the final chapter on Woody
Weeds.

ACKNOWLEDGMENTS

Information in this booklet has been compiled from many sources.
Much of it has come from studies or operations by the Branches of
Wildlife Research and Wildlife Refuges of the Bureau of Sport
Fisheries and Wildlife, U. S. Fish and Wildlife Service, cited in the
text as FWS, and much has been drawn from tests conducted by
State Federal Aid programs and by the Agricultural Research
Service in its cooperative studies with the Bureau of Reclamation.
Plant names, both common and scientific, have been based largely
on Neil Hotchkiss's Checklist of Marsh and Aquatic Plants of the
United States, FWS Wildlife Leaflet 210, 1950.

> Information on duck-food usefulness of various marsh and aquatic plants, as well as on their distribu-
tion, propagation, and environmental requirements, is given in Food of Game Ducks in the United States
and Canada, by A. C. Martin and F. M. Uhler, FLsh and Wildlife Service Research Report 30, 1951.

» Control of Aquatic Plants in Ponds and Lakes. Eugene W. Surber. FWS Fishery Leaflet 344, 1949.

GENERAL SUGGESTIONS

PREVENTIVE PLANNING

In creating new ponds, it is frequently possible to prevent seri-
ous weed-control problems by proper planning in the blueprint
stage. An especially important step in this direction is the provi-
sion for draining and reflooding of pond basins. Such arrange-
ments, as discussed in the section on Control By Water-Level Man-
agement, page 6, help avoid weed infestations or enable effective
elimination of them. One of the advantages is the fact that pest
plants can be plowed up or destroyed by other means while the
pond is drained.

Another construction feature that helps minimize weed difficul-
ties is the subdivision of large impoundments by means of dikes.
This makes possible more efficient management of water supplies,
both for weed control and for food production, especially when
there is provision for independent flooding and drainage of each
pool. Building the dikes wide enough to serve as roads permits
ready access for inspections, weed control, and other operations.

PROMPT ACTION

In weed control, an ounce of prevention is generally worth tons
of cure. Heading off serious infestations before they become well
established prevents deterioration of habitat and costly control
work later on. If the small colonies of waterchestnut found near
Alexandria, Va., in the early twenties had been eliminated at that
time, duck hunting, fishing, and other recreation on the Potomac
River would not have been ruined for a couple of decades, and
hundreds of thousands of dollars eventually spent on control of
the plant could have been saved. The principle of timely preven-
tion holds true for almost any marsh or pond. Destroying scat-
tered patches of invading weeds before they have a chance to
spread and become firmly established is the most efficient way of
controlling weeds. There are two important prerequisites, how-
ever: Ability to recognize the objectionable plants, and periodic
or seasonal inspections to spot early stages of infestation.

4 GENERAL SUGGESTIONS

SELECTING THE BEST METHOD

Local conditions must help decide which is the best method for
use on a particular weed problem. The kind of plant or plants,
extent of infestation, depth and controllability of water, accessi-
bility, and similar factors, need to be weighed against costs and
merits of the various methods that can be used.

If local water supplies and water-control structures are suffi-
ciently adaptable, adjustment of water levels often controls weeds
and results in an abundance of desirable plants. Disking, mowing,
or crushing provide satisfactory control of many weeds in marshes
that can support mechanical equipment. Pulling by hand, mowing
with a sickle, or spot treatment with herbicides are often practical
for elimination of small patches of weeds. Combinations of meth-
ods sometimes prove particularly effective. These include mow-
ing followed by flooding, burning and spraying, spraying and flood-
ing, or use of two different chemicals. More detailed comments
on various control procedures are given in the chapter on Methods
(page 6).

BEST SEASON FOR TREATMENT

Most marsh plants have a particular growth stage during which
control measures are likely to be especially effective — ^the vulner-
able stage. Frequently, though not always, the ideal time for treat-
ment is associated with the plant's flowering or fruiting, probably
because of depleted food reserves. Needlerush appears most sus-
ceptible in the spring during its flowering. Mowing of cattails at
two particular stages of growth yields a 95% to 100% kill, whereas
numerous mowings at various times may result in little or no real
control (see page 22).

In some plants, the ideal time for treatment shows no clear rela-
tion to the flowering or fruiting stages. Woody plants commonly
respond best when control measures are applied in later summer or
fall ; giant cutgrass is killed most readily about two months after
its flowering, during the period of maximum runner growth. Much
remains to be learned about the ideal time for treatment of different
species by different methods, and as more information is obtained
on these matters control programs will improve in efficiency and
economy.

GENERAL SUGGESTIONS 5

REPLACEMENT BY USEFUL PLANTS

The success or failure of control measures depends not only on
the extent to which weeds are eliminated but also on the kind, qual-
ity, quantity, and duration of the new plants that come in their
place. Where useful plants are already present in a suppressed
state, removal of weed competition is frequently all that is neces-
sary to ensure satisfactory improvement. Thus, in a Maryland
locality dominated by cattails, control by mowing resulted in good
volunteer growths of wild millet, wildrice, and rice cutgrass for
about a decade. However, satisfactory replacement cannot always
be left to nature. Sometimes, seeding or planting is necessary
soon after control operations to assure a good crop of useful plants
and prevent rapid reestablishment of weeds.

METHODS

CONTROL BY WATER-LEVEL AAANAGEMENT

Suppressing weeds by water-level management has been prac-
ticed in many places because of its practical advantages in econ-
omy and simplicity. The method requires sufficient water supply
and a suitable type of dam to maintain desired levels at different
seasons.

Various kinds of marsh weeds can be eliminated by drowning,
either as seedlings or in more mature stages of growth. Flowering
beds of lotus have been killed by a 30- to 36-inch increase in depth
lasting for 12 days. Mowed needlerush was destroyed by an 8-inch
rise in water level held for 3 weeks. Mowed cattails can be con-
trolled more effectively when water covers the cut bases. Most tree
and brush stumps lose their ability to sprout if covered by water
during one growing season. Many other species that are some-
what tolerant of flooding can be drowned if depth and duration are
sufficient.

An especially important water-level measure for improving vege-
tation is the seasonal drawdown. In certain instances, however,
the drawdown can be a liability rather than an asset since more
weeds or worse weeds can become established while water levels
are down. Potentialities of the drawdown for good or harm depend
upon local conditions — particularly upon plants, soils, and weather
of the area.

In places where dark-stained water, excessive acidity, turbidity,
troublesome weeds, carp, or similar factors limit production of
good duck foods, it is often advantageous to drain or partially
drain the pond basin during summer and raise wild or cultivated
seed-producing annuals. Gradual restoration of water levels in
the fall makes abundant food supplies readily available for ducks
and geese. If weedy perennials survive inundation through win-
ter and spring and become a nuisance, it may be advisable to omit
the drawdown in some years so the undesirable plants can be
drowned during continuous high water.

At the Patuxent Refuge, a system of alternating drawdowns on
two adjacent ponds, one pond being lowered one year and the other
the next, has worked particularly well. The plan has not only pre-

HERBICIDAL CONTROL 7

vented serious weed problems but has also provided good nesting
conditions adjacent to excellent food supplies. In the Southeast,
where control of one kind of weed sometimes simply clears the way
for a worse one, wild millet, browntop millet, corn, soybeans, and
other useful species are commonly planted in dewatered marsh
areas that have been cleared of weeds and prepared for seeding
by farm machinery.

Timing of water-level adjustments is especially important in
drawdowns involving production of native (noncultivated) plants.
The more rapid growth of desirable annuals, as compared with
most perennial weed seedlings, frequently makes it possible to
destroy unwanted plants and spare the good ones. At the Thousand-
Acre Marsh in Delaware, thousands of seedlings of phragmites,
hibiscus, swamp dock, pickerelweed, and arrow-arum have been
drowned while permitting survival of young plants of wild millet,
smartweeds, rice cutgrass, and fall panicum. If the timing of
water-level change is wrong, however, more harm than good is
likely to result. Excessive duration of a drawdown risks estab-
lishment of perennial-weed seedlings and serious spread of runner-
producing plants such as phragmites and giant cutgrass.

Temporary drainage has been found advantageous for control
of some weedy marsh and aquatic growths. Draining of beds of
alligatorweed or waterprimrose allows more thorough spraying
and better kill of the plants. Draining also facilitates mechanical
control and enables the burning of old dead stalks of phragmites
or needlerush prior to spraying. In some places, winter drainage
followed by freezing of the marsh floor has helped get rid of unde-
sirable plants. Flooding, following cutting or herbicidal control,
is frequently helpful in eliminating plants such as needlerush, hibis-
cus, and buttonbush.

HERBICIDAL CONTROL

The use of herbicides in marsh management has increased rap-
idly during the past decade, and it will probably continue to grow
as new and better chemicals become available. Several valuable
new herbicides released within the past few years have given fresh
impetus to work in this field.

DANGERS

Spraying marsh weeds with herbicides usually poses no direct
danger to man, livestock, fish, or wildlife. In this regard, plant-
control agents are much safer than other pesticides such as insecti-
cides or rodenticides. Formerly, dangerous substances such as

8 HERBICIDAL CONTROL

sodium arsenite and ammonium thiocyanate were used on marsh
plants to a limited extent, but they have been largely superseded
by herbicides that are safer and fully as effective.

Because gas or vapor drift from some spraying operations
has proved harmful to farm crops, there are strict laws^ on aerial
spraying of herbicides. There is also the possibility of injury to
nearby stands of duck-food plants. Coarse sprays, such as those
released by low-pressure equipment, are much less likely to drift
and cause damage to crop plants than fine sprays produced under
high pressure. Furthermore, herbicides consisting of salts or low-
volatility esters are safer than volatile esters as regards release of
toxic vapors from sprayed areas.

2,4 -D AND 2,4,5-T

Two of the most popular herbicides in marsh management are
the so-called phenoxy compounds, 2,4-D (2,4-dichlorophenoxyacetic
acid) and 2,4,5-T (2,4,5-trichlorophenoxyacetic acid). 2,4-D is
cheaper than 2,4,5-T and is used much more widely, but both are
comparatively low in cost, and both are effective on many kinds of
marsh plants.

Besides their economy and wide effectiveness, formulations of
2,4-D and 2,4,5-T are easy to apply, are somewhat selective in gen-
erally being more toxic to broadleaf plants than grasses, and are
noncorrosive to spraying equipment. The sensitivity of cotton,
tobacco, grapes, tomatoes, and other crops to these chemicals is an
important drawback to the use of herbicides in agricultural areas.

The most commonly used formulations of 2,4-D and 2,4,5-T are
amine salts and esters. Under favorable conditions, the salts and
esters give nearly equivalent results, though generally the former
are slower in showing effects. The salts dissolve readily in water
but not in oil, handle easily, and their sprays do not volatilize so
readily as esters. On the other hand, the fact that esters can be
dissolved in oils is an advantage in low-volume spraying, as in aerial
applications. Furthermore, oils such as kerosene and diesel oil
have some herbicidal value and their use as diluents enables better
residual adherence to floating-leaf aquatics like waterlilies or spat-
terdock and on other plants wet by rain shortly after spraying.

OTHER HERBICIDES

During the past decade, ammonium sulphamate (Ammate) and
TCA (sodium trichloroacetate) have been used to a limited extent

' Such laws are summarized in Herbicides Law Manual, compiled by the National Agi-icultural Chemi-
cal Association, 1145 19th St., N. W., Wasbiiistou 6, D. C.

HERBICIDAL CONTROL 9

in marsh management. Ammate is effective on many marsh and
swamp plants, but it is comparatively expensive. The main use-
fulness of TCA is on grasses, but in combination with 2,4-D it has
proved effective on needlerush. Both Ammate and TCA corrode
spraying equipment. The soil sterilants Borascu and Polybor-
chlorate have been used successfully for local, small-scale eradica-
tion of hard-to-control pests in unflooded sites, but costs are likely
to be prohibitive for extensive treatments with these chemicals.

A number of promising new herbicides have become available
recently. Among these is dalapon (2,2-dichloropropionic acid),
sold under the names Radapon and Dowpon. It has shown superi-
ority to 2,4-D for control of phragmites, maidencane, and certain
other narrow-leaf marsh plants. ATA (3-amino-l,2,4-triazole)
which is sold as Amino Triazole and Weedazol has important possi-
bilities for control of cattails and other plants. It is translocated
readily and inhibits chlorophyll production with the result that new
growth is often albinistic. Erbon (2-(2,4,5-trichlorophenoxy)-
ethyl-2,2-dichloropropionate) has given indication of being valu-
able as a temporary soil sterilant in sites that are not flooded for
two or three months. It is sold under the names Baron and
Novon. Certain formulations of silvex, (2-(2,4,5-trichlorophe-
noxy) -propionic acid) have given encouraging results on alligator-
weed, specifically Kuron and 2,4,5-T Propionic. Additional prom-
ising new herbicides include urea formulations such as monuron,
diuron, and neburon, and certain chlorobenzoic derivatives.

