Vogtia xr.Sillor . ..7 "^?~P.ODUCFD

PYPALID (LEPIDOPTERA) FOR THE CONTROL OF ALLIGATORWEED

IN T\IF. UNITED STATES

By
J-OHN LEE BR0V7N

A DISSERTATION PRESENTED TO THE
GRADUATE COUNCIL OF TPIE UNI\^RSITY OF FLORIDA
IN ?ARTIi\L FULFILLIIENT OF THE PvEQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPY

UT'IVEPSITY

1973

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS '. v

LIST OF TABLES vi

LIST OF ILLUSTRATIONS viii

ABSTRACT ix

INTRODUCTION. 1

LITERATURE REVIEW 3

Alligatorweed 3

Vogtia rnalloi 4

Biology 5

METHODS AND MATERIALS 7

Biology and Behavior of Vogtia raalloi 7

Fecundity and Fertility 7

Oviposition 7

Egg and Larval Mortality 8

Greenhouse Colony 9

Field Releases 10

19 71 10

19 72 15

RESULTS AND DISCUSSION 17

Biology and Behavior of Vogtia rnalloi 17

Fecundity and Fertility 17

ii

Page

Oviposition 19

Egg and Larval Mortality 19

Behavior , ^ . . . 20

Field Releases 22

19 71 22

1972 , . 29

ALLIGATORWEED PRODUCTIVITY 31

Introduction 31

Methods and Materials 31

Greenhouse Studies 31

Field Studies 33

Results and Discussion •■ 34

Greenhouse Studies 34

Field Studies. 35

INTERRELATIONSHIPS BETWEEN ALLIGATORVJEED AND TWO

INSECTS- Vogtia and Agasicles 38

Introduction 38

Methods and Materials 40

Results and Discussion 42

Insect-host plant relationship 42

Nutrient levels and alligatorweed char-
acteristics 48

Comparison of Lakes and Streams 51

Analysis of variance of Plant Char-
acteristics 53

SUMMARY 64

111

Page
LITERATURE CITED 68

BIOGRAPHICAL SKETCH 71

IV

ACKNOWLEDGMENTS

I wish to express my appreciation to the numerous
individuals who have assisted in this study.

I am grateful for the technical assistance given by
Ted Center, Joe Balciiinas, and Debbie Beasley both in
laboratory and field studies.

I wish to thank Neal Spencer and the U. S. Department
of Agriculture for providing space and the necessary equip-
ment to carry out this program.

I am also indebted to my committee chairman, Dr. Dale
Habeck, for his assistance in the preparation and submission
of this dissertation.

LIST OF TABLES

Table . Page

1. Egg placement by V. nalloi female moths caged

over alligatorweed „ 19

2. Nutrient levels in water tables used for
productivity studies 34

3. Net productivity of alligatorv'/eed in green-
house tests , 34

4. Field studies of alligatorweed productivity 37

5. Variables used in canonical correlations with

mean and standard deviations 43

6. Canonical variates in analysis 1 44

7. Canonical coefficients for individual varibles,
analysis I 44

8. Canonical variates in analysis II 46

9. Canonical coefficients for individual

variables, analysis II 47

10. Variables used in canonical correlations III.... 48

11. Canonical variates in analysis III 49

12. Canonical coefficients for individual

variables, analysis III 50

13. Computer reclassification of lakes and streams.. 52

14. ANOV for plant height 54

15. ANOV for internode length 55

VI

Table , Page

16. ANOV for internode diameter 57

17. ANOV for stem density 59

18. ANOV for leaf area missing..,. 61

19. Overall effect of variables on alligatorweed 63

VI 1

LIST OF ILLUSTRATIONS

Figure Page

1. Vogtia release sites, 1971 and 1972 11

2. Log number of Vogtia eggs laid plotted

against time ,. 18

^* Vogtia population build-up and its effect on

alligatorweed, Lake Alice, 1971 „ 23

^' Vogtia population buiJd-up and its effect on

alligatorweed, Lake Alice, 1972 26

Vlll

Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial
Fulfillment of the Requirements for the Degree of Doctor of Philosopl:

VQGTIA MALLOI, A NEWLY INTRODUCED PYRALID (LEPIDOPTERA)
FOR ""the control OF ALLIGATORWEED IN THE UNITED STATES

By

John Lee Brown
December 197 3

Chairman: Dr. Dale Habeck
Major Department: Entomology

Vogtia malloi pastrana was introduced into the United
States in the spring of 1971 as a biological control agent
of alligatorweed, Alternanthera philoxeroides (mart.)
Griseb. Vogtia populations were established and survived the
winter as far north as Columbia, S. C. , and as far south as
Fort Lauderdale, Florida.

Vogtia populations dispersed randomly from release
sites, and at one location in 1971 reduced the number of
aerial stems/ft.^ from 52.5 to 4.0 in four generations.

Studies reported herein include: a canonical analysis
which describes the insect-host plant relationship between

IX

two insects and alligatorweed, the relationship between
nutrient levels and alligatorweed grov^th, and a comparison
of alligatorweed growing in lakes and in streams; a
multivariate regression analysis which measures the signi-
ficance of each of 12 measured variables in influencing the
growth and spread of alligatorweed. Also included are
measurements of alligatorweed productivity in greenhouse
studies and in field plot studies during the spring and

i.

summer growth periods..

INTRODUCTION

Alligatorweed, Alternanthera philoxeroides (Mart.)
Griseb., a vascular aquatic plant in the family Amaranthaceae,
was probably introduced into the United States in the 1890 's
(Weldon 19 60) . In recent years alligatorweed has become one
of the most troublesome weeds in fresh water ecosystems in
the southeast, and by 1970 more than 66,000 acres were
infested (Gangstad and Guscio 1970) .

Alligatorweed may occur as a terrestrial plant in low-
lands, as a semi-aquatic plant along streams or lake shores,
or as an emersed aquatic plant. Floating mats which may
extend 50 feet or more out over the surface of the v/ater are
formed of interwoven stems rooted near the shore. The stems
are marked by nodes at intervals of about 2-6 in. each of
which is capable of producing a new plant.

Most aquatic plants cannot compete with alligatorweed
where it has been introduced and are quickly crowded out
(Weldon 1960) .

Alligatorweed is not as susceptible to herbicides as
some of the other aquatic plants, requiring application at
higher rates and shorter intervals.

The lack of a suitable chemical control and the alien
status of alligatorweed led to consideration of biological
control as a possible solution. In 1959 the U. S. Army

Corps of Engineers provided funds to the Agricultural
Research Service, USDA, to explore the possibilities of
importing natural enemies of alligatortveed capable of
suppressing its grov;th and spread (Vogt 1960, Zeiger 19 67,
Maddox et al. 1971). As a result of these studies, three
insects have been introduced into the southeastern region
of the United States. A Chrysomelid beetle, Agasicles
hygrophila Selman and Vogt, introduced in 19 64 has been
very effective in controlling alligatorweed in certain
areas (Zeiger 1967, Maddox et al. 1971). Amynothrips
andersoni O'Neill, a thrips, was introduced in 19 67, and
although still present in some reJ.ease areas, it has not
caused significant damage to alligatorweed (-Maddox et al.
1971) .

The release and establishment of the third insect,
Vogtia malloi Pastrana, the so-called alligatorweed stem
borer (Lepidoptera: Pyralidae: Fhycitinae), is the
subject of this dissertation.

LITERATURE REVIEW
Alligatorweed

Alligatorweed, also known as "Lagunilla" in South
America, was probably introduced into the United States
in the ballasts of sailing ships late in the nineteenth
century. It was first recorded in Florida in 1894 (Weldon
19 60) . Mohr (19 01) discovered the plant completely filling
a creek near Mobile, Alabama in September 189 7. He states
that the source of the introduction was from the West
Indies and Brazil.

According to Weldon (1960) , the species was first
described in 1826 as Bucholzia philoxeriodes Mart, and in
1897 it acquired its present name, Alternanthera
philoxeroides (Mart.) Griseb.

Vogt (19 60) studied over 1500 herbarium specimens of
Alternanthera and related genera including approximately
80 species of Alternanthera .

Penfound (1940) described the anatomy and the life
history of alligatorTA?eed along with its aquatic growth
habits in Alabama while Arceneaux and Herbert (1943)
reported on alligatorweed growth in the cultivated fields
of Louisiana.

Weldqn (1960) extensively reviewed the literature
concerning past chemical and mechanical controls for
alligatorweed.

Vogtia malloi

The moth Vogtia malloi (Pyralidae: Phycitinae) was
named in 1961 when Jose A. Pastrana described it as a new
genus and species. Vogtia may be recognized by the
following characteristics: large labial palpi, three
times the diameter of the eye, pointing forward with
loose, thick scales and an obtuse third joint; no maxillary
palpi are present; ocelli are present; the front wing has
smooth scales, a slightly curved edge, and ten veins. The
wingspan is 20-22 mm and the wings are straw-colored,
dashed with brown scales on the edge and tip of the wing.

This insect was discovered by George Vogt in his
surveys in South America for natural enemies of alligator-
weed for possible introduction into the United States.
Vogtia was one of the four insects considered by Vogt (1961)
as a major suppressant of alligatorweed in South America.
He found that Vogtia was almost coextensive with alligator-
weed, both geographically and ecologically occurring as far
south (La Plata, Argentia) and as far north (Georgetown,
British Guiana) as his survey extended.

Vogt (1961) reared Vogtia from Alternanthera
philoxeroides and from Alternanthera hassleriana, the plant
he considered most closely related to alligatorweed.

Field .observations in Argentina and starvation tests
in the laboratory (Maddox and Hennessey 1970) indicated
that Vogtia could not complete its life cycle on plants
outside the genus Alternanthera. Plants were selected for
feeding tests on the basis of their . taxonomic relationship
to or ecological association with alligatorweed, or their
economic importance. In field observations (1962-1967) 51
native and introduced species of plants were examined for
feeding damage by Vogtia and 30 plant species were examined
in starvation tests for larval feeding and survival (Maddox
and Hennessey 19 70) .

