UNIVERSITY OF FLORIDA
Agricultural Experiment Station
CONTROL OF THE
VELVET BEAN CATERPILLAR
J. R. WATSON
Fig. 11. Moth of velvet-bean caterpillar.
The Station bulletins will be sent free upon application to the Experiment
Station, Gainesville, Fla.
BOARD OF CONTROL
P. K. YONGE, Chairman, Pensacola, Fla.
T. B. KING, Arcadia, Fla.
E. L. WARTMANN, Citra, Fla.
W. D. FINLAYSON, Old Town, Fla.
F. E. JENNINGS, Jacksonville, Fla.
P. H. ROLFS, M.S., Director.
J. M. SCOTT, B.S., Vice Director and Animal Industrialist.
B. F. FLOYD, A.M., Plant Physiologist.
J. R. WATSON, A.M., Entomologist.
H. E. STEVENS, M.S., Plant Pathologist.
S. E. COLLISON, M.S., Chemist.
JOHN BELLING, B.Sc., Assistant Botanist and Editor.
C. D. SHERBAKOFF, Ph.D., Associate Plant Pathologist.
S. S. WALKER, M.S., Associate Chemist.
F. F. HALMA, B.S., Assistant Horticulturist.
H. L. DOZIER, B.S., Laboratory Assistant in Entomology.
JULIUS MATZ, B.S., Laboratory Assistant in Plant Pathology.
H. G. CLAYTON, B.S.A., Laboratory Assistant in Dairying.
K. H. GRAHAM, Auditor and Bookkeeper.
T. VAN HYNING, Librarian.
E. G. SHAW, Secretary.
L. T. NIELAND, Farm Foreman.
COMMITTEE ON PUBLICATIONS
John M. Scott, H. E. Stevens and B. F. Floyd.
1. This caterpillar is the only serious insect enemy of velvet beans in Florida.
2. The egg hatches in 3 days; the caterpillar grows 3 weeks; the pupa stage lasts
3. The insect does not survive the winter. Fields are reinfested each summer, by
moths from the south.
4. It has numerous natural enemies which should not be molested.
S. A flock of turkeys helps to control the caterpillars.
6. When early varieties of velvet beans, such as the Chinese, can be grown, a strip
around the edges of the field should be sown with the Florida variety as a trap crop.
7. This crop should be sprayed or dusted with lead arsenate every two weeks dur-
ing the caterpillar season, and when it is necessary the main crop should be similarly
8. Fields should be watched for the first appearance of moths and preparations
should be made to spray or dust.
Life history of the insect -------------.... -----------------------------50
Migration and distribution..----------.--------- -----------------------51
Food of the caterpillar ------------.. ----------------------------------52
Methods of control ------------------------------------------------------53
Early preparation -----..------------------------------------------53
Spraying and dusting..---------- ------.------------------------- 54
Formulas for spraying and dusting.-------------------------------55
Control by enemies ----------------------------------------------- 56
Cholera ------- -------------------------------------------------58
Control of the Velvet-Bean Caterpillar
BY J. R. WATSON
The only serious insect enemy of velvet beans in Florida is
the caterpillar of the moth, Anticarsia gemmatilis (Fig. 11), which
eats the leaves (Fig. 12). The damage from this insect is usually
Fig. 12. Leaf eaten by the caterpillars.
Florida Agricultural Experiment Station
severe and often disastrous, The entire seed crop is sometimes
destroyed. Aside from soil improvement, it is chiefly for the
seeds, or seeds and pods for winter forage that velvet beans are
grown in Florida. They are not generally used as green forage.
Since the plant normally produces much of its growth after late
August or September, the stripping of the vines at that time cur-
tails the soil-improving effect of the crop as well as its seed
The severity of the infestation varies in different years and
also in different fields and localities. This is due chiefly to the
activities of the insect's natural enemies. In general the severity
of injury increases southward, because the insects get an earlier
start in the southern part of the State.
