THE ECOLOGY OF THE MESQUITE TWIG GIRDLER,
ONCIDERES RHODOSTRICTA BATES, AND ITS
EVALUATION AS A BIOLOGICAL
CONTROL AGENT
by
KENITH LEE POLK, B.S.
A THESIS
IN
ENTOMOLOGY
Submitted to the Graduate Faculty of Texas Tech University in Partial Fulfillmient of the Requirements for the Degree of
MASTER OF SCIENCE
Approved
Accepted
May, 1971 AC. Sor rs ?07,1 no. 7e ACKNOWLEDGEMENTS
">0/r ^ I ara deeply indebted to Dr. D. N. Ueckert for his guidance, encourageinent, and help in collecting data for my thesis. I would also like to express my appreciation to the other members of my committee. Dr. C. R, Ward, Dr. D. Ashdown, and Dr. J. E.
George, for their helpful criticisra. I also thank the raany land owners whose cooperation made this project possible,
I am thankful to Drs. B. D, Burks and P. M. Marsh of the
U.S.D.A, Insect Identification and Parasite Introduction Research
Branch for specific determinations of parasites of the mcsquite twig girdler.
I would also like to acknowledge the State of Texas and Texas
Tech University for their financial assistance and support in the completion of this project.
11 TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ii
LIST OF TABLES vi
LIST OF FIGURES viii
I. INTRODUCTION AND LITERATURE REVIEW. . . 1
Literature Review ...... 2
Description , . 4
II. , METHODS AND PROCEDURES 6
Study Areas 6
Mesquite Twig Girdler 7
Life History and Behavior , 7
Number of Eggs/Branch, Percent Egg Hatch, and Diameter of Girdled P ranch 8
Elytra, Egg and Head Capsule Measure ments , . 9
Volume of Wood Consuraed by Larvae .... 10
Population Densities 10
Damage to Mesquite by Girdlers 11
Field Cage Studies 12
Parasitism, Predation, and Competition ... 14
Laboratory Rearing 15
iii Page
Biological Check Method 16
Effects of Temperature on Twig Girdler
Larvae 17
Branch Method of Girdler Colonization . . 19
Adult Method of Girdler Colonization ... 20
III. RESULTS AND DISCUSSION 23
Life History 2 3
Adult 23
Larvae 27
Pupae , , 28
Description 28
Adult 28
• Egg 30
Larvae 30
Pupae 31
Population Densities 34
Damage to Mesquite by Girdlers 37
Field Cage Studies , , 38
Parasitism, Predation, and Competition. . 41
Biological Check Method 51
Effects of Temperature on Twig Girdler Larvae 52
iv Page
Branch Method of Girdler Colonization ... 54
Adult Method of Girdler Colonization. ... 57
IV. SUMMARY ^^
LITERATURE CITED ^"^ LIST OF TABLES
Table Page
1. A comparison of sex ratios of O. rhodosticta Bates collected by hand and in light traps. Texas. Autumn, 1970 24
2. Population densities of adults of _0^ rhodosticta Bates on two 50-tree transects in Winkler- Ward and Garza Counties, Texas. Autumn, 1970 35
3. Mean numbers of mesquite branches girdled prior to 1969, in 1969, in 1970, and the mean numbers of adults of ^_^ rhodosticta Bates, on two 50-tree transects in both Garza and Winkler-Ward Counties, Texas. Autumn, 1970 36
4. Effects of population density of 0^_ rhodosticta Bates on damage to mesquite and egg laying habits in cage studies. Dickens Co. , Texas. Autumn, 1970 39
5. Parasites and predators of ^_^ rhodosticta Bates larvae dissected from mesquite branches col lected in Garza and Ward Counties, Texas, between March and September, 1970 45
6. Numbers of larvae of ^^ rhodosticta Bates and their parasites and predators dissected from mesquite branches collected in 1970 and 1971 in Ward and Garza Counties, Texas 47
7. Mean no. of larvae of jO_^ rhodosticta Bates per branch under five temperature regimes in Ward and Garza Counties, Texas. June 26 to August 28, 1970 53
8. Results of colonization attempts using branches in fested with larvae of O. rhodosticta Bates in 6 uninfested Texas sites using two methods of shading branches. Autumn, 1970 55
vi Table Page
9. Results and evaluation of adult releases of O. rhodosticta Bates collected in Winkler Co. , Texas, and introduced into 3 uninfested Texas sites in early October, 1970 58
10. A comparison of adult and larval methods of colonization of ^^ rhodosticta Bates in Texas. 1970 59
VLl LIST OF FIGURES
Figure Page
1. Adult mesquite twig girdler O^ rhodosticta Bates. Ward Co. , Texas. Autumn, 1970 29
2. Egg of mesquite twig girdler O. rhodosticta Bates. Ward Co. , Texas. Autumn, 1970 29
3. Last larval instar of mesquite twig girdler, O. rhodosticta Bates. Ward Co., Texas. Autumn, 1970 32
4. Frequency graph of head capsule measurements of 250 larvae of ^^ rhodosticta Bates. Measurements were made periodically during 1970. Ward Co. , Texas 32
5. Pupae of mesquite twig girdler O^ rhodosticta Bates. Ward Co. , Texas. Autumn, 1970, . . 33
6. Effect of O. rhodosticta Bates on mesquite at a population density of 50/tree: A, Tree be fore caging; B, Tree after all girdling had stopped; C, Tree with girdled branches re moved. Dickens Co. , Texas. Autumn, 1970 40
7. Trend in density fluctuations in interacting popu lations of larvae of ^^ rhodosticta Bates and their parasites and predators dissected from galleries in mesquite branches collected in 1970, 12 miles northeast of Post, Garza County, Texas 43
8. Trend in density fluctuations in interacting popu lations of larvae of O. rhodosticta Bates and their parasites and predators dissected from galleries in mesquite branches collected in 1970, 12 miles south of Wink, Ward County, Texas 44 viii Figure Page
9. Total numbers of insects reared from^ four laboratory cages, each containing 200 rpesquite branches girdled by ^_^ rhodo sticta Bates. Branches v/erc collectevi on April 8, May 28, July 2, and Au^.aist 13, 1970, 12 miles northeast of Post, Garza County, Texas 49
10. Total numbers of insects reared frorr:! four laboratory cages, each containing 200 mesquite branches girdled by O. rhodo sticta Bates. Branches were collecicd on March 25, May 27, July 29, and y'^ugust 17, 1970, 12 miles south of Wink, vV?. rd County, Texas. , 50
IX CHAPTER I
INTROJ3UCTION AND LITERATURE REVIEW
About 56 ])iillion acres of Texas rangeland are infested with mesquite, Prosopls glandulosa Torr. var. glandulosa, one-fourth of which has greater than 20% canopy cover (Smith and Rechenthin,
1964). Mesquite is a noxious plant on rangeland because it reduces herbage production! by competition for sunlight, moisture, and nutri ents. Thick stands of mesquite make it almost impossible for ranchers to gatiu.j- their livestock. Many present day chemical and mechanical control methods are often infeasible due to cost or ineffectiveness.
The inesquite twig girdler, Oncideres rhodosticta Bates, attacks the crowns of mesquite trees. Preliminary observations on the life history and description of this insect have been presented by Ueckert, Polk, and Ward (1971). This insect chews through the bark and living xylem layers of limbs 0. 5 to 2. 0 cm in diaraeter, killing the branch above the girdle. In several areas of Texas, mesquite twig girdlers have reduced the canopy cover of mesquite considerably, and the need for more detailed studies concerning its ecology, life history, and parasite and predator complex and its evalu ation as a biological control agent has been suggested. (Ueckert et al. , 1971). This study was begun in I969 to study the mesquite twig girdler as a possible biological control agent for mesquite. The purpose of this study was to determine the biology and ecology of the mesquite twig girdler and its effect on mesquite in Texas.
The use of insects in biological control has proven highly successful with other noxious perennial plants, including Klamath weed, Hypericum perforatum, and pricklypear, Opuntia spp. (Hollo- way, 1964).
Literature Review
Bates (1885) first described the mesquite twig girdler, Ci. rhodosticta Bates, in 1885 from Lerdo, Mexico, however, there is much confusion in the literature due to misidentification of this insect. Linsley (1940) reported the taxonomy of this insect and distinguished it by the presence of three callsoities in a transverse row on the pronotum, the pale anti-median fascia of the elytra, and the yellowish or tawny spots throughout the elytral pubescences.
High (1915) reported the habits of the huisache girdler,
Oncideres putator Thomson, but most of the published records to which he refers pertain to ^ rhodosticta Bates according to Linsley
(I94O). Swenson (1969) reported that the mesquite twig girdler inflicted an average damage of about 10% to mesquite trees during the autumn of I968 on the Spining Ranch in Garza County, Texas. Ueckert et al. (1971) estimated that about 90% of the mesquite trees in infested areas in the Trans-Pecos region of Texas had been attacked by this insect, and that about 40% of all limbs from 0.5 to 2,0 cm in diameter had been girdled.