WHEN AND HOW TO SPRAY

In general, herbicides are likely to be most effective under con-
ditions favorable to plant growth — sunny, warm, humid weather.
Spraying in early morning or midmorning usually results in good
penetration before afternoon or evening showers arrive. Water-
soluble herbicides should remain on plants at least 6 hours before
a rain. Spraying is easier, safer, and more effective when air
movement is limited.

Volume used depends on the method of spraying (from land,
boat, or plane), as well as on the height and density of the vegeta-
tion treated. Generally, spraying operations from land or boat
use 50 to 500 gallons per acre. Good coverage without wasteful
runoff requires careful operation with efficient equipment. In
many instances, low-volume spraying with large droplets has
proved as effective as thorough wetting with fine high-pressure
spray. Most commercial herbicides now include wetting agents
that increase adherence of sprays to leaves, but sometimes addition

10 HERBICIDAL CONTROL

of a detergent or "sticker" is needed to assure wetting of water-
repellent foliage. Household detergents and Tergitol 4 are suitable
for this purpose, but only enough should be used to give good adher-
ence. Commonly, the adherence of a spray solution is tested pre-
liminarily to appraise adequacy of wetting agents in it.

SPRAYING EQUIPMENT

Efficient herbicidal control of marsh weeds requires equipment
adapted to the scale or kind of operation involved. For small-,
medium-, or large-scale projects, many makes and types of spray-
ing equipment are available. For small areas and local "spot treat-
ments," single-nozzle hand sprayers are used, either afoot or on
boats. Power sprayers mounted on trucks, tractors, or boats are
employed commonly on projects of medium or large scale. Usually
such spraying outfits have booms with numerous nozzles, but in
some instances a single-nozzle hose is attached. For extensive
tracts, plane application is practical. This method has, among
other merits, the advantage of avoiding difficulties of land or boat
travel in soft muck or in areas of interspersed marsh and water.
A major part of herbicidal operations on Federal waterfowl refuges
now involves airplane spraying.

Air-thrust boats deserve mention here, though their usefulness
is by no means limited to weed-control work. These shallow-draft,
flatbottom craft driven by airplane propellers have proved practical
in weed-control reconnaissance and for spraying in comparatively
inaccessible areas, particularly in marshes that are interlaced with
shallow waterways.

Publications listed below can be consulted for suggestions on
spraying equipment:

1956. Sprayers for Weed Control. William G. Westmoreland and J. C. Ferguson.

North Carolina State Ext. Circ. 403.
1951. Weed Spraying Equipment. G. E. Page. Oregon Agr. Exp. Sta. Bui. 493.
1950. When Buying Weed Sprayers Consider Many Points. R. E. Larson and V. H.

Johnson. Minnesota Farm and Home Sci., 7(2).
1948. Chemical Weed-Control Equipment. Norman B. Akesson and W. A. Harvey.

California Agr. Exp. Sta. Circ. 389.
1948. Spray Equipment for Weed Control. Noel S. Hanson and John Steele. Univ.

Nebraska Ext. Circ. 174.

JUDGING RESULTS

Since most marsh weeds are perennials, success or failure of
efforts to control them cannot be judged reliably until the following
growing season. Furthermore, some herbicides cause inhibition of

HERBICIDAL CONTROL 11

growth well into the second year. Impressive top kills or brown-
outs produced by certain chemicals are frequently followed by vig-
orous regeneration from basal parts, either in the same year or the
next one. Rapid killing of plant tissues is often disadvantageous
by hindering translocation of toxic substances.

Another factor to be considered in judging success of herbicidal
operations is the rate at which treated weeds return to dominance.
Kinds of weeds that reestablish and spread rapidly need to be
eliminated 95% or more, else efforts toward their control will be
nullified in a short time.

CALCULATIONS

An important aid in figuring amounts of chemicals required for
spray solutions and in determining how to adjust spray equipment
to obtain particular dosages per unit area is the Handbook of Weed
Control Calculations, by John T. Maletic, USDI Bureau of Reclama-
tion, 1949. Additional help may be available from the following
formulas and tables, which were prepared with the help of Clark
G. Webster of the Patuxent Refuge.

CALCULATING AMOUNTS OF MATERIALS TO USE

The starting point for ascertaining amounts of ingredients
needed in spray solutions is the herbicide's concentration as re-
corded on the container label. Commonly the concentration is ex-
pressed in pounds of acid equivalent or active ingredients per
gallon and in percentages of acid equivalent or active ingredients
(often referred to as "active"). Having this information, it is
possible to determine the amount of commercial herbicide required
to make a solution with 1 pound of acid equivalent or active ingredi-
ents per 100 gallons or a 1-pound-per-acre application of acid
equivalent or active ingredients (table 1), or to make other specific
solutions. In preparing solutions other than for 1 pound per 100
gallons or per acre, or for other quantities than 100 gallons, cor-
responding fractions or multiples of figures in the tables can be
computed readily. It is desirable to use acid-equivalent figures
when they are available.

USEFUL CONVERSION FIGURES:

1 gallon = 4 quarts = 8 pints = 128 fluid ounrea = 3785 milliliters.

1 pint = 2 oxipa = 16 fluid ounces = 32 teaspoonfuis.

1 tablespoonfui = 3 teaspoonfuls = 15 milliliters.

1 fluid ounce = 29.6 milliliters.

1 pound = 16 ounces = 453.6 grams.

1 acre = 160 square rods = 43,560 square feet.

12 HERBICIDAL CONTROL

POUNDS-PER-ACRE FORMULAS; MAKING SPRAY SOLUTIONS FOR A PARTICULAR NUMBER
OF POUNDS PER ACRE OF ACID EQUIVALENT OR ACTIVE INGREDIENTS:

Using liquid herbicide :

The -pints needed per 100 gallons of solution =

wanted lbs, acid equiv. or act, ingr. per acre X 800
gals, epray per acre X lbs. acid eq. or act. ingr. in 1 gal.

Example: If a 2,4-D formulation containing 4 pounds acid equivalent per
gallon is to be applied at 5 pounds acid equivalent in 120 gal-
lons of solution per acre, the pints needed per 100 gallons of
solution are:

5 X800 ^„ . ^

■ or 8.3 pmts

120 X 4

The milliliters needed per gallon of solution =

wanted lbs, acid equiv. or act, ingred. per acre X 3785

gals, spray per acre X lbs. acid eq. or act. ingr. in 1 gal.

Using solid (dry) herbicide :

The pounds needed per 100 gallons of solution =

wanted lbs, acid equiv. or act, ingred. per acre X 100
gals, spray per acre X % acid eq. or act. ingred.

Example: If a formulation of dalapon containing 71% acid equivalent is
to be sprayed at the rate of 30 pounds acid equivalent in 200
gallons of solution per acre, the pounds of herbicide needed per
100 gallons of solution are:
30 X 100

200 X .71

or 21.1 pounds

POUNDS-PER-100-GALLONS FORMULAS; MAKING SPRAY SOLUTIONS WITH A PARTICULAR
NUMBER OF POUNDS OF ACID EQUIVALENT OR ACTIVE INGREDIENTS PER 100 GALLONS:

Using liquid herbicide :

The pints needed per 100 gallons of solution =

wanted lbs. acid equiv. or act. ingred. per 100 gals. X 8
lbs. acid equiv. or act. ingred. in 1 gal. herbicide
Example: If a 2,4,5-T formulation containing 4 pounds acid equivalent
per gallon is used to prepare a solution with 5 pounds acid
equivalent per 100 gallons, the pints needed per 100 gallons are:

5X8 _ .
— : — or 10 pmts
4

Using solid (dry) herbicide :

The pounds needed per 100 gallons of solution =

wanted lbs. acid equiv. or act. ingred. per 100 gals.

% acid equiv. or act. ingred. in the herbicide

Example: If an Ammate formulation containing 95% active ingredients

is used to prei are a solution with 60 pounds active ingredients

per 100 gallons of solution, the pounds per 100 gallons are:

60

or 63.2 poimds

0.95

HERBICIDAL CONTROL 13

Table 1. — Amounts of herbicide needed for certain solutions

To make a solution for a 1-pound-per-acre application of acid
equivalent or active ingredients :

When using a formulation with one of
the following strengths (representa-
tive of important herbicides) :

Pounds"of acid
equiv. or act.
ingr. per gal.

4.._.
3.34.

Percent

Acid
eq.

Act.
ingr.

£-f

'Then the amount of commercial her-
bicide needed for one acre in:

I

Liquid measure:

Dry measure:

Fluid
ounces

Pints

Ounces

Pounds

32

2

2.3G

38

22
20
32
16.8

1 41

1 26

2

1.05

To make a solution with 1 pound of acid equivalent or
active ingredients per 100 gallons or per 1 gallon :

When using a formulation with one of)
the following strengths (representa-
tive of important herbicides) :

Percent of — f

Pounds of acid
equiv. or act.
ingr. per gal.

Acid
eq.

Act. /
ingr.

4 .-

3.34.

71

79.3

'"56"'
95

/

/Then the amount of comme
bicide needed for:

if %

rcial her-

100 gallons:

1 gallon:

^ j Pints

Pounds

Fluid
ounces

Ounces

2

0.32
0.38

2.38

1.41
1.26
2.
1.05

0 22

0 20

0 32

0 17

\

14 MECHANICAL CONTROL

CALCULATING SPRAY APPLICATIONS

r Based largely on Suggested Guide for Chemical Control of Weeds, U. S. Agr. Res. Service Special"!
I Report 22-23, April 1956. Useful information on calculations and procedures in aerial spraying is I
Lgiven in How to Spray the Aircraft Way, USDA Farm. Bui. 2062, June 1954. J

The amount of herbicide delivered to a sample unit or area can be
determined preliminarily by a test run using plain water in the
spray tank. The method is applicable to various types of equip-
ment under particular conditions of nozzle adjustment, tank pres-
sure, and rate of progression. When a known acreage, fraction of
acreage, or other unit of area has been sprayed during a test run,
the amount of water used can be ascertained by measuring the
quantity required to refill the tank. In making calculations for
different operations, rate of progression or nozzle adjustment may
need to be altered if plants to be sprayed are of different height
or density.

The number of gallons per acre applied by a power sprayer, at a
given setting of nozzles, tank pressure, and rate of progression, can
be determined after a 220-yard test run as follows: Multiply the
number of gallons used by 66 and divide the product by the number
of feet in the sprayed strip width. Similarly, for a 110-yard test,
multiply the gallons by 132 and divide by the number of feet in the
sprayed strip width.

The number of pints per acre applied by a full 3-gallon hand
sprayer can be calculated by testing on a 10 x 10 foot (100 square
feet) area as follows : 1/2 pi^^t applied to 100 square feet equals 27
gallons per acre, 1 pint equals 55 gallons per acre, and one quart
equals 110 gallons per acre. Accordingly, if one wishes to apply
1 pint of commercial weed killer per acre, it should be mixed with :
27 gallons of diluent if the sprayer applies 1/2 piiit to 100 square
feet, 55 gallons of diluent if sprayer applies 1 pint to 100 square
feet, and 110 gallons of diluent if the sprayer applies 1 quart to
100 square feet.

MECHANICAL CONTROL

Besides mechanical equipment used in spraying, machinery of
several types is of practical value in controlling weeds in marshes
and swamps and in converting such areas into cultivated fields or
marshes attractive to waterfowl. Equipment for such purposes
includes bulldozers for felling woody plants, rotary brush cutters,
heavy harrows, disks, and rototillers for conditioning soil that con-
tains masses of tough roots, and mowers or crushers for control
of coarse marsh weeds. A principal limiting factor to wider use
of such equipment is its inability to function well on soft, boggy

BURNING 15

land or in places that include combinations of marshland and
water. So-called marsh buggies with their enormous, expensive
tires are only a partial answer to the need. If in the future a prac-
tical land-water vehicle of the "rollagon," "rhino," "weasel," or
marsh-buggy type comes into commercial production, there may
then be more extensive use of mechanical equipment for weed
control as well as for other marsh-development programs.

Heavy machinery of the bulldozer, disk, and harrow type is
being used on many Federal and State waterfowl refuges to pro-
duce planted food crops for the birds in marshy or swampy areas.
This appears to be one of the most practical ways of combating
persistent marsh weeds such as are particularly prevalent in the
Southeast.

Cutting of weeds with hay mowers and rotary mowers is feas-
ible on marshes that are not too wet. However, such mowing
usually needs to be timed at particular stages of growth if control
of the species is to be more lasting than the mowing of grass in a
lawn. Generally, crushing by special rollers or by tractors with
wide cleats is more likely to be effective in controlling marsh plants
than is cutting. Crushers devised and used by mosquito-extermi-
nation organizations in certain New Jersey counties have been
operated successfully for marsh management for many years.

