Influence of construction characteristics and home maintenance practice on subterranean termite infestation rates in Nor...


Material Information

Influence of construction characteristics and home maintenance practice on subterranean termite infestation rates in Northeastern Florida
Physical Description:
vii, 180 leaves : ill. ; 29 cm.
Richman, Dina Lynn, 1971-
Publication Date:


Subjects / Keywords:
Entomology and Nematology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Entomology and Nematology -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 2003.
Includes bibliographical references.
Statement of Responsibility:
by Dina L. Richman.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030456642
oclc - 81047160
System ID:

This item is only available as the following downloads:

Full Text






Copyright 2003


Dina L. Richman


The support of my loving family throughout my graduate schooling was

invaluable. I thank Steven for going through this degree with me and I thank Roxanne for

all her help in the end. I also express deep appreciation to my advisor, Dr. Philip Koehler,

for seeing me through the process. The thoughtful critique of my work by my dissertation

committee was greatly appreciated. I also appreciated the support of Dow AgroSciences

for a portion of my research.



ACKNOWLEDGMENTS .............................. ............... iii

ABSTRACT ................. ......... ......................... vi


1 INTRODUCTION ............................................1




INFESTATIONS .............................................. 117

6 SUMMARY AND CONCLUSIONS ............................... 131


A SERIAL DILUTION DERIVATIONS ............................... 134

B LABEL RATE CALCULATION .................. ................ ..136

THE EFFECT OF TIME ON CONTROL SOIL pH .................... 138

TREATED WITH TERMITICIDES ................................. 139



HOMEOWNERS .............................................156

NORTHEAST FLORIDA ....................................... 158

REFERENCES ....................................... .............. 160

BIOGRAPHICAL SKETCH ............................................ 180

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Dina L. Richman

August 2003

Chairperson: Philip G. Koehler
Major Department: Department of Entomology and Nematology

This research determined the effects of construction practices and home

maintenance on subterranean termite infestations in two phases. The first phase

investigated the effect of cement contamination on residual performance of termiticides in

the laboratory. The second phase examined the influence of home building and

maintenance practices on infestation rates of houses in three northeastern Florida


First, a laboratory bioassay with Reticulitermesflavipes (Kollar) was conducted 5

days after soil was mixed with portland cement to alter pH levels to 6, 7, 8, and 9.

Termites were confined to soil treated with 10x, Ix, 0.lx, or 0.01x the manufacturer's

suggested application rates for imidacloprid, fipronil, chlorpyrifos, bifenthrin, permethrin,

or cypermethrin. Results of additional bioassays at 5 and 10 months indicated that

alkaline soil pH significantly decreased residual efficacy of imidacloprid and fipronil at

0.1x label rates, and chlorpyrifos, bifenthrin, and permethrin at 0.01x the label rates.

Cypermethrin was not affected by soil pH.

Second, homeowner responses to a mailed survey allowed classification of

single-family houses built between 1994 and 1998 in Jacksonville Beach, St. Johns

County, and Flagler County according to their subterranean termite infestation history,

construction type, and maintenance characteristics. Pest control companies and county

records provided preconstruction soil chemical treatment information. Infestation rates

were 17.60% from Jacksonville Beach, 15.85% from St. Johns County, and 2.70% from

Flagler County. Subterranean termite infestations were strongly associated with older

houses with wood frames. Conditional logistic regression of covariates showed

relationships among factors known to be conducive to subterranean termite infestations in

order of their influence on infestation likelihood were political boundary > occurrence of

structural wood > repellent preconstruction termiticide treatment > termite access to

structure via foraging guidelines and lack of foundation perimeter inspection space.

I also evaluated the effect of the St. Johns County Building Code on infestation

rates of houses built after addition of a Termite Protection Ordinance. The 1996

Ordinance was enacted in order to decrease subterranean termite infestation rates. Poor

implementation of the Ordinance made it impossible to fully evaluate its impact.


Termites decompose organic matter (Collins 1989, Donovan et al. 2001) and help

maintain nitrogen and carbon cycles (Waller et al. 1989, Tayasu et al. 1997), and are

thought to be a key contributor to soil heterogeneity (Donovan et al. 2001). Their

diversity is influenced by habitat, with the highest number of species found in wet

lowland tropical forests (Eggleton 2000). Southern latitudes have more termite species

and endemism than their northern counterparts (Eggleton 1994). Eggleton (2000)

hypothesized that this trend is due to protection of southern regions from the effects of

glaciation. He also suggested that Isoptera originated in the Cretaceous Period (Mesozoic

Era, -144 mya) on Pangea, and that the subsequent evolution of groups was due to

continental separations (Eggleton 2000).

Within North America, the most speciose regions are Florida, northern Mexico

and the southwestern United States (Weesner 1965). Termites inhabiting these areas

include genera derived from Central and South Americas. According to Thome et al.

(1993), the only North American endemic is Zootermopsis Emerson. However,

Reticulitermes has the widest overall distribution in the region (Weesner 1965).

Within the United States, the genus Reticulitermes includes six species of

economic impact (Su and Scheffrahn 1990a). Reticulitermes arenincola Goellner can be

found from the Midwestern through the Northeastern U.S. and is considered to be of

small economic impact compared to other species (Edwards and Mill 1986).

Reticulitermes hesperus Banks represents the major termite pest in the western United

States (Kofoid 1934, Snyder 1954, Weesner 1965, Edwards and Mill 1986).

Reticulitermes tibialis Banks and R. hageni Banks, both found in the midwestern through

eastern United States, are also known to cause structural damage but are reported less

frequently (Kofoid 1934, Edwards and Mill 1986, Su and Scheffrahn 1990a). Both R.

flavipes (Kollar) and R. virginicus Banks are found in most of the U.S., except the

northwestern states (Su and Scheffrahn 1990a).

Reticulitermesflavipes is probably the most economically important termite in the

southeastern states (Su and Scheffrahn 1990a). The economic impact of R. virginicus,

however, is unknown. Reticulitermes virginicus infestations were misidentified as those

of R. flavipes because of morphological similarities between the two species (Su and

Scheffrahn 1990a). In cooperation with pest management professionals, data are being

collected to verify the economic impact of R. virginicus. Reports of R. virginicus

infestations are increasing in number, especially in Florida (Su and Scheffrahn 1990a),

suggesting R. virginicus may be as significant a pest as R. flavipes.

Biology of R. flavipes and R. virginicus

The new Reticulitermes colony is founded by the royal pair after their nuptial

flight, in which the fertile primary reproductive (alates) leave their colony seeking to

establish new colonies. In Florida, the eastern subterranean termite (R. flavipes) swarms

January to April, while the dark southern subterranean termite (R. virginicus) swarms

March to May (Scheffrahn and Su 1994). They both swarm during midday in sunshine,

and may also swarm during fall months (Scheffrahn and Su 1994).

The number of individuals within the termite colony varies (Thome et al. 1997).

Under laboratory conditions and without the intrusion of periodic sampling by

researchers, the mean size of 2-year old R. flavipes colonies bred from royal pairs was

387 226 and ranged from 51 to 984 individuals, excluding eggs (Thome et al. 1997).

Direct excavation of colonies has led to population estimates of 51,505 to 363,512

(Howard et al. 1982). Using mark-release-recapture techniques total Reticulitermes

colony populations have been estimated at 25,000 to 5 million (Esenther 1980a, Grace et

al. 1989, Grace 1990, Su et al. 1993a).

Primary Reticulitermes reproductive are rarely found in the field, and fecundity

of the founding female decreases over time (Thome et al. 1999). Brachypterous and

apterous neotenic (second and third forms, respectively) reproductive are more

frequently encountered than primary reproductive in the field (Snyder 1954). Both

supplemental types contribute to the total egg production within the colony (Pickens

1934, Snyder 1954, Buchli 1958, Pawson and Gold 1996, Thorne 1996). Neotenic

reproductive (also referred to as "supplemental reproductives) can differentiate within

their colony while the royal pair is still alive and reproducing (Snyder 1920, Noirot 1989).

Individual female neotenic reproductive are less fecund than alate-derived queens

(Snyder 1920). However, hundreds of these supplemental reproductive may occur within

the colony, so their cumulative egg contribution can exceed that of the primary queen

(Howard and Haverty 1980, Myles and Nutting 1988, Thorne 1996).

Reticulitermes colony formation may also occur through budding, although

budding does not always produce new colonies (Thorne et al. 1999). Thome (1999)

defined budding as occurring when satellite colony units form as aggregates of nestmate

groups that spend time separated from the primary reproductive. Supplemental

reproductive may differentiate within the new units (Thome 1999). Budding has been

reported in both R.flavipes and R. virginicus to occur either actively, when a critical mass

is reached (Snyder 1948); or passively, due to some physical separation (Pawson and

Gold 1996). When small groups of Reticulitermes nymphs were isolated from their royal

pairs, brachypterous and apterous neotenic reproductive formed within 4 (Pawson and

Gold 1996) and 10 months (Pickens 1934), respectively. Pawson and Gold (1996)

reported that R. flavipes and R. virginicus supplemental reproductive produced viable

progeny (eggs and nymphs) within 5 and 7 months, respectively.

Colonies of Reticulitermes have no permanent, central nest site. Thome et al.

(1999) described them as "mobile and amoeboid rather than sessile." Snyder (1936)

hypothesized that the entire colony may relocate or bud in response to temperature,

moisture, resource conditions, season, or stage of colony growth.

Reticulitermes and all other termites are considered to be eusocial. They have

overlapping generations within their colonies, colony members cooperatively care for

brood, and they have a division of labor that allows members to individually contribute to

a productive, adaptable society. Their division of labor constitutes a social structure

consisting of three castes: reproductive (discussed above), workers, and soldiers.

Workers feed the other castes, forage for resources, and participate in colony defense.

Soldiers are a defensive caste and comprise about 1-2% of R.flavipes (Howard and

Haverty 1980, Thorne et al. 1997) and R. virginicus (Pawson and Gold 1996) colonies.

Reticulitermes flavipes and R. vireinicus Feeding and Tunneling

Subterranean termites build underground tunnel networks which provide them

protection from predation and dessication while foraging. Tunnel construction is

influenced by the number of tunnels present, physical barriers, direction, and soil

moisture. Subterranean termites tended to tunnel in directions that evenly divide their

search area (Robson et al. 1995, Powell 2000, Campora and Grace 2001, Puche and Su

2001). They readily used pre-formed tunnels. Kirton et al. (1998) observed subterranean

termites using tunnels in corrugated cardboard. Pitts-Singer and Forschler (2000)

speculated termites would readily use tunnels made by other colonies or those formed by

decomposed roots or earthworms. Tests with wires placed in soil-filled arenas resulted in

R. flavipes following 47% of the encountered wires and R. virginicus following 42%

(Pitts-Singer and Forschler 2000). Reticulitermes prefer to construct tunnels directed

downward or horizontal versus upward (Pitts-Singer and Forschler 2000). In laboratory

tests, R. flavipes tunneled faster in sand and clay with moderate moisture (by weight) of

10 to 25% or 20 to 30%, respectively, compared to soils with lower or higher moisture

content (Gahlhoff 1999). While some research indicates wood volatiles act as kairomones

(chemical emitted by one species and perceived by another species which benefits; Gordh

and Headrick 2001) that guide subterranean termites to food placed in loose substrates or

open-air settings (Clement et al. 1988, Reinhard et al. 1997), Puche and Su (2001) found

no indication that R. flavipes workers were able to detect wood in sand over distance.

Consumption of a substrate by a subterranean termite colony is affected by colony

population density (Esenther 1980b, Lenz and Barrett 1984), vigor (Su and LaFage 1984,

Lenz 1985), substrate composition (Smythe and Carter 1969, Behr et al. 1972, Carter and

Dell 1981, Carter and Huffman 1982, Duryea et al. 1999), previous damage to substrate

by conspecifics (Delaplane and La Fage 1989b), temperature (Haverty and Nutting 1974),

and moisture (Delaplane and La Fage 1989a). Although Reticulitermes have been shown

to prefer rotted wood over sound wood (Schultze-Dewitz 1972, Amburgey 1979, Waller

et al. 1987), there is some uncertainty regarding termite preference for certain fungal

species, such as the white-rot genus, Ganoderma (Amburgey and Beal 1977, Waller et al.


Amount of wood consumed by individual termites is affected by placement and

type (Oi et al. 1996), volume (Waller and LaFage 1987, Hedlund and Henderson 1999,

Wang and Powell 2001), and size (Gentry and Whitford 1982, Waller 1988, Wang and

Powell 2001). Consumption incidence and quantity by individual termites is

unpredictable (Forschler 1996).

Researchers have assessed the minimum distance traveled by tunneling and

foraging termites with mark-release-recapture and mapping studies. Workers from a

single R. flavipes colony explored resources over a linear distance of up to 79 m and over

an area of up to 2,361 m2 (Oi 1994, Grace et al. 1989, Grace 1990, Suet al. 1993a).

Although habitat (undeveloped vs. residential) was not correlated with colony population

size (Su et al. 1993a), population size and territory seem to differ among Reticulitermes

found in different geographic locations. For instance, foraging populations and territories

from South Florida (Su et al.1993a) were larger than those described in Gainesville,

Florida (Oi 1994).

The Evolution of Subterranean Termite-Resistant Construction

Essentially, subterranean termite control is the process of preventing termites from

feeding on any cellulose-containing material in a structure. Understanding the

relationship between subterranean termite biology and construction and landscaping

practices is relevant to prevention and control efforts.

Wood Preservatives With Emphasis on Borate Treatments

One of the earliest records of preventative subterranean termite control dates back

to 1756 when creosote was first reported as a wood preservative (Rambo 1997). It

became widely used in 1865 when a pressure-treatment plant was built in Massachusetts.

However, creosote use decreased because it is explosive, causes skin and eye irritation,

smells bad, and resists paint. Commercial wood preservatives used after creosote use

ended included Bruce preservative (naphthol), sodium fluoride based Wolman salts, zinc

meta-arsenite (including chromated copper arsenate), pentachlorophenol, copper

naphthenates, cuprinol products, and borates (Rambo 1997). With the exception of

borates, these wood preservatives are no longer used for interior wood because of

environmental and human health concerns .

Borate wood preservatives protect wood from decay fungi and wood-destroying

insects (Williams and Amburgey 1987, Williams et al. 1990, Su and Scheffrahn 1991).

Borate treatments are currently applied to structures for preventive and remedial control

of termites. Toxicity to termites is probably due to interference with the digestive process

(Williams et al. 1990). Subterranean termites feeding on wood containing 100 ppm boron

died within 2-4 weeks (Jones 1991, Grace 1992, Williams 1997). However, >2,500 ppm

boric acid equivalent was needed to deter both C. formosanus and R. flavipes from

feeding on wood treated with sprayed on Bora-Care (Nissus Corp.; Su and Scheffrahn


Surface-applied borates in wood exhibit moisture-dependent mobility after the

first week of treatment (Schoeman et al. 1998). Significant wood diffusion was noted one

week post-treatment of DOT in wood having a moisture content >15%. Moisture content

of the wood was more important than wood species (Schoeman et al. 1998). With a

moisture level of<10%, borates will be deposited 3-10 mm into wood (Williams 1997).

DOT sprayed on the surface of southern yellow pine or Douglas fir having moisture

content of 10-15% will penetrate 6-12 mm into the wood (NPCA 1992).

Borate solutions are prepared in monoethylene glycol or water, to facilitate

diffusion of sprayed-on borate products through dry wood (Becker 1976). (This treatment

allows thorough penetration into new structural lumber, which has a normal moisture

content of 16-18% (Levi 1986)). For remedial control, subterranean termites may enhance

the diffusion rates of borate products in infested wood, since they transfer water to their

feeding sites (Grube and Rudolph 1999). Bora-care spray-on treatments protected pine

(Pinus spp.) from both C. formosanus and R. flavipes when wood contained > 2,500 ppm

boric acid equivalent (Su and Scheffrahn 1991). Pine and spruce are both commonly used

in Florida construction (pers. comm., H. T. White, Deputy Chief Building Inspector, St.

Johns County, Florida).

Unpalatable Wood

Subterranean termites have shown preferences for certain wood species over

others in laboratory tests (Smythe and Carter 1969, 1970, Behr et al. 1972, Mannesmann

1972, Carter and Smythe 1974, Esenther 1977, Carter 1979,Carter and Dell 1981). For


example, both R. flavipes and R. virginicus prefer pine species over melaleuca (Duryea et

al. 1999) and cypress (Smythe and Carter 1969, Carter and Huffman 1982). In general,

decayed and/or moist wood is preferred over sound, dry wood (Smythe et al. 1971,

Schultze-Dewitz 1972, Amburgey 1979, Waller et al. 1987, Delaplane and LaFage

1989a) and softwood is preferred over hardwood (Behr et al. 1972).

Soil Treatments

In the 1920s and early 30s, soil treatments were used frequently in southern

California as remedial treatments for subterranean termite infestations. The most common

treatment was a 10% sodium arsenite solution applied to the ground or in a trench dug

under or around infested houses at the rate of 1 gal per 100 ft2 (Randall and Doody 1934).