Closely related species and their distril.ution and host records were reported by Linsley (1940), Tliey inclado: (1) O^ pus tula ta
Lee. , which is abundant in the lower Rio Grande Valley of Texas where it girdles branches of Acacia farnesi-ma, Pithecolobium flexicaulis, Prosopis glandulosa, A cacia be rUaidieri, Parkinsonia aculeata and Mimosa lindheimeri; [2) 0_^ cin^'ulata (Say), which girdles hickory, pecan, persimmon, elm, poplar, sour-gum, bass-wood, honey locust, dogwood, apple, pear, peach, plum, citrus, eucalyptus and probably other trees; (3) O, texana Horn, which occurs in southern Texas where it girdles the twigs of Acacia and probably also Prosopis and Pithecolobium; and (4) O^ que reus
Skinner, which is found in southern Arizona where it attacks oaks,
Quercus spp.
Linsley (1940) reported that _CL rhodosticta Bates occurs in the southwestern United States and northern Mexico where it occurs on mesquite (Prosopis). Howard (1900) was the first to report this beetle's damage to naesquite; however, he misidentified the beetle as O. putator Horn according to Linsley (1940). Howard's report 4 was documented from specimens and information sent to him by
Mr. M, R. Wise, who speculated that the mesquite in Calabasas,
Arizona, would ultimately be exterminated due to excessive pruning which eliminated seed production. According to Linsley '1940),
Craighead (1926) reported this species from Sarcobatus. Essig
(1926) illustrated the adult, egg, larvae, and its injury to mesquite
under the narae O^. pustulatus Lee. (Linsley, 1940).
Description
The mesquite twig girdlers (Figure 1) belong to the family
Cerambycidae, subfamily Lamiinae, and tribe Onciderini. One of
the raajor characteristics of the tribe is that the front coxal cavities
are angulated on.the outer side and closed behind. The antennae
of the males are much longer than the body, while those of the females
are as long as the body.
The genus Oncideres was first described by Serville in 1885,
Linsley (1940) described the beetles of this genus as stout, cylindri
cal insects of medium to large size, dark gray or grayish brown
in color, and often colorfully marked. The antennae are as long as
or a little longer than the body and the thorax often has a spine on
each side. The legless larvae are cylindrical, shining borers, with
heads longer than they are wide and the anterior margin of the front beset with transverse rows of short carinae or ridges. The pronotum is raised, very shining, and finely striate; and the ampullae bear two, or rarely three, regular rows of tubercles. The genus Oncideres can be distinguished from Taricanus by the eyes not being so finely granu lated, the lower lobe of the eye is more elongate, and the antennae are slender in both sexes. The vertex of the head is flat instead of concave as in Taricanus. '^Linsley, 1940). CHAPTER II
METHODS AND PROCEDURES
Study Areas
Field studies were conducted in three areas which were infested by high population densities of twig girdlers: (1) 6 and 12 miles south of Wink in Winkler and Ward Counties, Texas; (2) 12 miles northeast of Post in Garza County, Texas; and (3) 6 miles south west of Quannah in Hardeman County, Texas.
The Winkler-Ward County study area is in the Trans-Pecos region of Texas and is located principally on the Ron Slaughter
Ranch. Vegetation in the area is typical of arid mesquite-creo sote bush associations. Soils in this area are Springer loamy fine sand or sandy loam, and the topography is nearly level. The average annual precipitation is about 30 cm. The mesquite is shrub-type and has a high percentage of the small limbs typically preferred as oviposition sites by the mesquite twig girdler.
The Garza County study area is in the Rolling Plains of Texas and is located on the Post-Montgomery Estate where the average annual precipitation is about 47 cm. Vegetation consists of buffalo grass, tobosa grass, sideoats grama, and scrubby mesquite and soils are Vernon clay loam with a 3-5% slope. Mesquite in this area is generally scrubby, but may be quite large in low lying areas.
6 The Hardeman County study area is in the Rolling Plains of
Texas and is located on the Joe Ed Steep Ranch. Vegetation con sists of tobosa grass, juniper, and mesquite. Tellman-Vernon soils are predominant and the topography is rolling. The average annual precipitation is about 63 cm and mesquite in this area is m.ostly tree-type with basal stems from 7 to 13 cm in diameter.
The twig girdlers at the Hardeman County study area were determined to be Oncideres cingulata (Say) instead of O. rhodosticta
Bates, thus research at this site was curtailed.
Mesquite Twig Girdler
Life History and Behavior
Mesquite twig girdlers were observed under field conditions to determine their feeding and mating habits and their response to light traps. Field cages were used to observe longevity, fecundity, and ovipositing habits.
Large numbers of mesquite twig girdlers were needed for various experiments in this study, thus the effectiveness of light traps for collecting specimens was determined. The girdler's response to artificial light at night was determined using light traps in Garza and Winkler Counties on October 3 and 8, respec tively, when the air temperature was above 70° F. Light used for these traps was supplied by a 100-watt bulb suspended from a 8
5-ft. tripod, and powered by a portable generator. Beetles were collected by placing a sheet under the tripod for background con trast. They were sexed by the differential length of the antennae.
Large numbers of mesquite twig girdlers were hand collected from mesquite branches. Their sex was determined and the numbers of males and females collected with light traps were compared with the numbers collected by hand. The chi-square test (Snedecor and
Cochran, 1967) was used to determine if the sex ratios were equal using these two collecting methods. The mean antennal length was also determined on both raale and female beetles.
Nuraber of Eggs/Branch, Percent Egg Hatch, and Diameter of Girdler Branches
On October 14 and 27, 1970, 60 recently girdled branches were collected at both the Garza and Ward County study areas to determine the mean number of eggs/branch, the time required for the eggs to hatch, and the percent egg hatch. The incubation period was determined by selecting branches in which beetles were oviposit ing and dissecting these oviposition sites daily in the lab until the eggs hatched. Sixty branches were dissected about three weeks after they had been girdled to determine egg hatch. It was assumed that this was an adequate incubation period. Dial calipers were used to determine twig diameters immediately above the girdles. Elytra, Egg, and Head Capsule Measurements
Elytra, eggs, and head capsules of mesquite twig girdlers were measured with a microscope equipped with an ocular scale at 6, 12, and 25 magnifications for these three measurements, respectively.
Scale values were converted to metric units with a micrometer.
Elytral and overall length naeasurements were made on mes quite twig girdlers hand collected in Winkler County in late Septem ber. Forty-five elytral measurements were taken from both male and female specimens. These measurements were used to compare male:female size ratio instead of the overall length to avoid errors due to flexing of the head and prothorax. The "t" test was used to determine if raales and females were the same size (Steel and Tor- rie, I960).
The length and width of 60 eggs were measured from branches collected in Garza and Winkler County study areas on October 14 and 19, respectively. Branches containing unhatched eggs were selected from representative trees at each study area. The bark was removed from the oviposition site, leaving the egg exposed, and measurements were made.
Head capsule widths of 250 mesquite twig girdler larvae, ranging in size from first instars to pre-pupal instars, were deter mined, Dyar's Law was used to attempt to determine the nuinber of instars in the larval development of this insect (1890), 10
Volume of Wood Consumed by Larvae
The volume of wood consumed by a mesquite twig girdler lar vae was deternained from 35 galleries. Branches, from which gird lers had emerged, were collected in Oarza County in raid-November,
At each emergence hole a sraall opening about 2 mna wide v/as cut the entire length of the gallery. Larval or pupal exuvae and frass were reraoved and individual galleries were filled with fine sand which was weighed and converted to a volume by applying a weight: volume ratio.
Population Densities
Population densities of mesquite twig girdlers were determined at the Winkler-Ward and the Garza County study areas during the fall of 1970. Fifty trees on each of two transects were permanently marked with wooden stakes at each of the two sites. Population density counts at the Winkler-Ward County site were made on Septem ber 24, October 7, 19, and 27, 1970. In Garza County, counts were made October 7, 14, 18, and November 5, 1970. Turnover in the twig girdler population was determined by marking the elytra of each beetle with a small dot of model airplane paint, using different colors of paint on different dates.
On each transect the following were recorded: (1) mean number of branches girdled prior to 1969; (2) mean number of 11 branches girdled in 1969; (3) mean number of branc^ies girdled in
1970; and (4) mean number of mesquite 1v ig girdh;rs present in
1970. These variables w^ere used to deto.miine if future population densities and danaage could be predicted from measurable statistics.
The statistical method outlined in vSinjdecor and Cochran (1967) was used to test for significant correlation coefficients between the following: Xj^, mean number of branches girdled prior to 1969, with
Yj^,mean number of branches girdled in 1969; X2,n'iean number of
branches girdled in 1969, with Y^j^aean nuinber of branches girdled
in 1970; X3,mean number of branches girdled in 19f'>9, with Y^, mean
number of twig girdlers present in 1970; and X4, mean number of
branches girdled in 1970, with Y^, mean number of beetles in 1970.