The injury is proportional to the size of the field if other con-
ditions are equal. Greater injury occurs in large fields, because
the caterpillars become so numerous that their natural enemies
(principally birds) cannot control them. In a small field, enemies
of the insect come in from the surrounding woods and fields and
usually keep them in check.
Fortunately the caterpillars do not appear in disastrous num-
bers until August or September in the large velvet bean growing
sections of the State. In October at least, and often in September,
they are brought under complete control by "cholera," a fungus
disease. Consequently it is necessary only to supply a little aid
to the natural enemies at a critical period.
Few farmers take any measures of control. They trust that
the velvet beans will be able to survive and produce some seed
in spite of the caterpillars. This bulletin is intended to show
that by taking advantage of the food preferences of the cater-
pillar and of its natural enemies, and by the judicious use of
poison, the damage can be reduced to such an extent that velvet
beans will be a dependable crop.
Altho it is a serious pest, the velvet bean caterpillar can be
controlled at a comparatively low cost. No one needs hesitate
to plant velvet beans on account of the ravages of this insect.
LIFE HISTORY OF THE INSECT
The eggs are small white, roundish bodies which are about
one-twelfth of an inch in diameter. (Fig. 13.) The majority of
them are laid on the lower surfaces of the mature leaves. The
egg hatches in about three days. The young larva is about one-
tenth of an inch long. It feeds on the leaves about three weeks,
during which it molts (casts its skin)
five times and grows to nearly two
inches long. After it is half grown
it is usually dark green with promi-
nent bright colored lines with darker
borders running lengthwise of the
body. (Fig. 14.) Many of the cater-
pillars, however, are pale green and
the lines are either indistinct or
entirely absent. The line along the
side is wider than the others and
is often pink or brown. The cater-
Fig. 13. Eggs on leaf. pillar has no conspicuous hairs. If
disturbed, it throws itself about violently until it reaches the
When full grown, the caterpillar enters the ground where it
constructs an earthen chamber in which, after a final molt, it
passes into the pupa stage.
The pupa is brown and
three-fourths of an inch
long. During September
the insect remains in the
pupa case about ten days Fig. 14. Fifth Stage Caterpillar.
before emerging as an adult moth. As the weather becomes
cooler the time is greatly lengthened, but in no case has the
insect been observed to remain in the pupa stage all winter.
The moth, too, is variable in color but is usually some shade of
gray or brown. A characteristic mark and one that will enable the
farmer to distinguish this moth from any other is the double line
that extends diagonally across both wings. (Fig. 11.) The moth
is about an inch and a half across the outstretched wings.
MIGRATION AND DISTRIBUTION
One of our most interesting discoveries concerning this in-
sect is that it is migratory like the moth of the cotton caterpillar.
It does not winter in North or Central Florida, but flies north
each summer from the southern end of the peninsula or perhaps
The most important practical result of this discovery is that
one can predict the coming of the caterpillars. Since the moths
are known to appear in a field before the caterpillars, the grower
can foretell almost to a day when the caterpillars will begin to
Florida Agricultural'Experiment Station
damage his crop. He needs only to be able to recognize the
moths and to watch for their appearance. Since the eggs hatch
in three days, and the caterpillars do little damage until after the
second molt, an abundance of moths in a field means that it will
be necessary to dust or spray in about twelve days. The grower
who finds his field swarming with moths should order his mate-
rials at once.
The flight of the moths northward can actually be recorded
and predicted in the same manner as the progress of a storm is
watched and predicted by the Weather Bureau.
FOOD OF THE CATERPILLAR
The writer has found the caterpillars feeding on but three
plants. In order of the severity of infestation, they are: velvet
beans (Stizolobium sp.), kudzu vine (Pueraria thunbergiana),
and horse beans (Canavalia sp.).
Some varieties and species of velvet beans are evidently pre-
ferred to others. The Florida velvet bean is always much more
severely damaged than the Chinese when the two are planted
side by side. On the Experiment Station grounds they fre-
quently occupy neighboring plots, where unusual opportunity
is afforded to study the comparative severity of infestation.