BURNING

By itself, fire is generally an ineffective control for marsh weeds.
Indeed, it sometimes favors undesirable species by removing ex-
cessive shading or other handicaps contributed by their own dead
stalks. However, fire can be an important adjunct to other means
of control, particularly in herbicidal treatment of dense growths
such as needlerush and phragmites. Preliminary burning in late
winter or spring enables more efficient use of herbicides by avoid-
ing wasting of the spray on dead stems. Burning several months
after treatment can also be beneficial by helping to eliminate weeds
that survive spraying and by opening up growing space for desir-
able plants. Wilson 1 found that burning of needlerush in winter
in North Carolina, following an apparently unsuccessful herbicidal
treatment earlier in the year, resulted in complete kill of the plants.
Myers^ working with the same species in Florida not only demon-
strated the value of preliminary burning to increased efficiency of
control by mowing or by herbicides but also demonstrated that

' See references under Needlerush on page 34.

16 BIOLOGICAL CONTROL

burning after treatment was beneficial provided it was delayed
until fall. Burning preliminary to herbicidal treatment of re-
growth is practiced in California on dense growth such as phrag-
mites and cattails. In tidal marshes, burning at intervals of 1 to
3 years has been found beneficial in controlling bushy growths.

BIOLOGICAL CONTROL

The only instances of man-arranged biological control in marshes
occur where muskrats, nutria, or livestock have been allowed to
reduce undesirable stands of plants. Cattail is plagued by stem
borers, lotus by leaf beetles, and other weeds by their special para-
sites, but at present man has not developed methods of capitalizing
on the usefulness of these minute control agents for marsh man-
agement.

MARSH WEEDS

Under the heading Marsh Weeds, 17 of the more important herba-
ceous pests of waterfowl habitat are discussed alphabetically by
their common names. Besides species that are typical marsh plants,
waterlilies, waterhyacinth, and other aquatics having leaves at or
above the water surface are included since their dense stands are
often implied in the designation "marsh." Their foliage is sprayed
with herbicides in much the same way as other marsh plants, and
in this regard they are distinct from submerged aquatics. Swamp-
inhabiting trees or shrubs are discussed under Woody Weeds on
page 52.

ALLIGATORWEED

BOTANICAL

The genus Altemanthera (also listed as Achyranthes) is largely
tropical but includes half a dozen species in the Southeast. Only
one member of the group, alligatorweed, (A. philoxeroides) also
called alligatorgrass or pigweed, is a serious pest in this country.
Besides having wide distribution in Central and South America,
it occurs along our lower coastal plain from Texas to North Caro-
lina. An infestation that appeared in the TVA area about 1933
has persisted and spread. In 1955, beds of alligatorweed were
found at Moyock, N. C, near Currituck Sound and the Virginia
line, the northernmost point of infestation thus far. Infestations
in Los Angeles, Calif., have recently been reported.

17

18 ALLIGATORWEED

It reproduces and spreads mainly vegetatively ; seeds are not
known to be produced in this country. Broken-off branch frag-
ments root readily, and stems may elongate as much as 200 inches
in one season. Alligatorweed grows rooted in mud but often ex-
tends considerable distances over water surfaces. It is tolerant
of flooding and, though mainly a fresh-water plant, endures mild
salinity.

IMPORTANCE

Alligatorweed is eaten some by baldpates, gadwalls, coots, musk-
rats, and nutrias, and to a limited extent by hogs and cattle, but it
has no value that compensates appreciably for its detriment It is
one of the worst pest plants of the Southeast, obstructing naviga-
tion in canals, bayous, and lakes, destroying beds of duck foods,
ruining fishing areas, and causing a serious malaria hazard in
some localities.

CONTROL

During the past quarter of a century, effective control of alliga-
torweed has remained an unsolved problem. Various methods
and means tried have proved costly and unsatisfactory. Recently,
however, certain new herbicides have given indication of useful-
ness for this purpose. Sterilants such as erbon have shown prom-
ise in unflooded sites, and several new herbicides, both alone and
in combinations, have given encouraging results on alligatorweed
in flooded places.

Among sites or conditions in which control is especially difficult
are continuously flooded waterways, growths partially buried under
silt, and heavily matted beds. Alligatorweed on unflooded flats or
on dikes can be killed somewhat more readily. Because of the
plant's ability to reinvade rapidly, control needs to be complete or
nearly so.

Crushing, cutting, and mechanical removal of beds, as was done
extensively in Louisiana formerly, provided expensive and only
temporary opening of waterways that were clogged by alligator-
weed. Though subsequent use of 2,4-D for this purpose reduced
operational costs, it too contributed comparatively little toward
extermination or enduring control of the plant in extensive areas.
The cooperative study conducted from 1948 to 1950 by the Boyce
Thompson Institute, the Corps of Engineers, and Tulane Univer-
sity reported that 2,4-D (preferably the amine salt) applied at 8
pounds per acre during the season of less active growth, from July
tc the following spring, gave nearly complete control of beds extend-
ing over water in experimental pools. However, general experi-
ence elsewhere with 2,4-D treatment has been discouraging. Usu-
ally, only temporary setback has resulted.

ALLIGATORWEED 19

During the past few years, several new chemicals have been
tested on alligatorweed at waterfowl refuges in the Southeast, and
cooperative studies by the State of North Carolina and the U. S. Fish
and Wildlife Service have tried numerous new compounds, both
alone and in various mixtures, on more than 100 experimental
plots in that State. Some of the more promising treatments are
indicated below under subheadings that denote the type of site.

IN FLOODED SITES

Treatments with the Weedone 2,4, 5-T Propionic form of silvex
at 40 pounds acid equivalent and neburon at 32 active per acre
have given good results. Mixtures of chemicals also have been
encouraging in some instances, the combinations proving to be
more effective than twice the amount of either component used
separately. Following are some of the more promising mixtures
with indication of poundage per acre of acid equivalent or active
ingredients: neburon at 16 with the Kuron form of silvex at 20;
TBA mixture at 6 with silvex at 20; dalapon at 33 with silvex
at 20.

IN UNFLOODED OR TEMPORARILY DRAINED SITES

Erbon has given effective control in North Carolina and else-
where with dosages of 40 pounds active per acre, and in some cases
as low as 20 pounds active, providing flooding does not occur for
2 months after treatment. Best results were obtained by treating
during the period of active growth. The sterilant effect of erbon
at these dosages disappears largely or entirely within 3 months.

ATA at 10 pounds applied at different stations during maximum
flowering has given inconsistent results. At 2 or 3 tons per acre,
Polybor-chlorate, Borascu, or sodium chlorate are capable of elimi-
nating alligatorweed in moist or comparatively dry places.

REFERENCES

1956. Portrait of a Killer. Jack Dermid. Wildlife in North Carolina. 20 (5).
1953. The use of 2,4-D in the Control of Water Hyacinth and Alligator Weed in the

Mississippi Delta, with Certain Ecological ImpHcations. Willis A. Eggler.

Ecology, 34 (2).
1952. Engineer Uses of Chemical Herbicides. Walter A. Carey. The Military

Engineer, 44 (299).
1950. Growth and Reproduction of Water Hyacinth and Alligator Weed and Their

Control by Means of 2,4-D. A. E. Hitchcock, P. W. Zimmerman, Henry

Kirkpatrick, Jr., and T. T. Earle. Contrib. Boyce Thompson Inst., 16 (3).
1940. The Biology of Achyranthes philoxeroides (Mart.) Standley. William T.

Penfound. American Midi. Nat., 24 (1).

20

ARROW-ARUM

BOTANICAL

Arrow-arum (Peltandra virginica) is widely distributed in the
East and is particularly plentiful in fresher parts of tidewater sec-
tions of streams and bays along the Atlantic Coast. It is a coarse
perennial that spreads mainly by means of its fleshy-coated seed-
like berries. Its large arrow-shaped leaves can be distinguished
from those of arrowheads and pickerelweed by the fact that the
latter two have all their main veins radiating from one point near
the base of the blade whereas in arrow-arum, several main veins
branch out from the midrib. Another species of arrow-arum,
P. glauca, having a white spathe and red berries, is restricted to
the South.

IMPORTANCE

The large seedlike berries are eaten by wood ducks, and this
may explain the local name duck-corn. In many places it is a
weed, using space that could be occupied by more desirable plants.

CONTROL

Complete spraying of foliage with 2,4-D (either salt or ester) at
a concentration of 6 pounds per acre during the flowering and early
fruiting period has given efl^ective control in Delaware. In that
State, the most susceptible stage occurs in the first half of June.

REFERENCE

1954. The Marshes of Delaware, Their Improvement and Presei-vation (p. 34).
John H. Steenis and others. Del. Board of Game and Fish Comm.

domingensis

T. glauca

BOTANICAL

This worldwide genus (Typha) has four species in the United
States, as mapped and named above. Cattails, also called flags or
tules, often dominate extensive marsh areas. They spread rap-
idly by means of rootstocks and also by means of minute, airborne
seeds which germinate readily on mud flats and occasionally in
shallow, clear water. The seeds can remain viable for at least
5 years.

IMPORTANCE

Cattails are excellent food and housebuilding material for musk-
rats, are eaten by nutria, serve as food to a limited extent for
ducks and geese in some localities, and provide nesting sites for
several species of waterfowl and rails ; the leaves have commercial
value for mats, chair seating, thatching, and barrel calking; and
the fluff from heads has been used as a substitute for kapok. On
the debit side, cattails are among the most widespread and abun-
dant weeds of waterfowl habitat and are serious pests on western
irrigation ditches.

22 CATTAILS

CONTROL

Cattail stands can be controlled effectively by herbicides, mow-
ing, or other mechanical means. The species of cattail and the
local habitat conditions commonly have a bearing on effectiveness
of treatments. In general, the optimum stage for treatment is
about the time flowers of the staminate spike are drying. On lim-
ited areas, invading seedlings can be eliminated effectively by
hand pulling.

New herbicides, particularly ATA and dalapon, have largely
supplanted 2,4-D in recent control operations. Though 2,4-D was
used widely for this purpose formerly, it was not particularly satis-
factory since a number of retreatments have usually been necessary.
Tests with the new herbicides will need to be continued for several
more years before positive conclusions can be made about the best
method of treatment for the different species in different environ-
mental conditions.

ATA has given uniformly good to excellent results in California
(Citrus Experiment Station, Riverside) and Maryland (FWS
Patuxent Research Refuge, Laurel) at rates of 5 to 10 pounds active
ingredient per acre. Dalapon at 20 pounds acid equivalent has
given complete control in Michigan tests (Grigsby et al., 1955)
and at the St. Marks Refuge in Florida, but in a Maryland site
periodically flooded by tidewater 40 pounds failed to kill all of the
plants. At present it appears that dalapon is much more likely
to be effective on comparatively dry sites than in flooded habitats.
In southern California, two dalapon treatments (May and Septem-
ber) at 10 pounds each, followed by midwinter burning of dead
stems, failed to prevent considerable regrowth of cattails. Mix-
tures of ATA (50% commercial formulation) and dalapon (S5%
commercial formulation) at respective approximate poundages of
3 and 6 have been used in 100 gallons of water applied at the rate
of 400 to 500 gallons per acre in southern California.

Cutting cattails close to the ground at two special stages of
growth in the same season resulted in 90% to 100% control in
Maryland, Tennessee, and Utah tests. The first cutting should be
made when pistillate spikes are well formed and at least two-
thirds full size but not mature enough to scatter viable seeds ; the
second about a month later, after regrowth has attained a height
of about 2 feet. Cutting at other times was found to reduce
greatly the percentage of kill. Success is likely to be greater in
areas where water is present and covers the cut stem-bases, yet
even where water is absent much of the time, as in some tidal or

CATTAILS 23

inland cattail marshes, mowing at the two specified stages can give
complete or practically complete control. The procedure has been
demonstrated successfully in various parts of the country, but its
use on a large scale requires mechanical equipment adapted to
mowing in marshy areas.

Mechanical crushing or breaking of cattail plants has given
results largely equivalent to those obtained by cutting. Drowning
(without cutting) is possible if bases of stems are submerged by
4 feet or more of water for more than 1 year, except for T. domin-
gensis which can endure slightly higher levels. Biological con-
trol favoring use by waterfowl may result from "eat-outs" in cat-
tail marshes by muskrats or nutria. A limited degree of natural
control has been caused by a boring-moth larva (Arzama sp.).
In some instances, it has seriously damaged or entirely eliminated
dense cattail stands.

