INFORMATION TO USERS

This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Pagg(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated with a large round black marl<, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal section~; with a small overlap. If necessai":', .sectioning is continued again - beginning below the first row and concinuing on until complete.

4. The majority of users indicate that the textual cont~nt is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "pho,tographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

Xerox University Microfilms 300 North Zeeb Road Ann Arbor. Michigan 48106 76-16,433 FUJII, Jack Koji, 1940- EFFECTS OF AN ENTOMOGENOUS NEMATODE, NEOAPLECTANA CARPOCAPSAE WEISER, ON THE FORMOSAN SUBTERRANEAN TERMITE, COPTOTERMES FOm10SANUS SHlRAKI, WITH ECOLOGICAL AND BIOLOGICAL STUDIES ON C. FORMOSANUS. University of Hawaii, Ph.D., 1975 Entomology

Xerox University Microfilms, Ann Arbor, Michigan 48106 EFFECTS OF AN ENTOMOGENOUS NEMATODE, l.J"EOAPLECTANA

CARPOCAPSAE WEISER, ON THE FORMOSAN SUBTERRANEAN

TERMITE, COPTOTERMES FORMOSANUS SHlRAKI, WITH

ECOLOGICAL AND BIOLOGICAL STUDIES

ON C. FORMOSANUS

A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAII IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN ENTOMOLOGY

DECEMBER 1975

By

Jack Koji Fujii

Dissertation Committee:

Minoru Tamashiro, Chairman Henry Y. Nakasone John W. Beardsley Martin Sherman Wallace C. Mitchell iii

ACKNOWLEDGEMENTS

I gratefully acknowledge Mr. Jerry Wakatsuki. Agriculturist of the Oahu Sugar Co., who allowed me to set-up termite traps in their cane fields; Dr. Asher Ota and Mr. Yuki Inouye of the Hawaii S\lgar

Planters' Association who were instrumental in setting-up termite traps on Kauai- -Grove Farm Plantation; Mr. Po-yung Lai for his valuable assistance in portions of this research.

I am also indebted to the Office of Naval Research Grant

NOOOl4-67 -A-0387 -006 which supported the present study. Special thanks are due to Iny wife, Gail and my sons, Scott and Todd for their patience and understanding. iv

ABSTRACT

The effects of an entomogenous nematode, the DD-136 strain of

Neoaplectana carpocapsae Weiser, in the Formosan subterranean

termite, Coptotermes formosanus Shiraki, were studied. Gross

symptomatology, histopathology and the course of infection of N.

carpocapsae in g. formosanns were determined. The primary mode

of nematode entry into termites anesthetized with carbon dioxide was

via the anus although they were also able to enter via the mouth.

Nernatodes were able to penetrate the termite alimentary tract into

the hemocoel in the regions of the fore-, mid- and hindgut. The

bacteria, Achromobacter neinato~ehilus Poinar and Thomas, associated

with the nematode invaded the termite Inuscle, hemolymph, fat body

and nervous tissues. The nematode in turn invaded the termite

nervous tissue, fat body, salivary gland. muscle tissue and sternal

gland.

There was a significant difference in the body weight of termites

from different colonies. Concentration-mortality studies of ~.

carpocapsae in g. formosanus from two colonies were studied. The

results indicated that weight differences between termites from different colonies had no significance in the susceptibility studies.

The LC values were 2,666 and 3,472 dauerlarvae, respectively. SO

LTSO values were also deterr.ained for termites exposed to various

concentrations of nematodes. v

Seasonal abundance of ~. formosanus workers, soldiers and

alate nymphs from two field colonies were studied over a 18 month

period. There was no definite seasonal pattern of the termite

populations in the colonies studied.

Growth stages based on pronotum width and the number of

antennal segments were determined for ~. formosanus workers,

soldiers and alate nymphs from two colonies. The soldier caste did

not have any growth stages and represented a terminal molt. Seasonal

fluctuation of the relative abundance of the worker and alate nymph

growth stages revealed a definite pattern of development for each

successive growth stage. Approximately 60 and 40 weeks were

required for the smallest growth stage to become the largest for the workers and alate nymphs respectively.

The rate of movement of the worker caste within their galleries in the field was determined by the use of stained termites.

The attractiveness of Douglas fir wood of various densities to the termites in the field was studied. Termites were attracted to the lighter, less dense wood. In addition, as expected, there was a direct correlation with termite numbers and wood consumption. vi

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS. .•..•.....•.....•••....•..•.•....••• iii

ABSTRACT. •.•••.••••••••.•••...••.•••.••••.•..•••••....•• iv

LIST OF TABLES. • •• .•. •• •. ••• •. • •. • •• ••••. •• •• •. •. •. . .• •• viii

LIST OF ILLUSTRATIONS. ..••••..•••••••••.••••..•••••••.•• xi

INTROnUCTIOI\l .•..•...... •.....•...... 1

MATERIALS AND METHODS

Collecting TerITlite s.•.•••••••.••••••••..•..•••.•.••.. . 11 Preparation of the InoculuITl. ••••••••••.•••.•.•.••••••.. 16 Histopathology and Course of Infection. ••.••.••..•••..... 18 Susceptibility Studi ."l •••••••••••••••••••••••••••••••••• 20 Ecological and Field Biology Studies Seasonal Abundance. .••••••••.•••.••.•••••.•..•.• 22 TerITlite Growth Stages

Workers. e ~ •••••• 'j. ••••• •• •• ••• ••• ••• ••• •• • 26 Soldiers ~ ...... 27 Alate NYIllphs ...... 27 Seasonal Fluctuation of the Growth Stages Worlters...... ~8

Alate Nymphs .. II ••••••••••••••••••••••••••• 28 TerITlite Movement Studies. •. •••• ••• •• •• •••. • •. ••. 28

RESULTS AND DISCUSSIONS

G ros s Syrnptolnatology . 35 Course of Infection and Histopathology.••••.•••.•••••.••. 36 Susceptibility Studies Foraging Worker Body Weights.••••••.•••••••••.•. 60 Carbon Dioxide Exposure on Recovery TiITle .•.••••. 62 Susceptibility of Workers to NeITlatodes LC50 Studies.. tJ •••••••••••••••• 0 ••••••••••• 65 LTSO Studies II •••••••••••• 69 Ecological and Field Biological Studies

Seasonal Abundance fool •••••••• 74 Foraging Workers . 76 Soldiers . 92

Alate NYlnphs 0 •••••••••••••• 100 vii

TABLE OF CONTENTS (Continued)

Page

General Biology...... 101 Foraging Workers. •• •• •.• •• •• ••• •• ••• .• ••• • 102 Soldiers...... 112 Alate Nynlphs ...... 114 Seasonal Fluctuation of the Relative Abundance of Foraging Worker Growth Stages .••••••••.•.• 120 Seasonal Fluctuation of the Relative Abundance of Alate Nymph Growth Stages. . •• •• •.•••. •••••. 125 Termite Movenlent Studies. • • ••••••••••••.••...••. 134 Relationship of Wood Weight and the Number of Foraging Workers. .•••••••••••••••.•••••••• 1. 38 Relationship of Foraging Worker Numbers and Wood Consumed •••••••••••••.•••••.•.••.• 144

SUMMARY AND CONCLUSIONS ..•••. 0. •••••••••••••••••••••• 149

LITERATURE CITED. ••.•.••••••••••••••••••••.•••••.••.•.• 155 viii

LIST OF TABLES

Table Page

I THE NUMBER OF COPTOTERMES FORMOSANUS SHIRAKI WORKERS OBSERVED IN SAGGITAL SECTIONS HARBORlNG NEMATODES AFTER VARIOUS EXPOSURE PERIODS TO THE DD-136 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER DAUERLARVAE••••..•.••...•.. 38

II LOCATION OF NEMATODES IN TERMITES AT HOURLY INTERVALS AFTER EXPOSURE .•.•.••••.••••.•.•.•••.. 39

III ANALYSIS OF VARIANCE OF A SPLIT-PLOT EXPERIMENT ON FORAGING WORKER BODY WEIGHTS OF COPTOTERMES FORMOSANUS SHlRAKI FROM THREE CO.·;~ONIES.•••...... 61

IV ANALYSIS OF VARIANCE OF A SPLIT-PLOT EXPERIMENT ON THE RECOVERY TIME OF COPTOTERMES FORMOSANUS SHIRAKI FORAGING WORKERS FROM THREE COLONIES EXPOSED TO CARBON DIOXIDE FOR VARIOUS PERIODS ..•••.•••..•.. 63

V MEAN RECOVERY TIME FOR COPTOTERMES FORMOSANUS SHIRAKI FORAGING WORKERS FROM THREE COLONIES EXPOSED TO FOUR CARBON DIOXIDE EXPOSURE PERIODS.•...... 63

VI PERCENTAGE MORTALITY OF UH-l AND EWA COPTOTERMES FORMOSANUS SHIRAKI WORKERS AFTER SEVEN DAYS EXPOSURE TO THE DD-136 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER DAUERLARVAE (COMBINED VALUES FOR 3 TESTS) ..... 66

VII CUMULATIVE CORRECTED MEAN PERCENT MORTALITIES OF UH-l AND EWA COPTOTERMES FORMOSANUS SHlRAKI WORKERS EXPOSED TO VARIOUS CONCENTRATIONS OF THE DD-136 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER DAUERLARVAE.••.•.••••.••••••••••••.••.•... 70 ix

LIST OF TABLES (Continued)

Table Page

VIII CALCULATED MEDIAN LETHAL TIME (LT ) FOR COPTOTERMES FORMOSANUS SHlRAiP WORKERS FROM THE UH-l AND EWA COLONIES EXPOSED TO VARIOUS CONCENTRATIONS OF THE DD-136 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER. •.••••••••.••••.••.••••••••.••. 75

IX TOTAL WORKERS, SOLDIERS AND ALATE NYMPHS OF COPTOTERMES FORMOSANUS SHlRAKI COLLECTED FROM FOUR TRAPS DURING THE COLLECTING PERIOD AT THE

UH-l SITE r _•••••••••••••••••••••••••••••• 77

X TOTAL WORKERS, SOLDIERS AND ALATE NYMPHS OF COPTOTERMES FORiv10SANUS SHlRAKI COLLECTED FROM TRAPS DURING THE COLLECTING PERIOD AT THE EWA SITE...... 83

XI FORAGING WORKER PER SOLDIER (W IS) OF COPTOTERMES FORMOSANUS SHlRAKI FOR EACH COLLECTING PERIOD DURlNG FEBRUARY 1971 THROUGH JULY 1972 AT UH-l AND EWA SITES ....•...•.••...... "'...... •. • •• 93

XII PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHlRAKI FORAGING WORKER GROWTH STAGES FROM THE UH-l COLONY. .•• •• . •• • .. • •. •. . •. .•• .. •. . .. 107

XIII PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHlRAKI FORAGING WORKER GROWTH STAGES FROM THE EWA COLONY ...... •...... •...•.•..••... .. 108

XIV PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHlRAKI SOLDIER GROWTH STAGES FROM THE UH-l AND EWA COLONIES •.•...•...... •.•.. 113 x

LIST OF TABLES (Continued)

Table Page

XV PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHIRAKI ALATE NYMPH GROWTH STAGES FROM THE UH-1 COLONY. ••••.••••.•••••••••••...••.....•. •• 115

XVI PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHIRAKI ALATE NYMPH GROWTH STAGES FROM THE EWA COLONy •...•....••...... •...•.•. 116

XVII ALATE PREFLIGHT PERIOD FOR SEVERAL

TERMITE SPECIES •...... •...... •.•••.•• 0 •• 135

XVIII DISTANCES TRAVELED WITHIN GALLERIES BY COPTOTERMES FORMOSANUS SHIRAKI FROM FOUR COLONIES...... •...... ••..•.. •. 139

XIX MEAN NUMBER OF COPTOTERMES FORMOSANUS SHIRAKI ON DOUGLAS FIR WOOD OF DIFFERENT WEIGHTS...... 140

XX AMOUNT OF DOUGLAS FIR WOOD CONSUMED BY COPTOTERMES FORMOSANUS SHIRAKI AFTER FOUR WEEKS EXPOSURE. •...... •...... •..•• •. 146 xi·

LIST OF ILLUSTRATIONS

Figure Page

1 Distribution map of the Formosan subterranean termite, Coptotermes formosanus Shiraki •. ..••...... 3

2 Dauerlarva of the DD-136 strain of Neoaplectana carpocapsae Weiser. s second stage cuticle ensheathing third stage dauerlarva. .•••.....••..••...... 9

3 Basic components of the term.ite trap. a five gallon can. b Douglas fir wood used as bait.

c plywood used as the trap cover .••..•.. e • •• •••• •• ••• •• 12

4 Coptotermes formosanus Shiraki crawling on tongue depressor ramps from one tray to collecting boards on another tray. • ...... 15

5 Coptotermes formosanus Shiraki. a worker. b soldier. c alate nymph. •....•••...•.•.•.•...... •.•. 24

6 Coptotermes formosanus Shiraki workers marked with fast green stain. Soldiers with dark head capsules do not feed and did not take-up the stain. •.•..... 30

7 Layout of traps with termites at the Kauai Coptotermes formosanus Shiraki colony. a initial release of marked termites. b second release of marked termites. ••....•....•...••. 32

8 a Layout of termite traps at the Waipio Coptotermes for-1TIOSanUs Shiraki colony. b layout of termite traps at the UH-2 .,9. formosanus Shiraki colony. •...... •••.•...... •..•..••. 33

9 Coptotermes formosanus Shiraki workers killed by the DD-136 strain of Neoaplectana carpocapsae Weiser. n nematodes completely devoured termite with the head capsule remaining. h nematodes in the termite head capsule. •..•.....•...•. 37 xii

LIST OF ILLUSTRATIONS (Continued)

Figure Page

10 Saggital section of Coptotermes formosanus Shiraki worker in the abdominal region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the rectum of the proctodeum ••••.••.••...... 42

11 Saggital section of Coptotermes formosanus Shiraki worker in the abdominal region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the coJ.on of thE:: proctodeum. p protozoa...... 42

12 Saggital section of Coptotermes formosanus Shiraki worker in the thoracic region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the proventriculus (p). m mesenteron...... 44

13 Saggital section of Coptotermes formosanus Shiraki worker in the thoracic region showing the DD-136 strain of NeoapleC';td.na ca~ocapsae Weiser (arrow) in the ~esophagous (0). s salivary gland. t thoracic ganglia. .••.• . ...•...•.•.•....•.•....•...... 44

14 Saggital section of Coptotermes formosanus Shiraki worker in the head region showing the DD-l36 strain of Neoaplectana carpocapsae Weiser (arrow) in the buccal cavity. In mandible. 1 labrum. p pharynx. o oesophagous...... 47

15 Saggital section of Coptotermes formosanus Shiraki worker in the abdominal region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the mesenteron. m malpighian tubules. p proctodeum. f fat bodies. .••..••••••.•..•.•...•... .. 47

16 Saggital section of Coptotermes formosanus Shiraki worker in the thoracic region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) penetrating the proventriculus (p). m malpighian tubule. mg mesenteron. s salivary gland. .•...... •.. 50 xiii

LIST OF ILLUSTRATIONS (Continued)

Figure Page

17 Saggital section of Coptotermes formosanus Shiraki worker in the abdominal region showing a portion of the DD-136 strain of N':~oaplectana carpocapsae Weiser (arrow) beginning to pe:"letrate the mesenteron (m). f fat bodies. c cuticle. h hindgut••.•.•...•.....•..••. 50

18 Saggital section of Coptotermes formosanus Shiraki worker in the head-thorax region showing a cross section of the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) dauerlarvae occurring externally in the cervical region. h head capsule t thorax.~ ..• e ••••••••••••••••••••••• " ••••••••••••••• 51

19 Invasion of the bacteria, Achromobacter nematophilus Poinar and Thomas (arrow), in muscle tissue (:m.) of the head capsule of Coptoterme~formosanus Shiraki worker...... 53

20 a Saggital section of Coptotermes formosanus Shiraki worker in the head region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) invading the subesophageal ganglion (sg). b Saggital section of the same region showing normal subesophageal ganglion (sg). t tentorium. cc circumesophageal connectives. 0 oesophagous...... 55

21 Saggital section of Coptotermes formosanus Shiraki worker in the thorax-abdomen region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) invading the fat body (f). p proventriculus. mg mesenteron. m malpighian tubule. s stomodeum... 57

22 Saggital section of Coptotermes formosanus Shiraki worker in the region of the thorax showing a cross section of the DD-l36 strain of Neoaplectana carpocapsae Weiser (arrow) invading an acini (a) of the salivary gland. p proventriculus. 1 base of termite leg. f fat body. •.•.•.•••....••.•..••...•... 57 xiv

LIST OF ILLUSTRATIONS (Continued)

Figure Page

23 The DD-l36 strain of Neoa.plectana carpocapsae Weiser (arrow) invading muscle tissue of Coptotermes formosanus Shiraki. b bacterial rods of Achromobacter nematophilus Poinar and Thom.as. n nuclei of muscle cells. ••.••...•. 59

24 Saggital section of Coptotermes formosanus Shiraki showing th~ sternal gland invaded by the DD-l36 strc.in of Neoaplectana carpocapsae Weiser (arrow). p posterior portion of the gland. 1 lurrwn of the gland. a anterior portion of the gland. n 4th abdominal ganglion with connectives. m malpighian tubule. s 3r.d abdominal sternite. f fat body. pc proctodeum... 59

25 Concentration-mortality curve for UH-l and Ewa Coptotermes formosanus Shiraki workers exposed to various concentrations of the DD-136 strain of Neoapleetana carpocapsae Weiser dauerlarvae .•..•....• .. 68

26 Time-mortality curves for UH-l Coptotermes £ormosanus Shiraki foraging workers exposed to various concentrations of the DD-l36 strain of Neoaplectana carpocapsae Weiser dauerlarvae ...... 72

27 Time-mortality curves for Ewa Coptotermes formosanus Shiraki foraging workers exposed to various concentrations of the DD-136 strain of Neoaph_~tana carpocapsae Weiser

dauerlarvae 0 • • • • 73

28 Seasonal abundance of Coptotermes formosanus Shiraki foraging workers, soldiers and alate nymphs from the UH-l colony during the sampling period. •. •••. .• .. •• .•. •. 80

29 Temperature and rainfall during the sampling period at the UR-1 colony .. o. ••••••••••••••••••••••••••••••••••• 82

30 Seasonal abundance of Coptotermes formosanus Shiraki foraging workers, soldiers and alate nymphs from the Ewa colony during the sampling period ••••.•.•••.•...•.. 86

31 Temperature a.nd rainfall during the sampling period at the Ewa colony...... 88 xv

LIST OF ILLUSTRATIONS (Continued)

Figure Page

32 Seasonal fluctuations of Coptotermes formosanus Shiraki worker/soldier ratio at the UH-l colony

during the sa:mpling period.••••..•.•...•...••..• n , • , ••• 96

33 Seasonal fluctuations of Coptotermes formosanus Shiraki worker/ soldier ratio at the Ewa colony during the sampling period. •••••••..••••••.•••••.•...•• 97

34 Frequency distribution of pronotum width with respective nmnber of antennal segment groupings of Coptotermes formosanus Shiraki workers fronl the UH-l colony...... 105

35 Frequency distribution of pronotum width with respective number of antennal segment groupings of Coptotermes formosanus Shiraki workers from. the Ewa colony 106

36 Regression of the mean pronotum widths on the number of antennal segments of Coptotermes formosanus Shirald workers from the UH-l and Ewacolonies 110

37 Frequency distribution of pronotum widths with respective number of antennal segment groupings of Coptotermes formosanus Shiraki alate nymphs from the UH-1 and Ewa colonies ..•.•..••••••••....•.. •. 117

38 Regression of the mean pronotum widths on the number of antenna1 segments of Coptotermes formosanus Shiraki alate nymphs from the UH-l and Ewa colonies...... 119

39 Seasonal fluctuations of Coptotermes formosanus Shiraki worker growth stages from February 1971 through July 1972 at the UH-l colony...••.•••••.••..••.•• 122

40 Seasonal fluctuations of Coptotermes formosanus Shiraki alate nymph growth stages from February 1971 through July 1972. . .•. •. • ••••. •.• .•. •. •• .• .. .•. .. .. 127 xvi

LIST OF ILLUSTRATIONS (Continued)

Figure Page

41 a Hollow tile fence with Coptoten.nes forrnosanus Shiraki alates emerging from a vertical crack (arrow) during a May 1972 swarm.. b Close-up of the vertical crack in the fence showing alates emerging. Note the abundance of soldiers guarding the open crack. •••...... •. 133

42 Regression of the mean number of Coptotermes formosanus Shiraki on wood of various weights four weeks after exposure...... •. •. .• .. . .. •.. .• .. •.. 141

43 a Douglas fir wood showing the abundance of lignified (1) summer wood. b Douglas fir wood showing the thickness of spring wood. c Douglas fir wood showing damage by Coptoterrnes formosanus Shiraki workers. Note 'Chat the termites fed primarily on the spring wood leaving the summer wood relatively untouched...... •...... • .. 143

44 Correlation between the amount of Douglas fir wood consumed and the number of Coptotermes formosanus Shiraki foraging workers. •...... 148 1

INTRODUCTION

The Formosan subterranean termite, Coptotermes formosanus

Shiraki, is one of 5 termite species established on most of the major islands in the Hawaiian chain. The other species include the lowland tree termite, Incisitermes irnrnigrans (Snyder); the forest tree termite, Neotermes connexus Snyder; the West Indian termite,

Cryptotermes brevis (Walker) and the Philippine milk termite,

Coptotermes vastator Light. The latter 2 species are known to inflict appreciable damage to wooden structures and other cellulose containing materials on the Islands. .f. vastator was reported from one infestation in Kaimuki, Oahu in 1966 (Bess, 1966) and has not been subsequently collected. However, the Formosan subterranean termite is considered by far the most devastating of all the species and is probably responsible for the bulk of the estimated 2 to 3 million dollar annual damage in Hawaii (Beal, 1967).