Damage to Mesquite by Girdlers
Estimates of the extent of injury to mesquite by the mesquite
twig girdler, 0_^ rhodosticta Bates, and the twig girdler, O^ cingulata
(Say) were made in late May of 1970 at the three study sites. Damage
by O, rhodosticta Bates was estimated in Ward and Garza Counties,
while damage by jO^ cingulata (Say) was estimated in Hardeman County,
The extent of girdling damage on 200 mesquite trees in 10 line tran
sects was estinaated at each site. Injury for each tree was the aver
age of two observers' estimates of the percent of the total volume of
the photosynthesizing area that had been killed by twig girdlers. • • 12
These estiiTiates were made on May 27, 28, and 29 for Ward, Garza, and Hardeman Counties, respectively. Damage estimates included girdling activity in 1969 and all previous damage.
Field Cage Studies
Field cages were used to confine beetles in an attempt to deter mine damage to mesquite from known population densities of adult twig girdlers. These cages were also used to determine fecundity, longevity and the number of branches girdled/female. This phase of the study was conducted 5 miles east-northeast of Dickens on the
7-Bar Ranch in Dickens County, Texas. The mesquite in this area had been controlled by chaining about 10 years prior to this study and the resprouts had a high percentage of stems of the size pre ferred by twig girdlers for oviposition sites. Three 6X6 ft, cages, constructed of wood and saran screen, were placed over three trees of approximately equal size.
The first cage trial was conducted in June, 1970 using twig girdlers,. O. cingulata (Say), which were collected on June 11, 12 miles east of Guthrie on the 6666 Ranch in King County, Texas.
Population densities of 26, 50, and 100/tree were used with a 1:1 sex ratio. The beetles were placed in the cages on June 12 and the cages were removed on July 24 and damage estimates were made. 13
A similar cage experiment was conducted in late September,
1970, usinp, mesquite twig girdlers, O, rhodosticta Bates, which were collected at the Ward County study area on Septem,ber 24,
Population densities of 10, 26 and 50 adults per tree were used, also at a 1:1 sex ratio. The beetles were collected as early after emergence as possible in order to obtain feraales which had not yet laid eggs. The cages were removed and damage estinaates were made on NoveiTiber 4,
A third cage experiment was conducted at the Garza County study site, A similar cage was placed over a small roes quite tree that had sustained heavy girdling for several years prior to this study in order to determine if girdlers would girdle branches of greater diameter at extrenaely high population densities. On June
11, 100 twig girdlers, O. cingulata (Say), collected on the 6666
Ranch in King County, Texas, with a 1:1 sex ratio were caged on the tree. The cage was reraoved on July 24, 1970, and damage estimates were naade. Twenty-six adult mesquite twig girdlers,
O. rhodosticta Bates, with a 1:1 sex ratio were caged in the same tree on September 11, 1970. These beetles were reared in the laboratory from mesquite branches collected at the Garza County study area. The cage was removed and damage estimates were made on November 5, 1970, 14
Permanent photo-points were established for each caged tree.
Pictures were taken: (1) before each tree was caged; '2) after girdling had ceased; and (3) after girdled branches had been removed. After girdling had ceased and the beetles were dead the following data were recorded for each tree: II) percent of crowm volume removed; (2) nunaber of branches girdled; (3) mean diaraeter of girdled branches; and (4) mean number of eggs laid per girdled branch. The "t" test (Steel and Torrie, i960) was used to determine if the mean numbers of eggs/branch were equal under the three population densities and if the mean diameters of the girdled branches were equal under these treat- raents.
Parasitisna, Predation, and Competition
The nuraerical relationship between twig girdler larvae and their parasites and predators was deterrained by periodically dis secting girdled branches frona Garza and Ward Counties, Each oviposition site or gallery was examined to deterraine if the lar vae had been attacked by a parasite or predator. The number of twig girdler larvae and the number of parasites and predators was recorded for each branch.
Larval parasitization or predation was determined by the presence of a parasite or predator in the gallery or the presence 15 of material produced by parasites or predators. This material, v.diich attached the parasite pupae to the gallery wall was usually deposited in the upper end of the gallery, above the oviposition site. Predators deposited several small black spots, probably feces, on the floor of the gallery.
Parasites or predators in the late larval instar or pupal stage were reared to the adult stage in the lab. The adults of these parasites and predators were sent to the Entoraology Research
Division, A, R, S. , U. S. D, A, in Beltsville, Maryland, for identification.
Laboratory Rearing
Laboratory cages v/ere used to study relationships between twig girdlers and other insects inhabiting girdled branches. Some of these insects were parasites or predators, while others corapete v/ith the twig girdler larvae for food and space. On each of four sampling dates, 200 girdled branches were collected from the Ward,
Garza, and Hardeman County study areas. The sanapling dates for each of these three counties were as follow : for Ward County --
March 25, May 27, July 29 and Aug. 17; for Garza County -- Apr. 8,
May 28, July 2, and Aug. 13; for Hardeman County -- March 31,
May 29, July 9, and Aug. 18. The smaller twigs of these girdled branches were reraoved and the raain stems were placed in 2 ft. 16
X 3 ft. rearing cages, made from aluminuir; ^•. indow screens and
covered with paper to exclude light, with ends i-nade of fiber board.
A pint jar was attached to the bottona corner of each cage to collect
positively phototactic insects as they emerged "rom the inesquite
branches.
Insects reared from the branches were periodically collected
from the jars and preserved in 70 percent elbanol. The total number
of twig girdlers reared from each cage was conipared with the
numbers of parasites, predators, and competitors to determine
their numerical and chronological relationship to the twig girdler
population.
Biological Check Method
DeBach and Bartlett (1964) staled that tlxC evaluation of the
effectiveness of natural enemies of all pests in a ecosystem is a
fundamental prerequisite to intelligently manipulate insect popu
lations ecologically. This concept seemed applicable In the evalu
ation of the effects of parasites on a potential biological control
agent of mesquite. A biological check method was therefore used
to evaluate the effects of parasites on natural populations of mes
quite twig girdlers. This study utilized the natural aggressiveness
of ants against parasites of honeydew-producing insects (Bartlett,
1961; DeBach, Fleschner, and Dietrick, 1951; Ilanders, 1951). 17
In this biological check study an atterapt was made to increase ant activity on girdled branches. It was assumed that the ants would
be aggressive against all parasites, regardless of their hosts. The
treatments applied in the field and in two replications, each consisted
of 50 girdled branches and included: (1) applying honey to girdled
branches to encourage ant activity, and (2) a non-treated control.
Four-dram plastic vials taped to the girdled branches above the
highest oviposition site were filled with honey, and vial caps were
held open about 3-4 mm with insect pins to exclude larger insects.
The honey was replaced periodically as it was removed frora the
vials by ants.
Honey was applied to 50 girdled branches in Ward and Garza
Counties on July 2, and 6, 1970, respectively. On August 28, 1970,
the treated branches and the non-treated branches were collected at
both sites. All oviposition sites and galleries were exarained to
determine the percent parasitization and the mean numbers of larvae/
branch. The "t" test (Steel and Torrie, I960) was used to determine
if the naean percent parasitization levels and the naean nurabers of
larvae surviving was equal under the two treatments.
Effect of Temperature on Twig Girdler Larvae
High (1915), working with Oncideres putator Thomson, observed
that when girdled huisache branches remain high in the top of trees 18 the larvae would die from lack of raoisture, while severed branches or those only a few feet above the ground apparently were under higher humidity conditions and therefore produced more adult beetles.
Based on High's observations, a study was designed to evaluate four methods of augmentation of native mesquite twig girdlers by larval habitat manipulations. Larval survival was used to determine the favorability of raicro-climatic conditions in plots established in Garza and Ward Counties, Texas, on June 26 and July 6, 1970, respectively. The treatments, each applied to 50 girdled branches, included: (I) hanging girdled branches in the canopy of mesquite trees; (2) placing girdled branches on the ground in the shade of mesquite trees; (3) placing girdled branches on mid-grass vegeta tion (tobosa and black grama); (4) placing girdled branches on bare ground or short-grass vegetation (buffalo grass and false-buffalo grass), and (5) a non-treated control. The mean number of live larvae/branch at the beginning of the experiment was determined by dissecting 50 branches from each study site.
Temperatures were recorded at positions corresponding to the
5 treatments to determine if temperature affected larval survival.
Two teraperature raeasurements were taken during the hottest part of the day (between 3:00 and 4:00 PM) in early and late July. On
August 28, 1970, these girdled branches were collected and dissected 19 in the lal; to determine the mean numbers of larvae surviving as well as percentages of larvae, pupae, and adults.
The ni^aiysis of variance and Duncan's new multiple range test (Steel and Torrie, I960) was used to determine if the mean numbers of larvae surviving under these treatments were equal.
A correlation coefficient (Snedecor and Cochran, 1967) was cal culated between the mean temperature and the mean number of larvae surviving under each treatment to determine the effect of tempera ture upon Ir.rval survival.