REFERENCES

1956. Control of Cattail by Various ChemicarTreatments. W. O, Lee and F. L.

Timmons. WWCC Res. Prog. Rep. Feb., pp. 81-82.
1956. Control of Cattail by Chemical Spray. B. H. Grigsby, W. A. Cutler, and C. A.

Reimer. Abstracts Charter Meeting Weed Society of America, pp. 56.
1955. Observations on the Control of Cattail, Typha spp, by Chemical Sprays.

B. H. Grigsby, C. A. Reimer, and W. A. Cutler. Mich. Quart. Bui., 37:3:

400-406.
1955. Amino Triazole Puts Cattail on the Run. Anon. Farm Journal, Southern

Ed., Nov.
1952. Results with Chemical and Mechanical Methods of Controlling Common

Cattail, {Typha laiifoUa). F. L. Timmons. 13th WWCC Res. Prog. Rept.
1949. Taxonomy and Distribution of North American Cattails. Neil Hotchkiss

and Herbert L. Dozier. American Mid. Nat., 41 (1).
1948. Cattail Control with Scythe and 2,4-D. C. H. Walker. The Progressive

Fish-Culturist, 10 (3).

24
COCKLEBUR

BOTANICAL

A recent classification of the cocklebur genus, Xanthium, places
all of the common broadleaf forms under the one species X. stru-
marium, but some botanists still recognize numerous species.
These coarse annuals grow in a wide variety of sites including
fields, roadsides, and different types of waste places, but they be-
come especially abundant on exposed floodplains of impoundments
subject to drawdown. The prickly burs bear two seeds.

IMPORTANCE

Rank growths of the plants on flood plains and impoundment
margins often occupy space that would support duck-food plants.
Also seedlings in the cotyledon stage are reported toxic to live-
stock. In their favor is the fact that seeds of cockleburs are eaten
tc a limited extent by mourning doves.

CONTROL

Cockleburs can be killed readily by herbicides. Formulations
of 2,4-D are effective at the rate of 1/2 pound acid equivalent per
acre. Recent tests at the Tennessee National Waterfowl Refuge
have indicated that silvex gives satisfactory results at 14 pound
per acre. Also, complete flooding for 12 to 15 days has been
found to kill cockleburs.

25
GIANT CUTGRASS

BOTANICAL

Giant cutgrass Zizaniopsis miliacea, also called whitemarsh, is
a tall perennial of semitropical America that extends northward a
short distance beyond the southeastern Coastal Plain. It grows
in freshwater coastal marshes from Virginia to Texas and is par-
ticularly abundant in former ricefields of South Carolina and
Georgia. Giant cutgrass also occurs inland in the Lower Missis-
sippi Valley region as far north as Reelfoot Lake in Tennessee,
where it covers nearly 2,000 acres.

This weedy grass grows on moist soil or in water up to 3 feet
deep, and for limited periods it tolerates deeper water. Its seeds
— generally produced rather sparsely — germinate readily on moist
mud flats. Runners are the principal means of spread, especially
in flooded areas. In late summer, stolons or runners develop
freely from the lower parts of the plant, some of them attaining
a length of 14 feet. Leafy buds and roots produced at nodes of
the runners establish independent plants at some distance from
the parent, and in this way giant cutgrass spreads over both
flooded and unflooded sites.

IMPORTANCE

Waterfowl feed on giant-cutgrass seeds where they are plentiful
and readily available, as along margins of openings of the marsh,
or where the growth is sparse. Such value, however, is slight
compared to its detriment to waterfowl habitat ; the plant's aggres-
sive growth and saw-edged leaf blades make it one of the most
troublesome marsh weeds in the South.

26 GIANT CUTGRASS

CONTROL
Giant cutgrass is difficult to kill, and no fully satisfactory, eco-
nomical way has been found to control it despite nearly a decade
of experimentation at Reelfoot Lake and elsewhere. Within lim-
ited areas, complete elimination is possible, but generally it is
difficult and costly. Methods of control used and the extent of
their effectiveness depend considerably upon whether standing
water is present. Degree of success also depends on whether
treatment is made during the most vulnerable stage, which is in
late summer or early fall, during maximum runner growth. This
susceptible stage occurs about 2 months after the main period of
flowering.

CONTROL ON DRAINED AREAS

Growing on land, giant cutgrass can be reduced or largely elimi-
nated by disking, grazing, or by herbicides. Winter freezing kills
the plants, particularly if preceded by dewatering and burning.

On drained sites, dalapon seems superior to TCA, Polybor-
chlorate, Chlorax Liquid, silvex, 2,4,5-T and several other herbi-
cides tested. At rates of 70 to 199 pounds acid equivalent per
acre, dalapon applied during the vulnerable period has given good
to excellent control. Monuron at the rate of 100 pounds active
per acre has been effective in controlling giant cutgrass on moist
land not subject to moving water.

CONTROL IN FLOODED AREAS

At Reelfoot Lake complete foliage coverage by Ammate at the
rate of 170 pounds active or by Polybor-chlorate at 200 to 300
pounds per acre during the vulnerable period has yielded good to
excellent control. Of the newer herbicides, dalapon applied at
the rate of 140 pounds acid equivalent per acre proved effective.
The results from treatments with silvex and certain other herbi-
cides have been inconclusive.

Underwater cutting can, under favorable circumstances, yield
at least a 90% kill, but the method has practical limitations for
extensive application. For best results, the cutting must be done
in the vulnerable period and should be in at least 6 inches of water.
A commercial underwater mower tested for this purpose proved
impractical since it could not operate in water less than 16
inches deep and was unable to cut the tough stems of plants over
30 inches tall. Allowing cut stems to remain where they fall
appears to increase the percentage of kill. Mowing at other than
the vulnerable stage is comparatively ineffective.

REFERENCES
1955. A Report on Experimental Control of Giant Cutgrass (Zizaniopsis miliacea),
i 1950-54. Charles K. Rawls, Jr. Mimeo report SE Assoc. Game and Fish
. Commissioners Conf., Daytona Beach.

HIBISCUS 27

1950. Waterfowl Habitat Improvement on Reelfoot Lake. John H. Steenia. Jour.

Tenn. Acad. Sci., 25 (I).
19-15. A Progress Report on the Marsh and Aquatic Plant Problem: Reelfoot Lake.

John H. Steenis and Clarence Cottam. Jour. Tennessee Acad. Sci., 20 (1).

HIBISCUS

BOTANICAL

This warm-climate genus of nearly 200 species has about 20 to
25 native representatives in the United States. Because of their
showy flowers, various kinds of exotic hibiscus are cultivated as
ornamentals in warm regions of the country. Only two or three
species are significant as weeds in waterfowl habitat, though sev-
eral others occur in marshes. The plants are perennials that re-
generate from basal buds in the spring. Their large seeds ger-
minate abundantly on mud flats exposed by receding water.

H. moscheutos, here treated as including palustris and oculi-
roseus, is the most important weedy hibiscus. It grows in fresh
or mildly brackish marshes from Massachusetts to Texas and
occurs inland locally in the East. In some localities the plant is a
weed of major importance. Its local names are rosemallow, hibis-
cus, hollyhock, and marshmallow.

H. militaris is primarily an inland species, occurring from Penn-
sylvania, Minnesota, and Nebraska, south to Florida and Louisi-
ana. Its chief abundance is along river margins and in adjoining
areas subject to periodic overflow. H. lasiocarpos also grows inland,
sometimes associated with H. militaris. Saltmarsh mallow, Koste-
letzkya virginica, is common in Atlantic and Gulf Coast marshes
but generally has lesser consequence as a weed.
IMPORTANCE

Hibiscus plants provide nesting sites for some marsh songbirds

28 HIBISCUS

and have attractive flowers, but there is little else to recommend
them. Their seeds are seldom eaten by ducks or other v^^ildlife,
and locally the plants are serious weeds in waterfowl and muskrat
habitat.

CONTROL
Experiments on eradication of this weed in Delaware and Mary-
land have shown that effective control can be obtained by mod-
erately light dosages of 2,4-D, by underwater cutting, or by
drowning. Valuable duck-food plants have become established in
test plots where hibiscus was eliminated. Reinfestation should
be prevented by avoiding, if possible, continuous drawdown dur-
ing the growing season. Where protracted lowering of levels is
unavoidable, temporary reflooding of the area will drown the
seedlings.

HERSrCIDES

Tests on H. moscheutos have indicated that 2,4-D (about 6 1/2
pounds per acre, 0.5% concentration, and 160 gallons per acre)
gives about 95% control if applied during full-flowering or early-
fruiting stages. Also, treatments with a mixture of 2,4-D at
4 pounds and TCA at 50 pounds per acre have given consistently
good control. Results have been nearly the same on H. militaris.

DROWNING

Mature plants of H. militaris and H. lasiocarpos at Isom Lake,
Tenn., were drowned by a permanent 3-foot rise in water level.
Similar results have been obtained in TVA impoundments. In
the Thousand-Acre Marsh near Delaware City, Del., H. moscheutos
was nearly eliminated by an increase in water level varying from
15 to 30 inches and lasting about a year.

Seedlings can be controlled readily by drowning. In Delaware
it was found that hibiscus seedlings can be drowned without de-
stroying seedlings of desirable species such as smartweeds and
Walters wild millet.

cuniNG

Small-scale tests on cutting of hibiscus by sickles in Delaware
and Maryland have shown that if the cut bases are submerged
continuously, one mowing can give a high degree of control. In
other situations, two cuttings may eliminate 95% or more of the
plants. The two cuttings may be timed 1 to II/2 months apart,
or the second cutting may even occur in the following year and
still be nearly equally effective.

REFERENCE
1954. The Marshes of Delaware, Their Improvement and Preservation (p. 35).
John H. Steenis and Others, Delaware Board of Game and Fish Comm.

29
LOTUS

BOTANICAL

The native yellow-flowered lotus, Nelumbo lutea, also called yon-
kapin, yonkanut, bonnets, or water chinkapin, is common in ponds
and lakes of the Mississippi Valley and occurs widely in the East.
A pink-flowered Asiatic species, N. nucifera, has been introduced
in various places but thus far has shown no tendency to become
a problem plant.

This umbrella-leafed aquatic grows in water up to 8 feet deep.
It tolerates drops in water level, but a sudden rise covering leaves
for at least 12 days will kill mature plants. Vegetative spread by
runnerlike rootstocks is rapid — about 250 feet of new growth
in 1 year has been reported. Banana-shaped tubers, filled with
starchy food, are borne on the runners. The acornlike seeds may
germinate shortly after maturity or may remain dormant and
viable for hundreds of years.

IMPORTANCE
A few species of ducks such as wood ducks and teal feed on
lotus seeds to a limited and generally insignificant extent — mainly
in late summer before the seedcoat has hardened. Tubers and
seeds were formerly eaten by Indians. In many localities the
plant is a serious pest, covering ponds with its rank vegetation
and precluding good growths of desirable duck-foods.

30 LOTUS

CONTROL

Control can be accomplished rather easily, by herbicides, drown-
ing, or mowing. The best time is the summer season, when the
plants are flowering freely. Eradication must be complete, else
rapid spread of surviving plants may soon reestablish beds.

In some localities, a biological means of control is evident;
larvae of the leaf beetle Donacia hypoleuca bore into petioles below
water level and sometimes destroy extensive beds in deep water.
Only once in a total of 10 years did this insect provide conspicuous
control of lotus at Reelfoot Lake in Tennessee.

HERBICIDES

A single herbicide spraying of 2,4-D (either salts or esters) at
2 to 4 pounds of acid equivalent per acre can give nearly complete
control of lotus.

DROWNING

Where water levels can be managed readily, drowning of lotus
is the easiest and most economical control procedure. A rapid
rise of 30 to 36 inches (to cover the leaves) is needed and must
be maintained at this high stage for at least 12 days. Plants are
more susceptible when flowering freely. Turbidity of water also
increases effectiveness. Drowning of lotus sometimes occurs nat-
urally, after heavy rains and flooding.

cuniNG

In general, mowing of lotus is likely to be less practical* than
the use of 2,4-D. Underwater mowing of leaf and flower stalks
in shallow water, less than 2V2 feet deep, is generally ineffective.
In water 3 to 5 feet deep, 80% to 100% control has been obtained
by two cuttings. Occasionally, as when the water is dark, one
cutting may be sufficient. A commercial weed cutter tested in
Reelfoot Lake was found to be impractical for this purpose be-
cause of stumps or snags present.

REFERENCES

1950. Waterfowl Habitat Improvement on Reelfoot Lake. John H. Steenis. Jour.

Tennessee Acad. Sci. 25 (1).
1950. Studies on the Use of Herbicides for Improving Waterfowl Habitat in Western

Kentucky and Tennessee. John H. Steenis. Jour. Wild. Mgt. 14 (2).
1950. Leaf Beetles vs. Lotus. John H, Steenis and Robert T. Mitchell. Jour.