Biology

Egg Eggs are deposited singly on the upper portion

of the aerial stems of alligatorweed. The average length
and width of Vogtia eggs was 0.69 mm and 0.37 mm respectively
(Maddox 1970) . When the egg was first deposited, it was opaque
white; as it aged it gradually changed from white to light
yellow to amber. The head capsule was visible through the
chorion usually after the second day. Maddox (1970)
reported the average incubation period was 3.6 days at
an average of 32.9°C and 41.4 % RH.

Larva According to Maddox (1970) , head capsule

measurements indicated that there were five instars. The
first stage larva was approximately 2.25 mm long with a dark
brown to black head capsule about 0.27 mm in diameter
(Maddox 1970) . In later stages the head capsule was brown
to light tan, and tan wavy longitudinal lines appear

on the dors.um and pleura of the thorax and abdomen. The
mature fifth stage larva averaged 13.7 mm long with an
average head capsule width of 1.25 mm (Maddox 1970). Larval
development required 24 days at 23°C and 64 % RH (Maddox
1970).

Prepupa — As with most holmetabolous insects, a rela-
tively inactive larval period immediately preceeded pupation,
The larva becomes shorter, thicker, and may appear greenish
in color at the time the cocoon was constructed.

Pupa — The pupa was amber or light tan when first formed,
It gradually turned a dark brown or almost black just prior
to emergence. Pupal length varied from 9 to 10 mm and the
width from. 1.5 to 2.0 mm (Maddox 1970). The- pupal period
averaged 9.5 days at 23.3°C and 49.7 % RH (Maddox 1970).

Adult--The adult is approximately 13-14 mm long v/ith
light tan scaled (both sexes) . Variation occurs in both
the size (females are usually larger) and in the definition
of the color patterns of the adult moth. Females often
have scales with a reddish tan hue forming indefinite
patterns, and the males often have black or gray scales
forming definite spots along the front margin and tip of
the wings .

METHODS AND MATERIALS
Biology and Behavior of Vogtia malloi

Fecundity and Fertility

In laboratory studies newly emerged adults were con-
fined in pint and gallon ice cream cartons covered with a
fine mesh net. One female and two males were placed in
each container and fed a 1:10 honey-v/ater solution. There
were 10 replications using the pint containers and 25 repli-
cations with the gallon containers. Eggs were deposited on
the net covers which were changed daily and the eggs counted.
The temperature and RH were maintained at 27 °C and 72 %.

To determine hatchability , as an indicator of fertility,
447 eggs collected from the moths in the gallon containers
and 1079 eggs collected from the pint containers were held
at 31 °C and 60 % RH for 6 days. After 6 days the unhatched
eggs were removed and counted.

Oviposition

To duplicate natural conditions in the placement of
Vogtia eggs for field releases it was necessary to deter-
mine the ovipositional preference of the female moth.

One male and two female adults were confined in a 15 X
15 X 18 in. screen cage which contained a bouquet of 10 to
15 alligatorweed stems in water.

Egg and Lar-val Mortality

Studies v;ere conducted at Lake Lawn, Orlando, Fla. to
determine the percentage of (1) the eggs oviposited that hatch
and (2) the larvae that survive and successfully enter the stems
of alligatorweed. One hundred Vogtia eggs were placed singly
with a camel's hair brush in the axils of therminal leaves
at the rate of 1 egg per stem, along about 30 linear feet
of a continuous mat of alligatorweed. These stems were
then marked with red flagging tape for later recovery.
Five days later, the number of surviving first instar
larvae was determined by locating the marked stems and
recording the number of plants injured within a radius of
about 10 in. around each flagged stem. *

In a separate study at Lake Alice Vogtia eggs in
groups of 50 or more attached to strips of cloth were
pinned to stems of alligatorweed to detect and collect
egg parasites. The eggs were left in the field for two
days, after which they were collected and held in the
laboratory for emergency of moth larvae on parasites.

Larvae, prepuase, and pupae collected from the Lake
Alice samples were removed to the laboratory where they
were observed for parasitism. In this study, approximately
250 larvae and 45 pupae or prepupae were collected.

Greenhouse Colony

Permission to release V. malloi in the United States
was obtained in 1970. Pupae or eggs were collected from
the Buenos Aires area of Argentina, an area with climatic
conditions similar to the gulf coast states, and shipped to
the USDA quarantine facility in Albany, California. One
generation of Vogtia were reared at the Albany laboratory
to exclude parasites and diseases. Eggs collected from
that laboratory were shipped to the USDA laboratory in
Gainesville, Florida where a greenhouse colony was estab-
lished on alligatorweed grov?ing in water tables to obtain
sufficient numbers for field release.

Vogtia was first released in Florida in the spring of
1971 and later in the year in Georgia, North Carolina, and
South Carolina.

In 1972 specimens were collected 450 km south of
Buenos Aires, the latitudinal equivalent of Richmond,
Virginia, to broaden the genetic base of populations
already established and perhaps obtain a more winter-hardy
strain.

Specimens from the southern area of Argentina were
released at Fort Jackson (May 10, 1972), Twin Lakes (July
12, 1972), and Garden's Corner, S.C. (May 20, 1972).
Vogtia from this colony v/ere also released in Tennessee
Valley Authority projects and in North Carolina by indiv-
iduals employed in these areas.

10

Field Releases

1971

During the siUBiner of 1971, Vogtia were released at the
following locations (Fig. 2): Lake Lawn, Orlando, Fla. ;
Lake Alice, Gainesville, Fla.; USDA Plant Introduction
Station, Savannah, Ga.; Black Lake, Melrose, Fla.; Fort
Pierce, Fla.; Ashapoo River at Hwy. 17, S.C.; Savannah
River, near Savannah, Ga.; unnamed ditch, Gainesville,
Fla.; Santee Reservoir, near Lone Star, S.C., and Greenfield
Lake, Wilmington, N.C. Populations at the Ashapoo River,
the Savannah River, and the unnamed ditch sites were
destroyed by flooding, and the Fort Pierce site was destroyed
by dredging.

Lake Alice, Fla. — At Lake Alice, on the University of
Florida campus, conditions were sim.ilar to those found at
Lake Lawne, i.e. high nutrient levels and a luxuriant
growth of alligatorweed. The Vogtia releases were made in
a small stream, about 20 ft. wide flowing into Lake Alice
from a sewage treatment plant. At the time Vogtia were
released, alligatorweed extended out 5-10 ft. from the
banks on either side of the stream for a linear distance
of about 100 ft. A similar area, about 150 ft. upstream
from the release area, was designated as a control area.
No Vogtia were released there and there was no attempt to
prevent the spread of Vogtia to this area.

On May 18, 1971 small strips of cloth (ca. 2 in. ), on
which Vogtia eggs had been deposited (a total of 64) , were

11

Wilmington

X Xort Pierce

XjFort Lauderdale

Fig. 2. Vogtia Release sites, 1971 and 1972

12

taped to apical tips of alligatorv/eed. A second release
on June 6 consisted of 81 eggs or newly hatched larvae
(ca. 10 of the eggs had hatched) . In this release the
strips of cloth were pinned to the aerial steins in a yd^
area near the shore. Additional releases of 200 eggs on
July 20 and 580 eggs on July 26 v/ere raade as in the second
release. These releases were made a fev7 yards upstream
from the first two.

The Vogtia population was sampled during the larval
period of each of the 4 generations v/hich occured at this
site following the original release (sample dates: July 5,
August 28, September 29, and October 26). A sample con-
sisted of a 1-ft.^ area of alligatorweed, f.rom which aerial
stems were removed and counted. Samples taken on July 5
were evenly spaced along arcs 1 yd. and 4 yds. out from the
point of release. Seven samples were taken along the 1 yd.
arc and 5 were taken along the 4 yd. arc. Subsequent
samples were taken by randomly throv;ing a ft. wooden frame
on the mat of alligatorweed at 15 ft. intervals along
either side of the stream.

On January 27 after the first frost the Vogtia popula-
tion was checked to determine mortality due to cold and
again on May 2 to determine winter survival. These samples
were taken by randomly selecting injured stems.

Lake Lawne, Fla. — Lake Lawne is a 33 acre lake located
in Orlando, Florida. Sewage effluent from the surrounding
community drains into the lake creating severe nutrient

13

pollution which contributes to the luxuriant growth of
alligatorweed extending out 50-60 ft. from the shore in
some areas at the time of the releases.

On April 5 and May 20, 1971, 232 and 150 eggs respec-
tively, were placed on alligatorweed in the northwest corner
of the lake. Small strips of cloth, each with 10-15 Vogtia
eggs attached, were taped to aerial stems. Five days later
the eggs and stems were inspected. On June 23, a 2 x 4 ft.
section of alligatorweed mat was removed from the lake and
replaced by a Vogtia- infested mat of the same size from the
greenhouse colony. The infested mat and the surrounding
alligatorweed were inspected weekly to determine the number
of Vogtia larvae persent. Releases of adult Vogtia were
made in the same general area on July 23, August 13, August
18, and August 26 of 10, 7, 40 and 21 moths, respectively.

The Vogtia population was monitored periodically by
counting the number of injured plants in the release area.

Black Lake, Melrose, Fla. — Alligatorweed growth on
this 20-25 acre lake was much less vigorous than at other
sites when Vogtia were released. The stems were smaller
and more fibrous and the alligatorweed supported several
(Sacciolepis striata (L.) was the dominant species) species
of grass and weeds. The lake was encircled by alligator-
weed mats 5-20 ft. wide. On August 20, 1971, 19 adults
were released in a 3-ft.2 screen cage placed on the
alligatorweed mat. Seven days later after the moths had
completed oviposition the cage was removed. No further

14

releases were made at this site. Counts of larvae or wilted
tops were made November 15, 1971 and January 28, February 11,
and April 17, 1972 from randomly selected sites.

Savannah, Ga. — The only release at the USDA plant
Introduction Station consisted of Vogtia eggs placed in 3
screened plots (5X6 ft.) containing terrestrially-rooted
alligatorv;eed by Mr. IVilley Durden on April 14, 19 71 (Willey
Durden, personal communication) .