Branch Method of Girdler Colonization
Six field introductions were made involving from 500 to 6, 000 girdled mesquite branches infested with mesquite twig girdler larvae,
These introduction plots were 30 m X 30 m and were in areas with no twig girdler activity in 1969. Two methods of shading girdled branches were used: (1) placing from 10 to 15 branches on tall grass under a mesquite tree (light shading): and [2) placing about
500 branches in a pile under a mesquite tree and covering them with cardboard, green mesquite branches, and leaves (heavy shading). The mean number of larvae/branch was determined for each introduction by dissecting oviposition sites on 25-50 represen tative branches. 20
Locations and dates of field introductions were as follow:
(1) 7-Bar Ranch, 5 miles east-northeast of Dickens, Dickens County,
Texas (May 20, 1970); (2 and 3) Sclprbauer Land and Cattle Company,
16 miles east-northeast of Andrews, Andrews County, Texas (June 8 and 18, 1970); (4) L-7 Ranch, 7 miles south of Crosbyton, Crosby
County, Texas (August 5, 1970); (5) Stephen Farms, 12 miles south of Draw, Borden County, Texas 'August 5, 1970); and (6) White
Farms, 3 miles east of Grassland, Lynn County, Texas '^August 5,
1970). At each site, 13 mesquite trees were permanently marked
in each of the four cardinal directions; the last tree marked in each
direction was about 35 m from the center of the introduction plot.
Population estiniates of adult twig girdlers and the number of branches
girdled/tree were determined on these 52 trees in late October,
The relative success of these introductions were determined by dividing the mean numbers of branches girdled/tree by the total
number of branches introduced. This number was then multiplied by
a constant, 10^ , to adjust success indices to a value between 0, 1
and 1, 000.
Adult Method of Girdler Colonization
Three field introductions, involving from 2,000 to 4, 100 live
adult mesquite twig girdlers, were made in locations with no twig
girdler activity in 1969. The locations and dates of these introductions 21 were: (1) Caswell Farms, 3 miles north of Draw, Lynn County, Texas
(September 28 and October 8, 1970); (2) Dennis Ranch, 16 miles
southwest of Post, Borden County, Texas (September 25 and 28,
1970); and (3) Koosman Ranch, 1 mile south of Dickens, Dickens
County, Texas (September 28, 1970).
The beetles used in these introductions were hand collected
by dislodging them from mesquite branches. The beetles were trans
ported to introduction sites in 3-gallon ice-cream cartons. Small
mesquite branches were placed in these containers to provide food
and resting areas, and to reduce the tendency of the beetles to dis-
raember each other before they could be released into mesquite trees
at the introduction sites. Population estimates and the mean number
of branches girdled/tree were raade on each of 52 permanently marked
trees, 13 in each of the four cardinal directions frora the center of 30
m X 30 ra introduction plots.
The relative success of these adult introductions was determined
by dividing the mean numbers of branches girdled/tree by the total
number of beetles released. This value was then multiplied by a
constant, 105 , to give a success index between 0. 1 and 1, 000,
Adult and branch colonization methods were compared by adjust
ing numbers of branches introduced to represent the numbers of twig
girdlers that would have emerged from these branches had they been
left intact on raesquite trees. The mean number of adult beetles that 22 eraerged from branches left intact on raesquite trees was determined by counting the number of eraergence holes present on 50 branches in Garza County on Noveaaber 6, 1970, after emergence of adults had ceased. Attempts were also made to collect similar data at the Ward-Winkler County study area in late November; however, emergence had caused severe breakage of limbs and counts were not possible. CHAPTER III
RESULTS AND DISCUSSION
Life History
Adult
Adult inesquite twig girdlers began emerging from galleries in late August and early September and continued until late Novem ber. Branches dissected August 28, 1970, at the Garza County study site contained 2% adults, 33% pupae, and 65% larvae. Branches dis sected at the Ward County study site on August 28, 1970, contained
14% adults, 47% pupae, and 37% larvae. Thus adults probably enaerge in Ward County one to two weeks earlier than in Garza County, which is located about 180 railes to the north.
Longevity-In field cages naost adults had died in 20 to 30 days; however, a few adult beetles lived raore than 45 days.
Behavior-Only female mesquite twig girdlers were observed girdling mesquite branches. Males were attracted to light traps significantly more than females (0.05 level). About 87% of the adult twig girdlers attracted to lights were males (Table 1). Both sexes behaved in a similar manner when disturbed. Instead of flying, the beetles dropped frora the mesquite tree to the ground and reraained motionless. They made a noise by rubbing the prothorax against
23 24
(U M rt ::3 in o cr 00 O to 1 • r-l CO .c ^ U X3 $^ x) (D -M (U C to -»->
X) XI (1) +J to O rce n CO OO lect e 00 r-H mal e r—1 O cu 0 O O W o pq X) CO Q) nJ (T5 r^ O +-> un o o •^ H O O W O '^ o O J-i P. P. o (d tt} ^1 ^1 UH O .2 X -M 4-> t^ -M O o 4-J O X 4-> to 1—1 .r: .ri O bJD fcuD •i-i .•-< tr-l CU JiU I—< I—1 ^ a cu ra t mn . ^ 4-> (I) :i cu (0 o < X "^-l o 00 CO O -l-cu> CJ P. C3 to -4-cuJ O o5 cd CD o o o CO ^ r—1 o O • r-l (1) P 0 O CU H O P^ +-> e to a 0 0^ >> u a u 0 O o • r-l u U o .1-1 ^1 U bJD o .1-1 • •-I cu cu t—I I—( cd r-l N u Pi ^^ O H ^ ^ O 25 the anterior margin of the elytra when captured. Both of these characteristics were presumed to be protective mechanisms. Feeding habits-Adult beetles fed for several days prior to mating and girdling on tender bark around the buds, thorns, and small limbs of mesquite. The adults did little obvious daraage to mesquite by feeding, and only at population densities of 100/tree in the Garza County cage experiment was there any noticeable damage. Mating- Mating was observed both at day and at night, even while feraales were involved in girdling or egg laying activities. Males and females were usually found on the same branch. Observ ing beetles raarked with paint showed that males frequently mated with more than one female. Fecundity- Cage studies on the 7-Bar Ranch in Dickens County, Texas, in September showed that the average number of branches girdled by raesquite twig girdlers varied from a low of 0. 96/feraale at a population density of 50 beetles/tree to a high of 1. 5/female at a population density of 26 beetles/tree. The average numbers of eggs laid/branch for Garza and Winkler Counties were 8. 1 jf 1.1 and 8. 2 + 1.1 (95% confidence intervals), respectively, with 97% and 98% egg hatch, respectively. Oviposition- Female beetles girdled branches before laying eggs, by chewing through the periderra, nontranslocating phloem, 26 translocating phloem, cambium, and far enough into the xylem to prevent the upward translocation of water and iiutrients. Branches died soon after they were girdled. The leaves remained on these girdled branches because an abscission zone did not form. These girdled branches averaged 9 ^aim jf 0, 05 mm Grange 6 to 15 ram) and 10 mm 4; 0. 06 mra (range 6 to 20 mna) in diameter in Winkler and Garza Counties, respectively (95% coniidence intervals). Meyer et al. (1971) stated that the small stems of raesquite trees were iinportant for storage of carbohydrates as starch gran ules. When these sraall branches were girdled by the mesquite twig girdler the anaount of stored carbohydrates available to the plant was likely reduced. The reduction of this stored food by the action of these insects may raake it necessary for the plant to use food reserves stored in the roots for budding and leaf development the following spring, thus weakening the plant. After the branch was girdled, the female made an incision in the bark with her raandibles and placed the egg between the bark and wood. The female faced downward while making the incision, then turned around and inserted the ovipositor so the egg was inserted parallel to and beneath her abdomen. After extracting the ovipositor, the incision was sealed with an amber colored secretion. It took about 20 to 30 minutes for the feraale mesquite Iwig girdler to lay each egg, while the girdling process required as much 27 as two days. The distance between oviposition sites varied with the size of the branch, but they were usually about 40 ram apart. The eggs hatched in 10 to 14 days. Larvae Feeding habits-The raesquite twig girdler larvae fed