The Formosan subterranean termite was originally described from specimens collected in Taiwan (Shiraki, 1909) and is generally considered a species of the Indo-Malayan or Ethiopian regions (Snyder,

1949). In addition to Taiwan, f. form.osanus has been recorded from

China, Okinawa, Japan, Celyon, South Africa, Mainland United States and Hawaii (Bess, 1970) (Figure 1).

The initial introduction of the Formosan subterranean term.ite into the Island of Oahu was probably by m.eans of m.erchant ships from. 2

Figure 1. Distribution map of the Formosan subterranean termite, Coptotermes formosanus Shiraki. 3

i (/) C=' 0 (J) 0 E '- .....0 ~.. en Q) '. .E I .0 '- Q) , ~ et- ..-0 O! 81 • 4

the Orient. For example, Ehrhorn (1919) reported an interception of

termites in shipping boxes for banana plants and yams from the

Philippines. It is generally believed that C. formosanus made its

initial establishment at the water front and radiated from this point.

The presence of C. formosanus in HawaIi was first recorded in

the Hawaiian literature by Swezey (J. 914) from an infestation he

discovered in the floor of the Kamehameha Chapel on December 5,

1913. Swezey also stated that Perkins described termites in Fauana

Hawaiiensis which could have been Coptotermes species. This

observation was confi:rl1."led later (Swezey, 1931) when he found

specimens of c. formosanus which were labeled Hvnolulu, 1907 or earlier in a collection of Hawaiian insE.r:ts sent to t.~e Bishop Museum in 1929 by Perkins. This 1907 date, therefore, is the earliest record of the Formosan subterranean termite in Hawaii.

Prior to 1920, 2 species of Coptotermes, C. formosanus Shiraki and C. intrudens Oshima, were thought to occur in Hawaii. The latter species was described by Oshima (1920) from specirnens collected in

Hawaii; however, the validity of thIS species was questioned although

Snyder (1924) in support of Oshima, postulated that c. intrudens may be a new species derived from C. formosanus after it was intToduced into Hawaii from the Orient. However, Light (1926), after examining the 2 species, decided that the difference between C. intrudens Oshima and C. formosanus Shiraki was not sufficient to warrant species status 5 and synonymized C. intrudens with C. formosanus Shiraki.

Subsequent to the establishment of C. formosanus on Oahu, the termite steadily dispersed to the outer islands. By 1931, approximately 24 years after the initial introduction, C. formosanus appeared and was established on Hawaii, Kauai and Lanai (Fullaway,

1925; 1929; 1932). The Formosan subterranean termite was also found on the Island of Maui in 1933 but was quickly eradicated; however, the termite reappeared in 1963 (Davis, 1964). The

Formosan subterranean termite was recently discovered at Kipu,

Molokai. Moreover, C. formosanus was reported from the 3~uthern

Gulf States on the U. S. Mainland (Scott, 1966; Beal, 1967).

Current control methods for the Formosan subterranean termite are based on 2 basic principles, eo g., preventing the termite access to buildings and if this fails, destruction of the colony. The latter method is difficult since the nest is often located away from the building and is often deep within the soil. There is no reliable method, at present, to detect the colony. Colony destruction is generally considered the most reliable control procedure although preventative techniques still leaves much to be desired. These techniques consist of preconstruction sanitation, "termite-proof" construction of foundations and substructures, metal shields to expose termite activity, pressure treated wood and soil treatment.

Soil treatment generally involves drenching the soil with residual 6 chlorinated hydrocarbons such as chlordane, heptachlor, dieldrin or aldrin. These insecticides have been used extensively in Hawaii, but close surveillance m.ust still be maintained since the termites can penetrate treated soil and have done so in Hawaii despite treatment.

Bea1 and Smith (1971) reported that C. formosanus was able to penetrate dieldrin-treated soil under laboratory conditions. This technique, therefore, although a standard practice in termite control, is not very satisfactory. Iv10reover the long residual nature of these very toxi.c insecticides and their deleterious impact on the environment make it imperative that other more suitable methods of ter-mite control be sought.

One of the alternatives to the use of chemicals is the utilization of insect pathogens such as nematodes. The earliest observation of nematodes parasitizing termites was that of Merrill and Ford (1916) who found head inhabiting nematodes, described by Cobb as Diplogaster labiata, in the termite, Leucotermes lucifigus Rossi, in Kansas.

They found that when the nematode infestation was heavy, the termites died. Pemberton (1928) examined termites from the islands of Celebes and North Borneo and found them to be parasitized by a rhabditoid nematode. These nematodes, always immatures, were found in the head of live termites. He felt that the death of the termites resulted from the maturing nenlatodes. Pemberton attempted to infect Q. formosanus with these head inhabiting nematodes but did not succeed. 7

This study, in part, deals with the possible use of an

entomogenous nematode, the DD-136 strain of Neoaplectana

carpocapsae Weiser in controlling the Formosan subterranean termite.

The DD-136 nematode was originally found in 1954 parasitizing codling

moth larvae, Carpocapsa pomonella (L. ) in an apple orchard at Stevens

City, Virginia (Dutky and Hough, 1955). This undescribed nematode

was designated as DD-136, the accession number given at the time of

collection.

A year earlier in 1953, Weiser (1955) also had found a nematode parasitizing C. pomonella in Czechoslovakia and described it as

Neoaplectana carpocapsae. There were some questions as to whether

DD-136 and N. carpocapsae were the same species. Schmiege (1964) indicated that the 2 nematodes were probably the same species while others (Dutky, ~t. al., 1964; Jackson, 1965; Weiser, 1955) stated there were differences between the 2. Poinar (1967) found that both nematodes were able to interbreed and proposed calling the DD-136 nematode, the DD-136 strain of N. carpocapsae and the original N. carpocapsae, as the Czechoslovakian strain of N. cal'pocapsae.

Although the DD-136 strain of N. carpocapsae was originally found in a lepidopterous host, it has been known to parasitize insects from other Orders such as HYmenoptera, Diptera, Coleoptera,

Hemiptera, Homoptera, Ol'thoptera and Isoptera (Dutky, 1959). In addition, the DD-136 nematode was reported parasitizing a noninsect 8

arthropod belonging to the Class Symphyla (Swenson, 1966). This wide

host range makes this nematode a possible biological control agent for

many insect pests (Welch and Briand, 1961a, b; Tanada and Reiner,

1962; Schmiege, 1963; Moore, 1965; Reed and Carne, 1967; Jaques,

1967; Creighton, et. ale, 1968; Jackson and Moore, 1969; Nash and

Fox, 1969).

The third stage of DD-136 (Figure 2) which is ensheathed within the second stage cuticle is responsible for infecting its host. This

stage is commonly called, the infective juvenile, ensheathed juvenile or dauerlarven (dauerlarvae) as coined by Fuchs (1919). In the present study the third stage nematode will be called the dauerlarva.

The dauerlarvae can be stored for long periods without nourishment.

In addition, the dauerlarvae are associated with a bacterium (Dutky and

Hough, 1955).

Poinar and Thomas (1965) isolated and described the bacterium as Achromobacter nematophilus. Later studies indicated that A. nematophilus was found in the ventricular region of the intestiona1 lumen of the dauerlarvae (Poinar, 1966). The association between the nematode and bacteria is considered a mutua1istic one (Poinar and

Thomas, 1966) since the namatode requires A. nematophilus to reproduce. The bacteria, in turn depends on the nematode for dispersal, protection and a mechanism for entering the host body cavity for its multiplication. Since the DD-136 strain of N. 9

Figure 2. Dauerlarva of the DD-136 strain of Neoaplectana carpocapsae Weiser. s second stage cuticle ensheathing the third stage dauerlarva. 10

carpocapsae c.nd its associated bacteriUlTI was found to be a potential

biological control agent for the Formosan subterranean termite,

studies were undertaken to determine the effects of this nematode in

C. formosanus.

In addition, a thorough search of the literature revealed that there was a paucity of information on the ecology and field biology of

Coptotermes formosanus. Since any good control method must have a

sound ecological and biological base, studies were conducted to obtain this basic information. Population samples were obtained to determine the phenology of the various castes that occurred in the foraging areas.

Although definite instars in the life cycle of the different castes of C. formosanus could not be determined, growth stages which may be considered to be similar to instars were determined. The seasonal fluctuations of the stages of the foraging workers and alate nymphs were recorded. In addition, studies were conducted on wood consumption and movement of the workers within their galleries. 11

MATERIALS AND METHODS

Collecting Termites.

A simple trap was improvised to collect C. formosanus in the

field. The trap consisted of 3 basic parts; a 5 -gallon chemical reagent

can, bait composed of wood bound together with wire and trap cover

(Figure 3). Initial trap construction involved the removal of the can's

lid. The can was then thoroughly washed to insure that no chemical

reagent remained. A rectangular hole about 8 x 15 cm was cut out of the center portion of the can's bottom through which a por.tion of the bait protruded. The bait consisted of 4 pieces of pine wood measuring approximately 2. 5 x 8.0 x 30.5 cm which were stacked on each other with the inner 2 pieces of wood evenly staggered together about 5 cm below the outer 2. The 2 inner pieces of wood, ther.efore, protruded through the hole in the bottom of the can while the bulk of the wood remained within the can. Quarter inch outdoor plywood, 30. 5 cm2, was utilized to cover the trap. A rock approximately 4 to 5 lbs. or an object of comparable weight was placed on the cover to secure it in place.

Trap placement depended on the loc2tion of the termite nest or their foraging galleries. Since C. formosanus nests are difficult to locate, foraging tubes or other signs of termite activity were sought by turning over boards, etc. When evidence of termites were found, wooden stakes approximately 30.5 cm long were driven into the ground. 12

a b c

:F'igure 3. E3.sic components of the termite trap. a five gallon can. b Douglas fir wood used as bait. c plywood used as the trap cover. 13

The stakes were arranged approximately 90 cm apart in a rectangular

grid and were checked periodically. Generally the termites located

the wooden stakes within a week.

Stakes harboring termites were replaced with traps. After

rernoving the st;).kes, a hole was dug around the stake and the trap was

set in place so that approximately 10 cm of the trap was buried in the

soil. In cases where termites were not attracted to the traps, water

was poured into the traps to increase the moisture in the trap which

attracts the termites. The wood in the traps was harvested when a

large number of termites had accumulated. This number depended on

several factors such as colony size, amount of wood in trap, and the

length of trap exposure to the termites. When the traps were placed in

an area where a large termite colony existed, the wood could be

harvested in 3 to 4 weeks.

The wood from the harvested traps was placed in polyethylene

wastebaskets and transported to the laboratory. Empty traps were

replenished with new wood. In the laboratory, 3 white enamel trays measuring 47 x 51 x 2 cm were placed on the work table side by side.

The wood harboring termites was placed on the first tray while the wire binding the wood together were cut. The termites were jarred from the wood into the second tray by tapping the board with the handle

end of a 12-inch screwdriver. This procedure was continued on the

remaining boards. Four pieces of wood measuring approcimately 2.5 14 x 10.0 x 30.5 cm were cleaned and dipped in tap water then stacked on one another and spaced with small pieces of a tongue depressor. A ramp, consisting of a moistened tongue depressor was placed between the termites on tray 2 and the wood on tray 3 (Figure 4). The termites in tray 2 were apparently attracted to the ll'lOist tongue depressors and they traveled along the ramp to the moist wood on tray 3. After a short period, most of the termites congregated on the wood on tray 3 leaving the debris on tray 2. The tongue depressors were used over and over since the termites deposited the trail following pheromone on the tongue depressors which enhanced termite attrnction.

The termites on the moist wood were then jarred onto another tray using the same m.ethod mentioned previously. These termites were relatively free from debris and were considered 'Iclean".

"Clean" termites were either used for test purposes or were held in the laboratory for further use. Termites that were held in the laboratory were housed in either plastic shoe or sweater boxes.

Moist paper toweling was placed on the bottom of the plastic containers and pine wood, cut to size, were then stacked on top of the paper towel. Small pieces of tongue depressors were placed between the wood allowing for greater surface area for the termites. The plastic containers vv"ith termites were placed on a shelf which was covered with black plastic so that the termites were kept in the dark. Periodically water was added to the plastic containers to insure that the termites 15

Figure 4. Coptotermes forrnosanus Shiraki crawling on tongue depressor ramps from one tray to collecting boards on another tray. 16

were maintained in a moist environment.

Preparation of the Inoculum.

The original DD-136 strain of Neoaplectana carpocapsae Weiser

dauerlarvae were obtained from Dr. S. R. Dutky of the USDA

Entomological Research Division, Beltsville, Maryland in June 1968.

This stock inoculum was increased by infecting the late instar greater

wax moth larvae, Galleria mellonella (L.).

The technique used to mass rear G. mellonella in the laboratory

was a modified method used by Dutky, et. al (1962). Larvae of the

greater wax moth were reared on an artificial diet consisting of 130 ml

honey, 130 ml glycerine, 75 ml distilled water, 227 gms Gerber's

mixed cereal, 205 gms Kellogg's Special K, 33 gms soya flour and O. 5

gm Vanderzant's insect vitamin mix. To prepare the medium., the

Kellogg's Special K flakes were blended into a powder and thoroughly mixed with the soya flour. This preparation was incorporated into the

Gerber's mixed cereal. In a separate container the honey and glycerine were combined. The insect vitamin mix was dissolved in the distilled water a~d mixed with the honey-glycerine solution. The liquid portion of the diet was then thoroughly r_ended into the solid portion.

The rearing containe r consisted of a wide-mouthed gallon jar.

Enough rearing medium was placed in the container to cover the bottom

2 inches. Approximately 30 adultG. mellonella were placed with the 17 medium for a period of 24 hours and then removed. This was

sufficient time for the adults to deposit their eggs on the medium. The

jars were covered with a 20 mesh wire screen and the jar lid. A

circular hole 9 cm in diameter was cut out from the center of the lid.

The rearing containers were stored on shelves and held at a mean temperature of 26. 9 0 C (80.4of) with a m.ean high of 27. 60 C (81. 6of) and a mean low of 2 6.20 C (79. 10 F). The relative humidity was maintained at a mean of 52. 3% with a mean high of 59. 9% and a mean low of 44. 7%.

Late instar G. mellonella larvae were infected with the dauerlarvae using a method similar to that described by Dutky, et. a1.

(1964). However dead wax-moth larvae were held in petri dishes lined with filter paper for not more than 5 days to allow the nematodes to multiply before placing them on the nematode collecting traps.

The nematode collecting trap was a modification of the one described by Dutky, et. ale (1964). Petri dishes without covers having the 3 point support were wrapped externally with a Whatman No. 2 qualitative filter paper (18.5 cm circles) and were placed with bottoms up in a stainless steel roasting pan (5 x 25 x 34 cm) and autoclaved for

30 minutes. After cooling, approximately 50 dead g. mellonella larvae were placed on each inverted petri bottom.

A O. 1% formaldehyde-distilled water solution was then added to the roasting pan to a level of a fourth to a half the height of the 18

inverted petri bottoms. Emerging dauerlarvae from the G. mellonella

larvae crawl into the formaldehyde solution and were harvested twice a

week. New formaldehyde solution was added to the pan after each

harvest. The stock suspension of nematodes were cleaned by allowing

the nematodes to settle in a flask then drawing the supernatant off with

a 50 ml syringe. New formaldhyde solution was added to wash the

nematodes. This washing process was repeated at least 3 time s.

Nematode counts were made by placing 99 m.l of a 20% glycerine­

water solution in a 100 rnl volumetric flask. One ml of the stock

suspension was then added to the volumetric flask and thoroughly mixed

with the glycerine solution which kept the nematodes su.spended. Three

replicates of 5 - -0. 1 ml samples were withdrawn from the volumetric

flask and placed on microslides for counting. After the dauerlarvae

from the O. 1 ml samples we re counted, the mean was calculated from

which the concentration of the stock suspension was estimated. Fifty

:ml of the stock suspensions, containing up to concentrations of 2 x 104

nematodes per ml, were stored in 250 ml Erlenmeyer flasks. These

flasks were plugged with cotton and stored at approximately 6 to 7oC.

Histopathology and Course of Infection.

Worker termites were exposed to the dauerlarvae in an inoculation chamber similar to the one used for G. mellonella larvae.

Dauerlarvae held at 6°C were brought to room temperature, agitated, and 2 ml of the stock suspension (8.5 x 103 daurerlarvae/ml) was 19 drawn and evenly distributed on the filter paper in the chatnber.

Termites were carefully aspirated in grcups of 100 and placed in 50 ml beakers and anesthetized with carbon dioxide. These anesthetized termites were then placed in the chamber with the nematodes. The chambers were housed in plastic shoe boxes (9 x 17 x 30 cm) containing moistened paper toweling to prevent dessication.

Two tests were conducted. In the first test workers from the

UH-l colony ~ere anesthetized with carbon dioxide for 4 minutes prior to exposure to the nematodes. Ten live termites were randomly selected and sacrificed at 12 -hour intervals for a period of 48 hours after exposure. In addition dead termites found at the l2-hour intervals were also selected for sectioning. In the second test, UH-l foraging workers were exposed to carbon dioxide for a period of 4 minutes prior to exposing them to the nematodes and 5 termites were randomly selected at hourly intervals .for a period of 27 hours for sectioning.

Termites selected for sectioning were immediately placed into

Carnoy's fixative (6: 3:1) for 3 hours. They were then transferred to

70% ethyl alcohol until dehydrated using the n-Butyl-alcohol technique described by Smith (1943). The termites were impregnated with

Paraplast (melting point 56-570 C) in a vacuum oven. Saggital sections were cut at 10 microns and the sections were stained with Patay's triple stain (Gray, 1958). All observations on sectioned termites were ZO

nlade with the aid of a Wild cOnlpound nlicroscope with Kohler

illunlination. A Nikon EMF photonlicrographic attachnlent was

utilized to prepare figures in this study.

Susceptibility Studies.

Susceptibility studies were conducted on field collected foraging

workers fronl the UH-I and Ewa colonies. Ternlites fronl the

respective colonies were trapped and cleaned as nlentioned earlier.

Clean workers were randonlly divided into groups of 50 with an

aspirator and placed in 8 dranl shell vials. Vials containing ternlites

were placed in a glass anesthetizing chanlber fed with a continuous

flow of COZ. The ternlites were anesthetized with COZ for 8 minutes prior to exposing thenl to the various concentrations of DD-I36

dauerlarvae.

A serial dilution was nlade with the stock suspension (11,000 dauerlarvae/nlI) resulting in the following concentrations: 5,500;

2,570; 1,375; 688; 344; 17Z; and 86 dauerlarvae per nlI. Enough dauerlarvae stock was prepared so that each concentration for the entire test canle fronl the sanle source.

The ternlite workers were not fed a specific dosage of the dauerlarvae; therefore, the ternl concentration was used instead of dosage. Ternlites were exposed to a concentration of dauerlarvae that were e'l"enly distributed over an area of 63. 6Z CnlZ, the area of the 9 Cnl filter paper circle on the bottOnl of the inoculation chanlber. A sinlilar 21

chamber used to infect the Q. mellonella larvae with dauerlarvae was

used in the present test.

Eight concentrations (22,000; 11,000; 5,500; 2,750; 1,375; 688;

2 344; and 172 dauerlarvae/63. 62 cm ) were tested on both UH-l and Ewa workers. Desired nematode concentrations were placed in the chamber by drawing 2 ml of the stock suspension (room temperature) into a pipet and evenly distributing the suspension on the 9 cm filter paper circle. Anesthetized termites were placed in the treated chambers and swirled evenly to distribute the workers on the filter paper.

Controls for each test were inoculated with 2 ml of O. 1% formaldehyde- water solution.

Each test consisted of 3 groups of 50 workers from each colony which were treated with each concentration. There were eight concentrations and each test was replicated 3 times. Mortality was observed and re.corded daily for 7 days. All dp.ad termites were removed. Mortality for all tests was corrected for natural mortality according to Abbott's formula (Abbott, 1925). The regression equations and median-letha,l concentrations (LC50 ) and median-lethal time (LT50 ) with their fiducial limits were determined by probit analysis as described by Finney (1952) with the aid of the Fortran IV probit analysis program developed by Daum (1970) for use in the IBM

Systems 360/30 computer. 22

Ecological and Field Biology Studies.

Seasonal abundance. The seasonal fluctuations of population densities of workers (Figure Sa), soldiers (Figure 5b) and alate nymphs (Figure

5c) were studied for a period of approximately a year and a half in 2 colonies. One colony was located at the base of a hibiscus hedge on the University of Hawaii Manoa campus and was designated as the UH-l colony. The other colony, the Ewa colony, was located under a flume situated in an Ewa sugarcane field owned by tlle Oahu Sugar Company.

Termite traps mentioned earlier were utilized as a method of sampling the various termite castes from both colonies. Four traps arranged iLL ~ straight line approximately 1 m apart were set-up at each location.

Wood used as bait for these traps consisted of 4 commercial

Douglas fir pieces (2.5 x 10.2 x 30.5 em), select merchantable, with

4 sides smooth which were cut from the same board. The 30.5 em pieces were then placed in a drying oven held at 75 0 C with circulating air for a duration of 2 weeks to remove moisture in the wood.

Each piece of wood was weighed individually on a Mettler top­ loading balance immediately after removing them. from the drying oven.

