'!.i *.. J
'i .. . f
U. S. DEPARTMENT OF AGRICULTURE,
BUREAU OF ENTOMOLOGY-BULLETIN No. 85, Part IV (Revised).
L. 0. HOWARD, Entomologist and Chief of Bureau.
PAPERS ON CEREAL AND FORAGE INSECTS.
W. HARPER DEAN,
Agent and Expert.
ISSUED APRIL 18, 1911.
GOVERNMENT PRINTING OFFICE.
t .. .
:. .. ... .
E:..:: Lk .
B UREA U OF ENTOMOLOGY
L. 0. HOWARD. Entomologist and Chief of Bureau.
C. L. MAfLATT, Entomologist and Acting Chief in Absence of Chief.
.R. S. CLIFTON, Executivre Assistant.
W. F. TASTET, Chief Clerk.
F. H. CHITTENDEN, in charge of truck crop and stored product insect inv restigations.
A. D. HOPKINS, in charge oj.forest insect investigations.
W. D. HUNTER, in charge of southern field crop insect investigations.
F. M. WEBSTER, in charge of cereal and forage insect investigations.
A. L. QUAINTANCE, in charge of deciduous fruit insect investigations.
E. F. PHILLIPS, in charge of bee culture.
D. M. ROGERS, in charge of preventing spread of moths, field work.
ROLLA P. CURRIE, ill charge of editorial work.
MABEL COLCORD, librarian.
CEREAL AND FORAGE INSECT INVESTIGATIONS.
F. M.. WEBSTER, in charge.
GEO. I. REEVES, C. N. AINSLIE, J. A. HYSLOP, WV. HARPER DEAN, V W. R. WALTON,
T. D. URBAHNS, V. L. WILDERMUTH, R. A. VICKERY, T. H. PARKS, HERBERT T.
OSBORN, agents and experts.
W. J. PHILLIPS, E. 0. G. KELLY, GEO. G. AINSLIE, PHILIP LUGINBILL, entomological
NETTLE S. KLOPFER, preparator.
a Resigned March 1, 1910
Introduction ............... ...............................................
History of the sorghum midge in America....................................
D distribution ..............................................................
Investigations carried on in Louisiana and Texas during 1908 and 1909........
H ost plants ...............................................................
Descriptions.......... ............................................ ..- ....
The adult ..............................................................
The pupa ....................................
The "cocooned larva".........................
Life history and habits...........................
Location of the egg............................
Location of the larve ...........................
Location of the pupa.........................
Number of generations ........................
The life cycle .............. ........................
Length of life of adult .........................
Relation of Johnson grass to the midge problem.....
Destruction of Johnson grass ............................................
Clean harvesting.............. ........................................
Destroying heads of first crop..........................................
Bagging heads .......................................................
.... ... 48
.- -...... 52
...- ... 53
ILL US TRAT IONS.
PLATE I. Fig. 1.-Early Johnson grass heading and blooming on wet banks of
sewage canal, San Antonio, Tex. Fig. 2-.-Rubbish left
by removal of sorghum shocks in a field near San An-
I. Fig. 1.-Harvested field of sorghumn near San Antonio, Tex., show-
ing scattered heads. Fig. 2-.-Harvested field of sor-
ghum near San Antonio, Tex., in which the sorghum
midge is hibernating. Fig. 3.-Fence line bordering a
by removal of sorghmum shocks in a field near San Antonio-, Tex., allowed to grow
toup in Johnson grass. .......................... ......... 54
II. Fig. 1.-Htarvested field of sorghum near San Antonio, Tex., show-
ingFIG. 20. Sorghumscattered heads: Partly destroyed by English sp2.-arrows, invested field of sor-by
ghum near San Antonio, Tex., in which the sorghum
midge is hibernating. Fig. 3.--Fence line bordering a
sorghum field near San Antonio, Tex., allowed to grow
up in Johnson grass. . . .. . .. .. .. 58
FIG. 20. Sorghum heads" Partly destroyed by English sparrows, infested by
sorghum midge (Contarinia sorghicola), healthy ................... 40
21. Base map showing present distribution of the sorghum midge .....---... 42
22. Growing sorghum head, bagged and tagged after natural infestation by
the sorghum midge .............................................. 43
23. Breeding cage suspended over growing sorghum...................... 44
24. The sorghum midge (Contarinia sorghicola): Adult male and details.. 45
25. The sorghum midge: Adult female and details, eggs, larvae, pupa... 46
26. The sorghum midge: "Cocooned larvae," or hibernating form ......... 48
27. Sorghum heads, showing four successive stages in the opening of the
sheath or "boot" .............................................. 50
28. Sectional views of the sorghum seed during the flowering stage ....... 51
29. Aprostocetus diplosidis: Adult female- ................----..........--..... 56
30. Aprostocetus diplosidis: Adult male ...........----....................... 56
31. Tetrastichus sp.: Adult female ...................................... 57
U. S. D. A., B. E. Bul. 85, Part IV (Revised).
PAPERS ON CEREAL AND FORAGE INSECTS.
THE SORGHUM MIDGE.
(Contarinia [Diplosis] sorghicola Coq.)
By W. HARPER DEAN,
Agent and Expert.
Sweet sorghum, aside from its use in making molasses, ranks as
one of the most important forage crops grown in the United States.
It is highly prized as a green food for cattle and horses and is well
adapted to entering the composition of silage. Several crops are
sometimes produced during the season, the last fall crop generally
being cured as a dry winter fodder. In Louisiana and Texas, while
this crop is grown practically over the entire States, no large areas
are cultivated, but it is found in small, isolated blocks ranging in
extent from one-fourth acre upward.