toward the oviposition scar and opened it after about three months of feed ing. The larvae used this opening to expell frass frora the gallery and were observed feeding on only one side of the branch thus, severely weakening the branch at that point. Many weakened branches broke during high winds, probably killing the larvae directly or indirectly. Behavior-The larvae exhibited several behavioral character istics. In the later stages of developraent when galleries were dis sected the larvae were usually found with their venter toward the bark. The larvae used sraall strips of wood to plug up any sraall holes that were raade in the galleries. When the larvae fed within about 3 mra of another gallery they usually veered to one side. The larvae of the mesquite twig girdler did not cut an emer gence hole in the bark nor did they prepare a pupal cavity as do some other species of Oncideres. A prepupation stage occured in the larva in which the head capsule became deflexed, the naouthparts became hypognathous, and moveraents characteristic of the pupae 28 were exhibited. Peristaltic-like movements and many dorsal spines were used by the pupae to renaove the larval exuvium after it splits immediately caudad to the head capsule. Pupae Description-The pupae were very raobile and moved inside the galleries using the elongated last abdominal segment which has a row of sraall spines around its perimeter, by pressing the spines against the sides of the gallery. The pupal stage occurred in late August and early September and lasted for about 14 days. Description Adult Adult raesquite twig girdlers were black in color with a raid-transverse grey band across the elytra. Each elytron was dotted with about 35 to 40 small patches of orange pubescence (Figure 1), The raean length of the male and feraale elytron was 10. 1 mm + 0, 3 mm ('range 8. 0 to 12. 1 mra) and 10. 6 mm j- 0. 3 nam (range 8, 0 to 12. 0 nana) (95% confidence intervals), respectively, while the naean overall lengths were 16, 3 and 15. 1 mm, respectively. Male and fen^iale elytral lengths were not significantly different (0. 05 level). Males were distinguished from females by the length of the antennae. The average antennal length in males was about 30, 3 mm, 29 Figure 1. Adult raesquite twig girdler O. rhodosticta Bates. Ward Co., Texas. Autumn, 1970. Figure 2, Egg of mesquite twig girdler 0_^ rhodosticta Bates, Ward Co. , Texas, Autumn, 1970. 30 as compared to 18. 4 mm in the females. The distal joint of male antennae were about four times longer thnn that of the feraale. The length of this segment in the male varied from 3. 2 to 6. 5 mm, while in the fenaale, it varied from 0, 9 to 1,5 mm. Males can also be distinguished frona femahjs by the shape of the last ventral abdonainal segment. The posterior margin of this segment in the male is indented or slightly "V" shaped, while in the feraale this margin was not "V" shaped and was broader. The eggs were elliptically ovate in shape, about four times as long as w^ide (Figure 2), of creanay white color, with a mean length of 4. 3 mm +_ 0. 06 nam (range 3. 8 to 4. 7 mm) (95% confidence interval). Their mean width was 1. 1 ram +_ 0. 02 nana (range 0. 9 to 1. 3 mm) (95% confidence interval). The eggs were slightly pointed on the end away from the oviposition site. Just before the egg hatched the yellowish embryonic larvae could be seen through the chorion. The head developed toward the oviposition site, and the dorsum developed adjacent to the bark. Larvae The niean length of newly hatched larvae, which were yellowish white in color, was 3. 2 nana ('range 2. 6 to 3, 6 mm). The prothorax was about 1 ram wide and was whiter than the re si of the body. 31 The head was light brown, while the mandibles were black, (Figure 3). Just before pupation, the larvae were soiaewhat scrabei- form and about 26 nam long (Figure 3). Each abdominal seg ment possessed two callous-like ridges on its dorsuna and ven ter; these ridges were probably used to stabilize the larvae while feeding. Out of 250 head capsule measurements there were no distinct size groups, therefore it was inapossible to deternnine the number of larval instars by use of a size-frequency graph (Figure 4). Ap parently Dyar's Law^ does not apply to this species. Volume of wood consumed-A raesquite twig girdler larva ate 1. 44 cc of wood (range 0. 91 to 2, 1 cc). The sixiall volume of wood eaten indicated that these larvae were very efficient in extracting essential nutrients from mesquite wood. Pupae The pupae were about 20 mm long ^Figure 5), and were white when they first enierged frona the last larval skin; however, they gradually became darker until transfornaation into the adult stage. The eyes darkened first, then the mandibles, vertex of the head, and prothorax in rapid succession. 32 Figure 3, Last larval instar of mesquite twig girdler, O. rhodosticta Bates. Ward Co. , Texas. Autumn, 1970. 0. 31 Head capsule diameter (mm) Figure 4. Frequency graph of head capsule measurements of 250 larvae of 0_^ rhodosticta Bates, Measure- ments were made periodically during 1970, Ward Co. , Texas. 33 Figure 5. Pupae of mesquite twig girdler O^ rhodosticta Bates. Ward Co. , Texas. Autumn, 1970, 34 Population Densities The highest population density of mesquite twig girdlers occur red on September 24, 1970, the first sampling date (Table 2). Sam pling iT\ay have begun too late to determine peak population densities, and it was observed that the number of twig girdlers/tree decreased as the season progressed. Sampling probably should have begun in early September and continued through early November. A turn over in the population occurred within a 10 day period, thus the standing crop of beetles cannot be used as a reliable estimate of the total effective population. Previous and current girdling damage and current population densities on transects in Garza and Ward Counties are shown in Table 3. There was no significant correlation between: (1) the mean number of branches girdled/tree prior to 1969 and the mean number of branches girdled/tree in 1969; (2) the mean nunaber of branches girdled in 1969 and the mean number of branches girdled in 1970; (3) the mean number of branches girdled in 1969 and the mean number of beetles/tree in 1970; and (4) the mean number of branches girdled in 1970 and the mean number of beetles/tree in 1970 (0. 05 level) (Table 3). Thus, previous damage records or population densities may not be reliable variables for predicting future trends. The twig girdler population appeared to be decreasing in Garza County and increasing in Winkler and Ward Counties. 35 Table 2. Population densities of adults of ^ rhodosticta Bates on two 50-tree transects in Winkler-Ward and Garza Counties, Texas. Autumn, 1970. Mean no. girdlers/tree Garza Co. Winkler-Ward Co. Transect Transect Sampling no. no. dates 1 2 1 Sept. 24 - 1/ 4. 7 Oct. 7 0. 8 - 3. 5 2. 5 Oct. 14 0. 5 0. 55 22.. 88 1. 1 Oct. 27 0. 2 0. 2 0. 6 0. 1 Nov. 5 - 0. 1 — Total no, 74 44 583 185 1/ Dashes indicate that no data were collected, 36 Table 3. Mean numbers of raesquite branches girdled prior to 1969, in 1969, in 1970, and the mean numbers of adults of 0»^ rhodosticta Bates, on two 50-tree tran sects in both Garza and Winkler-Ward Counties, Texas. Autumn, 1970. Va riables Mean no. Mean no. Mean no. Location branches branches Mean no. branches and girdled girdled beetles girdled transect prior 1969 in 1969 in 1970 in 1970 Garza A 7. 3 5. 9 1. 5 0. 9 Garza B 7. 7 7. 1 0.9 1.4 Winkler- WardA2. 8 1.8 11. 6 13. 7 W inkler- W ard B3. 0 7. 0 3. 7 9.0 37 Inadequate naethods for determining population densities in 1970 probably accounted for insignificant correlation between this and the mean number of branches girdled/tree in 1970, Damage to Mesquite by Girdlers Damage estimates showed that 34% _+ 3% and 36% + 3% (95% confidence intervals) of the crowns of mesquite trees had been killed by natural populations of mesquite twig girdlers, O, rhodo sticta Bates, in Ward and Garza Counties, respectively. In Harde- naan County, 17% j- 2% (95% confidence interval) of the crowns of mesquite trees had been killed by the twig girdler, O. cingulata (Say), Variations .in the degree of damage were attributed to size or growth habit of the raesquite trees, persistence of the infesta tion, and population densities of girdlers. In Hardeman County all three factors appeared to account for less insect damage. Trees were larger, the infestation was fairly recent and the population density appeared to be low, Mesquite trees in Garza County were larger than those in Ward County and mainly single-stemmed; however they were damaged more than the multiple-stemmed mesquite in Ward County, There fore, the higher damage estimates in Garza County may have been the result of a longer period of infestation, or population densities 38 in Garza County in 1969 and previous years naay have been higher than those in Ward County. Field Cage Studies Adult twig girdlers, O^ cingulata (Say), collected in June 1970 on the 6666 Ranch in King County, killed 15%, 5 5%, and 75% of the crown volume of three caged raesquite trees at population densities of 26, 50, and 100/tree, respectively. The number of branches girdled/female was 1. 