Greenfield Lake, Wilmington, N.C. — Vogtia eggs and
larvae were placed in a small stream 20-30 ft. wide flowing
into the northern end of the lake. The stream was completely
covered by a solid mat of alligatorweed for several hundred
feet. On August 2, 1971, a 4-ft.'^ section of mat was removed
and replaced by a Vogtia- infested mat of the same size from
the greenhouse colony. The mat was wrapped in 6 mil poly-
ethylene and a small amount of v/ater added for transporting.

Santee Reservoir, S.C. — Vogtia were released on the
south side of the reservoir, near the town of Lone Star.
Alligatorweed at the release site appeared as rooted plants
in shallow water. The stems were small and fibrous, a
condition often observed in rooted alligatorweed. On August
3, 1971, a 2-ft. section of Vogtia- infested mat from the
greenhouse colony was placed in the stand of alligatorweed
after a section was removed.

15

Field Releases

1972

Releases were made in 19 72 at the folloxving locations:
Fort Jackson (Upper Legion Lake, Coluinbia, S.C.), Twin Lakes,
S.C., Peeples' Pond (Garden's Corner, S.C.), and a roadside
ditch (Fort Lauderdale, Fla.) (Fig. 2).

Fort Jackson, S.C.— Upper Legion Lake is a small man-
made lake (10-15 acres) ringed with alligatorweed extending
out 5-20 ft. from the shore. On May 10, 72 larvae were
released on the east side of the lake, by placing them in-
dividually, with a camel's hair brush in the terminal leaves
of alligatorweed stems.

During the summer, 1972, the Vogtia release site was
siobjected to malathion fog, used in mosquito abatement, and
to Diquat at 2 lbs. /acre sprayed in alternating strips.
The Vogtia population was sampled September 11 and November
15, 1972 and June 16, 1973.

Twin Lakes, S.C. — Twin Lakes consists of two ponds,
about 5 acres each, connected by an old riverbed and located
near Summerville, S.C. Alligatorweed normally covers all
but a small area in the center of both ponds.

On July 12, 100 adults were released in the norther-
most pond where alligatorweed was 16 to 18 in. tall in
relatively dense stands. The Vogtia population was sampled
on September 12 and November 15, 1972.

Garden's Corner, S.C.-- Peeples Pond is approximately
10 acres in size and ringed with alligatorweed extending

20 - 30 ft._ out from the shore. On May 20, 150 Vogtia
eggs were placed on the aerial stems of alligatorweed and
25 adults v/ere released.

Both Agasicles and Vogtia populations were sampled on
July 13, and September 11, 1972 and June 16, 1973.

Fort Lauderdale, Fla. — On April 19, 85 first instar
larvae collected in the wilted tops of alligatorweed at the
Melrose, Fla. (Black Lake) site were released in the road-
side ditch (20 to 30 ft. wide) alongside the Sunshine State
Parkway 1/2 mile north of exit 16. Three months later,
July 18, 125 adults from the greenhouse colong were released
in the same general area. Alligatorweed grew luxuriantly
along the ditch for approximately 0.5 mile on either side
of the release site.

The Vogtia population was sampled July 18 and September
20, 1972 and April 26, 1973.

RESULTS PJ'^D DISCUSSION
Biology and Behavior of Vogtia mallei

Fecundity and Fertility

According to Maddox and Hennessey (1970) , the female
moth lays an average of 267 eggs with 6, 34, 18, 15, 12, 6,
5, and 4% of the total laid from the first through the eighth
day, respectively.

In laboratory studies 25 female moths held in gallon
containers laid only 132 eggs (average) . The adults used in
the test were taken from a crowded greenhouse colony and may
reflect a reducted activity due to crowding. This study confirme(
Maddox' s findings that the greatest percentage (69) of the
eggs were laid on the second day following emergency (Fig. 1) .
Although some eggs were laid on the first day, most were
apparently infertile and did not hatch.

Only 84.3% of the eggs collected from moths in gallon
containers hatched. Of 1079 eggs from moths in pint containers,
the percentage was only 49.7%, presumably from a lack of mating
in the small containers.

A mated female, held for observation in a pint con-
tainer, laid 47 eggs in one period which began at 11:00 PM.
The average time interval between eggs was 17.3 seconds.

17

18

Y axis
-2.0

log
of

eggs
laid

1.6

1.2

Linear regression analysis
y = a(o) + a(l) X

p.s

>

\ 1

<

0.4

\

L_»

l_.. ™..

\

X axis

DAYS

Fig. 1. Log number of Vogtia eggs laid plotted against
time.

19

This moth was very active dujring cviposition, flitting
around the container and stopping only long enough to
deposit an egg.

Oviposition

Most of the eggs (60.6%) V7ere deposited on the under-
side of the terminal leaves, with progressively fev/er eggs
being placed on leaves of the lower nodes (Table 1) . These
data confirmed Maddox ' s earlier report that eggs are de-
posited on the underside of a leaf in the first through
fourth pairs of apical leaves on the host plant.

Table 1. Egg placement by V. malloi female moths caged

over alliga

ton

,\?eed.

Margin

M,

idrib

Axil

Total

% Total

Terminal

17

19

24

63

60.6

leaves

Below 1st

3

8

8

19

18.3

node

Below 2nd

4

4

7

15

14.4

node

Below 3rd

2

1

4

7

6.7

node

26

32

43

104

Egg and Larval Mortality

The expected egg viability as indicated by laboratory
studies was 85%. The number of damaged stems recorded
indicated that 41% of the eggs produced larvae which caused
visible injury to the alligator^/zeed stems, leaving 44% of
the Vogtia eggs or larvae lost to predation or other

20

environmental factors.

Several specimens of the egg parasite Trichogramma
perkinsi Girault (identified by L. R. Ertle USDA) were
collected from the Vogtia eggs attached to alligatorweed
for 2 days .

Six internal parasites belonging to 2 species of
ichneumonids were reared from the Vogtia larvae and pupae
collected from Lake Alice: 4 were Gambrus bituminosus
(Cushman) and the other 2 were G. ultimus (Cresson) (determi-
ed by R. W. Carlson USNM) . Both species have been collected

from a number of Lepidoptera in the United States (Muesebeck
et al. 1951) .

Behavior

Upon hatching, the first stage larva enters and girdles
the alligatorweed stem, usually at the second internode.
Later in the season, the aerial stems may become tough and
fibrous, and the young larvae may descend on a silk strand
to the axillary shoots growing farther down the stem which
are still tender (Brown and Spencer 1973) .

Within a few hours, the girdled top is wilted, and
v/ithin a few days it usually breaks off at the node just
below the girdle and falls onto the supporting mat of
alligatorweed with the young larva still inside. The
larva soon leaves the dead top and bores into another stem,
or the same stem at a lower level, and the process is repeated,

As the larva matures the process of girdling and feed-
ing on the interior of the hollow stems above the girdle
may be repeated raany times (as many as 9 stems were reported

21

destroyed by a single larva, Maddox 1S)70) . The mature larva
is known as a roving stem borer because of its characteristic
feeding habit. The larger the larva, the farther down the
stem it girdles, and frequently mature larvae are found
feeding below the water line, protected from drowning by
the air-filled cavity of the hollow stems.

When the Vogtia larva bores into a stem it quickly
seals the entry hole with silk strands to make it water-
proof. The exit holes however are never sealed leaving
the plant open to attack by other organisms.

Just prior to pupation the mature larva enters an
internode and seals off both ends at the nodes. A small
"window" is chewed in the stem wall, leaving only a thin
layer of the outer epidermis intact. The larva spins a
cocoon and orients itself so that the head region is
adjacent to the window, through which the adult will
eventually emerge.

22

Field Releases

1971

Lake Alice, Fla. — Larval counts taken at Lake Alice
(July 5) along arcs 1 yd. and 4 yds. from the release area .
were not significantly different, and no directional move-
ment was indicated; the mean number of larvae/ft.^ along
these arcs were 5.4 and 5.2, respectively. The standard
deviation of insect counts in the 12 samples was 2.84.
The population, v/ithin this limited area was distributed
with X = 5.3, S^ = 8.06, S^ /X = 1.52, and K = 10.19.
Adult dispersal appeared to be random. Counts from later
samples indicated that the population expended at a nearly
uniform rate in all possible directions v/ith each new
generation. Population density increased to 9.4 larvae/ft.
by the second generation in the immediate release area and
then began a steady decline as the attractiveness of the
alligatorweed was reduced and the number of available egg
deposition sites diminished (Fig. 3) .

During the first two generations the area of infesta-
tion was not significantly expanded based on observations

of wilted tops. The mean number of healthy aerial stems of

2
alligatorweed/f t. was reduced from 52.5 to 36.9. An

invasion of the test area by the southern beet webworm,

Herpe to gramma bipunctalis (F. ) , occurred during the second

generation v/hich resulted in partial defoliation of many of

the aerial stem.s. Although v/ebworm damage was temporary,

it may have influenced the observed movement of V. malloi

23

.iJ/swais lYiHav

o

o

o

o

o

in

"^

ro

(N

rH

o

1

1

!

T-

7

1

/ 1 ^

/

1

CN

c

/

'

I

•

U

0

-p

>

/

/

O

o

i

/

/

/

/

/

a)

•

/

/

/

o^

-P r~-

•H c?i

CN

J

/

/

CM

•

B

^

/

/

•

Eh

(0 0)

o

/

/

/

/

Id-up
e Ali^

r

/

/

/

n bui

o ^

/

tN

/

CO

(T3 S

/

\

Cvi

:3 0

to

a-P

i

/

S

•

0 (0

i

H

•

o

Q. tr

1

/

H

D

-H

1

CO

<;

to

rH

^

I /

■H
-P

rH

(0

V

0
>

/

/

00

•"a*

tN

in

n

-H

.giiC/aVAHYTE

24

adults from the release area to the nearby control area
during the second generation, a distance oC 150 yds.

In the third generation sanples (September 29) , the
number of larvae, ft. had dropped to 6.3, v/ith the nunber of
healthy aerial stems reduced to 17.4/ft.^. By the middle
of the fourth generation (October 26) , the mean number of
larvae and aerial stems/ft. had dropped to 3.4 and 4.0,
respectively. With this damage the alligatorweed mat was
beginning to break up in the release area and was no longer
increasing. Other semiaquatic plants began to move in over
the submersed portion of the alligatorweed mat.