In parts of a great many sorghum-growing States the seed do not
mature a profitable crop, and while this may be attributed, and
rightly, too, to a number of causes, it is safe to say that in the majority
of cases the sorghum midge, Contarinia sorghicola Coq. (figs. 24, 25),
is directly responsible for the damage to the seed.
Such destructive agencies as various pathological diseases, the
English sparrow, the moth Nigetia sorghiella, and the rice weevil
(Calandra oryza L.) all help to curtail the number of sound, mature
seed produced, but by far the most destructive agency that has been
observed by the writer is this minute fly, the midge, which breeds
in swarms from the time the first heads have bloomed until the last
have been killed by cold.
An examination of damaged seed in sections where the midge is
known to occur will reveal the minute larvae of this fly lying close
alongside the ovary, which is blackened and shriveled (fig. 20, b),
while the ovary in healthy mature seed is plump and white (fig. 20, c).
Injury by the English sparrow can be readily distinguished upon
examination from injury by other agents. Such heads have a shat-
C. F. I. I., April 18, 1911.
CEREAL AND FORAGE INSECTS.
tered, frayed appearance, resulting from the seed being torn from
the outer glumes (fig. 20, a), while those seed injured by the midge
have a dark, flattened appearance (fig. 20, b), and minute pink and
white larvae can be seen, sometimes as many as six of them, lying
close alongside the ovary. These are the larvae of the sorghum
HISTORY OF THE SORGHUM MIDGE IN AMERICA.
The earliest reference to the sorghum midge occurs in 1895 in a report
by Mr. D. W. Coquillett, of the Bureau of Entomology, United
FIG. 20.-a, Sorghum head partly destroyed by English sparrows; b, sorghum head in-
fested by the sorghum midge (Conlarinia sorghicola), showing characteristic flattened
appearance of spikelet; c, sorghum head with matured healthy seed. (Original.)
States Department of Agriculture, in which he described the appear-
ance of several heads of sweet sorghum received from Dillburg and
Montgomery, Ala. These heads contained a large number of seed
which had failed to mature and which had apparently been destroyed
by the larvae of some species of Cecidomyiidae. However, only the
empty pupal skins attached to the spikelets of the seed were present,
so no clue to the identity of the species was given. This note was
THE SORGHUM MIDGE.
made October 2, 1895, and nothing further was heard from this
insect until September 26, 1898, when Mr. Coquillett received a sec-
ond lot of heads from College Station, Tex. The latter contained
the larvae of this insect, as well as a large number of the flies them-
selves. Mr. Coquillett decided that these belonged to an undescribed
species, and forthwith he published the first technical description of
the midge under the name Diplosis sorghicola, new species.a
In a special article in Science, published January 17, 1908, Prof.
Carleton R. Ball, of the Bureau of Plant Industry, United States
Department of Agriculture, discussed his experiments with the
midge in Louisiana and Texas during the spring of 1907. These
experiments resulted from an investigation of the causes of sterility
of sorghum seed.
Working under the direction of Prof. Wilmon Newell, of the
State Crop Pest Commission of Louisiana, Mr. R. C. Treherne con-
ducted a systematic investigation of the midge during the spring
and summer of 1908, and later published a brief summary of the
Prof. Glenmn W. Herrick, entomologist at College Station, Tex.,
contributed to the information on this insect in an article pub-
lished in Entomological News.c
At the time when official recognition was taken of the importance
of this insect by the Bureau of Entomology, Prof. F. M. Webster,
under whose department this investigation fell, was unable to begin
the work through lack of available field assistants. In order that
no time should be lost in making this beginning, however, Professor
Webster arranged with Prof. Wilmon Newell, of the State Crop
Pest Commission of Louisiana, to pursue a cooperative investigation
until such time as he could relieve him of the bulk of the work.
Accordingly, Professor Newell assigned one of his assistants, Mr.
R. C. Treherne, to the work, which was systematically conducted
during the spring and part of the summer of 1908 until the writer
was assigned the problem under the direction of Professor Webster.
On July 25, 1908, Mr. Treherne discontinued the work and at that
point it was taken up by the writer at Baton Rouge, La.
Professor Newell kindly allowed the writer unlimited access
to his most complete laboratory and offices, and these were used
as headquarters during the Louisiana investigation. In many
other ways, too numerous to mention, he contributed to the progress
of the work. It is therefore evident that but for this hearty coopera-
a Bul. 18, n.s., Div. Ent., U. S. Dept. Agr., p). 82, 1898.
b Thirty-ninth Annual Report of the Entomological Society of Ontario, 1908.
c Entomological News, vol. 20, No. 3, pp. 116-118, pl. 7, Mar., 1909.
42 CEREAL AND FORAGE INSECTS.
tion of Professor Newell the investigation of the sorghum midge
by the Bureau of Entomology would have been much delayed. I
Acknowledgment is due Prof. C. S. Scofield, of the Bureau of
Plant Industry, who rendered valuable assistance to the investigation
in Texas by placing the experiment farm near San Antonio, Tex.,
at the service of the writer for growing varieties of sorghum used .j
in the experiments.
During 1908, in Louisiana, and 1909, in Texas, the investigation
was continued, and the results constitute the basis for this report.
The distribution of the sorghum midge in the United States, as given
under this chapter, does not necessarily include all infested territory.
-Zy "a a .... ..
