Endophytic Fungi 259from 28 species of woody plants representing
24 families and 14 orders of angiosperms at BCI
(N=9 leaves from three individuals per species in
the understory of late secondary forest). Studies
in Puerto Rico and Guyana have had similar
results (Gamboa and Bayman 2001, Cannon and
Simmons 2002). The proportion of endophyte-
infected leaves appears to increase from the arc-
tic to the tropics (Arnold and Lutzoni 2007),
although most plant communities have not yet
been sampled. In the temperate zone, the fre-
quency of endophyte infections is influenced by
precipitation, humidity, elevation, irradiance, and
air pollution, but the roles of these factors have
not been fully assessed in the tropics. In partic-
ular, tropical savannas and dry forests – as well
as the forest canopy in moist or wet forests –
represent unique environmental conditions
imposed by high irradiance, high temperature,
and geographic congruence with endophyte-rich
forests. Plants in these communities may offer a
wealth of novel endophyte species.
Neither seedlings nor leaves of tropical trees
typically contain culturable endophytes at emer-
gence, but colonization proceeds rapidly given the
presence of airborne inoculum and high rela-
tive humidity or wettin gof leaf surfaces by dew,
rain, or fo g(Arnold and Herre 2003). At BCI,
infection rates (defined here as the proportion
of leaves containin gendophytic fun gi) increase
to nearly 100% of leaves as foliage matures.
Field experiments have shown that endophytes
are present in more than 80% ofTheobroma cacao
leaves within 2 weeks of leaf emergence dur-
in gthe early wet season at BCI (Arnold and
Herre 2003). Leaf toughness does not influence
endophyte colonization: youn gand mature leaves
can be colonized with equal frequency (Arnold
and Herre 2003).
Endophyte infections within leaves are typi-
cally quantified by determinin gthe proportion of
small leaf fragments (typically ca. 2 mm^2 ) that
yield endophytes in culture (Box 15.1). Propor-
tions of leaf area colonized by endophytes differ
amon gtropical sites and host species, althou gh
all tree species examined to date at BCI have
consistently high densities of endophyte infec-
tion (>95% of tissue segments; Arnold 2002).
Rodrigues (1994) found that 25% of leaf pieces
were colonized by endophytes in fronds ofEuterpe
oleracea(Arecaceae) in seasonally inundated sites
in Amazonia. Gilbertet al.(2002a) recovered
endophytes from 20%, 92%, and 80% of 3 mm^2
leaf fragments from three mangrove species in
Panama (Avicennia,Rhizophora, andLaguncularia,
respectively), with infection frequencies parallel-
in gthe salinity of water on leaf surfaces. Lod ge
et al.(1996) found a high infection rate in Puerto
RicanManilkara bidentata(Sapotaceae; 90–95%
of 2 mm^2 leaf fragments), as did Gamboa and
Bayman (2001) for leaves of Guarea guidonia
(Meliaceae) in Puerto Rico (>95%). Apparently
healthy leaves contain numerous, independent
infections, rather than systemic or otherwise
extensive growth of hyphae (Lodgeet al.1996).
The biomass resultin gfrom any given infection is
very low, such that each leaf represents a densely
packed mosaic of diverse endophyte species (Lodge
et al.1996). Mature leaves have a higher infec-
tion density than do younger leaves, reflecting the
accumulation of numerous, independent infec-
tions as leaves age, and the differential prolif-
eration of favored species as leaves approach
senescence (Arnoldet al.2003). Synthesizing
the results from several field studies in central
Panama, Herreet al.(2005b) suggested that
most leaves are saturated by endophytic fungi
(i.e., contain endophytes in 100% of 2 mm^2 tissue
segments) within 3–4 weeks after emergence.
Unfortunately, methodological artifacts often
prevent comparisons amon gstudies, limitin gour
ability to assess the role of abiotic factors or
microhabitat characteristics in influencin gendo-
phyte abundance. For example, Gamboaet al.
(2002) demonstrated a strong, inverse relation-
ship between the size of leaf fragments used in
culture and the number of fungi isolated from
leaves. Only studies with similar leaf fragment
sizes, media, culturin gconditions, and surface-
sterilization protocols can be compared, and
standardization of these methods is needed.TROPICA LENDOPHYTE
DIVERSITY
Hawksworth (1991) estimated global diversity
of fungi at 1.5 million species, drawing from a