The early varieties have usually flowered before the cater-
pillars become abundant.
Some notes on the comparative amount of damage done to
different varieties, or species, of Stizolobium when planted side
by side, were made September 9, 1913, at the Station Farm. Four
varieties, Wakulla, Alachua, Yokohama and Florida, were used
in the test. Wakulla is a very early variety, and matures at the
same time as the Yokohama, the earliest of the genus. Alachua,
another selection from a cross, matures one or two weeks earlier
than the Florida. There were three rows of each kind, and they
stood in the field in the order given in the following table, which
shows the comparative damage to the different varieties:
Variety Maturing Damaged by the Caterpillar
Wakulla Very early Little
Alachua Late Considerably
Yokohama Very early Very little
Wakulla Very early Little
Florida Very late Heavily
Wakulla Very early Little
Alachua Late Badly
Wakulla Early Slightly
Florida Very late Very heavily
Wakulla Very early Little
Alachua Late Badly
Yokohama Early Slightly
Wakulla Very early Hardly touched
The preceding discussion applies only to those cases where
the varieties are grown close together. When a large field con-
taining thirty or forty acres of Chinese velvet beans is compared
with another large field of Florida velvet beans, there is less dif-
ference in the damage. Even in this case, however, there is
usually a difference in favor of the Chinese.
METHODS OF CONTROL
EARLY PREPARATION NECESSARY
1. The farmer should begin to fight this pest at planting
time. If Chinese or early Georgia velvet beans are as suitable as
the Florida, the main crop can be planted to them or to some of
the new early varieties originated at the Experiment Station,
such as the Osceola and the Wakulla. Because of early maturity
and probably less attractiveness to the moths, these are damaged
less severely than the Florida velvet. If the tendency of the
Chinese to shell is a serious objection (as when it is used as a cat-
tle food) one of the other kinds, such as the early Georgia, should
be planted. As a further protection some of the Florida velvet
beans should be planted in the vicinity to attract the moths away
from the early varieties. This trap crop should be distributed
about the fields so that it will not be too far away (certainly not
more than an eighth of a mile) from any part of the main field.
The trap crop should be planted in accessible places so that it can
be readily sprayed or dusted.
2. A flock of turkeys will consume vast numbers of cater-
pillars and other insects, especially grasshoppers.
3. Birds, wasps, and skunks should not be molested. All are
useful destroyers of insects. Birds and skunks feed on grass-
Florida Agricultural Experiment Station
hoppers also, which, after the caterpillars, are the most destruc-
tive insects in a velvet bean field.
4. A careful watch should be kept for the first moths. The
farmer can distinguish this insect from any other common Flor-
ida moth by the (usually double) diagonal line which stretches
across both wings and turns up to the apex of the fore wing
(Fig. 11), the appearance of the under side of the wings, and the
peculiar darting flight. These moths may be expected during
July in South Florida, during August in Central Florida, and dur-
ing late August and early September in the extreme northern and
western sections of the State.. When the moths are noticed in
large numbers in the fields, it is probable that the beans will need
to be sprayed or dusted after twelve days or two weeks. The
presence of moths should, therefore, be a signal to the farmer to
obtain spraying materials.
CONTROL BY SPRAYING AND DUSTING
The application of lead arsenate or zinc arsenite is the best
means known for controlling the caterpillar. It will be well to ob-
tain these in the powdered form, because the powder is more uni-
form in composition than the paste, especially when the paste has
lost some of its water. Paris green should not be used on velvet
beans, as they are easily burned. A dosage of paris green strong
enough to kill a large percentage of the caterpillars is sure to
severely damage the vines. Even with the dosage of lead arse-
nate recommended here, the leaves will be burned sometimes.
However, this burning will be confined to old leaves that have
almost fulfilled their mission, and no serious damEge will result.
