Endophytic Fungi 265ENDOPHYTES AS MUTUALISTS
A third and non-exclusive possibility is that plants
benefit from the presence of endophytes in their
tissues. Evolutionary theory provides little support
for this hypothesis, as highly diverse symbionts
and contagious spread are typically associated
with parasitic or pathogenic lifestyles (Bull 1994).
However, there are numerous examples of diverse,
horizontally transmitted organisms that inter-
act mutualistically with hosts (e.g., pollinating
insects, mycorrhizal fungi, and root-nodulating
bacteria; see Herreet al.1999). The potential for
endophytes to improve the fitness of hosts, and to
do so in a host-specific manner, raises a series of
questions of interest to community ecologists. I
address a few of these below.
DO ENDOPHYTES ACT AS
ENVIRONMENTALLY ACQUIRED
IMMUNE SYSTEMS?
Herbivores and pathogens are important agents
of density dependence (Carsonet al.Chapter 13,
this volume) and have played an important role
in plant evolution, as demonstrated by the diver-
sity and variation in chemical and structural plant
defenses in tropical plants (Coley and Barone
1996). Both herbivores and pathogens have the
potential to interact closely with foliar endophytes
through the host plants that they share. Could
endophytes provide a cryptic defense against
antagonists?Thishypothesishasbeenpresentedin
various forms by numerous authors (e.g., Carroll
1991), yieldin gfour possibilities re gardin gmodes
by which endophytes may contribute to host
protection.
Do endophytes provide novel chemical
defenses for hosts?
Endophytes may impart direct chemical defense to
plants by producin gsecondary compounds that
deter insects and inhibit pathogenic organisms
(see Saikkonenet al.1998). The ability of endo-
phytes to secrete substancesin vitrothat limit
the growth of other microbial species, including
pathogens, has contributed to current enthusiasm
re gardin gbioprospectin gand biolo gical control
with endophytic fungi (e.g., Strobel and Daisy
2003, Gunatilaka 2006). In the context of her-
bivory, this mode of defense is exemplified by
the alkaloids produced by clavicipitaceous endo-
phytes of temperate grasses (Clay and Schardl
2002), and has been demonstrated in a few hor-
izontally transmitted endophytes of woody plants
(e.g., endophytes that produce compounds toxic
to spruce budworms; Findlayet al.2003). Arnold
et al.(2003) suggested that endophytes of trop-
ical trees serve as acquired immune systems,
actin gin concert with intrinsic leaf defensive
chemistry when young, and in place of those
defenses in mature leaves. However, the poten-
tial for these low-biomass infections to manifest
major chemical signatures in foliage has not been
assessed.
One intriguing hypothesis is that large quanti-
ties of chemical output per endophyte may not be
needed to defend host tissues. Carroll (1991) pro-
posed that endophytes protect hosts via a mosaic
effect, whereby endophytes create a heteroge-
neous chemical landscape within and among
leaves. As a result, parts of a genetically uni-
form plant would differ unpredictably in terms
of palatability or quality for herbivores, and in
terms of infectivity for pathogens. This hypothesis
is compellin gbut has not yet been explored.Do endophytes activate host
defenses?Systemic acquired resistance has lon gbeen rec-
ognized in plants (Agrios 1997) but there is
currently no evidence for systemic protection of
tropical plants as a function of endophyte infec-
tion. Arnoldet al.(2003) raised endophyte-free
seedlings, inoculated a subset of leaves on each
seedlin gwith endophytes, and then inoculated
endophyte-infected and endophyte-free leaves
with a virulent foliar pathogen (Phytophthora). In
that study, the presence of endophytes in some
leaves did not protect other leaves on the same
plants from severe pathogen damage. That study
did not examine pathogen damage in seedlings
that had no endophytes, raisin gthe possibility
that some systemic defense occurred but was not