'a a\ I-'1--'Vi^
,r-.-- ...4.._,"'. ,,,--,. "
/ -- -- I -:B et / _.. .'
s* 1. ; "i ^"y--v
# t L.
I --- I >: i /,-"r I v
FIG. 21.-Base map showing present distribution of the sorghum midge. Dots indicate infested local
r-- ---- --.-*... I- i .. _.. ~
.... l---"i; k' --
Through examinations made in person in the fields or through the
actual rearing of the midge from heads received from various sections,
these data have been collected and will of course be added to from
time to time. The sorghum midge is not known to exist west of the
one hundredth meridian, and its extreme southern point of distribu-
tion in the United States is Brownsville, Tex. The accompanying
map (fig. 21) shows the general distribution as know-n at this writing.
INVESTIGATIONS CARRIED ON IN LOUISIANA AND TEXAS DURING 4
1908 AND 1909. I
During the years 1908 and 1909 the writer pursued the investigate RP
..... I4 I
I 2- I1
tion of the sorghum midge in Louisiana and Texas. Theso investi-
gations were in the main confined to the field, although certain laborat-h
actua remain o themidg frmhasrcieS ro aiu etbs
FIG. 21i.-Bs map showing presgent ra distribution oftesogu inide otsndiate tinste localng
Thrurigh texyamiatos m98ande inpeso9 i the fiieldpusuor throuvstgh the'
time tof tie The sorghum midge inLusianot knond Toexist whest ofntesi
gtions inete Uinthed States cnise Brownsvied, atex.uThe acompainyinga-
THE SORGHUM MIDGE.
tory methods were employed to get at the more obscure phases of the
In Louisiana it was found that, owing to the excessive humidity,
infested sorghum heads when brought into the laboratory from the
field almost invariably molded and rotted before observations upon
the emergence of the adults could be made, and, furthermore, such
conditions would not yield true
results. For this reason, then- A
and the same applies to Texas- ..i
all records of life history were
made in the field upon growing
heads and subject to absolutely "
In brief, the method employed
in the study of the various forms
of the midge was to select a num-
ber of sorghum stalks in the field
whose heads had not. broken the
boot or protecting sheath and
which were therefore not in-
fested by the midge. Over these
unbroken heads were placed
waterproof paper bags, the tops
of which were gathered securely
about the stalks well below the
head and securely tied with
string, to which tags were at-
tached. These bags were allowed
to remain over the unopened ,
heads until by observation the
latter had broken the sheath i .
and the spikelets were in a con-
dition to receive an infestation
by the midge. Then they were
removed from the heads, and
these were watched until females
were aFKrj. 22.-Growing sorghtm 1 head, bagged and tagged
were actually seen to oviposit after natural infestation by the sorghum midge.
within the glumes. When the This method is also practical for protecting seed in
n the field from damage by the midge. (Original.)
natural infestation was well un-
der way a note was made upon the small tag attached to the stalk
below the head, giving the date and the hour of the first egg deposi-
tion. (See fig. 22.)
At various times these heads were cut and dissected in the labo-
ratory and the oldest form of the midge found therein recorded,
79536-Bult. 85, pt 4-11- 2
CEREAL AND FORAGE INSECTS.
which gave accurate data upon the time required for the various
stages in the life cycle.
It is a matter of regret that this system was not discovered during
the work in Louisiana, only one similar method being employed.
This consisted in suspending a breeding cage from a small scaffold
over the growing head and artificially introducing the adult midges
from a bell jar. (See fig. 23.)
In Louisiana great difficulty in these life-history studies arose from
the depredations of the Argentine ant (Iridomyrmex humilis Mayr).
FIG. 23.-Breeding cage suspended over growing sorghum. Three sides of the
cage are covered with cheese cloth; the fourth is fitted with movable glass.
With this arrangement artificial breeding of the sorghum midge can be
observed under absolutely natural conditions. (Original.)
These ants repeatedly gained access to breeding cages in the labora-
tory and destroyed the results by extracting the midge pupe from
the apex of the spikelet just prior to emergence, and also by capturing
the adults which had succeeded in emerging.
In addition to the many varieties of sweet sorghum, the sorghum
midge is known to infest broomcorn, kafir, Johnson grass, and milo.
THE SORGHUM MIDGE.
In one instance the writer reared a single adult from the common
foxtail grass (Setaria glauca)a and Mr. George G. Ainslie has also
reared the midge from the grass Sieglingia seslerioidlesA
In the investigation of this problem many varieties of sweet sor-
ghum were observed in their relation to infestation, among which
are Early Amber, Gooseneck, Sapling, and Sumac with some mem-
bers of the durra group; and while there is some difference in the
degree of infestation of these varieties it has not been observed to
be sufficiently great to merit the recommendation of any of them as
FIG. 24.-The sorghum midge ( Contarinia sorghicola): a, Adult male: b, antenna
joints of same: c, head, frontal view. a, Greatly enlarged: b, c, highly mag-
resistant varieties-all being infested to such an extent that they
would have failed to produce a profitable crop of seed.
The following is the original description of the species by Mr.
Antenne of the male as long as, of the female almost one-half as long as, the body,
in both sexes composed of 14 joints; joints 3 to 14 in the female each slightly con-
stricted in the middle, each except the last one greatly constricted at the apex into
a Baton Rouge, La., September 14, 1908.
bClemson, S. C., August 15, August 31, November 3, and September 8, 1908.