Contrary to the general rule the young foliage of velvet beans
is less easily burned than the old. A young and vigorous leaf
is evidently more able to withstand the poison.
Owing to the late appearance of the caterpillar and the al-
most sure development of "cholera," there is usually not more
than a month during which the grower will need to protect his
crop. It is not always necessary to treat the entire field. If the
most severely infested portion is treated, the birds will congre-
gate on the untreated portion and often hold the caterpillars in
The total cost of spraying at the Station in September, 1915,
was $1.10 an acre, while dusting at the same time and place cost
80 cents an acre for one application. It has never been necessary
to repeat the spray. At least two careful dustings are required
for the same protection, which makes the cost 50 cents an acre
more than spraying. But in order for spraying to be practical, a
good barrel spray-pump and water must be available. It is
usually difficult to drive Ihru a velvet bean field with a wagon,
altho in many cases the grower can leave a road every hundred
feet at planting time. The damage done by driving thru the vines
when they are running over the ground without support is not
as great as might be supposed. A week later it will hardly be
FORMULAS FOR SPRAYING AND DUSTING
Not more than twelve ounces of powdered lead arsenate (or
a pound and a half of the paste) to fifty gallons of water can be
safely used. Even with that small amount one should put a
pound and a half of quick-lime (or two quarts of fresh lime-sul-
phur solution) in the water and should keep the liquid well
agitated while spraying.
For a spray we recommend:
Lead arsenate, powder -----------.. ----------..12 ounces
Quick-lime __------.---------------------------. 1 pounds
Water .-------------------------..----------- 50 gallons
If the paste form of lead arsenate is used, take 24 ounces instead
of 12. This amount should suffice for nearly an acre.
The dry arsenate when used as a dust should be mixed with
about four times its volume of air-slaked lime. A coarse burlap
bag is tied to each end of an eight-foot pole, and filled with the
mixture. A man on a mule then takes the pole with the bags
and rides across the field, dusting the plants by constantly jarring
the pole. At least fifteen pounds of the mixture (three pounds of
lead arsenate or zinc arsenite) should be applied to the acre.
DUSTER MORE SATISFACTORY THAN BAGS
A more even and satisfactory method of spreading the dust
is by means of a "blower" or dusting machine.. Even a careful
man using the bags and pole will cover scarcely more than half
of the surface of the leaves and will get the dust too thick in
places; our experience has been that with ordinary labor but
little more than a third of the leaves are dusted. More time is
required to cover the field with a dusting machine, but the added
thoroness more than repays the added cost of labor. A careful
man is able to do nearly as thoro work with the duster as he is
with a spraying outfit and at a smaller cost.
Florida Agricultural Experiment Station
There are several makes of "knapsack dusters" which cost
ten dollars or more. These are best operated by a man on foot
who can cover a strip about twenty feet wide by dusting on both
sides. If there is any wind, it is better to dust only on the lee-
ward side to avoid inhaling the mixture. It is better to walk
across the field in a direction at right angles to the wind. A
large acreage will justify the purchase of a dusting machine. Of
course with a duster that will throw a sufficient amount one can
do more efficient work.
Dusting should be done in the early morning or after a
shower, while the vines are wet. The mixture sticks so well that
much of it remains after a heavy rain. It will be necessary to
redust every ten days or two weeks as long as the caterpillars are
abundant in order to cover the new growth which will have put
out. On the Experiment Station grounds we have never found
it necessary to make more than three applications.
If half of the caterpillars can be poisoned their numerous
enemies can usually be trusted to destroy a good percentage of
those that escape. In fact these enemies are always the real
controllers of an outbreak. The farmer with his arsenate only
helps them a bit at a critical time. Except in especially favor-
ably located fields, such as small ones near woods, it will not do
to depend entirely upon these enemies. Such a policy may mean
the loss of an entire crop, and will usually mean a reduction in
yield, which will be much more costly than the application of
After one or two rains it will be perfectly safe to allow stock
to eat the poisoned vines. As stock is usually not turned in until
the pods are mature, months after the application of the poison,
there can be no possible danger of poisoning the animals even
if there has been no rain meanwhile. All the leaves which were
poisoned will have died and fallen, carrying the poison to the
ground where it soon loses its potency. Usually the pods will
not have appeared at the time the poison is applied and conse-
quently will carry no poison.
CONTROL BY ENEMIES
The caterpillars have many natural enemies. One of the
most important is the "rice-bird," also called "blackbird," or "red-
and-buff-shouldered-marsh-blackbird." These collect in great
flocks in infested fields. Other birds, especially mocking-birds,
eat many of the caterpillars. It is probably on account of birds
alone that small patches of velvet beans planted near woods
usually escape with little injury.
Lizards, especially the "chameleon" (Anolis), feed eagerly
upon the caterpillar. The Anolis is commonly seen climbing
over the vines in velvet bean fields. They doubtless consume
a great number of the caterpillars.
Polecats or skunks are frequently found in the velvet bean
fields and probably feed on the caterpillars and pupae, since they
are fond of insects. They are among the most useful of wild
animals in this respect.
Wasps of certain species carry off many caterpillars with
which to stock nests for their grubs.
Perhaps the most important insect enemies of the caterpillars
are certain species of predaceous bugs (Figs. 15 and 16). These
bugs are abundant in
\ / velvet bean fields, and \
\ / are commonly seen
with caterpillars im-
paled on their beaks,
or slowly and stealth-
ily stalking their prey.
Since they attack most-
ly the smaller caterpil-
lars they do a great Fig. 16. Euthrynchus
deal of good, as they floridensis.
Fig. 15. A predaceous many in a day. Morever, by destroying
bug (Mutica grandis). the young caterpillars the bugs save more
velvet bean leaves than they would if they took the older
caterpillars which have already done most of their damage.
A small bluish carabid beetle (Callida decora) (Fig. 17) is
active in destroying eggs and young caterpillars.
It is frequently seen running actively over the I
A number of predaceous enemies also prey
upon the pupae in the ground. Common among
these are moles and large carabid beetles Say's \
Hunter. (Fig. 18.) The former is seldom seen,
but its tunnels are everywhere under the vines.
The latter is nocturnal and is found during the Fig.17. The lit-
day under the dead leaves. It, too, is seldom tie blue hun-
seen, altho it is common and highly bene- ter (Callida de-
Florida Agricultural Experiment Station
Probably the only practical measure the farmer can take to
aid these:natural enemies of the caterpillars is to see that they
are unmolested. Birds, wasps and
skunks, which are commonly per-
secuted, should be protected. They
are among the farmer's best allies.
It is true that skunks have an uri-
fortunate appetite for poultry, but
the poultry can be kept safeguard-
ed at night.
Turkeys are fond of insects of
all kinds, and, because they are
prone to wander, are particularly
valuable on the farm. If possible
a farmer should keep a flock for
their insectiverous value, even if
they do not bring large returns at
Dragonflies capture many of the Fig. 18. Say's Hunter (Calosoma
CONTROL BY CHOLERA
By far the most efficient check on the increase of this pest
is a disease called "Cholera" by farmers. This is caused by the
fungus (Botrytis rileyi) (Fig. 19). In October, 1914, and again
in 1915, and also in previous
years, this fungus almost ex-
terminated the caterpillars in
the fields around Gainesville.
Less than one-tenth of one per
cent escaped. On the Experi-
ment Station grounds where
they had been numerous
enough to destroy much of
the crop, the caterpillars be-
came scarce in one week.
Fig. 19. Caterpillar killed by "cholera."
This is not unusual, but oc-
curs almost every year. Sooner or later the fungus appears and
nearly exterminates the caterpillars, tho it is often too late to
save the crop. After it becomes established in the field, the fungus
seems to control the insect for the remainder of the season. The
fungus to become epidemic seems to require a cool, prolonged,
rainy period, such as usually occurs in late September or October.