Wild. Mgt. 14 (4).
1945. A Progress Report on the Marsh and Aquatic Plant Problem; Reelfoot Lake.

John H. Steenis and Clarence Cottam. Jour. Tennessee Acad. Sci. 20 (1).
1944. The Biology of the American Lotus, Nelumbo liUea (Willd.) Pers. Thomas F.

HaU and William T. Penfound. American Mid. Nat. 31 (3).

31

MAIDENCANE

BOTANICAL

Of the 170 Panicum species listed in the United States, maiden-
cane, P. hemitomon, is the most serious pest plant in waterfowl
habitat. In this country, it is confined almost entirely to fresh-
water marshes of the South Atlantic and Gulf Coasts. In Louisi-
ana, where the plant is known as paille fine or canouche, it con-
stitutes a large proportion of the sod in floating marshes.

Maidencane grows on moist soil or in shallow water. Many
aerial stems are sterile, but limited seed production is offset by
rapid growth of rootstocks.

IMPORTANCE

The plant has some value to muskrats for housebuilding and
food and is eaten by nutria, but much more significant is the fact
that it is a hard-to-control competitor of duck-food plants in the
South.

CONTROL

Because of its rapid spread by rootstocks, control of maiden-
cane needs to be complete or practically so. It is more difficult to
destroy when growing in water than on drained land.

At present, the most promising prospect for effective control
of maidencane has come from recent preliminary tests with
dalapon on Federal refuges in the Southeast. At 20 pounds acid
equivalent per acre, in early treatments (late March or April) at
the Savannah (Ga.) and Loxahatchie (Fla.) Refuges, it has given
complete or nearly complete kills. Older chemicals such as 2,4-D,
TCA, and Ammate have proved impractical because of large
amounts required, and 1 to 3 tons of soil sterilants such as Bor-
ascu or Polybor-chlorate are needed for extermination in de-
watered units. On drained sites, maidencane can be eliminated
largely or entirely in the course of a couple of years by a series
of diskings or by diskings combined with herbicide treatments.

32
NEEDLERUSH

BOTANICAL

Many kinds of rushes (Juncus) grow on moist soil or in water
in various parts of the United States, but needlerush {J. roemeri-
anus) is the only one of major importance as a weed in waterfowl
habitat. This tall, needle-pointed plant dominates thousands of
acres of brackish coastal marshlands from Maryland to Texas.
It is perennial and spreads by rootstocks as well as by seeds. The
common rush (J. effiisus) frequents meadows and other moist
sites but generally is of lesser importance as a weed.

IMPORTANCE

Needlerush provides excellent cover for rails and is also useful
as cover for black ducks and other waterfowl in some places.
However, its minute seeds are eaten only rarely by ducks, and
in many marshes it limits usefulness of the habitat for waterfowl
and muskrats.

The weed status of this plant varies in different marshes or
in different parts of the same marsh. Besides being beneficial
as cover in some flooded sites and detrimental in many others, it
is neither harmful nor useful wherever standing water is so
limited in extent or duration that waterfowl would not use the
area even if needlerush growths were destroyed.

NEEDLERUSH 33

CONTROL

Studies on needlerush control have yielded encouraging results,
but much remains to be learned about best methods to use in dif-
ferent places and about prospects for suitable replacement. In-
dications are that the plants cannot spread or reestablish as
readily in sites flooded most of the time as compared with better-
drained locations. The most vulnerable period is in the spring,
during flowering and early stages of fruiting.

Different results have been obtained in different places with
similar control treatments, as indicated in the three main areas of
needlerush study : North Carolina, Florida, and Maryland. The
variations in susceptibility appear to depend on both regional and
localized factors. Another peculiarity in the response of needle-
rush to chemicals is the fact that herbicidal concentrations have
to be increased when applying a large gallonage per unit area,
as with power sprayers. This contrasts with the usual situation
with marsh plants in that ordinarily the requirements for pound-
age of herbicides per unit area remain the same regardless of
carrier volume.

In North Carolina, 27 pounds acid equivalent of 2,4-D isopropyl
ester per acre were required for effective control (Barber, 1952).
At the Chassahowitzka Refuge in northern Florida, 95% to 99%
elimination of needlerush resulted from airplane treatment using
16.7 pounds acid equivalent of 2,4-D ester in 5 gallons per acre
(Myers, 1955). In Maryland, both salt and ester formulations
of 2,4-D have yielded over 95% control at rates as low as 14
pounds acid equivalent per acre in 200 gallons of carrier (either
water or oil) when applied during the vulnerable period (Steenis,
Webster, Nicholson, 1954).

The Maryland tests involved use of hand-pressure cylinder
sprayers. Treatments at other than the vulnerable period gave
erratic results. Slightly superior results were noted with 2,4-D
esters in oil, probably owing to better coverage plus some herbi-
cidal effect of the oil. TCA alone was not satisfactory for needle-
rush control but with a combination of 2,4-D at 8 pounds acid
equivalent per acre and 25 pounds of TCA the results were bet-
ter than with 2,4-D alone. In treatments with 2,4-D alone by
power-sprayer equipment, using over 400 gallons of water per
acre, it was found necessary to increase the herbicidal concentra-
tion to 27 pounds per acre. Also in mixed treatments with a
power sprayer, it was necessary to increase the concentration of
2,4-D to 14 pounds per acre and TCA to 50. Monuron, acting
as a soil sterilant, was used successfully at 20 pounds active per

34 NEEDLERUSH

acre in areas where there was no extensive water movement for
3 months following treatment. Dalapon showed limited promise
at rates of 45 pounds or more active per acre. Mixtures of it
with other herbicides have not been tried on needlerush.

REFERENCES

1955. Management of Needlerush Marsh at the Chassahowitzka Refuge. Kent E.
Myers. SE Assoc. Game and Fish Commnrs. Conf.

1955. Improvement of Marshes by Weed Control. John H. Steenis, Clark G.
Webster, and William R. Nicholson. Proc. NE Sect. Wildlife Soc.

1954. Progress Report on Needlerush Control. John H. Steenis, Clark G. Webster,
and William R. Nicholson. FWS mimeo report.

1952. Experimental Control of Juncus roemerianus with Herbicides in North Caro-
lina. Yates Barber. N. C. Wildhfe Res. Comm. mimeo report.

1951-54. Comments by Kenneth A. Wilson in N. C. P-R Quarterly Reports on
Project 6-R-F, June 1951, September 1953, and June 1954.

35

PHRAGMITES

BOTANICAL

Phragmites (Phragmites communis) is a tall, plumed marsh
grass native to Asia, Africa, and Europe as well as to North Amer-
ica. Though widespread in this country, it is largely absent from
the Southeast, except in gulf-coast marshes. Local names include
giant or common reed, cane, feathergrass, or foxtail (north At-
lantic coast), roseau or roseau cane (Louisiana and eastern
Canada), and carrizo or carrizo cane (Southwest).

Dense stands of this robust perennial sometimes extend over
many acres. The plant spreads mainly by long runners either
underground or aboveground. In shallow water, phragmites can
extend its growth if the runners are able to obtain anchorage.
Above-ground runners have been known to grow 25 feet in 1 year.
It had been assumed that the plants produce viable seed only
rarely, yet seedlings have been noted commonly in the Middle At-
lantic Region.

IMPORTANCE

Margins of phragmites beds or stands that are not too dense
may provide useful cover for waterfowl and locally they serve as
cover for pheasants, rabbits, and other game. Runners and aerial
stems have some usefulness as food for muskrats and are valuable
for their house-building. Hunters frequently make duckblinds

36 PHRAGMITES

with the stems. In the Southwest the plant has been used for
thatching, matting, cordage, and nets.

On the debit side, phragmites is one of the most common weeds
in waterfowl habitat and one of the most difficult to control.
Though most stands of this plant occur on comparatively high, dry-
parts of marshes, above the important zone for duck foods, it is
also a serious competitor with useful plants, particularly in areas
where water levels are not constant. Once established on de-
watered flats, it is likely to persist indefinitely and has been known
to survive in water a foot or more deep for at least 8 years. In
the West, phragmites is a pest along irrigation canals, and in vari-
ous parts of the country it is a serious fire hazard near industrial
installations.

CONTROL

Destruction of phragmites is difficult because of its tough,
durable rootstocks. Killing the aerial stems by burning, mow-
ing, or spraying frequently results in only a temporary setback.
The plant is particularly resistant to control measures when it
is growing in water. Its most vulnerable stage for treatment
is during the time of pollen release, a 3-week period occurring in
midsummer to early fall depending on location. In most cases,
best results have been obtained by treatment during the vulner-
able period, but the effectiveness of dalapon is not limited to this
season.

Dalapon and ATA, either alone or in combinations, have recently
shown ability to control phragmites effectively. Ammate and
2,4-D have proved unsuitable for this purpose, and TCA, though
better, is not satisfactory. On comparatively dry sites, phrag-
mites can be eliminated by repeated mechanical operations, such
as disking.

HERBICIDES ON COMPARATIVELY DRY SITES

Dalapon treatments have given good results on terrestrial
stands of phragmites at various times during the growing season
from spring when new shoots are about 3 feet high until early
fall when the plants are full grown and in their most vulnerable
stage. Application of 30 to 35 pounds acid equivalent of dalapon
at any time during this 3- to 4-month period has provided good
control. ATA at 16 pounds active per acre was also effective if
applied during the pluming-flowering stage, a period of 4 to 5
weeks. A mixture of dalapon at 15 pounds and ATA at 2 pounds
has given good kills at various stages of growth, from early to

PHRAGMITES 37

late summer. The soil sterilant, erbon, was successful at 40
pounds active per acre, and monuron at the rate of 40 pounds
active per acre was fairly effective if disked into the ground.
Operational programs using dalapon or mixtures of dalapon and
ATA applied in approximately 400 gallons of water per acre by
power sprayers are now being conducted in New York, New Jer-
sey, and Delaware.

HERBICIDES ON FLOODED OR CONTINUOUSLY WET SITES

Dalapon alone has proved unsatisfactory in flooded habitat,
regardless of poundage used or stage of plant growth. However,
treatment with dalapon at 25 to 35 pounds combined with ATA
at 8 to 10 pounds gave good results when done during the vulner-
able stage. ATA alone was sometimes successful when used at
20 to 24 pounds in the vulnerable period. A recent airplane treat-
ment with a mixture of dalapon and ATA in 15 to 20 gallons of
water and using triethanolamine as dispersion agent has given
encouraging indications. This method of control can have special
importance as a means of reducing or eliminating phragmites in
comparatively inaccessible flooded areas.

MECHANICAL METHODS

A mowing of phragmites in August 1951, followed by burning
and two diskings that fall, resulted in limited spring survival rang-
ing from 5% to 20% at the Fish and Wildlife Service Killcohook
Refuge near Salem, N. J. The plants were growing on a fill depos-
ited from dredgings of the Delaware River. After a third disking,
cultivated crops were planted. In New Jersey mosquito-control
programs it has been found that crushing and crimping by tractor
wheels equipped with angle-iron cleats is more destructive to
phragmites and other marsh plants than mowing.

At the Delta Duck Station in Manitoba, 3 years of summer mow-
ing have been reported as effective in transforming phragmites
beds into meadows on peaty soil which becomes dry in late summer.
However, mowing on wet, nonpeaty soils elsewhere has given little
or no control. Underwater cutting on small plots during the flow-
ering season in Delaware and Maryland studies was effective, but
the method has limited practicability.

BURNING

Burning appears to injure phragmites to only a minor extent
except in localities where the plants are growing on peat. At the
Delta Duck Station in Manitoba, late-summer burning not only

38 PHRAGMITES

destroyed phragmites beds but also created useful sloughs by burn-
ing into the peat. Intensive burning continued for several years
in the Imperial Valley of California and in gulf -coast marshes has
proved unsuccessful in suppressing this weed. However, winter
burning before spraying can be advantageous in making spraying
operations easier.

FLOODING

Mature stands of phragmites cannot be controlled by raising the
water level, but their seedlings can be drowned this way. Also a
12-inch depth of water during summer production of runners
causes them to float on the surface instead of becoming anchored
and thus prevents extension of an established stand.

REFERENCES

1957. Use of Herbicides to Increase Game Food in Solid Stands of Phragmites.

Robert A. Beck. Proc. N. E. Wild. Conf.
1956. Progress Report on Phragmites Control. John H. Steenis. Weed So. Am.