Although laboratory studies showed topically applied sodium arsenite was toxic to

subterranean termites, soil treatments had high failure rates (Randall and Doody 1934). In

response to these failures, 2 to 6% solutions applied at 10 to 50 gal per 100 ft2 were

evaluated. Control recommendations then changed to using large amounts of dilute

solutions to obtain a thicker layer of treated soil around structures (Randall and Doody

1934). The posting a conspicuous permanent warning sign stating the ground had been

poisoned by sodium arsenite was also recommended (Randall and Doody 1934). One

report strongly cautioned against the use of any arsenical soil treatments due to their high

toxicity (Randall and Doody 1934).

The 1940s saw the emergence of soil termiticides such as the chlorinated

hydrocarbon DDT (dichloro-diphenyl-trichloroethane) and the cyclodienes chlordane,

aldrin, heptachlor, and dieldrin. DDT at 2% in sandy loam persisted at least 33 years,

killing 61% of the C. formosanus in bioassays with 24-year old soil (Grace et al. 1993).


Cyclodienes persisted over 35 years (Kard et al. 1989). Dieldrin in sandy loam at 0.30%

killed 99% of the C. formosanus, even after the soil had been exposed to the environment

for 33 years (Grace et al. 1993). While persistence was a desirable trait, it caused

environmental concern and eventually led to the withdrawal of all cyclodiene termiticides

from the U.S. market by 1988.

Less persistent termiticides, such as chlorpyrifos and the pyrethroids, are currently

used by pest management professionals. Chlorpyrifos and the pyrethroids cypermethrin,

permethrin, and fenvalerate were 100% effective against subterranean termites for 9 and 6

years, respectively, in Florida soil (Kard 2001). Chlorpyrifos use is currently being phased

out due to the concerns arising from the Food Quality Protection Act of 1996. Pyrethroids

repel termites from treated areas without causing significant mortality (Su et al. 1993b).

The newer, slow-acting toxicants are nonrepellent and include imidacloprid (Kuriachan

and Gold 1998) and fipronil (Osbrink et al. 2001). In field tests, both have provided at

least 5 years of 100% protection of wood within concrete slabs on Florida soil (Kard


Soil termiticides protect structures by either repelling or killing termites that enter

the treated area. In studies with C. formosanus, Su et al. (1982) defined three types of

termiticides: type I was instantly repellent, type II was slowly repellent, and type 1I was

nonrepellent. All currently available pyrethroids are considered to be type I termiticides

(Su et al. 1982, Su et al. 1995). At termiticidal application rates, type I termiticide

treatments will result in low mortality due to their repellent effect.

Chlorpyrifos is acutely toxic to subterranean termites and was grouped with the

type I termiticides (Su et al. 1982). Su et al. (1982) observed termites entering the

chlorpyrifos-treated area for the first few days. They reported that avoidance of the

treatment did not occur until "there were many dead and decaying individuals in the

treated area," and speculated that secondary repellency occurred due to the decomposing

bodies of dead termites within the treated area (Su et al. 1982, Su and Scheffrahn 1990b).

At termicidal application rates, type II termiticide treatments will result in initial mortality

of individuals entering the treated area, and subsequent low mortality when termites begin

to avoid the treatment.

Imidacloprid, fipronil, and chlorfenapyr are considered to be slower-acting than

both pyrethroids and chlorpyrifos and are nonrepellent (Kuriachan and Gold 1998,

Osbrink et al. 2001, Wagner et al. 2003). At termiticidal application rates, these soil

treatments should result in high mortality of exposed individuals away from the treated

area. Slow-acting imidacloprid prompted the USDA-Forest Service to make the first

protocol change. Before imidacloprid field tests, wood plots either passed or failed

according to whether or not damage was noted. Imidacloprid soil treatments resulted in

delayed mortality of termites, which allowed them to cause some surface etching of wood

before death. Once the delayed effect of imidacloprid was detected, plots no longer

strictly either passed or failed. Instead, wood in plots were rated on a scale similar to that

of the American Society for Testing Materials (ASTM), which allowed for acceptable

surface etching (Wagner et al. 2003). A second protocol change occurred after some field

experience with fipronil soil treatments caused suspicion of the treatment being spread

from termite to termite. Before fipronil, control plots were placed near treatment plots in

tests. Due to lack of activity in control plots after 3 to 5 years, fipronil tests were redone

with control and treatment plots separated (Kard 2001).

Population Control With Baits

Currently registered baits contain either hexaflumuron, diflubenzuron, or

sulfluramid. The most extensively studied active ingredient, hexaflumuron, has been

reported to eliminate subterranean termite activity at several locations with several

species (Su 1994, Su and Scheffrahn 1996, Peters and Fitzgerald 1999, Sajap et al. 2000,

Rojas and Morales-Ramos 2001, Grace and Su 2001). Researchers found no differences

in C. formosanus consumption rates of bait matrices containing 250 ppm diflubenzuron,

hexaflumuron, or chlorfluazuron, and all laboratory colonies died within 9 weeks (Rojas

and Ramos 2001). Foraging populations of C. formosanus were reduced 65 to 98% within

one year by baits containing 600 ppm sulfluramid (Su et al. 1991a).

Principles of Termite Prevention

Soil treatments were first used for remedial rather than preventive measures.

Prevention recommendations were later developed and consisted of three basic principles:

1) structures should be inaccessible to termites, 2) structural wood should be unpalatable

to termites, and 3) termite foraging in the vicinity of the structure should be discouraged

(Brown et al. 1934). These principles are part of the National Pest Control Association's

(NPCA) Approved Reference Procedures for Subterranean Termite Control, the St. Johns

County Termite Protection Ordinance, and the current Florida Building Code (FBC)

(NPCA 1980, SBCCI 1994 and 1997, FBC 2002).

The NPCA identified 50 basic elements of construction which must be considered

in termite control in their Approved Reference Procedures for Subterranean Termite

Control (1980). These included areas where termites could enter a structure unseen

through cracks, such as brick veneer below grade on a frame house or through a concrete

foundation. NPCA (1980) also identified construction elements in which wood is in close

proximity to the ground, such as supported slabs of basementless wood frame houses and

earth filled wooden porches, among their list of construction elements. For each of the 50

cases, the NPCA identified the condition that allowed for subterranean termite infestation

and the difficulty in treating the area. For example, for floating or supported slab floor

construction of wood frame houses, termites may gain access through the block voids of

the foundation, through the space between the slab and a block, or through a space

between the slab and exterior cladding. NPCA warned of the hazard in treating this type

of construction due to heating tubes or service lines in the slab at unknown locations.

Additionally, insulation strips or forgotten grade stakes may be hidden from view by a

sole plate or the slab itself, thereby preventing its identification as an infestation channel.

Regarding the first principle, the Termite Investigations Committee recommended

that wood not contact the ground, untreated wood above the ground be supported by

either concrete or wood that has been chemically treated, voids in masonry units be

avoided and all cracks be filled with cement, ample ventilation be provided within the

substructure, adequate soil drainage be provided beneath and around the structure, and a

metal shield be provided as a barrier to runways under structural wood (Brown et al.

1934). Use of metal termite shields has since been reclassified as an inspection aid rather

than a control measure, since termites will build tubes over the shields (Su and

Scheffrahn 1990a, Lewis 1997).

Physical barriers that can be placed around foundation supports and building

cavities (thereby making the structure inaccessible to subterranean termites) and cannot

be crushed under the weight of slabs include granite or basaltic rock particles, wire mesh,

and insecticide-impregnated plastic. Rock particles consist of either crushed gravel or

sand (Grace and Yamamoto 1993, Tamashiro et al. 1987, Ahmed and French 1996, Lewis

et al. 1996). Su and Scheffrahn (1992) found that R.flavipes could not tunnel through

rock particles ranging from 1.18 to 2.80 mm (effective particle size will vary with termite

species). Rocks this size prohibited interstitial termite movement and are too big for the

termites to manipulate with their mandibles (Ebeling and Pence 1957).

Australian Termi-mesh was developed from stainless steel wire in the early 1990s

to act as termite barrier installed during the construction process (Lenz and Runko 1994).

The mesh excludes termites because head capsules cannot fit through the screen (Grace et

al. 1996, Kard 1998). Therefore, the mesh size needed to exclude termites will vary

depending on species. It was originally intended to be a complete layer under the concrete

slab (Ewart 2001). The high cost of quality steel was prohibitive and limited sales. To

overcome this limitation, Termi-mesh sought and gained Australian recognition that a

properly constructed slab, with predictable cracking sites made by scoring the slab, could

also be a component of the barrier system (Ewart 2001). Mesh installation was then

targeted to the perimeter, slab penetrations, and slab joints. This allowed the Australian

company to grow its market share by reducing installation cost (Ewart 2001).

Use of the slab as a barrier, without any added mechanical or chemical devices,

has also been explored. Australian researchers investigated subterranean termite

movement through cracks in slabs of varying widths. They reported that cracks must be a

sufficient size to accommodate the termite head capsule, plus additional width to allow

body movement during tunnel construction (termites lined cracks with building (fecal)

material; Lenz et al. 1997). For example, Coptotermes acinaciformis, an Australian pest

species, was able to penetrate 1.5 mm-wide cracks in concrete slabs (Lenz et al. 1997).

Similar tests have not been published for North American subterranean termite species,

but R. flavipes workers should presumably be able to travel through slab cracks of at least

their body width, 1.11 mm (head capsule width of 1.03 mm; Su et al. 1991b). Cracks of

this size may occur during normal settling of a structure over time (NPCA 1980).

Impasse (Syngenta Crop Protection, Greensboro, NC) is an insecticide-

impregnated vapor retarder which contains the pyrethroid lambda-cyhalothrin within its

polymer layers to repel termites from cracks in the slab or gaps created around utility

penetrations (Harbison 2003). It is installed before the slab is poured and is intended to

provide at least 10 years of protection (Harbison 2003). This product became available in


Addition of termiticides to concrete has been investigated. Reticulitermes did not

tube over hardened concrete when either dieldrin or chlordane were added at the time of

concrete mixing (Allen et al. 1961, Beal 1971). These insecticides, however, have been

removed from the market due to environmental concern.

Referring to the second principle, the Termite Investigations Committee

recommended using wood pressure-impregnated with chemical preservatives (discussed

above) known to be toxic to termites (Brown et al. 1934). In areas where complete

protection was not feasible, the Committee recommended use of either sound seasoned

heartwood of an unpalatable species (discussed above); or wood treated with toxic

chemicals by methods other than pressure-impregnation, such as spray, brush, or dip.

Today's structural lumber is provided by softwood trees rather than hardwood. Therefore,


chemical treatment of wood is more common than using the hardwood of an unpalatable


As part of the third principle, the Termite Investigations Committee advised

homebuilders to provide adequate soil drainage, remove all cellulose sources (e.g.

stumps, roots, construction debris) in or on the ground near the structure, provide

substructural ventilation for crawl spaces, and locate and destroy colonies based on

swarming sites around the structure (Brown et al. 1934). For ventilation, the Committee

specifically called for an opening of 2 sq ft for each 25 linear feet of structural perimeter

(Brown et al. 1934). The Committee also recommended that wood floors have at least 18

inches of clearance between the joists and ground (Brown et al. 1934). The present-day

Florida Building Code (FBC) calls for at least 1 sq ft for each 150 sq ft of crawl space

that can be reduced by 10% if a vapor barrier is installed. Vapor barriers are required for

slab-on-grade construction (FBC 2002). The FBC presently mandates at least 8 inches of

clearance between the joists and ground (FBC 2002).

Termite protection principles for new construction were present in building codes

as early as 1923. That year, Burlington, Iowa, had provisions in its building code to

protect homeowners from poor construction that might lead to termite infestation (Snyder

1935). In 1927, the Pacific Coast Building Officials adopted recommendations of the

Bureau of Entomology and Plant Quarantine ("Bureau") for structural prevention of

termite damage as part of their uniform building code (Snyder 1935). Honolulu adopted

similar recommendations in 1928 (Snyder 1935). In 1934, both the Federal Housing

Administration and the Home Owner's Loan Corporation wrote specifications for

prevention of termite damage to woodwork of buildings (Snyder 1935). The 1935

resolutions of the Building Officials of America supported adoption of the Bureau's

structural termite prevention recommendations for the eastern United States (Snyder


Urban entomologists Smith and Zungoli influenced national building codes in the

late 1990s (Zungoli 1999) after results from a survey of 225 South Carolina pest control

companies indicated the below grade installation of rigid board insulation (RBI) increased

risk of subterranean termite infestations and undetectable structural damage (1995a).

Although termites gain no nutritional value from RBI (Hicken 1971), they will readily

tunnel through the material (Guyette 1994, Smith and Zungoli 1995b, Rambo 1998).

They reported that >12% of those surveyed had been involved in litigation concerning the

issue (Smith and Zungoli 1995). RBI was not visible in 82% of the inspected South

Carolina houses that reportedly contained RBI and had a history of unsolved termite

problems (Smith and Zungoli 1995b). Damage to these houses was randomly located

throughout the structure, as opposed to being concentrated in areas of excessive moisture

or areas that are difficult to treat (Smith and Zungoli 1995b). Language prohibiting the

installation of RBI below grade and specifying a 6 inch clearance between RBI and earth

was adopted by the Council of American Building Officials (CABO), the Standard

Building Code Congress International (SBCCI), and the International Code Council (ICC)

in response to the RBI survey and inspection reports (Zungoli 1999).

In April 1996, St. Johns County, Florida, enacted construction regulation to

reduce subterranean infestations. After convening with his appointed Subterranean

Termite Treatment Committee, County Building Official Roland Holt, amended the

building code to include a Termite Protection Ordinance, which was intended to decrease

subterranean termite food sources, eliminate hidden termite access into structures,

increase efficacy of the chemical barrier underneath and around structures, and provide

documentation of preconstruction termiticide treatment (Table 3-1). (The Ordinance is

discussed in Chapters 3 and 4.) Several Florida counties later adopted similar regulations.

Most of the St. Johns County Termite Protection Ordinance was approved for inclusion in

the new statewide Florida Building Code, effective March 1, 2002.

Statement of Purpose

This research was conducted to examine the relationship between construction

and subterranean termite infestation rates. First, the effect of concrete contamination on

the residual performance of termiticides was determined. In laboratory experiments, the

hypothesis that elevated soil pH due to masonry cement contamination degrades soil

termiticide performance was tested using subterranean termite mortality as a chemical

degradation indicator. Second, failure rates of termite prevention practices in three

Florida counties were determined and correlation of termite infestations with certain

building construction elements before and after building code changes was attempted.




An effective soil termiticide is expected to protect a structure from subterranean

termite infestation for at least 5 years (Kard et al. 1989). Despite this industry standard,

treatment failures still occur within that period. For example, 15.85% of houses 2-6 years

old in St. Johns County, Florida, that received a soil termiticide treatment at the time of

construction were infested as reported by homeowners (Chapter 3). Many factors can

influence treatment failures, including termite abundance, termiticide concentration, type

of active ingredient, product formulation, soil type, and chemical degradation (Macalady

and Wolfe 1983, Felsot 1989, Forschler and Townsend 1996, Gold et al. 1996).

Residual activity of soil termiticides in known to decline over time. In Florida

concrete slab tests, chlorpyrifos (Dursban, Dow AgroSciences, 10,000 ppm) failed in one

of ten plots after 9 years (Kard et al. 1989). Cypermethrin (Prevail FT, FMC Corp., 500

ppm) failed in one of ten plots after 5 years, and permethrin (Dragnet FT, FMC Corp.,

500 ppm) failed in two of ten plots after 4 years (Kard et al. 1989, Wagner et al. 2003).

Key factors that affect pesticide degradation include chemical, photochemical, and

microbial degradation, leaching, run-off, volatilization, bioaccumulation in plants and

animals (Rao et al. 1993), and size of area treated (Su et al. 1999a). Soil degradation rates

have been determined for chlorpyrifos, bifenthrin, permethrin, and cypermethrin when



applied at termiticidal concentrations. In laboratory and field trials, half-lives in different

soils ranged from 23 to 462 days for chlorpyrifos (analytical grade, Sigma, Racke et al.

1988; Dursban, Dow AgroSciences, 1,000 ppm in soil (wt/wt): Di et al. 1998, Baskaran et

al. 1999, Su et al. 1999b, Murray et al. 2001); 5 to 1,410 days for bifenthrin (analytical

grade, FMC Corp., 100 ppm in soil (wt/wt), Baskaran et al. 1999; Biflex FT, FMC Corp.,

31 ppm in soil, Su et al. 1999b); 22 to 45 days for permethrin (Dragnet FT, FMC Corp.,

50 ppm in soil (wt/wt), Su et al. 1999b); and approximately 12 days for cypermethrin

(Prevail FT, FMC Corp. 30 ppm in soil (wt/wt), Su et al. 1999b).