A one-way analysis of variance showed that the reduc-
tion in the number of aerial stems between sample dates
was statistically significant (0.01 level) but there was
no significant difference between insect counts for these
dates, probably due to the wide variation between samples.

The January 27, 1972 data consisted of superficial
observations only. The lack of alligatorweed stems due to
the cold and to Vogtia and Agasicles dam.age made detailed
samples impractical. Ten first and second instar larvae
collected from wilted stems in 2 sm.all mats of alligator-
weed were not in diapause, but were actively feeding.

On April 25, 1972, the alligatorweed had attained a
height of 15 in. when the Vogtia population was sampled.
The mean number of larvae/ft.^ was 3.3 at the release site;
therefore a substantial number of Vogtia survived the v/inter.

25

In 19 72 the Lake Alice population duplicated the 19 71
response, i.e. the Vocjtia population peaked at about the
same density and then began a decline (Fig. 4) .

^^•^ke Lawne, Fla. — The first releases made in this
area (April 5 and May 20, 1971) were probably unsuccessful
since no damaged stems were found, a result of the newly
hatched larvae becoming entangled in the tape. The second
attempt to establish a Vogtia population in Lake Lawne
(June 23) was likewise unsuccessful; no larvae were found
after the first week. Most of the Vogtia present in the
infested alligatorweed were probably drowned due to flood-
ing of the cut stems. A few larvae were found the following
week but they apparently did not survive to reproduce. On
July 23, 10 adults were released at the same location from
which a small larval population became established. Three
additional releases were made in August to supplement the
population.

The Vogtia population did not expand its numbers

rapidly. The population remained at low levels (ca. 0.1

2
larvae/ft. ) and larvae were never found more than 50 yd.

from the point of release during the summer of 1971. At

Lake Lawne where no frost occurred the alligatorweed

remained green throughout the winter. On May 21, 1972 the

population had increased to 3.7 Vogtia larvae/ft. in the

release area.

Black Lake, Fla. — Only one release of 19 adults was

made at this site on August 20, 1971. On November 15

26

.i.^/SRaoLs ^vIHav

c

0

+J

u

<D

M-l

W-l

•

Q)

CM

r-

U)

en

4J H

•H

OJ

'd

o

c

•H

m

rH

<

a

13

cu

1

^

Ki

to

rH

l-q

•H

:3

•>

Xi n3

(D

C

0

0

5

•H

S-l

-P

o

to

-p

nH

m

3

en

cx

H

0

rH

Q^'-i

,I,j/aVAHVT

•H

4J

O
>

«

Cn
■H

27

numerous injured alligatorweed srems were found \ip to 50 ft.
from the point of release and a few injured plants were
found about 100 ft. away. By January 23, 1972-, the Vogtia
population had expanded to all areas infested, by alligator-
weed in the lake. Vogtia larvae v;ere active and still
feeding inside aerial stems recently killed by frost; most
of these were early instars. On February 11 the larvae
were found feeding below the frost-killed aerial stems on
green stems that v/ere protected from frost. By April 17,
1972, the Vogtia population had attained a density of about
4.0/ft. around the entire lake.

By the end of the summer the alligator>/eed at Black
Lake had been so reduced in area and density of the stems
that in many areas of the lake it was no longer the dominant
species. The suppression of alligatorweed at this site v/as
probably due to a combination of factors, i.e. insect
pressure, interspecific plant competition, and low nutrient
levels in the water.

Savannah, Ga. — A Vogtia population became established
and survived the 1971-27 v/inter despite killing frosts on
terrestrially-rooted alligatorweed in experimental plots.
Mid-winter sampling failed to locate the overwintering
insects (burden, personal communication) . In South America,
Vogtia v;as found during the winter in the roots of terres-
trial alligatorweed (Vogt 19 60) .

*Willey C. Durden

28

Greenfield Lake, N.C. and Santee Reservoir, S.C. —
Vogtia has not been recovered at the 2 northermost release
sites. The method of release, i.e. the transfer of infested
mats, was proved unsatisfactory at Lake Lawne and may be one
reason for lack of establishment at these sites.

29

Field Releases

1972

Fort Jackson, S.C. — By the end of 1972 only small
patches of alligatorweed remained at this site because of
the herbicidal treatment. The effect of removing most of
the. alligatorweed on the Vogtia population was to concen-
trate the insects on the remaining mats of alligatorweed.

Samples taken November 15 indicated a population density

2
of 5 larvae/ft. . Earlier samples taken September 11

showed only 1 larva/ft. .

On June 16, 1973 after a severe winter in this area,

the Vogtia population was 0.6/ft. and the alligatorweed

was killed back to the water in most mats. A few protected

areas had healthy alligatorweed, and in one such area adult

Agasicles were found which had presumably survived the

winter as well.

Twin Lakes, S.C. — The September 11 sample date showed

2
1.0 Vogtia larva/ft. with moderate defoliation by Agasicles

(approximately 50% of the leaf area missing) . By November

15 the Vogtia population had increased to 2.6/ft. . This

area was not sampled in 1973 because of spring flooding.

Garden's Corner, S.C. — Garden's Corner was the only

site where Agasicles stripped all the foliage and many of the

small shoots from the aerial stems of alligatorweed after a

Vogtia population had become established. The Vogtia

population remained low but v/as not elimianted. V7hen the

Vogtia release was made (May 20, 1972) the Agasicles adult

30

9

population was 5 adults/ft.". By July 13 the alligatorweed
was under severe stress and the Agasicles population was
11 adults/ft. and the Vogtia population was 0.66/ft. .

Samples on June 16, 197 3 indicated the Vogtia popula-

2
tion density was 1.6 larvae/ft. after a severe v;inter for

that area.

Fort Lauderdale, Fla. — Samples on July 18 indicated

that the Vogtia population had attained a density of 0.43

2 9

larvae/ft. from the first release. A population of 3.4/ft.

was recorded on September 20, 1972. In both cases the
sampling was done within approximately 100 yds. of the
release site. The population density declined as the dis-
tance was increased from the release area. *

On April 26, 1973 just after spring growth had begun

2
the population was 2.25/ft. . Therefore the population

was probably not reduced significantly during the winter -

period. This was expected since frost did not occur in

this area.

ALLIGATORWEED PRODUCTIVITY

Introduction
To evaluate the effects of stress caused by biological
control agents to growing plants, comparisons must be made
between that plant growing undisturbed, and that plant with
the additional stress factors added. Environmental condi-
tions, nutrient levels, and many other factors play a
significant role in regulating a plant's grov;th and spread.
This is especially true of plants such as alligatorweed
that are adapted to a wide range of conditions. With these
limitations in mind, attempts v/ere made to determine the
net rate of biomass production in alligatorweed growing
under favorable greenhouse conditions, and in four lakes
which had varying levels of injury from biological control
agents.

Methods and Materials
Greenhouse Studies

Alligatorweed v/as grown in a 4 X 12 ft. X 6 in. green-
house table divided into three 4X4 ft. sections. The two
end sections were covered v/ith 6 mil polyethylene and
filled with tap water to a depth of 5 in. On December 13,
9.5 lbs of alligatorv'/eed from the floating portion of a mat
of weeds taken from Lake Alice, Gainesville, Fla. were

31

32

placed in the water-filled sections. The middle section
contained a Henry J. Green hygrothernograph, which recorded
the temperature and relative humidity at the approximate
water level.

Artificial lighting, about 3000 ft. candles, was
supplied by 12 Sylvania 40w VHO cool v/hite florescent
bulbs and 24 f 75w incandescent lamps attached to a light
board and suspended 11 inches from the table top. An
Intermatic model TlOl timer controlled the 12 hour photo-
period.

Each plant section was initially filled with tap
water to v/hich 150g milorganite and 200g of 20-20-20 soluble
plant food was added. On January 4 another 180g of the
same fertilizer plus 45g of 1.82% iron supplement were
added to each section.

Half of the growing mat in each section was removed
on January 10, the remainder on January 16, terminating the
study. To reduce experimental error, alligatorv/eed was
taken from alternate sides of the table. The plant material
was allov/ed to drain for 5 minutes and then weighed on a
Fairbanks Morse portable platform scales.

Water samples were taken at the beginning of the
experiment and on January 4. These samples v;ere analyzed
for common nutrients, i.e. nitrates, phosphates, potassium,
chlorides, magnesium, manganese, iron, copper, and calcium
by the Soils Laboratory, IFAS, University of Florida.

The mean daily temperature was 29°C., the mean daily RH

33

was 50.6%, and the pH of the water was 7.5. The water tem-
perature on the first sample date was 26°C.

Field S^yudies

Standing bioraass samples were taken in early spring,
before much growth had occurred (March 19, 20, 22, and 23),
in early summer (May 2, 10, 20, and 24), and in midsummer
(July 13, 18, 27, and August 1) at two lakes (Black Lake,
Melrose, Fla. and Garden's Corner, S.C.) and three streams
(stream flowing into Lake Alice, Gainesville, Fla. , roadside
ditch, at Ft. Lauderdale, Fla., and Twin Lakes, S.C.).
Alligatorv/eed characteristically grows very rapidly in the
spring and then considerably slower during the rest of the
growing season. These samples represent both growth periods
for alligatorv/eed at these sites.

A stainless steel cylinder was constructed to extract
a 1.2 ft.^ core sample from the floating mats of alligator-
weed. One end of the cylinder was sharpened to facilitate
cutting through the mass of alligatorweed, and the other
end was covered with 1/4 in mesh hardware cloth. A sample
was removed from the water, after cutting, by inverting
the sampling device. The sample was allowed to drain for
5 minutes and then emptied into a plastic bag and v;eiqhed
with a hand-held scales. The same mat of alligatorweed was
used for all sam.ple periods at each location. All samples
were taken half way between the shoreline and the leading
edge of the mat.