3, 2. 6, and 1, 1 at the three population densi ties, respectively. In this study the highest population density caused the greatest damage, while each female girdled more branches under the intermediate population density, Mesquite twig girdlers. O^ rhodosticta Bates, collected in late September in Ward County killed 10%, 75% and 65% of the crown volumes of caged mesquite trees at population densities of 10, 26, and 50/tree, respectively (Table 4), ^Figure 6), The number of branches girdled/female was 1, 6, 1, 15, and 0, 96 at the three popu lation densities, respectively. The mean diameter of the girdled branches did not increase as population densities were increased in these field cages. The mean diameter of the branches girdled in these cages was 9 mra (range 6 to 12 mra), which agreed with the results obtained under natural conditions. Thus, mesquite twig girdlers probably would not girdle larger branches if natural popu lation densities were to be significantly increased. 39 X Ccj »s X CO -.^ ^ CU (D CH 4-> t—l ^-1 (\5 (T5 r\3 t^ CU X cu d ?H O VrH C -r-l o r< n3 bJO s quit e > '0 . • r.1 00 c> (0 rO -^ ^ I—< o ,« ^ jH S CU IT) X cu u a cu ^ o > ^ to 2X CD cu r^ -M H .r.1 (tJ X fC| - -~^ >> I—I CNJI % 1 o o LO r-l o3 U CO '^ LO O o • •-I to cu bD 5 .r-l stict a C5 CU i^ (U ^ 2fii ^1 o rO X rM .r-l o o CO M • r-l u Q O $:: 6\ ^ cu V4H Q) to >-i o •-' cu X A -^ "" >s 7i X JH 4J -l-> ? ^ (U ••-' to -M to 1—1 (ti X ?^ 00 LO 4-> CO cu t—I o -^ bJD den : cu cag e c .1-o1 .1-n1 cti -M CO d CO r—1 +-> cu r-J ..-1 rid CO +j o rt ^ cu in >^ r£3 o vO X g X 1-7 CD CU ^ t s o f d ayin g r-l O U ,-H •t-i 1—) cu ^., ftJ I—1 ^ bJD ^1 O !*1 bD •-l-l W cu CO se x ^' w ^1 cu (U ' r-l o '^ CM o LO ^ ° CU o (VJ H rQ 40 *-a A. B. C. Figure 6. Effect of O. rhodosticta Bates on mesquite at a popu lation density of 50/tree: A, Tree before caging; B, Tree after all girdling had stopped; C, Tree with girdled branches removed. Dickens Co., Texas. Autumn, 1970. 41 The cage experiment in Garza County, designed to determine if twig girdlers, O. cingulata (Say), would girdle larger branches if population densities were extremely high showed that no stems over 19 mm in diameter were girdled even at a population density of 100/tree. Sixteen smaller branches were girdled and eggs were laid in some of the larger branches, even though they were not girdled. When this experiment was repeated in September, using the mesquite twig girdler, O^ rhodosticta Bates, at a population density of 26/tree, one branch, 23 mm in diameter, was girdled. These beetles were reared in the insectary from branches collected in Garza County; thus, the beetles had done no previous girdling. Both experiments showed considerable feeding danaage with popu lation densities at these levels. Parasitism, Predation, and Competition The larva is the life stage of the mesquite twig girdler usually attacked by both parasites and predators. No parasitized eggs were observed, and the pupae were attacked by predators on very few occasions. In Garza County, by August 28, 1970, about 34% of the original 8. 1 larvae/branch had died of unknown causes, 22% were killed by parasites or predators, and 44% remained alive. The percent of the larval population that died from exposure to 42 excessive heat when girdled branches broke off and fell on the ground due to high winds was not determined; however observations on many occasions implied that this factor is frequently important under West Texas conditions. The number of larvae/branch decreased as the year progressed and there was a corresponding increase in the number of parasites and predators/branch (Figure 7). By early November, 20% of the larvae had emerged as adults, 8% had not pupated, and 72% had died from attack by parasites or predators or unknown causes. In Ward County, by August 28, 1970, 55% of the original 8. 2 larvae/branch had died of unknown causes, 15% were killed by para sites or predators, and 30% remained alive (Figure 8). The number of larvae/branch decreased and the number of parasites and predators increased as the year progressed. By mid-December, 24% of the larvae had emerged as adults, and 76% had died from attack by para sites or predators or other causes. A wide variety of parasitic Hymenoptera were dissected from galleries of raesquite twig girdler larvae from Ward and Garza Counties. These parasites were found in about equal numbers at both areas. Five families, 8 genera, and 10 to 12 species were identified as parasites of mesquite twig girdler larvae 'Table 5). Apparently all of these parasites attack the larvae in about equal numbers except for Tenuipetiolus n. sp, and braconids, which were much less numerous. 43 CO O « Tl <1^ CU "5 'r*:^ r—I to 5 ^ p. u cr -^-^ CO "^ p. O (ti CU cu P. P. (U fcUO fH ^1 .^ CU o o CO $:: cu CO .r-i CU < CU u cu • r-l CO c bJO cvJ / • r-l CO I—I / O •> CO n tT5 •I-I •4-1 •4-i cu u rr;^ O (U CTJ CO \ 4-> f-l st i \ -t-J • I-I 0 \ t) (J 4-» •4-> CU H Xo to ri (U (T3 \ r—1 to X -y X U-{ rCu o \ CO (U X (U >> u • •-I 0 f-i X u o r^ P. cu (J u CO O to CO X f~l o cti ' 0) ^ CD O N fd >> X •+-> o O (T5 ^1 oJ ^ o CI) X o P. .id •4-) X cti to ra s >-l rv a U-l P ^ O fTi CtJ H o cti rO OH 1—< PI u -A P^ cu >-l bJO •I-I I 1—I—f 1~ 1—\—r "1—I—r vO CM CX) ^ vO CM CO xf' O vO CM CO ^ o o vO vO LT) LO T::t< xf -^ ro CO to to rj CU (U (U • r^ I—I o •*" PI X (U So ™ (ti o r- ^ CQ u o • p cti ^ ,-1 w r-l ^ p X • u•-1 cu .r-l d; r-l UJ o X H to 0 cu cu X to ^-^ p^ c • r-l cu -tlj .—I o o rr-J r-H C *-4 rSii cti u to o o (U ^ ^ u B • o to o O X cti CO cu c; ^1 CU H ri^ ^ ^ • r-l C3 o -M +oJ CO (ti d, tti 0 X '^ ^1 ^ X Q) to v' X rlJ r^ 0 J^ r! u p X P. (U cti (ti cu u 'U I—1 CO X ^+H to . ^ 3^ 0) (ti •f-Pi o 4-> Cti to .TI CO "-• O Cti X cl CO (U _rl pq p o (ti cl ti o ..-1 te s (U f-l >-! :3 CD • r-l u -M (ti rj O (ti to cti P.^^ CO Cti > ^-1 fti (ti PH r-H PI < 00 (U ^1 p t)X) I • r-l T" I I—I—rn—I—I—I i I \ I I r '^ TfOvDCOOO^ O s£)CM OO^OvO CMOO vO v£» in LO'^'^'^ corOcMfvlcM 'qou-e-iq 3;Tnbs9iii/s.io;-Bp9jd pu^ sa^is-ejied JToq:^ pu^ o^Ajisj j9Xp.iT§ §T/>^:^ Jo jaquinu u^aj^ 45 Table 5. Parasites and predators of O. rhodosticta Bates larvae dissr:cted from mesquite branches collected in Garza and V\^ard Counties, Texas, between March and September, 1970. Family Species Parasites Chalcedectidae Euchrysia hyalinipennis Ashm. Pte r omalidae Lampoterma sp. Eupelnaidae Eupelraus bruchivorus (Cwfd. ) Eurytonaidac Eurytonaa magdalidia Aslina. Eurytoma dorcaschemae Ashna. Eurytoma sp. Tenuipetiolous n. sp. Braconidae Odontobracon nigriceps Cam, Iphiaulax sp. Glyptocolastes sp. Predators Cleridae Enoclerus sp. unidentified sp. 46 The stage and condition of parasites or predators of the mes quite tvvig girdlers at various dates are shown in Table 6. In Ward County branches were dissected periodically from March 25, 1970, to Ar.i'il 8, 1971. In Garza County branches were dissected period ically frona March 31, 1970, to January 15, 1971. The highest incidence of parasitism occurred between mid-April and early July. Laboratory rearing-The date or season that twig girdlers were attacked by parasites or predators could not be determined by cage rearing trials. Very few girdlers emerged from branches in the cages. More adult girdlers emerged from those branches collected in August; thus these cages w^ere apparently not suitable for rearing mesquite twig girdlers. The low numbers of beetles reared is probably due to low temperature and the unnatural position of the branches in the cage. Many insects were reared from girdled raesquite branches, some of which were parasites and predators, v/hile others are believed to compete with twig girdler larvae for food and space. The nunierical relationship between mesquite twig girdlers and other insects occupying girdled branches is shown in Figures 9 and 10 for branches collected in Garza and Ward Counties, respectively. The species of insects reared frona branches collected in Ward and Garza Counties were the same, but their numbers differed greatly. All Colcoptera, except Cleridae, compete with ^ rhodosticta Bates 47 u-l 1 o CO to • r-l p it e X X (U CO (ti to cti u (ti cu 1 OunrOOr-iOOOO ^ .