Termiticides protect structures by either killing or repelling termites (Forschler

1999). Nonrepellent termiticides, such as chlorpyrifos (Dursban, Dow AgroSciences; Su

et al. 1982 and 1995, Jones 1988) and imidacloprid (Premise 75, Bayer Environmental

Science; Kuriachan and Gold 1998, Gahlhoff and Koehler 2001) allowed termites to

penetrate treated soil but caused high mortality within 7 days. Greater than 90% of

exposed Reticulitermesflavipes (Kollar) workers died within 7 days after penetrating 2.5

cm into 5 cm of soil treated with 100 ppm of chlorpyrifos (Dursban, Dow AgroSciences;

Su et al. 1995). Reticulitermesflavipes completely penetrated 10 mm of soil treated with

10 ppm imidacloprid (Premise 75, Bayer Environmental Science; Gahlhoff and Koehler

2001). Pyrethroids, such as permethrin, cypermethrin, and bifenthrin, are strong repellents

that caused < 30% mortality of R. flavipes workers that were exposed for 7 d to 2.5 cm

treated segments but did not tunnel into them (Su et al. 1982 and 1995). Five cm of soil

treated with any of these pyrethroids at only 1 ppm was enough to completely repel

termites (Su et al. 1995). Degradation below this threshold concentration can result in

termites penetrating treated soil and accessing a structure.

Soil pH can also affect termiticide degradation. Alkaline pHs caused degradation

of chlorpyrifos (Racke et al. 1994, Baskaran et al. 1999), imidacloprid (Sarkar et al. 1999,

Zheng and Liu 1999), fipronil (U.S. EPA 1996, Bobd et al. 1998), and pyrethroids

(Camilleri 1984) in solutions. Gold et al. (1996) reported that 95% of the chlorpyrifos

(Dursban TC, Dow AgroSciences) applied at 1,000 ppm in soil (wt/wt) of pH 8.2 had

degraded within 2 years. Alkaline soils (pH 7.1 to 9.6) did not affect 50 ppm imidacloprid

(Bayer Environmental Science) and 100 ppm bifenthrin (FMC Corp.), both >98% purity

(Baskaran et al. 1999). Gold et al. (1996) studied the persistence of six termiticides for 5

years under field conditions and ranked them in the following order based on their

persistence: permethrin > fenvalerate > bifenthrin > chlorpyrifos > cypermethrin >

isophenphos. At soil pHs of 6.4 to 8.2, all termiticides originally applied at manufacturer

recommended rates were found at less than half of their original concentrations at 1 year

post-treatment (Gold et al. 1996). The lowest mortalities of R.flavipes caused by

termiticide treatments occurred in alkaline soils (pH 7.1 to 8.2) compared to treatments in

slightly acidic (pH 6.4) soil (Gold et al. 1996).

Construction practices may be an additional influence on termiticide performance.

Masonry cement contamination in the vertical treatment zone around the perimeter of

new construction may increase termiticide degradation. Portland cement is the most

common masonry cement used in brick veneer mortar and cement foundations (Allen

1999). This cement consists of calcium carbonate and/or magnesium carbonate and, if

mixed with moist Florida soil (Alachua County: loamy, siliceous, thermic Arenic

Paleudults; Thomas et al. 1985), will probably increase soil pH when it hydrolyzes to

calcium and/or magnesium hydroxide. The purpose of this study was to determine how

much portland cement is needed to raise the pH of fine loamy sand to alkaline levels and

also to determine the effect of elevated pH on degradation of various concentrations of

termiticides (chlorpyrifos, imidacloprid, fipronil, bifenthrin, permethrin, or cypermethrin)

as indicated by 24 h mortality of R.flavipes.

Materials and Methods

Field Soil

Field soil was collected from 27 homes in Gainesville, Florida (A horizon soil:

loamy, siliceous, thermic Arenic Paleudults; Thomas et al. 1985) for the purpose of

determining the pH range for laboratory bioassays. Nine newly completed structures (<2

weeks old), nine 5-year old structures, and nine 10-year old structures were included. The

nine structures in each age group were divided so that one-third of each were covered by

either stucco, siding, or brick veneer. Soil 10 cm from structural walls was collected

using a metal pipe (10.16 cm ID) inserted 20 cm deep into soil. Individual soil samples

were shaken loose from the pipe into labeled, resealable bags. For each structure, one soil

sample was taken from a randomly selected area.

Laboratory Bioassay Soil

Soil (approx. 10 lbs.) was collected from Gainesville (Alachua County), Florida,

oven-dried (177C for 24 h), and sieved (Fisher No. 16, Pittsburgh, PA). Soil was fine

loamy sand (A horizon soil: loamy, siliceous, thermic Arenic Paleudults; Thomas et al.


Soil pH Determination

Field soil. Soil pH for samples taken from next to structures was determined with

i pH-meter (Fisher No.13-620-290: Pencil Thin Gel Filled Combination Electrode with

BNC, Hanna, Pittsburgh, PA) as per the protocol of the University of Florida Institute of

Food and Agricultural Sciences Soil Testing Laboratory. This method requires that 33 g

soil (25 mL) and 50 mL distilled water be stirred together in a paper cup. One subsamples

of each soil sample was used. The mixtures were left to stand for 30 min, then stirred

again. Once the soil settled, the pH was determined.

Laboratory bioassay soil. The pH was determined for 3 subsamples of untreated

bioassay soil as described for field soil. After initial pH determination, a weighed

arbitrary amount of portland masonry cement (Quikcret, Atlanta, GA), type "S", was

then added to each 33- g soil sample, stirred, and left to stand for 30 min before pH re-

determination. Soil and cement were handled this way until the quantities of cement

needed to raise soil pH to 6, 7, 8, and 9 were determined.


Six termiticides, Dursban TC (44% chlorpyrifos; Dow AgroSciences,

Indianapolis, IN), Premise 2 (21.4% imidacloprid; Bayer Environmental Science, St.

Louis, MO), Termidor SC (9.1% fipronil; Aventis/BASF, Montvale, NJ), Talstar (7.9%

bifenthrin; FMC Corp., Philadelphia, PA), Demon TC (25.3% cypermethrin; Syngenta

Corp., Greensboro, NC), and Prelude (25.6% permethrin; Syngenta Corp., Greensboro,

NC), were used in this study. Distilled water was used to make all insecticide dilutions

and served as the control treatment (0 ppm insecticide). Three identical stock dilutions

were prepared in 100 mL volumetric flasks for each termiticide to achieve the highest

concentration tested when 3.3 mL of dilution was applied to 33 g soil (representing 10x

the calculated termiticide concentration after application to soil at label rate). (See

Appendix A for exact termiticide amounts added to distilled water). Three 1:10 serial

dilutions were then made from each stock dilution

Laboratory Bioassay Soil Treatments

Thirty-three g of soil were placed in individual plastic weigh boats (13 by 13 cm).

The appropriate amount of portland cement was added to each weigh boat to prepare each

pH level (6, 7, 8, and 9). Each stock dilution line served as a replication so that there were

3 replicates of each treatment combination: (6) termiticides, (5) concentrations, and (4)

pH levels) for a total of 360 treatment units.

To incorporate the desired amount of active ingredient into the soil, 3.3 mL of

dilution was added to 33 g soil/cement in weigh boats and stirred to uniformly moisten

the mixture (10% moisture; Gahlhoff 1999) and attain concentrations of 10,000, 1,000,

100, 10, and 0 ppm (wt [AI]:wt soil) for chlorpyrifos and permethrin; 5,000, 500, 50, 5,

and 0 ppm for cypermethrin; 600, 60, 6, 0.6, and 0 ppm for fipronil and bifenthrin; and

500, 50, 5, 0.5, and 0 ppm for imidacloprid. These concentrations include the

manufacturers' recommended application rates (see Appendix B for label rate

calculations for soil treatments) and equal 10x, Ix, 0. Ix, and 0.01x the label rates in soil.

The treated soils were air-dried in ahood for at least 5 d to allow solvents in the

formulation to evaporate.

In addition to the treatment units, three controls (soil + cement) were prepared

with distilled water for use as an assay for each soil pH (6, 7, 8, and 9) at each time

interval (1 day, 5 months, and 10 months). Termites were not added to the controls. This

resulted in a grand total of 396 units (360 bioassay units + 36 controls soil pH assay


Original weights were determined for all soil termiticide treatments and controls

(soil + cement + liquid + weigh boat) and the first bioassay began 24 h later. After each

bioassay, termites were removed, and soil was replaced and reweighed. Replaced soil was


reused at 5 and 10 month intervals. Distilled water was added every 7-10 days to maintain

the original wet weights over the 10-month period.

Weigh boats were stacked in plastic tubs (one tub per chemical) with steel mesh

screen between layers. Tubs were loosely covered with aluminum foil and kept in the

dark at ambient room temperature and humidity.


Termites from three R. flavipes field colonies of were collected from PVC

ground-tubes and plastic paint bucket traps on the University of Florida campus,

Gainesville, Alachua County, Florida.

One colony was collected as described by Powell (2000) from a PVC ground-tube

trap (11.5 cm diam. by 30.5 cm long) containing corrugated cardboard strips (236.0 cm

long by 15.2 cm wide; Hesco, Waverly, FL), buried approximately 12 cm in the ground,

and covered with a PVC cap (12.5 cm diam.; NIBCO, Elkhart, IN).

Termites from the other two colonies were collected as described by McManamy

(2002). Each trap consisted of a plastic paint bucket (19.5 cm by 22 cm; Venture

Packaging Inc. Monroeville, OH) monitor with lid buried in the ground. Holes (2 cm)

were drilled in the sides and bottom of the buckets for termite access. The traps contained

a roll of corrugated cardboard (236.0 cm long by 15.2 cm wide; Hesco, Waverly, FL) and

a block of wood (3.5 cm thick, 8.5 cm wide, 14 cm long; Pinus spp.) with six grooves (3

mm wide, 1 cm deep, 1 cm apart) on one side to encourage termite exploration. The

bucket was then concealed with groundcover.

Cardboard containing termites were removed and placed in clear plastic boxes (ca.

30 by 19 by 10 cm) containing a layer of damp paper towels and covered with a close-

fitting lid. Boxes were labeled with colony name, collection date, and stacked inside 32-


gallon polyethelyne storage container (ca. 36 by 30 by 80 cm). Insects were held for up to

10 d at 23 lC and 94 5% RH.

Data Collection and Research Design

Field soil. A single soil pH was determined for each structure. Each experimental

unit for field soil pH determination consisted of field soil and distilled water in a weigh

boat. This was a 3 x 3 factorial design (3 claddings x 3 ages).

Laboratory soil pH assay for bioassay soil. For controls, pH was determined for

three 33 g samples of clean, dry soil, without cement. The pH of laboratory bioassay

soil/cement mixed with distilled water was determined immediately after treatment and

again after 5 and 10 months for 3 randomly chosen units for each of the four starting pH

levels. Each experimental unit consisted of soil, cement, and distilled water in a weigh

boat, replicated three times.

Bioassay. Bioassays were conducted at 5 days, 5 months, and 10 months after

soil treatment. Each termite colony served as one replicate (n=3). For each termiticide a

group of ca. 10 worker termites (undifferentiated larvae of at least the third instar), one

each from three separate colonies, were each placed on 7 g treated soil and held in the

dark in covered 29.57 mL plastic cups. Number of dead termites was counted at 24 h. All

termites then removed from treated soil and the soil was returned to its respective weigh

boat for reuse.

The experiments for each termiticide were a 5 (concentrations) x 4 (pH) x 3 (time

intervals) factorial design replicated three times. Each experimental unit consisted of ca.

10 termites and treated soil/cement mixture in a plastic cup.

An additional bioassay was conducted at 10 mo. for 0, 5, and 50 ppm

imidacloprid in soil in which termites were held in covered cups for 7 d, instead of 24 h,

so that full effect of imidacloprid could be realized. The experimental units were finely

sprayed with distilled water every 24 h to prevent dessication. Three replications were

made and mortality was recorded at 7 d. This was a 3 (concentrations) x 4 (pH) factorial


Data Analysis

Field soil pH sssav. The pH means, minimums, and maximums for soil located

adjacent to Gainesville structures of different exterior claddings and different ages were

determined. A two-way analysis of variance (ANOVA) was used to determine if there

were significant interactions between the main effects: cladding type and age (a= 0.05;

SAS Institute 2000).

Laboratory soil pH assay. The relationship between cement weights and soil pH

(6, 7, 8, and 9) at 5 d was determined by linear regression. Changes in pHs over time for

pH assay soil were analyzed by one-way ANOVA, and mean pH levels at 0, 5, and 10

months were separated using Scheffe's Test (a= 0.05; SAS Institute 2000).

Bioassa. For each termiticide, a three-way ANOVA was performed to determine

significant interactions of main effects (pH, termiticide concentration, time interval) on

termite mortality. When three-way interactions were significant and variation occurred

among mortality means, two-way ANOVAs were used for each termiticide to determine

the effect of concentration and pH on termite mortality at specific time intervals. Mean

termite mortality at each time interval for each concentration and pH level for

cypermethrin was reported because the three-way interaction was not significant. When

two-way interactions were significant, one-way ANOVA was performed for each

termiticide concentration and time period to determine differences between pH

treatments. When the two-way interaction was not significant, mean percentage termite

mortality due to concentration was separated within time periods. At 10 mo., additional

ANOVAs were performed to determine the effect of concentration and pH on mortalities

of termites confined for 7 d on soil treated with 0, 5, and 50, ppm imidacloprid. Means

were separated using Scheffe's Test (SAS Institute 2000). Percentage mortality was

arcsine square-root transformed before all ANOVA, and a-level of 0.05 was used for all

analyses. Mortality served as a measure of residual activity of soil termiticides.

The relationship between time period and pH on mean percentage termite

mortality was estimated by linear regression (SAS Institute 2000) for the highest

termiticide concentrations at which pH significantly affected mortality. Non-overlap of

95% confidence intervals of slopes were considered to be significantly different.


Field Soil pH Assay

The pH of soil taken from 27 Gainesville structures ranged from 5.20 for 10-year-

old stuccoed houses through 10.10 for new houses with brick veneer (Table 2-1). Soil

was alkaline for all 5 year old structures. There was no significant interaction between

exterior cladding and structural age, and neither main effect was significant (Cladding:

df=2, MS=2.9633, F=2.82, P=0.0863; Age: df=2, MS=1.5478, F=1.47, P=0.2562;

Cladding x Age: df=4, MS=1.4544, F=1.38, P=0.2797; Error: df=18, MS=0.0526; Table


Laboratory Soil pH Assay

Soil pH before treatment was 5.6. The soil required less than 1% of its weight in

cement to increase pH to 9.07 (Table 2-2 and Fig. 2-1). Soil pH did not significantly

change during the first 5 months of being held in the laboratory. Soils of pH 7.02, 8.02,

and 9.07 significantly decreased to below neutral after 10 mo. Soil at pH 6.01

significantly increased to 6.53 at 10 mo. (See Appendix C for ANOVA table.)


No termites died from the control treatments. Cypermethrin caused >93%

mortality at all three time periods for subterranean termites exposed to treated soil for 24

h compared with the other five termiticides. There were no significant interactions

between cypermethrin concentration, soil pH, and time (Table 2-3). Only concentration

had a significant effect on mortality of termites exposed for 24 h to cypermethrin-treated

soil (Table 2-3). A two-way ANOVA was not performed for the cypermethrin treatments

because the three-way interaction was not significant. Residual activity of cypermethrin

was not significantly decreased by pH and 2 93.33% of the treated termites died with 24 h

for all pH levels, cypermethrin concentrations and time intervals (Table 2-5).

All interactions and main effects significantly affected mortality of termites

confined on soil treated with the other five soil termiticides: bifenthrin, permethrin,

chlorpyrifos, fipronil, and imidacloprid (Appendix D). For bifenthrin and permethrin, all

termites died within 24 h for all concentrations at 5 d. The interaction between

concentration and soil pH, and main effects, significantly affected mortality at 5 and 10

months for bifenthrin and at 10 months for permethrin (Table 2-4) Residual activity of

both bifenthrin and permethrin at 0.01x the label rates (0.6 and 10 ppm, respectively) was

significantly reduced by pH (Table 2-4). For bifenthrin, residual activity decreased at both

the 5 and 10 month intervals (Table 2-6 and Appendix E), while permethrin's activity

was significantly reduced at 10 months (Table 2-7 and Appendix E).

All termites exposed to chlorpyrifos died within 24 h for all concentrations at 5 d.

The interaction between concentration and soil pH significantly affected mortality of

termites confined to chlorpyrifos-treated soil at 5 and 10 months (Table 2-4). Soil

concentrations of chlorpyrifos >0. lx the label rate were not significantly affected by pH

as evidenced by complete termite mortality (Table 2-8 and Appendix E). Ten ppm

chlorpyrifos (0.01x label rate) killed all termites at 5 d, but residual activity significantly

declined in soil made alkaline by cement (Table 2-8 and Appendix E).