Abstracts.
1956. Control of Phragmites with Dalapon. B. E. Day and A. W. Swezey. Down

to Earth, spring issue.
1954. Progress Report on Studies on the Control of Phragmites. John H. Steenis,

Clark G. Webster, and Robert A. Beck. FWS mimeo report.
1942. Phragmites Management. Edward Ward. Trans. 7th N. Am. Wild. Conf.
1932. The Development and Use of Machinery for Crushing Grasses, Reeds, and

Sedges on the Salt Marshes. John P. Peterson. N. J. Mosquito Exterm.

Assoc, Proc.

PICKERELWEED

BOTANICAL
Pickerelweed (Pontederia) in the United States has been classed
either as one species, P. cordata with a variety lanceolata, or as two
species. The lanceolate-leaf form is restricted mainly to the South.
Pickerelweed has wide range in the East, but its main abundance
and most extensive beds are in fresh or slightly brackish water
along the Atlantic Coast and in the southern part of the Great
Lakes area. The spikes of blue flowers are distinctive, and so are
the somewhat heart-shaped leaves with rounded bases and no prom-
inent veins. The plants are perennial from thick rootstocks.

IMPORTANCE

Seeds of pickerelweed are seldom eaten by ducks, and the plants
are, in general, liabilities rather than assets in waterfowl habitat.

CONTROL
Limited experiments with pickerelweed in Delaware have indi-
cated that the best control known at present is a double treatment :
a foliage spraying of 2,4-D at the rate of 8 pounds acid equivalent
per acre during the flowering period (late June in Delaware) fol-
lowed 4 to 6 weeks later by a treatment with Ammate at 60 pounds
active per acre.

REFERENCE

1954. The Marshes of Delaware, Their Improvement and Preservation (p. 34).
John H. Steenis and others. Delaware Board Game and Fish Comm.

40
SPAHERDOCK

BOTANICAL

Some botanists recognize numerous species of Nuphar in the
United States, whereas others treat the group as one species,
N. luteum, with several varieties. Spatterdock, also known as
cowlily, mulefoot (South), or yellow pondlily, is common in fresh-
water ponds, lakes, and slow streams in various parts of the coun-
try. The plants are able to grow in 5 feet or more of water, and
their seedlings can develop in approximately that depth. Spread
is by seeds and by rootstocks.

IMPORTANCE

Ringnecked ducks, wood ducks, and, to a lesser extent, other
waterfowl feed on spatterdock seeds ; in breeding areas, the plants
can serve as useful brood cover, and they harbor insects upon which
young ducks feed. The rootstocks are considered about a third-
rate food for muskrats. In many locations it thrives under con-
ditions of turbidity or fluctuating levels that preclude desirable
vegetation, but in other places it is a serious competitor with more
useful waterfowl plants.

CONTROL

Spatterdock is fairly difficult to destroy, and fully satisfactory
methods have not yet been developed. The plant can be controlled
by herbicides, or by cutting provided it is done at the right time.

SPATTERDOCK 41

Tlie most vulnerable stage of growth has been determined to be
the period of maximum blooming and early fruiting. Young seed-
lings are easily destroyed by dewatering, but killing mature root-
stocks by this means is a slow process, especially if the stems are
covered by silt.

HERBICIDES

A limited recent testing with ATA in New Jersey at 8 pounds
acid equivalent per acre has given fairly encouraging results.
Combination treatments with 2,4-D and Ammate have consist-
ently yielded over 90% control in Tennessee and Delaware. A
2,4-D application at the rate of 6 pounds acid equivalent per acre
during the time of heavy blooming and early fruiting was fol-
lowed a month or 6 weeks later by spraying with Ammate at 80
pounds active per acre. Such combination treatments have been
more effective than repeated sprayings with either 2,4-D or
Ammate.

MOWING

Underwater cutting of spatterdock in about 4 feet of water at
Reelfoot Lake showed, among other things, that a first cutting
during heavy flowering and early fruiting followed by a cutting
of regrowth about a month later gave 80% to 90% control. In
shallower water, a third cutting about another month later was
necessary for equivalent results. Cuttings made without regard
to the vulnerable period are generally ineffective.

REFERENCES

1950. Waterfowl Habitat Improvement on Reelfoot Lake. John H. Steenis. Jour.

Tennessee Acad. Sci., 25 (1).
1945. A Progress Report on the Marsh and Aquatic Plant Problem: Reelfoot Lake'

John H. Steenis and Clarence Cot tarn. Jour. Tennessee Acad. Sci., 20 (1)

42

WATERCHESTNUT

BOTANICAL

Waterchestnut (Trapa natans), a native of Europe and western
Asia, was found in a lake near Scotia, New York, about 1884.
Since then it has spread to various points in New York State and
has become established in Massachusetts, Vermont, and Maryland.
Presumably it will continue to appear in new places. Other names
for the plant are water caltrop, waternut, bullnut, and shingoda.

Waterchestnut grows in ponds, lakes, river estuaries, and slow-
moving streams at depths varying from a few inches to about 15
feet. It thrives particularly at depths of 2 to 5 feet and favors
muddy bottoms with high organic content. Though typically a
freshwater plant, waterchestnut tolerates slightly brackish condi-
tions such as those in parts of the tidal Potomac where the chlo-
rine content during the average growing season ranges from about
1,700 to 3,500 parts per million of water.

Floating rosettes, commonly several of them produced on
branches of the same plant, top the flexible stems. Leaf stalks
are inflated, but not as conspicuously so as in waterhyacinth. In
early summer, small white flowers develop in the rosettes, and
these in turn produce large nutlike seeds armed with four sharp
spines having minute retrorse barbs. Most of the seeds germi-
nate in the following spring, but some may remain dormant and
viable for years. The empty seed shells float to the surface and
become a menace to bathing beaches.

WATERCHESTNUT 43

Detached, well-developed rosettes are capable of taking root on
muddy margins and maturing a crop of seeds. The plant is
strictly an annual, and the comparatively small number of seeds
produced, commonly a dozen or more, are its sole means of sur-
vival. Nevertheless, a small infestation located on the Potomac
near Alexandria, Va., in 1923 extended itself 35 miles in 15 years
— its beds blanketing much of the shallow water in this distance.

IMPORTANCE

Some use is made of the seeds as human food in Asia and
Europe, but in this country the plant appears to be wholly a lia-
bility. The "waterchestnut" served in American versions of Chi-
nese dishes is not Trapa natans but Eleocharis dulcis, an entirely
different plant. Dense beds of waterchestnut can impede or pre-
clude both commercial and recreational navigation, destroy use-
fulness of areas for bathing, fishing, and waterfowl hunting, and
increase hazards from malarial or nuisance mosquitoes. It is not
only a serious pest at present but also menaces waterways not
yet invaded. Once in a river system, its eradication is costly.
Control work on waterchestnut in the Potomac River required
more than 20 years of extensive operations costing several hun-
dred thousand dollars, and additional years of patrol and cleanup
activity will be necessary before extermination is complete.

PRESENT EXTENT

NEW YORK STATE: Waterchestnut is estimated to cover about 2,500
acres in the Hudson and Mohawk Rivers. It appears rather
hopelessly entrenched in these streams because danger to nearby
crops limits safe and effective control. The plant has been largely
or entirely eliminated in a number of small water areas near
these two streams, but meanwhile it has invaded Lakes Keuka
and Champlain.

MASSACHUSEHS: Beds in or near the Sudbury and Concord Rivers are
reported to have been brought largely under control during the
past decade, but it will be years before the weed is entirely
eliminated here. An infestation in a pond on the Mount Holyoke
College campus at South Hadley is a threat to the Connecticut
River.

VERMONT: Watcrchestnut was found on the Vermont side of Lake
Champlain's southern end in 1950, presumably introduced by
freight barges. Limited-scale control efforts are under way.

44 WATERCHESTNUT

MARYLAND: Large beds of waterchestnut have recently (1954) been
found in the Chesapeake Bay area, near the mouth of the Gun-
powder River. Extensive cutting and spraying operations were
conducted here in 1955, but only limited reduction was evident in
1956. Full control will probably not be obtained for a good many
years, and meanwhile nearby areas valuable to waterfowl are
endangered.

CONTROL

Since the plant is an annual, the main aim in control operations
is to prevent maturing of new crops of seeds. It is these which
maintain and spread infestations. All feasible means of control
should be used. For extensive, dense beds, herbicidal spraying is
likely to be practical. Underwater mowers are valuable, both for
extensive cutting of beds and to clear paths for spraying equip-
ment. Hand pulling and disposal of plants on land is advisable
where growths are sparse or beds are small.

Timing of operations appears important. Cutting and pulling
should begin early in the season, about the time small rosettes
have reached the surface and before flowering and seed-setting
have commenced. The best time for spraying is about midsummer
when many of the leaves are forced above the water by the pres-
sure of rosettes on each other. These conditions favor direct con-
tact with the herbicide. Spraying after seeds are developed is
likely to be too late.

After infestations have been suppressed, regular patrolling
must be maintained for an indefinite number of years to prevent
reestablishment. The maximum longevity of waterchestnut seeds
is not known, but it is evident they can survive under silt for at
least 10 years.

HERBICIDES

Tests by the New York State Department of Conservation in
1950 showed that though treatments with 2,4-D at rates of 2 or
4 pounds of acid equivalent per acre can be fairly effective, better
or more uniform results were obtained at 8 pounds per acre.
Amine and ester formulations of 2,4-D were found nearly equiva-
lent in effectiveness. New York State has limited its control
operations largely to amine-salt formulations to avoid danger of
damage to crops and has made applications at the rate of 8 pounds
acid equivalent in about 50 gallons of water or #2 fuel oil per
acre. The Massachusetts Department of Conservation has used
a low-volatile ester of 2,4-D at about 4 pounds acid equivalent in

WATERCHESTNUT 45

50 gallons of #2 fuel oil per acre. In a limited part of the Gun-
powder River infestation, the Maryland Board of Natural Re-
sources appears to have obtained good results in 1955 (under re-
portedly ideal spraying conditions) v^^ith amine salt at the rate of
about 2 pounds acid equivalent per acre in 25 gallons of v^^ater
containing a detergent. Recently, however, the poundage has
been increased.

Follow-up control within the same season may be necessary.
Spray is likely to be less effective on foliage which is largely
under water as in sparse stands or when waves wash over the
rosettes. Airboats have proved useful for travel over beds dur-
ing spraying operations.

MOWING

Cutting stems about a foot below the water surface can be
very helpful in suppressing waterchestnut infestations. Hockney
underwater mowers or other types of equipment, including sickles,
can be used for the purpose. Mowing should begin early in the
season, about the time small rosettes appear at the surface. Cut-
ting at this time is easier and may involve less regeneration. At
least one follow-up mowing is likely to be necessary within the
same season. Mowing after the rosettes have begun to set seeds
is not likely to be worth while since some seeds may have already
dropped to the bottom and others can mature on drifting rosettes.
In small operations it may be possible to remove cut rosettes and
destroy them, but in larger operations this is not practical.

REFERENCES

1955. Experimental Control of Water Chestnut {Trapa nalans) in New York State.

Ralph H. Smith. New York Fish and Game Jour.
1955. Water Chestnut Threatens Disaster to Maryland Water Areas. Maryland

Tidewater News, 12 (1).
1955. Another Waterchestnut Infestation. Alexander C. Martin. Jour. Wild.

Mgt., 19 (4).
1950. Water Chestnut Control Project Report and P-R Quarterly Reports on Projects

10-Dl, 2, and 3, Massachusetts Dept. Cons.
1949. Potomac River and Tributaries at and below Washington, D. C. ; Ehmination

of Waterchestnut. 81st Cong. House Document 113.
1949. (approximately). Danger: The Water Chestnut. State of New York, Dept.

of Cons.

46

WATERHYACINTH

BOTANICAL

The tropical genus Eichhornia (also listed as Piaropus) in-
cludes three species in South America besides waterhyacinth,
E. crassipes. The latter is a pest in North, Central, and South
American countries, and in Australia, India, and Java, Africa,
and other areas. In the United States it is confined by climate to
the Southeastern Coastal Plain and a few California localities.

The plant thrives either free-floating, buoyed by bladderlike
bases of the leaf stalks, or attached by roots to muddy margins.
Rapid propagation occurs by offsets of small new plants. Three
Louisiana plants are reported to have multiplied into 3,000 plants
within 50 days by this means. Waterhyacinth also produces
viable seeds, but propagation by seedlings appears to be negli-
gible except on floating mats. The plant is also called hyacinth,
lily, water orchid, and river raft.

IMPORTANCE
Waterhyacinth *s beautiful flowers and curious form make it a
popular aquarium specimen, but in other respects it is a weed.
Its hindrance of navigation is so great that nearly 15 million dol-
lars have been spent in efforts to control the pest during the past
40 years. In addition, waterhyacinth causes much harm by
destroying usefulness of fish and wildlife habitat, hinders drain-
age, and increases malaria hazard. A 1947 report (Lynch and
others) on the harmfulness of waterhyacinth and alligatorweed

WATERHYACINTH 47

states that "The wildlife crop these weeds destroy each year in
Louisiana is valued at $14,727,000." It has been estimated that
the total annual loss to Louisiana agriculture, fish and wildlife,
navigation, drainage, and public health amounts to nearly 38 mil-
lion dollars. Because of this economic drain, efforts are being
made to reduce or eliminate waterhyacinth on thousands of acres
of its present range.