CEREAL AND FORAGE INSECTS.
a short petiole, a few bristly hairs not arranged in whorls scattered over each joint;
in the male, joints 3 to 14 are each greatly constricted, slightly before the middle,
and again at the apex except in the case of the last joint, the constricted portions
arc as long as the thickening at the base of each joint; each of the thickened por-
tions bears a whorl of bristly hairs. In the living insect the head, including the palpi,
ij yellow, antennae and legs brown, thorax orange red, the center of the mesonotum
and a spot crossing the pleura and enlarging on the sternum black, abdomen orange red,
wings grayish hyaline. The first vein reaches the costa noticeably before the middle
of the wing; third vein nearly straight, ending slightly below the extreme tip of the
wing, the basal portion of this vein, where it joins the first vein, distinct; fifth vein
forked slightly before the middle of the wing, its anterior fork ending nearly mid-
way between the tip of the pos-
_Al terior fork and the apex of the
-:. .( / third vein. Length nearly 2 mm
FIG. 25.-The sorghum midge: a, Adult female, dorsal view;
b gntpnnnl inints ,f cnnmn" c ,tkrv.m d riillr alrolnnod lorma
The male (fig. 24, a), as
indicated in Mr. Coquil-
lett's description, is pro-
vided with antenna which
exceed in length those of
the female and which differ
greatly in the structure of
their joints, while on the
wing the former charac-
teristic serves to distin-
guish the sexes; also, the
movements of the male are
much quicker than those
of the female.
The female (fig. 25, a) to
superficial examination, at
rest or on the wing, appears
more robust than the male
and at the same time her
as found in cocoon, showing characteristic cecidomyiid movements are much more
"breastbone"; (, larva in early stage; /f, pupa. a, d,e,f, deliberate and slow. She
Greatly enlarged: 6, c, highly magnified. (Original.) is provided with a delicate
is provided with a delicate
hairlike ovipositor, capable of great extension during the process of
oviposition. Often its length exceeds that of the entire body. The
following measurements of average male and female midges serve to
illustrate the comparative dimensions of the two sexes:
TABLE I.-Comparative measurements of males and females of the sorghum midge
(Con tarinia sorghicola).
Measurement. A female Adult male.1;
Milli- : Milli-
Length of body (antennae meters. meters. I
excluded)............. 1.78 1.22
Length of antenna ........ .843 1.05
Length of ovipositor...... 1.64 ...........
Total length ............. 4.36 2.27 H
Width of thorax..........
Length of wing...........
THE SORGHUM MIDGE.
The above measurements are taken from average-sized adults
and, while these latter vary considerably, especially in the male, the
measurements given represent with a fair degree of accuracy the
average comparisons between the two sexes.
(Fig. 25, c.
The egg is a delicate, elongated, cylindrical structure. One end
tapers to a fine hairlike point, which l)efore oviposition is often twice
the length of the egg proper. The color is pale pink or yellow.
This tapering end of the egg appears to be a special construction,
the purpose of which is, at this writing, unknown. The end of
the egg proper is capped with a ferrule-like appendage from which,
in gradually diminishing (diameter, this projection extends. This
appendage is not dissolved in xylol, alcohol, or any fluid in which
the egg proper is dissolved, and does not appear to be the same
material as eggshell. When examined within tlhe abdomen of
the female, the eggs are found in an unbroken chain of groups of
three or four, the threadlike appendages projecting toward the
distal end of the ovipositor and apl)parently loosely joined at their
extremities, so that the appearance suggests small bunches of eggs
attached to a common point by the ends of their delicatee append-
ages. When laid, the appendage gradually shrivels and dries until
it contracts to a third of its original length or less. The egg is
about 0.15 mm. in length.
(Fig.. 25, d, ,.
The newly hatched larva closely resembles the egg in appear-
ance, it often being difficult to determine just when the egg stage
ceases and the larval stage begins. In color the newly hatched
larva runs between a pale yellow and a pale pink. It is uniformly
broad throughout the entire body length. As growth continues
the color changes from a pale pink to pink, then red, and when full
grown it is often deep re(l. The body to superficial examination
appears uniformly cylindrical, but under the microscope is seen
to taper perceptibly at the head and( posterior extremity. Only
after the last molt can the characteristic cecidomviid "breastbone"
(fig. 25, d) be distinguished.
Repeated measurements of lairvn in different stages of growth
give uniformly regular figures. Table II gives the measurements of
a number of larvae taken from seed at different stages of growth
from newly hatched to full grown.
CEREAL AND FORAGE INSECTS.
TABLE II.- Measurements of larva' of the sorghum midge.
1................ .. . .
3 . . . . . .. . . . . . .
5 ............ . .
( .. . . . . . . .. . .
Larva No.- Length.
7 ........................ 0.6
8 ...................... .89
9. ........................ .903
10 ....................... 1.0
11 ................. ..... 1.5
While the above examinations do not refer to larvae of the same
generation, they serve to illustrate the gradual growth from the
time of hatching until just before transformation into pupme. These
figures can safely be taken as representative of the measurements
of a single larva from date of hatching until
Full grown. In this instance the larval stage
covers eleven days.
WlWhen newly formed the pupa is uniformly
deep red in color, while just before emergence
FIG. of the adult the head and appendages turn
FIG. 2ti.-The sorghum midge:
Ct'ocooned larv;-," thei hi- (lark and finally black, while the abdomen
lrnating form of the midge. remains a deep red. There is often a delicate
Much enlarged. (Original.)
c cocoon surrounding the pupa before the latter
lhas worked its way to the apex of the spikelet for the emergence of
the adult. It is evident that such are found upon those pupw
derived from "cocooned larvae." This thin covering is very loosely
attached to the pupa and has never been found by the writer after
the pupa has left its initial position alongside the ovary.
THE "COCOONED LARVA."