34

Results c:nd Discussion

Greenhouse Studies

Although the study began December 13, the actual growth
period of the alligatorweed did not begin until several days
later. When the alligatorv;eed was transferred to the green-
house table the aerial steins were broken and disoriented to
such an extent that growth was effectively stopped until nev;
stems were produced. Therefore the first samples at 28
days showed little change (Table 3) , since the plants had
been actively growing for 10 days or less. The increase in
biomass during the first peri.od was 64%. The second sample
period, lasting only 6 days increased its biomass by 48%.

Nutrient levels (Table 2) were much higher than those
normally found in natrual lakes or streams. In Lake Alice,
for example, only phosphates were higher than in these water
tables (2.98 ppm) . Iron was 0.25 ppm, nitrates were 5.3 ppm,
chlorides were 25.0 ppm, and the pH was 7.75 on May 2, 1972.

Table 3. Net productivity of alligatorweed in greenhouse
tests

Wet v/eight of samples (lbs)

Initial wt . 28 days 34 days % increase

Rep. 1 9.5 14 19 200

Rep. 2 9.5 '11 18 189

35

Table 2. Nutrient levels in water tablc-s used for productivity
studies (ppm)

Sample No3 P K CI Mg Mn Fe Cu Ca

Rep. 1 22.0 0.25 70.9 92.0 1.43 .075 1.80 0.20 0.90
Rep. 2 22.0 0.30 70.0 92.0 1.28 .075 1.90 0.20 0.82

tap water 4.3 0.0 0.0 22.5 0.73 0.0 0.0 0.10 1.10

Field Studies

Alligatorweed productivity in the field did not approach
that in the greenhouse. The Tv/in Lakes site had a high
percent increase, but the growth period was 52 days, compared
with 34 days for the greenhouse study (Table 4).

The early summer daily increment factors for the five
sites were .104, -.003, .082, .035, and .028 for Twin Lakes,
Fort Lauderdale, Garden's Corner, Lake Alice, and Black Lake,
respectively. The daily increment factor for the greenhouse
alligatorweed was .26, or 2 1/2 times the hignest increment
factor recorded in the field.

On a per acre basis, the Twin Lakes growth rate was
equivalent of 3,775 lbs of biomass produced per acre per day.
In the 52 day period an acre of alligatorweed would produce
193,300 lbs of new growth. That is approximately tv/ice as
productive as the most productive aggricultural crop, in half
the time. Corn grown on muck soils produces 2-30 tons of
silage in about 100 days (Cunha and Rhodes 1965) .

36

At Garden's Corner during the summer sample period, and
at Fort Lauderdale, to a lesser extent, the standing biomass
was reduced. The reduction at Garden's Corner corresponded
with a heavy attack of the Agasicles beetle;, in which the
aerial stems of alligatorv^eed were completely stripped of
their foliage and small stem.s. The reduction at Fort Lauderdale
was less dramatic and may have been due, in part, to severe
interspecific plant competition.

37

— >1

. m

m U

X!\

iJ O

^ <

LO

o

u>

>.D

r-

VD

'ir

(Ti

\

r--

r-

CN

r-

00

rH

LO

O

-tJ •

j-^

CM

r-

CTi

00

o

CN

rH

G 4J

n

r-{

CN

H

rH

1

(1) IS

1

g

(D<N

Vj .

U +J

C ^

•H CM

>irH

'=1'

in

CN

CN

00

t^

CO

rH X,

o

n

og

CO

LO

CN

o

o

(N

■H .

<-i

o

o

o

o

O

o

o

r^

to 4-1

•

•

•

•

•

•

•

•

en

C S

1

1

iH

^

>

4J

•H

>

cu

•H

w

-P

fO

O

Q)

d

U

n3

O

o

CN

^

n

■^

o

VD

\o

0

c

rH

'*

OJ

r^

CN

■^

1

U

•H

iH

1

a

oV>

13

QJ

Q)

^

M

O

w

-P

>1

(0

re

&i

13

-H

(N

O

rH

—\

■^

CO

r-

G-i

tH

•

LO

-^T

VO

^

LO

ix;

lO

CO

iH

0

(13,

•^

o

m

iH

0

13

OJ

iH

o

CO

"^S"

r-

O

>.

CN

•H

rvj

CN

w

•H

(t!

>i

♦

<u

P

2

r3

Cr.

>1

rH

>i

rH

1

•H

0)

S

P

fO

>-(

(0

13

Tl

G.

1

1

<

s

n

S

t-D

o

:3

n

1

1

1

r

CN 00

4J

(1)

ro

CN

rH

o

o

CM

rH

r-i

(0

r-l

rsi

CM

CM

01

rH

a

•

•

O

•

•H >,

13

E

h

!-l

c

U

>,

S-l

>!

M •-{

iH

(0

nj

(C

3

m

(0

(li

(0

a :3

(D

m

s

?

I-)

'r-'

21

s

2

<: to

•H

•

d

fa

•-

^ 0)

U

^rH

1

U)

0) rH

•

<U fa

u •

Q)

U rH

cn

X

CD fO

•

c

^

■H -H

w

(0 -

13 rH

'^

0

m

■H >

» ^

J 0)

3 fa

•H

J

< C/l

c v^

CO

fd

0)

-p

•

<u

OJ (U

A^ O

fj -

rH

(d

c u

(1) c

«

13 C

U Sh

0)

^

o

■H

M -H

m

^^ ^-1

f3 rH

. rH

«0

0

5 •

a rrj

rH

(0 0

rH t)

4J ni

EH

h^

&H C/D

r^ U

fa

U U

m S

fa 13

INTERRELATIONSHIPS BETWEEN ALLIGATORWEED
AND TVv'O INSECTS— Vogtia and Agasicles

Introduction

Pioneering plants or opportunists such as alligatorweed
may because of their tolerance to wide environmental condi-
tions appear in different areas with very different growth
and morphological characteristics. Alligatorweed in some
areas may produce numberous , short, fibrous stems while in
other areas the aerial stems may be much less dense in
number and taller, thicker, and more rapidly growing. Under
these conditions, it becomes increasingly difficult to
evaluate the host plant-insect relationship, and almost
meaningless to compare insect population data between the
areas. Therefore a great number of samples (287) from all
the release areas have been combined to provide an overall
estimate, or comparison, of the effect of a given parameter
on alligatorweed with the effect of all other parameters
considered at the same time. The variables representing
the two insects have been evaluated with and without the
influence of the other variables being included in the model.

Two stastical methods were used to shov; the interactions
between the Vogtia and Agasicles populations, between each
of these insects and alligatorv/eed, and between available
nutrients and alligatorv/eed growth.

38

39

In the canonical correlations program, 24 variables
were grouped into sets. The sets were then compared and
significant relationships were defined. The second program
was a full multivariate regression model which explained
the significance of each of the measured variables in
influencing alligatorweed growth.

In canonical correlations, a set of independent vari-
ables may be compared with a set of dependent variables to
find the linear con±)ination of variables in each set which
when correlated is maximum. The resultant variable is
known as a canonical variate. If some linear relationship
between the sets of variables still remains unaccounted for
by the first set of canonical variates, the process of find-
ing new linear combinations that would best account for the
resideual relationships between the sets can be continued.
This process can go on until there are no significant
linear associations left. Each canonical variate is
orthogonal to (or unrelated to) other canonical variates.
The chisquare test is used to evaluate variates for signi-
ficance.

In canonical correlations, variables in one set may
be combined to predict maximally the variations of the
variables in the other set. This process was utilized to
compare data frcm lakes and streams, and to predict the
number of samples that should occur in using different sets,
lakes or streams (See Morrison 0.961) for detailed discussion
of canonical correlations and tests of significance) . All

40

statistical, methods were analyzed by computer program, BND,
SPSS (Nie, 197 0) on the University of Florida North East
Regional Data Center's IBM 370/165.

Methods and Materials
The Vogtia and Agasicles sampling procedure was described
in the section on colonization, i.e. a ft. frame was used to
determine the population density, as well as to gather plant
material for stem measurements. Sample sites were selected
either by a random process (Lake Alice, Black Lake, and Lake
Lawn) or by taking samples at preselected 15 feet intervals
(Garden's Corner, Ft. Lauderolale , and Twin Lakes). Aerial
stems v/hich were selected for measurement were representative
of the sample in both height and diameter. In most cases,
the uniformity of the samples made multiple plant measure-
ments unnecessary.

Measurements for each of the sample sites included the
following: stem height, length of the fourth internode,
diameter of the fourth internode, stem density, percent
leaf area missing, Agasicles/ft. , vogtia/ft.'^ (expressed
as a total and as each of the instars as a separate variable) ,
pH, alkalinity, iron (ppm) , turbidity (jtu), chlorides (ppm) ,
phosphates (ppm), nitrites and nitrates (ppm), hardness,
water depth, and whether the sample was taken from a lake or
stream. The fourth internode was chosen for plant measure-
ments because it was the first fully developed internode.

41

Water samples were analyzed for nutrients in the laboratory
with the use of a Hach water testing kit.

A total of 287 samples were taken from the six following
release sites, three of which were lakes and three were
streams. The number of samples taken from each site is shown
in parenthesis: Lake Alice, Gainesville, Fla. (76); Black
Lake, Melrose, Fla. (155); Lake Lawn, Orlando, Fla. (34);
Peeples Pond, Garden's Corner, S.C. (2); Roadside ditch.
Ft. Lauderdale, Fla. (17); and Twin Lakes, Summ.erville ,
S.C. (3), (Fig. 2).

42

Results and Discussion

Insect-Host Plant Relationship

In the first analysis, the set of plant characteristics
was compared with the set of variables representing numbers
of Agasicles and Vogtia present and a measure of Agasicles
feeding damage. As a result of these correlations, 3 new
canonical variates were formed (Table 5 and Table 6). Each
of these new variates was significant at the .01 level of
probability, indicating that 3 independent and significant
linear relationships existed.