The weights of the 4 pieces of wood was totaled, recorded and written on the side pieces making up the bait with a permanent, waterproof felt pen and tied together with copper wire in the same manner mentioned earlier. These bundles of wood were stored until they were required to 23

Figure 5. Coptotermes formosanus Shiraki. a worker. b soldier. c alate nymph. 24

a

b

c 25

replenish traps.

Sampling from the Ewa and UH-1 colonies was initiated by placing the preweighed bait into the traps on February 2 and 9, 1971,

respectively. Sampling was terminated on July 25, and 31, 1972 from the Ewa and UH-1 colonies. The bait was exposed for 2-week periods prior to sampling during the initial 6 samples. In the remaining 16 samples, the wood was exposed for 4-week periods which provided a better sample.

Total number of workers, soldiers and alate nymphs present in the 4 traps at each location were recorded. Wood harvested from each trap was placed in individual polyethylene wastebaskets and transported to the laboratory. The harvested wood was replaced with new weighed wood.

In the laboratory the termites were "cleanedll in the manner mentioned previously. Alate nymphs and soldiers that were associated with the workers were aspirated individually and counted with the aid of a hand counter. After their numbers were recorded, they were stored in 70% ethyl alcohol for further studies.

Workers were generally so numerous that hand counting was too time consuming. Therefore, the number of workers was estimated volumetrically by pouring termites into a 10 ml graduate cylinder to a volume of 4 ml. Several of these 4 :Lnl aliquots were counted to estimate the number of workers per ml. This figure was used to estimate the 26

total population. Once the workers were counted and recorded, they

were also stored in 70% ethyl alcohol.

Termite growth stages--Workers. To determine the UH-l and Ewa

worker growth stages, the workers preserved in 70% ethyl alcohol

from the seasonal abundance study were used. Subsamples were

obtained from samples collected at approximately 2 -month intervals

from February 1971 through May 1972. This period provided

sufficient time to insure that the termites sampled repr~sented the

worker growth stages.

Glass tubing (5 mm dia. ) approximately 25 cm long was used as

a pipet to draw the subsamples from the jar containing the termites.

The 5 mm diameter glass tubing was sufficiently large enou.gh to

accommodate the largest worker so that all sizes could be drawn without any bias. However, before drawing a sample, the jars were thoroughly agitated so that the termites were evenly suspended to

obtain a random size distribution. From each collection period,

subsamples were taken from all 4 traps where possible.

Workers drawn into the glass tubing were released in a straight line on several layers of Kimwipes to absorb the alcohol. Starting at one end, a total of 50 workers were individually placed in 2 rows on a microslide coated with a thin layer of vaseline to support the termites in an upright position. Care was taken to avoid puncturing the termite integument during the transfer pTocess. Punctured termites were 27 discarded since they shriveled up.

The greatest pronotum width was measured from the dorsal aspect with the aid of an occular micrometer and a dissecting scope using a lOX occular and a 3X objective. Measurements were taken from the pronotum since it was well sclerotized and easy to measure.

The number of antennal segments on both antennae were also counted.

Counts were eliminated in cases where individuals had unequal number of antennal segments since the segments were either fused or broken off. Since additional segments were initiated at the third antennal segment, there were instances when the determination of segmentation was not clearly defined in this region. If the suture on the third segment was not well defined or if there were any doubt as to whether there was segmentation, the segment was considered as one.

Soldiers. Soldiers that were collected during the termite seasonal abundance study from the UH-I and Ewa colonies were subsampled and measured in the same manner as the workers to determine if there were any growth stages. However subsamples consisted of 25 individuals from either 1 trap and/or 4 traps from each sample collection. In counting the number of antennal segments, the method used with the workers was applied to the soldiers.

Alate Nymphs. Alate nymphs from the same colonies were also used to ascertain the number of growth stages. Unlike the workers and soldiers, the alate nymph population was so low that all individuals trapped were 28

IYleasured. The num.ber of antennal segm.ents and the pronotum. width

were rp.corded for each individual in an identical fashion as the previous

castes.

Seasonal fluctuations of the growth s'tages.

Workers. The fluctuation of the worker growth stages was studied only

from.the UB-1 colony by obtaining subsam.ples that '\-vere originally

collected for the seasonal abundance study. Subsam.ples of 50

individuals were m.easured in the sam.e m.anner as the growth stage

determ.inations; however, sam.ples were taken from. the 4 traps that

were collected at all the 2-and 4-week intervals.

Since it was found that the pronotum. width was correlated with the

num.ber of antennal segm.ents, only the pronotum. widths were tak en.

Alate Nym.phs. Data obtained for the UB-I alate nymph growth stages

was used to dete rm.ine the seasonal fluctuation of the alate nymphs.

Term.ite m.ovem.ent studies.

To determ.ine termite m.ovement, the workers from several

colonies were m.arked by feeding them. a vital dye, 0.5% fast green, or a fluorescent dye, 10% Glo-check (The Rinchem Co., Inc., Phoenix,

Ariz.). Two mi of the dye solution was evenly distributed over one of

2 filter papers (Whatm.ans #2--9 Cl.'TI circles) in a 9 cm. petri dish.

Approxim.ately 500-1,000 clean termites were placed in each petri dish

containing the dye treated filter paper. The termites were exposed to the dye for at least 24 hours before they were released. Stained 29

terITIites (Figure 6) retained their dye within their aliITIentary tract for

approxiITIately 3 days before it was excreted. Both dyes appeared to

have no deleterious effects on the terITIites.

To insure that cOITIpatibility was ITIaintained aITIong stained and

nOrITIal individuals in the colony, terITIites used for each release were

obtained froITI the colony being tested. TerITIites froITI 3 colonies

designated as Kauai, Waipio and UH-2 were used in the present study.

The Kauai colony was situated in a Grove FarITI canefield on the island of Kauai. The general layout of this plot is illustrated in

Figures 7a and 7b. This plot consisted of 3 rows (A, B and C) that were 15 feet apart. Rows A, Band Chad 10 traps that were spaced 20 feet apart; however, only those traps containing terITIites were illustrated. An irrigation ditch approxirnately 1 ITIeter deep extending into the canefield was situated between traps nUITIbered 6 and 7 and was connected to the ITIain irrigation ditch in front of row C.

The Waipio colony was located between a road an irrigation ditch in an Oahu Sugar Co. canefield on the Waipio peninsula on the island of

Oahu. The traps were arranged in 4 rows, A, B, C and D, with one trap in row A (Figure 8a). Traps in each row were placed approxiITIately 1 ITIeter apart; however, there were spaces,in each row without traps since they were not placed where wooden stakes were not attacked. Each row in turn was also 1 ITIeter apart.

The UH-2 colony was located on the University of Hawaii NIanoa 30

Figure 6. Coptoterm.es form.osanus Shiraki workers m.arked with fast green stain. Soldiers with dark hea.d capsules do not feed and did not take-up the stain. 31

Figure 7. Layout of traps with termites at the Kauai Coptotermes formosanus Shiraki colony. a initial release of marked termites. b second release of marked termites. KAUAI COLONY RELEASE ~ELEASE AO 0 0 ~ 0 ~ 0 b'-2.0·~ B 0 0 0 C 0 1 . 2 3 4 5 6 7 8 9 10 IRRiGATION DITCH a ROAD

KAUAI COLONY

AO 0 0 0 0 RELEASE . 0 ~20~ B . 0 0 0 b C 0 0 0 0 1 2 3 4 5 6 7 8 9 IRRIGATION DITCH b ROAD W N 33

WAIPIO COLONY

IRRIGATION DITCH

A 0-"-RELEASE '-- B 00 0 0 0 """'\ C 00 00 0 O. D 000 0 ~ 1 2 3 4 5 6 7 8 9 10

ROAD a

BUILDING

UH-2 COLONY . BUILD. AO 0 0 0 0 0 0 0 RELEAtE ·0 B 0 \) C2-39 0 0 0 0 cO 0 0 0 0 0 0 1 2. 3 4 5 6 7 0 0 () o-TREE b ROAD

Figure 8. a Layout of termite traps 'at the Waipio Coptotermes formosanus Shiraki colony. b layout of termite traps at the UH-2 C. forrrlOsanus Shiraki colony. 34

campus. Traps in this plot were arranged in a rectangle consisting of

3 rows of 7 traps (Figure 8b). The rows and traps were 1 meter apart.

Two sides of the plot were bordered by a road and the other 2 by buildings.

Initially 2 batches of fast green stain treated termites, approximately 2, 000 and 4,500, were released at the Kauai colony in traps A-5 and A-9, respectively. Stain treated termites were allowed to circulate within their galleries for 8 -1/2 hours. In a second trial at the Kauai colony, approximately 20, 000 Glo-check treated termites were released in trap B-8. The exposure period in this test was 6 hot:rs.

Two releases each cons:sting of approximately 40, 000 fast green stain treated termites were made at the Waipio and UH-2 colonies in traps A-5 and B-2, respectively. At the UH-2 colony the even numbered traps were collected 1/2 hoar after the release, and the odd numbered traps were collected 1 hour after the releas e. Stain treated termites released at the Waipio colony were allowed to circulate within their galleries for 2 hours.

Termites in the collected traps were cleaned as mentioned earlier.

The presence of one or more stain treated termites from the traps indicated termite movement. Furthest traps from the release trap harboring stained termites indicated the distance traveled by the workers. 35

RESULTS AND DISCUSSIONS

Gross Symptomatology.

Symptoms elicited by Coptotermes formosanus workers parasitized by the DD-136 strain of Neoaplectana carpocapsae were very obvious and could be easily distingu.ished. Lethargic and sluggish movement typified the initial symptoms elicited by the diseased termites. As parasitization by the nematode progressed, the legs of the termite became paralyzed. In many cases the paralysis seemed to progress from the mesothoracic and metathoracic legs forward. The termite was able to pull itself around on its prothoracic legs. When these moribund termites were no longer able to IIwalk, II they remained standing with the only movement being a slight quivering of the legs and a very slow vertical movement of the antennae. These termites eventually ended up on their backs, sides or upright and remained motionless unless poked with a probe object which caused a slight movement of the legs and antennae.

Symptoms characteristic of other insect diseases such as discoloration, changes in size and shape, and digestive disturbances were not observed in the diseased termites. However a peculiar response was elicited by the healthy termites to the diseased individuals.

The healthy termites, in some way, were able to detect the presence of the parasitized termites. The healthy termites immobolized the diseased individuals by chewing off their legs and antennae. The 36 im.m.obolized term.ites were then piled together in what could be called a "graveyard." Moreover the healthy term.ites sealed-off the dead term.ites by covering them. with m.asticated filter paper.

Course of Infection and Histopathology.

At 12-hour intervals after exposure to the nem.atodes, term.ites that were still alive and those killed by the nem.atodes were fixed for sectioning. The sectioned m.aterial revealed that inc reasing the exposure period did not lead to an autom.atic increase in the num.ber of term.ites attacked by the nem.atode. Even after being exposed for 48 hours, there were still the sam.e num.ber of term.ites that did not harbor any nem.atodes. Apparently, those term.ites that were not invaded were not attacked even if the exposure period was greatly increased. On the other hand, with increased exposure, those that were attacked were attacked by m.ore nem.atodes so that there was a substantial increase in the m.ortality of the term.ites attacked by the nernatodes (Figure 9).

These term.ites had nem.atodes in both the alim.entary tract and in the hem.ocoel. These results are sum.m.arized in Table 1. Since the results indicated tha.t som.e nem.atodes had already penetrated the alim.entary tract as early as 12 hours after exposure, a second experim.ent was conducted to determ.ine how early penetration occurred.

The results of the second experim.ent are sum.m.arized in Table 2.

The prim.ary m.ode of entry of the DD-136 dauerlarvae into the alim.entary tract apparently was through the anus. This conclusion was 37

Figure 9. Coptoterrnes formosanus Shiraki workers killed by the DD-136 strain of Neoaplectana ca:rpocapsae Weiser. n nematodes completely devoured termite with the headcapsule remaining. h nematodes in the termite headcapsule. TABLE~ I. --THE NUMBER OF COPTOTE&''\1.ES FORMOSANUS SHlRAKI WORKERS OBSERVED IN SAGGITAL SECTIONS HARBORING NEMATODES AFTER VARIOUS EXPOSURE PERIODS TO THE DD-136 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER DAUERLARVAE

Exposure Live Termites Dead Termites Pel'iod Without Alirnentary Without Alirnentary (Hrs. j N Nematode Tract Hemocoel N Nematode Tract Hemocoe1

12 10 9 1 0 2 0 2 2

24 10 7 3 1 5 0 5 5

36 8 6 2 2 25 0 25 25

48 10 9 1 0 9 0 9 9

w 00 TABLE II. --LOCATION OF NEMATODES IN TERMITES AT HOURLY INTERVALS AFTER EXPOSURE

Alimentary Tract Hemocoel Exposure Termites Foregut Midgut Hindgut Period Without Buccal Oesophagous, Proven- (hrs.) N Nematodes Cavity Crop triculus Paunch Colon Rectum Head Thorax Abdomen 1 5 3 - - - - 1 1 1 2 5 3 - 1 - 1 1 1 3 5 4 1 4 5 2 - 1 - 1 1 - 1 1 1 5 5 1 1 - - 1 2 - 2 6 5 3 - - - - - 1 2 7 4 2 - - -- - 1 1 8 4 2 - - - - - 1 2 9 4 1 - - - - - 1 1 10 5 4 - - - - - 1 1 11 4 2 - - - - 2 12 5 1 - -- 1 1 3 1 13 5 3 - --- 1 2 1 - 1 14 5 3 - - - - 1 1 15 5 3 - - -- - 1 1 16 5 2 - - - - - 1 2 - - 1 17 5 2 - - - 1 - 2 1 18 5 4 - .. - - 1 19 4 1 - -- - 2 - 1 1 1 20 4 0 - - - 1 1 2 1 21 5 2 - - - 1 - 2 2 22 5 2 - - 1 - - 1 1 23 4 2 - - 1 - - - 2 - 1 24 4 3 - - 1 1 - 1 1 1 1 25 5 1 1 1 2 1 3 2 2 1 1 26 5 0 3 2 2 - 2 1 - 2 1 27 5 0 - 1 1 2 2 1 1 2 3 2 w -.D 40

based on the fact that the majority of the sectioned termites had

nematodes in the reetmJ:l (Figure 10) and/or in the colon (Figure 11)

from the very first hour until the experiment was terminated at 27

hours. There were a few termites that had nematodes in the foregut at

the 2 to 5 hour periods but these apparently were able to enter when the

termite was anesthetized. Except for these few instances, nematodes

were not found in the anterior regions, i. ~., mouth, oesophagus or

proventriculus until 22 hours after exposure (Figures 12, 13 and 14).

In the termites, the anus may be the normal site of entry for the

nematodes. The nematodes reappeared in the buccal cavity after 25

hours at about the time the nematodes had penetrated the termite's

hemocoel. The termite becomes moribund and its movements were

greatly retarded or stopped enabling the nematodes to enter the buccal

cavity. Although there was a possibility that the nematodes may have

been ingested when the termites fed on the filter paper or when they

preened each other, this should have resulted in the nematodes being

in the buccal cavity for the entire period since the termites are

constantly engaged in such activity.

In contrast to the situation in the termites, Poinar (1967)

reported that with last instar Galleria mellonella the dauerlarvae entered the alimentary tract via the buccal cavity and the anus; however, he concluded that the ~ ~ route was the most common mode of entry since most of the dauerlarvae were found in the crop and midgut. 41

Figure 10. Saggital section of CoptoterITles forITlosanus Shiraki worker in the abdoITlinal region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the rectUITl of the p rotodeuITl.

Figure 11. Saggital section of CoptoterITles forITlosanus Shiraki worker in the abdoITlinal region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the colon of the protodeuITl. p protozoa. 42 43

Figure 12. Saggital section of Coptoterrnes formosanus Shiraki worker in the thoracic region showing the DD -136 strain of Neoaplectana carpocapsae Weiser (arrow) in the proventriculus (p) m mesenteron.

Figure 13. Saggital sections of Coptotermes formosanus Shiraki worker in t..h.e thoracic region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the oesophagous (0). s salivary gland. t thoracic ganglia. 44 45

In third and fourth instar larvae of Aedes aegypti, the dauerlarvae were

only found to enter via the mouth (Welsh and Bronskill, 1962).

Nematodes wert~ found in the midgut of c. formosanus workers

(Figure 15) as early as 2 hours, but it was not until approximately 17 hours that they appeared in the midgut with any regularity. Bronskill

(1962), on the other hand, reported that in A. aegypti the dauerlarvae

reached the midgut lumen within 2 minutes after exposure. This short period was probably due to the mosquito larvae actively seeking and feeding on the nematodes. In G. mellonella, Poinar (1967) noted that at least 10 hours were required before a nematode was found in the midgut.

The dauerlarvae after entering the alilnentary tract, shed their protective second stage cuticle. Although the cast skins could be found in some host species, in the termite the cast skins could not be differentiated from the gut contents. The nematodes then started penetrating into the hemocoel through the gut wall. In the termite, the dauerlarvae were able to penetrate any part of the gut including the proventriculus (Figure 16), midgut (Figure 17), paunch and rectum.

This differs from A. aegypti where the dauerlarvae only penetrated the proventriculus, and those that failed to penetrate the proventriculus seemed unable to enter the body cavity elsewhere (Bronskill, 1962). In

G. mellonella on the other hand, Poinar (1967) reported that penetration occurred in the midgut. Although no definite study of the mode of penetration has been made, most authors agree that it is accomplished by mechanical pressure. 46

Figure 14. Saggital section of Coptotermes formosanus Shiraki worker in the head region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the buccal cavity. m mandible. 1 labrum. p pharynx. o oesophagous.

Figure 15. Saggital section of Coptotermes formes anus Shiraki worker in the abdominal region showing the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) in the mesenteron. m. malpighian tubules. p proctodeum.. f fat bodies. 47 48

Dauerlarvae were found to penetrate the gut wall of g. formosanus

as early as 4 hours; however, it was not until 23 and 24 hours that the

nematodes were found consistantly in the thoracic and head regions of

the hemocoel. Under controlled conditions where G. mellonella larvae

were microfed, Poinar (1967) first observed nematodes in the hemocoel

after 11 hours. With~. aegypti larvae, nematodes were found in the

hemocoel as early as 10 but usually within 20 minutes (Welsh and

BronskHl, 1962).

Although nematodes frequently occurred externally in the inter­

segmental folds of the abdomen and in the cervical region (Figure 18),

direct penetration through the cuticle was never observed in the termite.

Moreover, nematodes entering the insect hemocoel via the spiracle and tracheal system as reported by Weiser (1964) was not apparent in the present study. Weiser (1962) also reported a darkening of the gut wall at the point of penetration and later this became a dark zone around the body of the dead host among 2 Neoaplectana species parasitizing curculionid and cerambycid beetles. This phenomena was not detected in the nematode infected termite.

Since the nematodes occurring in the termite's hemocoel were prinlarily found in the head and thoracic regions, the nematodes apparently could more easily penetrate the foregut or were carried or inigrated to these regions after penetrating the alimentary tract. One might expect that penetration would occur in the hindgut or the midgut 49

Figure 16. Saggital section of Coptotermes formosanus Shiraki worker in the thoracic region showing the DD-l36 strain of Neoapleetana carpoc1.psae Weiser (arrow) penetrating the proventriculus (p). m malpighian tubule. mg mesenteron. s salivary gland.

Figure 17. Saggital section of Coptotermes formosanus Shiraki worker in the abdominal region showing a portion of the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) beginning to penetrate the mesenteron (m). h hindgut. f fat bodies. c cuticle. 50 51

Figl~re 18. Saggital section of Coptotermes formosanus Shiraki worker in the head-thorax region showing a <::ross section of DD­ 136 strain of Neoaplectana carpocapsae Weiser (arrow) dauerlarvae occurring externally in the cervical region. h head capsule. t thorax. 52

since the nematodes were most commonly found in the hindgut a.nd the midgut does not have any cuticular lining to harnper penetration. The

sectioned termites, however, did not reveal any changes in the gut wall to indicate where the nematodes had penetrated. The presence of extensive muscle tissue in the head and thorax may have some bearing as to the concentration of nematodes in these regions as compared to the abdominal region. As mentioned earlier, Pemberton (1928) found termites parasitized by rhabditoid nematodes which also occurred in the head region. There is also the possibility that the nematode migrated up the alimentary tract from the hindgut to the foregut and then penetrated the gut wall.

Achromobacter nematophilus which is stored in the alimentary tract of the dauerlarvae and released when the latter enters the termite hemocoel, was not detected until 24 hours after exposure. The bacteria initially invade the muscle tissue in the head capsule (Figure 19). The bacterial rods either occurred singly or in pairs. After 36 to 48 hours of exposure, the bacteria were found throughout the body cavity invading muscle, hemolymph, fat body and nervous tissues. The nematodes in turn, were found feeding on the termite l s nervous tissue

(Figure 20), fat body (Figure 21), salivary gland (Figure 22), muscle tissue (Figure 23) and sternal gland (Figure 24).

c. formosanus workers exhibited no defense reactions to the dauerlarvae either by deposition of minute particles of melanin in 53

Figure 19. Invasion of the bacteria, AclJ.romobacter nematophilus Poinar and Thomas (arrow), in muscle tissue (m) of the head capsule of Coptotermes formosanus Shiraki worker. 54

Figure 20. a Saggital seetio:l of Coptotermes formosanus Shiraki worker in the head region show:l.ng the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) invading the subesophageal ganglion (sg). b Saggital section of the region showing normal subesophageal ganglion (sg). t tentorium. cc circumesophageal connectives. o oesophagous. 55

a 56

Figure 21. Saggital section of Coptotermes formosanus Shiraki worker in the thorax-abdomen region showing the DD-l36 strain of Neoaplectana carpocapsae Weiser (arrow) invading the fat body (f). p proventriculus. mg mesenteron. m malpighian tubule. s stomodeum.