For the fipronil treatments, only concentration significantly affected termite

mortality at 5 d. The concentration-pH interaction was significant at 5 and 10 months

(Table 2-4). Fipronil killed all termites in soil treated at 6 ppm (0. lx label rate) at 5 d, but

then mortality decreased significantly at alkaline pHs at 5 and 10 months (Table 2-9 and

Appendix E).

Imidacloprid concentration significantly affected termite mortality at 5 d, and the

concentration-pH interaction and both main effects were significant at 5 and 10 months

(Table 2-4). Imidacloprid applied at the label rate (50 ppm) killed all termites within 24 h

at 5 d (Table 2-10). For treatments aged 5 and 10 months, termite mortality was

significantly reduced at all pH levels, with a mean of only 7.22% of termites dying from

the label rate treatment at 10 mo. (Table 2-10 and Appendix E). However, at the 10 mo.

and label rate all confined termites died within 7 d at all pHs (Table 2-10). No termites

died within 24 h from 5 ppm imidacloprid (0. Ix label rate) aged 10 mos., but 30.56% of

the termites died after 7 d confinement on soil from the same treatment. (See Appendix F

for ANOVA table for all termiticides.)

Linear regression analysis of termite mortality after 24 h exposure to treated soil

aged 5 mo. indicated an inverse relationship between pH and termite mortality for

bifenthrin, chlorpyrifos, fipronil, and imidacloprid treatments (Fig. 2-2). Confidence

intervals of slopes for these four treatments overlapped at 5 mo. (Fig. 2-2). Five ppm

cypermethrin killed all termites at all pH levels (Fig. 2-2). At 10 mo., there was an

inverse relationship between pH and termite mortality for bifenthrin, permethrin,

chlorpyrifos, fipronil, and imidacloprid treatments (Fig. 2-3). Five ppm cypermethrin

killed all termites within 24 h. pH had the greatest effect on residual activity of

termiticides, at the concentrations presented in Figs. 2-2 and 2-3, in the following order:

imidacloprid > fipronil > chlorpyrifos = bifenthrin > permethrin > cypermethrin.


Field Soil pH Assay

The pH minimums of soil adjacent to new structures covered by either stucco or

siding agreed with the normal pH range, 4.5 through 6.5, of Alachua County fine, loamy

sand (Thomas et al. 1985). However, since alkaline soils are not normally found in the

Alachua area, the high pH soil adjacent to new brick veneer work may have been due to

the cement powder becoming mixed with the soil during the construction process. If so,

this would result in varying levels of cement contamination throughout the soil. Cement

mixed with moist soil raises soil alkalinity and may lead to subsequent hydrolysis of

applied termiticides, thereby rendering them impotent (Naumann 1990).

Soil pHs 6, 7, 8, and 9 were chosen for the bioassay based upon the results of the

field soil pH assay. Soil pH of 10 was not tested because soil from only one new brick

veneer structure of that pH was detected after bioassay testing began.

Laboratory Soil pH Assay

The change in the control soil pH may have occurred due to a combination of

microbial action and cation exchange capacity of the soil. Although soil was baked at a

sterilizing temperature before use, it was not kept in a sterile environment. Therefore

microbes could have aided in shifting the soil pH back towards its original slightly acidic

level (Brady and Weil 1999). The biggest change in soil pH over the 10-month period

was in soil that started at 9.07. This change was probably due to the cation exchange

capacity of soil, which increases with soil pH (Brady and Weil 1999). Also, water was

added to soil every 7-10 days so that evaporation occurred. This wet-dry regime also

could have contributed to the slow pH change over time.


The lowest concentrations that killed all termites exposed to treated soil for 24 h

at 5 d were 5, 0.6, and 10 ppm cypermethrin, bifenthrin, and permethrin, respectively.

Similar results were reported by Smith and Rust (1990) who found 100% mortality of R.

hesperus Banks confined on treated soil for 3 h. However, increased soil pH significantly

reduced mortality of termites confined to soil aged for either 5 or 10 mo. and treated with

either bifenthrin or permethrin at 0.6 and 10 ppm, respectively. These concentrations

represent 0.01x the label rate for soil treatments. Efficacy of bifenthrin significantly

decreased at the 5 month interval, while mortality of termites held on permethrin-treated

soil did not decrease until the 10 month interval. Su et al. (1999b) also found that

bifenthrin (formulated as Biflex) degraded faster than either permethrin (formulated as

Dragnet) or cypermethrin (formulated as Prevail). Their study, however, did not

differentiate soil pH levels. In this study, termite mortality from 0.01x the label rate of

both pyrethroids bifenthrin and permethrin decreased as soil pH increased.

In contrast to the other tested pyrethroids, cypermethrin soil concentrations

surprisingly were not affected by soil pH within the 10 month period. The fluoride in the

bifenthrin theoretically should have made it more stable in alkaline soil than either

permethrin or cypermethrin which do not contain a fluoride constituent (Naumann 1990).

Perhaps the different residual activities of these pyrethroids were affected more by their

inert ingredients (Harris 1972, Naumann 1990) than by their active ingredients.

Chlorpyrifos at > 10 ppm killed all termites within 24 h at all pH levels at 5 d.

Smith and Rust (1990) reported that chlorpyrifos in soil at 50 ppm killed 100% of

treatment-confined R. hesperus within only 7 h. Over the 10 month period, however,

alkaline soil pH significantly decreased residual activity of the lowest tested

concentration of chlorpyrifos in soil, 10 ppm (0.01x label rate). (Chlorpyrifos soil

concentrations >100 ppm were not affected by pH.) This parallels a study conducted by

Racke et al. (1994), who reported chlorpyrifos could have a half-life of only three months

at 10 ppm in the alkaline soils of Florida and Texas, while a concentration of 1,000 ppm

would retard hydrolysis. Murray et al. (2001) found that the stability of chlorpyrifos at

1,000 ppm is unaffected by natural soil alkalinity. The slower degradation of higher

concentrations of chlorpyrifos probably occurred because the degradation product, 3,5,6-

trichloro-2-pyridinol, is antimicrobial and retards further degradation by microbes

(Somasundaram et al. 1990). Both permethrin and cypermethrin, before degradation, in

sandy loam at 5 ppm were also antimicrobial (Naumann 1990).

Imidacloprid (Boucias et al. 1996, Kuriachen and Gold 1998, Gahlhoff 1999,

Ramakrishnan et al. 2000, Gahlhoff and Koehler 2001) and fipronil (Gahlhoff 1999,

Osbrink et al. 2001) have been shown to be nonrepellent, slow-acting contact poisons. In

this study, they were the only treatments that caused less than 100% mortality of termites

at 5 d. Boucias et al. (1996) found at least three days exposure to soil treated with 50-100

ppm technical imidacloprid were needed before termites began dying. For fipronil,

Osbrink et al. (2001) reported approximately 10-fold higher LT0s forR. virginicus

(Banks) compared to lethal times for termites exposed to chlorpyrifos, permethrin, or

cypermethrin. Also, fipronil in sand, soil, and clay at 5 ppm required 5 to 10 d to cause

100% mortality of Coptotermesformosanus Shiraki in test tube tunneling assays (Osbrink

and Lax 2002). Therefore, the 24 h exposures used in this study probably was not enough

time for full expression of termite mortality from these slow-acting termiticides. All

termites exposed for 7 d to label rate imidacloprid (50 ppm) at 10 mo. died at all pH

levels. However, with only 24 h exposure full expression of imidacloprid toxicity was not

realized, and pH seemed to affect termite mortality. Also, 5 ppm imidacloprid (0. Ix label

rate) killed 30.56% of the termites after 7 d confinement on treated soil, but no termites

died within the first 24 h of confinement. Although no 7 d trials were run with fipronil, a

longer exposure (>24 h) to treated soil should result in higher termite mortality (Osbrink

and Lax 2002).

Chlorpyrifos (Racke et al. 1994) and pyrethroids (Naumann 1990) are degraded by

hydrolysis under alkaline conditions. However, chlorpyrifos is more water soluble

(chlorpyrifos water solubility [WS] -0.73 ppm; Su et al. 1999b) than permethrin (WS -

0.006 ppm; Su et al. 1999b) and cypermethrin (WS <0.0001 ppm; Su et al. 1999b) and

could have degraded faster than the two pyrethroids. In termiticide degradation studies

using treated soil under miniature slabs, Tamashiro et al. (1987), Su et al. (1993b and


1999b), and Gold et al. (1996) reported faster degradation of chlorpyrifos compared with


Toxicant adsorbs more readily to sand particles than clay or silt particles (Harris

1972). In general, increased organic matter (Smith and Rust 1990 and 1993) or increased

silt or clay (Henderson et al. 1998b, Forschler and Townsend 1996, Gold et al. 1996)

decreases insecticide efficacy. The greater surface area of colloids found in silt and clay

increases chemical binding sites for termiticide adsorption and renders the toxicant as

unavailable to termites. A substrate with many colloidal surfaces will cause decreased

efficacy of applied termiticides (Smith and Rust 1990, 1993, Forschler and Townsend

1996, Gold et al. 1996) compared to the same termiticide applied to sand. The Alachua

County fine loamy sand used in my study was 4-10% clay, 5-20% silt, and 70-85% sand,

with 2-4% organic matter (Thomas et al. 1985). Adsorption of the different termiticides

to soil particles, and their subsequent bioavailability to termites, may have been

influenced by the chemicals' size, shape, conformation, configuration, polarity,

polarizability, pH, charge distribution, and water solubility (Harris 1972). Additionally,

the nature and properties of the organic and inorganic soil colloids could have also

influenced termiticide adsorption to soil particles (Harris 1972).

Residual activity of very low initial soil concentrations for five of the termiticides

tested were affected by alkaline pH. All affected concentrations were 0.Ix to 0.01x of the

label rates. These rates, however, are important because uniform distribution of

termiticide in the soil is unlikely. The amount of termiticide recovered significantly

varied between samples taken from different locations within the same treatment area

after chemical and soil were blended in a concrete mixer (Kard and McDaniel 1993, Gold

et al. 1996). In samples taken only 24 hours after mixing soil and termiticide, Kard and

McDaniel (1993) found 777 to 1071 ppm of chlorpyrifos (Dursban) originally applied to

soil at a calculated concentration of 1,000 ppm; 374 to 568 ppm permethrin (Dragnet)

was recovered from soil originally treated at 500 ppm; and 347 to 513 ppm cypermethrin

(Demon) was recovered from soil originally treated at 250 ppm. Recovery variability was

probably caused by non-uniform distribution of termiticide in the soil, resulting in some

areas with high termiticide concentration and others with low concentrations.

Several factors influence the distribution of termiticides in soil. Horizontal

penetration of termiticides in soil varied based on soil type. Beal and Carter (1968) found

more than 11% of the dieldrin originally applied to Northern Florida soil moved

downward 3.75 inches within 24 hours, while only 0.03% of the dieldrin moved to the

same depth in Arizona soil (dieldrin applied at 2 pt/ft2 sprinkled on top of soil).

Differential vertical penetration of termiticides depended upon the chemical (Bennet et al.

1974), sample distance from the injection point (Davis and Kamble 1992, Davis et al.

1993), and quantity interacting with application pressure (Davis and Kamble 1992).

Low concentrations of termiticide, caused by non-uniform distribution, may occur

in conjunction with cement added to the soil in the treated area. Soil moisture would

hydrolyze the cement and free enough hydroxyl ions to increase soil pH, resulting in

termiticide breakdown. An additional concern is that disruption or movement of treated

soil due to construction activity may also lower the concentration of chemical. According

to this laboratory study, addition of portland cement to soil treated with low

concentrations of chlorpyrifos, imidacloprid, fipronil, bifenthrin, and permethrin

degraded the termiticides as indicated by low termite mortality. Contamination within the

termiticide treatment area around new construction could easily be reduced by use of a

drop cloth to catch scrap cement.

Table 2-1. Range of soil pH levels for soil within 10 cm of structures covered by different exterior claddings located in Gainesville,
Mean soil pH SEM (Min. to Max.)
Structure Age Stucco (n=3) Siding (n=3) Brick Veneer (n=3)
< 2 weeks 6.47 0.45 (5.60 7.10) 6.77 0.64 (5.50 7.60) 8.90 0.72 (7.60 10.10)
5 years 7.80 0.26 (7.30 8.20) 7.83 0.35 (7.20 8.40) 8.57 0.26 (8.10 9.00)
10 years 7.80 0.26 (5.20 8.40) 7.43 0.69 (6.60 8.80) 7.27 0.46 (6.50 8.10)

ANOVA resulted in no significant interaction between cladding and age, and no significant differences in mean soil pH between
houses of different claddings or ages. (Two-way ANOVA: Cladding df=2, MS=2.9633, F=2.82, P=0.0863; Age df=2, MS=1.5478,
F=1.47, P=0.2562; Cladding x Age df=4, MS=1.4544, F=1.38, P=0.2797; Error df=18, MS=0.0526)

Tabld 2-2. Amount of portland cement needed to raise pH of 33 g fine loamy sand from 5.60 to 9.07 and the change in pH of aged
sand held in the laboratory and treated with distilled water
mg cement Starting pH Ending pH (Mean SEM)
added % of sand n 1 Day 5 Months 10 Months
0.0 --- 3 5.60 0.02 ----- -----
15.0 0.05 3 6.01 0.02a 6.34 0.19ab 6.53 0.06b
47.6 0.14 3 7.020.01a 7.21 0.1lab 6.580.16b
148.0 0.45 3 8.02 0.01a 8.02 0.15a 6.69 0.12b
291.0 0.88 3 9.07 0.01a 8.93 0.09a 6.97 0.04b
Means within a row followed by the same letter are not significantly different (a = 0.05, Scheffe's Test). See Appendix C for the
ANOVA table.

Table 2-3. Three-way ANOVA of the effects of concentration, pH, and time on the mortality of termites confined on soil treated with
cypermethrin (Demon TC)
Source df MS F P
Cone (C) 4 17.6534 11084.70 <0.0001
pH 3 0.0011 0.71 0.5471
Time (T) 2 0.0030 1.87 0.1592
pHXT 6 0.0011 0.71 0.6411
CXpH 12 0.0011 0.71 0.7830
CXT 8 0.0030 1.87 0.0715
C X pH XT 24 0.0011 0.71 0.8319
Error 120 0.0016

Percentage termite mortality was arcsine square-root transformed before analysis. No two-way ANOVA of cypermethrin was required
because 3-way interaction was insignificant.

Table 2-4. Two-way ANOVA of the effects of concentration and pH on the mortality of termites exposed for 24 h to soil the same
day, or 5 and 10 mos, after it was treated with currently registered termiticides
Treatment Time Source df MS F P
Bifenthrin 5 Mo. Cone (C) 4 6.3498 1437.56 < 0.0001
(Talstar) pH 3 0.0541 12.24 <0.0001
CXpH 12 0.0541 12.24 <0.0001
Error 40 0.0044

10 Mo. Conc (C) 4 6.4627 1201.62 < 0.0001
pH 3 0.0981 18.25 < 0.0001
CXpH 12 0.0981 18.25 <0.0001
Error 40 0.0054

Permethrin 10 Mo. Cone (C) 4 6.6425 3756.62 < 0.0001
(Prelude) pH 3 0.0388 21.92 <0.0001
CXpH 12 0.0388 21.92 <0.0001
Error 40 0.0018

Chlorpyrifos 5 Mo. Cone (C) 4 5.5533 515.01 < 0.0001
(Dursban TC) pH 3 0.0641 5.94 0.0019
C X pH 12 0.0641 5.94 < 0.0001
Error 40 0.0108

10 Mo. Conc (C) 4 5.5574 56926.50 < 0.0001
pH 3 0.1618 1657.17 < 0.0001
C X pH 12 0.1618 1657.17 < 0.0001
Error 40 0.0001

Table 2-4 Continued
(Termidor SC)

Time Source
5 Days Cone (C)

5 Mo.

Conc (C)

10 Mo. Cone (C)

Imidacloprid 5 Days Cone (C)
(Premise 2) pH
1 Day C X pH

5 Mo. Cone (C)

10Mo. Conc (C)



















< 0.0001

Table 2-4 Continued
Treatment Time Source df MS F P
Imidacloprid 10 Mo. Conc (C) 2 7.5674 6057.89 <0.0001
(Premise 2) pH 3 0.0087 7.00 0.0015
7 days CXpH 6 0.0087 7.00 0.0002
Error 24 0.0012

Percentage termite mortality was arcsine square-root transformed before analysis. Control mortality was 0%. No two-way ANOVA of
cypermethrin was required because 3-way interaction was insignificant (Table 2-3).