^ ^1 ^1 X -—1 fv) •—< (\J '-O o ^ cu (ti +J (ti a, p. cti O X cu X u p -M CO p. cti H (ti •r-l (ti (ti r (J CO ?H nc h cocNjr-or^ooor^ s I I I CO IT) '^^ y^ u (ti CU o (ti 1 t^ r^ vo a^ r- cu (ti cu -^ ^ +3 o p. <1^ X_ Or-S P rH (ti ,« ^ n t f r to •r.l or s b r CU _ cu -t-i o^cjoovbooooo 4-J X o (ti I i i (ti cu ^1 X 1 CN3 '-' « o cu cu PH ^1 (U PH (ti t—! 4-) I—< o o • CO -^ .r.1 o o CU CO to ?-i •^Ot^CXJOO OOsDOO CO p •4-> rP o o "~^, oor^r^cNjcN] oooo o o X cu p -i-> p -^1 o rP (ti cti (ti 1 OrH.—IrHrHOOOO I I I r-< an d rP o M-oi pcu o .tJ cu vO'^OO'-^^CMr^OLOOD i - (^ rv-) o Cr^ 00 a n : IT) r- vO IT) '^ (^ T rva ^ inLOcorsi'—'cNjooh-r^LO > ^ CO [•di e ranc h u <^ (ti iT) IT) o ^H ^ cu , rP d (McvjirisDoOLOvXJiriOLn fvj (M (M CM vD cu r^ • r-l o o • r-l CO +-> Z ti g (ti (ti P u o <-* ^ 0) o rQ cu ^ CO u o U vD (ti ^ " 2 S s' CU tsl T-i ^-1 u ^-1 u >> ^ rQ ^ P. P. >5 (ti (ti Mar . 18 , Oct . 27 , Jan . 28 , Anr . 8 ' Mar . 25 , Jul y 31 , ' <^ < ;^ H^ Ma y 27 , ' Jul y 6 , ' 7 collecte d War d Co . Locatio n ^ <, < ^ Aug . 28 , Aug . 17 , ' an d dat e H branche s O 48 O to TH -»-> .^ C .H X d) to (ti CO (\J t-- o o o U (ti CU (\3 00 •—I ^ ^1 X cu (ti ^ p^ p i^ (U cu .ti P CU O CO 4-) hD ^1 (ti (ti cti ^ o in o o o CU u CO PH f^ g PI g +j to X P cu cu (U i^ bJD rP U CO ^ CO 00 (X) o o o ^^ "J ?^ r^ vo r^ PI (ti (y •^ rQ PI CO U (U O o H-> (M O O O O (ti CSJ u X cu cu PH u p. (U u o cu p J_> VH H (NJ (M (M rH "-^ O ( ) CO t^ 00 1—I p to X +3 ^ cti O Hi (ti X ™ CD PH u PH cu o (ti ^1 p (ti (U CU rP O 00 CvJ O O vO p rH (ti c-> •^ Tti ir^ 00 CO o X (ti X > ti (V^ CO CO r^ CX5 vD O (U .H U (ti (ti J-I &JD -4-> -• rO (ti rP to X cu cu (U X rP -)-> ti in un o o o o (ti cu fV] CM t-O vO v£) IT) P O o u 5:1 .H P X • H J_, Cti 4-> p '^ cu P 0 0 0 .r-l o 0 0 r-- r^ r-- r- CO O r- U CO X r- - .^ ^ (U p CO CO CO I C'i 'J V ^^ (ti ^ o CM ..FQ Er cc CO >> O o >-i T^ CO O O cti •4-i o o (ti cu ^ '0 to ^ • ^—) -i- p •-^ >>00 o o .H rOCM ^ to 00 O O r' ra u "^ r-H ^" N o rH X r' Cti Q .-H .^ g ^ CjflOO u p ^ to rH r. o _^ • r—* -, - z X CU ^1 f:-. ^ -f^ P. O ^ ^ • rH CO r—I H ^^ (U O O o -B a"« (M C rH to XHLJ P .S (U [^^ P f^ rP '^ • H (J cu cti (ti H-» p; Cti [^ X ?-^ O o ^ O • H Cti CU •*-> H o PQ rH O PH (ti CU (U O PH (ti P o I cu cu O X (1) (D cu cu cti (ti 0) • I-H to u X cti g u • H X rH P (U X o rP cu O (ti X to (U CM 0o0 o S irt O cti O I rP « o (ti 0) to .. 0) r- "^ "^ < o O to cu •jj o X CO p p o (ti I-H IT) (ti x; «+H o to rH^ to u r-H H (ti cu O ?H o ao X> ^ P ^^-iO rHQ ^1 X (M .s Z ^ J-I >^ H-> X bJO (ti C O cu CO J-I to -^ o • rH con i le i cu j-i u cu CU te r to CU (U ti (ti cti rP PH rP P rP ^1 4-> • rH o r-H rP rP o o PQ H-> o O fQ o Q o o 51 for food and space, with bostrichids probably the most imporant competitor. The larvae of clerids are presumed to be predacious on both mesquiti-, twig girdler larvae and all other immature insects within girdled branches. Ichneumonids, chalcids, braconids, other Hymenoptera and Diptera are parasites of insects within girdled branches; however, no ichneumonids were dissected from twig girdler galleries. From both sites, numerous bostrichids of the genera Xylobiops and Amphicerus were reared, with Xylobiops emerging in much greater nunabers than Amphicerus. Ichneumonids were reared in the greatest nuiaibers in Garza County (Figure 9), while in Ward County, chalcids v/ere the most numerous parasites (Figure 10). Biological Check Method The mean number of twig girdler larvae in mesquite branches was not significantly increased nor w^as the mean number of parasites and predators significantly decreased by the biological check method used in this study (0. 05 level). Parasites probably attack the larvae in the spring (Table 6), thus the biological check method which was begun in July had little or no effect on parasites and predators of the larvae. The biological check method naight have been effective if ant activity had been increased between April and June; however more branches would probably have to be used to detect significant differ ences between the treatment and the control. 52 Effect of Temperature on Twig Girdler Larvae Survival of mesquite twig girdler larvae was not successfully increased by the methods of habitat manipulation used in this study. In Ward County, girdled branches left intact, in shade on the ground, or hanging in mesquite trees, had significantly more surviving lar vae than branches placed on tall grass, or on bare ground (0. 05 level) (Table 7). In Garza County girdled branches left intact con tained significantly more larvae than any of the treatments, while those placed on the ground in the shade and those hanging in trees contained significantly more than those on tall grass or those on bare ground (0, 05 level) ''Table 7). Branches nearer the ground were exposed to higher temperatures, and had higher larval mor tality. Higher temperatures did not decrease the time required to conaplete the life cycle. Thus, if mesquite twig girdler larvae infested branches are used for colonization, they should be shaded or placed in a situation simulating natural conditions. There were no statistical correlations between the maximum air temperatures corresponding to the five twig placement areas and the mean numbers of larvae surviving under these treatments (0. 05 level); however, it appears obvious that excessive temperatures reduced larval survival. The data obtained in this study did not substantiate High's theory (1915). Our data indicated that unless severed branches were shaded, the larvae would die from excessive 53 Table 7. Mean no. of larvae of ^ rhodosticta Bates per branch under five temperature regimes in Ward and Garza Counties, Texas. June 26 to August 28, 1970. Mean Number of Larvae/Branch 1/ In shade Hanging On tall On bare Location Control on ground in tree grass ground (County) (1) (2) (3) (4) (5) Maximum temp. 40°C 42°C 40°C 46°C 61°C Ward 2. 32 1. 92 1. 86 0. 12 0 Garza 3. 52 1. 88 1. 58 0 0 1/ Means underlined by the sanae line are not significantly dif ferent at the 5% level, ^Duncan's new-multiple range test). 54 heat, even though this study did not produce statistically conclusive results. Temperatviics insid.. tlie galleries would have been a more meaningful meafi^^reiTient and might have resulted in a negative cor relation between temperature and the mean number of surviving lar vae. Branch Method of Girdler Colonization Placing infested branches under the shade of mesquite trees resulted in 3. 5 to 15 times more effective introductions than shading branches with carc'board (Table 8). Covering branches with card- boa.rd may have cooled the branches so much tiiat the larvae did not develop. Even those branches collected in early August and covered with cardboard did not produce many adult beetles, although these branches renaained under the more favorable natural conditions for a longer period. This study indicated that if branches are used for introduction purposes they should be collected in May. This would serve two pur poses; first, by collecting the branches early, only a few parasites and predators wo\ild be introduced; second, when the branches are collected early the twig girdler's gallery is not large enough to weaken the branch, and the branches can be collected without being broken, resulting in death of the larvae. Perhaps, adult emergence 55 (ti 0 0 0 0 o to o U U cu U U U u ti (ti (ti (ti rP X X X N Is) S) to bC O a M ;H O • r-l ti u u J-I u p ^1 (ti (ti (ti (ti (ti (ti X H-> Cti o o ^ ^ ^ O rP u o o u rQ to V4H ^ o cu P cti (ti > M VH ^ p riS W) rP •H o o o (M in '^ (ti rH CU CO CM' CO > rP (ti X CM' in X "^ > P CU ^ •^ o (0 (U to v^ X S o to •^ rP d de r in P cu o o P . X rH 0 «, '^ (ti X r^ P3 rP X o to JH P^^ 4-> Cti (U P (ti rP rP • H CO H-> to p JH CU CU CU s a JH XC U bJD ria 'X X PH (ti cu 0) o H-J CU a (ti X cti (ti cu o (U g a o • FH P ^4-1 r-H (M X in X ti to X u o • H a) J-I CcuU X ^H to rP (ti ^ (U p JH rH CU o 0 • H P rQ to X rO (ti X cu ^1 to JH ^I rP • H (U a rQ (ti CJ bJD rP :3 ^ O p to o c O cti (U O rP J-I rP cti JH o rQ o • H ti rO 4-> X r-: ^ o (ti CM (ti ^H P cu o "^ 6 r-H p H-> cu 0 rQ (ti ^H H-> Z p * > • H O j> in o (U •4-> 1—1 p in 5 ti JH O ti CU g (ti rP bJD U o (U • H 00 ^ cu u t:M :^ ffi cu ^ -B rH .o CM CO| Cti H 57 might be higher if the branches were placed vertically, and well above the soil surface to stinaulate the natural micro-climate. Adult Method of Girdler Colonization The introduction of adult beetles on the Dennis Ranch in Borden County resulted in about 5 times more girdled branches than similar releases