Figure 22. Saggital section of Coptotermes formosanus Shiraki worker in the region of the thorax showing a c ros s section of the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) invading an acini (a) of the salivary gland. p proventriculus. 1 base of termite leg. f fat body. 57 58

Figure 23. The DD-136 strain of Neoaplectana carpocapsae Weiser (arrow) invading muscle tissue of Coptotermes formosanus Shiraki. b bacterial rods of Achromobacter nematophilus Poinar and Thomas. n nuclei of muscle cells.

Figure 24. Saggital section of Coptotermes formosanus Shiraki worker showing the sternal gland invaded by the DD-136 strain of Neoaplectana carpocapsae Weiser (arrow). p posterior portion of the gland. 1 lumen of the gland. a anterior portion of the gland. n 4th abdominal ganglion with connectives. rn malpighian tubules. s 3rd abdominal sternite. f fat body. pc proctodeum. 59 60

continuous layers over the nematode as reported by Bronskill (1962) in

A. aegypti larvae or in any other detectable way. Perhaps as in the

case with G. mellonella, the combined killing action of the nematode

and its associated bacteria, A. nematophilus, did not provide sufficient

tilne for the cells to encapsulate the nem.atode (Poinar, 1967).

Susceptibility Studies.

Foraging worker body weights. Size variation was observed among the

foraging workers from 3 c. formosanus colonies designated as UH-1,

UH-3 and Ewa. Since size differences could have some effect on the

susceptibility of the worker to the DD-136 nematode, foraging workers

from the 3 colonies were weighed. From each colony, workers were

randOlnly divided into 4 lots, each containing 6 groups of 50 workers.

These groups were weighed.

An analysis of \.3"riance of a split-plot design was conducted

(Table 3). Results of the analysis indicated that there were no

significant differences in the body weights within the lots from each

colony at the 95% level. The mean body weight of a single foraging

worker from the UH-1, UH-3 and Ewa colonies were 2.478 x 10- 3 + -3 -3 0.0001 g, 3.581 x 10 ± 0.0009 g and 4.584 x 10 + 0.0001 g,

respectively. These weights were significantly different (R = 0.05)

from each other.

It is surprising to discover that the foraging workers from different colonies differ so significantly in size. The reasons for this 61

TABLE III. --ANALYSIS OF VARlANCE OF A SPLIT-PLOT EXPERIMENT ON FORAGING WORKER BODY WEIGHTS OF COPTOTERMES FORMOSANUS SHIRAKI FROM THREE COLONIES

Sum. of Mean Source d. f. F Squares Square

Main Plots: Locations 2 5, 321. 19 2,660.595 2, 992. 795~o:~ Reps. 5 6.38 1. 276 1.435 Main Plot E r ror 10 8.89 0.889

Sub-Plots: Lots 3 8. 31 2.770 2.369 Lots x Locations 6 3.32 0.886 0.757 Sub-Plot Error 45 52.62 1. 169 62

difference, whether it is due to food, age of colony, soil conditions,

genetic, etc. are not known.

Carbon dioxide exposure on recovery time. Preliminary studies

revealed that a larger number of termites were infected if they were anesthetized with CO2 before they were exposed to the nematodes. This was thought to be due to the dauerlarvae having a better opportunity to actively enter the termite while it was anesthetized. If this hypothesis was correct, the longer the termite was anesthetized, the greater would be the infection rate. Thus studies were conducted to determine what CO2 exposure period would result in prolonged anesthesia of the termites. Since the size of the termites could have a direct effect on recovery time, termites from the 3 colonies used in the size studies were tested. Analysis of variance of a split-plot design was conducted

(Table 4). The results indicated that C. formosanus workers from the

3 colonies exposed to CO2 for various periods differed significantly in their recovery time. Moreover there was a significant difference between colonies. Therefore, an additional test was conducted by using

6 replications in groups of 50 termites from 3 colonies (UH-l, UH-3 and Ewa) were exposed to CO for 1, 2, 4 and 8 minutes. Recovery 2 time was recorded when approximately 50% of the termites began to actively move about. This data was analyzed by utilizing the Duncan's multiple range test (Table 5). There were no significant differences in the mean recovery time for UH-l workers exposed £l'om 1 to 8 minutes 63 TABLE IV. --ANALYSIS OF VARIANCE OF A SPLIT-PLOT EXPERIMENT ON THE RECOVERY TIME OF COPTOTERMES FORMOSANUS SHlRAKI FORAGING WORKERS FROM THREE COLONIES EXPOSED TO CARBON DIOXIDE FOR VARIOUS PERIODS

Sum. of Mean Source d. f. Squares Square F

Main Plots: Colonies 2 92,325.00 46,162.50 95. 590~o:~ Reps. 5 1,458.33 291.67 0.603 Main Plot Error 10 4,829.17 482. 92

Sub-Plots: Exposure Periods 3 28,523.61 9,507.87 12. 767~:o:~ Exposure x Colonies 6 7,188.89 1,198.15 1. 608 Sub-Plot Error 45 33,512.50 744.12

TABLE V. --MEAN RECOVERY TIME FOR COPTOTERMES FORMOSANUS SHlRAKI FORAGING WORKEFS FROM THREE COLONIES EXPOSED TO FOUR CARBON DIOXIDE EXPOSURE PERIODSa

Term.ite Exposure Periods (m.in.) Colony 1 2 4 8

UH-l 6. 12a 6.49a 6.46a 6.43a

UH-3 4.S1b 5.33b 5. 68ab 5. 86ab

EWA 4.42b 4.42b 5.l8b 5. 32b

a Mean recovery tim.e followed by letters vertica.lly and underscore by bars horizontally represent significant differences (95% level) when treated by Duncanr s m.ultiple ra.nge test. 64 to COZ. This was as expected since the recov~ry time from anesthesia

should be directly correlated to the depth of anesthesia. Once the termite is completely anesthetized, increasing the length of exposure

should have no effect on recovery time.

UH-3 workers recovered in 4.51 and 5. 33 minutes for 1 and Z minute COZ exposures respectively, and this difference was significant at the 95% level. This may indicate that between 1 and Z minutes of

CO exposure are required to completely anesthetize the UH- 3 Z termites. There was an increa.se in the recovery time of 5. 33, 5.68 and 5. 86 minutes corresponding to Z, 4 and 8 minutes COZ exposures, respectively, but this increase was not significantly different at the

95% level.

The mean recovery time for the Ewa workers for the 1 aud Z minutes exposures and the Z, 4 and 8 minu.te COZ exposure periods showed no significant differences at the 95% level. There was, however, a significant difference between the land 4 minute COZ exposures.

The duration of anesthesia for the smaller workers from the UH-l colony was significantly longer than that of the larger workers from the UH- 3 and Ewa colonies with the exception of the 4 and 8 minute COZ exposures of the UH-3 individuals. Workers from the UH-3 and Ewa colonies, however, elicited no significant c1.ifferences in their recovery time to all 4 COZ exposures (Table 5). 65

TerIl1ites are, in general, able to withstand higher

concentrations of CO2 than Il1any other insects. The differences in

recovery tiIl1e exhibited between colonies, are difficult to explain al­

though the obvious differences between the colonies is in the size of the

individu.al worker. Size, of course, is a function of age, nutrition,

genetics, etc.

Susceptibility of Workers to NeIl1atodes.

The susceptibility of g,. formosanus foraging workers from the

UH-l and Ewa colonies to the DD-136 strain of Neoaplectana

carpocapsae was determined in the laboratory. The mortality of the

UH-l foraging workers exposed to the various concentrations of dauerlarvae for 7 days are presented in Table 6. The LC50 for the UH-l foraging workers was calculated to be 2, 666 dauerlarvae per inoculation chamber with fiducial liIl1its (at the 95% level of confidence) of 2,146 to 3, 312 dauerlarvae. The regression equation for the line was Y :: O. 9352 + 1. 148lX, and the standard error of the slope was

O. 110 (Figure 25).

The Il10rtality of the Ewa workers 7 days after exposure to each dauerlarvae concentration is sUIl1marized in Table 6. The LC50 was

3,472 dauerlarvae with fiducial limits (at the 95% level of confidense) of I, 940 to 6, 704 dauerlarvae per chamber. The regression equation for the line was Y = -0.4361 + 1. 5868X, and the standard error of the slope was 0.190 (Figure 25). TABLE VI. --PERCENTAGE MORTALITY OF UH-l AND EWA COPTOTERMES FORMOSANUS SHlRAKI WORKERS AFTER SEVEN DAYS EXPOSURE TO THE DD-l36 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER DAUERLARVAE (COMBINED VALUES FOR 3 TESTS)

Volume of dauerlarvae Number of dauerlarvae Total number of Mean percentage m.ortality suspensionper inocula- 7 days after treatment per inoculation cham.ber workers treated tion cham.ber (m.l) UH-l Ewa - 2 22,000.0 450 95.8 96.0

2 11,000.0 450 85.8 72.0

2 5,500.0 450 67.2 49.6

2 2,750.0 450 49.4 38.8

2 1,375.0 450 26.0 30.2

2 688.0 450 19.4 29.4

2 344.0 450 12.2 21. 2

2 172.0 450 7.2 16.0

2 0.0 450 1.6 4.8

0' 0' 67

Figure 25. Concentration-mortality curve for UH-l and Ewa Coptotermes formosanus Shiraki workers exposed to various concentrations of the DD-136 strain of Neoaplectana carpocapsae Weiser dauerlarvae. 68

AlllVIHOW IN3:>H3d 000000 o it' ~ III ~ ,." N ,..

z 0

o Lr----,----r----r---~--_r_--r_-__,r_-__r'O,.. o in 0 U1 q III t' &ri &ti ret 'Ilt tv') AlllVIHOW !180Hd 69

Since the Ewa workers were considerable larger and recovered more quickly from C02 anesthesia, it was thought that Ewa ternlites would also have a higher LC50 value than the UH-l workers. Howe-ver, the data indicated that there were no significant differences between the concentration-mortality response of the Ewa and UH-l workers to the

DD-136 dauerlarvae. The significantly longer period that the UH-l workers were inactivated by the CO2 when initially exposed to the nematode did not significantly increase the mortality.

LT50 study. In addition to the LC50 determinations, LT50 (time required to kill 50% of the treated individuals) for several concentrations for UH-l and Ewa workers exposed to the nematodes were also determined. Table 7 summarizes the daily cumulative, corrected mean percentage mortalities for a total of 7 days. The experiment was terminated after 7 days since there was only a negligible change in the total mortality after this period. Only the top

2 and 3 concentrations for the Ewa and UH-l respectively resulted in mortalities greater than 50% 7 days after they were initially exposed although the 2 next highest concentrations for each colony had cumulative mortalities close to 50%. Mortality in the lowest 4 concentrations for each colony did not exceed 27%.

Since the upper 4 and 3 concentrations for the UH-l and Ewa termites, respective ly, surpassed or approached 50% mortality after

7 days exposure, these data were subjected to probit analysis. TABLE VII. --CUMULATIVE CORRECTED MEAN PERCENT MORTALITIES OF UH-l AND EWA COPTOTERMES FORMOSANUS SHIRAKI WORKERS EXPOSED TO VARIOUS CONCENTRATIONS OF THE DD-136 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER DAUERLARVAE

Number of Days After Treatment Cone., dauer­ UH-l EWA larvae per inoculation chamber 1 2 345 6 7 1 2 345 6 7

22,000 17.146.157.766.184.894.795.7 25.0 45. 6 72. 4 82. 5 90. 7 94. 1 95.8

11,000 7.150.662.867.972.679.985.6 22. 8 36.2 48.4 58. 5 62. 9 68. 3 70. 6

5,500 5.840.345.248.857.662.666.6 15.5 25.7 34.7 39.3 42.3 44.7 47.0

2,750 2.026.837.040.143.845.448.5 13.6 21.2 29.0 31.0 33.5 35.1 35.8

1,375 0.5 11.1 18.0 19.6 21.0 22.4 24.8 7.8 14.3 18.0 22.4 23.3 26.5 26.6

688 0.9 3.6 10.0 12.9 15.0 16.4 18.0 4.5 10.1 16.3 20.0 21.5 23.6 25.6

344 0.5 3. 1 5. 6 7.4 8.9 8.9 10.8 4.5 9. 8 12. 5 14. 9 15. 8 17.0 17.0

172 0.0 0.5 1.6 2.7 3.4 4.5 5.6 0.9 4.2 7.4 9.9 10.0 10.2 11.7

-.] o 71

Figures 26 and 27 summarize the time-mortality relationships for UH-l and Ewa workers. The calculated regression line for the UH-l workers for respective concentrations were as follows: Y = 3. 9216 + 2. 9893X;

Y = 3.9570 + 2. 535lX; Y = 3.8531 + 1. 9467X and Y = 3.6223 + 1. 7482X.

With the exception of the highest concentration, mortality among the UH-l workers was very low after the initial day of exposure.

However, there was a dramatic increase in mortality in the second day. Thereafter, with some minor fluctuation, the mortalities were in close G'.greement with the calculated line (Figure 26). The high mortality occurring on the second day apparently was due to the fact that most of the susceptible termites were attacked on the first day.

Mortality, however, was expressed the next day since approximately

24 hours are required before Achromobacter nematophilus kills the term.ite.

The calculated regression lines for the respective concentrations for the Ewa workers were as follows: Y = 4. 1489 + 3.0075X; Y =

4.1823 + 1. 6487X and Y = 3.9679 + 1.l992X. Unlike the UH-l terntites, mortality was relatively high the first day for all 3 concentrations indicating that the termites apparently acquired the dauerlarvae soon after initial exposure. Moreover there was no sharp increase in the mortality rate after the first day as in the UH-l terlnites. The daily mortalities for each dosage level were in close agreement with the calculated line (Figure 27). 72

8.0 98

95 7.5 UH-1 FORAGING WORKERS 90

6.0 80 >- >­ !:: 5.5 70 !::...... J < i$ 60~ ~ o o ~ 5.0-r 50~ f- ... 40% a w o U 8: 4.5 30ei A. mII ...... 1\I 22.000 DAUERLARVAE Y=3.9216+2.9893X . 20 4.0 0.. ---6 11.000 DAUERLARVAE Y=3.95"10+2.5351X 10 OD······O 5.500 DAUERLARVAE 3.5 Y =3.8531 + 1.9467X 5 A It,. 2.750 DAUERLARVAE 'A Y =3.6223+1.7482 X 3.0 3 l' 2 4 6 8 10 20 DAYS AFTER TREATMENT

Figure 26. Time-m.ortality curves for UH-l Coptotermes formosanus Shiraki foraging workers exposed to various concentrations' of the DD-136 strain of Neoaplectana carpocapsae Weiser dauerlarvae. 73

98

,t 95 ~

EWA FORAGING WORKERS H' ,1 .-GG 90 .> . ,I 80 J •• >­ ,f/} tfJPC· 70 !:: >- 5.5 -I ...... , .'" « - " ~.II 60~ ~ •• 0:: I o 0 5.0 , -­ 50~ ~ ,'Av •••.0 Ie­ I .G 40Z-- w , ._6 u -~4.5 , .0 30~ , 0- w ~ A. Q. .r·· 0---022.000DAUERLARVAE 4.0 0-0·" ••.0 Y =4.1489 + 3.0075X . .- Oe•••• O 11.000 DAUERLARVAE 3.5 y= 4.1823+1.6487X " .6._4 5.500 DAUERLARVAE 5 " V =3.9679+ 1.1992X 3.0 3 2 4 6 8 10 DAYS AfTER TREATMENT

Figure 27. Tim.e-m.ortality curves for Ewa Copt<;>termes forrnosanus Shiraki foraging workers exposed to various conc.entrations of the DD-136 strain of Neoaplectana carpocapsae Weiser dauerlarvae. 74

Table 8 represents the calculated LT50 values for UH-l and Ewa c. formosanus workers with 95% fiducial limits. UH-l workers treated

vrith 22, 000 dauerlarvae per chamber had a significantly shorter LT50

than those UH-l individuals treated with 5,500 dauerlarvae but there

were no differences from those treated with II, 000. However, in the

Ewa colony, there was significant differences in LT50 values between

the 22, 000 and the II, 000 dauerlarvae concentrations.

Ecological and Field Biology Studies.

Seasonal abundance. Quantitative data on termite phenology are very

sparse. Observations on the seasonal movement such as soil penetration by termites in relation to the dry season (Grasse and

Noriot, 1948; Hill, 1942) and swarming activities of termites (Nutting,

1969) are a few of the studies recorded in the literature dealing with seasonal termite activity. Sands (1965) with Trinervitermes ebenerianus and Bodot (1967) with 1'. trinervius conducted extensive studies on the populations of these mound building W. African termites.

They found that constructing, foraging and flight activities corresponded to seasonal weather patterns.

Most population studies with termites, however, understandably involved those species that were easily detected and observed such as mound builders. Species such as C. formosanus, which are so cryptic in their habits tho,t their colonies are difficult to detect and even more difficult to observe, were not studied. Moreover, until recently, a 75

TABLE VIII. --CALCULATED MEDIAN LETHAL TIME (LT50) FOR COPTOTERMES FORMOSANUS SHlRAKI WORKERS FROM THE UH-I AND EWA COLONIES EXPOSED TO VARIOUS CONCENTRATIONS OF THE DD-136 STRAIN OF NEOAPLECTANA CARPOCAPSAE WEISER

Concentration, 95 Percent Fiducial Dauerlarvae LT50 Colony Limits per Inoculation (Days) Lower Upper Chamber

22,000 2. 3 1.8 2.8

11,000 2.6 2.0 3. 2 UH-l 5,500 3. 9 3.0 5. 1

2,750 6. 1 ~:~ 4.6 10.9

22,000 1.9 1.7 2.2

EWA 11,000 3. 1 2.9 3. 3

5,500 7. 2~:< 6. 3 8. 3

'I'.f. Projected LT50 values. 76 technique to keep field colonies under prolonged continuous observation was not available. With the developInent of the terInite trap

(TaInashiro, et. al., 1973), an intensive study of the phenology of g,. fOrInosanus was initiated.

The seasonal abundance of workers, soldiers and alate nyInphs of c. fOrInosanus froIn the UH-l and Ewa colonies was followed for approxiInately 18 Inonths froIn February 1971 through July 1972.

Results of these studies are sUInInarized in Tables 9 and 10 and Figures

28 and 30. Weather data during the saInpling period at the UH-l and Ewa colonies are represented in Figures 29 and 31 respectively. These data were gathered froIn the U. S. Weather Service. For the initial 6 saInples froIn both colonies, the wood was exposed for 2 weeks while the reInaining 16 saInples, the wood was exposed to the terInites for 4 weeks.

UH-l foraging workers. The Inean worker population in 4 traps of the

UH-l colony during the first 6 saInpling periods which were collected at

2-week intervals was ~O, 400 ± 9,900. Except for a few saInples, the number of workers in the traps were relatively consistent. A low of

6, 300 workers was recorded in late April 1971; however, in the following saITlple collected in May 1971, the nUInber of worker s increased to 36, 900. The low nUInber of workers in the late April 1971 sample may have been due to the heavier wood used in the traps.