Tabld 2-5. Mean percentage mortalities (24h) of termites held on soil treated with cypermethrin (Demon TC)
Soil Concentration n pH 5 Days 5 Months 10 Months
5,000,500, 3 6 100.00 0.00 100.00 0.00 100.00 0.00
and 3 7 100.00 0.00 100.00 0.00 100.00 0.00
50 ppm 3 8 100.00 0.00 100.00 0.00 100.00 0.00
3 9 100.00 0.00 100.00 0.00 100.00 0.00

12 Mean 100.00 0.00 100.00 0.00 100.00 0.00

3 6 100.00 0.00 100.00 0.00 100.00 0.00
3 7 100.00 0.00 100.00 0.00 93.33 6.67
3 8 100.00 0.00 100.00 0.00 100.00 0.00
3 9 100.00 0.00 100.00 0.00 97.67 2.33
12 Mean 100.00 0.00 100.00 0.00 97.75 1.72

3 6 0.00 0.00 0.00 0.00 0.00 0.00
3 7 0.00 0.00 0.00 0.00 0.00 + 0.00
3 8 0.00 0.00 0.00 0.00 0.00 0.00
3 9 0.00 +0.00 0.00 + 0.00 0.00 + 0.00
12 Mean 0.00 0.00 0.00 0.00 0.00 0.00

5 ppm

0 ppm

Table 2-6. Mean percentage mortalities of termites held on soil treated with bifenthrin (Talstar)

Soil Concentration
6 ppm

0.6 ppm

0 ppm


5 Days
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

100.00 0.00
100.00 0.00
100.00 0.00
IMnn i 0 nn

5 Months
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

63.33 1.93a
48.90 1.10ab
20.10 7.59bc
In nn '77c

10 Months
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

74.23 7.76a
40.00 10.00ab
18.90 6.75bc
1 12 -) 713

12 Mean 100.00 0.00 35.58 6.80 33.84 8.68

3 6 0.00 0.00 0.00 0.00 0.00 0.00
3 7 0.00 0.00 0.00 0.00 0.00 0.00
3 8 0.00 0.00 0.00 0.00 0.00 0.00
3 9 0.00 0.00 0.00 0.00 0.00 0.00
12 Mean 0.00 0.00 0.00 0.00 0.00 0.00

Percentage termite mortality was arcsine square-root transformed before analysis.




Table 2-7. Mean percentage mortalities of termites held on soil treated with nermethrin (Prelude)

n pH
3 6
3 7
3 8
3 q

5 Days
100.00 0.00
100.00 0.00
100.00 0.00
01 l0 A 0- fn

5 Months
100.00 0.00
100.00 0.00
100.00 0.00
1nn00 n 00nn

10 Months
100.00 0.00
100.00 0.00
100.00 0.00
10nn r nnn

Soil Concentration
10,000, 1,000,
100 ppm

10 ppm

0 ppm

Percentage termite mortality was arcsine square-root transformed before analysis.

3 9 10 0. A 0f0
12 Mean 100.00 0.00 100.00 0.00 100.00 0.00

3 6 100.00 0.00 100.00 0.00 51.10 1.10a
3 7 100.00 0.00 100.00 0.00 31.13 2.94ab
3 8 100.00 0.00 100.00 0.00 17.77 2.23bc
3 9 100.00 0.00 100.00 + 0.00 5.57 2.94c
12 Mean 100.00 + 0.00 100.00 0.00 29.39 5.19

3 6 0.00 0.00 0.00 0.00 0.00 0.00
3 7 0.00 0.00 0.00 0.00 0.00 0.00
3 8 0.00 0.00 0.00 0.00 0.00 0.00
3 9 0.00 0.00 0.00 0.00 0.00 0.00
12 Mean 0.00 0.00 0.00 0.00 0.00 0.00

Table 2-8. Mean percentage mortalities of termites held on soil treated with chlorpyrifos (Dursban TC)

Soil Concentration
1,000, and
100 ppm

10 ppm

0 ppm

n pH
3 6
3 7
3 8
3 9

5 Days
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

5 Months
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

10 Months
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

12 Mean 100.00 0.00 100.00 0.00 100.00 0.00

3 6 100.00 0.00 100.00 0.00a 100.00 0.00a
3 7 100.00 0.00 88.90 11.10ab 100.00 + 0.00a
3 8 100.00 0.00 74.47 4.00 ab 52.23 2.23b
3 9 100.00 0.00 54.43 15.54b 28.90 1.10c
12 Mean 100.00 0.00 79.45 6.63 70.28 9.31

3 6 0.00 0.00 0.00 0.00 0.00 0.00
3 7 0.00 0.00 0.00 0.00 0.00 0.00
3 8 0.00 0.00 0.00 0.00 0.00 0.00
3 9 0.00 + 0.00 0.00 0.000 00: 0.00
12 Mean 0.00 0.00 0.00 0.00 0.00 0.00

Percentage termite mortality was arcsine square-root transformed before analysis.

Tabl6 2-9. Mean percentage mortalities of termites held on soil treated with fipronil (Termidor SC)

Soil Concentration
600 and
60 ppm

6 ppm

0.6 ppm

0 ppm

n pH
3 6
3 7
3 8
3 9

5 Days
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

5 Months
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

10 Months
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

12 Mean 100.00 0.00a 100.00 0.00 100.00 0.00

3 6 100.00 0.00 86.67 1.93a 97.77 2.23a
3 7 100.00 0.00 92.23 2.23a 57.77 2.23b
3 8 100.00 0.00 92.00 1.00a 44.53 5.90bc
3 9 100.00 0.00 18.67 7.22b 21.10 2.20c
12 Mean 100.00 0.00a 72.39 9.53 55.29 8.52

3 6 62.20 1.10 66.67 0.00a 83.33 3.84a
3 7 53.33 2.67 66.67 0.00a 61.13 5.57ab
3 8 63.43 2.02 65.47 1.23a 52.23 2.94bc
3 9 62.00 4.00 1.00 1.00b 22.23 8.00c
12 Mean 60.74 1.48b 49.97 8.53 54.73 7.00

3 6 0.00 0.00 0.00 0.00 0.00 0.00
3 7 0.00 0.00 0.00 0.00 0.00 0.00
3 8 0.00 0.00 0.00 0.00 0.00 0.00
3 9 0.00 0.00 0.00 0.00 0.00 0.00
12 Mean 0.00 0.OOc 0.00 0.00 0.00 0.00

Percentage termite mortality was arcsine square-root transformed before analysis.

Table 2-10. Mean percentage mortalities of termites held on soil treated with imidacloprid (Premise 2)

Soil Concentration
500 ppm

50 ppm

5 ppm

0.5 and
0 ppm

n pH
3 6
3 7
3 8
3 9

5 Days
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

5 Months
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

10 Months (1 day)
100.00 0.00
100.00 0.00
100.00 0.00
100.00 0.00

10 Months (7 days)

12 Mean 100.00 0.00a 100.00 0.00 100.00 0.00 ---

3 6 100.00 0.00 71.10 10.93a 10.00 0.00a 100.00 0.00
3 7 100.00 0.00 54.43 4.43ab 11.10 1.10a 100.00 0.00
3 8 100.00 0.00 39.97 3.33b 7.77 4.00ab 100.00 0.00
3 9 100.00 0.00 32.20 1.10b 0.00 0.00b 100.00 0.00
12 Mean 100.00 0.00a 49.43 5.19 7.22 1.59 100.00 0.00

3 6 44.43 1.13 28.87 7.27a 0.00 0.00 43.33 3.33a
3 7 47.77 2.23 12.23 2.93ab 0.00 0.00 28.90 1.00ab
3 8 44.43 1.13 8.90 2.20b 0.00 0.00 25.57 2.94b
3 9 46.67 1.93 0.00 0.00c 0.00 0.00 24.43 4.43b
12 Mean 45.83 0.84b 12.50 3.60 0.00 0.00 30.56 2.65

3 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
3 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
3 8 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
3 9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
12 Mean 0.00 0.OOc 0.00 0.00 0.00 0.00 0.00 0.00

Percentage termite mortality was arcsine square-root transformed before analysis.




z pH=6.08 + 0.01
7 o .. r=0.91


0 100 200

Cement (mg)

Figure 2-1. Relationship between amount of cement in soil and pH of mixture.

)5 (mg cement)

300 400
300 400

-# Chlorpyrifos 10ppm (/100)
-* Imidacloprid 50 ppm (label)
--- Fipronil 6 ppm (1/10)
-1.-. Bifenthrin 6 ppm(1/100)
----Permethrin 10 ppm. Cypenmethrin5 ppm(1/100)

Y=177.18-18.88(pH) -.I Y=147.80- 13.
r2 =0.88 (-24.33.-13.43) -. r2 =0.73 (-16
---- +

"'. L

6 7 8 9


Figure 2-2. Mortalities (24 h) of R. flavipes confined on soil of pH 6, 7, 8, or 9, treated with termiticides, and aged 5 months.
Confidence intervals (95%) for slopes are in {).

-- -Chlorpyrifos 10 ppm (1/100)

--- Imidacloprid 50 ppm (label)

- Fipronil 6 ppm (1/10)

.1. Bifenthrin 6 ppm (1/100)

S--*--Permethrin 10 ppm (1/100)

Cypermethrin 5 ppm (1/100)

Figure 2-3. Mortalities (24 h) of R.flavipes confined on soil of pH levels 6, 7, 8, or 9,treated with termiticides, and aged 10 months.
Confidence intervals (95%) for slopes are in { }.

o 50


6 7 8 9



Although subterranean termites are ecologically important as decomposers of

cellulose material, they are considered pests in urban habitats. Their cryptic habits make

control difficult and allow them to cause hidden damage to structures. The elements

necessary for subterranean termite infestations are moisture, cellulose, and structural


Subterranean termites are vulnerable to desiccation and require access to a supply

of moisture. Water content of R. flavipes nest material was reported to range from 16.3 to

67.7% and relative humidity within their galleries was reported to be 100% (Sponsler and

Appel 1990). If away from the high humidity within the colony, most termites can only

survive in air close to its water vapor saturation point (Rudolph et al. 1990). Inside the

colony, moist soil and fecal deposits throughout the galleries maintain the required high

moisture (Potter 1997). Eastern subterranean termite workers are -75% water (Sponsler

and Appel 1990) and maintain their water content by either withdrawing capillary water

from the soil or drinking free water (Rudolph et al. 1990). In the urban environment,

factors that may provide high soil moisture, and create ideal habitats for subterranean

termites around buildings, include mulch, sprinkler irrigation, roof run-off, gutter

downspouts, and air conditioner condensation drip lines (Haagsma and Rust 1995).


Subterranean termites can survive in a variety of substrates and moisture contents.

Haverty (1979) reported 75 to 82% survival of R.flavipes in 8-week laboratory bioassays

using sand (9 to 17% moisture by weight), vermiculite (100 to 500% moisture by weight),

and combinations of the two (43 to 63% moisture by weight). Vermiculite substrates were

less suitable for R. virginicus, but survivorship was highest in the sand and combination

substrates (65 to 77% and 75 to 79% survival, respectively; Haverty 1979).

The moisture requirement of subterranean termites is met partially by their own

metabolism and by the moisture diffusing through the soil in their galleries (Rudolph et

al. 1990). Ambient relative humidity and moisture content of occupied wood are less

important for maintaining water content of individual termites that are still connected to

their humid underground galleries (Rudolph et al. 1990). Subterranean termites use

tunnels to explore the soil environment. Grube and Rudolph (1999) speculated that the

feces and soil lining of their tunnels serves to conserve humidity.

When in soil with an acceptable moisture content, subterranean termite workers

also transfer moisture to their feeding sites (Su and Scheffrahn 1991 (obs.), Kirton et al.

1998 (obs.), Grube and Rudolph 1999). Additionally, they can form aerial colonies as

long they have a constant source of moisture (Ratner 1963, Potter 1994), such as from a

roof leak. Reticulitermesflavipes workers survived without any soil contact >189 days in

pine blocks with 26% wood moisture content (McManamy 2002). Areas in a structure

where moisture may build up due to condensation, such as kitchens and bathrooms, are

particularly susceptible to subterranean termite infestations (Snyder 1948).

Subterranean termites will feed on any palatable cellulose-containing material in

and around a structure, including living and dead trees, form boards or grade stakes, or

the wood frame and contents of the structure itself. Even wood preservatives may not

prevent damage to structures since Formosan subterranean termites have been known to

excavate through wood treated with chromated copper arsenate despite losing many

workers (Wilcox 1984). Su and Scheffrahn (1991) found that subterranean termites

consumed the centers of wood beams that were treated with 2,500 ppm borate spray. In

choice tests with untreated wood, Duryea et al. (1999) found both R. flavipes and R.

virginicus preferred feeding on sapwood versus heartwood. Feeding on mulches by

subterranean termite workers was also investigated. No-choice tests with mulches

indicate that both R. flavipes and R. virginicus did not feed on melaleuca mulch, while

feeding and survivorship was high for pine, cypress, and eucalyptus mulches (Duryea et

al. 1999).

Subterranean termites can enter structures through cracks in the foundation

(Meder 2000), or from behind exterior cladding stucco that extend below grade(NPCA

1980), and/or insulation (Smith and Zungoli 1995a and b). Coptotermes acinaciformis

workers crawled through 1.5 mm-wide cracks in concrete slabs (Lenz et al. 1997). Cracks

this wide may occur during normal settling of a structure over time (NPCA 1980).

Different slab types may also play a role in allowing termites to enter a structure, as

supported and floating slabs have more access points for termites than single-pieced

monolithic slabs (NPCA 1980). Cracks in the mortar of brick masonry or cracked stucco

shrinking away from the frame could also direct foraging termites in a structure (Snyder


In Florida, structural protection against subterranean termites begins with soil

termiticide treatments during the construction of a house. This preconstructionn"

treatment is done in several steps: first, horizontal and interior vertical soil treatments are

completed within the formed foundation before the concrete slab is poured; second,

exterior vertical treatments are applied to soil adjacent to where any slabs are to be

poured; finally, another exterior vertical treatment is placed around the perimeter of the

house after completion of landscaping. The soil termiticides currently used are

categorized into three groups (Su et al. 1982): type I, primary toxic repellents, such as

fenvalerate, bifenthrin, permethrin, and cypermethrin (Su and Scheffrahn 1990b); type II,

toxic but not necessarily repellent, such as chlorpyrifos (Su et al. 1982, Su and Scheffrahn

1990b), which has been withdrawn from the market due to environmental concerns; and

type HI, slow-acting toxicants which are nonrepellent, such as imidacloprid (Kuriachan

and Gold 1998), fipronil (Osbrink et al. 2001), and chlorfenapyr (Wagner et al. 2003).

Most soil termiticides are required to provide wood protection for 5 years in at least 3

United States Department of Agriculture Forest Service (USDA-FS) field sites before

EPA and state registrations are granted (Kard et al. 1989, Wagner 2003). (Fast-tracked

candidates may require fewer years.) Nevertheless, preconstruction treatment failures are

common due to improper treatments, disturbance of the treated soil, or construction that

may provide termites hidden access into the structure (pers. obs.). Also, homeowners may

disrupt termiticide barriers by digging up treated soil or by allowing gutter downspouts or

air conditioner condensation to drip or discharge onto termiticide treated soil. Termites

can breach treated soil through untreated gaps (Forschler 1994, Kuriachan and Gold


Mandatory preconstruction treatment is a very recent requirement by the State of

Florida. The new state-wide building code went into effect in March of 2002 and was

modeled after a change in the building code of St. Johns County (St. Augustine area,

northeast Florida). The chief building official of St. Johns County appended a Termite

Protection Ordinance (TPO) to the county code, effective April 28, 1996, in an effort to

reduce the following conditions that are conducive to subterranean termite infestations:

wood-to-ground contact, hidden termite access to structures, and moisture around the


The effects of the St. Johns County building code change, and subsequently the

state-wide Florida Building Code, on reducing subterranean termite infestations are

undocumented. The purpose of this study was to use a survey to determine how the TPO

of St. Johns County affected subterranean termite infestation rates, construction types,

and house maintenance characteristics. County and pest management professional (PMP)

treatment records were inspected to evaluate the effects of house age, political boundary,

preconstruction chemical soil treatment, and other factors relating to soil moisture,

cellulose around the foundation, and hidden subterranean termite access on structural

infestation rates.

Materials and Methods

Survey Population

The survey was conducted in three coastal, contiguous Northeast Florida counties.

Political boundaries were Flagler County, St. Johns County, and Jacksonville Beach in

Duval County. The survey was mailed to all single-family houses built between 1994 and

1998, totaling 12,868 surveys (7,147 to St. Johns County, 5,113 to Flagler County, and

608 to Jacksonville Beach). Surveys were considered to be delivered to homeowners if

they were not returned unopened and marked as undeliverable by the postmaster.

Delivered surveys totaled 12,027.

Houses built between 1994 and 1998 in St. Johns County were used in the survey

to determine the effect of the county building code change, effective in 1996, on

subterranean termite infestation rates. Houses from the northern and southern adjacent

counties, Duval and Flagler, respectively, built within the same time period were chosen

for comparison.

Survey Method and Format

A written survey, cover letter, and business-reply postage-paid return envelope

were mailed in May, 2000 to all single-family homes built between January 1, 1994 and

December 31, 1998. For identification, each survey was marked with the building permit

number corresponding to the mailing address. The cover letter was on University of

Florida letterhead, briefly explained the survey rationale, and identified the University of

Florida as the research institution, and provided contact information. It requested

homeowners to complete and return the survey in the envelope provided. To induce

homeowner participation, the text of the cover letter included an incentive: construction

and home maintenance practice recommendations for reducing termite infestations

resulting from this research were to be mailed to survey participants (Appendix G).