The "cocooned larva" is closely allied in structure and function to
the "flaxseed of the Hessian fly. The delicate envelope is somewhat
elliptical in shape, quite flat, and of a muddy-brown color. It is found
close against thle ovary itself within the delicate palet. Examined
through the microscope the envelope is semitransparent, containing
a larva about two-thirds grown and surrounded by a clear proto-
plasmic fluid. In this formn the segmentation of the larvae is dis-
cernible. yet the structures seem faintly outlined and embryonic
in contrast to the naked normal larva. The function of the "co-
cooned larva" is to perpetuate the species over winter and is the
THE SORGHUM MIDGE.
true hibernating form from which the spring pupTe are derived.
These enveloped larvTe also appear greatly flattened and are ex-
tremelv thin and delicate in structure. In addition they are them-
selves semitransparent, the protoplasmic contents being visibly
LIFE HISTORY AND HABITS.
Emergence of the adults takes place as a rule during the early
morning hours during the warmest weather, and later, when the
weather turns cool, later in the morning and continuing well into the
day. While there is almost a continual emergence in the field from
early morning until late in the afternoon, the bulk of emergence occurs
as stated, varying with the change of season.
The pupa, having worked its way from the initial position along-
side the ovary to the apex of the spikelet, protrudes about two-thirds
of its length. The operation of casting off the pup)al skin begins almost
immediately. The abdomen is seen to twist with a backward and
forward motion, the head and thorax likewise performing the same
motions until the pupal skin is gradually split open its entire length,
along the dorsum or venter, and sometimes both. The skin does not
necessarily part in a well-defined line on either the dorsal or ventral
surface, but is often torn raggedly over its entire surface. The lead
of the adult is gradually thrust through the opening, and then finally
the entire body. The legs assist materially in this operation by push-
ing back the clinging pupal skin free from the body. When free,
the adult is very moist and weak and cl;ngs to the outer glumes for
about ten minutes before it, has become sufficiently strong and tlhe
wings have dried so as to permit flight. About fifteen minutes elaipses
from the time the adult begins to free itself from the pupal skin until
it is on the wing. The cast-off pupal skin remains clinging to the apex
of the spikelets. A count of a whole season's emergence, numbering
many thousands, fixes the proportion of males to females as three of
the latter to one of the former.
Immediately after drying, the male takes wing and hovers about
the seed head, awaiting the appearance of the females. When the
latter have emerged, copulation at once takes place, more often before
they have sufficiently dried to fly. When the drying process of the
female is complete, she begins to oviposit within the seed glumes, and
this operation continues until she has laid her quota of eggs, when
death follows. The operations of copulation and oviposition are
very rapid and are kept up repeatedly until both male and female
CEREAL AND FORAGE INSECTS.
OVIPOSITION. .q i
The process of oviposition, as carried on immediately after emear ,4:
gence and copulation, is short and rapid. The females crawl..%
carefully over a head for a few seconds until they find a spikelqti
which presents the best adaptabilities for egg laying. Oviposition :
begins before the flowers appear and continues as long as the glumes.
remain flexible, probably from five to seven days. When the female
has selected the spikelet she takes up a position upon the apex, her
abdomen elevated slightly above the tip, and immediately extends
FIG. 27.-Sorghum heads, showing four successive stages in the opening of the sheath or "boot/'a
her ovipositor, pushing it within the spikelet until it is fully extended;
then, with a rapid pistonlike motion, she places the egg. It is
doubtful if more than one egg is deposited at a time by a single
female in a given spikelet. Dissections of glumes observed to have
been oviposited in by but one female, and but once, have not, during
the writer's investigations, revealed more than one egg.
However, it is no uncommon sight to observe several females follow
one another in quick succession, ovipositing within the same spikelet.
Numerous examinations of infested spikelets have revealed as many as
a See footnote, p. 51.
THE SORGHUM MIDGE.
eight or ten stages of the midge against one ovary, all the stages rang-
ing from unhatched eggs to fully developed pupae, showing that egg
deposition is kept up until the hardening of the glumes at the apex
Another important phase of egg deposition, which accounts for the
irregular emergence of the adults from a head, is that of the long
period during which a head is in condition to receive the eggs of the
midge. As has already been pointed out, oviposition may continue
for several clays; this in itself gives rise to a
large number of midge forms of different stages
of growth on a single ovary.
Again, oviposition continues upon a head
from the time the first spikelets are visible
through the opening sheath until after the head
is entirely exposed. Figure 27 represents four
stages of development of the head.a About f
four days are required for the head to corn-
pletely emerge, during which time the seed
n a 6 1
have been infested by the midge as fast as they c'
were accessible to her ovipositor. a
LOCATION OF THE EGG. a
The location of the egg, (fig. 28) varies, inas-
much as the condition of the glumes varies at 6A !
different stages of development, and conse- 4' :1 l
quently a female ovipositing just before the
glumes close would not place her egg as far Th. 2s.-Sectional views of
down as the female ovipositing immediately the sorghum seed during the
after the shedding of the blooms. Generally flowering stage: a, First
t) outergLume; b, second outer
speaking, the eggs are found near the apex of glume: c, inner glume: d,
the ovary, but the writer has found them second innerglume: lodi-
cufles. X's indicate points
located in practically every part of the inner at which eggs of the sor-
seed structure. It is therefore dependent upon foghun midge are commonly
seed found. (Original.)
the stage of seed development as to where the
egg will be found. If infested during the flowering stage, the female
usually inserts her ovipositor between the first and second inner
glumes, and in such cases the egg will be generally found somewhere
near the apex of the ovary sticking to one of the glumes. In one
instance, at San Antonio, Tex., the writer observed females ovi-
positing in glumes which had shed the flower and the apices of which
were too hard to admit of oviposition at that point. In this case
the females were observed to crawl over the glumes and then
insert the ovipositor into the crevice formed by the first outer glunme
a Fig. 27, b and c, are not normal, owing to the drying of the material before the
drawings could be made.