Heavier wood is usually more lignified and harder and is not preferred 77

TABLE IX. --TOTAL WORKERS, SOLDIERS AND ALATE NYMPHS OF COPTOTERMES FORMOSANUS SHIRAKI COLLECTED FROM FOUR TRAPS DURING THE COLLECTING PERIOD AT THE UH-l SITE

Date Alate Collected Workers Soldiers Nymphs

Feb. 23, 1971 16, 000 888 ° Mar. 9, 1971 21,200 802 0

Mar. 23, 1971 19, 900 1,057 0

Apr. 6, 1971 21,300 1,285 0

Apr. 20, 1971 6, 300 338 0

May 4, 1971 36,900 1,893 0

Jun. 1, 1971 18,800 1, 511 0

Jun. 29, 1971 42,100 2,735 43

Jut. 27, 1971 44,900 2,448 19

Aug. 24, 1971 39,800 2,476 18

Sep. 21, 1971 40,800 2,029 6

Oct. 19. 1971 38,800 938 4

Nov. 16, 1971 29,900 1,473 13

Dec. 14, 1971 49,500 1,949 11

Jan. 11, 1972 61,500 2,020 55

Feb. 15, 1972 46,000 825 45

Mar. 14, 1972 23, 300 196 65

Apr. 11, 1972 15,100 692 0 78

TABLE IX. --(Continued) TOTAL WORKERS, SOLDIERS AND ALATE NYMPHS OF COPTOTERMES FORMOSANUS SHlRAKI COLLECTED FROM FOUR TRAPS DURING THE COLLECTING PERIOD AT THE UH-l SITE

Date Alate Collected Workers Soldiers NYnlphs

May 9, 1972 24,500 1,686 0

Jun. 6, 1972 47,500 2,097 0

Jul. 4, 1972 39,600 2,030 1

Jul. 31, 1972 19,800 3,536 1 79

Figure 28. Seasonal abundance of Coptotermes formosanus Shiraki foraging workers, soldiers and alate nymphs from the UH-1 colony during the sampling pedod. 20,000 ° WORKERS 0,.....°" 10,000 ° /0- --0-0-0 / ° °

,0-0.. 5,000 0 1"-0 "0 '0 0/ -0, o \ ,/ '

1,000 ° IoU SOLDIERS B -' Co ~ 500 « V) ~. m \ D a:::: IoU Co "'C/' ' rJJ/ III / =: IoU a:l ~ 100 m ::;) \/Ill Z z 50 13 « IoU ~

10 ~ /~...... ~/~

" ALATE NYMPHS 5 ~-~ ~"""""'A '"A~I 1 l l ~-~ 'I o iii iii iii IIIIIII I j II FEB MAR APR MAY JUN JUl AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUl 00 1971 1972 o 81

Figure 29. Temperature and rainfall during the sampling period at the UH-1 colony. 30

29 28 ,...... 22 I , 27 I\ 20 I \ I\ 26 , \ 18 ,~ JJ 25 , , 16 I , ~ 24 14 i :;) I ., u =;: . , ~ 0:: 23 , 12 w r I Q. .~ I ' 1'\. ./ MEAN MIN. TEMP. ", . ~ ~ 22 \ I ~ , 10 I- , « , ~~I 0:: 21 ,~ \ I s..J , I \ I :=: \ ..... _, I 20 i I l" I ." I 6 e • I I I , I \ " I I I I' I \ RAINFALL' \ I \ 19 J '" \, I \ I \ I 4 /fIIIlO / \ "J \ ,--,.." - \ ~ ,------2 18 / \., I 17 o FEB MAR APR .MAY JUN JUL AUG SEP oel NOV ~EC JAN FEB MAR APR MAY JUN JUL 1971 1972

co tV 83

TABLE X. --TOTAL WORKERS, SOLDIERS AND ALATE NYMPHS OF COPTOTERMES FORMOSANUS SHIRAKI COLLECTED FROM TRAPS DURING THE COLLECTING PERlOD AT THE EWA SITE

Date No, of Alate Collected Traps Workers Soldiers Nyrn.phs

Feb. 16, 1971 4 1,300 39 6

Mar. 2, 1971 4 3,600 121 11

Mar. 16, 1971 3 600 52 0

Mar. 30, 1971 4 2,300 158 9

Apr. 13, 1971 4 1,000 81 1

Apr. 27, 1971 4 500 58 0

May 25, 1971 4 3,800 73 11

Jun. 22, 1971 4 10,900 376 373

Jul. 21, 1971 4 4, 2·00 168 80

Aug. 17, 1971 3 2,600 129 28

Sep. 14, 1971 4 10.800 367 224

Oct. 12, 1971 4 2,900 155 4

Nov. 9, 1971 2 800 25 0

Dec. 7, 1971 3 1,200 40 2

Jan. 4, 1972 3 700 35 2

Feb. 8, 1972 2 200 34 2

Mar. 8, 1972 1 900 17 0

Apr. 4, 1972 1 600 37 0 84

TABLE X. --(Continued) TOTAL WORKERS, SOLDIERS AND ALATE NYMPHS OF COPTOTERW...ES FORMOSANUS SHIRAKI COLLECTED FROM TRAPS DURING THE COLLECTING PERIOD AT THE EWA SITE

Date No. of Alate Collected Traps Workers Soldiers Nym.phs

May 2, 1972 2 1,400 46 3

May 30, 1972 2 1,500 59 3

Jun. 28, 1972 3 2,500 29 2

Jut. 25, 1972 2 1,100 34 12 85

Figure 30. Seasonal abundance of Coptotermes formosanus Shiraki foraging workers, soldiers and alate nymphs from the Ewa colony during the sampling period. 3,000 0\ 0 ./ /~ORKERS 1,000 o /0 o 0...... 0 0, '0_0-0 '\ 500 o 0 . 0-0, LU l\ ..I o /\ Q. ~ oct o \ V) o &:t: 100 o w 0" Q. ffi 50 o~ El~OLDIERS ca El 13, ...... _, ~ :;) ~ ~ / D • Z m-E'! \ ...... Z \/\'n,.....,EI m-m

A 5 . ALATE NYMPHS I A A \

'¥.~~'¥-./ ~. ?-~-~"Y""~ 1/ 6 b. A A )' j o I II ,IIII, I~i-j '''11 III FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL 1971 1972

00 0' 87

Figure 31. Temper'3.ture and rainfall during the sampling period at the Ewa colony. io

29 28 27

26 25

24 P23 w ~D:22 18 t- ~21 w -16 0.. ~20 w 14 i. .... u 19 12 ~ ~ O~ 18 1 Z 17 8D:<

~ 16 \ 6:3 /-',RAJNFAll 15 I 4 e " ,-\ ' \ I 14 • \..J ,I ' 2 , \" , __II .... \ l ----, 0 00 FEB MAR APR MAY JUN JUl AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL 00 1971 1972 89

by the termites. The mean wood weight in the traps collected in late

April 1971 was 1, 193 gms while the mean wood weight in the May 1971

traps was 939 gms. The mean wood weight for the 4 earlier samples

was 1,049 gms.

The first 4-week sample, which was collected in early June 1971, yielded 18,800 workers. This was considerably lower than the following 5 samples which averaged 41, 300 workers. This low worker count was due to the accidental placement of cedar wood in 2 or the 4 traps. Cedar apparently is not preferred by the termites since these 2 traps had only 46 and 27 workers while the other 2 traps containing

Douglas fir had 11,400 and 7, 300 workers.

From June to November 1971, an average of 36,400 ± 13,100 workers were collected in the traps with only minor fluctuations in the numbers collected. However, from mid-November to mid-February

(Figure 28) there was an apparent increase in the number of workers in the traps. A total of 49, 500 in December 1971, 61,500 and 46,000 in

January and February 1972, respectively, was collected. There was a concomitant increase in the number of alate nymphs in January,

February and March 1972 (Figure 28). Since the workers feed and care for the alate nymphs, it appears logical for the foraging worker population to increase when the alate nymph population increases.

Moreover, the alate nymphs occurring in the December 1971, January,

February and March 1972 samples were all late instar individuals which 90

require m.ore food than the younger individuals occurring in the sam.ples

prior to Decem.ber 1971. It is also known that alate nym.phs require

large stores of fat and protein which enable them. to m.ature and initiate

new colonies after the swarm.. That the worker population began to

decrease in March, April and May 1972 to 23,300, 15,100 and 24,500,

respectively, seem.s to substantiate the role of foraging workers in tending the alate nym.phs since the April and May 1972 sam.ples had no alate nym.phs.

The foraging workers increased to 47,500 and 39,600 in June and early July 1972, respectively, following their low population densities.

However, the last sam.pling period in late July 1972 showed a decrease in the num.ber oj. foraging workers to 19,800.

Ewa foraging worker~. The populations of term.ites in the Ewa traps were consistently lower than those 01 the VH traps (Table 10). In addition, there were great fhjl.ctuations from. sam.ple to sam.ple whether they were collected on a biweekly basis, as were the first 6 sam.ples, or collected on a m.onthly basis, as were the rest of the sam.p1es. A high of

3, 600 and a low of 490 workers were recorded during this period. The num.ber of alate nymphs and soldiers also reflected the sam.e fluctuations

(Figure 29). Highs of 2,725 and 2, 100 workers per trap were recorded in June and Septem.ber 1971, and a low of 100 per trap was collected in

February 1972. In addltion, even the traps attacked reflected variability. There were som.e traps that the term.ites ceased to attack 91 after about a year of sanlpling. In SOnle of these traps the ternlites

returned in subsequent sanlples, while in others they failed to reappear.

These fluctuations and variations could not be correlated with any obvious physical or biological phenonlena. Fronl June 1972, the worker population increased to a high of 2,500. This increase in foraging workers also corresponded to a slight increase in alate nynlphs.

The nUnlber of workers collected fronl the UH-1 colony was consistently higher than fronl the Ewa colony. This nlay nlean that the

UH-1 colony cOnlpared to the Ewa colony was (1) nlore active, (2) nluch larger, (3) had less wood available so they had to concentrate on the traps, (4) had the traps placed closer to the nlain nest, or (5) any cOnlbination of these factors.

Although fewer in nUnlbers, the individual worker fronl the Ewa colony was larger than those fronl the UH-1 colony. Individual size, however, according to SOnle workers, is inversely correlated with colony vigor (Nakaginla, : Q59, 1960; Nakaginla, et. al., 1964 and

Shinlizu, 1962). According to these workers, an old declining colony is rnade up of older workers which are larger and less vigorous than those in an active growing colony. In addition to size, they correlated the nUnlber of antenna1 segnlents with age of a colony. The Ewa colony contained nlany workers with 14 segnlented antennae while the UH-1 colony had very few such workers. The great nlajority of workers fronl the UH-1 colony had 13 segnlented antennae. 92

Soldiers. The proportion of soldiers in both the UH-l and Ewa sam.ples

except for one or two instances, rem.ained relatively constant in spite

of the great fluctuations in the total num.bers of term.ites trapped

(Table 11; Figurel:> 32 and 33). The proportion of soldiers in the traps

could have been affected by m.any external factors such as:

disturb&nce of the traps just before collection; a large increase in the

num.ber of workers without a corresponding increase in the num.ber of

soldiers; breaks in the tubes in the traps and by internal changes that

m.ay occur when alates congregate for flight. Since the colonies were

well established and stable, the total num.ber of soldiers in the colony

should not vary drastically from. m.onth to m.onth. However there m.ay

have been m.ovem.ent and concentration in different parts of the colony

at different tim.es of the year.

The m.ean worker/ soldier (W IS) ratio for the entire sam.pling

period for the UH-l colony was 26. From. February to Septem.ber 1971 '

the W /S ratio fluctuated between 12 and 26. In the October 1971 sam.ple

the W /S ratio began to increase in January 1972 from. 30 to a peak of

119 in March 1972. Following this peak, the W /S ratio fell back within the " n orm.alll range during April through July 1972. The lowest W /S

ratio was 6 which was recorded from. the last sam.ple collected in late

July 1972.

Furtherm.ore, out of a total of 86 sam.ples, there were 36 that had 1 or m.ore alate nym.phs. Sam.ples harboring alate nym.phs had a TABLE XI. - -FORAGING WORKER PER SOLDIER (W IS) OF COPTOTERMES FORMOSANUS SHlRAKI FOR EACH COLLECTING PERIOD DURING FEBRUARY 1971 THROUGH JULY 1972 AT THE UH-I AND EWA SITES

UH-l EWA Date No. of All Date No. of An Collected Traps Traps Range Collected Traps Traps Range

Feb. 23, 1971 4 19 15-32 Feb. 16, 1971 3 33 26-77

Mar. 9, 1971 4 26 13-36 Mar. 2, 1971 4 30 6-43

Mar. 23, 1971 4 19 13-22 Mar. 16, 1971 3 11 6-49

Apr. 6, 1971. 4 16 12-63 Mar . 30, 197 J. 4 15 9-15

Apr. 20, 1971 4 19 12-33 Apr. 13, 1971 4 12 9-25

May 4, 1971 4 19 14-35 Apr. 27, 1971 3 8 3-18

Jun. 1, 1971 4 12 8-18 May 25, 1971 3 23 11-31

Jun. 29, 1971 4 15 10-24 Jun. 22, 1971 4 29 17-46

Jul. 27, 1971 4 18 15-30 Jul. 21, 1971 4 25 11-36

Aug. 24, 1971 4 16 11-54 Aug. 17, 1971 3 21 16-36

Sep. 21, 1971 4 20 16-22 Sep. 14, 1971 4 28 20-42 ~ w TABLE XI. --(Continued) FORAGING WORKER PER SOLDIER (W IS) OF COPTOTERMES FORMOSANUS SHIRAKI FOR EACH COLLECTING PERIOD DURING FEBRUARY 1971 ThROUGH JULY 1972 AT THE UH-l AND EWA SITES

UH-1 EWA Date No. of All Date No. of All Collected Traps Traps Range Collected Traps Traps Range

Oct. 19, 1971 4 41 19-78 Oct. 12, 1971 4 19 13-24

Nov. 16, 1971 4 20 17-24 Nov. 9, 1971 2 34 15-61

Dec. 14, 1971 4 25 21-53 Dec. 7, 1971 3 29 26-29

Jan. 11, 1972 4 30 26-55 Jan. 4, 1972 3 21 13-45

Feb. 15, 1972 4 56 44-180 Feb. 8, 1972 3 5 5-6

Mar. 14, 1972 4 119 31-341 Mar. 8, 1972 1 55 - ----

Apr. 11, 1972 4 22 12-32 Apr. 4, 1972 1 16 - ----

May 9, 1972 4 14 11-36 May 3, 1972 2 31 22-43

Jun. 6, 1972 4 23 20-41 May 30, 1972 2 25 8-26

Jul. 4, 1972 4 20 16-26 Jun. 28, 1972 3 86 32-447

Jul. 31, 1972 4 6 3-17 Jul. 25, 1972 2 32 24-54 -.0 ~ 95

Figure 32. Seasonal fluctuation of Coptotermes formosanus Shiraki worker/soldier ratio at the UH-l colony during the sampling period. 120 o 110

100 UH-1 COLONY 0 90 ~ -0:: 80 0::: W -0 70 ....I 0 ~ 60 0:: w o ~ e:::: 50 0 :=. 40 C)a: C) 30 <: e:::: 0 20 o /0'0\ """ '0_0_0-0.....0 C..... 10 '0 o 0 FEB MAR APR MAY JUN JUl AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL 1971 1972

-..0 0' 100

90 -....0 o 80 EWA COLONY eeQ:: Cl:: ~ 70 A c -I °60 ~ o 50 o ffi~ 0::: 0 0 ~ 4 ~ 30 0'0 /O,~ C)- '\, /0,/0...... 0,0 ° ~ 20 °'0 (I \ o.... \,0"0 o 10 o '0 o .0 FEB MAR APR MAY JUN -lUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL 1971 1972

Figure 33. Seasonal fluctuations of Coptotermes formosanus Shiraki worker/soldier ratio at the Ewa colony during the sampling period.

-..0 -.J 98

mean W" /S ratio of 41 with a mode of 22 a.nd a range of 7 - 341. The

remaining 50 samples that were devoid of alate nylnphs had a mean W /S

ratio of 29, a mode of 19 and a ~ange of 3-180. This seems to indicate

that there are fewer soldiers to workers when the alate nymphs are

present in the traps. However a t-test indicated that there was no

significant difference in the W /S ratio between the samples with and

without alate nymphs.

In general, the Ewa soldier population fluctuated directly with the

foraging worker population as with the UH-l colony. The mean W /S

ratio throughout the sampling period was 28. Most of the W /S ratios

recorded fell between 19-34 which was considered the normal range

for the Ewa colony (Figure 31).

The W /S ratios peaked above SO in May 1971, March and June

1972. Unlike the UH-l colony these peaks did not correspond with an

increase in the alate nymph population. However when the Ewa traps

were analyzed, there were 31 samples that harbored 1 or more alate nymphs and these samples had a mean W /S ratio of 27, mode of 26 and a range of 5 -46. The remaining 31 collections without alate nymphs had a mean W /S ratio of 23 with a mode of 14 and a range of

3-77. Although the Ewa mean W /S ratio of those traps with alate nymphs was slightly higher than those traps without alate nymphs (not significant, t-test), the difference between the mean W /S ratios was not as great as the UH-l samples. 99

Nakagima, et. al. (l964), in Japan, observed 2 instances during a

year where the soldiers outnumbered the workers in a laboratory

reared colony of,9. formosanus. A ratio of 2 workers to 3 soldiers

was noted during periods when the population of soldiers was the

greatest. This situation, however, never occurred in this study. The

number of soldiers at no time exceeded the number of workers in the

UH-l and Ewa colonies. There were some instances with other C.

formosanus colonies in Hawaii where there were more soldiers than

workers in the traps. This situation existed only when the wood in the

trap was literally hollowed out by the foraging workers. Under this

condition the foraging workers leave the devoured wood while apparently some soldiers remain behind. Perhaps this situation occurred in the study conducted by Nakagima, et. al. (l964).

Logically it would seem that a larger population of soldiers than workers could not exist for long periods in a colony due to the dependence of the soldiers on the workers.

Bouillon (l970) listed the W /S ratio of 17 species of termites.

Only 1 of the species in the Family Kalotermitidae had a W /S ratio of 15.

The remaining species belonged to the most advanced group of termites,

Family Termitidae, and the W /S ratio among these species ranged from

20 to 300. Nakagima et. al. (1964) reported that the W /S ratio for laboratory reared g,. formosanus was approximately 10 while the ratio of field populations was 19. The mean W /S ratios for the UH-l and Ewa 100

c. formosanus colonies in the present study was 26 and 28,

respectively.

Alate nymphs. The alate nymphs of ~. formosanus made their initial

appearance in late June 1971 in the UH-1 colony (Figure 28). The total

number of alate nymphs trapped in the June 29, 1971 sample was 43.

Twelve weeks following the initial peak, the numbers of alate nymphs

gradually decreased to a low of 4 in October 1971. The number of alate

nymphs then began to increase again with some fluctuation to a high of

65 after a period of 20 weeks in March 1972. Following this peak the

number of alate nymphs fell to zero in the next sample period in April

1972. Twelve weeks elapsed before alate nymphs occurred in the June

and early July 1972 samples.

In the Ewa colony alate nymphs were collected more frequently

throughout the sampling period than in the UH-l colony (Figure 29).

During the period from February to May 1971, the numbers of alate

nymphs fluctuated between zero and 11. The peak numbers occurred

from June to September 1971, ranging from 28 to 37 per sample. In

October 1971 the number of alate nymphs fell drastically to 4 and to

zero in November 1971. Two alate nymphs occurred during December

1971 to early February 1972. Twelve weeks elapsed before the alate

nymphs began to increase to a high of 12 on the last sample.

A discussion of the fluctuation of the numbers of alate nymphs will be presented later. 101

General biology. Although the Formosan subterranean termite is

economically the most important pest in Hawaii, its biology is virtually

unknown. This is, no doubt, due to its cryptic habits and its extremely

long life cycle. A detailed study of its biology requires not only a long

period of time but also techniques to make periodic observations without

seriously disrupting the colony. Although it was beyond the main scope

of this study, some observations were made on the biology of C.

forrnosanus.

c. forrnosanus is a social insect with a highly developed caste

system consisting of reproductives, workers and soldiers. The

reproductive caste is further divided into 3 categories (Miller, 1969).

The first form reproductive (primary or imago) are the winged form or

alates which can disperse and initiate new colonies. The second form

reproductives (alate nymphs, brachypterous neotenies or nyrnphoids)

are nymphal stages of the first form that are capable of becoming

reproductively active without leaving the nest. The third form

reproductives (apterous neotenies or ergatoids) are functional males

and females that look like large workers externally but are more

pigmented.

The workers are sterile males and females without visible

differentiation toward alates or soldiers. They do most of the work of the colony such as tending the queen, the eggs, the young, the soldiers, the alate nymphs, building the nest and galleries, and foraging for food. 102

Soldiers are characterized by defensive adaptations such as their

sickle-shaped mandibles and well sclerotized head. They also produce

a white sticky substance which is expelled through a pore in front of

their head, the fontenell, when disturbed. Their primary function is to

protect the colony from foreign invaders.

In Hawaii, the alates generally swarm during the months of April,

May and June at dusk when the humidity is high and when the air is

relatively still. After they swarm, they land, drop their wingo (dea1ate)

and pair off. They then find a suitable place, such as cracks or crevices in wooden structures where moisture is readily available, to initiate their colony. After entering a suitable niche, the entrance is sealed and they mate, The first eggs are laid about 5 days after mating (Bess,

1970). Initially the primary reproductives care for the eggs and young workers. As these early groups of workers mature, they assume the task of maintaining the colony. Subsequently, the queen virtually

becomes an "egg-laying machine. II At least 4 to 5 years are required before a colony becomes large enough to cause substantial dan~age to a house.

Most of the damage is caused by the older or later instar foraging workers. The actual number of these older instars in the foraging area is not known since many workers remain within the nest.

Foraging workers. An attempt was nlade to determine the nymphal instars of the foraging workers that were collected from the UH-l and 103

Ewa colonies, but this proved to be difficult since there was no known way to sepa.rate one instar from another. Moreover, because of their cryptic habits, and the fact that they consumed their caste skin, molting or the debris of molting could not be observed. However, it was found that the workers in the traps could be segregated into a number of

"growth stages" based on the number of antennal segments and the pronotum width. There was a correlation between the width of the pronotum and the number of antennal segments. The larger termites tended to have more antennal segments although there was overlap of the pronotum widths. Since it was not known whether these antennal groupings represented instars or just variations within an instar, a quantitative study was made to determine how consistent these groupings were in the workers.

The number of antennal segments and pronotum width measurements were obtained from 1, 324 and 736 foraging workers from the UH-l and Ewa colonies respectively. Six antennal groupings were observed in the UH-l workers ranging from 11 to 16 antennal segments.

Five antennal groupings ranging from 12 to 16 segments were observed in the Ewa workers. Those workers with 11 antennal segments were designated as stage 1, 12 antennal segmented individuals stage 2, etc.

Figures 34 and 35 represent the fl'equency distribution of the observed pronotum width measurements for each corresponding antennal segment grouping from the UH-l and Ewa colonies. 104

The data indicated that there were variations in the worker

pronotum widths in each antennal group. There were considerable

overlapping between stages and in the sixth stage from both UH-l and

Ewa colonies the range of pronotum widths fell completely within the

range of the previous stage. There were very few sixth stage

individuals in the UH-l colony but there was a significant number in the

Ewa colony.

In order to determine whether the antennal groupings and pronotum widths conformed to Dyars' law (Dyar, 1890) and did represent growth

stages statistically, linear regressions were calculated (Snedecor, 1966) from the data in Tables 12 and 13. The regression equations that were calculated for the UH-l and Ewa foraging workers were Y = O. 0669X ­

0.2703 and Y = 0.0690X - 0.1984, respectively.