at two other locations (Table 9). There is no obvious explan ation for this difference nor the difference in the numbers of girdlers/ tree. The beetles were collected in the same locality and released on the same date at both the Koosman and Dennis Ranches, there fore the females in both introductions had done about the same amount of girdling prior to being collected. The success indices for the Caswell Farms and Koosman Ranches were about equal even though over twice as many beetles were released at the Casw^ell Farms. The branch and adult colonization attempts are compared in Table 10. These comparisons were made on the basis of the number of adult girdlers the branches would have produced had they been left intact and under natural conditions. On this basis the adult release method was raany times more successful than the infested branch method. Both colonization methods would have advantages and disadvan tages. Adult beetles would need to be collected before the females have laid their eggs, while branches can be collected over a period 58 cu ^ (U o ^; o ti CO X rH O ^1 ti d) cu o -J-> cti (ti r-i u CU CU X cu o ^ bc O CO to cu cu to ti cti o PQ to • rH '**'*^ tti -•-> rH to Cti to cti :^ (U p CU rH xa s (U uat i u ^1 rQ rH cu • H Cti H b£ to a > o to rH 0 CU - h- CU ti rP X ti 6 I-H Cti O o • H an d u cti 4-> JH JH (ti to ^1 1^ -M CU CU r-H r-i r-H rP rQ Z ^ o (ti 0 X o > to ti H-> a rP o (U • H J-I O (U u cti . ti ^1 PltH ^ O (ti P * K 0 ^ « (ti o ti O c" (ti rP r-H o (ti w (U bJD C^ O rH to • H • H O (U • H cu a ^ cu H-i ^ X >H d) I-H cti ti a ^ X U to ti C u o o rQ O .H Cti o (ti >> CU o H O H^ Q PQ ^ Q 59 Table 10. A comparison of adult and larval methods of coloniza tion of O. rhodosticta Bates in Texas. 1970. La rvae Adults Evaluation Evaluation Location index 1/ Location index .5/ L-7 Ranch 0. 2 Caswell Farms 56. 0 Crosby Co. Lynn Co. Winder TariTis 0. 6 Dennis Ranch 280. 0 Lynn Co. Borden Co. Stephens Farms 0. 8 Koosman Ranch 59. 0 Borden Co. Dickens Co. 7-Bar Ranch 0. 6 Dickens Co. Scharbra.cr Ranch 2. 5 Andrews Co. Scharbauer Ranch 3.0 Andrews Co, Mean no. branches girdled/52 marked trees 1/ X 10 No. branches introduced X no. adults produced/branch in control. 2/ Mean no. branches girdled/52 marked trees „ ,^5 No. of adults introduced. 60 of several months. Adult beetles must be collected during the brief period prior to girdling and ovipositing activities, thus requiring sevo ral collectors over a short period of time; however, once the beetles have been collected they can easily be released. Large numbers of branches can be collected with a small amount of man power; however, branches are bulky and hard to handle and they also must be protected from heat by sinaulating the natural micro climate as nearly as possible. CHAPTER IV SUMMARY Adult mesquite twig girdlers began emerging in late August and early September. Peak population densities occurred in mid- September and population densities ranged from 0. 1 to 4. 7 girdlers/ tree. These beetles lived from 20 to 30 days. Each female girdled 1 to 2 branches and deposited about 8 eggs/branch, with an eventual brood success of 20 to 24%. the mean diameter of girdled branches was about 9. 5 mm 'grange 6 to 20 mm). Under natural conditions mesquite twig girdlers, O. rhodo- sticta Bates, killed up to 36% of the crown volume of mesquite trees. The removal of the branches from the tree probably reduced the amount of stored carbohydrates available to the plant, thus weaken ing it. In cage trials, at a population density of 26/tree, these girdlers killed 75% of the crown volume of mesquite trees. Increas ing the population density of beetles, did not result in an increase in the mean diameter of girdled branches. Thus, girdlers probably would not attack larger branches even at high population densities. Parasites and predators killed about 20% of the potential brood, while about 55% to 60% of the brood died from unknown causes. Ten to 61 62 twelve species of parasites and two species of predators were found attacking the twig girdler larvae. An experiment designed to aug ment natural populations of girdlers by using ants to discourage parasites was probably unsuccessful because it was implenaented subsequent to the parasite attack. Larval survival was not increased by the larval habitat mani pulation methods used in this study. Unless the branches were pro tected from excessive heat the larvae died. In this study the mean number of larvae surviving/branch under natural conditions and in shaded branches were significantly higher than in non-shaded branches. Natural population of mesquite twig girdlers were extremely localized and isolated geographically. Twig girdlers were released into uninfested areas, using both adults and larvae. Introductions using adults appeared much more successful than larval introductions; however, it will take several years to determine if colonization attempts were successful. The activity of the mesquite trig girdler has been shown to partially control mesquite in several areas in Texas; however, several factors must be considered in evaluating this insect as a biological control agent. In heavily infested areas girdlers reduce the crowns of mesquite thus decreasing its shading effects. This pruning may also reduce the amount of stored carbohydrates available, thus 63 decreasing the vigor and ability of mesquite to compete with other plants for wal.:r and nutrients. Girdler activity has also been observed to cremate conditions favorable for the attack of mesquite by other insects such as flat-headed borers (Buprestidae). The naesquite twig girdler's inability to completely control mesquite may be expl^ilrv d in several ways: (1) the broad success of this insect is fairly high compared to that of other insects; however, the fecundity is very low; and (2) these insects tend to exist in localized infestations arjd are slow to disperse. While girdling activity causes an inimediate decrease in the crown volume of mesquite, this dam age may eventually increase the basal or crown area due to sprout ing of subterranean and aerial lateral buds. Studies concerning the effect of sustained heavy twig girdler activity on the total carbohydrate reserves and vigor of mesquite, and on forage production would contribute significantly to the evalua tion of this insect as a biological control agent. LITER/\TURE CITED Bartlett, B. R. 1961, The influence of ants upon parasites, preda tors, and scale insects. Ann. Entomol, Soc. Amer. 54:543-51, Bates, H. W. 1885. Class Insecta, Order Coleoptera, Tribe Longicornia. Biol. Cent, Amer., Coleopt. 5:367. Craighead, F. C, 1923. North American cerambycid larvae. Can. Dep. Agr. Bull, 27 (n, s. ). DeBach, P. , and B. R. Bartletl. 1964. Methods of colonization recovery and evaluation, p. 402-26. J[n_ P, DeBach, (ed. ) Biological control of insect pests and weeds. Reinhold Publishing Corp, , New Yr.rk. DeBach, P., C. A. Fleschner, and E. J. Dietrick. 1951. A biological check method of evaluating the effectiveness of entomophagous insects. Jour. Econ. Entomol. 44:763-5. Dyar, H. 1890. The number of molts of Lepidopterous larvae. Psyche. 5;420. Essig, E. O. 1926. Insects of western North America. Macmillan k Co. , New York. 460 p. Flanders, S, E. 1951. The role of the ant in the biological control of homopterous insects. Can. Entomol. 83:93-8. High, M. M. 1915. The huisache girdler. U. S, Dep, Agr. Bull, 184. p, 1-9. HoUoway, J. K. 1964, Projects in biological control of weeds, p. 650-70. In P. DeBach, (ed, ) Biological control of insect pests and weeds. Reinhold Publishing Corp. , New York. Howard, L. O. 1900. A troublesome twig girdler of the southwest. U.S. Dep. Agr. Div. Entomol. (n. s. ) Bull. 22:94-5. Linsley, E. G. 1940. Notes on Oncideres twig girdlers. J. Econ. Entonaol. 33:562. 64 65 Meyer, R. E. , H. L. Morton, R. H. Haas, E. D. Robins and T. E. Riley. 1971. Morphology and anatomy of honey mesquite. Agr. Res. Service, U. S. Dep. Agr. Bull. 1423. 186 p. Smith, N. H. , and C. A. Rechenthin. 1964. Grassland restoration-- The Texas brush problena. U.S. Dep. Agr. Soil Conserv. Service, Temple, Texas. Snedecor, G. W. , and W. G. Cochran. 1967. Statistical methods. Sixth Edition. Iowa State Univ. Press. 593 p. Steel, R. G. D. , and J. H. Torrie, I960. Principles and procedures of statistics. McGraw-Hill Book Coiaipany, Inc. , New York. 481 p. Swenson, W. H. 1969. Comparison of insects on mesquite in burned and unburned areas. M.S. Thesis. Texas Tech. Coll. , Lubbock, Texas. 62 p. Ueckert, D. N. , K. L. Polk, and C. R. Ward. 1971. Mesquite twig girdler: a possible means of mesquite control. Jour. Range Manage. 24:116-18.