52 CEREAL AND FORAGE INSECTS.
overlapping the second outer glume and at a point midway between I
the apex and base of the glumes. An immediate dissection of these
seeds revealed the eggs situated upon the inner surface of the second
outer glume almost in view from a superficial examination. It is
extremely (doubtful if larvae hatching under such conditions would
find their way to the ovary.
LOCATION OF THE LARVAE.
When hatched the larvae immediately make their way to the ovary
and( are invariably found lying directly against the latter within the
delicate palet. Their position remains unchanged throughout growth,
their length being parallel to the ovary and the head pointing to the
apex of the glumes. The larvae, expanding their full length close
against the ovary, sap the juices. A faint discoloration of the ovary
takes place at the point of contact with the larva shortly after the
latter has taken its normal position. This discoloration deepens per-
ceptibly as larval growth increases, and during the full-grown stage
the larva is set in a tiny depression, caused by the draining of the
plant juice by the larva at that point.
LOCATION OF THE PUPA.
The pupa is formed in exactly the same position as has been occu-
pied by the larva during its growth and development. The head is
directed toward the apex of the glumes, to which point it works itself
preparatory to emergence.
NUMBER OF GENERATIONS.
There are no well-defined broods or generations. From early spring
until late fall the midge may be found in any stage from egg to adult.
THE LIFE CYCLE.
The greatest of difficulty has attended the determination of the
periods required for the egg, larval, and pupal stages. Under the
most careful manipulation newly deposited eggs, when located and
examined, invariably shriveled and failed to hatch before the spikelets
opened and the eggs were exposed to atmospheric influences; conse-
quently, attempts to watch newly deposited eggs until the date of
hatching have been so far unsuccessful. The same difficulties apply
to raising larvae to maturity under artificial conditions. The pupam,
however, are more successfully handled, as exposure to air does not ;
seem to affect their (development and the emergence of the adult.
The method finally adopted consisted of permitting heads to become
infested under natural conditions in the field, then bagging them,
and later dlissecting the spikelets at various intervals. Thus, with
a large number of heads infested and examined at different periods,
THE SORGHUM MIDGE.
the approximate time required for the three stages was secured. At
Baton Rouge, La., the only successful attempt to secure the total
number of days in the life cycle gave 234 (lays from the time of
natural oviposition to the emergence of the adults. This develop-
ment took place during an average daily mean temperature of 79 F.
and an average daily mean humidity of 74.3. In San Antonio, Tex.,
the development from egg to adult required 14 d(lays during an aver-
age daily mean temperature of 84.7 F. and an average daily mean
humidity of 67.5. In the latitude of San Antonio, Tex., generally
speaking, the egg stage will cover from 2 to 4 days, the larval stage
from 9 to 11 days, and the pupal stage from 3 to 5 days, depending
upon the temperature and humidity. A very wide range in length.
of time for the various stages has been recorded, but during the
normal temperature and humidity conditions in this latitude from 14
to 20 days are the average.
LENGTH OF LIFE OF ADULT.
In confinement, when no opportunity is afforded for oviposition,
the length of life of male and female is approximately 24 hours, while
females, when allowed to oviposit, live longer, generally about 48
hours. The length of the life of the female is largely dependent upon
the number of eggs she is capable of laying-death following shortly
after the egg supply has been exhausted. Females were found upon
dissection to contain from a dozen to upward of a hundred eggs.
So far no feeding of adults has been observed. Close observation
has failed to reveal a single instance in which either male or female
partook of nourishment, their activity being confined solely to copu-
lation and oviposition.
In the spring the midge appears with the first Johnson grass and
sorghum, and, as this grass heads considerably before the cultivated
sorghum, it may be said that by the time the latter has headed the
midge has become sufficiently abundant on the grass to make the first
sorghum infestation a heavy one. In the latitude of San Antonio,
Tex., the first midges to be found during the season of 1909 were found
actively ovipositing in Johnson grass on May 14. At this date the
neighboring sorghum had not headed, and it was not until June 19 that
the first brood emerged from the sorghum, which puts the d(late of this
first infestation at approximately June 5.
Throughout the entire season the midge is found active upon its
hosts; in fact, as long as heading Johnson grass and members of the
sorghum family can be found in infested localities the midge is certain
to be present also. During the winter, when the activity of the midge
has apparently ceased, a few recurring warm days suffice to bring out
54 CEREAL AND FORAGE INSECTS. i
the adults in considerable numbers, and as long as there are any heads &
in which to oviposit they will continue to breed, although without the
regularity of the warm season, their development being entirely
dependent upon a sufficiently high temperature. On November 25,
1908, the writer discovered, in a field near Grand Prairie, Tex.,
several stalks of sorghum which had been allowed to stand after the
last crop had been harvested. Examination of seed in these heads
revealed a number of midge pupae, which, after remaining in the warm
office at Dallas for 24 hours, yielded a number of adults. Later these
heads were sent to Washington, where they were kept throughout the
winter at outdoor temperature, which was very low, and a careful
dissection of these heads during the month of January showed that all
the pupae had transformed into adults during the time the heads were
in the heated room at Dallas, leaving nothing but the true hibernating
"cocooned larvae" to carry the species over winter. It appears from
this that artificial heat will not develop "cocooned larvae" that have
been found late in the season.