The regression lines (Figure 36) show the mean pronotum widths for each antennal grouping for the UH-l colony conformed closely to the calculated line while there was more variation evident in the Ewa colony.

Part of the variation could have been due to the smaller sample size obtained from the Ewa colony. The regression coefficients for both lines 0.0669 for the UH-l workers and 0.0690 for the Ewa workers were very similar indicating that the pronotum widths of workers from both colonies increased proportionately at approximately the same rate.

The data seem to indicate that the antennal groupings do represent growth stages in the life cycle of the worker (Tables 12 and 13).

Although molting was not observed, there was a correlation between the 105

2701 260 UH-1 FORAGING WOKERS 230 1 2201 190

180

170

100

90 >­ u wz 80 ;:) S 70 0:: w. ov

50

40

30

20

10 o

PRONOTUM WIDTH (mm.) 11 12 13 14 15 16 ANTENNAL SEGMENTS

Figure 34. Frequency distribution of pronotum width with respective number of antennal segment groupings of Coptotermes formosanus Shiraki foraging workers from tle UH-l colony. 106

190

180 EWA FORAGING WORKERS 170

130

120

110

100

u> 90 % IoU a:::l 80 UJ ....0=: 70

60

50

40

30

20

10

0 ltl.ot--co N"l:t.oCO OM.o~ ~~.ot--t-- t--t--cococo cocococo .. PRONC>TUM 'WiDTH' (mm·.)· .. 12 13 14 15 16 ANTENNAL SEGMENTS

Figure 35. Frequency distribution of pronotum width with respective number of antennal segment groupings of Coptotermes formosanus Shiraki foraging workers from the Ewa colony. 107

TABLE XII. --PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHIRAKI FORAGING WORKER GROWTH STAGES FROM THE UH-1 COLONY

Growth Stages

I II III IV V VI

Number of Antennal 11 12 13 14 15 16 Segments

Range of Pronotum ----- .4494- · 5457- . 6099- .7062- .7383- Width (mm) .5778 · 6741 .7704 • 8025 . 8025

Mean Pronotum .4494 .5250 · 6210 .6934 .7438 .7650 Width (mm)

l'v1ode Pronotum .4494 .5136 .6420 .7062 .7383 .7704 Width (mm)

St. Dev. of Pronotum - ---- .0202 .0310 .0199 .0158 .0242 Width

Growth Ratio --- -- 1. 1682 1. 1828 1. 1166 1. 0727 1. 0285

N 6 146 600 444 122 6

Percent of Sample . 5 11. 0 45.0 34.0 9.0 0.5 108

TABLE XIII. --PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHIRAKI FORP...GING WORKER GROWTH STAGES FROM THE EWA COLONY

Growth Stages

I II III IV V VI

Number of Antenna! 11 12 13 14 15 16 Segments

Range of Pronotum - -- -- .5457- . 6099- .7383- . 8025- .8346- Width (rom) . 6099 .7383 .8667 .8988 .8988

Mean Pronoturn - ---- .5618 . 7695 • 7973 .8398 . 8696 Width (mm)

Mode Pronotum - - --- .5457 .7062 .8025 .8346 .8667 Width (mm)

St. Dev. of Pronotum ----- .0243 .0350 .0266 .0200 .0165 Width

Growth Ratio ------1.3687 1.0361 1. 0533 1. 0355

N 0 8 73 247 286 122

Percent of Sample 0 1.0 10.0 34.0 38.0 17.0 109

antennal groups and the width of the pronoturn indicating that these

stages were achieved by successive molts. In some of the Termitidae,

workers molt without any obvious growth except for the antennae and

legs while in the Reticulitermes (Rhinotermitidae) the workers may go

through an indefinite number of stages. ~. formosanus appears tCJ

resemble the more advanced Termitidae in this characteristic.

Nakagima et. ale (1964) found the Japan~. formosa.nus workers

reared in the laboratory had 12 to 16 antennal segments with 2

individuals in the sample with 17 antennal segments. Therefore, there

was a general agreement in the number of antennal segments of C.

£ormosanus workers both in Hawaii and in Japan.

A significant finding which emerges when the workers from the

Ewa and UH-l colonies are compared is the fact that the average Ewa

worker is larger and has one more antennal segnwnt than the UH-l

workers. The reason or reasons for these differences are not fully

understood. One possibility may be nutritional since the Ewa termites

were feeding on a wooden structure supporting a flume in a cone field.

The lumber was rotted by fungus which may have enhanced the

nutritional value of the wood (Sands, 1969). Moreover they had a

constant moisture source from leaks in the flume.. The UH-l termites

on the other hand were feeding primarily on dead hibiscus bushes.

Another explanation which is actually in 2 parts is based on findings of Nakagima (1959; 1960), Shimizu (1962) and Nakagima, et. al.

(1964). First they felt that body size and weight of Q. formoscmus 110

0.9 FORAG ING WORKERS

-:- ~ 0.8 -::c -I­o ?; 0.7 ~ ::J I-o z ~0.6 0.. z

0.4 1..-_""- ..... -.&. -01'-- -.J .A.-_ol! 11 12 13 14 15 16 ANTENNAl SEGMENl'S

Figure 36. Regression of the mean pronotum width on the number of antennal segments of Coptotermes formosanus Shiraki foraging workers from the UH-l and ~wa colonies. 111

depended on the vitality of the colony. The more active growing

colonies had small workers while in colonies where the vitality had

decreased, the workers were large. According to these researchers,

this was partially due to the fact tha.t an increased burden is placed on

the worker in an actively growing colony. This they felt prevented the

workers from becoming large. Secondly they stated that worker size

was correl~ted to the stage of development of the colony.

Shimizu (1962) stated that C. formosanus colonies in Japan could

be categorized into the following stages on the basis of their strength and vitality; incipient stage, middle stage, accomplished stage, terminal stage and resting nest. These categories were based on the percentage of workers with a given number of antennal segments (14-17) in each nest. Counts (Table 12) had indicated that in the UH-1 colony

45% of the workers had 13 segments and 34% had 14 segments in the antennae. In the Ewa colony (Table 13) 34% had 14 and 38% had 15 segments in the antennae. According to the classification developed by

Shimizu, the UH-1 colony was in its middle stage and the Ewa colony in the accomplished or terminal stage. The Shimizu classification probably cannot be applied in its entirety in Hawaii since the colonies do not overwinter or nest. Moreover, the classification is based on laboratory colonies and may not be indicative of the situation in the field. The antennal criteria probably is more reliable than size alone.

Nutrition and temperature seem to have r.aore influence on size than ,'. 112

the stage. Some highly active colonies in Hawaii are composed of large

individuals. Moreover measurements of the soldiers and alate nymphs

indicate that the difference in antennal segments may be more

ascribable to genetic differences in the colonies.

Soldiers. The number of antennal segments and the widths of the

pronotum of soldiers from the UH-l and Ewa colonies are presented in

Table 14. Almost all of the 1, 000 soldiers examined (98.4%) from the

UH-l colony had 14 segments in the antennae. Only 1. 6% had 15 segments

and there were no individuals with other than 14 or 15 segments. The

widths of the pronota were very similar in both the 14 and 15

segmented individuals.

In the Ewa colony, however, there were 4 different antennal

groupings in the 429 soldiers examined (Table 14). They were found to

have from 13 to 16 antennal segments. The great majority (86.2%),

however, had 15 segmented antennae. There were 1. 4% with 13, 5. 8%

with 14 and 6. 6% with 16 segmented antennae. The pronotal widths,

surprisingly, were similar for all of the soldiers whether they had 13

or 16 segments in their antennae.

In both the UH-l and Ewa colonies, the soldiers, in contrast to

the workers, almost all had the same number of antennal segments.

Although there appeared to be different growth stages according to the

number of antennal segments, most of the soldiers were in the same

stage and not evenly divided as in the workers. This would seem to TABLE XIV. - -PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHlRAKI SOLDIER GROWTH STAGES FROM THE UH-1 AND EWA COLONIES

Number of Antennal Segments

UH-1 Colony Ewa Colony 14 -1.5 13 14 15 16

Range of Pronotum .6741- . 7383- . 8667- . 8346- . 8346- .8667- Width (mm) .8343 .8667 .8988 .9309 .9630 .9309

Mean Pronotum . 7589 .7884 . 8881 .8834 .8908 .8954 Width (mm)

Mode Pronotum .7704 .8025 .8988 . 8988 .8988 .8988 Width (mm)

St. Dev. of Pronotum .0265 .0331 .0166 .0247 .0185 .0202 Width

N 984 16 6 25 370 28

Percent of Sample 98.4 1.6 1.4 5.8 86.2 6.6 I-' I-' W 114

indicate that the predominant soldier stage was a terminal stage and

those soldiers with lTIore antennal seglTIents metamorphosed from

workers already having as lTIany segments as the predominant stage.

The fact that soldiers can develop from workers has been observed in

Reticulitermes 1ucifugus (Rossi) by Buchli (1956). As in the case of the

workers, the UH-1 soldiers were smaller than the Ewa soldiers.

Alate nymphs. A total of 276 and 261 alate nymphs from UH-l and Ewa

colonies, respectively, were trapped, observed and measured. The

results, presented in Tables 15 and 16, indicatE: there were definite growth

stages. Four antenna1 groupings were observed from the UH-l colony which ranged from 14 through 18 seglTIents; the 15 segmented individuals were not present in the samples. Five antennal groupings were found among the alate nymphs from the Ewa colony. These antennal groups ranged from 15 through 19 segments. The frequency distribution of observed pronotum width measurements for each corresponding antennal segment grouping from the UH-l and Ewa colonies are presented in Figure 37.

To obtain the lTIissing growth stage from the UH-l colony, (the 15 segmented antennal individuals) a method outlined by Taylor (1931) was used. The pronotum width for this stage was deterlTIined by utilizing the ratios derived by the division of an actual measurement by the stage immediately following it. The ratios between the 16 and 17 seg:mented individuals and the 17 and 18 segmented individuals were averaged. This 115 TABLE XV. --PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FOFMOSANUS SHlRAKI ALATE NYMPH GROWTH STAGES FROM THE UH-1 COLONY

Growth Stages

I II~~ III IV V VI~~

Number of Antenna1 14 15 16 17 18 19 Segments

Mean Pronotum .7729 . 8416*~:~ .9259 1.0246 1. 1205 ----- Width (mm)

Mode Pronotum .7704 - ---- · 9309 1. 0272 1. 1235 ----- Width (mm)

Calculated Pronotum . 7615 .8493 · 9371 1.0249 1. 1127 - ---- Width (mm)

St. Dev. of Pronotum .0198 - ---- .0232 .0403 .0598 ----- Width

Range of Pronotum . 7383------· 8667- . 9309- 1.0914------Width (mm) .8025 .9630 1. 1235 1. 1556

Growth Ratio ----- 1.0889 1. 1002 1. 1066 1. 0857 - -- --

N 13 0 32 25 206 0

~:~ Stages II and VI were not found in the samples.

~~>;~ Pronotum width derived by utilizing the Dyar ratio of increase as outlined by Taylor (1931). 116

TABLE XVI. --PRONOTUM WIDTH MEASUREMENTS OF COPTOTERMES FORMOSANUS SHIRAKI ALATE NYMPH GROWTH STAGES FROM THE EWA COLONY

Growth Stages

I'"',' II III IV V VI

Number of Antennal 14 15 16 17 18 19 Segments

Mean Pronotum - -- -- • 8667 .9523 1.0686 1. 1745 1.2244 Width (mm)

Mode Pronotum ------.9630 1. 0593 1. 1877 1. 2198 Width (mm)

Calculated Pronotum - ---- .8697 . 9635 1. 0573 1.1511 1.2449 Width (mm)

St. Dev. of Pronotum - -- -- • 0321 .0389 .0258 .0308 .0242 Width

Range of Pronotum - ---- . 8346- .8998- . 9951 - 1.0914- 1. 1556- Width (mm) .8998 . 9951 1. 0914 1. 2519 10 2519

Growth Ratio ------1.0988 1.1221 1.0887 1.0424

N 0 2 6 38 146 69

,f, ',' Growth stage I observed in the UH-l colony was not present in samples from the Ewa colony. 117

131

50 UH-1 ALATE NYMPHS

10 o Ig; g ~ ~ ~:! ~ t"'l -0 0" N In 0" N ~ ~ ~ 0: 0: 0: 0: ~;~...... ~ ": .. . PRONOiUM WIDTH (mm~ --- ,. 15 16 17 18 ANTENHAL SEGMENTS

50 EWA ALATE NYMPHS 40 >- ~30 LII ;:) 020 LII· .....ac­ 10 o

...... PRONOTUM WIDTH (mm.) 15 16 17 18 19 ANTENNAL SEGMENTS

Figure 3 rt. Frequency distribution of pronotum width with respective number of antenna1 segment groupings of Coptotermes formosanus Shiraki alate nymphs from the UH-l and Ewa colonies. 118

ratio was then m.ultiplied by the m.ean pronotum. width of the 16 antennal

segm.ented individuals to determ.ine the m.issing pronotum. width.

The growth of the alate nym.phs were subjected to a regression

analysis (Tables 15 and 16). The regression equations that were

calculated for the growth stages of the UH-l and Ewa alate nym.phs were

y = O. 0878X - 0.4677 and Y = O. 0938X - 0.5373, respectively. The

observed pronotal widths for both UH-l and Ewa alate nym.phs were in

close c,greem.ent with the calculated values (Figure 38). The Rlope of

both lines were very sim.ilar indicating a parallel rate of increase in

the pronota. Although the Ewa alate nym.phs again were larger than the

UH-l counterparts, the difference was not as large as in the workers

and soldiers.

The growth ratios between the m.ean pronotum. width of each

growth stage for the alate nym.phs from. the UH-l and Ewa colonies were about the sam.e indicating that the growth of the pronotal width increased in a geom.etric p:rogression (Tables 15 and 16).

Therefore, the workers of.9. form.osanus apparently all continue to m.olt even after they becom.e recognizable as workers. In fact, som.e workers apparently do not reach a fixed term.inal stage and m.ay m.olt until they die. There are no obvious changes with each m.olt except for increase in antennal segm.ents and a slight increase in size.

The alate nym.phs also go through a series of stages which also can be recorded by increases in antennal segm.ents and by an increase 119

1.3

1.2 ALATE NYMPHS

-~ E -E ~.1 :c..... -o 3: 1.0 ~ ::l t- O Z 00.9 0::: 0.. Z

0.7 I----I-__---I ...L.-__--L.__---J. .J----J 14 15 16 17 18 19 ANTENNAl SEGMENTS

Figure 38. Regression of th~ mean pronotum width on the number of antennal segments of Coptotermes .formosanus Shiraki alate nymphs from the UH-l and Ewa colonies. 120

in size. Surprisingly, there are also no obvious changes in the alate

nymph. The wing pads do not seem to increase greatly in length.

However, the alate nymph does go through one final molt to become an

adult so that there is a terminal molt.

The soldiers, Oil the other hand, apparently do not molt, after

they become soldiers. Therefore, the great majority of the soldiers

all had the same number of antennal segments. Those few soldiers

that had different numbers of antennal segments apparently

metamorphosed from workers earlier or later than the usual stage.

The soldier represents a terminal molt.

Seasonal fluctuations of the relative abundance of foraging worker

growth stages. The relat~ve abundance of UH-l worker growth stages fluctuated considerably throughout the sampling period as shown in

Figure 39. Stage 1 workers (11 antennal segments) were the least abundant of all the growth stages. These individuals first appeared in the traps in June 1971 and reached a peak of 7.5% of the total sample in

July 1971. The proportion fell to 2.5% in the following sample and dropped even lower to 1% or less in all subsequent samples.

Surprisingly, however, there always were a few stage 1 individuals in the sample.

The stage 2 individuals, which were also never very abundant, w~nt from a low of 0.5% in late April 1971 to a high of 28.5% in August

1971. Thereafter, the proportion gradually decreased to 5% in the last 121

Figure 39. Seasonal fluctuations of Coptotermes formosanus Shiraki worker growth stages from February 1971 through July 1972 at the UH-l colony. 65 UH -1 FORAGING WORKERS , '='"...~ .~ 60 o 0 .STAGE 1 •, -:.·..4/ ~ 1I·...•.....'a STAGE 2 55 ....////..~ STAGE 3 0---- STAGE 4 0;·0..... 0.\ A...... STAGE 5 I · \ ,/·.... 50 II ~ , ~A ~ ~ , V 45 ~ AA • ~O "/0.l ,~, 40 Y\'A',V"YeAV, . /' ':A V +, '\: t- 35 .. '\\ ~ 0 I I z...., '\ fl·", I '\ O-"O~II' ~~'~""A , ...... ,..#"P.~ l u30 .. .. A " "', \ V, '" A a=::...., \::' ~ /' ~O if! ,,~ I .. : ,.~ If.... • ~ 0 A_\.• A. 25 ~~~ ~ ~ i\ = ":.~~ ..... 0 A;·V·.... '." ., '.'. ..• ~ -·V AV .•" , 20 ..'A' •'.'10 •= ~~~ .,.. : ."---"'" 'V'. • • ••, '.~" A v 15 : :: • • -" .'.-. A • •"· '. ·0.~...•• · '"'" ,••, ., '" " •M...'.eo ••. .6. • G .''''''11I Ill".. • '. • ... M " ',. • 'O. ~ '_' _'" 10 M •"_ • ...... - • • • -. • «.' n •• •••.'::MA ~.•• .'.'.• A O.• : •••• II ", A ....iII...... '"Iiilljpo. • 0 II","" .4oee..I), I 5 'A" '. ..A •••A -.A•·B.. ~$1Il ...."1lI•, ",,'A""" 0 .....°__ "A...A. OIZlllllllt A Ill.~~ .;'" 0- ....0_"'_ o

t-' FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL N 1971 1972 N 123 sample.

The proportions of workers with 13 antenna1 segments (stage 3) fluctuated between 22% and 40% of the total in the initial 10 samples.

From. September 1971 the proportion of the stage 3 workers increased until they c01nprised 62% of the sample. This occurred in November

1971 and the proportion of stage 3 remained at this high level until

February 1972. In March 1972 the proportion dropped to 30.5% and continued to decline until it comprised only 13% of the workers in the last sample taken in late July 1972.

The stage 4 workers were generally the most abundant in most of the samples. They made up the major proportion of workers in the first 8 samples. There were a few samples, however, where they did not predominate. They went from a high of 46.5% of the sample in

March 1971 to a low of 22.5% in November 1971. This low proportion was maintained for the next 4 samples and then abruptly increased to

45.5% in March 1972. The proportion remained in the 50% region in subsequent samples.

Stage 5 workers made up 6. 5% of the first sample in February

1971. There was more or less a steady increase in the number of this stage to slightly over 30% of the samples in June 1971. There was, however, a definite decline in the numbers of stage 5 workers in May

1971. Twelve weeks later the proportion fell back to 5% and remained at approximately this level until October 1971. The proportion then 124 declined further to 2% and remained at this level until March. In March

1972 the stage 5 foraging worker proportion increased abruptly to 23.5% but again as in the previous year's sample, the relative abundance of stage 5 again dropped in May. The numbers of stage 5 workers again increased in the subsequent 2 samples. The reaSO:1 for the reduction in

May are not known.

If one follows the fluctuation of the various worker growth stages, from the UH-l colony, a definite pattern of development can be observed.

Five significant peaks were observed throughout the sampling period as shown in Figure 39. These peaks each corresponded to the 5 worker growth stages. The first peak corresponding to stage 1 workers occurred in July 1971. The second peak which occurred in August 1971 was related to stage 2. The third peak lasted from November 1971 to

February 1972 and represented stage 3 workers. The fourth peak representing stage 4 began in April 1972 and lasted 16 weeks to July

1972. The final peak for the stage 5 workers just began to increase in

July 1972.

This sequence of peaking appear to indicate that each growth stage developed into the successive growth stage. Surprisingly only

4 weeks elapsed between stage 1 and 2 whereas 16 weeks were required for stages 2 and 4. Unfortunately the peak duration for stage 5 was not determined due to the termination of the experiment. However the peak of stage 5 that occurred in April and May 1971 indicate that the peak may 125

last at least 8 weeks.

From this study it appears that 60 weeks are required for the

stage 1 workers to become stage 5 workers. Moreover if one notes the peaking of stage 5 workers in 1971 and 1972, a total of 60 weeks elapsed between the waning peak in June 1971 to the waxing peak in July 1972

(Figure 39).

According to Nakagima, et. al. (1964), Q. formosanus im.m.atures having 12 antennal segments were considered larvae which were individuals that were capable of differentiating into any of the 3 primary castes. The workers having 11 antennal segments appear to be those individuals that remain close to or within the nest since they onl'} appear in small numbers in the foraging area. Therefore, it appears that stage 2 workers normally begin the foraging activity. The 7. 5% peak of the stage 1 workers in July 1971 (Figure 39) may represent workers that may have wondered from the nest.

Throughout the entire sampling period, stage 4 appeared to be the most abundant (Figure 39). However when the worker proportion was at the highest level, the stage 3 were the most prominant in the samples. Stages 3 and 4, therefore, are responsible for most of the foraging activity and perhaps caring for the alate nymphs and soldiers in the foraging area.

Seasonal fluctuation of the relative abundance of the alate nymph growth stages. The fluctuations of the relative abundance of the 4 alate 126

nymph growth stages w·~re observed from the UH-1 colony. Growth

stage 2 mentioned earlier was not present in the salup1es observed.

Figure 40 graphically illustrates the actual number of alate nymph

growth stages present in the traps during the sampling period.