If we examine the coefficients of the individual
variables (Table 7) , we can explain the combinations of
variables making up each of these associations. The first
canonical variable represents the most important of the
linear associations, and in this case there is a positive
association between plant height, internode length, inter-
mode diameter and the number of Agasicles and Vogtia , and
a negative association between the same plant characteristics
and Agasicles feeding. Therefore, both Vogtia and Agasicles
tend to occur more frequently in alligatorweed with larger,
taller stems. Vogtia, having a larger coefficient, probably
favors these conditions more than Agasicles . The negative
association between the plant characteristics and Agasicles
damage was expected. The more damage done by Agasicles ,
the more stress is placed on the growing plants , resulting
in smaller unthrifty stems. *

43

The second canonical variate of less importance than the
first, but equally as significant, consisted of a negative
association between Agasicles feeding and plant height, and
internode diaraeter. This may appear to be in contradiction
of the first canonical variate; however, the canonical
variates must be considered separately. Therefore, 2 rela-
tionships exist, one superior to the other. When plants
are heavily attacked by Agasicles , the upper internodes may
be damaged sufficiently to cause them to be dropped off
resulting in a condition described by this relationship.

Table 5. Variables used in 'canonical correlations with
mean and standard deviations.

Variable Mean Standard Dev. Cases

SET I

Plant Height 11.895

Internode Length - 2.2250-: •.

Internode Diameter .12007 ■ •

No. Stems 51.212

Percent damage 15.174

SETII

No. Agasicles 1.536

No. Vogtia 3.515

1st Inst. Vogtia 1.268

2nd Inst. Vogtia 0.651

3rd Inst. Vogtia 0.498

4th Inst. Vogtia 0.456

5th Inst. Voatia 0.418

3.90

287

1.321

287

.05771

287

21.85

287

26.328

287

2.634

287

4.138

287

2.095

287

1.219

287

0.876

287

0.918

287

0.280

287

44

Table 5 continue

:d.

Variable

Mean

Standard

Dev.

Cases

Vogtia Prepupae

0.066

0.288

287

Vogtia Pupae

0.177

0.465

287

Table 6. Canonical Variates in Analysis I

Number of Canonical Chi-Square Degrees of

Variate Sets Freedom

1 156.99149 12**

2 75.90101 6**

3 12.57912 2**

** indicates significance at .01

Table 7. Canonical Coefficients for Individual Variables,
Analysis I

Can. Var . 1 Can. Var. 2 Can. Var . 3

1 Plant Height

2 Internode Length

3 Internode Diam.

4 No. Stems

5 % Leaf -Area Missing

6 No. Agasicles

7 No. Vogtia

0.345*

-1.152

-0.469

0.487

-0.022

0.628

0.329

0.726

0.188

-0.027

-0.535

0.906

-0.638

0.704

-0.554

0.321

0.394

0.979

0.645

0.443

-0.700

♦Variables underlined contributed most significantly to
the linear relationship.

45

The th.ird canonical variate, likewise independent of
the first 2 and of less importance, consisted of a positive
association between the number of Agasicles and internode
length and the number of stems. We can conclude from this
that adult Agasicles tend to occur more frequently in dense
stands of alligatorweed with samller stems, and the latter
condition is negatively correlated with the number of Vogtia
and with Agasicles feeding. The Agasicles feeding and stem
size association is in agreement with the first canonical
variable but much smaller in magnitude. The positive
correlation of Vogtia with stem size is expected since in
small stems the full grown larvae may be as large in diameter
as the stem itself and, therefore, not protected from para-
sites and predators or environmental factors.

It should be noted that 63.96% of the variation is
e3q)lained by canonical variable 1, 30.92% by canonical,
variable 2, and only 5.12% of the variation is explained
by canonical variable 3. The most important associations,
therefore, are probably described by canonical variables
1 and 2.

In the second analysis, the same 4 plant characteristics
i.e. plant height, internode length, internode diameter, and
number of stems were used to compare with the number of
Agasicles, Agasicles feeding damage and each of the 7 immature
stages of Vogtia present.

In this correlation, 4 new canonical variables were
formed (Table 8). The first 2 were significant at the .01

46

level of probability. The third was significant at the .10
level of probability and, therefore, will not be considered
in this discussion. The fourth variate was not significant.
The first variate incorporates 61.22% of the total variation
and the second 29.87%.

Table 8. Canonical Variates in Analvsis II.

Number of Canonical
Variate Sets

1
2
3

4

Chi-Square

195.18961

94.64266

23.28067

5.74528

Degrees of
Freedom

36**

24**

14

6

■* indicates significance at the .01 level of probability

An inspection of the canonical coefficients for the
individual variables (Table 9) reveals a positive associa-
tion between stem diameter and internode length with the
5th instar Vogtia. This relationship, as explained in the
first analysis, has been observed in the field in areas where
alligatorweed was under severe stress. During temporary
stress periods, the aerial stems may become small, v/ith the
stems making up the mat remaining large. In this case,
the larvae can survive in the mat unless flooding occurs
which may drown the larvae. Another relationship occurring
in the first canonical variable is a neaative association
between Agasicles feeding and stem size. This relationship,

47

also apparent in the first analysis, is a result of Agasicles
damage producing stress in the alligatorweed which resulted
in smaller, less thrifty plants.

The second canonical variate shov/s a negative associa-
tion between Agasicles feeding and plant height. As dis-
cussed earlier, reduced plant height with increased Agasicles
damage is a direct result of Agasicles attack. Since the
plant suffers more damage in the top few internodes, the
more severely attacked plants are much more prone to break-
ing over, and to dropping these upper internodes.

Table 9. Canonical Coefficients for Individual Variables,
Analysis II

Variable Can. Var. 1 Can. Var. 2

1.156
0,200

-0.561
0.578

-Q..g55

SET I

Plant Height

0.036

Internode Length

0.550

Internode Diam.

0.387

No. Stems

-0.220

% Leaf Area Missi

ng

-0.359

SET II

No. Agasides 0.273 -0.195

1st Instar Vogtia 0.187 0.141

2nd Instar Vogtia 0.028 -0.114

3rd Instar Vogtia 0.021 -0.145

4th Instar Vogtia 0.237 -0.090

48

Can. Var. 1

Ca

n. Var.

2

0.544

-0.075

0.051

-0.147

0.119

-0.076

Table 9 continued.

Variable
5th Instar Vogtia
Vogtia Prepupae
Vogtia Pupae

Nutrient levels and alligator^.veed characteristics

In a similar program the same plant characteristics
were compared with the physical and chemical parameters
associated with water quality at each site. The second set
of parameters were: pH, alkalinity, iron, turbidity,
chlorides, phosphates, nitrates, hardness, depth of the
water, and the lake or stream origin of sample (Table 10) .

Table 10. Variables used in canonical correlations III (IN)

Variable Mean S.D. Cases

SET I

Plant Height ' 11.89

Internode length 2.225

Internode diameter .1201

No. stems 51.21

■ 3.90

287

1.321

287

.0577

287

21.85

287

SET II

pH

7.02

0.99

alkalinity

74,14

74.21

iron

0.23

0.29

turbidity

36.13

2 0.15

chlorides

18.48

12.2

287
287
287

287

237

phosphates

1.32

1.75

nitrites & nitrates

3.68

1.35

hardness

99.21

•120.45

49

Table 10 continued.

Variable Mean S.D. Cases

287
287
287

These canonical variates produced four linear relation-
ships which were significant at the .01 level of probability
(Table 11) . As in the first program, each of these relation-
ships is independent of the others. The first and most
important association is a negative correlation between
water depth and plant height and a positive correlation
between plant height and turbidity. Since larger coeffi-
. cients occur between water depth and plant height those are
likely the more important variates in this linear relation-
ship. Therefore, as water depth increases, plant height
tends to decrease. Secondarily, as turbidity increases
plant height also increases. The latter relationship
probably reflects nutrient pollution which would tend to
produce larger, more vigorous stems (Table 12) .

Table 11. Canonical variates in analysis III

Variates Chi-sauare D.F.

1 526.21 40**

2 233.11 27**

3 76.46 16**

4 24.03 7**

**

Indicates significance at the .01 level of probability

50

The second canonical variable shows a second positive
relationship between plant height and turbidity, independent
from the first. In this case however, the major emphasis
should be placed on the positive association between inter-
node diameter and water hardness, where as water hardness
increases, so does the diameter of the aerial stems.

Table 12. Canonical coefficients for individual variables,
Analysis III

Variable

C

an

. Var.

1

C

an. Var.

2

C

an. Var.

3

C

an. Var. 4

SET I

plant Ht.

-0.033

-0.701

-0.168

-0.784

internode

L,

0.681

0.129

-0.145

-1.343

internode

D,

-0.430

■ 1.086

0.779

1.204

No. Stems

0.124

-0.230

1.022

-0.245

SET II

pH

0.130

-0.038

0.861

-0.053

alkalinity

-0.150

-0.226

0.808

-0.697

iron

0.129

0.185

-0.216

0.563

turbidity

■ -0.773

-0.651

0.777

-1.003

chlorides

-0.042

-0.088 •

0.056

0.140

phosphates

0.053

0.450

0.899

0.375

nitrates ,

-0.047

-0.113

-0.745

-0.117

nitrites

hardness

0.005

0.981

-1.740

0.612

depth

1.376

0.369

1.281

0.306

Underlined coefficients contributed most to the linear
relationship.

51

The third canonical variable indicates a negative
relationship between the number of stems and water hardness.
This, in conjunction with the second variable suggests that
wate^-iiardness is normally associated with stems of larger
diameter and of lower density as opposed to a larger number
of stems of smaller size. Also important in this relation-
ship is the positive association between water depth and
stem density, indicating that as depth increases stem density
increases as well. Another part of this association is a
positive relationship between stem diameter and pH , phosphates
and alkalinity, indicating that as these variables increase,
the diameter of the stems increase similarly.

The fourth variable shows a negative relationship
between turbidity and internode diameter similar to the
second variable, and a positive relationship between tur-
bidity and internode. length.. The latter relationship would
be expected since a positive relationship has been shown
between turbidity and plant height. Internode length and
plant height are almost certainly related.

Comparison of lakes and streams

To determine if there was statistically significant
difference between alligaterweed growths in lakes and
streams a canonical discriminant analysis was utilized to
compare the 287 samples and group them according to their
origin (Cooley and Lohnes 1970) . Three of the four plant
characteristics were significant between the two groups at
the .01 level of probability. These were: internode

52

length, internode diameter and stem density. The samples
were then weighted according to the relative importance of
these three characteristics and reclassified through the
use of the canonical discriminant function (Tab]c 13).