As previously stated, the true hibernating form of the midge is the
"cocooned larva." Although naked pupae derived from normal
larvae can be found during the winter months, this "cocooned larva"
is the one form which, if the heads are subjected to extreme cold,
will perpetuate the species. Normal pupae will stand considerable
cold, and later, upon being exposed to sufficiently high temperature,
will emerge, but the "cocooned larvae," when once they have been
subjected to cold, will remain over winter until spring and produce
cocooned pupae and, later, adults. Therefore we can only say that
the sorghum midge hibernates as "cocooned larvae" and naked pupme,
though the preponderance of the former during the winter is very
marked. In addition to the instance cited by the writer at Grand
Prairie, Tex., Prof. Glenn W. Herrick records an instance in which,
after a freeze sufficient to kill the sorghum and kafir, he brought in
infested heads of the latter, from which, after they had been exposed
to the temperature of a heated room, adult midges emerged in large
numbers. Professor Herrick found that the normal pupae and larva
in these heads had not been killed by the freeze.
The occurrence of larvae upon the seed during the winter months
does not indicate the wintering of the midge in this stage, but is
attributable to the habit which the sorghum has of continuing to put
out a number of branch heads during recurring warm days when the
temperature does not become sufficiently low to kill the plant itself.
As already pointed out, the midge likewise will emerge irregularly dur-
ing the winter months, and as these heads present the opportunity,
oviposition takes place. The normal larvae are then formed and during
Bul. 85, Part IV (Revised Bureau of Entomology, U. S. Dept. of Agriculture.
FIG. 1.-EARLY JOHNSON GRASS HEADING AND BLOOMING ON WET BANKS OF SEWAGE
CANAL, SAN ANTONIO, TEX. (ORIGINAL.)
FIG. 2.-RUBBISH LEFT BY REMOVAL OF SORGHUM SHOCKS IN A FIELD NEAR SAN ANTONIO.
THE SORGHUM MIDGE. 55
the winter it is no uncommon thing to find them within the glumes,
but the writer has not found a single instance in which normal
larvae occurred in heads that had been formed during the summer
and allowed to stand for a sufficiently long time to yield midges from
all possible infestation; only occasional naked pupae and a predomi-
nance of "cocooned larvae" are contained in the latter.
These "cocooned larvae" are formed in the seed during the entire
breeding season of the midge. As early as June, in the latitude of
San Antonio, Tex., these forms were found, and there seems to be no
regularity in their habit of emerging. After all emergence has ceased
from normal naked larvae these forms develop into cocooned pupae
and emerge irregularly. Ofttimes a cocooned pupa is found upon
a seed along with a normal naked larva. Just what controls the
development of this form has not been discovered. Attempts to
induce hibernation artificially by subjecting these to low temperatures
and later placing them in a warm room have been unsuccessful.
RELATION OF JOHNSON GRASS TO THE MIDGE PROBLEM.
From what, has been said previously in regard to the midge in rela-
tion to Johnson grass, it is a self-evident fact that this grass furnishes
the key to the situation. Johnson grass allowed to remain over win-
ter in and about sorghum fields carries the midge until spring, and
being the first to head and bloom, gives the midge a good start, and
by the time the sorghum is headed there is a large brood of midges
from the grass ready to infest it. (See P]. I, fig. 1.) Johnson grass
is generally considered one of the greatest pests on the farm, and its
function as a host for the sorghum midge serves as but another indict-
ment against it. It is no uncommon sight to find sorghum fields
from which the last crop has been harvested, with Johnson grass
growing and heading in the fence corners (P1. II, fig. 3) and even in
the fields (PI. II, fig. 2).
The midge in certain localities is abundantly parasitized by a
S small black hymenopterous parasite determined as Aprostocetus
diplosidis Crawf. (figs. 29, 30) and by a smaller parasite determined
by Mr. Crawford as Tetrastichus sp. (fig. 31). The latter is known
to be both primary and secondary, but it is more likely to be primary
in its relation to the midge. These latter parasites are reared from
infested sorghum heads along with the predominant Aprostocetus
diplosidis. Although the predominant parasite is very aggressive
and parasitizes the midge very actively it does not become suffi-
S ciently numerous to materially check the midge until late in the
56 CEREAL AND FORAGE INSECTS.
summer, when the second and third crops of sorghum are heading.
These crops are found to mature upward of 90 per cent of sound seed, A
while the earlier crops are a total failure. Late in summer emer-
gence of these parasites and midges from infested heads is approxi-
mately in the proportion of 6 of the former to 1 of the latter.
Only one observation has been recorded as to the feeding habit.
of this parasite. On August 1, 1908, the writer observed the parasites
clustering upon the leaves of the spined Amaranthus (Amaranthus spi-
nosus), commonly known as "careless weed." Investigation revealed
that these followed in the wake of some leaf-eating beetles which
abraded the leaves, from the torn edges of which the juice oozed.
This the parasites fed upon, following the leaf eaters as they changed
FIG. 29.-Aprostocetus diplosidis: Female. Greatly FIG.30.-Aprostocetusdiplosidis: Male.
enlarged. (Original.) Greatly enlarged. (Original.)
their point of attack. These beetles were afterwards determined as
Disonycha collata Fab. and D. glabrata Fab.
The distribution of Aprostocetus diplosidis is generally the same
as that of the sorghum midge, although in some sections where the
latter abounds the parasite is not found.a
The method of parasitism can be seen readily in the field. The
parasite crawls slowly over the infested heads and then, apparently
locating a larva, takes up a position upon the spikelet, the head
toward the apex of the latter, and arching the abdomen drives the
ovipositor through one of the glumes to the interior.