Although samplin~ began in February 1971, the alate nymphs did

not appear in the traps until late June when all 4 stages appeared in

the samples. There were 12 stage 1, 27 st<:1.ge 2, 3 stage 4 and a single

stage 5 individual in the June sample. The numbers of stage 1 nymphs

dropped to 1 in the July sample, and disappeared from all subsequent

samples except for a single nymph which was collected in January of the following year.

Stage 3 nylnphs, which were the most abundant in the initial

sample also quickly disappeared from the subsequent samples except for a single individual which was collected in the spring of the following year. Stage 4 increa.sed to 9 and 10 individuals in the following 2 consecutive 4 week sampling periods but fell to 3 and zero in the following 2 consecutive 4 week sampling. Thereafter stage 4 did not appear in the traps until July 1972 with 1 individual.

Stage 5, the penultimate stage observed, was also represented by 1 individual in late June 1971 but fell to zero in the following sample. However, in August 1971 the numbers of stage 5 nymphs increased and continued to increase with some fluctuations. In samples of January to March 1972 the numbL.'s of the stage 5 alate nymphs reachad a peak. In April the stage 5 alate nymphs abruptly fell to zero 70 UH -1 ALATE NYMPHS 65 A ••••• A STAGE 1 o f)O o __"II STAGE 3 o...... 0STAGE4 o 55 o -0STAGE 5 50 \o G:: 45 &II ED 40 :Ie ~ 35

30 II 25 ,Ii, 20 B , ,i I, 15 , .,

10 'A \ I; -. '0"·,,,0~ .... ;-0 •,.: ..~-, 0 .. 5 ,I N'-- ..' l ~A"""'O I: ""V?J VIII" ~~~~II.~.-,---"f'I~~ FEB MAR APR MAY JUH JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL 1971 . 1972

Figure 40. Seasonal fluctuations of Coptoterm.es fonnosanus Shiraki alate nym.ph growth stages ..... N from. February 1971 through July 1972. -J 128

and Jid not appear in subsequent samples.

According to Nutting (1969), the production and development of the

alates appear to be determined by pheromones and nutrients which are

exchanged by grooming and tropholaxis. There appears to be little

evidence that environmental factors influence the production of alates.

Sands (1965), however, noted that alate development in Trinervitermes

ebenerianus Sjostedt in Nigeria was accelerated by moisture resulting from the rainy season. In the present study, rainfall was directly

related to the presence of alate nymphs in the traps (Figures 28 and

29). The atmospheric temperature during the sampling period (Figure

29) at the UH-l colony appeared to have no effect on the presence of the alates.

Although pheromones and nutrients play an important role in the development of the alates, the alate developmental cycle appears to be synchronized to their swarming exodus. Swarming of~. formosanus usually occurs during April, May and June in Honolulu. Alate development from the UH-l colony appeared to be synchronized to this swarming period. Since the sampling period in the present study was not over a period of many years, it is not definitely known if the flights from an individual colony occur annually or biannually. However, the presence of one stage 1 individual in Jan uary 1972 seems to indicate that the alates are produced annually once the colony attains the stage where alates are produced. The results from the alate nymph 129 sampling indicate that one primary brood of alates is produced annually, although other species such as Microcerotermes parvus (Haviland)

(Kemp, 1955) and Reticulitermes lucifigus Rossi (Buchli, 1958) have 2 broods of alates per year.

Growth stage fluctuations presented in Figure 40 illustrate the progressive alate nymph development. However, there were peculiarities in that stage 2. alate nymphs were not present in the samples. Furthermore, it was odd that stage 1 and 3 peaked during the same sample period. One would normally expect the younger stage

1 to peak before stage 3. Other than these oddities, the development from stages 3 to 4 proceeded in a logical manner similar to that of the foraging workers. As the stage 3 population decreased, stage 4 increased and as stage 4 decreased stage 5 began to increase. This seems to indicate that one growth stage develops into the following growth stage.

The data indicate that the development of the alate nymphs from stages 1 through 5 require at least 40 weeks. The developmental period for stages 1 and 3 could not be determined since they were present at peak levels in their initial appearance. Stage 4 required approximately 12 weeks to develop while stage 5 matured after 20 weeks.

The presence of stages 1, 3 and 4 in relatively low numbers when compared to stage 5 suggests that these stages develop away from the foraging sites. It appears that after differentiating into the reproductive 130

line, stages 1, 3 and 4 remain in the ne st with the eggs and young

undifferentiated termites.

It is interesting to note the large number of the stage 5 alate

nymphs (Figure 40) found during December 1971 through March 1972.

There are several speculative rE:asons why this situation may have

occurred. One possible reason may be due to a herding response from the rest of the colony. This herding phenomenon was recorded in the literature for another Rhinoterminae species, Reticulitermes lucifigus

(Buchli, 1961). Gregarious tendencies among maturing alate nymphs and alates have also been recorded in the literature. Calaby and Gay

(1956) working with Australian Coptotermes species noted that the gregarious nature of the alates was due to short term movement to warmer galleries.

There are several observations to indicate that the stage 5 alate nymph is the penultimate stage. These are (1) after stage 5 peaked in

March 1972, they did not appear in the following samples, (2) there were no alate nymphs with 19 or more antennal segments in any samples, and

I (3) stage 5 individuals disappeared from the samples in April indicating that they may have molted and moved to some other part of the colony.

Present evidence indicates that the adults remain the colony for a while after molting.

The reason the dates did not reappear in the traps can be explained by the fact that the wood in the traps in the present study was replaced 131

every 4 weeks. This did not allow the workers sufficient time to

hollow-out the wood for the alates to congregate in. The dates are

usually concentrated in some aerial portion of the colony, "staging

areas, " from where they can emerge quickly in large numbers. Wood

in traps from other colonies that have not been replaced fo r several

months during the swarming period were literally hollowed out and

contained numerous alates. That the swarming of the alates involves

the emergence of a large number of alates in a short time has been

observed many times. This, of course, indicates that a large number

of alates are concentrated in the area close to the emergence or exit

hole. Figure 41 depicts the emergence of c. formosanus from a hollow

tile wall in May 1972 in Honolulu. The ala-tes literally "boiled-out!' of

a crack in the tile.

Since the a1ates in the present study probably became adults in

April 1972 (April is generally considered the beginning of the primary

swarming period for C. formosanus in Honolulu) and since swarming

occurred a month later in May 1972, a preflight period existed. During this period the alates prepared for the mass exodus. The hardening and

darkening of the cuticle requires 2 days in Reticulitermes lucifugus upon molting to the imago stage (Buchli, 1961). Moreover, Nutting

(1969) reported that the neuromuscular flight mechanism required 9 to

14 days to develop after ecdysis in Marginitermes hubbardi (Banks).

Although the UH-1 a1ates possibly remained in the colony for 132

Figure 41. a Hollow tile fence with CoptoterInes forInosanus Shiraki alates eInerging froIn a vertical cra,ck (arrow) during a May 1972 swarIn. b close-up of the vertical crack in the fence showing a1ates eInergillg. Note the abundance of soldiers guarding the open crack. 133

a 134

o.pproximately 8 weeks, this preflight period may have been extended if

the correct conditions for flight did not occur during that time. Table

17 lists several species of termites with their respective alate preflight

period.

Termite Movement Studies.

There has been several reports in the literature concerning the

. gallery systems of Coptotermes species. Ratcliffe and Greaves (1940)

found that galleries of C. lateus (Froggatt) radiated over an area of 1. 5

acres from a central mound. Gay (1946) unearthed galleries of C.

frenchi Hill from a primary nest to an infested house 70 feet away. In

Hawaii, Ehrhorne (1934) reported that C. formosanus constructed

foraging galleries 165 feet long and from 1 to 10 feet deep. More

recently, King and Spink (1969) reported C. formosanus galleries

which covered an area of about 1. 4 acres in Louisiana. These galleries

were connected to food sources more than 200 feet apart with gallery

depth ranging from 2 to 46 inches. Although there have been studies

conducted on the gallery system of C. form.osanus and how far they

radiate, there have been no studies to determine the rate of terr.llite

movement within these galleries. Therefore, studies were conducted to

determine the rate of movement and the distance C. formosanus foraging

workers radiated within their galleries in Hawaii.

In this study, the straight line distance (SLD) covered by the dye

treated foraging workers, between the release trap and the furthest TABLE XVII. --ALATE PREFLIGHT PERlOD FOR SEVERAL TERMITE SPECIES

Preflight Family Subfamily Species Period Reference

Mastotermitidae Mastotermes darwiniensis Month or Hill, 1942 Froggatt more

Kalotermitidae Kalotermitinae Paraneotermes simp1icico~nis Several weeks Nutting, 1966a (Banks)

Kalotermitidae Kalotermitinae Pterocermes occidentis (Walker) Several weeks Nutting, 1966b

Hodotermitidae Termopsinae Zootermopsis laticeps (Banks) Several weeks Nutting, 1965

Hodoteunit idae Hodotermitinae Anacanthotermes ochraceus Six months Clement, 1956 (Burmeister) or more

Rhinotermitidae Heterotermitinae Reticulitermes 1ucifigus Rossi 4-57 days Buchli, 1961

Rhinotermitidae Coptotermitinae Coptotermes formosanus 2 months or Shiraki more

Termitidae Amitermitinae Amitermes hastatus Haviland Month or Skaife, 1954 more

Termitidae Termitinae Macrotermes nata1ensis Month or Ruelle, 1964 (Haviland) more

Termitidae Nasutitermitina~ Trinervitermes spp. Several weeks Sands, 1965

Termitidae Nasutitermitinae Tenuirostritermes t~nuirotris Several weeks Weesner, 1953 (Desneux) ------_ ...... w U1 136 trap in which treated termites ""vere recovered, was nleasured. Although the SLD probably did not represent the actual distance covered by the termites since the galleries probably did not extend directly from trap to trap, it did represent a minimum distance traveled.

The initial release of fast green treated termites was made at the

Kauai colony in traps A-5 and A-9 (Figure 7A). Eight and one half hours after the release, all of the traps that were harvested contained marked termites. The maximum distance travelled by the termites was

160 feet. A second release of fluorescent dye treated termites was made at the Kauai colony in trap B-8. After a period of 6 hours one or more stained termites were found in all traps harboring termites except traps

C-5 and C-6 (Figure 7B). This study revealed that the treated termites were able to cover a SLD of approximately 140 feet in 6 hours. This was the maximum distance m.easurable by the experimental design.

Marked termites possibly would have been detected at a greater distance if more traps had been placed further out.

The irrigation ditch between traps 6 and 7 in the Kauai C. formosanus colony was no barrier to the foraging workers. Since the irrigation ditch was approximately 3 feet deep, the termites must have constructed foraging galleries at depths greater than 3 feet. Moreover the gallery appeared to be impervious to water since the ditch was regularly flooded for long periods.

The third release of fast green treated termites was made at the 137

t. UH-2 colony in trap B-2 (Figure 8B) to determine the rate of dispersal in a short period of time. Results from this study revealed that the foraging workers covered a SLD of approximately 6. 7 feet in a half an hour and 9. 5 feet in one hour. The rate of dispersal of the termites at the half hour sample was considerably greater than for the 1 hour sample. This probably was caused by the removal of the even numbered traps after the half hour exposure period. This not only disrupted the galleries but also resulted in the removal of the trap furthest from the release point. Had this trap been available at the 1 hou!' sample, it probably would have had marked termites indicating that the termites traveled 12. 3 feet in an hour.

The fourth release of fast green treated termites was conducted at the Waipio colony. Treated termites were released in trap A-6

(Figure 8A) and allowed to disperse for 2 hours. Results from this release indicated that the termites covered a SLD of 17. 3 feet in 2 hours. At the 2-hour sample, however, all of the traps at the fringes of the experimental area had marked termites. Again, this distance therefore, represents the maximum distance measureable under the conditions of the expe:dment.

The results from the mark release study to determine the SLD traveled by foraging workers of c. formosanus within their galleries are summarized in Table 18. The SLD covered by termites within a 1- hour period ranged from approximately 8. 6 to 23. 3 feet with a mean of 138

of 16. 4 feet.

Relationship of wood weight and the number of foraging workers. During

the course of the population studies of the UH-l colony, studies were

concurrently conducted on the attractiveness of Douglas fir wood of

various densities to the termites. Pieces of wood that were

approximately equal in size were ovendried, weighed, and exposed to

the termites for 4 weeks. The number of foraging workers were then

counted from each trap. The results of this study are sumlnarized in

Table 19 and Figure 42.

The wood was grouped in 100 gm increments into 5 classes, 850,

960, 1,050 and 1,240 gms and the mean number of workers extracted

from each of the classes were 13,400, 10, 700, 9,000, 9,200 and 7,800,

respectively. These data were subjected to a regression analysis from

which Y = 24, 183. 5 - 13. 5X was calculated. A negative regression was

obtained (Figure 42) indicating that there were more workers in the lighter wood. The termites exhibited a definite preference for the lighter wood even though the wood was from the same species of tree.

The density of the wood depends in part on the proportion of spring to summer wood. Spring wood generally has large cells and proportionately, a small amount of wall substance per unit volume.

Summer wood, on the other hand, has small cells and a large amount of wall substance per unit volume (Esau, 1965). Figures 43a illustrates wood that has a relatively large amount 01 spring wood and Figure 43b 139

TABLE XVIII. - -DISTANCES TRAVELED WI'l'HIN GALLERIES BY COPTOTERMES FORMOSANUS SHIRAKI FROM FOUR COLONIES

Furthest Number Time Before Distance Distance Traveled Colony Termites Recapture Traveled in One Released (Hrs. ) (feet) Hour (ft. )

Kauai 2,000 8.5 160.0 18.8

4,000 8. 5 80.0 9.4

Kauai 20,000 6.0 140.0 23. 3

UH-2 40,000 0.5 6.7 13.4

1.0 9.5 9. 5

Waipio 40,000 2.0 16.0 8.0 TABLE XIX. --MEAN NUMBER OF COPTOTERMES FORMOSANUS SHIRAKI ON DOUGLAS FIR WOOD OF DIFFERENT WEIGHTS FOUR WEEKS AFTER EXPOSURE

Mean Number Mean Wood Range Wood Range Foraging N of Foraging Weight (gm.s) Weight (gms) Worker Numbers Workers

6 850 832-874 13,400 5,500-27,800

12 960 912-995 10,700 2,000-30,000

13 1,050 1,008-1,098 9,COO 2,100-15,000

8 1,130 1,106-1,158 9,200 1,600-20,000

4 1,240 1,210-1,258 7,800 3,400-11,000

..... ~ o 141

14.000

o 13,000

12,000 .. en ex w ~ ex o 11,000 ~

C) Y =24,183.5019 -13.5277X ~ C) «10.000 ex o..... o

7,000

800 900 1,000 1,100 1,200 1,300 MEAN WOOD WEIGHT (gms.)

Figure 42. Regression of the mean number of Coptotermes formosanus Shiraki on wood of va rious weights 4 weeks after exposure. 142

shows wood that has more sununer wood. Although both spring and

summer wood contain cellulose, the latter has a higher concentration

of lignin which in part accounts for the high density of summer wood.

Studies conducted by Wolcott (1946, 1947) indicated that lignin

acted as a natural repellent to m.any species of termites and that they

were unable to digest it. Although c. forrnosanus was not completely

deterred from attacking wood having more lignin, there seemed to be a

preference for the pieces having more spring wood (Figure 43b). The

termites fed primarily on the spring wood and avoided the surnrner

wood (Figure 43c).

Resistance of wood to termite attack, however, involves many factors other than the proportion of lignin in the wood. Smythe and

Carter (1970) reported that Reticulitermes flavipes (Kollar) survived better and ate more on ponderosa pine sapwood while survival and feeding was lower when they fed on heartwood. The toxicity of the heartwood may be attributed to extractives present in this region such as tannis, gums, resins, pigments, salts of organic acids and other materials (Robbins, et. al., 1964). Heartwood, however, is not always more resistant to termites than sapwood. Schulze-Dewitz (1965) reported that Scot pine and Douglas fir sapwood were more resistant to

R. lucifugus than heartwood. Moreover, even after the sapwood was stored for 8 years, it was more resista.nt than the heartwood was at the beginning of the tests. 143

, \

\ I \ \ \ '

I r / 'b

Figure 43. a Douglas fir wood showing the abundance of lignified (1) summer wood. b Douglas fir wood showing the thickness of spring wood. c Douglas fir wood showing damage by Coptotermes formosanus Shiraki workers. Note that the termites fed primarily on the spring wood leaving the summer wood relatively untouched. 144

Due to individual variation of trees to termites attack, Gay, et. aI., (1955) suggested that samples have to be obtained fron1. many trees of the same species to determine the termite feeding suitability. Becker

(1961) further stated that care must be taken to sample various portions of the tree trunk. Another aspect of termite feeding behavior that may influence wood suitability are the gut inhabiting protozoa. Seifert and

Becker (1965) mentioned that the kind and numbers of the microoganisms in the termite's alimentary tract primarily determines the enzyme content of the termite's gut which, in turn, determine what substance can bE. digested or detoxified.

Relationship of foraging work.er numbers and the amount of wood consumed. There has been some work reported in the literature concerning the influence 01 wood hardness and weight loss due to termite feeding (Behr, et. aI., 1972) and feeding responses to sound wood by subterranean termites (Smythe and Carter, 1969, 1970).

However these studies were conducted in the laboratory where a few termites are placed on a small piece of wood. There has been no studies conducted to dete rmine the amount of wood consumed by C. formosanus naturally occurring in the field. This particular study attempted to determine the relationship between the numbers of termite present and the amount of wood consumed in a naturally occurring colony in the field.

The data for this study were extracted from the populatlon data 145

for the UH-l colony. Douglas fir wood used in the traps was washed in tap water after the termites were removed and counted. The amount of

wood eaten was determined by drying the wood for a constant period of

2 weeks and weighing the sample. As expected, the amount of wood

consumed was directly correlated with the number of workers feeding on the wood (Table 22). A correlation analysis yielded a correlation coefficient of O. 936 (Figure 44). However, the amount eaten by an individual termite did not differ with differing numbers. Although there was some variation, the total amount eaten per termite did not differ significantly. 146

TABLE XX. --AMOUNT OF DOUGLAS FIR WOOD CONSUMED BY COPTOTERMES FORMOSA"N"US SHIRAKI AFTER FOUR WEEKS EXPOSURE

Number of Workers Amount Wood Amount Wood (x 103) Consumed (gms) Consumed N by Individual Mean Range Mean Range Termite (gms)

1 2.0 ------28.4 ------14.2

2 3.4 3.3-3.4 79.6 64.9-135.3 23.4

1 5.5 ------80. 1 ------14.6

7 6.2 6.0-6.5 103.5 67.1-135.3 16.7

4 7. 3 7.0-7.6 91. 1 72.6- 99.4 12.5

1 8. 8 ------168.0 ------19. 1

1 9. 3 ------105. 1 ------11. 3

3 10. 3 10.1-10.5 173. 1 146.6-204.7 16.8

3 11.3 11.0-11.5 206.6 131. 9-145. 7 18. 3

3 12.5 12.4-12.8 241.6 205.1-270.8 19. 3

2 14.5 14.4-14.5 239.8 214.4-265.3 16. 5

1 15.0 ------262.7 ------17.5

1 17.4 ------284.3 ------16.3

1 18. 9 ------233.0 ------12. 3

2 23.9 20.0-27.8 447.3 349.5-545.1 18.7

1 30.0 ------376.7 ------12.6 147

Figure 44. Correlation between the amount of Douglas flr wood consumed and the number of Coptotermes formosanus Shiraki foraging workers. 500 o

o

o r =0.936

r 6 8 10 12 14 160 18 20 .. 22 24 26 28 30 32 3 MEAN NUMBER OF FORAGING WORKERS (x10 )

...... ~ 00 149

SUMMARY AND CONCLUSIONS

The Formosan subterranean termite, Coptotermes formosanus

Shiraki, Hawaii I s most destructive stxuctural pest, was accidentally introduced on the island of Oahu in 1907 or earlier. Since then C. formosanus has spread to all the major Hawaiian islands.

Chemical control is the primary method 01 controlling C. formosanus in Hawaii although preventative structural construction is also utilized. A potential alternative control method is the use of microbial agents such as nematodes. This study, in part, dealt with the effects of an entomogenous nematode, the DD-136 strain of

Neoaplectana carpocapsae Weiser, in the Formosan subterranean termite.

A thorough search of the available literature revealed that there was a paucity of information regarding the ecology and field biology of

C. formosanus. Thel'efore studies w'ere also conducted to determine the seasonal abundance of the various castes of C. formosanus. In addition, the general biology of the Formosan subterranean termite field populations was studied which included the relative abundance and seasonal fluctuations 01 the worker and reproductive caste growth stages. Additl.Onal studies were undertaken to determine the rate of termite movement within their galleries; the relationship between termite numbers, wood weight and wood consumption.

Gross syn1.ptorns elicited by C. formosanus workers parasitized 150

by the DD-136 strain of N. carpocapsae were typified by lethargic and

sluggish movement. As the disease progressed, the legs became

paralized, and there was light movement of the legs or antennae.

These moribund termites were eventually killed and ended-up on their

backs, sides or upright. Surprisingly the healthy termites were able

to detect diseased individuals under laboratory conditions. Diseased

termites were immobolized by the healthy termites by chewing off their

legs and antennae and were placed in a pile and covered with masticated

filter paper.

Termites that were exposed to dauerlarvae and prepared for

sectioning at various time intervals revealed that increasing the

dauerlarvae exposure period did not lead to a suspected increase in

parasitization rate. Further studies indicated that the anus was the

primary mode of entry into the alimentary tract although the

dauerlarvae also entered the alimentary tract via the buccal cavity.