Table 13. Computer reclassification of Lakes and Streams.

Original data set

Lakes Streams

191 96

Reclassified according to plant characteristics
Lakes Streams

Lakes 67 29 » 70%

Streams 30 161 = 84%

Reclassified according to physical and chemical properties
of the water

Lakes Streams

Lakes 93 3 = 97%

Streams 0 191 = 100%

Using plant characteristics 67 of the 96 lake samples
were classified correctly; 29 were reclassified as stream.
samples thereby producing an accuracy of 70 percent. The
stream samples were reclassified with an 84 percent accuracy
(161 of the 191 stream samples) . Thirty of the stream
samples v;ere reclassified v/rongly as lake samples.

The procedure of reclassification was repeated using
the parameters which measured water quality, all of which

53

were significant at the .01 probability level: PH , alkalinity,
iron, turbidity, chlorides, phosphates, nitrates, hardness,
and water depth. The result was an overall accuracy of 99
percent. Ninety-three of the 9 6 lake samples were reclassi-
fied correctly v/ith only three appearing as stream samples
(97% accuracy) . The stream samples were reclassified v;ith
100 percent accuracy (Table 13) .

The author is cognizant of the fact that with the
small number of sample sites these conparisions may not
represent a true picture of lakes and streams in general;
they are, however, valid for this study.

Analysis of variance of Plant Characteristics

Two analysis of variance tables were constructed for
each of the variables representing plant characteristics;
one which showed the statistical significance of the
dependent variable with all independent variables included
in the model, and the other an analysis of the dependent
variable with only the independent variables representing
the insect populations included in the model.

In each analysis, the sums of squares v/as broken down
using partial sums of squares to shov; how each of the
independent variables contributed to the variation of the
dependent variable. In each case the F test was employed
to determine statistical significance.

As shown in table 18 plant height was significant at
greater than the .01 probability level in both analysis.
The partial sums of squares indicate that iron, turbidity,

54

water depth, and Vogtia were significant at .01 in contribut-
ing to the variation of plant height. Alkalinity was signi-
ficant at the .05 probability mvtjj.. In the second model
which included only the insect -rariables, both Agasicles
and Vogtia were sicrnificnnt at .0]_ (Table 14).

Table 14. ANOV for nlant

Source

df

:^. o

rp.s F Prob F

Regression

12

23.69

1.97 27,

,14 0.0001

Error

274

19.92

0.07

corrected total

286

43.61

Source

DF

Partial SS

F

Prob F

pH

1

0.116

1.59

0.205

Alkalinity

1

C.3u0

4.12

0.040

Iron

1

0.562

7. 74

0.005

Turbidity

1

1.S79

25.83

0.000

Chlorides

1

0.01

0.18

0.677

Phosphates

1

0.03

0.34

0.57

Nitrates

1

0.00 3

0.04

0.84

Hardness

1

0.06

0.81

0.63

Water depth

1

5.38

74.01

0,00

Lake or Stream

1

X

0.04

0.49

0.509

Vogtia

1

0.46

6.33

0.01

Agasicles

1

0.22

3.07

0.08

55

Table 14 continued.

ANOV

for Plant Height with nutrient effects removed

Source

DF SS MS F Prob F

Regression

Error

2 3.93 1.99 14.25 0.0001
284 39.63 0.14

Corrected Total 286 43.61

Source

DF Partial SS F Prob F

Agasicles 1 ' 1.53 10.97 0.0014

Vogtia

1 3.12 22.34 0.0001

The analysis in which internode length was the dependent
variable and with the sane independent variables the analysis
of variance indicated significance at the .01 level (Table 15)
The partial suras of squares shov7ed only one statistically
significant independent variable, Agasicles (.01). In the
insect only model both Agasicles and Voatia were significant
at the .01 probability level.

Table 15. ANOV for internode length

Source

DF SS MS F Prob F

Regression 12 149.96 12.50 9.79 0.0001
Error 274 349.84 1

Corrected total 286 499.80

z.i

56

Table 15 continued.

Source

DF

Partial SS

F

Prob F

PH

1

0.061

0.05

0.821

Alkalinity

1

0.670 ■

0.52

0.524

Iron

1

1.351

1.0 6

0.305

Turbidity

1

0.509

0.40

0.535

Chlorides

1

1.082

0.85

0.639

Phosphates

1

0.696

0.55

0.532

Nitrates

1

0.253

0.20

0.661

Hardness

1

1.659

1.30

0.254

Water depth

1

0.341

0.27

0.612

Lake or Stream

1

0.448

0.35

0.561

Vogtia

1

2.892

2.26

0.129

Agasicles

1

8.304

6.50

0.011

ANOV for Internode

length

with nutrie

nt effects

removed

Source

DF

SS

MS

F

Prob F

Regression 2

Error 284

Corrected Total 286

81.65 40.82 27.73 0.0001
418.15 1.47
499.80

Source

DF

Partial SS

F

Prob F

Agasicles
Vogtia

1
1

11.32

58.05

7.69
39.43

0.0061
0.0001

57

The analysis of variance with internode diameter as
the dependent variable indicated significance at the .01
probability level both v/ith all independent variables
included and v/ith only the independent variables for insects
included. The partial suras of squares with all variables
included showed significance at .0,1 for water depth and
Agasicles. The partial suras of squares with only the insect
variables included resulted in Agasicles not being statis-
tically significant and Vogtia significant at the ,01
probability level (Table 16) . The reversal in significance
^^^ Agasicles is probably due to the correlation of Vogtia
with stem diameter (the simple correlation coefficient
was . 51) . <

Table 16. ANOV for internode diameter

Source

DF

SS

MS

Prob F

Regression 12
Error 274

Corrected Total 286

44.42

3.70

50.84

0.18

95.26

19.95

0.0001

Source

DF

ial SS

F

Prob F

0.001

0.00

0.95

0.144

0.77

0.62

0.040

0.21

0.65

0.001

0.00

0.95

0.033

0.18

0.68

pH 1

Alkalinity 1

Iron 1

Turbidity 1

Chlorides 1

Table 16 cdntinued,

58

Source

DF

Phosphates 1

Nitrates 1

Hardness 1

Water depth 1
Lake and Stream 1

Vogtia 1

Aqasicles 1

tial SS

F

Prob F

0.661

3.56

0.06

0.442 •

2.38

0.12

0.540

2.91

0.08

1.708

9.20

0.00

0.523

2.82

0.09

0.211

1.14

0.29

0.950

5.12

0.02

ANOV for internode dianeter with nutrient effects removed

Source

DF

SS

MS

Prob F

Regression

2

15.76 . 7.88

Error

284

79.49 0.28

Corrected

286

95.26

Total

Source

DF

Partial SS

28.16

0.0001

Prob F

Agasicles
Vogtia

1
1

0.592
13.58

2.11
48.53

0.1429
0.0001

Table 17 shows the analysis of variance using stem
density as the independent variable with the same independ-
ent variables previously listed. Stem density was significant

59

at the .01 probability level in both analysis. The partial
sums of squares shows significance at .01 for the following
variables: pH, alkalinity, nitrates, hardness, and water
depths' Significance at .05 was indicated for phosphates,
lake or stream variable, and Vogtia, The partial sums of
squares with only insect variables included shov/ed signifi-
cance at .01 for Vogtia and no significance for Agasicles
(Table 17) .

Table 17. ANOV for Stem density,

Source

DF

SS

MS

F

Prob F

Regression

12

384

.94

32,

,08

8.9 7

0.0001

Error

274

980

.08

3.

,58

Corrected
Total

286

1365

.02

Source

DF

Partial

SS

F

Prob F

pH

1

40.77

11.40

0.00

Alkalinity

1

27.51

7.69

0.01

Iron

1

6.66

1.86

0.17

Turbidity

1

7.06

1.97

0.16

Chlorides

1

0.10

0.03

0.86

Phosphates

1

14.88

4.16

0.04

Nitrates

1

•

64.42

18.01

0.00

Hardness

1

71.90

20.10

0.00

Water depth

1

53.29

14.90

0.00

Lake or Stream 1

14.43

4.03

0.04

Table 17 CQntinued,

60

Source

DF

Partial SS

F

Prob F

Vogtia
Agasicles

1
1

18.72
5.21

5.2 3
1.46

0.02
0.23

ANOV for Stein density with nutrient effects removed.

source

DF

SS

MS

Prob F

Regression 2 141.05
Error 284 1223..97

Corrected total 286 1365.02

70.52 16.36 0.0001
4.31

Source

DF

Partial SS

Prob F

Agasicles
Vogtia

1
1

2.777
126.53

0.64
29.38

0.5710
0.0001

The last ANOV table was constructed for the variable
representing leaf area missing, in which case this was the
dependent variable. Leaf area was significant at .01 in
both analyses, i.e. in both the limited model and when all
independent variables were included in the model (Table 18) .
The partial sums of squares indicated that the following
independent variables were significant sources of variation
at the .01 probability level: Alkalinity, iron, nitrates,
Vogtia and Agasicles . Turbidity and phosphates were

61

similarly significant but at the .05 level. The partial
sums of squares in the limited model shows both Agasicles
and Vogtia to be significant at the .01 probability level,

Table 18. ANOV for Leaf area missinq.

Source

DF

S;

S

MS

F

Prob F

Regression

12

616

.05

51,

.34

10.29

0.0001

Error

274

1366

.48

4,

.99

Corrected

286

1982

.53

Total

Source

DF

Partial

SS

F

Prob F

pH

1

0.54

0.11

0.74

Alkalinity

1

33.52

6.72

0.01

Iron

1

134.37

26.94

0.00

Turbidity

1

23.71

4.76

0.03

Chlorides

1

2.42

0.48

0.51

Phosphates

1

25.86

5.18

0.02

Nitrates

1

34.72

6.96

0.01

Hardness

1

0.05

0.01

0.92

Water depth

1

9.70

1.94

0.16

Lake or Stream

1

*1.18

0.23

0.63

Vogtia

1

36,86

7.39

0.01

Agasicles

1 •

91.99

18.44

0.00

ANOV for leaf damage with nutrient effects removed

62

Table 18 continued.