The species of Tetrastichus referred to, while not proved to be pri-
mary in conjunction with Aprostocetus diplosidis, is certainly primary
upon the midge ini some instances. During the early part of the
season, when only Johnson grass is available as a host for the midge J
a Fayetteville, Ark., and Neodesha, Kans., have not been recorded as sections i
abounding in the parasite, although the sorghum minidge occurs quite abundantly.
THE SORGHUM MIDGE. 57
and the predominant parasite is not found, these parasites may be seen
actively ovipositing through the outer glume of the sorghum seed in
precisely the same manner as has been described with reference to
Aprostocetus diplosidis. At the same time the writer has reared this
parasite from Setaria glauca infested by another species of midge.
The pupae of Aprostocetus diplosidis and Tetrastichus sp. are found
occupying the same position within the spikelets as is taken by the
pupae of the sorghum midge, viz, directly against the ovary within
the delicate palet, the head directed toward the apex of the seed.
These pupae are not enveloped in the larval skin of the midge, but
are naked. While microscopic examinations of sectioned midge
larvae have not been made for the purpose of studying the develop-
ment of these
parasites, it is
evident that f
these parasites has $e -
oviposit within et-:. 4i
midge larvae in
all stages of de-
velopment. Ex- -'
observed to have
been visited by
parasites has re-
vealed, in some
instances, newly FIG. 31.- Tetrastichus sp.: Female. Greatly enlarged. (Original.)
while in other cases half-grown or even full-grown larvt e were
By far the most important predaceous enemy of the midge is the
Argentine ant (Iridomyrmex humilis Mayr) occurring in Louisiana.
These ants swarm over the sorghum stalks and heads in the fields and
whenever they find midge pupw project ingfrom the apex of the glumes
theyseize the latter in their mandibles and carry them off to their nests.
The fly Psilopodinus flaviceps Aldrich has been observed by the
writer to prey upon the adults of the midge. These flies rest upon a
sorghum blade conveniently near a head and dart out frequently,
seizing an adult and devouring it immediately.
On July 15 the writer saw a hummingbird (probably Trochilus
alexandri, according to Prof. F. E. L. Beal, of the Bureau of Bio-
logical Survey) hovering about the heads of sorghum, which were
at the time swarming with midges. To all appearances it was feeding
upon the sorghum midge. Professor Beal states that small spiders
and minute insects are often found in the stomachs of hummingbirds.
." .: ..
58 CEREAL AND FORAGE INSECTS. .
REMEDIAL MEASURES. :\
DESTRUCTION OF JOHNSON GRASS.
The destruction of Johnson grass is one of the most vital factors i
in midge control. The mere cutting of this grass is not sufficient. '
It should be burned over wherever discovered, and such areas w
plowed in the spring to prevent an early crop of heads. jl
The practice of allowing Johnson grass to grow within and around !
areas planted to sorghum is sure to furnish ideal conditions for early
and late infestation by the midge, while the small isolated patches ofi
the grass in fence corners will carry the species over winter in the seed.
Careless methods of harvesting the sorghum crop are largely
responsible for damage by the midge. It has been pointed out that
stalks allowed to stand in the harvested fields will continue to send out ii
until late in the winter branching heads, which furnish breeding
possibilities and, later, hibernating material. Again, when the crop
is harvested, the stubble should be burned over after all loose heads i
have been collected and burned. Such heads allowed to lie upon the
harvested fields over winter harbor the hibernating midges until the
following spring. In many sections it is customary to stack the har-
vested sorghum stalks in the field for winter use as a dry fodder.
This practice, as can be readily understood, furnishes unlimited
possibilities for the midge to successfully pass the winter. (See PL
I, fig. 2, and PI. II, fig. 1.
DESTROYING HEADS OF FIRST CROP.
Inasmuch as the early crop of seed is practically destroyed by the
midge and the second crop matures a very large percentage of sound
seed, it is possible that the practice of destroying the first crop of seed
and retaining the last crop will yield better results and at the same .
time eliminate a very great percentage of midges. The fumigation of
thrashed seed and storing it in tight receptacles would possibly prove
effective in reducing the number of emerging adults from seed stored
for planting purposes.
When a small crop of seed is desired for planting purposes it will be
found practical to protect the heads from the midge by bagging as
illustrated in figure 22. This should be done before the heads have- .
broken through the protecting sheath, i. e., before the stage illustrated
in figure 27, a, and the bags allowed to remain until the seed are i
mature and hardened. Of course this method is not practical of ,:|
application on a large scale, but when a small crop is desired it will :
be found to suit the purpose admirably. ::"
Bul. 85, Part IV (Revised:, Bureau of Entomology, U. S. Dept. of Agriculture.
FIG 1.-HARVESTED FIELD OF SORGHUM NEAR SAN ANTONIO, TEX., SHOWING SCATTERED
HEADS. (ORIGINAL. )
FIG. 2.-HARVESTED FIELD OF SORGHUM NEAR SAN ANTONIO, TEX., IN WHICH THE SORGHUM
MIDGE IS HIBERNATING. (ORIGINAL.)
Fli. 3.-FENCE LINE BORDERING A SORGHUM FIELD NEAR SAN ANTONIO, TEX., ALLOWED
TO GROW UP IN JOHNSON GRASS. (.ORIGINAL.)
UNIVERSITY OF FLORIDA
3 1262 08928 9184
-- .. ^y........ ...--
"i .;" I',.
." .. '.. .. .......
A : 5"" : ....... ....
.." :.... .....
. ~.. .
f *.: i.:
..... : *' ^
JH 'E :!jS i
"":; I t
HI : ,