336 
PACIFIC SCIENCE, Vol. XIII, October, 1959 
Aluminum is thought to function in plant 
maturation and seed setting (Hutchinson, 
1943 ) and in water uptake. The work relating 
aluminum to the blue pigments in plants 
(Chenery, 1948), especially in Hydrangea, is 
well known. The element has been thought 
to exercise a strong influence in plant compe- 
tition in pastures (Shorland, 1934). Alumi- 
num accumulation is thought to be a primitive 
phytogenetic character in plants (Webb, 
1954) because it is generally confined to the 
Archichlamydeae and the primitive sections 
of Metachlamydeae. High aluminum concen- 
trations are rarely found in monocotyledons 
(Chenery, 1949) and Webb records no posi- 
tive Gramineae of 16 species tested. In the 
pteridophytes, accumulators are again con- 
fined to more primitive families and Lyco- 
podium has been extensively studied to relate 
aluminum content and alkaloid molecular 
weight. Hutchinson (1943) considered the 
evidence in Lycopodium to suggest "that the 
capacity to accumulate the element has been 
developed more than once in the genus." 
Although some workers have reported a 
requirement for the element in trace amounts 
(Hutchinson, 1945) by some plants, and 
stimulation from aluminum, especially of the 
ferns (Yoshii, 1939), it is not usually con- 
sidered to be one of the essential nutrients. 
The more frequently reported condition is 
that of aluminum toxicity (Gilbert and 
Pember, 1935; McLean and Gilbert, 1926; 
Rees and Sidrak, 1956). This effect is usually 
related to soils below pH 5.5 (Chenery, 1951; 
Nagata, 1954) and is frequently correlated 
with high concentrations of manganese and 
free oxides or soluble aluminum compounds 
in leached podzolic and latosolic soils (Ellis 
and Truog, 1955; Perkins et al., 1955). 
Magistad (1925) reported toxicity on an alka- 
line soil from a soluble aluminate and demon- 
strated the insolubility of aluminum in 
circumneutral soils, nutrient solutions, and 
water. 
The free hydrated oxides of iron and alu- 
minum have been shown to be more effective 
in fixing phosphorus added to Hawaiian soils 
than are the crystal lattice clays of either the 
2:1 or 1:1 types (Chu and Sherman, 1952). 
Free aluminum oxides do occur in mont- 
morillonite clays (Ellis and Truog, 1955) 
however, and do account for most of the 
phosphorus fixation in soils of this type. 
Nagata (1954) has shown this effect to be 
more pronounced on calcicolous plants and 
Saeki and Okamoto (1954), studying the pure 
aluminum-phosphorus system, showed com- 
plete fixation to occur only when the P/ A1 
ratio was less than unity and that iron and 
aluminum systems showed an almost identical 
trend. Perkins et aL (1955) attributed phos- 
phate-fixing soluble aluminum to decompo- 
sition products of clay minerals. 
Wright and Donahue (1953) definitely 
show aluminum to interfere with the phos- 
phorus metabolism by precipitation on the 
root surfaces although Wallihan (1948) con- 
cludes that aluminum does not interfere with 
the activity of phosphorus in the tops of 
ladino clover plants. Rees and Sidrak (1956) 
found aluminum induced phosphorus defi- 
ciency in barley but not in spinach or Atriplex. 
Hou and Merkle (1950) reported little cor- 
relation between pH and aluminum content 
even though accumulator plants are usually 
calcifugous and have a greater content of 
either aluminum or manganese than "acid- 
indifferent” plants. 
METHODS 
Collection Procedure 
Samples of a selection of plants from three 
of the Hawaiian Islands were collected during 
the winter of 1957-58 (October-March) . The 
principal areas of collection were those soil 
series currently under investigation as poten- 
tial sources of alumina (gibbsite). About 500 
grams of fresh plant material were collected of 
mature leaf, frond, or shoot unless otherwise 
specified. Special attention was given to some 
of the species known to be accumulators of 
aluminum and in many cases the same species 
were collected in areas of nonaluminous soils. 
