BIOLOGICAL CONTROL BY 
INTRODUCTION 
Biological control through introduction of 
new agents has been the most popular and 
successful of the methods utilized to date. In 
one sense, it is the easiest to apply of the 
three methods and can be attempted with a 
minimum of understanding of the plant pest or 
agent to be introduced, However, to be utilized 
fully, thorough knowledge of the plant and its 
parasites is required. 
Introduction of phytophagous parasites is 
contingent on their proved specificity and 
safety. These elements involve an understand- 
ing of the parasite host-finding-and-host-se- 
lection behavior, plus the requirements for 
development, reproduction, and maintenance 
of the species. Most of this work has been 
with insects. 
Studies indicate that phytophagous insects 
perceive and accept the plant host with the help 
of certain plant chemicals; and that these 
chemicals confer specificity, often according 
to taxonomic plant groupings (Harris, 1963, 
1964; Andres, 1965). When these chemicals 
are present throughout a relatively wide range 
of plants, the phytophagous parasites found on 
one species can sometimes be used against a 
weedy species within the same group, but 
outside the current distribution of the parasite. 
Thus, the search for biotic agents may be 
extended to areas other than where the pest 
weed grows and may be especially useful in 
the search for parasites attacking native weeds. 
Within the group of plants circumscribed by 
chemical stimuli that aid the insect in finding 
and feeding upon the host, we often discover 
a smaller number of plant species to which the 
insect is limited for one reason or another. 
Sometimes these limitations appear to be 
nutritional. Adult weevils of Microlarinus 
spp., currently being used to combat 
puncturevine (Tribulus terrestris L.), are 
polyphagous in the absence of their normal 
host, but develop odcytes and oviposit only in 
the presence of Tribulus spp. or closely related 
genera of the Zygophyllaceae (Andres and 
Angalet, 1963). Adult weevils of Phrydiuchus 
spp., under consideration for control of Salvia 
aethiopis L., will feed to varying degrees on 
several labiates, but can develop eggs only 
78 
on certain species of Salvia plus the related 
Lycopus europaeus L, and Melissa officinalis L. 
Within the host range delimited by feeding 
and nutritional requirements, the species sus- 
ceptible to attack may be further proscribed 
by factors of plant-growth habit, form, and 
texture. Weevils of Phrydiuchus spp. did not 
oviposit on plants otherwise chemically and 
nutritionally adequate to support them if the 
plants lacked a basal rosette. In addition, 
larvae of Phrydiuchus spp., the progeny from 
weevils forced to oviposit on certain plants, 
failed to penetrate the stem or exited after 
brief attempts at feeding (Andres, 1965). 
Those insects that do not feed directly on 
the host plant prior to oviposition are also 
guided to specific plants by chemical as well 
as physical stimuli, which also govern the 
selection and ovipositional behavior. 
These are only a few of the factors that 
govern host specificity, but they are the ones 
currently relied upon to guarantee the safety of 
introduction. Evaluation of the feeding response 
of free living larvae and adults has been em- 
phasized in this regard and has followed two 
approaches, In island areas with limited agri- 
cultural crops, such as those in Hawaii, 
starvation tests are carried out with several 
economically important plants regardless of 
their taxonomic relationship to the pest weed. 
If little or no feeding occurs, release of the 
parasite is generally approved. Exclusion tests 
of this type have been used in most of the con- 
trol programs already described. 
For release in areas of more diversified 
agriculture, we attempt to delimit the plants, 
eliciting the feeding response by exposing the 
phytophagous parasite to species of varying 
taxonomic relationship to the weed pest. We 
assume that if plants closely related to the 
weeds donot contain the proper phagostimulant, 
the more distant species would be even less 
likely to elicit feeding and thus are safe from 
attack. We often include a minimum of related 
and nonrelated economic plants to eliminate 
special concern. Wecan also attempt to identify 
chemically the feeding stimulant present inthe 
host plant and to correlate its presence or 
absence with the actual feeding results. This 
second approach contributes more to the funda- 
mental knowledge of insect feeding habits than 
the limited exclusion tests. 