Homeowners were promised anonymity with regard to any publications resulting from

this research. All mailed material was photocopied text on white paper. Survey question

answers were either multiple choice, fill in the blank, or list format. The homeowner was

asked to mark those aspects which occurred on or around the house (Fig. 3-1).

Survey questions regarding type of foundation, frame, and exterior cladding were

chosen to categorize houses and illustrate construction trends over political boundaries

and time. Questions relating to subterranean termite access, wood proximity to the

foundation, and water near the foundation were chosen based upon factors that influence

subterranean termite infestation likelihood. The questions about exterior perimeter

inspection gap, air conditioner condensation drip line, gutter downspout, and irrigation

and lighting lines were phrased to reflect the TPO appended to the St. Johns County

Building Code, effective April 28, 1996, which mandated a minimum distance of 2 feet

from the foundation for these structural features. (See Table 3-1 for a list of modifications

to the St. Johns County Building Code. The TPO was amended again in 1998 (Table 3-1).

Features mandated by the second amendment were not applicable to this research.

House Classifications and Survey Answers

Houses were classified according to political boundary, age, building code (1996

TPO) for houses in St. Johns County, and vertical preconstruction chemical soil

treatment. Political boundary was determined by address. Age was defined as year home

construction began and was determined by county-issued building permit number.

Building code for houses in St. Johns County was also determined by building permit

number. Vertical preconstruction soil chemical treatment information was available for

53.60% of the respondent houses (from pest control companies and county building

records). (More records contained vertical treatment information instead of horizontal

treatment information, so only vertical treatment data was included for analysis.) Survey

response and infestation rates are provided for houses classified according to political

boundary, age, and building code for St. Johns County houses. Infestation rates are also

provided for houses classified by vertical preconstruction treatment.

Infestation rates are also provided for houses according to responses to questions

regarding construction type and maintenance (Fig. 3-1). Questions regarding construction

included identification of foundation type, frame type, and exterior cladding type.

Maintenance questions included those regarding inspection gap, bathroom plumbing

access port, sprinklers, and identification of items within 2 feet of the foundation.

Construction type frequencies among houses were categorized by political

boundary, age, building code for houses in St. Johns County, and preconstruction soil

chemical type. Cladding types defined as stucco included those covered with shell-dash.

Siding exteriors included all paneling (i.e. vinyl, aluminum, wood, concrete). Brick

veneer included houses covered with stone veneer.

Influence of the Building Code Change on St. Johns County Houses

Influence of the St. Johns County Building Code change was evaluated for three

house characteristics: gutter downspout discharge length, air conditioner condensation

drip line discharge length, and bottom foundation perimeter inspection gap around houses

with exterior cladding other than veneer. Those characteristics were the only building

code changes for which the mailed survey was used to verify implementation.

Tvyes of Subterranean Termite Evidence in Infested Houses

Infested houses were categorized by type, location, and time infestation evidence

appeared, as noted on the survey by respondent homeowners. Homeowners may have

indicated more than one type and/or location of infestation evidence. Types of infestation

evidence were either alates, damage, or mud tubes. Locations included doorframe,

window frame, bathroom, kitchen, fireplace, foundation, roof, and inside walls. Time for

evidence to manifest was defined as amount of time, in years, from start of house

construction until evidence of subterranean termites was noticed. Among infested houses,

amount of time until evidence of infestation was classified by both preconstruction soil

chemical treatment type and type of evidence.

Pest Control Company Public Relations

Homeowner knowledge of the pest control company that pre-treated their home

was compared among houses of different ages and different political boundaries. Contract

renewal rate was determined for homeowners who knew who their preconstruction

treatment pest control company was.

Survey Verification

The survey was evaluated for bias and robustness. Survey bias was detected by

comparing both house classifications and survey answers from voluntarily returned

surveys with respective classifications and survey answers collected from 100

homeowners surveyed by telephone. Homeowners contacted by telephone had received

the mailed survey but did not complete and return it. Clarification of survey questions to

homeowners was provided if needed. Telephone surveys were divided so that 50

telephone calls were made to St. Johns County and 25 calls each were made to Flagler

County and Jacksonville Beach. These homeowners were chosen by random number

generation (SAS Institute 2000).

Survey robustness was evaluated by determining correct answers to survey

questions regarding construction type and maintenance characteristics verified during site

inspections of 35 infested premises belonging to St. Johns County homeowners who

voluntarily answered the mailed survey. Inspections were conducted July through August

2001 for all premises to which access was granted by homeowners. With the exception of

year construction began, respective homeowner survey answers were verified by both

inspecting each of the 35 premises and interviewing homeowners. Construction years

were identified by building permit number, not by respondents' answers.


Respondent houses were classified as "infested" if homeowners answered "yes" to

the question of infestation on either the returned survey or telephone survey. Houses for

which homeowners answered "no" or "don't know" to the question of infestation were

classified as "non-infested", indicating that homeowners had not found evidence of

subterranean termites in their homes. Houses for which that question was left unanswered

were not included in the analysis. Answers from telephone surveys were pooled with

answers from voluntarily returned surveys for all analyses, except those relating to survey


Differences in proportions among classification frequencies and answers to

individual survey questions regarding construction and maintenance from both

voluntarily returned and telephone surveys were detected at the 10% significance level for

either the chi-square test or Fisher's exact test (SAS Institute 2000) when total number of

houses to be included in analysis was >200. Analysis was performed at the 5%

significance level when total number of houses to be included in analysis was < 200.

Fisher's exact test was performed in lieu of chi-square tests when tables contained

expected cell frequencies < 5 (Schlotzhauer and Littell 1997).


Survey Response Rates

Of the delivered surveys, 2,486 (20.67%) were completed by homeowners and

voluntarily returned. Response rates significantly differed among both political boundary

and house age (Table 3-2). Among political boundaries, St. Johns County returned the

highest proportion of delivered surveys (22.36%). Within St. Johns County, however,

response rates did not vary by building code (Table 3-2). Within the entire surveyed area,

there was a significant trend among homeowners of newer houses to return more

delivered surveys than homeowners of older houses.

Infestation Rates

Infestation rates significantly differed among political boundary, house age, and

building code for St. Johns County houses (Table 3-2). While St. Johns County and

Jacksonville Beach to the north had similar infestation rates (15.85% and 17.60%,

respectively), Flagler County (southemly adjacent to St. Johns County) had a lower

infestation rate among surveyed houses (2.70%). Among houses of different ages, newer

houses tended to have lower infestation rates. The largest increase in infestation rate

occurred between 3 and 4 year old houses; 4 year old houses had 3x the rate of 3 year old

houses. Within St. Johns County, infestation rates also significantly differed by building

code, probably due in part to the confounding effect of house age.

A highly significant association occurred between vertical preconstruction

treatment and infestation rates: 15.89% of houses treated with repellent chemicals were

infested compared to a 9.00% infestation rate among those treated with nonrepellent

chemicals (Table 3-3). (Vertical preconstruction treatment information was available for

1,386 out of the 2,586 survey respondents. Therefore, only 120 out of the 300 infested

houses were included in this analysis.) Further statistical analysis revealed significance to

lie within the repellent chemicals with significantly different infestation rates among

termiticides within the group (Table 3-3). Conversely, nonrepellent treatments did not

have significantly different infestation rates, possibly because almost all treatments

(96.3%) were chlorpyrifos.

Amount of time needed for evidence of infestation to be found by homeowners of

houses treated with repellent compared to nonrepellent soil termiticides was not

significantly different at the 5% significance level (P = 0.0716, Fig. 3-2). However, a

trend was observed for repellent chemicals to fail sooner than nonrepellent chemicals.

The median and mode for time to evidence of subterranean termites was between 2 and 3

years for repellent termiticides and between 3 and 4 years for nonrepellent termiticides.

Infestation rates also differed significantly among certain construction and

maintenance features (Table 3-4). The infestation rate of wood frame houses was almost

five times higher than that of concrete block houses, and houses covered with

combination siding/stucco had the highest infestation rate among the different cladding

types (23.53%). Infestation rates were significantly higher for houses with wet exterior

walls due to sprinklers, and for houses that had air conditioner condensation drip lines,

gutter downspout discharge, or trees/shrubs/stumps within 2 feet of the foundation.

Access to inspect bathroom plumbing was also significantly associated with infestation


The infestation rate of concrete slab houses (11.66%) was not significantly

different from the rate among crawl space houses (11.63%). Presence of a perimeter

inspection gap, gutter type, and sprinkler/irrigation/lighting lines, mulch, wooden fence

post, and firewood within 2 feet of the foundation all had no significant effect on

infestation rates.

Construction Types

Foundation, frame, and exterior cladding types significantly differed among

political boundaries (Table 3-5). Most respondent houses (98.10%) were concrete slab-


on-grade, with the highest frequency in Flagler County. This county also had the highest

frequency of above-ground basements (0.24%), and Jacksonville Beach had the highest

frequency of crawl space foundations (2.48%). For wall construction, 53.55% of houses

were made of wood frames. However, while there were more wood frame than concrete

block houses in both St. Johns County and Jacksonville Beach, the opposite occurred in

Flagler County. Among exterior cladding types, stucco was most prevalent (66.58%).

While the majority of houses in both Flagler and St. Johns Counties were stuccoed

(70.42% and 66.69%, respectively), Jacksonville Beach had almost the same number of

houses with stucco cladding (37.29%) as with exterior siding (33.90%).

Houses classified by age had significantly different proportions of foundation and

frame types, while exterior cladding types did not differ (Table 3-6). The frequency of

concrete slabs increased slightly in newer construction: 97.95% in 6 year old houses to

98.68% in 2 year old houses. The frequency of concrete block walls, however, increased

more dramatically in newer construction: 35.16% in 6 year old houses to 53.71% in 2

year old houses.

Within St. Johns County, most houses had concrete slabs (97.64%), with wood

frames (72.34%), and stucco cladding (66.69%). Only wall construction significantly

differed by building code (Table 3-7), with a trend towards more concrete block houses

and less wood frame houses in newer construction.

Among both repellent and nonrepellent preconstruction soil treatments, most

houses had concrete slabs (98.41%), wood frames (68.72%), and stucco (67.51%; Table

3-8). However, proportions of both wall construction and exterior cladding types

significantly differed among treatments. More wood frame houses were treated with

repellent than nonrepellent termiticides (73.87% versus 63.64%, respectively). More

stucco houses were treated with nonrepellent than repellent termiticides (69.61% versus

65.40%, respectively).

Influence of the Building Code Change on St. Johns County Houses

The new St. Johns County building code contained provisions to increase

discharge distances of both gutter downspouts and air conditioner condensation drip lines

from the foundation, and mandate an inspection gap around the bottom perimeter of

houses covered with cladding other than brick or stone veneer. Results of the survey

indicated that gutter downspout discharge distance did not change in accordance with the

building code, while both air conditioner condensation drip line discharge distance and

presence of an inspection gap did (Table 3-9). However, even though the code changes

seemed to influence drip lines and inspection gaps, >43% of the surveyed houses in St.

Johns County built after the TPO took effect did not meet the new code requirements.

Types of Subterranean Termite Evidence in Infested Houses

The evidence most frequently cited by homeowners was subterranean alates

(47.77%), and the most frequently cited area was a window frame within the interior of

the house (23.39%; Table 3-10). Types of evidence were found in significantly different

proportions around and inside houses. Alates were found more often on an inside window

than any other location. Most damage was reported in door frames, adjacent to slabs and

inside windows and interior walls. The highest proportion of mud tubes were associated

with foundations.

Alates, damage, and mud tubes (>28%) were most frequently found by

homeowners between 3 and 4 years after construction began (Fig 3-3). Evidence of


infestation was noticed within one year of construction in approximately 7 to 10% of the

infested houses.

Pest Control Company Public Relations

Most homeowners were able to identify the pest control company that treated their

home during the construction process (88.53%) and most renewed that contract (82.23%;

Table 3-11). Homeowner knowledge of their pretreatment pest control company

significantly differed among house ages and political boundaries (Table 3-11). More

homeowners of newer than older houses and more homeowners in St. Johns County than

both Flagler County and Jacksonville Beach knew the pest control company that pre-

treated their homes.

Owners of infested houses within the surveyed area relied on professional pest

control services to protect their homes from further subterranean termite damage. The

survey indicated that 97.83% (270 out of 276 owners of infested houses who answered

the survey question) called for professional remedial service and 95.56% maintained the

warranty on that work. According to homeowners who answered the question, mean

treatment cost was $1,075.78 120.22, with a median of $850, and a range of $0 (work

was covered by the original preconstruction contract) to $8,500. Mean damage repair cost

was $907.73 139.39, with a median of $400, and a range of $0 to $5,000.

Survey Verification

Among houses classified by political boundary, significantly higher proportions of

voluntarily returned surveys came from Flagler and St. Johns Counties versus phone

surveys, while a lower proportion of surveys were returned by homeowners in

Jacksonville Beach than those surveyed by phone (Table 3-12). Chi-square analysis found


no significant difference in proportions of voluntarily returned surveys and phone surveys

classified by age. Proportions of both foundation types and exterior cladding types

significantly differed, with most types occurring in greater proportions among voluntarily

returned surveys than phone surveys (Table 3-12).

Among other survey questions, bias (indicated by higher proportions of positive

answers among phone surveys than returned surveys) was found for air conditioner

condensation drip line, wooden fence post, firewood, trees/ shrubs/ stumps, pest control

contract renewal, and the question of subterranean termite infestation (Table 3-13).

Homeowners answered "don't know" most frequently (20.29%) to the question regarding

bathroom plumbing inspection access.

As for survey robustness, mean percentage of correct answers was 84.08% for

structural physical attributes verified by site inspections (Table 3-14). All homeowners

correctly identified their homes' type of foundation, frame, and cladding, and all were

able to identify whether or not firewood was < 2 feet from their foundation. The least

reliable survey question, of which only 20% of homeowners answered correctly, was

presence or absence of access to inspect bathroom plumbing (Table 3-14). Most

homeowners were unable to identify their bathroom inspection ports. The second least

reliable question was that of construction year. Among all surveyed homeowners, only

64% answered this question correctly after verification using building permit numbers.

Therefore true construction year as indicated by permit number was used to classify these

houses correctly, instead of using survey answers.


Survey Response Rates

While many factors affect mail survey response rates, salience of the

questionnaire topic to the survey audience was one of the strongest (Heberlein and

Baumgartner 1978, Goyder 1982). When included as a predictor variable used during

their meta-analysis (quantitative analysis of contribution of independent factors to

response rates) of 98 independent published studies, Heberlein and Baumgartner (1978)

reported their salience scale to be directly correlated with return rates. For example, non-

salient mailed surveys had 42% response while surveys judged to be salient had 77%

return (Heberlein and Baumgartner 1978). However, salience is debatable among social

scientists, since it is more subjective rather than objective (Goyder 1982, Fox et al. 1988).

Subterranean termites appeared salient in St. Johns County, considering the

history of the area. Public concern over the termite problem in the County led to

amendment of the St. Johns County Building Code with the TPO (Shaheen 1996).

Therefore, it is not surprising that the return rate from St. Johns County was higher

(62.71%) than those of Flagler County (33.27%) and Jacksonville Beach (4.02%; Table

3-2). Perhaps St. Johns County homeowners knew about subterranean termites before,

either from personal experience or from the local media. The difference in response rates

from homeowners of different-aged houses was probably due to a lower number of

owners of older houses who could correctly answer questions. Perhaps they misplaced or

forgot specific information about their homes which led them to disregard the survey.

Survey response rate, 20.67%, was low compared with those published from other

mail surveys. However, direct comparison with those response rates is difficult, since

social scientists send their surveys to target populations. For example, surveys sent

exclusively to professionals and university-educated persons achieved approximately

18% and 5.5%, respectively, higher return than surveys sent to others (Goyder 1982).

Male populations yielded higher returns than female-only or mixed populations, and

responses were 13% lower in urban areas than in rural or mixed populations (Goyder

1982). This survey did not target any specific demographic. The surveyed population,

although not assessed, was assumed to be a mixed group; it included both urban and rural

residents, and men and women of different ages and education levels.

In addition to salience, other factors that increase mail survey response rates

include repeated contacts (Goyder 1982, Fox et al. 1988), inclusion of a stamped reply

envelope (Linsky 1975, Armstrong and Lusk 1987, Fox et al. 1988, Yammarino et al.