CONTROL

Herbicidal treatment has proved so effective on waterhyacinth
that mechanical methods employed formerly have been largely
abandoned. The plant can be killed rather readily by light or
moderate dosages of 2,4-D formulations except where floating
mats are so dense that herbicide spray does not penetrate into
lower layers.

Recommendations resulting from experiments conducted co-
operatively in Louisiana, 1948-50, by the Boyce Thompson Insti-
tute, the Corps of Engineers, and Tulane University, included the
following: Use of the alkanol amine salt of 2,4-D (for combined
economy and low hazard to crops) at 8 pounds of acid equivalent
per acre; spray concentration of 1.2^ and 80 gallons of solution
per acre in low-pressure spraying; in high-pressure treatment
from a helicopter, 2 gallons per acre of the 8-pound acid equiva-
lent formulation direct from the drum. Most effective killing and
sinking of treated plants was found to occur in the period of rela-
tively slow growth, from August to March.

Current practice in waterhyacinth control deviates consider-
ably from the above recommendations. The amount of acid
equivalent actually applied per acre in current control programs is
as low as 1 or 2 pounds, the operators believing that these amounts
are effective and adequate. The Research Committee of the South-
ern Weed Conference has recommended (1957) use of "1 to 4
pounds of 2,4-D acid equivalent in 2 to 150 gallons of water" for
waterhyacinth control.

In recent years, the Florida Game and Fresh Water Fish Com-
mission cleared more than 20,000 acres of waterhyacinth infesta-
tion on lakes and ponds in the State. This was done on a coopera-
tive basis in which the landowners paid costs of chemicals,
amounting to about $3 per acre, and the Commission footed the
rest of the bill which was about $5. Aerial applications were
made with a V^-gallon per acre of a 3.34 pounds acid equivalent
ester formulation of 2,4-D, and treatments from boats (mainly
airboats) involved % gallon per acre of a 4 pounds acid equivalent

48 WATERHYACINTH

amine of 2,4-D in water. Since 1948, the Louisiana Wildlife and
Fisheries Commission has conducted extensive control programs
on waterhyacinth, and at present such activities have covered
nearly half of the infested sections of the State. Efficiency in
operations has been increased substantially recently by using
lighter, lower-pressure spraying equipment and larger, more pow-
erful barges. In a 1955 report on waterhyacinth investigations,
the Florida Everglades Experiment Station in cooperation with
the U. S. Agricultural Research Service and the Central and
Southern Florida Flood Control District has advocated the use of
1 pound acid equivalent of 2,4-E) amine per acre. The report in-
dicates general expectancy that a 90% to 95% kill will result from
the initial treatment.

REFERENCES

1957. Water Hyacinth Control. Pete Button. Louisiana Conservst., Jan.

1955. Hyacinth Control Report. Pete Button. Louisiana Conservst., May-June.

1955. Noxious Vegetation Control. Bon R. Luethy. Florida Game and Fresh

Water Fish Coram. Fed. Aid Bui. 2, May.
1955. Controlling Water Hyacinths with 2,4-B in South Florida. J. C. Stephens,

V. L. Guzman, and C. C. Seale. Everglades Sta. mimeo report 55-12, Ma}'.
1954. Water Hyacinth Control. Pete Button. Louisiana Conservst., February.
1954. America's Most Breadful Flower. James Polling. Colliers Mag., July.
1953. The Use of 2,4-B in the Control of Water Hyacinth and Alligator Weed in the

Mississippi Belta, with Certain Ecological Implications. Willis A. Eggler,

Ecology 34 (2).
1952. 2,4-B V8 the Hyacinth. Bon R. Luethy. Florida Wildlife 6 (3). August.
1952, Engineer Uses of Chemical Herbicides. Walter C. Carey. The Military

Engineer 44 (299).
1950. Control of Water Hyacinth. P. W. Zimmerman, A. E. Hitchcock, Henry

Kirkpatrick, Jr., and T. T. Earle. Boyce Thompson Inst. Prof. Paper 2 (9).
1950. Growth and Reproduction of Water Hyacinth and Alligator Weed and Then*

Control by Means of 2,4-B. A. E. Hitchcock, P. W. Zimmerman, Henry

Kirkpatrick, Jr., and T. T. Earle. Boyce Thompson Inst. Reprint 688.
1949. Water Hyacinth: Its Growth, Reproduction and Practical Control by 2,4-B.

A. E. Hitchcock, P. W. Zimmerman, Henry Kirkpatrick, Jr., and T. T.

Earle. Boyce Thompson Inst. Reprint 669.
1948. The Biology of Water Hyacinth. WiUiam T. Penfound and T. T. Earle.

Ecol. Monol. 18 (4).
1947. Effects of Aquatic Weed Infestations on the Fish and Wildlife of the Gulf

States. J. J. Lynch, J. E. King, T. K, Chamberlain, and Arthur L. Smith, Jr.

FWS Spec. Sc. Rep. 39.
1944. The Economic Status of the Water Hyacinth in Louisiana. James Nelson

Gowanloch. Louisiana Conservst. 2 (9), August.

49
WATERLETTUCE

BOTANICAL

Waterlettuce (Pistia stratiotes) is a floating, semitropical plant
limited to the Gulf region. Generally, it is much less abundant
and important than its frequent associate, waterhyacinth. How-
ever, there is growing need for information on how to control
waterlettuce since it is less susceptible to 2,4-D than water-
hyacinth and its beds tend to spread in places where water-
hyacinth has been killed. Water solutions of 2,4-D, such as are
used in waterhyacinth control, are not effective on the velvet-
surfaced leaves of waterlettuce. Consequently, oil carrier is
required.

50

WATERLILIES

BOTANICAL

The two closely related species of white-flowered waterlilies,
Nymphaea odorata and tuberosa, occur commonly and often abun-
dantly in ponds and lakes of the Eastern States. Three other
species have less importance as weeds because their distribution
i? limited. Banana waterlily, N. mexicana, is a valuable duck food
locally along the Gulf and South Atlantic Coasts.

IMPORTANCE

Waterlilies, whether growing wild or planted in ponds or for-
mal pools, are attractive to the eye, and their seeds have some
usefulness as food for ducks. As long as their beds remain
sparse they are not likely to be objectionable, but frequently water-
lilies become so dense as to shade out submerged duck-food plants
and interfere with fishing and boating.

CONTROL

At present, no method of control has proved uniformly effective
on waterlilies in different places. Results with herbicides, includ-
ing various formulations of 2,4-D, or with mowing have gener-
ally required repeated operations in two or more years. Excellent
small-plot kills have been obtained with 2,4-D isopropyl ester at
the rate of 2 pounds acid equivalent per acre on the Lacassine
Refuge in Louisiana. The treatments were made in May on mixed
growths of waterlilies and watershield {Brasenia schreberi) with
the result that nearly pure stands of watershield dominated the
plots.

Experiments with new herbicides are under way at several sta-
tions but at this time no conclusive results are available.

51

WATERPRIMROSES

BOTANICAL

The Southeast, the region in which waterprimroses are fre-
quently nuisances in waterfowl habitat, has nine species of this
genus (Jussiaea). Some of the nine are too limited in distribu-
tion or too sparse to be serious weeds. Two of the most abundant
and troublesome ones are J. diffusa and /. grandiflora. Though
some species grow erect in moist soils, others root on muddy-
margins of ponds or ditches and produce dense beds over water
surfaces.

IMPORTANCE

Waterprimroses often do harm by competing with desirable
plants; their seeds have limited local value as duck food.

CONTROL

The most vulnerable period for waterprimroses is during maxi-
mum flowering. /. diffusa can be controlled fairly easily by vari-
ous 2,4-D formulations when applied at the rate of 4 pounds acid
equivalent to 100 gallons of water, sometimes by a single treat-
ment if the beds are not too dense. /. grandiflora is somewhat
more resistant and generally requires several treatments with
2,4-D. In Maryland, good results were obtained from spraying
with 2,4-D ester at 8 pounds acid equivalent in 100 gallons of oil,
applied during the vulnerable stage. Water-carried sprays re-
quired wetting agents, and results were not as good as with oil.

WOODY WEEDS

In many localities, dense growths of trees or shrubs hinder
production of desirable marsh vegetation. Woody weeds are prob-
lems in seasonally flooded swamps of the Southeast and various
other regions, in bogs of northern States, and in thousands of
acres of Southwestern lowlands that are covered by jungles of
saltcedar, willows, and poplars.

Control of woody growths in swamps can be based on chemical
or mechanical means, on water-level adjustments, or on combina-
tions of these methods. Chemicals most practical for the purpose
are 2,4-D, 2,4,5-T, "brush killer" mixtures of the two, Ammate,
and under certain conditions, urea compounds. Herbicidal treat-
ments include foliage spraying, basal spraying on the lowest foot
or two of tree trunks, notch treatment consisting of placement
of the chemical in notches of tree trunks, spraying or painting of
stumps with herbicides, and ground injections to poison roots.
Ground sterilants and basal treatments are of limited importance
where flooding causes dispersion of the chemical.

Mechanical control by means of heavy disks, brush cutters, or
similar equipment can be effective in sites suitable for such oper-
ations. One of the most effective ways of controlling woody
growths is by cutting, followed with flooding of stumps for one
growing season. Very few woody weeds can survive this treat-
ment.

An important limiting factor in the control of woody weeds,
particularly when applying herbicides to foliage, is insufficiency
of knowledge of the stage of growth when control can be applied
most effectively. For many species of woody plants, this vulner-
able period can be defined in general terms as occurring in late
summer or early fall during maximum foliage and before the
leaves begin to deteriorate. The best dates for making applica-
tions on a particular species in a given locality must be ascer-
tained for maximum efficiency. Applications made during other
than the most vulnerable period are likely to necessitate extra
treatments and costs.

52

WOODY WEED 53

Because of their importance in waterfowl habitat management,
alders, buttonbush, saltcedar, and willows are discussed separately
on the following pages. Additional trees and shrubs, mainly of
lesser importance, are covered briefly under the heading Other
Woody Weeds. Information on methods and materials for con-
trolling various woody plants can be obtained from the references
below.

REFERENCES

1955. Brush Control on Forest Lands. Walter G. Dahms and George A. James.

USDA Pac. NW. For. and Range Exp. Sta. Res. Paper 13.
1954. Controlling Hardwood Sprouts with Foliage Sprays. S. Clark Martin and F.

Bryan Clark. USDA Central States For. Exp. Sta. Tech. Paper 145.
1954. Controlling Woody Plants with 2,4,5-T; 2,4-D and Ammate. S. Clark Martin,

J. M. Nichols and Dayton L. Klingman. Univ. of Missouri Agr. Exp. Sta.

Bui. 615.
1953. Proceedings Lake States Forestry Clinic on Using Chemical Controls in Forest

Management. USDA Lake States For. Exp. Sta. Misc. Report 21.
1952. Herbicide Effects in Connecticut. Frank E. Egler. Jour. Forestry 50 (3).
1951. Chemical Control of Brush and Tree Growth for the Lake States. Paul O.

Rudolf. USDA For. Service Misc. Report 15.
1951. The Use of Chemicals to Control Inferior Trees in the Management of Loblolly

Pme. L. E. Chaiken. USDA SE For. Exp. Sta. Paper 10.
1950. Studies on the Use of Herbicides for Improving Waterfowl Habitat in Western

Kentucky and Tennessee. J. H. Steenis. Joiu-. Wild. Mgt. 14 (2).

54

ALDERS

BOTANICAL

Of nine alder (Alnus) species native to the United States, three
commonly produce thickets in bogs or swamps or occur as fringes
on the margins of ponds. These are rugosa (formerly incana)
and serrulata (formerly rugosa) in the East, and tenuifolia in
the West. Though alders occur as far south as Florida, their
principal abundance is in north temperate parts of the country.

IMPORTANCE

The value of alders for wildlife cover is limited, and their detri-
ment to waterfowl habitat is considerable, particularly in im-
pounded and dewatered units.