Dauerlal'vae were found in the midgut as early as 2 hours, but it

was until 17 hours after exposure that they were found in the midgut

with any regularity. The dauerlarvae were able to penetrate the alimentary tract into the hemocoel as early as 4 hours after exposure; however, approximately 24 hours were normally required. There was no evidence that the dauerlarvae were able to penetrate the alimentary tract into the hemocoel in the stomodeum, mesenteron and proctodeum.

The initial detection of the bacterium, Achromobacter lSI

nematophilus, in the termite hemocoel was 24 hours after exposure to

the nematodes. After 48 hours of exposure the bacteria in the

sectioned termites were found throughout the hemocoel invading muscle,

hemolymph, fat body and nervous tissues. The nematodes in turn

invaded the termite nervous tissue, fat bodies, salivary gland, muscle

tissue and the sternal gland.

The body weight of termite workers from several colonies

differed significantly. This weight difference was considered in the

susceptibility tests. It was furthe l' noted that termites anesthetized

with CO2 prior to dauerlarvae treated increased termite parasitization.

Moreover it was found that after a critical CO2 exposure period, the

recovery time was not significant with increased CO2 exposure.

Generally the smaller termites required a longer period to revive

after the critical CO exposure period than the larger termites. 2 There was no significant difference in the susceptibility of

termites from the UH-l and Ewa colonies exposed to various

concentrations of dauerlarvae. The LCSO for the smaller UH-I termites was slightly less than the larger Ewa termites, 2,666 and

3,472 dauerlarvae respectively.

The LTSO studies indicated that UH-l workers treated with a conceD.tration of 20,000 dauerlarvae per chamber had a significantly

shorter LTSO thCl,n those treated with 5, SOO nematodes, but there was no significant difference from those treated with 11,000. However the 152

LT50 differed significantly between concentrations of 22, 000 and 11, 000

dauerlarvae for the Ewa treated workers.

The abundance of the foraging workers from the UH-l and Ewa

colonies during the 18 months sampling period revealed that there was no defInite seasonal pattern. However, there were more termites trapped from the UH-l colony than the Ewa colony. Moreover there was greater fluctuation from sample to sample in the Ewa colony when compared to the UH-l colony. The greater worker sample size from the UH-l colony may be due to the colony being 1) more active, 2) much larger, 3) less available wood resulting in concentration in the traps, 4) close to the traps, or 5) any combination of the above.

The soldiers from the UH-l and Ewa colony remained relatively constant throughout the saInpling pe riod. The mean W /S ratio for the

UH-l colony was 26 with a range of 3 to 119 and 28 with a range of 5 to

86 for the Ewa colony.

There was no definite seasonal pattern in the abundance of the alate nymphs. Perhaps a longer sampling period would reveal a cyclic pattern. Moreover, it appears that there was no correlation between the weather conditIons and the abundance of alate nymphs. There were, however, alate nymphs collected during almost all the sampling periods from the Ewa colony while they occurred primarily from June to

March in the UH-l colony.

An atternpt was made to determine C. formosanus worker 153

instars; however this was difficult since each instar could not be

observed in the laboratory as they developed. Therefore, "growth

stages" for the foraging workers were determined based on the number

of antennal segments and the pronotum width. There was a close

relationship between pronotum width and the number of antennal

segments. Six and 7 antennal groupings were recorded from the UH-l

and Ewa colonies respectively.

Soldiers did not have any growth stages. It appears that the

soldier caste is derived from the worker caste at a certain growth

stage depending on the colony. Soldiers from the UH-l colony had

predominantly 14 antennal segments while the majority from the Ewa

colony had 15. Prior to the mature soldier stage, there is a presoldier

stage.

Growth stages were also determined among the alate nymphs from

both UH-l and Ewa colonies. The Ewa individuals were again larger

than the UH-1 individuals. Regression analysis indicated that although

the Ewa alate nymphs were larger than the UH-l alate nymphs, both

developed at a parallel rate of increase.

Seasonal fluctuation of the relative abundance of the foraging worker growth stages from the UH-l colony revealed definite pattern

of development from one stage to the successive stage. Sixty weeks were required for stage 1 workers to become stage 5, the largest worker stag~ in the SalYlples. A shnilar pattern of development was 154 observed in the seasonal fluctuation of the relative abundance of the

UH-l alate nymph growth stages. Forty weeks were required for the development of the smallest to the largest alate nymph growth stage.

Moreover the alate nymph development corresponded to the swarming of the alates during the year of the present study.

Studies were conducted to determine the rate of movement of the worker caste within their t:nderground galleries by the use of stained termites. This study revealed that termite workers traveled a distance of approximately 8.6 to 23. 3 feet with a mean of 16.4 feet per hour.

The attractiveness of Douglas fir wood of various densities to the termite workers was also studied. Results indicated that the workers were attracted to the lighter, le3s dense wood. This was apparently due to the high lignin content in the dense summer wood which is a natural repelent to many species of termites. In addition, as expected, there was a direct c0rrelation with termite numbers and the amount of wood consumed. 155

LITERATURE CITED

Abbott, W. S. 1925. A method of computing the effectiveness of an insecticide. J. Econ. Entomol., 18: 265-267.

Andrews, E. A. 1911. Observations on te:rmites in Ja.maica.; J; Anim. Behav., 1 (3): 193-228.

Beal, R. H. 1967. Formosan invader. Pest Control, 35(2): 13-17.

Beal, R. H. and V. K. Smith. 1971. Relative susceptibilities of Coptotermes formosanus, Reticulitermes flavipes, and R. virginicus to soil insecticides. J. Econ. Entomol., 64(2): 472-475.

Becker, G. 1961. Beitrage zur Prufung und Beurteilung der naturlichen Dauerhaftigkeit von Holz gegen Termiten. Holz als Roh-Werkstoff, 19: 278-290.

Behr, E. A., C. T. Behr and L. F. Wilson. 1972. Influence of wood hardness on feeding by the Eastern subterranean termite, Reticulitermes flavipes (Isoptera: Rhinotermitidae). Ann. Entomol. Soc. Amer., 65(2): 457-460.

Bess, H. A. 1966. Notes and exhibitions. Proc. Haw. Entomol. Soc., 19(2): 136.

19'70. Termites of Hawaii and the oceanic islands. In "Biology of termites" (K. Krishna and F. M. Weesner, eds.), Vol. 2: 449-476. Academic Press, New York.

Bodot, P. 1967. Cycle saisonniers d'activite collective des termites des savanes de Basse Cote d'Ivoire. Insectes Sociaux, 14: 359­ 388.

Bouillon, A. 1970. Termites of the Ethiopian region. r.rt "Biology of termites" (K. Krishna and F. M. Weesner, eds.), Vol. 2: 153­ 280. Academic Press, New York.

Bronskill, Joan F. 1962. Encapsulation of rhabditoid nematodes in mosquitoes. Can. J. Zoo1., 40: 1269-1275.

Buchli H. 1956. Le cycle de developement des castes chez Reticulitermes. Insectes Sociaux, 3: 395 -401. 156

1958. L'origine des castes et les potentialites entogeniques des termites europeans du genre Reticulitermes. Ann. Sci. Nat. Zool. BioI. Animale, 20: 263-429.

1961. Les relations entre la colonie maternelle et les jeunes imagos ailes de Reticulitermes lucifugus. Vie et Milieu, 12(4): 628-632.

Calaby, J. H. and F. J. Gay. 1956. The distribution and biology of the genus .Coptotermes (Isoptera) in Western Australia. Aust. J. Zool., 4(1): 19-39.

Clement, G. 1956. Observations sur L'Anacanthotermes ochraceus Burm. (Isoptera). Bull. Soc. Ent. France, 61(5-6): 98-103.

Creighton, C. S., F. P. Cuthbert, Jr., and W. J. Reid, Jr. 1968. Susceptibility of certain coleopterous larvae to the DD-136 nematode. J. Invert. Pathol., 10(2): 368-373.

Daum, R. J. 1970. Revision of two computer programs for probit analysis. BulL Entomol. Soc. Amer., 16(1): 10-15.

Davis, C. J. 1964. Notes and exhibitions. Proc. Haw. Entomol. Soc., 18(3): 336.

Dayar, H. G. 1890. The number of molts of lepidopterous larvae. Psyche, 5: 420-422.

Dutky, S. R. 1959. b.1SCct rnicrobiology. Advan. Appl. Microbiol., 1: 1'75-200.

Dutky, S. R. and W. S. Hough. 1955. Notes on a parasitic nematode from codling moth larvae, Carpocapsa pomonella. (Lepidoptera, Olethreutidae). Proc. Entomol. Soc. Wash., 57: 244.

Dutky, S. R., J. V. Thompson and G. Cantwell. 1962. A technique for rearing the greater wax moth. Proc. Entomol. Soc. Wash., 64: 56-58.

1964. A technique for the rnas s propagation of the DD-136 nematode. J. Insect Pathol., 6: 417 -422.

Ehrhorn, E. M. 1919. Notes and exhibitions. Proc. Haw. Entomol. Soc., 4(1): 83. 157

1934. The termites of Hawaii, their economic significance and control and the distribution of termites by commerce. In" Termites and termite control" (Kcfoid, et. al., eds.). 2nd ed., : 321-333. Univ. of Calif. Press, Berkeley.

Esau, K. 1965. Plant Anatomy, 2nd ed. John Wiley & Sons, Inc. New York. 767pp.

Finney, D. J. 1952. "Probit analysis, a statistical treatment of the sigmoid response curve," 2nd ed., Cambridge Univ. Press, Cambridge.

Fuchs, A. G. 1915. Die Naturgeschichte der Nematoden und einiger anderer Parasiten; des ~ tropgraphsu L. ; des Hylobius abietis L. Zool. Jahrb., 38: 209-222.

Fullaway, D. T. 1925. Notes and exhibitions. Proc. Haw. Entomol. Soc. , 6 (1): 19.

Fullaway, D. T. 1929. Notes and exhibitions. Proc. Haw. Entomol. Soc. , 7 (2): 205.

1932. Notes and exhibinons. Proc. Haw. Entomol. Soc. , 8 (1): 7.

Fuller, C. 1915. Observations on some South African termites. Ann. Natal. Mus., 3(2): 329-505.

Gay, F. J. 1946. A case of house infestation by a tree dwelling colony of Coptotermes frenchi Hill. J. Counc. Sci. Ind. Res. Aust., 19: 330-334.

Gay, F. J.; T. Greaves, F. G. Holdaway and A. H. Wetherly. 1955. Standard laboratory colonies of termites for evaluating the resistance of timber, timber preservatives, and other materials to termite attack. Aust. Comm. Sci. Ind. Reso Organ. Div. Forest Prod. Bull., 277: 1-60.

Grasse, P. P. and C. Noirot. 1948. La "climatisation" de la termitiere par ses inhabitants et Ie transport de I' eau. C. r. hebd. Seance Acad. Sci. Paris, 227: 869-871

Gray, P. 1958. IIHandbook of basic microtechnique, II 3rd ed. McGraw-Hill Book Co., San Francisco. 158

Hill, G. F. 1942. Termites (Isoptera) from the Australian region. Council Soc. Ind. Res., Melbourne, Aust.

Jackson, G. J. 1965. Differentiation of three species of Neoa.plectana (Nematoda: Rhabditida), grown axenically. Parasitol., 55(3): 57i-578.

Jackson, G. J. and G. E. Moore. 1969. Infectivity of nematodes p Neoaplectana species, for the larvae of the weevil, Hy10bius pales, after rearing in species isolation. J. Invert. Pathol., 14(2): 194-198.

Jaques, R. P. 1967. Mortality of five apple insects induced by the nematode DD-136. J. E con. Entomol., 60: 741-743.

Kemp, P. B. 1955. The termites of north-eastern Tanganyika: Their distribution and biology. Bull. Entomol. Res., 46: 113-135.

King, E. G., Jr. and W. T. Spink. 1969. Foraging galleries of the Formosan subterranean termite, Coptotermes formosanus, in Louisiau3.. Ann. Entor'J.ol. Soc. Am., 62(3): 536-542.

Light, S. F. 1926. In "Termites or white ants in Hawaii," by D. T. Fullaway, The Haw. Forester and Agric., 23(3): 69.

Merrill, J. H. and A. L. Ford. 1916. Life history and habits of two new nematodes parasitic on insects. J. Agric. Res., USDA, 6(3): 115-127.

Miller, E. M. 1969. Caste differentiation in the lower termites. In "Biology of termites" (K. Krishna and F. M. Weesner, eds.), Vol. 1: 283-310. Academic Press, New York.

Moore, G. E. 1965. The bionOluics of an insect-parasitic nematode. J. Kanas Entomol. Soc., 38: 101-105.

Nakajima, S. 1959. Studies on the determination of the termite nest. Report of technical institute of Japanese state ownership railway cor-po (In Japanese).

1960. Studies on the synthetical judgment of the vitality of the Formosan termite nest. Report of technical institute of Japanese state ownership railway corp. (In Japanese). 159

Nakajima, S., K. Shimizu and Y. Nakajima. 1964. Analytical studies on the vitality of colonies of the Formosan termite, Coptotermes form.osanus Shiraki. II. Seasonal fluctuatlons on the external characters of the workers, the ratio of caste-member and carbon dioxide in the nest of a colony. Bull. Fac. Agric., Univ. Miyazaki, 9(2): 222-227.

Nash, R. F. and R. C. Fox. 1969. Field control of the Nantucket pine tip moth by the nematode DD-136. J. Econ. Entomol., 62(3): 660-663.

Nutting, W. L. 1965. Observations on the nesting site and biology of Arizona dampwood termite Zootermopsis laticeps (Banks) (Hodotermitidae). Psyche, 72(1): 113-125.

1966a. Colonizing flights and associated activities of termites. 1. The desert damp-wood termite, Paraneotermes simplicicornis (Kalotermitidae). Psyche, 73: 131-149.

1966b. Distribution and biology of the primitive dry-wood termite Pterotermes occidentis (Walker). Psyche, 73: 165-179.

1969. Flight and colony foundation. In 11 Biology of termites" (K. Krishna and F. M. Weesner, eds.), Vol. 1: 233-282. Academic Press, New York.

Oshima, M. 1920. A new species of immigrant termite from the Hawaiian Islands. Proc. Haw. Entornol. Soc., 4(2): 261-264.

Pemberton, C. E. 1928. Nematodes associated with termites in Hawali, Borneo and Celebes. Prac. Haw. EntomoL Soc., 7(1): 148-150.

Poinar, G. 0., Jr. 1966. The presence of Achromobactor nematophilus Poinar and Thomas ill the infective stages of a Neoaplectana Spa (Steinernematidae, Nematoda). Nematologica, 12: 105-108.

], 967. Description and taxonomic position of the DD-136 nematode (Steinernematidae, Rhabditoidea) and its relationship to Neoaplectana carpocapsae Weiser. Proc. Helm. Soc. Wash., 34: 199-209. 160

Poinar, G. 0., Jr. and G. M. Thomas. 1965. A new bacterium, Achromobacter nematophilus sp. nov. (Achromobacteriaceae: Eubacteriales) associated with a nematode. Int. Bull. Bact. Nomen. Taxon., 15: 249-252.

1966. Significance of Achromobacter nematophilus Poinar and Thomas (Achromobac­ teraceae: Eubacteriales) in the development of the nematode DD-136 (Neoaplectana sp., Steinernematidae). Parasitol., 56: 385-390.

Ratcliffe, F. N. and T. Greaves. 1940. The subterranean foraging galleries of Coptotermes lacteus (Froggatt). J. Coun. Sci. Ind. Res. Aust., 13: 150-161.

Reed, E. M. and P. B. Carne. 1967. The susceptibility of a nematode (DD-136) for the control of some pasture insects. J. Invert. Pathol., 9(2): 196-204.

Robbins, W. W., T. E. Weier and C. R. Stocking. 1964. Botany an introduction to plant science, 3rd ed. John Wiley & Sons, Inc., New York.

Ruelle, J. E. 1964. L'essaimage de Macrotermes natalensiR Haviland dans 1a region de Leopoldville (Isoptera, Macrotermitinae). In "Etudes sur les termites Africains'l (A. Bouillon, ed.): 231-245. Masson, Paris.

Sands, VV. A. 1965. Mound population movements and fluctuations in Trinervitermes ebenerianus Sjostedt (Isoptera, Termitidae, Nasutitermitinae). Insectes Soiaux, 12(1): 49-58.

1969. The association of termites and fungi. In IIBiology of termites" (K. Krishna and F. M. Weesner, eds.), Vol. 1: 495-524. Academci Press, New York.

Schmiege, D. C. 1963. The feasibility of using a neoapleetanid nematode for contI'ol of some forest pests. J. Econ. EntomoJ.., 56: 527 -431.

1964. A note on the taxonomy of an undescribed insect parasitic nematode in the genus Neoapleetana. Parasito!., 54(2): 233-236. 161

Schultze-Dewitz, G. 1965. Abhangigkeit des Versuchsergebnisses bei der Termitenprufung von der Prufdauer. Beitrage zur Ent., Bd. 15, Heft 1/2: 127-134.

Scott, C. M., Jr. 1966. Formosan subterranean termite found in Louisiana and Texas poses a threat to Georgia and other southern states. J. Ga. Entomol. Soc., 3: 31.

Seifert, K. and G. Becker. 1965. Der chemische Abbau von Laub-und Nadel holzarten durch verschiedene Termiten. Ho1zforschung, 19: 105-111.

Sen-Sarma, P. K. 1962. Some observations on swarming in nature and colony foundation under laboratory conditions in Odontotermes assmuthi (Holmgren) at Dehra Dun. (Isoptera: Termitidae). Beitrage zur Entomologie, 12(3-4): 292-297.

Shimizu, K. 1962. Analytical studies on the vitality of colonies of the Formosan termite, Coptotermes formosanus Shiraki. 1. Analysis of the strength of vitality. Bull. Fac. Agric. Univ. Miyazaki, 8(1): 106-110.

Shiraki, T. 1909. On the Japanese termites. Trans. Ent. Soc. Japan, 2: 229-242.

Skaife, S. H. 1954. Caste differentiation among termites. Trans. Roy Soc. S. Africa, 34: 345 - 35 3.

Smith, S. G. 1943. Techniques for the study of insect chromosomes. Can. Entomol., 75: 21-34.

Smythe, R. V. and F. L. Carter. 1969. Feeding responses to sound wood by the eastern subterranean termite (Reticulitermes flavipes (Isoptera: Rhinotermitidae). Ann. Entomol. Soc. Amer., 62: 335-337.

1970. Feeding responses to sound wood by Coptotermes formosanus, Reticulitermes flavipes and R. virginicus (Isoptera: RhinotermitidaG). Ann. Entomol. Soc. Amer., 63(3): 841-847.

Snedecor, G. W. 1966. Statistical methods. Iowa State Univ. Press, Ames, Iowa.

Snyder, Thomas E. 1924. Whence the termites of Hawaii? Proc. Haw. Entomol. Soc., 5 (3): 381-384. 162

1949. Catalogue of the termites (Isoptera) of the world. Smithsonian Misc. Call., 112: 1-490.

Swenson, K. C. 1966. Infection of the garden symphy1an, Scutigerella immaculata, with the DD-136 nematode. J. Invert. Pathol., 8(1): 133-134.

Swezey, O. H. 1914. Notes and exhibitions. Proc. Haw. Entomol. Soc., 3(1): 27.

1931. Miscellaneous notes. Proc. Haw. Entomol. Soc., 7(3): 483-484.

Tamashiro, M., J. K. Fujii and P. Lai. 197.3. A sL"fllple method to observe, trap, and prepare large numbers of subterranean termites for laboratory and field experiments. Env. Entomol., 2(4): 721-722.

Tanada, Y. and C. Reiner. 1962. The use of pathogens in the control of the corn earworm, Heliothis zea (Boddie). J. Insect Pathol. , 4: 139-154.

Taylor, R. L. 1931. On IIDyar t s rule" and its application to sawfly larvae. Ann. Entomol. Soc. Amer., 24: 451-466.

Weesner, F. M. 1953. The biology of Tenuirostritermes tenuirostris (Desneux) with emphasis on caste development. Univ. of Calif. (Berk.) Publ. Zoo1., 57(4): 251-302.

Weiser, J. 1955. Neoap1ectana carpocapsae n. sp. (Anguillulata, Steinernematinae), novy cizopasnik housenek obalece jab1ecneho, Carpocapsa pomonella L. Vestnik Ceskoslovenske Zoologicke spolecnosti, 19(1): 44-52.

1962. Uber die Benutzung der Nematoden zur biologische Schadlingsbekampfung. Internat. Congo of Entomol. (lIth) Vienna. Proc. Vol. 2: 880-882.

1964. Protozoonosen und Nematodenbefall bei Insekten. Colloq. Intern. Pathol. Insectes, Paris, 1962. Entomophaga, Memoire, 2: 67 -75.

1966. Nemoci Hmyzu. Academia Prague, 554pp. 163

Welch, H. E. and L. J. Briand. 1961a. Field experim.ents on the use of a nem.atode for the control of vegetable crop insects. Proc. Entom.o1. Soc. Ontario, 91: 197-202.

1961b. Tests of the nem.atode DD­ 136 and an associated bacterium. for control of the Colorado potato beetle, Leptinotarsa decem.lineata (Say). Can. Entom.ol., 43: 759-763.

Welch, H. E. and J. F. Bronskill. 1962. Parasitism. of m.osquito larvae by the nem.atode, DD-136 (Nem.atoda: Neoap1ectanidae). Can. J. Zoo1., 40: 1263-1268.

Wolcott, G. N. 1946. Factors in the natural resistance of woods to term.ite attack. Carib. Forester, 7(2): 121-134.

1947. The perm.anence of term.ite repellents. J. Econ. Entom.ol., 40(1): 124-129.