1991), postage type (Linsky 1975, Yammarino et al. 1991), incentives (Linsky 1975, Fox

et al. 1988, Yammarino et al. 1991), sponsorship (Heberlein and Baumgartner 1978, Fox

et al. 1988), and length and appearance of the questionnaire (Heberlein and Baumgartner

1978). Anonymity and confidentiality did not to influence response rates (Yammarino et

al. 1991, Groves et al. 1997). Researchers should be able to increase their response rates

by mailing a cover letter that includes appeals or incentives with a survey of less than four

pages (Yammarino et al. 1991). Surveys longer than four pages reduced response rates by

7.8%, while both appeals and reply envelopes increased response rates by 4.7% and 7.9%,

respectively (Yammarino et al. 1991). Institutional versus commercial sponsorship also

increased survey return rates (Fox et al. 1988). For institution-sponsored surveys,

inclusion of stamped/metered return envelopes increased response rates by 6.1%

(Yammarino et al. 1991). However, first class return postage, as opposed to the more

cost-effective business reply rate, was reported to increase response rates by 9%

(Armstrong and Lusk 1987). Green paper versus white paper resulted in a range of

effects, varying from a decrease in response rates of 5.6% to an increase of 9.1% (Fox et.

al 1988). This survey was moderately salient, institutionally-sponsored, less than four

pages, and included an incentive and a business-reply return envelope. Response rates

might have increased had follow-up phone calls been made to homeowners who did not

return the survey, or if first-class stamps been included instead of business-class rate.

Infestation Rates Among House Classifications and Construction Types

The three surveyed areas each have similar environments and are on the Atlantic

coastline. For each, sandy soils predominate to a depth of 76.20 cm (30 in), average

temperature is approximately 20.56C (69" F), and average rainfall is approximately 1.35

m (53 in) (Readle et al. 1983, Readle et al. 1997, Watts et al. 1998). The climate is

subtropical with long, warm, humid summers and mild winters. Approximately 60% ot

the rainfall occurs June through October (Readle et al. 1983, Readle et al. 1997, Watts et

al. 1998).

Given the close proximity of houses to each other within the surveyed area, the

differences in infestation rates in Jacksonville Beach / St. Johns County versus Flagler

County was unexpected (Table 3-2). These may have been due to termite pressure, which

was not measured and may have been influenced by previous land use.

Another explanation for these different rates may be differences the in

construction types, as each had different proportions of foundations, wall construction,

and exterior claddings. (Foundation type, however, was not significantly associated with

infestation rates.) Wall construction also differed among political boundaries. Flaglei

County, for example, had more houses made of concrete block walls than wood frames,

which is opposite of St. Johns County and Jacksonville Beach. The strong association

between frame type and infestation status indicated that the lower infestation rate in

Flagler County could have been due to the decreased number of wood frame houses

compared with the other two areas.

Houses covered with a combination siding/stucco cladding had the highest

infestation rate. While this may have been an artifact of small sample size of these

houses, of important note is that Jacksonville Beach, which had the highest overall

infestation rate, also had the highest percentage of siding/stucco houses.

The construction process itself may have disrupted existing subterranean termite

colonies by destroying either part of or the entire colony during land clearance. Homes

with evidence of infestation in less than 3 years may have been built over an existing

colony. Any portion of the colony not decimated during construction could continue as a

satellite of the original colony. Pawson and Gold (1996) reported that R.flavipes and R.

virginicus pseudergates separated from colonies differentiated into supplemental

reproductive and produced eggs and nymphs within five and seven months, respectively.

Therefore, it is not surprising that infestations were reported shortly after construction


Additionally, higher infestation rates occurred in older houses probably because

foraging termites are more likely to find an entrance to a structure with time. Henderson

et al. (1998a) reported that, after 13 months, approximately 50% more stakes placed in

areas around structures assumed to be conducive to subterranean termites (prone to high

moisture or had readily available cellulose sources) were attacked versus stakes placed in

undirected patterns. Presumably, termites would have continued to forage in conducive

areas and may have eventually gained structural access.

Another explanation for the differences in infestation rates of older verses newer

houses has to do with wall construction. While almost 5x more wood frame houses were

infested than concrete block houses, less of the newer homes were built with wood

frames. Therefore, since infestation is more likely to occur in wood frame houses, perhaps

infestation rates were higher for older houses because more of them had wood frames.

Wood frames of slab-on-grade houses are usually within 20 cm (-8 in) of the

ground (Allen 1999). The close proximity of wood frames to the ground and the cryptic

nature of subterranean termites with their ability to crawl through small cracks (Lenz et

al. 1997) make untreated wood in frame houses easily accessible.

The higher infestation rates in St. Johns County according to building code is

probably an artifact of time. The new building code took effect in 1996, four years before

this survey was conducted. The decrease in rates, from 27.79% for houses built under the

old code, to 7.10% for houses built under the new code (Table 3-2), reflected the

decreased rates of 4 year old houses versus 2 to 3 year old houses. Also, the higher

proportion of wood frame versus concrete block houses built under the old code may

have also influenced infestation rates in older houses.

Survey results revealed that repellent soil termiticides were less effective than

nonrepellents as pre-treatments. Among repellents, excluding both bifenthrin and

fenvalerate, which had small sample sizes, cypermethrin performed better than

permethrin. The three nonrepellents performed similarly, although this may have been an

artifact of small sample sizes for fipronil and imidacloprid.

Variable performance levels were reported for repellent termiticides as a group by

Wagner et al. (2003) in Florida USDA-FS field tests. For concrete slab tests,

cypermethrin (0.5%) provided 5 years of 100% efficacy, fenvalerate (0.5%) provided 3

years, permethrin (0.5% Dragnet) provided 6 years, and bifenthrin (0.062%) provided 16

years (Wagner et al. 2003). Nonrepellent termiticides performed consistently, however, as

chlorpyrifos (0.5% and 1.0%), imidacloprid (0.05%), and fipronil (0.06%) all provided at

least 5 years of 100% protection of wood within concrete slabs (Kard 2000). (These

formulations and concentrations reflect the majority of repellent termiticides and

concentrations used for vertical preconstruction treatments of respondent houses

(Appendix H)).

Different infestation rates among the two termiticide treatment types could have

also been caused by wall construction bias, as wood framed houses comprised 73.87% of

repellent treatments, but only 63.64% of the nonrepellent treatments (Table 3-8).

Proportions of brick exterior claddings on repellent-chemical houses were also higher

than on nonrepellent-chemical houses, which may have also contributed to bias. Brick

sidings had higher infestation rates than houses with no brick, excepting siding / stucco

combination houses (Table 3-4). Additionally, variability in either repellent or

nonrepellent treatments could have resulted from applicator errors or formulation

differences. Therefore, different proportions of frame and exterior cladding types,

possible human error, and formulation differences make it difficult to conclude with

certainty if nonrepellent chemicals out-performed repellent chemicals.

Statistical analysis of the 120 infested houses for which termiticide treatment type

was known indicated that there was no significant difference at the 5% significance level

between the time for evidence to manifest in houses treated with repellents versus

nonrepellents. However, there seemed to be a trend for homeowners of repellent

termiticide-treated houses to find termite evidence sooner than their nonrepellent cohorts.

This data must be interpreted cautiously since more homeowners surveyed by telephone

indicated their homes were infested compared to those who voluntarily returned the

survey by mail.

Difference in time to failure, if any, could have been due to the nature of the

chemicals. Repellents form a barrier between the structure and the termite colony,

without killing the insects. Nonrepellent treatments kill the termites after they contact the

treated soil. Reticulitermesflavipes foragers continuously tunnel, and will change

tunneling direction after contacting soil treated with repellent chemicals (Forschler 1994).

This could ultimately lead termites to find a gap in the treatment. The minimum gap size

in a repellent soil treatment that R. flavipes located varied between 3 and 4 cm (Kuriachan

and Gold 1998), a width that could easily result from soil disturbance, such as that caused

by gutter downspout discharge or growing tree roots. Perhaps, repellent termiticide

failures were seen sooner than the nonrepellent termiticides because low termite mortality

enabled more surviving foragers to locate gaps. Another explanation could be that some

pyrethroids were applied at the low label rate which did not obtain 5 years of control in

USDA-FS field trials.

Infestation Rates Among House Maintenance Characteristics

Maintenance characteristics that were moisture-related and inquired of

homeowners on this survey included if sprinklers wet the house, gutter type, air

conditioner condensation drip line length, gutter downspout length, and if mulch is < 2 ft

from the foundation. Only sprinklers, gutter downspouts, and air conditioner

condensation drip lines were significantly associated with infestation rates (Table 3-4).

For the sprinklers and drip line, consistent irrigation during hot weather probably

maintained the soil moisture necessary for termites to continue exploiting a resource

(Haagsma and Rust 1995). Water from the gutter downspout, may have eroded enough

treated soil to allow termites access to a structure without affecting the colony. Lack of

gutters was not significantly associated with infestation rates perhaps because bare roofs

may cause rainwater to spread at the roof drip line, which is outside the area of vertical

soil treatment. Lack of significant association between infestation and mulch near the

foundation was unexpected, since mulched soils are presumed to buffer against

temperature changes and conserve moisture more then bare soils (Fraedrich and Ham

1982, Smith and Rakow 1992). This buffering effect may be superficial; according to a

study done by Long et al. (2001), neither organic nor inorganic mulch affected either

temperature or moisture level in the soil at 12 cm below the surface. This survey did not

discern between organic and inorganic mulches -- all houses with mulch near the

foundation were pooled. According to NPMA (1999), mulch supposedly discourages

termite activity by speeding soil drainage and drying.

Among structurally-associated cellulose sources in the survey, wall construction

had the only significant association with infestation. Firewood, wooden fence post, and

mulch were not significantly associated with infestation. This suggests that neither

firewood nor a wooden fence post led to infestation in houses aged 2 through 6 years,

although these features occurred with low frequency. Also, although organic mulch is a

cellulose source, it provides inadequate nutrition for subterranean termites (Duryea et al.

1999, Long et al. 2001), perhaps leading the termites to feed on other materials.

For house characteristics which allow for visual identification of termite

infestation or are related to termite access, only bathroom plumbing access port and

tiees/shrubs/stumps near the foundation were significantly associated with infestation.

The bathroom plumbing access port was marginally significant (P = 0.0979) and the high

rate of don't know's and non-answers (20.53%) made this surveyed characteristic

unreliable because homeowners may not have understood the question. For the

trees/shrubs/stumps, perhaps the root systems served as guidelines to cracks in building

foundations. Decomposed root systems from stumps may have created pre-formed

tunnels. Irrigation and lighting lines in the ground near a structure, although potential

guidelines for termites, were not significantly associated with infestation. This is

consistent with results of Pitts-Singer and Forschler (2000), who found that laboratory

cultures of both R. flavipes and R. virginicus almost always followed pre-formed tunnels,

but did not follow wires as readily. Not surprisingly, presence of a foundation perimeter

inspection gap was also not significantly associated with infestation rate, probably

because it only allows early detection and is not a deterrent.

Influence of the Building Code Change on St. Johns County Houses

St. Johns County enacted the TPO in hopes of reducing subterranean termite

infestations in new houses (Shaheen 1996). This survey indicated that gutter downspout

length did not change when the new code took effect, and 63.08% of houses built after

April 1996 had downspouts that did not comply with the new law. There were also less

houses built with air conditioner drip lines near the foundation after the code changed, but

43.01% still did not meet code the requirement. Significantly more houses were built with

the required foundation perimeter inspection gap after the new code went into effect than

were built before the code change. However, only 44.06% of the houses met this building

code specification.

These results indicate that surveyed St. Johns County houses were not constructed

according to the Ordinance for two out of the three code requirements evaluated in this

survey. This was verified by site inspections of 35 infested premises in St Johns County,

as explained in Chapter 4 (Table 4-2). In fact, none inspected houses that were built in

1997 met the Ordinance requirements for any measurements (Chapter 4). If the new

building code requirements did not influence these visible characteristics, compliance

with code requirements not visible after construction was completed, such as burial of

wood away from the structure or veneer bearing ledge poured integrally with the main

foundation, may have also been low (See Table 3-1 for St. Johns County Building Code


Types of Subterranean Termite Evidence in Infested Houses

Homeowners cited alates as the most frequent evidence of subterranean termite

infestation. In a survey of Kentucky homeowners, less than 19% mentioned mud tubes as

evidence of infestation, while 56% mentioned damaged wood and 39% mentioned

swarms (Potter and Bessin 2000). Therefore, it is not surprising that damage and alates

were more frequently cited than mud tubes in this survey.

In an unpublished phone survey (Shimberg Center, Policy and Management

Research, 2001) commissioned by the Florida Department of Agriculture and Consumer

Services (FDACS), alates were the most common sign of infestation. Mud tubes were

cited as the second most common sign of evidence, and damaged wood the third. This

differs from the current survey, in which homeowners indicated evidence as occurring

from highest to lowest: alates > damage > mud tubes. Differences could be due to sample

size: 300 (out of 2,586 survey respondents) versus 41 (out of 602 interviews) for the

FDACS survey. Also, the definition of damage may have differed between the two

surveys. Both damaged wood and exit holes in walls made by swarmers counted as

damage in the current survey, while the FDACS survey included only damaged wood.

Most homeowners found termite evidence inside their homes associated with a

window. This is not surprising since alates are attracted to light (Snyder 1948). On

outside facing walls, exterior landscaping may have disturbed the perimeter soil treatment

and allowed termites to cause damage or construct tunnels. For houses where the

evidence was found on the exterior, most was in or near a door frame, including those in

garages and patios. These door frames are associated with abutting slabs or driveways

where cold joints are formed. Termite infestations in door frames may be the result of

inadequate termiticide application or disturbance of termiticide treated soil prior to the

pouring of secondary slabs abutting the main foundation.

Pest Control Company Public Relations

Most homeowners (88.53%) were able to identify the pest control company that

treated their home during the construction process, and most renewed their contract with

that company (82.23%; Table 3-11). Differences due to age of house and political

boundary in the percentages of homeowners who knew the company that did the

termiticide pre-treatment were probably due to local building codes. The St. Johns

County TPO required a Notice of Treatment to be secured on or near the fuse box in the

house. Therefore, it was not surprising that a greater percentage of owners of newer

houses knew the company. Also, as expected, more homeowners in St. Johns County

knew of their original pest control company compared with those living in the other

surveyed areas.


The FDACS survey (Shimberg Center, Policy and Management Research, 2001)

also reported a high percentage (80.2%) of homeowners who knew their homes were

treated for subterranean termite control at the time of construction. However, only 41.8%

were able to name the treating company. This survey differed in that it was conducted

state-wide with homes ranging from 1 to 10 years old.

Owners of infested houses within the surveyed area relied on professional pest

control services to protect their homes from further subterranean termite damage, as

demonstrated by the high percentage of homeowners who called for professional service

(97.83%) and maintained the warranty on that work (95.56%). In a nation-wide survey of

1,100 homeowners, Bayer Crop Science reported that only 72% of the homeowners who

had a termite problem called for professional service (Harbison 2000). Perhaps the

discrepancy between the two surveys was due to the differences in the two geographical

areas. These numbers, however, demonstrate that professionals have a large share in the

termite control market, in spite of availability of over-the-counter consumer products for

subterranean termite control.

Survey Verification

Bias occurred among political boundary and was due to different ratio of

voluntarily returned versus phone surveys for the three areas (Table 3-12). For future

surveys, phone surveys among different political boundaries should be in equivalent ratio

to mailed surveys. Among construction types, bias occurred for foundation and exterior

cladding types and was probably caused by the different proportions of construction types

among political boundaries.

Among survey questions, bias occurred for air conditioner condensation drip line,

wooden fence post, firewood, trees, shrubs, and stumps, and pest control contract

renewal, and the question of subterranean termite infestation (Table 3-13). Although

explanation of survey questions was minimal while conducting phone surveys, the bias

detected by the chi-square tests indicate that perhaps an explanation was given during

phone surveys that was not given by the mailed survey. The significantly lower

infestation rate among homeowners who voluntarily returned the survey (11.65%

infested) versus those surveyed by phone (18.00% infested) indicated that perhaps

infestations were underestimated. Homeowners who had infestations in their current

homes may have been more inclined to reveal this while on the phone as opposed to

marking it on the mail-in survey.

This survey had a robustness of 84.08% for structural physical attributes that were

verified during site inspection of 35 houses (Table 3-14). Most homeowners were able to

correctly answer the survey questions (> 77%), with the exception of bathroom plumbing

access and year construction began. The numbers of homeowners who answered "don't

know", left the question unanswered, and answered incorrectly indicate that further

explanation of this character would be needed if used on future surveys. Additionally, the

researchers would have to record information for each bathroom in the house. Also, for

reliability of future surveys, the year construction began should be identified by building

permit number, or other county-issued number. Some homeowners surveyed by phone

needed further explanation to correctly answer this question, and homeowners who

voluntarily returned the survey may have actually answered with their move-in date.

Without inclusion of these two questions the survey robustness would have been 89.01%.

Overall Survey Remarks

This survey indicates that subterranean termite infestation was strongly associated

with older wood frame houses. Although the data are not conclusive, houses treated with

different repellent preconstruction termiticides had significantly different infestation rates

and failed sooner than nonrepellent termiticides.