Source DF SS MS F Prob F

Regression 2 343.94 171.97 29.81 0.001

Error 284 1638.59 5.77

Corrected Total 286 1982.53

Source DF Partial SS F Prob F

Agasicles 1 288.82 50.06 0.0001

Vogtia 1 108.97 18.82 0.0001

In table 19 the overall effect of each of the independent
variables on the plant characteristics as a whole is shown.
Like the previous programs, F values v/ere calculated using
all independent variables in the first model and secondly
with only the insect variables included. Pillai's Trace
with the F statistic was used as a test for significance.
When all variables v;ere included only one independent
variable lacked significance, chlorides. Variables with
larger F values and therefore more important sources of
variation were: Iron, turbidity, nitrates, hardness, water
depth, and Agasicles. With only the insect variables
included, both Agasicles and Vogtia were significant at
greater than the .01 probability level.

63

Table 19. Overall effect cf variables on alligatorweed

Variable

DF

F

Prob F

PH

5

and

270

3.15

0.0091

Alkalinity

5

and

270

3.07

0.0104

Iron

5

and

270

6.42

0.0001

Turbidity

5

and

270

10.54

0.0001

Chlorides

5

and

270

0.38

0.8624

Phosphates

5

and

270

4.76

0.0006

Nitrates

5

and

270

8.44

0.0001

Hardness

5

and

270

5.92

0.0001

Water depth

5

and

270 '

26.91

0.0001

Lake or Stream

5

and

270

3.98

0.0020

Agasicles

5

and

270

8.89

0.0001

Vogtia

5

and

270

3.57

0.0042

Overall effect of insects with nutrient effects removed

Variable

DF

F

Prob F

Agasicles
Vogtia

5 and 280
5 and 280

16.19
15.08

0.0001

0.0001

SUMMARY

Vogtia malloi was introduced into the United States in
the spring of 1971 with populations successfully established
in 1971 and 1972 at the following locations: Upper Legion
Lake, Columbia, South Carolina, Twin Lakes, South Carolina,
Garden's Corner, South Carolina, Black Lake, Melrose, Florida,
Lake Alice, Gainesville, Florida, Lake Lawne, Orlando, Florida,
and a roadside ditch, Fort Lauderdale, Florida.

Populations of Vogtia expanded randomly as alligatorweed
in the release area became damaged and less desireable.
Samples of the parasite populations at Lake Alice and Lake
Lawne indicated that the Vogtia populations were not seriously
reduced and only general parasites were observed.

Both Vogtia and Agasicles demonstrated their ability for
surviving frost and cold temperatures. At Columbia, South
Carolina both insects survived the 1972-73 VJinter with below
freezing temperatures recorded for 17 of the 2 8 days in the
month of February. With this degree of cold tolerance it
is probable that these insects can extend their range through-
out the northern limits of alligatorv/eed infestations.

At Gardenfe Coner the Vogtia population was reduced but
was not excluded even when Agasicles had completely defoliated

64

65

the alligatorweed and almost destroyed it. When the Agasicles
population subsided, as expected during the hot nonths of
summer, the Vogtia population began to build up, keeping the
alligatorweed under stress for the entire growing season.

While Agasicles populations remained low in south Florida,
presumably because of high temperatures, the Vogtia population
was apparently unaffected, and will therefore probably surpass
Agasicles in its influence on alligatorweed grov/th in the
southern regions.

The use of canonical correlations, in conjunction with
a multivariate analysis has enabled us first to shov/ the
interrelationships which occured between Vogtia and Agasicles ,
between each of these insects and alligatorv/eed, and between
the physical and chemical characteristics of the v;ater and
alligatorweed, and secondly, to determine the significance
of these factors in controlling alligatorweed growth.

As expected, a significant relationship existed between
Agasicles and plant height. Agasicles characteristically
begins feeding on the upper leaves and items of alligator-
weed resulting in shorter plants. Similarly, Agasicles adults
occured less frequently in heavily damaged alligatorweed, as
did Vogtia larvae, a probable result of the food supply being
reduced.

In this study, as stem density increased Vogtia occurred
less frequently and Agasicles occurred more frequently. Since
alligatorweed may occur in relatively dense stands of stems

66

with small diameters, this indicates the preference of
Agasicles for the younger leaves, and with the positive
correlation between Vogtia larvae and stem size, the in-
ability of Vogtia larvae to survive in very small stems.
This is also indicated by the tendency of both populations
to occur in greater numbers with tall healthy plants.

As the depth of the water increased, aerial stems
tended to be shorter and more numerous, and as turbidity,
iron, and alkalinity increased plant height increased.
Also, as water hardness, pH , and phosphates increased the
plants tended to be larger. This simply indicates that
alligator^veed grows better in shallow, fertile ponds as
opposed to deep, clean water. <

Samples from lakes and streams were more accurately
seperated according to characteristics of the water than
to plant characteristics, suggesting that a greater differ-
ence occurs between lakes and streams than between stands
of alligatorweed growing in these areas.

The overall effect of the independent variables on
the plant characteristics as a whole resulted in all
variables, except chlorides, being significant at 0.01
The most important of these were: iron, turbidity, nitrates,
hardness, water depths, and Agasicles , as indicated by
larger F values.

The potential for seasonal control of alligatorweed
has been clearly demonstrated by both Agasicles alone
(Zeiger 1967) , by Vogtia, and by Vogtia and Agasicles in

67

combinations. However, the resiliency of alligatorweed in
re-establishing itself is truly remarkable, and for this
reason no attempt has been made to predict the final
impact of these insects in reducing the overall area of
infestation of alligatorv/eed.

In many areas, the mats of alligatorweed have increased
over the years such that they now frequently occur 8-10 in.
in depth. Under these conditions, herbicides may be required
to facilitate insect presure if alligatorweed is to be reduced
drastically in a short period of time.

LITEFi; CITED

Acreneaux, G. and L. P. Herbert. 1943. Preliminary studies
of periodic flaming as a means of controlling Johnson
grass and alligatorweed on sugarcane lands. The Sugar
Bull. 22: 21-23.

Brown, J. L. ana N. R. Spencer. 19 73. Vogtia malloi, a

newly introduced Phycitine moth (Lepidoptera: Pyralidae)
to control alligator^-zeed. J. Environ. Entomol. 2(4):
519-523.

Cooley, W. W. and P. R. Lohnes . 1970. Multivariate Data
Analysis. John Wiley and Sons. 359d.

Cunha, T. J. and G. N. Rhodes. 1965. Beef cattle in

Florida. Bulletin 28. Florida Department of /igricul-
ture.

Gangstad, E. 0. and F. J. Guscio. 1970. Research planning
conferences for aquatic plant control. Tech. Rep.,
Chief of Engineers, U.S. Army Corps of Engineers.

Maddox, D. M. and R. D. Hennessey. 19 70. The biology and
host range of Vogbia malloi Pastrana. Unpublished
manuscript, Agricultural Research Service, USDA, Albany,
California.

Maddox, D. Zl. , L. A. Andres, R. D. Hennessey, R. D. Blackburn,
and M. R, Spencer. IP" nsects to control alligator-
weed, an xnvaaer of aquatic ecosystems in the United

CQ

69

States. Bioscience 21: 985-91.
Mohr, C. 1901. Plant Life of Alabama. Vol. 6, 921p.
Morrison, D. F. 1967. Multivariate statistical nethods.

McGraw-Hill Co. Nev; York. 338p.
Muesebeck, C. F. W., K. V. Krombein, and H. K. Townes .

19 51. Hymenoptera of Anerica North of Mexico. U.S.

Dep. Agric, Agric. Mono. no. 2. 142 0 pp. U. S. Govt.

Printing Office, Washington, D. C.
Nie, N. 1970, Stastical package for the Social Sciences.

McGraw-Hill Co. New York. 338p.
Pastrana, J. A. 1961. Una nueva Phycitidae (Lep.) parasito

de la "Lagunilla". Rep'. Arg. Publ. Tec. 71, Inst. Nac.

Tec. Agropecuar., p 265-272.
Penfound, W. T. 1940. The biology of Achyranthes

philoxeroides (Mart.) Standley. American Midi. Nat.

24: 248-252.
Vogt, G. B. 1960. Exploration for natural enemies of

alligatorweed and related plants in South America.

U. S. Dept. Agric, Agric. Res. Serv. , Entomol. Res.

Div. , Special Report PI-4, 58p.
Vogt, G. B. 1961. Exploration for natural enemies of

alligatorweed and related plants in South America.

U. S. Dep. Agric, Agric. Res. Serv., Entomol. Res.

Div., Special Report PI-5, 50p.
Weldon, L. W. 19 60. A summary review of investigations

on alligatorweed and its control. USDA, Agric. Res.

Serv., Crops Res. Div., CR 33-60, 41p.

70

Zeiger, C. F. 19 G7. Biological control of alligatorweed
with Agasicles n. sp. in Florida. Hyacinth Control
J. 6: 31-34.

BIOGRAPHICAL SKETCH

John L. Brown was born November 17, 19 44 in Weiner,
Arkansas. He attended Weiner Public Schools from 1950
to 1962 and after graduation he enlisted in the U.S. Air
Force. In 19 67 he enrolled in the school of Agriculture
at Arkansas State University. He received the Bachelor of
Science degree in 1969, and enrolled at Iowa State Univer-
sity. He received the Master of Science degree in entomo-
logy in 19 71 and transferred his studies to the University
of Florida where he is completing the requirements for the
degree of Doctor of Philosophy.

John L. Brown is married to the former Delena Y.
Newmans and has two children, a son of 10 years and a
daughter of 5. • •■ "

He is a member of the Entomological Society of America
and the Florida Anti-mosquito Association.

He was employed by the Florida Division of Health in
January, 1973 as director of the stable fly control program.

71

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.

.-^iM-^^^^i

B. David Perkins
Assistant Professor of
Entomology

This dissertation was submitted to the Dean of the College
of Agriculture and to the Graduate Council, and was accepted
as partial fulfillment of the requirements for the degree of
Doctor of Philosophy.

December, 19 73

Dean, Graduate School