CONTROL

Tests by various organizations in different parts of the country
have shown that alders can be controlled fairly satisfactorily by
2,4,5-T, 2,4-D, or Ammate, though commonly two or more treat-
ments are necessary for eradication. Basal spraying (thorough
wetting of lowest 2 feet of trunks) was found effective on A. ru-
gosa Sit the Dunbar Forest Experiment Station, Mich. (Day,
1952). A solution containing 40 pounds of 2,4,5-T ester to 100
gallons of diesel oil gave 90% to 100% kill and 80 pounds of
2,4,5-T to 100 gallons of oil resulted in practically 100% control
in either summer or winter. A commercial mixture of 2,4-D and
2,4,5-T gave nearly as good results, but 2,4-D alone was not as
effective unless used at a 20% strength.

At the Patuxent Refuge a recent series of treatments involving
complete spraying of alder foliage with 1% 2,4,5-T or 2% silvex
in mid-August yielded effective control. Basal spraying of the
Pacific Coast tree alder (A. rubra) with 32 pounds of 2,4,5-T in
100 gallons of diesel oil gave a 96% kill (Ruth and Bernsten,
1956).

BUTTONBUSH 55

REFERENCES

1956. Chemical Basal Treatment to Control Red Alder. Robert H. Ruth and Carl
M. Bernsten. USDA Pac. NW For. and Range Exp. Sta. ReP. Note 128.

1952. Controlling Alder With Basal Stem Sprays. Maurice W. Day. Mich. State
Agr. Exp. Sta. Quart. Bui. 34 (4).

BUTTONBUSH

BOTANICAL

Our one native species of buttonbush, Cephalanthiis occidentalis,
is widely distributed in swamps and other wet places. It is com-
mon throughout the East and is particularly plentiful in the Lower
Mississippi region ; in the Southwest and California it occurs only
locally. The shrub tolerates fluctuating water levels, shade, and
acidity, and frequently develops dense, extensive thickets. Other
names are buttonwillow, buttonwood, and elbowbush.

IMPORTANCE

In many places, buttonbush does not deserve classification as a
weed. Frequently it is not abundant enough to be a serious com-
petitor with useful plants, and in some places the duck food
(seeds) and cover provided justify its presence. In many other
places, however, habitat can be improved for waterfowl by reduc-
ing or eliminating buttonbush thickets.

CONTROL

Mechanical eradication of buttonbush has been successful in
colder parts of its range. The water level is drawn down as low
as possible before freeze-up, and when the ice becomes safe the
plants are cut with a rotary brush mower. Later, the water level
is restored to keep the cut stems inundated during the growing sea-
son. The effectiveness of this technique depends on sufficient
depth and duration of submergence.

56 SALTCEDAR

The plant is rather difficult to kill with herbicides. This is par-
ticularly true when it is on exposed margins or other unflooded
sites. Effective control resulted from the use of Ammate at 60
pounds to 100 gallons of water in the Kentucky Woodlands
Refuge in flooded habitat (Steenis, 1950).

2,4,5-T is more effective on buttonbush than 2,4-D. Recent
tests on buttonbush control at the Patuxent Refuge in mid-August
have shown promising preliminary indications from treatments
with combinations of 50 pounds TCA and 4 pounds 2,4,5-T.

REFERENCE

1950. Studies on the Use of Herbicides for Improving Waterfowl Habitat in Western
Kentucky and Tennessee. J. H, Steenis. Jour. Wild. Mgt., 14 (.2).

SALTCEDAR

BOTANICAL

Several species of tamarisks (Tamarix), of about 75 that are
native to the Old World, have been cultivated as ornamentals in
this country. Of these, saltcedar (T. pentandra) , has become a
serious pest in the West during the past half -century. Normally
it is a shrub or small tree, but under favorable conditions it
grows 25 feet or higher. On reservoir margins, river deltas, and
other places that are moist part of the year, saltcedar forms
heavy thickets, sometimes pure stands extending for miles. More
frequently it is mixed with cottonwoods and willows. When their
tops are killed or cut, saltcedar trees regenerate readily by
sprouts. They also reproduce effectively by seed, frequently thou-
sands of seedlings appearing in one season.

SALTCEDAR 57

IMPORTANCE

Saltcedar has become the No. 1 wetland weed of the Southwest.
The tree is not only a detriment to waterfowl habitat on thou-
sands of acres of overflow and river-bottom land, but it also
causes serious economic loss (through transpiration) of great
quantities of water needed for irrigation, and harms grazing in-
terests by competing with grasses or other useful vegetation. In
addition, dense stands of saltcedar along some stream courses have
caused flood damage to adjoining land by impeding flow at high-
water stages. Bees get nectar from the flowers, cattle use the
thickets for shelter, and in some river basins the plant is consid-
ered desirable for erosion control, but on the whole it is much
more of a liability than an asset.

CONTROL

While saltcedar has been controlled quite effectively along irri-
gation and drainage canals by high-volume spray applications of
2,4-D, investigations by several agencies since 1951 have thus far
found no satisfactory, effective method of eliminating it from
river deltas, reservoir margins, flood plains, and other areas where
extensive thickets exist. Instead, tests have made it clear that
the plant is difficult and costly to control in these situations by
either chemical or mechanical means. Results with 2,4-D or
2,4,5-T sprays or with combinations of the two have been incon-
sistent. Generally, repeated treatments have been necessary, and
ordinarily fair to good results have been obtained only at dosages
of 2.5 or more pounds per acre. Basal spraying has been moder-
ately successful.

Hundreds of acres of Federal refuges have been cleared of the
pest in the past few years by use of heavy equipment (bulldozer,
bush-and-bog disk, and rotary tiller) and by follow-up mainte-
nance. In general, sod or cultivated crops must be sown after
clearance, and in addition saltcedar seedlings need to be sprayed
periodically to prevent reestablishment of thickets. Mechanical
control followed by spraying of regrowth is generally regarded
as the best means of combating saltcedar, as indicated in point 11
of the following summary from a report (Arle, 1957) by H. Fred
Arle of the Bureau of Reclamation :

1. Saltcedar is more difficult to kill on flood plain situations than along irrigation
channels and streams.

2. Single spray operations have never given satisfactory total plant kill of adult
ealtcedar and only rarely have two repeated treatments eUminated 80% or more of the
plants.

58 WILLOWS

3. Periodic spraying of infested areas with 2,4-D and 2,4,5-T will defoliate salt-
cedar and in this manner reduce transpiration losses.

4. Applications of 2,4-D and related materials appear more effective on young
regrowth following mechanical destruction than on adult saltcedars.

5. Apphcation rates of less than 2.0 pounds per acre have generally given poor
results.

6. Low volatile esters of 2,4-D or combinations of 2,4-D and 2,4,5-T have been
consistently more effective than amine or sodium salts of 2,4-D.

7. Dormant apphcations of 2,4-D and 2,4,5-T esters have shown promise in the
control of saltcedar.

8. Mechanical means, although expensive, are useful in the eradication of salt-
cedar, especially in areas near cotton or other crops susceptible to 2,4-D.

9. Mechanical control must be exercised at least yearly to eliminate regrowth from
root sprouts and seedlings.

10. Saltcedar is more difficult to kill with 2,4-D and related materials than most
willows, cottonwoods and other woody phreatophytes.

11. Mechanical clearing followed by spraying of young regrowth with 2,4-D or a
mixture of 2,4-D and 2,4,5-T at 2.5 pounds or more per acre repeated as necessary
once or twice a year appears to be the most effective and practical method now known
for controlling saltcedar.

REFERENCES

1957. A summary of results of experimental and field trials pertaining to the control
of saltcedar {Tamarix gallica). H. Fred Arle. Proc. Amer. Geophysical
Union.

1957. History of the Introduction of Saltcedar into the United States. Ciu-tis W.
Bowser. Proc. Amer. Geophysical Union.

1957. The Phreatrophyte Problem. T. W. Robinson. Proc. Amer. Geophysical
Union.

1951. Report to Saltcedar Interagency Council. Interagency Task Force, Albu-
querque, N. Mex.

WILLOWS

BOTANICAL

The numerous species of willows (Salix) that invade water-
fowl habitat in various parts of the continent are particularly
troublesome where fluctuating water levels favor establishment of
seedlings. Most willows grow rapidly and are capable of propa-
gation by seeds, by shoots from roots, or by broken-off twigs or
branches rooting in the mud.

IMPORTANCE

Willows have minor value to wildlife and to man in various
ways. In the North, elk, moose, deer, beaver, and several species

WILLOWS 59

of grouse, feed extensively on the bark, twigs, and buds. In re-
gions of the West where cover is scarce, willows sometimes provide
useful protective or nesting cover for game birds. Locally, wil-
lows have been used for controlling soil erosion on stream banks.
Their light, pliable wood has some value for cooperage and other
purposes. On the other hand, besides crowding out duck-food
plants in waterfowl areas and necessitating the spraying of sev-
eral hundred acres of willows on Federal refuges each year, wil-
lows cause serious water losses on margins of western reservoirs
and irrigation canals.

CONTROL

Herbicides, particularly 2,4-D formulations, are the most prac-
tical means of destroying willows. On black willow {S. nigra),
control ranging from 95% to 100% may result from a late-summer
foliage spraying with dosages of 4 pounds 2,4-D to 100 gallons
of water. Effectiveness of treatment varies with the species, and
in time it may be possible to give specific recommendations for
each of the more important kinds of willows. Glossy-leaved spe-
cies are likely to require a "sticker" in the spray. Injection treat-
ments (into cuts on trunks) and basal spraying are generally not
recommended for willows except for larger species such as the
black willow or in instances where specimens are too tall or too
few to justify foliage spraying.

Control recommendations based on studies by various organiza-
tions in western States have been summarized in the 13th Western
Weed Conference Research Progress Report (Butler, 1952). Some
of the principal findings reported are as follows:

2,4-D is more effective on willows than the more expensive 2,4,5-T.

Both the amines and the esters of 2,4-D are satisfactory at comparatively low rates.

Aerial spraj'ing using large droplets has been effective in applications as low as 3
pounds of acid equivalent in one gallon of oil per acre.

Early summer applications seem best for foliage sprays.

Old willows are likely to require treatment for 2 or 3 successive years. The dead
tops can be felled and burned more readily during or following the second winter after
the original treatment. Such removal is desirable for efficient treatment of sprouts
and for final eradication.

For land spraying of foliage, 100 to 400 gallons of solution per acre may be required
to wet foliage effectively. In contrast to this, airplane spraying of willows in the
Strawberry Valley in Utah was effective with 3 pounds of acid equivalent in one gallon
of solution per acre.

A hehcopter proved very satisfactory for spraying willows along canals near Malta,
Mont.

REFERENCE

1952. Undesirable Woody Plants on Irrigation Systems and Irrigation Lands. C. C.
Butler. Res. Prog. Report, 13th Western Weed Conf.

60

OTHER WOODY WEEDS

Plants discussed under this heading are given only limited
space because of their lesser importance and because, in some in-
stances, little is known about their control.

Along permanently or seasonally flooded areas in western Ken-
tucky, small or medium-sized sprout growths of red maple {Acer
ruhrum) , green ash (Fraxinics lanceolata) , and sweet gum {Liquid-
o.mbar styraciflua) have been controlled satisfactorily with various
formulations of 2,4-D salts and esters at the rate of 4 pounds per
100 gallons of water applied in late summer (Steenis, 1950).
Equivalent results were obtained with Ammate at 75 pounds per 100
gallons. However, in better-drained sites, such treatment produced
only partial control of these species. Trumpetvine or cowitch
{Tecoma radicans), a woody creeper that is a problem plant on
islands and seasonally flooded margins of TVA impoundments and
other Southeastern reservoirs was controlled readily with the same
dosage of 2,4-D or with Ammate when treatment was made in
early fall (August 15 into September).

Recent tests in Maryland have shown that both of the common
saltmarsh shrubs, groundselbush {Baccharis halimifolia) and high-
tide-bush {Iva frutescens), can be controlled by 2,4-D formula-
tions at about 4 pounds per 100 gallons if treated in late summer
or early fall. Also, these plants and bayberry {Myrica) were
killed readily in coastal North Carolina by monuron pellets scat-
tered sparsely around the bushes.

In Delaware, the semi woody weed decodon {Decodon verticil-
latus) has been controlled readily by 4 pounds of 2,4-D per 100
gallons. The treatment proved effective during different parts
of the growing season, from June into August.

Recent studies on leatherleaf {Chamaedaphne calyculata) by
the State of Wisconsin have indicated that this common shrub of
northern bogs, can be controlled up to about 90% by a single
summer treatment of 2,4-D or 2,4,5-T at 4 pounds acid equivalent
per 100 gallons. Higher poundages of the herbicides yielded
higher percentages of kill.

REFERENCE

1950. Studies on the Use of Herbicides for Improving Waterfowl Habitat in Western
Kentucky and Tennessee. John H. Steenis. Jour. Wild. Mgt. 14 (2).

•^ U. S. GOVERNMENT PRINTING OFFICE 1958—429176

WHOI Libi;

|"iKHSE "OOOSS