Changes to the St. Johns County Building Code, as written in 1996 and 1998,

may not decrease the County's infestation rate due to incomplete compliance. Better

compliance and additional code requirements should lower the subterranean termite

infestation rates. This surveyed evaluated only three factors addressed by the Code, two

of which were significantly associated with infestation rate: air conditioner condensation

drip lines and gutter downspout length. A third moisture-related factor, sprinklers wetting

the house, was not addressed by the Code but was significantly associated with infestation

rate. Additionally, wood framing and cladding type, which were both significantly

associated with infestation rate, were not addressed by the new Code. These factors

should be considered in future Code amendments. The new Code, however, increased

consumer awareness of original pretreatment pest control company.

Table 3-1. Modification to the St. Johns County Building Code, the Termite Protection Ordinance (TPO).

Subject Standard Building Code' TPO 2, TPO 3,
April 28. 1996 July 22. 1998

Natural wood, vegetation, dead roots, Material shall be removed

stumps, trash, debris of things of similar
nature that could reasonably be expected to
attract organisms destructive to structures

Construction materials made from wood
(grade stakes, forms, contraction spacers,
tub trap boxes, plumbing supports, posts,
organic or termite susceptible construction

from foundation area and
fill shall be clean.
(Sec. 2301.1.2)

Loose wood, debris and
wood forms shall be
completely removed from
all spaces under building.
(Exception: wood
members of pressure-
treated or naturally
resistant wood) {Sec.
2301.1.3 & 2304.2.21

Debris inside cells and cavities of masonry Not specified.

Bottom perimeter inspection space of
foundation sidewall exterior

6 in space required
between wood siding and
earth. (Exception:
pressure-treated or
naturally resistant wood)
(Sec. 2304.2.5)



Not allowed to remain on building lot or
underground within 5 ft of structure. Fill
shall be clean. (Exception: living trees
outside the roof line) (Sec. 2304.6.1.1)

Materials must be placed in obvious
location for later removal. Material cannot
be willfully buried within 15 feet of
building. (Exception: posts of pressure-
treated or naturally resistant wood, <8 in
dimension, installed >6 in from structure
for inspection and retreatment)
(Sec. 2304.6.1.2}

Free of debris before placement of
concrete. (Exception: inorganic plugs and
clean earth fill) (Sec. 2304.6.1.3)

6 in space required below any wood,
siding, felt, wire lath, sheathing, foam
board or expanded polystyrene (included
stuccoed framed posts or columns) down
to top of sod, mulch, or soil. (Exception:
paint or stucco adhered directly to
masonry foundation) (Sec. 2304.6.2.1)

Same, but 4 in
space above
concrete or
paving allowed.


Table 3-1 Continued


Bearing ledge for brick or stone veneer

Standard Building Code'

Not specified

TPO 2,
April 28. 1996
Shall be poured integrally with the concrete
foundation. (Exception: veneer may be
carried by a structural metal member
secured to foundation sidewall, 6 in
inspection required above soil, sod, or
mulch) {Sec. 2304.6.2.2)

Sleeves around piping through concrete slab-
on-grade floors

Placement of decks, fences, patio, planters,
or other building components

Any material not harmful to Must be non-cellulose, tightly closed at both
concrete permitted. ends by wire twists, tape wraps, or stainless
{Sec. 1907.3.1) steel hose clamps.
(Sec. 2304.6.2.3)

Not specified.

Pre-construction soil chemical treatment for Not specified, at discretion
prevention of subterranean termites, of local building official.

Material for plumbing traps or other boxed- Not specified.
out spaces

Must not abut foundation sidewall so as to
obstruct 6 in inspection space. (Exception:
components with >18 in ground clearance)
(Sec. 2304.6.2.4}

Initial horizontal treatment to occur after all
excavation, backfilling, and compaction is
completed. Retreatment is mandatory if any
soil disturbance occurs after initial
treatment. Manufacturer recommendations
on label must be followed. (Sec.




Must be created with permanently placed Same.
metal, solid plastic, or masonry forms of
adequate depth so as not to disturb soil
treatment. {Sec. 2304.6.3.2)

July 22, 1998
Same, but a
physical termite
barrier must
bridge wall sill
plate to mortar

Table 3-1 continued.
Subject Standard Building Code TPO 2, TPO 3,
April 28, 1996 July 22 1998
Vapor retarder Placed beneath slab, Chemically treated soil covered by vapor Same.

minimum 0.152 mm
polyethylene with joint
lapped 152 mm and
sealed, or other approved
material. (Sec. 1909.21

Concrete overpour or mortar excess
accumulated along exterior foundation

Pre-construction soil chemical treatment
under adjacent slab, vertical soil chemical

Discharge distance for condensate drain
lines, condensing units, roof downspouts,

Not specified.

retarder within 1 h of treatment.
Retreatment of soil mandatory if
disturbance occurs. (Sec. 2304.6.3.3)

Shall be removed prior to vertical soil
treatments. (Sec. 2304.6.3.4)

Not specified, at discretion Horizontal treatment shall be applied
of local building official, under all exterior concrete on grade within
1 ft of the primary sidewalls. Also a
vertical treatment shall be applied around
the completed structure, after landscaping
and irrigation installation. Disturbance of
the treatments will require retreatment of
the disturbed area. (Sec. 2304.6.3.5)

Not specified.

Condensate drain lines, condensing units,
roof downspouts shall not place water <2
ft of the structure sidewall. Gutters
required on all eaves <12 in horizontal
projection. Sprinkler heads shall not apply
water <1 ft of structure wall. (Sec.



No water <2 ft
from structure
wall. Gutters on
eaves <16 in

Table 3-1 continued.
Subject Standard Building Code' TPO TPO ,
April 28, 1996 July 22, 1998
Pre-construction soil chemical treatment Not specified. Weather-tight and obvious treatment Same.
notice certificate (for each treatment) posted on
jobsite, copies provided for building
permit holder and building permit file.
(Sec. 2304.6.4.1)

Homeowner notification of pre- Not specified. Permanent sign of at least 3x5 in posted in Same.
construction soil chemical treatment a conspicuous place in the garage or
company, contact information, and within 3 ft of the electric panel box.
necessity for reinspection Exterior signs must be weatherproof.
{ Sec. 2304.6.4.2)

Preconstruction soil chemical treatment Not specified. Occupant or title holder shall receive an Same.
warranty and renewal unlimited 1 yr retreatment and repair
warranty with an optional 4 yr renewal
provision. {Sec. 2304.6.4.3)

1. Termite-protection related sections adapted from the 1994 Edition of the Standard Building Code by Southern Building Code
Congress International, Inc. (SBCCI 1994).
2. Adapted from the Termite Protection Ordinance appended to the 1994 Edition of the Standard Building Code by Southern Building
Code Congress International, Inc. (SBCCI 1994).
3. Adapted from the Termite Protection Ordinance appended to the 1997 Edition of the Standard Building Code by Southern Building
Code Congress International, Inc. (SBCCI 1997).

Table 3-2. Response and infestation rates of houses classified by political boundary, age, and St. Johns County Building Code
Classification N Delivered % Response 2 P Response % Infested P Infested
Political Boundary
Flagler County 4541 18.21 < 0.00014 2.70 < 0.00015
Jacksonville Beach 513 19.49 17.60
St. Johns County 6973 22.36 15.85
Average of all houses 12027 20.67 11.60
Age of House6
2 years (built in 1998) 2603 22.59 0.00707 2.14 < 0.0001
3 years (built in 1997) 2453 21.73 4.70
4 years (built in 1996) 2459 20.13 12.43
5 years (built in 1995) 2270 19.82 17.87
6 years (built in 1994) 2242 18.73 25.68
Average of all houses 12027 20.67 11.60
St. Johns County Building Code 9
Old 2981 22.81 0.4322 '1 27.79 <0.0001"
New -3992 22.02 7.10
Average for all houses 6973 22.36 15.85

1. Number of surveys delivered. Surveys were consider to be delivered if they were not returned (unopened) as undeliverable by the postmaster.
2. Percentage of voluntarily returned surveys only.
3. Homeowners who answered "don't know" to infestation status were considered to have no evidence of infestation and included in the "no"
category. Percentage includes both voluntarily returned and phone surveys.
4. P- value for chi-square test (X=29.2787, df=2).
5. P- value for chi-square test (X=98.5207, df=2).
6. True construction year based on building permit numbers.
7. P- value for chi-square test (X2=14.0807, df=4).
8. P- value for chi-square test (X2=182.1942, df=4).
9. Old = houses built 01 Jan 1994 27 Apr 1996. New = houses built 28 Apr 1996 31 Dec 1998.
10. P- value for chi-square test (X2=0.6168, df=l).
11. P- value for chi-square test (X2=126.0174, df=l).

Table 3-3. Infestation rates of houses classified by pre-construction soil chemical treatment type.

Classification or Survey Response N % Infested P
Pre-construction chemical classification 2
Repellent 686 15.89 <0.0001'
Nonrepellent 700 9.00
Average for all houses 1386 12.41
Repellent chemical treatments (no. of pest control companies)
Bifenthrin (1) 1 0.00 < 0.00014
Cypermethrin (15) 546 12.82
Fenvalerate (1) 9 44.44
Permethrin (3) 130 26.92
Average for all houses (20) 686 15.89

Nonrepellent chemical treatments (no. of pest control companies)
Chlorpyrifos (9) 674 9.20 0.7434 4
Fipronil (1) 1 0.00
Imidacloprid (4) 25 4.00
Average for all houses (13) 700 9.00

1. Homeowners who answered "don't know" to infestation status were considered to have no evidence of infestation and included in
the "no" category.
2.Vertical pre-construction treatment soil chemical information obtained from pest control companies and county building records.
3. P- value for chi-square test (X2=14.9706, df=l).
4. P-value for Fisher's exact test. Fisher's exact test is performed when the conditions for a chi-square test are not met. It is especially
appropriate for tables with expected cell frequencies < 5 (Schlotzhauer and Littell 1997).

Table 3-4. Infestation rates of houses among survey responses to construction types and maintenance-related questions
Survey Response N % Infested X2, df P
Foundation Type
Above-ground basement 3 0.00 --- 1.0000
Concrete slab 2531 11.66
Combination concrete slab and crawl space 3 0.00
Crawl space 43 11.63
Total 2580 11.63
Wall Construction
Logs 2 0.00 ---- <0.00012
Concrete block 1172 3.84
Steel 22 0.00
Wood 1379 18.49
Total 2575 11.65
Exterior Cladding Type
Brick 310 15.16 -- <0.00012
Brick and siding 39 15.38
Brick and stucco 49 10.20
Siding 379 10.03
Siding and stucco 85 23.53
Stucco 1717 10.72
Total 2579 10.07
4 6" inspection gap
Yes 847 10.98 1.0884, 1 0.3391
No 1611 12.41
Total 2458 11.96
Sprinklers wet house
Yes 838 13.72 5.1267, 1 0.0236
No 1672 10.65
Total 2510 11.67

Table 3-4 Continued
Survey Response N % Infested X2, df P
Full 1343 12.29 2.8793, 1 0.2370
Partial 222 13.51
None 994 10.36
Total 2559 11.65
Air conditioner condensation drip line
Yes 1280 12.81 3.5217,1 0.0606
No 1302 10.45
Total 2582 11.62
Gutter downspout
Yes 1047 14.90 18.4601, 1 <0.0001
No 1535 9.38
Total 2582 11.62
Access to inspect bathroom plumbing
Yes 486 13.79 2.7401, 1 0.0979
No 1569 11.03
Total 2055 11.68
Sprinkler heads, irrigation or lighting lines
Yes 1508 12.33 1.8065, I 0.1789
No 1074 10.61
Total 2582 11.62
Yes 1861 11.87 0.4268,1 0.5136
No 721 10.96
Total 2582 11.62
Wooden fence post
Yes 292 11.99 0.0444, 1 0.8332
No 2291 11.57
Total 2583 11.61

Table 3-4 Continued
Survey Response N % Infested X2, df P
Yes 42 4.76 ------ 0.2236
No 2542 11.72
Total 2584 11.61
Trees, shrubs, stumps
Yes 1742 12.63 5.4573, 1 0.0195
No 843 9.49
Total 2585 11.61

1. Homeowners who answered "don't know" to infestation status were considered to have no evidence of infestation and included in
the "no" category.
2. P-value for Fisher's exact test. Fisher's exact test is performed when the conditions for a chi-square test are not met. It is especially
appropriate for tables with expected cell frequencies < 5 (Schlotzhauer and Littell 1997).

Table 3-5. Frequencies of foundation, wall, and exterior cladding types classified by political boundary
Frequency for political boundary (% ')
Total (% Flagler County St Iohns County Jacksonville Beach P2
Foundation Type
Above-ground basement 3 (0.12) 2 (0.24) 1(0.06) 0 0.0244
Crawl Space 43(1.67) 5(0.59) 35(2.18) 3(2.48)
Concrete Slab 2531 (98.10) 843 (99.06) 1570 (97.64) 118(97.52)
Combination crawl space and concrete 3 (0.12) 1(0.12) 2 (0.12) 0
slab 2580 851 1608 121
Total number of homeowners who
Wall Construction 2(0.08) 1(0.12) 1(0.06) 0 <0.0001
Log home 22(0.85) 2(0.24) 20(1.25) 0
Steel frame 1172(45.51) 744 (87.53) 423 (26.36) 5 (4.17)
Concrete block 1379 (53.55) 103 (12.12) 1161 (72.34) 115(95.83)
Wood frame 2575 850 1605 120
Total number of homeowners who
Exterior Cladding Type
Brick 310(12.01) 88(10.33) 214 (13.30) 8(6.78) < 0.0001
Brick and siding 39(1.51) 16(1.88) 19(1.18) 4(3.39)
Brick and stucco 49(1.90) 18(2.11) 29(1.80) 2(1.69)
Siding 379(14.70) 114(13.38) 225 (13.98) 40(33.90)
Siding and stucco 85 (3.30) 16(1.88) 49(3.05) 20(16.95)
Stucco 1717 (66.58) 600 (70.42) 1073 (66.69) 44 (37.29)
Total number of homeowners who 2579 852 1609 118

1. Percentage within column group.
2. P-value for Fisher's exact test. Fisher's exact test is performed when the conditions for a chi-square test are not met. It is especially
appropriate for tables with expected cell frequencies < 5 (Schlotzhauer and Littell 1997).

Tah1P 3-6 Freniipnriea nf fmniinatinn un1Iand Inxlptprinr rladdrino tvnn' rIaaaifindl hv hnii.p aoep

Frequency for House Age (Years) (% 2)
Total (% 2) 2 3 4 5 6 P

Foundation Type
Above-ground basement
Crawl Space
Concrete Slab
Combination crawl space and concrete
Total number of homeowners who
Wall Construction
Log home
Steel frame
Concrete block
Wood frame
Total number of homeowners who
Exterior Cladding Type
Brick and siding
Brick and stucco
Siding and stucco
Total number of homeowners who

3(0.12) 0(0.00)
43(1.67) 7(1.15)
2531 (98.10) 600 (98.68)
3(0.12) 1(0.16)

6 (1.08)
545 (98.55)

0 (0.00)
12 (2.35)
499 (97.65)
0 (0.00)

0(0.00) 2(0.46) 0.08173
11(2.35) 7(1.59)
457 (97.44) 430 (97.95)
1(0.21) 0 (0.00)

2580 608 553 511 469 439

2 (0.08) 0 (0.00)
22 (0.85) 4 (0.66)
1172 (45.51) 326 (53.71)
1379 (53.55) 277 (45.63)
2575 607

310 (12.01) 84(13.84)
39(1.51) 9(1.48)
49(1.90) 13(2.14)
379(14.70) 78 (12.85)
85 (3.30) 15 (2.47)
1717 (66.58) 405 (67.22)
2579 607

8 (1.45)
282 (51.18)
260 (47.19)

54 (9.82)
8 (1.45)
86 (15.64)
376 (68.36)

\7 (1.37)
220 (42.97)
284 (55.47)

55 (10.74)
8 (1.56)
75 (14.65)
20 (3.91)
340 (67.19)

0(0.00) 0(0.00) <0.00014
2 (0.43) 1(0.23)
190(40.68) 154 (35.16)
275 (58.89) 283 (64.61)
467 438

57(12.13) 60(13.64) 0.81935
6 (1.28) 9 (2.05)
7 (1.49) 11(2.50)
73 (15.53) 67 (15.22)
18 (3.83) 13 (2.95)
313 (65.74) 283 (63.64)
470 440

1. True construction year based upon building permit numbers. For houses aged 2 years, construction began in 1998; for houses aged 3
years, construction began in 1997; for houses aged 4 years, construction began in 1996; for houses aged 5 years, construction
began in 1995; and, for houses aged 6 years, construction began in 1994.

Table 3-6 Continued
2. Percentage within column group.
3. P-value for chi-square test (X2=14.0012, df=8).
4. P-value for Fisher's exact test. Fisher's exact test is performed when the conditions for a chi-square test are not met. It is especially
appropriate for tables with expected cell frequencies < 5 (Schlotzhauer and Littell 1997).
5. P-value for chi-square test (X2=14.2173, df=20).

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd