THE EFFECT OF DEPTH OF ROOTING ON CITRUS ROOT STRUCTURE
AND WATER ABSORPTION'
WILLIAM S. CASTLE
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
William S. Castle
No man is an island, entire of itself;
every man is a piece of the continent,
a part of the main; if a clod be
washed away by the sea, Europe is the less,
as well as if a promontory were, as well
as if a manor of thy friends or of thine
own were; any man's death diminishes me,
because I am involved in mankind; and
therefore never send to know for whom the
bell tolls; it tolls for thee.
The author is indebted to many people who were involved in the success-
ful completion of his graduate training; to Dr. A. H. Krezdorn, the chair-
man of his supervisory committee, who by deed and word constantly challenged
the author to strive for the highest professional standards and contributed
freely of his time and counsel; to Dr. R.C.J. Koo, co-chairman of his
supervisory committee, who was instrumental in planning this research; and
to Drs. J. Bartholic, N. Gammon Jr. and J. Soule who were invaluable in
providing the author opportunities to discuss his research.
The author wishes to thank all the above mentioned members of his
supervisory committee for their advice and support during the preparation
of this manuscript and Miss Rose Ann Abernathy who typed it.
Appreciation is also extended to several members of the Soil Dept.,
Drs. J.G.A. Fiskell, L.W. Zelazny, T.L. Yuan, D. Graetz and L.C. Hammond,
who although not members of the author's supervisory committee, willingly
provided laboratory equipment and/or advice on matters pertaining to the
soil aspects of this research. The author is particularly grateful to
Dr. Hammond who provided laboratory facilities and guided the author
through the collection of certain soil physical data.
This research was made possible by a graduate assistantship provided
through the Dept. of Fruit Crops and the assistance of the Tennessee
Valley Authority Educational Fund.
The author wishes to express his sincere gratitude to his wife,
Eileen. It is only with an appreciation of the many hot hours spent
collecting field data, the many days spent working in the laboratory,
and the many lonely nights, that the author-gives credit to his wife for
her contribution to this manuscript.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS . . . .......... . . iv
ABSTRACT .. . .......... . . ix
INTRODUCTION .. . . . . .... ........ . 1
LITERATURE REVIEW . . .. ..... . . . . . 3
The Citrus Root System . ...... . . . . .. 3
Control of Root Development . . . . .. .. 3
Citrus Root Distribution ...... . . . . 3
Citrus Root Anatomy and Morphology . . ... . .. 7
Root Physiology . . . . .. . . . . . 11
Preliminary Remarks . . ... ..... ... . 11
Theories of Ion and Water Absorption . . . ... 12
Relationship of Root Anatomy and Morphology
to Ion and Water Absorption .. ... . . 16
The Soil Environment and Its Effect on Root
Structure, Function and Distribution . . . . .. 22
Soil Texture .. .. .... ........... . 23
Soil Moisture .. .. ... ........ . . 25
Soil Aeration ... .. ............. 27
Soil Nutrients .. .... .. . . ... 29
Soil Temperature ......... ... .... 31
Soil pH . .... ...... . .. . . . 33
Soil Microorganisms . .... . . . .. 34
The Influence of Depth of Rooting on Root
Structure and Function .. . ....
Anatomy . . . . . . . .
Water Absorption . . . . . . . .
Nutrient Absorption . . . . .
MATERIALS AND METHODS . . . . . . .
Description of the Experimental Site . . . .
Experimental Units . . . . . . .
Field Sampling Procedure . . . .
Soil Moisture Depletion .....
Soil Samples for Chemical and Physical Analyses
Root Samples ..... . . . . . .
Sample Preparation and Analysis . . .
Soil Chemical Analysis . . . .
Soil Physical Analysis . . . . . .
Root Samples . . . . . . .
Statistical Analysis ... . .
RESULTS . . . .
Characterization of the Soil Environment
Physical ....... . . . .
Chemical . . . . . . . .
Soil Mineral Nutrient Content . . . .
Macronutrients .. . . ....
Micronutrients . . . . ....
Aluminum . . . . . . .
. . 37
. . 37
. . 40
. . 40
. . 40
. . 41
. . 43
. . 44
. . 44
. . 44
. . 61
. . 64
. . 64
. . 76
. . 76
Rootstock Water Use .. ... ....... . . . 83
Fluctuations in Soil Moisture Content . . ... .. . 83
Analysis of Rootstock Water Use . . . ... .... 87
The Relationship of Root Distribution to Soil
Water Depletion . .. .. ..... . ... 93
Root Anatomy . . . . ....... .. . . ... 104
Root Morphology . . . . . . . . . .. . 113
DISCUSSION . . . ........ . . . . . 120
Rootstock Water Use ... . . . . . . . . 120
Relationship of Root Distribution to Water
Absorption .. . ......... . .. 120
Root Efficiency . . . ... . . . .. 121
Validity of the Moisture Extraction Data . . . .. 124
Root Structure ...... . . . . .... . 126
Rootstock Adaptation . . .. .... . . 127
The Soil Environment .. . ............. . 130
Chemical . . . . . . . . . ... . 130
Physical . . . . . . . . . .. . 132
SUMMARY AND CONCLUSIONS . . ... ... ........ 134
LITERATURE CITED..... . .... ........ . ... .. 137
BIOGRAPHICAL SKETCH . .. ... .. . .... . 154
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
THE EFFECT OF DEPTH OF ROOTING ON CITRUS ROOT STRUCTURE
AND WATER ABSORPTION
William S. Castle
Chairman: Dr. A. H. Krezdorn
Co-Chairman: Dr. R.C.J. Koo
Major Department: Fruit Crops
The effect of rootstock root distribution on water absorption and
citrus root morphology and anatomy was investigated. Experimental units
were trees on 4 rootstocks, rough lemon, sweet orange, sour orange and
'Rusk' citrange in an 'Orlando' tangelo rootstock trial located among a
commercial tangelo orchard in Lake County, Florida. Soil samples were
taken from the experimental site for chemical and physical analysis.
Soil water depletion was studied by the neutron moderation method.
Selected water extraction data were compared to root distribution data
obtained from the same trees in a previous study.
The results indicated that soil water loss was directly proportional
to root quantity as measured by the dry weight of feeder roots. The
general pattern of moisture depletion for trees on all stocks showed
that significant losses in soil moisture occurred near the surface first,
followed by gradual increases in the contribution of deeper roots as the
surface moisture was depleted.
Meaningful differences were found in root efficiency among depths
and rootstocks when soil water depletion was compared to root distribution.
The magnitude of the ratios of these 2 factors emphasized the significance
of deep roots especially when surface moisture was rapidly lost as a
result of a high root density. It appeared that differences in root-
stock adaptation to the soil environment could be explained on the basis
of inherent differences in the root distribution of trees on the 4 root-
stocks and on an interaction of these differences with the physical
characteristics of the soil environment. This interaction seemed to
have been the source of differences in root effectiveness in water uptake.
The morphology and anatomy of field-sampled roots of each rootstock
and from several soil depths were remarkably similar on the basis of
observations made both with the scanning electron microscope and with a
light microscope. Several features were observed which had possible
significance to citrus root function. Root hairs were found on roots
from every rootstock and depth; radial elongation of groups of hypoder-
mal cells and bark cracks were frequently noted; and the epidermis
generally deteriorated and was shed at a short distance from the root
The chemical environment of soil from the experimental site changed
considerably in pH, organic matter content, cation exchange capacity and
mineral ion content with depth. A subsoil clay horizon differed in
physical and chemical characteristics from the sandy soil. The soil
environment at all depths appeared to be able to sustain considerable
root growth. Soil physical data demonstrated the low water-retention
characteristic of sandy soil.
The existence of meaningful differences among root systems of agri-
cultural plants has long been recognized. Many laborious excavations have
revealed the rooting characteristics of both agronomic crops and fruit
trees including citrus. Nevertheless, many years of research effort
regarding the water and nutritional relationships between roots and their
environment have yielded little direct information on the functional
significance of root distribution. Most knowledge of the physiology of
root systems has been obtained from studies on individual excised roots
or small greenhouse plants in sand or solution culture. The few field
studies have been largely limited to field crops which have relatively
shallow roots and lend themselves more readily to study of their entire
Root systems of fruit trees are diverse, often deep, and inherently
difficult to investigate. Attempts to determine if any relationships
existed between fruit tree root systems and certain scion characteristics
have led to some gross observations and scientific evidence indicating
that the size and density of a root system was related to tree size and
fruit yield (31,217). In the past, it has generally been accepted that an
increase in tree size was accompanied by a proportional increase in fruit
yield and the size of the root system; however, size of root system may
refer to lateral and vertical spread or quantity of roots. In the lat-
ter case, a positive relationship of citrus yield to feeder root quantity
was demonstrated but a point was reached beyond which other factors began
to limit yield (34).
A similar role for depth of rooting in plant growth and yield has
also been supported by field observations and limited experimental data.
Surveys of soils in deciduous fruit production indicated that both tree
size and yield are favorably influenced by greater rooting depths. Studies
of citrus trees also imply that deep, well-distributed root systems are
essential for optimum growth and productivity. In Florida, the largest
trees on well-drained sands, were associated with deep root systems (38),
Moreover, the vertical development of root systems was not always asso-
ciated with an increase in root quantity indicating some significance to
depth alone. Trees that develop in the shallower soils of Florida are
smaller and less productive where root growth is restricted.
It is not clearly understood whether differences in root systems are
related to plant growth and yield; however, it is apparent that current
economic and production requirements dictate the need to improve the
efficiency of plant functions. Therefore, it is essential that the
physiological significance of root distribtuion be investigated.
The objectives of this research were to study the water absorption
pattern of 4 citrus rootstocks as it varies with depth of rooting and soil
moisture stress and characterize the root environment of the same rootstocks
as it varies with depth and relate this information to root anatomy and
The Citrus Root System
Control of Root Development
There are basically 3 forms of root systems among vascular plants,
but only the tap root type is common in woody plants such as fruit trees
(66). Root development is however strongly modified by environment (127).
It is often difficult to ascertain the relative influence of heredity or
environment, although one factor appears to predominate over the other
in some instances.
Heredity seems to exert its greatest influence upon root development
during the early stages of tree growth (127) For example, substantial
inherent differences were evident in the root systems of young citrus
trees budded to several rootstocks (188). Genetically stable character-
istics of the root system are maintained as seedlings or young, budded or
grafted trees grow older but other characteristics are increasingly
modified by environmental factors (127).
Citrus Root Distribution
The physical difficulties inherent in studying the entire, often
deep root system of a tree have hindered the accumulation of knowledge
about such root systems. One of the traditional methods for examining
root distribution until recently was the trench method (2,4,15,17,22,104,
164,165,215). This procedure required a considerable amount of laborious
excavation that discouraged broad sampling. Newer methods such as the
use of radioisotopes (85,151,152,153,185,186) and measuring soil mois-
ture depletion (4,30,46,121,207) have eased the task of measuring root
General information on citrus root distribution has been reported
from most areas of the United States where citrus is grown (2,30,32,33,
34,50,75), yet there are very few detailed descriptions of citrus root
systems. Nevertheless, it was apparent from early root studies that
Citrus and related species had a tap root system with the potential to
root very deeply and extensively (50,75,145,171,215) and whose extent
was strongly modified by the soil environment (2,76,77,78,149,172).
The most detailed and comprehensive descriptions of citrus root
distribution have come from studies conducted in Florida. An early study
reported the vertical and lateral spread of 4 to 6 year old nursery trees
of sweet orange budded on several rootstocks (188). Striking differences
were found in the character of the root systems; however, the results
were representative of young, closely spaced trees. It was not known
whether the root distribution of older trees was similar.
Ford was the first to quantitatively examine citrus root systems in
detail (72,73,74,75,76,77). Comparisons were made on the basis of the
dry weight of small roots collected from cores of soil taken at selected
depths. Root systems of several common Florida rootstocks, rough lemon
(C. jambhiri Lush.), Celopatra mandarin (C. reshni Hort. ex Tan.), sour
orange (C. aurantium L.) and sweet orange ft. sinensis L. (Osbeck)] in
the deep sandy soils typical of the central "Ridge" area of Florida were
examined (75). Rough lemon rooted as deeply as 5.5 m in contrast to 3 m
reported earlier (72) Trees on rough lemon were found to have as many
as 50 percent of their feeder roots below 75 cm in the soil with 15 per-
cent below 3 m. Ford consistently found in repeated sampling that trees
on rough lemon in deep sandy soil were deep rooted with numerous well-
distributed fibrous roots (72,73,74,75,77) Lateral roots were equally
extensive, often reaching out to 8 m from the trunk. Ford later showed
that lateral roots might extend a distance of 18 m (81). A report from
California confirmed the general nature of the rough lemon root system
but strong development of the tap root was not observed (217).
Sour orange roots occasionally penetrated as deeply as those of
rough lemon but differed in root distribution. There were fewer roots
below 75 cm in the soil and more above, especially in the 0 to 25 cm zone.
Sweet orange roots were found to a depth of 3.5 m. Trees on this root-
stock had the highest total weight of feeder roots for any rootstock
studied. Cleopatra mandarin compared favorably to rough lemon in depth
of rooting and total feeder root weight. Both sweet orange and Cleopatra
mandarin had a larger proportion of their feeder roots near the surface
than did rough lemon (75).
Ford's studies revealed that citrus trees are capable of extensive
root development in a favorable environment. These results were confirmed
and expanded in a recent study where root distribution was sampled twice
over a 2-year period (38). This study included the 4 rootstocks examined
by Ford plus Palestine sweet lime (C. limettioides Tan.), several citrange
(C. paradisi x C. recticulata) cultivars, and 2 selections of trifoliate
orange (Poncirus trifoliata Raf.). Trees in this experiment had a common
scion and were planted in a site with fairly uniform sandy soil underlain
by clay at 210 to 240 cm. General characteristics of the root systems
were similar but trees on sweet and sour orange and Cleopatra mandarin
rootstocks were not as deep rooted as reported by Ford. Total feeder
root weights reported by Ford were generally smaller for all rootstocks.
The root system of trees on sweet lime was similar to that of trees
on rough lemon. Other studies supported this observation (150,162,188).
Trees on the trifoliate rootstocks were shallow rooted, 3 m, but they had
a large number of feeder roots with the majority at less than 75 cm deep.
Total feeder root weights were comparable to those of rough lemon. Trees
on 'Rusk' citrange were consistently the shallowest rooted, 2 5 m, and
had the smallest total feeder root weights. Those on 'Troyer' citrange
were intermediate in depth of rooting and number of feeder roots, while
those on 'Carrizo' citrange were similar to rough lemon. The lateral
extent of rooting could not be accurately determined because the root
systems overlapped in some areas (38).
Drastic modifications in the root systems described above can occur
as a result of changes in root environment. Root growth in the Indian
River area of Florida is severely limited by a high water table. The
depth to the water table in this area necessitates bedding the land to
provide an adequate volume of soil for root development. Depth of root-
ing is often limited to 0.7 to 1 m and 75 percent of the roots are found
in the top foot of soil (76,81,172). Nevertheless, these root systems
have a larger number of feeder roots per unit volume of soil as compared
to the root systems of trees growing in deep sands (172).
It can be concluded that citrus trees even under moderate environ-
mental restrictions have characteristic root systems which can be defined
quantitatively. These characteristics are probably lost in environments
where root development is severely limited.
Citrus Root Anatomy and Morphology
Citrus trees form a tap root system. The primary root emerges from
the seed coat and continues to grow and enlarge forming the tap root
from which the main lateral roots arise. Secondary laterals along with
the other main roots eventually form the framework as the root system
expands. The smallest laterals branch and rebranch to support and form
a network of "bunch-like" groups of fine fibrous or feeder roots.
Gross anatomy of an individual citrus root is similar to that
depicted in textbooks (66) as typical of a woody dicotyledenous plant.
The primary body of elongating roots consists of several distinct tissues
and regions or zones of functional significance, including the root apex
(root cap and meristematic area), zone of elongation and zone of differ-
entiation and maturation.
A typical citrus root apex has 3 histogens. The basipetal root initial,
or plerome, gives rise to the stele or vascular cylinder. The histogen
adjacent to the plerome is the periblem which gives rise to the cortex.
A common initial, the dermatogen-calyptrogen, gives rise to the epidermis
and root cap, respectively at the root apex. The root cap is almost
always present and easily recognized but is most prominent in rapidly
The 3 root initials are not easily discerned in the root apex of the
primary root but they can be identified in fibrous roots. The tissues
arising from each initial are distinct at a small distance from the his-
togens. The meristematic organization in the root apex of Citrus is
different from that of the pear, apple and peach (65,154,174,191).
1This discussion is taken primarily from the work of Hayward and Long
(109), Schneider (189) and Cossman (48). Contributions from researchers
other than those mentioned are cited in the discussion.
The epidermis can be identified in the proximal portion of the
meristematic region. It is generally 1, and occasionally 2, cells thick.
Several different cell types appear in groups randomly distributed around
the root as the epidermis matures. The outer tangential wall becomes
thickened in 1 type while there is only secondary thickening and the
wall remains very thin in another one. The epidermis is eventually
sloughed off or ruptured by secondary growth and replaced by a periderm.
The tissue adjacent to the epidermis is the cortex. The outer layer
of the cortex forms a hypodermis (exodermis) in the zone of elongation.
The hypodermis is usually 1 cell thick and composed of cells considerably
larger in volume than those of the epidermis. The outer periclinal walls
become impregnated with some material of which there is disagreement as
to its nature. It is agreed, however, that the inner walls become strongly
suberized. The remainder of the cortex consists of thin-walled parenchyma
tissue with an occasional sclerid or crystal. Peripheral cortical cells
are generally smaller and thinner-walled with a denser cytoplasm.
The inner most single-cell layer of the cortex matures further behind
the root apex in the zone of elongation and in the zone of differentiation
into a specialized tissue called the endodermis. Conspicuous casparian
strips, which are a continuous band of suberin in the transverse and
radial walls of individual cells, appear in the mature region of the root.
Unsuberized endodermal cells, called passage cells may occur opposite
The plerome gives rise to the stele which contains the provascular
strands. The pericycle forms from the outer one or 2 layers of stelar
cells and lies adjacent to the endodermis. Alternating strands of proto-
xylem and protophloem can be seen interior to the pericycle. The proto-
xylem strands are arranged in the shape of a multipointed star with the
protophloem lying between the xylem strands adjacent to the pericycle.
Primary roots typically have 6 to 8 protoxylem strands. The larger lateral
or framework roots commonly have 3 to 5 strands while smaller laterals
and fibrous roots have only 2 or 3 strands. The relationship of root
size to the number of xylem strands has been observed in other fruit
The central tissue of fibrous roots is composed of metaxylem. The
center of the stele in larger roots with polyarch xylem contains a
parenchymatous pith interspersed with metaxylem particularly near the
Differentiation of the provascular elements begins at different
levels in the stele. Young protopholem elements can be observed in the
zone of cell division and more clearly in the zone of elongation near
the root tip but protoxylem elements do not appear before the zone of
differentiation. This pattern of vascular maturation is common in woody
Secondary development of the citrus root closely resembles that of
woody dicotyledons (66). Lateral roots frequently have extensive growth
with little or no secondary enlargement occurring in the small fibrous
roots. All tissues external to the pericycle are sloughed off in roots
where considerable secondary growth has taken place. A transverse section
of an older, large root will thus show a periderm, secondary phloem,
cambium, secondary xylem, and pith. The xylem is diffuse porous. Vestiges
of the cortex may remain in roots with less or no secondary growth although
one or more periderms may be formed.
Secondary growth can result in the formation of longitudinal growth
cracks. Tissues exterior to the new cambium are stretched beyond their
elastic limit when the vascular cylinder enlarges and a cambium forms.
The hypodermis proliferates locally by reverting to a meristematic state
and forms "hypodermal absorbing areas" in some areas.
The most controversial facet of citrus root morphology has been the
development of root hairs. Early general studies on citriculture indicated
that citrus roots do not produce root hairs (116); however, this view has
subsequently been disproven (48). Root hairs are finger-like projections
from individual epidermal, and rarely hypodermal cells, which arise in
the region of elongation. Their formation is apparently an adaptive
response in that they may occur at different distances from the root
apex in response to factors such as light, soil pH, temperature, and mois-
ture and rate of root growth (93,126). Evidence proving the existence of
root hairs has been largely obtained under laboratory or greenhouse
conditions. There is little information available regarding their for-
mation and response under field conditions.
Another relatively uninvestigated aspect of citrus root morphology
is the mycorrhizal relationship of roots with different Endogone species
of fungi. The fungal hyphae penetrate the root cortex, but this relation-
ship has not been shown to alter the anatomy or morphology of citrus
There is a paucity of information regarding the comparative anatomy
and morphology of citrus roots collected from different rootstocks and
from several depths in the soil. Differences were noted in the one instance
where shallow roots of several rootstocks were compared as to cell size,
mode of lignification of the pith, suberization in the endodermis,
thickening of epidermal cell walls, and the shape of protoxylem strands
Several comprehensive reviews in the field of root physiology have
been published in recent years (20,28,29,137,203). These reviews have
encompassed the many facets of the subject, which have been studied in
great detail, generating a large amount of literature and making it
difficult to survey root physiology in its entirety. The present review,
therefore, will be focused on the physiology and anatomy of water and ion
uptake, preceded by a few remarks regarding procedures and new terminology
which have arisen from these studies.
Most knowledge pertaining to root physiology has been obtained from
investigations of excised herbaceous plant roots. Very few studies have
employed roots of woody plants. Thus, it is commonly assumed that the
information obtained from, for example corn roots, is applicable to
woody plant species.
The object of many absorption studies has been to determine answers
to questions such as how ions and water move to the plant root, by what
mechanisms and pathways do ions and water move radially to the xylem, how
are root function and anatomy related, and what factors influence them.
It is generally agreed that the first step in the absorption of
mineral ions is the free diffusion of ions into the "apparent free
space" (A.F.S.) (63,126). The A.F.S. is composed of 2 parts, "outer"
and "Donnan" free space (63,126). Both represent extracellular space
within the root that is in direct contact with the soil solution. A
continuum is formed in fact by which ions are able to freely diffuse
from the soil solution into the root, primarily the cortex region. The
"Donnan" free space is that part of the A.F.S. where ions are adsorbed
onto the negative exchange sites present in cell walls.
There are several unresolved questions regarding the A.F.S. It is
not agreed how far it extends into the root; however recent evidence
indicates that the endodermis is probably the inner boundary (137).
Secondly, there is a question as to the boundary of the A.F.S. on a
cellular level. The plasmalemma is often considered the first barrier
of a cell (63,126), while others consider the tonoplast as the primary
The "Donnan" free space has been implicated in ion uptake. It has
been suggested that it qualitatively affects ion absorption as a result
of the cation exchange (C.E.C.) sites in the cell walls (196,197,210,
211). Roots with a high C.EC. favor the uptake of divalent cations over
monovalent ones (197,224). Nevertheless, the importance of the root C.E.C.
has been challenged. It is uncertain how reliable C.E.C. measurements
are and the supposed influence of C.E.C. on the selectivity of ion
absorption is not well established (63). The "Donnan" free space may be
important, though, at extremely low ion concentrations (63).
Theories of Ion and Water Absorption
An important function of roots is to supply the aboveground parts
of the plant with water and mineral nutrients. Ions and water must move
through the soil to the root to perform this function. Ions then move
from the root surface into the A.F.S. of the cortex where they are
accumulated. These ions, and water, then move laterally into the xylem
where translocation takes place (63,126).
Events occurring prior to ion movement into the stele, and their
subsequent absorption by the xylem, are reasonably well defined and
agreed upon. Two processes, first, ion absorption by the cortical cells,
which requires crossing the plasmalemma, and, second, radial transfer of
absorbed ions to the xylem, are occurring simultaneously in that region
of the cortex where the A.F.S. exists (63,126). The first event takes
place independent of water movement. The second process is affected,how-
ever, by the rate of water uptake.
The first activity presents 2 pathways for radial movement of ions.
Ions may travel from vacuole to vacuole or through the cytoplasm of
adjacent cells or the ions may move through the A.F.S. if they are not
absorbed. Free diffusion beyond the endodermis is usually considered
to be prevented by the deposition of suberin (43,63,126).
The first pathway, within the cell, is not too important because the
vacuole is considered an ion "sink" where ions are deposited (126,137).
It has been shown that ions do not leave the vacuole rapidly (28). The
second pathway is considered the primary route. Ions move from cell to
cell through the plasmodesmata. This pathway is called the symplasm as
opposed to the apoplast or the A.F.S. (29). Each pathway, including
free diffusion to the endodermis, has some significance, but the sym-
plastic path is thought to be the most important.
Ions are carried to the vascular cylinder via the symplasm where
they are transferred to the conducting xylem elements and translocated to
the shoot. It is not known how the ions gain entrance into the xylem.
This problem was recently reviewed by Brouwer (28). It is apparent that
some form of metabolic activity is involved (63). Earlier, it was thought
that the endodermis actively secreted ions into the stele (63,126).
Ultrastructural studies have not shown the endodermis to be particularly
suited for secretion (18). Nevertheless, the endodermis is considered an
effective physiological sheath, which in the suberized state, prevents
free diffusion of both water and ions into and out of the stele (43,48).
There are essentially 4 hypotheses which attempt to explain ion
absorption and translocation. They are generally in harmony, as mentioned
previously, with the major divergence involving the processes occurring
in the vascular cylinder.
The oldest and most frequently cited theory is that of Crafts and
Broyer (49) which proposes that ions are actively accumulated by cells
in the well-aerated cortex, move symplastically into the stele and
"leak" out. Oxygen-deficient stelar cells are less able to retain ions
and therefore they escape and accumulate in xylem cells. This theory
has been subjected to criticism on several grounds. One of the strongest
arguments raised against it has been the demonstration by electron micro-
probe analysis that stelar cells often contain a higher ion concentration
than cortical cells (137).
Laties and Budd (136) modified the Crafts-Broyer theory by suggesting
that an inhibitor prevented the stelar cells from retaining ions. This
hypothesis is similar to Crafts-Broyer and has been criticized on the
Hylmo (117) proposed that xylem cells with living protoplasts
accumulate ions that are released upon the death of the protoplast.
There is little experimental evidence to support this hypothesis, Iowever,
it was found in corn roots that ion uptake was associated with the
number of living xylem cells in the stele. Both uptake and the number
of living cells decreased basipetally from the root apex (6).
These 3 theories presume that some form of active or metabolic,
energy-requiring process is responsible for the absorption of ions
across the plasmalemma in the cortex. This process is the basis for the
fourth theory, the carrier theory of Epstein (63,64), Evidence such as
the effect of temperature, oxygen and metabolic inhibitors all indicate
that some form of active process is occurring (63). Salient features
of this hypothesis include:
(1) A carrier molecule located in the plasmalemma unites with an
ion exterior to the cell and carries it across the membrane
where it is released into the cytoplasm.
(2) Michaelis-Menton kinetics are applicable.
(3) The kinetics of the mechanism imply an enzyme-like situation
and ion selectivity.
(4) Two mechanisms are involved, one operating at low ion concen-
tration and the other at high ion concentration.
(5) The mechanism requires Ca in order to maintain the semipermeable
state of the plasmalemma.
This hypothesis proposes that ions are accumulated in the cortex,
move through the symplasm into the stele, and then are transferred out
of stelar cells by carrier molecules into adjacent xylem cells. This
view has not been completely accepted. Sites of the high and low con-
centration mechanisms have been vigorously debated (63,135). Also, the
carrier molecule has not as yet been physically or chemically identified
The physiology of water uptake has not been as difficult to eluci-
date as the uptake of ions. Water moves through the soil and plants
largely in response to decreasing gradients of water potential (205).
The gradient within the plant is dependent upon the rate of water loss
due to transpiration. Therefore, water absorption is primarily a depen-
dent and passive process. Water follows the path of least resistance
within the root and moves by mass flow through the A.F.S. Water does
not, however, flow unhindered as evidenced by the frequently observed
lag of absorption behind transpiration. Water movement is resisted in
the root, chiefly due to the need for water at some time to pass
through living cells (126). A primary point of resistance is the endo-
Ion uptake has been reported as being directly proportional to the
rate of water uptake; however, the considerable amount of evidence
regarding the validity of this effect is conflicting (28,29).
Relationship of Root Anatomy and Morphology to Ion and Water Absorption
Absorption of water and mineral nutrients by a root is related, in
part, to its anatomy and morphology. Evidence for this relationship
has been obtained primarily by examining the uptake of various radio-
isotopes and water by isolated segments along intact roots. This infor-
mation is then correlated with changes in anatomy at the respective
sites along the root.
Numerous studies have shown that many roots have a region charac-
terized by a large accumulation of ions, a second region where a large
number of ions are absorbed and translocated and a third region where
the activity of the region is not firmly established and varies con-
siderably with the plant species (126,184,219). The first region
corresponds to the meristematic root tip. Meristematic cells are small,
filled with cytoplasm, and closely spaced, with little intercellular
space. Movement of water and ions is therefore strongly impeded. More-
over, it has been suggested that slow movement of water and nutrients
in this region restricts activity of the root tip and is, thus, the
primary factor limiting root growth (126). It is commonly noted in uptake
experiments, despite the apparent impermeability of the root apex, that
this region is the site of maximum ion accumulation (184,219). Earlier
investigations led to the false conclusion that this region was paramount
in the supply of nutrients to the plant (126). It is now thought that
accumulation in this region occurs as a response to high metabolic needs
of the root apex (126).
The second uptake site corresponds to the location in the root where
elongation has essentially been completed and differentiation is begin-
ning. The cortex is well formed with large vacuolate cells here. The
endodermis has casparian strips and suberization has begun. Mature
xylem elements are present. The root is most active in simultaneous
absorption and translocation at this level (219).
The endodermis continues to suberize and the cell walls thicken
beyond the site of maximum uptake and translocation. The epidermis, and
hypodermis if present, suberize. The endodermis, epidermis and hypodermis
represent possible barriers to water and ion movement. Tissues external
to the pericycle are shed when secondary growth occurs and a periderm
forms. Suberized cells formed from the periderm and vascular cambium
may then function as barriers at this stage. Little or no absorption is
thought to occur at this point; however, it has been shown that some
absorption does occur in this region in trees (1,125,128). The volume
of soil occupied by unsuberized roots is too small to supply the water
needs for an entire tree. Therefore, it is significant that some tree
roots, for example, citrus, can absorb water through lenticels and growth
cracks in the surface of older root tissue (52,110).
The relative positions of the regions described above are thought to
be similar for most roots. Actual distances from the root apex are not
known for most plants for these regions and appear to be dependent upon
plant species and the rate of root growth (126).
A recent report (43) exemplifies the present manner in which the
relationship of root function to anatomy has been approached. In this
study, the experimental units were the intact roots of barley plants.
Root segments were isolated in a solution of radioisotopes and their
absorption pattern recorded. Cereal and other agronomic crops are
commonly used in this type of experiment. The root system of barley,
like those of other monocotyledonous plants, consists of 2 types of roots;
seminal roots or those of embryonic origin and nodal adventitiouss) roots
that arise from the growing shoot. Laterals arise from the seminal and
nodal axes. They differ in anatomy, the seminal axes containing only 1
long vessel in the stele. Nodal axes have several vessels in which the
individual elements are shorter (101).
The results of the investigation, which are listed below, confirm
earlier findings (19,183,184,185) and add some new information.
(1) Ion uptake was related to root volume more so than to any other
(2) Seminal axes absorbed the most PO4 per unit volume with the
absorption peak for all roots located 2 to 5 cm behind the root
(3) The point of maximum absorption corresponded to the point where
cell elongation was complete and mature vascular tissue was
(4) Uptake for some ions, viz. Ca, was dependent upon root anatomy.
(5) The symplasm, by way of plasmodesmata, extended through the endo-
(6) Sites of maximum ion and water uptake were essentially identical.
(7) Removal of the cortex (stele stripping) eliminated some barrier
to ion and water uptake thought to be the endodermis.
Many experiments of this nature point out the need to study water
and nutrient uptake on the basis of an individual root. Survey of the
literature shows that there is a virtual absence of any comparable research
involving species of fruit trees. However, the relative location and
anatomy of the site of active uptake and translocation for some tree
roots is similar to that reported for common experimental plants such as
barley (48,65,109,154,174,189,191). Consideration must then be given,
assuming that all roots have a zone of active absorption and translocation
of minerals and water, to: (1) anatomical features of this region that
seem to have some functional significance and (2) how these features may
differ between plants, thus affecting the functioning of their roots.
Nearly every aspect of the root anatomy in the active absorption
zone has some meaning in terms of root function. The epidermis in the
suberized state has been proposed as an important ion barrier (28,48);
however, the endodermis is considered to be the primary barrier regard-
less of its actual role. The importance of the epidermis in explaining
absorption differences between plants must consider factors other than
its physical presence. Most plant roots have an epidermis but each
epidermis does not have the same dimensions nor does each one exhibit
the same degree or rate of impregnation with suberin or lignin (48,126).
Quantitative and qualitative differences in suberization of the epider-
mis have been observed between several species of citrus (48). Moreover,
citrus has a hypodermis which suberizes and thus may enhance the barrier
effect of the epidermis.
The cortex is the primary accumulation site for ions. Thickness of
the cortex and the proportion of individual cell contents may be impor-
tant factors. An increase in the thickness of the cortex provides a
larger absorption surface area. Then, an ion may follow 1 of 2 pathways
once it is absorbed, either through the symplasm or into the vacuole.
Cells with large vacuoles may be more competitive for ions thus reducing
the radial flux to the xylem (28).
The endodermis is generally agreed to serve as a passive barrier to
water and ion movement due to casparian strips (43,126). Two features
of the endodermis are, therefore, significant. First, the time and rate
of formation of casparian strips and subsequent suberization may be
important. Second, unthickened endodermal cells (passage cells) are
frequently found adjacent to xylem arches. These cells represent per-
meable sites in the endodermis. They have been observed in Citrus (48).
The significance of anatomy in the root vascular cylinder at the
level of active uptake is more controversial than any other zone in the
radial pathway of ion and water transfer (137). The stele in this zone
is composed of parenchyma cells and mature xylem and phloem elements.
Several hypotheses have been proposed which attempt to explain the
function of these tissues. Briefly, it is thought that ions are released
either passively or are actively secreted into the xylem. Active secre-
tion is considered a reversal of the ion absorption that occurs in the
cortex. Therefore, the ions must cross a membrane, either the plasma-
lemma of parenchyma cells adjacent to metaxylem elements (137) or the
same membrane of the xylem elements themselves (137). Xylem parenchyma
cells have been implicated because they have a known large capacity to
accumulate ions (137). The xylem elements have been implicated on the
basis of a recent report which showed a positive relationship between
the amount of ion absorption along a root and the number of xylem elements
that had retained their cytoplasm (6). The actual mechanism of ion
transfer to the xylem is unknown, but it is not unreasonable to assume
that the certain stelar cells are involved. Quantitative differences in
these cells within the stele may be responsible for differences in
physiological behavior (137).
An area of the root which has not been as fully investigated as the
regions closer to the root apex lies above the root hair zone and the
site of active water and ion uptake. Suberization and lignification of
primary tissues or periderm and cambium formation during secondary growth
supposedly prevent the easy movement of water and minerals in this region
The physiological significance of this supposed impermeable area
has not been determined. It is known that water and minerals may enter
through this region in some plants (1,110,125,128). Entrance of these
materials into this area is probably due to pathways opened by emerging
lateral roots, growth cracks, and lenticels (126). Also, the observation
that high rates of water flux increase the width of the absorbing zone
significantly is indicative of the complexity of uptake. Thus, it seems
probable that the contribution to absorption by this zone may be con-
siderable, especially by the suberized roots of trees (1.,125,128).
Considerable attention has been given to root hairs and root volume
because of their possible relationship to water and ion uptake. Root
hairs are found in many woody plants although rarely in large numbers
under field conditions (16,47,111,179,180). It is difficult to obtain
field samples for study because of the fragile nature of root hairs.
Development of root hairs is influenced by root environment and their
presence can add to the root surface in contact with the soil. There is,
however, no consistent evidence indicating that root hairs are important
for either water or mineral ion absorption (47,126,157), Root volume
has been the physical dimension of a root most consistently related to
uptake in controlled environments (19,43,183,184,185). One cannot assume
that the same relationship would be equally applicable in a field environ-
ment where many variables could influence root volume,
The Soil Environment and Its Effects on
Root Structure, Function and Distribution
The edaphic environment of a root system is complex and highly
variable and has a profound but meaningful influence on the development
of root systems and root growth. The task of identifying the components
of this environment and studying their relationship to plant roots has
been difficult. The numerous physical, chemical and biological aspects
do not function independently but are strongly interrelated. The plant
root is continuously responding to and "integrating" changes in its
surroundings. Thus, researchers have often been restricted to studying
only 1 or 2 factors under artificial conditions. The real cause of a
noted response is not always known with certainty in these circumstances,
as in the field. Nevertheless, there is ample evidence to support some
general observations regarding specific soil factors.
The physical entity which supports plant root systems is the soil,
Texture is an important soil characteristic based on the percentage
particle-size distribution of the solid mineral matter. Pore size and
distribution, another significant physical characteristic of soil,
results from the spatial arrangement of the particulate matter. The
texture of a soil is intimately related to soil moisture and aeration.
Mineralogical attributes of the various particle fractions greatly
influence soil chemistry. Thus, soil texture exerts a rather broad
influence over the soil environment in addition to influences that result
directly from its own physical nature. Indirect effects of texture upon
roots are discussed where appropriate in later sections. This discussion
is devoted primarily to the direct physical interaction of soil texture
and root growth.
Field observations have shown that certain soils or soil layers
limit root growth. Tobacco and corn roots were restricted by tillage
pans (56,71,119). Citrus roots were prevented from further extension by
a subsoil clay layer (37,75). The direct cause of this restriction is
not always known, but it is generally agreed that "soil strength" has a
significant role. Root growth here is mechanically impeded.
The ability of soil to limit root extension is approximated by
determining "soil strength" with a penetrometer (60,96,98). A thin
probe constructed to be physically similar to a root is driven into the
soil and the resistance measured. The force required to penetrate the
soil is related to texture and bulk density. Root extension ceases if
the soil offers a greater resistance than the maximum pressure a root
Soils containing a large percentage of fine particles such as clays,
have higher bulk densities and are more resistant to root penetration
than sandy soils. Clay soils are less susceptible to compression and
have a greater number of small pores. Controlled experiments that
employed materials with known pore sizes showed that roots did not
penetrate pores with diameters smaller than those of young roots (220)
Soil density is related to texture and may impede root growth as
described above. Studies of bulk density showed that this factor was
negatively correlated to root extension (205). In one case (208), an
attempt was made to establish a "threshold bulk density." Increases
in bulk density in the form of tillage pans and other compactions due
to cultural operations restricted root penetration (56,71,119).
Mechanical impedance affects root structure and distribution. When
growing roots encounter strong resistance in the soil, they ramify and
individual roots become shortened, stubby and often expand radially in
the region behind the meristem (11,12,60,71). The number of root hairs
may increase and act as miniature anchors as the root exerts pressure on
the soil (60). Lateral and tap root development may be severely reduced
Changes in root structure resulting from mechanical impedance reduce
the rate of water and mineral absorption because root growth is restricted
Water is essential for all stages of plant growth and development
and is normally obtained from the soil through the root system. Root
growth is affected as well as the growth and yield of the whole plant when
there is insufficient moisture in the soil profile. Plant roots are also
unable to function when soil pore spaces are filled with water. Plants
are subjected to a "physiological drought" under these circumstances,
where water absorption is inhibited indirectly as a result of reduced
aeration (126). This phenomenon is not completely understood, but it
is apparent that roots are able to survive when inundated by water as
evidenced by their growth in aerated solution culture.
Plants vary in their ability to withstand wet or dry habitats and
also differ in the anatomy of their roots (67). Roots of plants from
wet environments frequently have a larger cortex with a greater per-
centage of intercellular space (126). Some differences in cortical
porosity have been demonstrated for several species of citrus (143).
Furthermore, the soil moisture status affected quantity and length
of root hairs (48).
Roots develop a characteristic anatomy which reflects the moisture
status of their environment, but other changes occur in response to
daily and seasonal moisture fluctuations. The primary response is the
cessation of root elongation as has been reported during seasonal dry
periods when root growth stops despite an otherwise favorable environment
(48,162,171,176,179). Failure of roots to continue elongating is thought
to result from the inability of a plant to maintain the necessary water
potential gradient to cause water movement into the root (205).
There are few reports regarding the quantitative effects of soil
moisture stress on such factors as plant growth or yield. Nevertheless,
some evidence is available implying basic differences between plants in
their response to soil water deficits. Such differences may be anatomical
or physiological. Studies with citrus trees have repeatedly demonstrated
the greater ability of trees budded to rough lemon and sweet lime root-
stocks to extract water from the soil (42,161,162,163). Roots of these
rootstocks have an inherently high osmotic value and are therefore able
to maintain water movement during greater stresses compared to other
rootstocks (48); however they also have larger root systems.
Soil moisture stress modifies root distribution. Irrigation water
is used to supplement rainfall to prevent water deficits in most areas
where cultivated plants are raised. It has long been debated whether
depth of rooting and root proliferation are controlled by irrigation.
Rooting depth was not increased, in one case (115), in contrast to more
recent reports (33,35) showing that less frequently watered trees rooted
deeper. Studies with citrus trees showed that frequency of irrigation can
significantly alter root distribution (33,35). Those trees which were
subjected to frequent additions of water had a large percentage of roots
near the surface. Those less frequently irrigated had more evenly dis-
tributed root systems. Roots growing in soil with a continuous supply
of water tend to branch and ramify throughout the available volume of
soil, while roots growing in drier soil are prone to extension of indi-
vidual roots rather than the formation of laterals (53).
An adequate supply of 02 is an essential prerequisite for good root
growth. Poor soil aeration limits the metabolic activities of roots,
eventually reducing plant growth. Soil 02 concentration is usually
lowered while the CO2 level increases during root and microbial respira-
tion. Thus, the 02 and CO2 composition of the soil atmosphere commonly
changes reciprocally and the sum of the 2 gases parallels that of the
aerial atmosphere (21,205). Soil 02 and C02 also show seasonal fluctua-
tions due to changes in soil temperature and water content (21,24).
Moreover, CO2 content generally increases with depth (21,24).
Oxygen consumed by root respiration must be replaced and CO2
released must be removed in order to avoid permanent damage to the root
system. Many plant roots are not affected by high 02 content. They are
capable of functioning within a wide range of 02 concentrations (23).
The critical 02 concentration may vary, however, according to the phase
of root activity (25,140). It has been suggested that the rate of 02
supply and CO2 removal are more significant than their content in the
Gas exchange occurs by diffusion and mass flow (95). The effective-
ness of these processes is related chiefly to soil texture. Soils with
a loose texture and structure readily permit gas movement (21,205). This
is evidenced by the analysis of air samples collected at different depths
from these types of soils (21). The gas composition of the samples was
very similar to that of the above ground atmosphere (21). Finer textured
soils have a larger number of small pores that restrict gas movement and
consequently, allow CO2 to accumulate to high levels.
Soil water content is an important factor in gas exchange. Oxygen
diffuses considerably faster through air than water. There is very little
gas exchange when the total pore space of a soil becomes filled with water.
Roots must rely on their inherent capacity to tolerate flooded or essen-
tially anaerobic conditions. Thus, it has been suggested that deeply
rooting trees possess in their deep roots an ability to tolerate low soil
02 (126,201,202,221,222). Research with small citrus plants demonstrated
some rootstock differences during controlled flooding tests (79). Rough
lemon, 'Rangpur' lime, and 'Carrizo' citrange had satisfactory tolerance.
Sour orange, sweet orange, and trifoliate orange were damaged by flooding.
Cleopatra mandarin was the least tolerant of all rootstocks tested. Root
damage has also been shown to result from the microbial production of
toxic substances, such as H2S, under anaerobic conditions caused by
Some plants, such as those growing in bogs, are adapted to wet environ-
ments and have a well-developed internal transport mechanism for supply-
ing 02 to their roots. This has been verified with radioactive tracer
studies (10,97). Circumstantial evidence indicates that a similar system
is present in some mesophytic plants (97). Differences in root anatomy,
especially gas space as has been found in some citrus rootstock seedlings,
may be significant (143).
Root function is impaired during periods of deficient aeration (105).
Moreover, root efficiency also falters during periods of unusual 02 demand.
For example, root 02 requirement is greatly increased as a result of the
stress encountered upon growth into a mechanically stronger soil layer.
Roots also require additional 02 as soil water potential decreases. Water
uptake is reduced if this need is not met.
Experiments with citrus and avocado seedlings illustrate the effect
of deficient aeration on ion uptake. Absorption of all ions was reduced
except for N. This nutrient was supplied in the NO3 form. Liberation
of 02 during its reduction was thought to enhance its uptake (130,131,
The 02 status of the soil indirectly influences the absorption of
certain micronutrients. Iron chlorosis in sweet orange was associated
with 02 deficiency and limed soils (212,214). There is considerable
uncertainty as to whether an 02 deficit per se or CO2 was responsible for
the observed effects (213,214).
The morphogenetic effect of 02 is manifested primarily through the
activity of apical and lateral root meristems. Low soil 02 stimulates
branch root formation and depresses root elongation (90). Root hair
formation requires adequate aeration (47,48,93).
The primary source of mineral nutrients for plants growing in soil,
is the solution bathing the roots (9). The quantitative and qualitative
composition of this solution is constantly changing as nutrients are
absorbed and replaced by natural processes and fertilization. Those changes
resulting from normal root activity, do not generally have any unfavorable
effects upon the roots. Undesirable consequences on root growth and
function are usually due to the accumulation of excess salts in the soil,
and specific ion effects.
Root growth is reduced and root maturation hastened when the total
salt concentration of the soil solution reaches an inordinate level. The
high osmotic potential of the external solution prevents water absorption
and seems to reduce root permeability. Plant root permeability, neverthe-
less, is apparently affected only temporarily and roots recover after
a short period of time. Rate of water uptake, however, is then less than
the original rate (126). Ability of a plant to tolerate high salt levels
is a function, in part, of the species. Considerable research has shown
that Cleopatra mandarin and Rangpur lime have high salt tolerance, while
rough lemon, many mandarins, tangelos, sweet oranges and sour orange
have moderate tolerance. Sweet limes, trifoliate orange and citrange
cultivars generally are intolerant to salt (39).
There are several well-established specific ion effects. Excessive
quantities of either Cu, P, N, C1 or B are known to damage citrus roots
(35,39,84,193,198,199). Roots from high Cu soils differ in color from
healthy ones (26,69). They are shriveled, corked and lack an active meri-
stem. Lateral roots were reduced to stubby outgrowths. Copper accumulates
in the hypodermis, endodermis and pericycle. These accumulations were
thought to "plug" the root and interfere with root function. The number
of feeder roots was decreased and their distribution modified by excess
P and N in the soil (35,84,192,198,199). Similar responses were noted in
Fe deficient soils (83). Root hair formation was reportedly depressed
by high Ca (47).
Ca and B are recognized as being essential for root growth (41,126).
Roots have failed to penetrate soil in which Ca is absent (88,114). Neither
nutrient is apparently readily translocated within the root system.
Therefore, they must be available in the root environment. Roots become
short and stubby and eventually die when either Ca or B is absent (41).
Other nutrients, however, such as K, are translocated via the root system
from soil areas adequate in K to roots growing in soil containing no K
The overall fertility level in a soil affects root quantity and dis-
tribution. Trees growing in infertile soils often have a greater root:shoot
ratio as compared to those growing in fertile soils. Differences in this
ratio have been interpreted to indicate that less root growth or perme-
ation of the soil is needed to supply the required nutrients in the lat-
ter case (175,176,177,178,181).
Changes in nutrient concentration and balance indirectly influence
roots through changes in soil pH, structure and modifications of the
microbial population and their activities, in addition to direct effects,
Most plants are adapted to a wide range of temperatures. They are
able to function satisfactorily within this range but optimum growth is
generally restricted to only a small part of the spectrum. Moreover, the
temperature response of a plant varies with species, stage of develop-
ment and 02 supply (126). It is also possible that the temperature
range, and that part of it optimum for growth, are different for the
roots and the aerial parts of a plant. The soil temperature reportedly
most favorable for root growth of deciduous fruit trees is 25C (126).
Citrus trees have maximum root growth rates at slightly higher tempera-
tures, which are species dependent (155,156,159,160,171).
Root growth and function are not greatly impaired b' soil tempera-
ture fluctuations until very high or low temperatures are reached. Cool
or cold soils halt or severely reduce root extension and maturation (218).
Roots of herbaceous annual plants generally deteriorate and die when
growth stops. Woody plant roots, however, often only become quiescent
and resume growth when the environment is again favorable (103). Roots
of many deciduous tree fruits cease growth at soil temperatures rang-
ing from 4 to 10oC; however, death does not occur until much lower tempera-
tures are reached.
The reported "vital" temperature for citrus is 130C (155,216).
At this temperature, most physiological activities appear to cease,
including root extension. It is not known whether this temperature can
be equally applied to both roots and shoots or to all species of citrus.
Nevertheless, in California, soil temperatures below 130C apparently
caused root growth to be periodic (171). Root growth is not normally
limited by low soil temperatures in other areas, such as Florida (59).
Soil temperatures well above the optimum also prevent root growth
(44). Only a small quantity of active roots was found in the top 30 cm
of soil in some deciduous fruit orchards during July and August because
of high soil temperature and rapid loss of moisture (168). A scarcity
of citrus feeder roots near the surface has also frequently been observed
in Florida (37,75) where the surface soil temperature may reach 330C
during summer months (59).
Changes in root activity are not entirely associated with soil
temperature (112). For example, some citrus root growth was observed
in California during the winter months (51). Soil temperatures and
some internal factors apparently interact in controlling the growth of
individual roots, but water stress and temperature often override the
intrinsic controls (104,107,171,227). Further evidence obtained from
observations of roots grown in uniform environments supported this
Temperature extremes that do not cause root death normally affect
only the rate of root activity in water and ion uptake (36,94,100,124,
134,148). Thus the transpiration rates of young, budded citrus trees
were reduced at root temperatures above and below the optimum (139,161).
Citrus and apple trees continued to absorb N at temperatures as low as
0 to 5C but at reduced rates (3,14,40,209). Greenhouse studies with
citrus have shown marked effects of variations in soil temperature and
rootstock on the scion leaf mineral content. Significant differences in
absorption and translocation were noted for most nutrients (159).
Root structure is affected by temperature changes primarily as a
result of modifications in differentiation. Peach and apple roots grown
at temperatures above their optimum, 200C, had tips containing fewer
primary cells, were smaller in diameter and vascular tissue had differ-
entiated closer to the apex (153). Cortical cells were suberized and
many appeared dead. Roots grown below optimum temperatures were white
and succulent. Differentiation occurred at a greater distance from the
root tip. Other plants have an opposite reaction: root extension
proceeds at a faster rate than maturation at higher temperatures com-
pared to lower temperatures where maturation follows root growth closely
Temperature indirectly influences root morphology particularly
the occurrence of root hairs. The zone of elongation, where root hairs
emerge, is lengthened and persists longer when root maturation is slowed
by either high or low temperature (153).
Solution culture studies have shown that many plants are capable
of satisfactory growth over a wide range of pH values (7,8). The pH
range for adequate root growth in soil is more limited, however, and
the primary effect of pH changes is indirect, Generally, a soil pH
near 6.0 is considered optimum for most plants because large departures
from this value render essential nutrients unavailable. Also, at very
low pH values other elements such as aluminum and manganese become
soluble in toxic amounts.
A change in pH is by definition a change in H+ concentration. Thus
fluctuations in soil pH do not preclude a direct effect on roots due to
H+ toxicity. Smith (169,194) has been a particularly active proponent
of this hypothesis. He contended that most pH experiments were poorly
designed and were unable to separate direct from indirect effects of pH.
Smith (194) studied the ion uptake pattern of citrus seedlings growing
in solution culture with a pH of 4.5 or 6.0 and ample available nutrients.
Fewer ions were absorbed at the lower pH and root growth ceased. Other
researchers have conceded a possible toxic effect of H+ at very low pH
values but there is less agreement about its direct effect at pH values
near neutrality (7,8,99,170).
Citrus root growth and morphology were modified by pH in the experi-
ments of Smith (194). Root elongation was severely reduced or stopped
below pH 4.0 to 4.5. Roots became thickened and turned brown but were
not killed. Roots were white, smooth and elongated freely at pH 5.0.
Root hairs were most prevalent at pH 6.0, which allowed rapid root
growth. Root growth was proportionately reduced without any apparent
deviation in morphology above pH 6.0.
The micro-environment immediately adjacent to roots, or rhizosphere,
is the habitat for a large and diverse population of microbes, primarily
bacteria, fungi, and actinomycetes. These organisms are involved in
a large number of activities which include important mineral transfor-
mations, various types of associations and specific pathogenic infections
of roots. Recent emphasis in rhizosphere investigations has been to
determine how these microbes affect the supply of nutrients to roots and
the characteristics of roots that are related to water and nutrient
The microbial composition of the rhizosphere has been used to gather
information regarding the root microenvironment and root physiology.
For example, isolating anaerobic bacteria from well-aerated soil indicated
a localized 02 deficiency (20). Furthermore, many plant roots seem to
exude a mucilagenous material, called mucigel, that is often cited as
a dominant factor affecting the qualitative nature of the microflora
(20,102,118). The mucigel may be colonized by specific microbes depend-
ing upon its chemical nature. Mucigel has been observed on citrus roots
Many essential elements are mineralized or rendered unavailable by
microbial activity, especially that of bacteria. They are particularly
significant in N transformations. Certain species which form nodules
modify root morphology and function but bacteria with the exception of
the N-fixing species do not generally affect roots other than by inter-
cepting nutrients which might otherwise normally be absorbed by the root.
Fungi are virtually ubiquitous in the rhizosphere and are responsible
for many pathogenic infections. Some form mycorrhizae which greatly
alter certain roots of Pinus and related gymnosperms (20,106). This
ectomycorrhizal association causes pine roots to form an apparent di-
chotomy. Root hairs are absent. Fungal mycelia form an external sheath
and penetrate the cortex up to the endodermis. The function of this
fungal association remains moot but considerable evidence supports a
role in the uptake of immobile ions such as P when the native fertility
is low. This suggested role is manifested, in view of the absence of
root hairs, through the increased effective absorption volume and sur-
face area resulting from infection (20,123,126,127).
Other types of fungi form endomycorrhizas (92,146). The fungus
invades the cortex and does not form a sheath. There is little or no
change in root form. The nature and extent of this association has been
largely overlooked and is not well understood. Citrus forms this type
of association without any apparent change in root anatomy (120,147).
Root hairs are reportedly absent in endotrophic roots as with the
Two soil-borne organisms, burrowing nematode (Rhadopholus similus
Cobb) and citrus nematode (Tylenchulus semipenetrans Cobb), are serious
pests of citrus.
Burrowing nematode occurs to considerable depths in deep sandy soils
in Florida where it is of cardinal interest; however, high soil tempera-
tures and rapid loss of moisture prevent a large population from develop-
ing in the surface 2 to 3 feet (62,82). Burrowing nematode attacks young
feeder roots causing a quantitative decrease in the number of these
roots as compared to healthy trees (58,74). Nematodes enter a root in
the zone of elongation or at the root tip. Growth ceases if the root
apex is damaged. The primary tissue damaged is the cortex. Other tissues
are invaded but not suberized or lignified cells (58). Certain apparent
morphological changes have been noted, one of the most interesting being
the induction of local proliferation of the pericycle when penetrated by
a nematode (58).
Citrus nematode is a serious pest in California and is common in
Florida. It poses a more serious threat because it can form cysts.
Also, it occurs throughout a soil profile. A study conducted in Cali-
fornia showed some basic differences in entry sites and damage caused
by this nematode compared to burrowing nematode (190). The citrus nema-
tode penetrates older fibrous root tissue and does not normally feed
deeper than the outer layers of cortical cells. This nematode does
not appear to have any profound physiological or morphological influence
because invaded roots are able to initiate new roots in the vicinity of
infection. Invaded areas are readily isolated by wound periderm.
The Influence of Depth of Rooting
on Root Structure and Function
Most roots used for anatomical studies have been obtained from young
plants grown in a controlled environment. Those roots collected from
more mature plants growing in the field have been taken from shallow
depths in the soil. No reports describing the anatomy of roots collected
from below the soil surface were found in the literature.
A deep and extensive root system is generally believed to be a
drought survival mechanism for trees (89,123,166,226). Many deep-rooted
plants appear to remove water in proportion to the quantity and activity
of their feeder roots when soil moisture stress is absent and the soil
profile is uniformly moist (4,5,46,142,207). Deep-rooted plants often
have a large percentage of their roots located at shallow depths (4,57,
108,144,225). Therefore, soil moisture is depleted first at this level.
Reduced availability of water in this region is then compensated for by
increased absorption at successively greater depths in the soil. It is
only under these circumstances that deep roots have been shown to markedly
increase their rate of water uptake and substantially contribute to the
total water needs of a plant (205).
There are conflicting statements in the literature concerning the
response of roots growing in soils of different water potential to total
plant water requirement (86). Nevertheless, data from irrigation (87)
and "split-root" (27,138) studies have generally supported the conten-
tion that when roots in one area are stressed, water uptake increases
in other root zones where the water potential is higher. The water uptake
of deep roots in frequently watered plants is usually smaller than that
of less frequently irrigated ones in irrigation experiments. The root
system of a small plant is divided in half with each part placed in an
environment (soil or solution) that differs in water potential in "split-
root" studies. Root response to water stress can be observed in this
The moisture depletion pattern of citrus root systems is reportedly
similar to the typical case described above; however, these studies have
not always been supplemented by a study of root distribution (40,91,121,
122,187). Thus, it is not known if water absorption occurred in propor-
tion to root quantity. A deviation from this relationship was apparent
in one study where moisture depletion and root' distribution were compared
(30). Deep roots were found to be more efficient, i.e., they absorbed
more water per unit weight of root compared to surface roots. Moreover,
the water absorption pattern for the same rootstock varied with soil texture.
Available information regarding the relative extent to which nutri-
ents are absorbed from different depths in the soil is meager. Early
investigations were limited to those of Weaver (54) who placed fertilizers
in different zones in the soil, separated the zones with wax seals per-
meable to roots, and then recorded plant mineral content and yield.
These investigations were conducted with relatively shallow-rooted
plants but Weaver showed that moderate quantities of nutrients were
absorbed by deep roots.
The method of Weaver has been criticized because it was subject to
limitations imposed by changes in root structure and function due to
localized high nutrient concentrations (152). The development of radio-
active tracer techniques has greatly enhanced the investigation of
nutrient uptake with depth, but research by the method was again with
grain crops whose root systems are rather shallow, fibrous, and other-
wise dissimilar in many respects to those of tree crops (13). Uptake
of nutrients with depth, as determined through the use of tracers,
follows somewhat the same pattern as water but appears to be consider-
ably more complex (61,151,152,153,158). Uptake of 32P and 45Ca generally
decreased with root concentration and depth; however, it was clearly
evident that the selective absorption or perhaps translocation abilities
of roots from different depths, were dissimilar. The uptake pattern of
the tracers was attributed to inherent factors rather than external ones.
Further study revealed the difficulty of interpreting tracer uptake data
due to the strong influence of soil water content and the lack of root
distribution data. Radioactive isotopes have been used with tree crops
but largely limited to studies with young plants.
MATERIALS AND METHODS
Description of the Experimental Site
Samples for this study were collected from an experimental planting
of 'Orlando' tangelo trees located in Lake County, Florida. The plant-
ing consists of an approximately equal number of trees on each of 11
rootstocks, spaced at a distance of 4.5 x 9.0 m. The trees were set out
in 1961 as part of a commercial tangelo orchard using a completely
randomized design. The orchard and the experimental trees have received
standard commercial care to maintain fruit production.
The soil of the experimental site, Astatula (formerly Lakeland) fine
sand, is a typical soil of Florida's ridge area that is planted to
citrus. It is classified according to Great Soil Group as an Entisol.
It is well-drained and has low water and mineral nutrient retaining
properties. The soil is underlain at depths of 2.5 to 3.5 m by a finer
textured horizon of undetermined thickness. The clay content of this
horizon varies from 10 to 38 percent (37).
Trees in the experimental planting differ considerably in their
adaptation to the environment of the site. Therefore, 4 rootstocks were
chosen for study on the basis of known differences in tree size, fruit
production and root system characteristics. These were rough lemon,
sour orange, sweet orange, and 'Rusk' citrange, representing the 2 major
rootstocks of Florida, and 2 of lesser importance, respectively. These
trees were used for investigation of soil water depletion and root
anatomy and morphology. In addition, soil samples from the root environ-
ment of each rootstock were taken for certain physical and chemical
Field Sampling Procedure
Soil Moisture Depletion
Moisture changes in the soil profile of each rootstock were measured
with a Troxler neutron probe, Model 1603, and scaler, for 1 year begin-
ning in April, 1973. Aluminum access tubes were installed just inside
the drip line of 16 trees, 4 trees of each rootstock (Fig. 1). One tube
was installed per tree at either a NE, SE, SW or NW location. Readings
were taken at 30 cm depth intervals, beginning at 30 cm, to a depth of
3.6 m. Trees on rough lemon were measured to a depth of 3.9 m because
they were generally deeper rooted as compared to those on other root-
stocks. Also, there was a gradual increase in the depth to the clay
horizon from the south to the north end of the planting where the trees
on rough lemon were located. The neutron probe measures the water con-
tent by volume of a sphere of soil approximately 30 cm in diameter.
Therefore, it could not be used to measure the surface 15 cm of soil
without serious error. Water content of the 0 to 15 cm zone was deter-
mined gravimetrically. Combined samples of 3, 2.5 x 15 cm cores of soil
were collected around each access tube.
Fig. 1. Planting plan of the experimental site showing the location of
various rootstocks for 'Orlando' tangelos. Trees used in this
study ( --I) and access tube locations ( ) are indicated.
Key to Rootstock Symbols:
Palestine sweet lime
'Christian' trifoliate orange
'Rubideaux' trifoliate orange
'English Small' trifoliate
Soil moisture content was determined twice weekly from 1-minute
readings taken at each depth. A value as percent soil water by volume
was obtained from the instrument calibration curve using a ratio of
these readings to the mean of 3 standard readings, i.e., readings taken
with the neutron probe above ground and in its shield.
A calibration curve for this instrument was prepared at another
location prior to its use in this study. Extensive recalibration at
the experimental site for this study did not indicate any change in
Soil Samples for Chemical and Physical Analyses
Access tubes used for measuring soil water content were installed
by boring holes of slightly smaller diameter before the tubes were
inserted. Soil obtained from these borings was collected in 30 cm
intervals and retained for chemical and particle-size analysis. In
addition, undisturbed cores in 6 cm brass cylinders, 3 cm deep, were
collected at depths of 7.5, 20.0, 45.0, 60.0, 120.0, 150.0, 210.0,
240.0 and 270.0 cm from several locations to determine the soil moisture
characteristic curve for the experimental site.
Roots of the selected rootstocks were obtained by excavation near
the tree drip line in April 1973, for anatomical study and observation
under the scanning electron microscope (SEM). Relatively undisturbed
soil samples containing roots were taken from 2 trees on each rootstock
from the surface, the subsoil clay horizon, and at an intermediate depth.
The same trees were not used for root sampling and the soil water study.
Sample Preparation and Analysis
Soil Chemical Analyses
All field samples were prepared for analysis by air drying for 48
hours and sieving through a 2 mm screen. An outline of the various
determinations made on these samples is shown in Table 1.
Soil Physical Analyses
Particle-size analysis. The soil of the experimental site is essen-
tially pure sand as determined by a previous study (37). Therefore, only
the texture of subsoil clay samples was determined using the hydrometer
Moisture characteristic curve. The moisture characteristic curve
of the field samples was determined through the use of a special porous-
plate pressure cell and a procedure developed by van Bavel and Reginato
(206). This required collecting field samples in special brass cylinders
as previously described, to accommodate data collection by the pressure-
cell method. Briefly, the procedure involved placing the soil cores with
retaining cylinder, in the pressure cell (commercially known as a Tempe
cell), saturating the soil, then closing the cell and connecting it to a
regulated source of compressed air. The moisture characteristic curve,
i.e., a plot of volumetric soil water content versus soil matric potential,
was obtained by adjusting the air pressure to a preselected value,
allowing the soil core to come to equilibrium, then weighing the cell.
Appropriate data were obtained by repeating this procedure at a number
of preselected air pressure values. Air pressure was gauged as cm of
water by bubbling air through a column of water the required height.
e q C 4 o
) ZOin S0
0- -4 U CC1
Q) 40 C1
0 0 u x O 3
0 r mr O
-4 En Z r ca 4
.O C) C C C 0 0 0 C r
SIi O CC O C M
C 0 ae a) = Iz C)
0 0 Sl E l M CC m1
a -- r CCC)
)o rI 4eq azme r = o
c o C a C C ZO4B
a 0 4.4f b C
C) O ~ .-4-r C1 mf S C )
O o c m n m I a oi
-4,0-- 0 0 0 4 0* C -
; O Ca 0 O0 r r<
-4 o 0 a a
0 mw 'a o Q r
Ca 41 t) o
SI IC z V C C
) -h OC N
*-0 0 P C) 0CC)
E C 0 O4 CC CC
C C) C C O O CC
O CC C CO CC C '
o 0 C o
Om m O
T-I CO ) C)
. CD I TI B0
o ^U nj
O -r O
0 O fl
P C )
,- Q CJ
Subsamples of the soil cores were used to determine the 15-bar water
content by the pressure-plate method when the pressure-cell phase was
Anatomy. Roots were washed in water to free them of adhering soil
particles. Approximately 20 root tips, 20 mm long from each rootstock
and depth, were selected, divided into 5 mm segments, fixed in 50 per-
cent formalin-acetic acid-alcohol (5-5-90), embedded in paraffin, sec-
tioned at 10% and stained with safranin-fast green by standard pro-
Scanning electron microscopy. Roots were washed and selected as
described above then fixed in a solution of 6 percent glutaraldehyde in
0.1 M phosphate buffer, pH 6.8, for 2 hours. This was followed by
rinsing in several changes of buffer, and post-fixing in 2 percent Os04
in buffer for 2 hours. Segments were dehydrated in a graded water-acetone
series, infiltrated with a freon solvent and critically point dried with
Freon 13. Selected segments were mounted on viewing stubs and coated
with a gold-palladium alloy. A set of roots collected from seedlings
grown on filter paper was similarly prepared. Roots were examined with
a Cambridge Mark II SEM at 20 KV.
Certain soil moisture data were submitted to analysis of variance.
Total soil profile water losses for selected periods during the year
of study were analyzed as a randomized complete-block design using time
as a block. Soil water content data for the period October 10 to 20,
1973, were analyzed as a 4 x 12 factorial experiment using rootstock
and depth as the variables. The analysis was repeated using rootstock
replicates, which represented access tube locations, as blocks. A 4 x 13
factorial analysis was performed on feeder root weights of the same
trees. Significant differences between treatments were determined using
Duncan's Multiple-Range test at the 5 percent level. Correlations were
made between feeder root weight and water use.
Characterization of the Soil Environment
Moisture relationships of a soil are controlled largely by soil
texture, A particle-size analysis performed earlier (37) showed the
soil of the experimental site, Astatula fine sand, is a sand with small
amounts, 1 to 2 percent, of silt and clay size particles. A clay hori-
zon which lies under this sandy soil beginning at a depth of approximately
210 cm contained 21 to 34 percent clay, generally increasing in clay
content with depth. This horizon is part of the Citronelle Formation
and kaolinite is the dominant clay mineral (68).
The moisture release curve for soil from the experimental site
(Fig. 2) shows the water retention capacity for the Astatula sand is
very low. The water content at only 80 cm of pressure potential was
reduced from its maximum value to only 5 percent by volume. A small
amount of organic matter in the surface 23 cm (Table 5) affected the
saturation moisture value without influencing the remainder of the mois-
ture release curve to a significant extent.
Release of moisture in the clay horizon (Fig. 2) is more gradual,
although the initial water content was similar to that for the sand.
The release curve for both the sand and clay horizon shows that a
small change in water content is associated with a much larger change in
soil matric potential after an initial quantity of water is removed.
E C3 l
I~ I C" 4
c m 0
0 a C)
~f) I fL
% -IN] NOD Iiivm IIDi]Nfl1OA
Values from the moisture release curve (Fig. 2) were used to deter-
mine the amount of water commonly referred to as "readily available"
(RAW). Field capacity moisture content was obtained from the appropriate
curve (Fig. 2) at a matric potential of 60 cm. Resulting values indi-
cated 10 cm of water was available in the soil profile to a depth of 210 cm
(Table 2). The clay horizon from 210 to 330 cm contained an additional
11 cm of available water.
Ability of water to move through a soil is largely a function of
water content and pore size. Hydraulic conductivity for the sand and
clay horizon (Fig. 3) is shown to parallel their respective release
curves with regard to changes in volumetric water content. Water moves
at a rate 100 times faster in the sand than in the clay at water content
values near saturation. Conductivity changed rapidly in the sand as
water content decreased.
The sandy soil has an average bulk density of 1.56 g/cm3 and thus
has approximately 41 percent pore space (Table 3). Mean bulk density
of the clay horizon was 1.73 g/cm3, corresponding to a porosity value
of 33.7 percent. Size and distribution of the pores in the sand and
clay horizon is presented in Table 4 and Fig. 4, respectively. Pores
in the sandy soil showed an unbalanced distribution with the majority of
the pores having the larger diameters. These pores retained the major
portion of the soil water but also drained under relatively low tension
(Fig. 2). Soil obtained from the clay horizon had few large pores.
Pore sizes were more evenly distributed (Table 4) and therefore the
soil released water at a more gradual rate (Fig. 2).
Table 2. Selected soil moisture values for the experimental site.
Depth Moisture Values, percent by Volume Available Soil Moisture
cm F.C.z 15-bar cm
Table 2 continued.
Depth Moisture Values, percent by Volume Available Soil Moisture
cm F.C.z 15-bar cm
270.0 29.23 21.30
330.0 35.65 25.24
MEAN 29.35 20.34 10.81
Z"Field Capacity" value was the moisture content at 60 cm pressure from
Fig. 3. Hydraulic conductivity-water content relationship for Astatula
fine sand and the subsoil clay horizon from the experimental
- 10-4 .
3 5 10 15 20 25 30 35 40 45 50 55
VOLUMETRIC WATER CONTENT--%
o 0 -22.5 (m
v 22.5- 210
Table 3. Soil bulk density and porosity for the experimental site.
0-23 23-210 210-330
Bulk Densityz, g/cm3
1.55 1.58 1.73
Soil Porosity, percent
41.5 40.4 33.7
Assumed particle density of 2.65
37A C Od VCOI CO 0/0
viWm0A 3dd "IVIOl AO o/o
Organic matter content of the soil profile was highest at the sur-
face, 0.77 percent, and decreased sharply with depth (Table 5). Little
or no organic matter was detected at depths greater than 180 cm. Cation
exchange capacity of the soil was very low and reflected primarily the
presence of either organic matter or clay minerals. Exchange capacity
of organic matter is approximately 200 meq/100 g. The surface 30 cm of
soil had an exchange capacity of 1.49 meq which corresponds to the 0.77
percent organic matter content found. This indicates that the cation
exchange capacity of the sand is primarily due to organic matter present.
The 3 to 4 meq of exchange capacity in the subsoil layer results from
the presence of kaolinite.
Soil pH was measured in water, 0.01 M CaC12 and 1.0 N KCI. Mean
pH in water ranged from 6.49 at the surface to 5.45 near the clay hori-
zon where the pH continued to decrease another unit with depth (Table 6).
The range in pH included values as low as 3.85. Measurement of soil
reaction in CaC12 theoretically represents the true field pH value because
a 0.01 M salt solution approximates the concentration of the soil solu-
tion in the field. Mean pH values in CaC12 as compared to H20 values
were consistently reduced but the general trend with depth was not altered.
Measurement in 1.0 N KC1 reduced the pH values even further, but again,
without affecting the general depth trend. Measurement in KC1 theoretically
eliminates fluctuations in pH in the field due to changes in soluble salt
concentration because of the high ionic strength of the KCI solution.
These values represent maximum pH development because all exchangable ions
are replaced particularly Al.
Table 5. Soil organic matter content and cation exchange capacity values
for experimental site.
Depthz Organic Matter Range Cation Exchange Capacity Range
cm percent meq/100 g
15 0.77 0.47-1.05 1.49 0.95-2.25
45 0.38 0.19-0.75
75 0.20 0.11-0.36 0.52 0.25-1.50
105 0.14 0.08-0.18
135 0.11 0.08-0.14 0.24 0.10-0.33
165 0.10 0.07-0.14
195 0.08 0.04-0.10 4.00 4.10-5.52
225 <0.05 3.76 3.56-3.96
225 3.88 3.31-4.93
285 3.36 2.28-4.48
ZEach depth value represents the center of a 30 cm interval.
Table 6. Soil pH values for experimental site.
Depthz Mean Values Range
cm H20 CaC12 KC1 H20" CaC12 KC1
15 6.49 5.95 5.79 6.25-6.83 5.58-6.10 5.38-6.20
45 6.11 5.57 5.28 5.70-6.45 4.95-5.93 5.00-5.50
75 6.01 5.46 5,27 5.70-6,28 5.08-5.73 5.08-5.73
105 5.91 5.36 5.22 5.43-6.30 5.05-5.60 4.85-5.55
135 6.01 5.26 5.05 5.40-6.58 5.03-5.48 4.73-5.35
165 5.94 5.29 5.15 5.13-6.38 4.65-5.73 4.63-5.50
195 5.45 5.19 4.86 4.68-5.78 4.80-5.75 4.30-5.40
225 5.11 4.58 4.55 4.25-5.80 3.98-5.20 3.85-5.15
255 4.82 4.36 4.32 4.08-5.53 3.90-5.00 3.55-5.15
285 4.62 4.29 4.23 4.05-5.75 3.90-5.30 3.75-5.05
315 4.70 4.22 4.17 4.15-5,85 3.73-5.28 3.65-5.08
345 4.49 4.10 4.20 3.85-5.78 3.48-5.40 3.58-5.20
ZEach depth value represents the center of a 30 cm interval.
Soil Mineral Nutrient Content
Soil nutrient analyses were conducted primarily to characterize the
soil environment and not to study the influence of rootstock on nutrient
absorption. However, samples were selected for analysis by rootstock,
in such a manner that the mean value for all samples could be considered
as representative for the levels of the soil nutrients in the entire
area the experimental trees were located. Results are given by root-
stock and a mean for all samples, shown as a dashed line, is also given.
The soil levels of NH4 and NO3 ions are shown in Figs. 5 and 6
respectively. All NH4 values were less than 10 ppm. The mean value
was approximately 5 ppm and this concentration was uniform with depth.
NO3 ion is mobile, moves primarily in the soil solution and was relatively
uninfluenced by exchange phenomena. Highest NO3 levels were found at
the surface where they ranged from 8 to 30 ppm. NO3 level decreased
rapidly with depth to less than 4 ppm with little variation between root-
stocks. Downward rate of water movement was hindered by the subsoil
clay horizon, hence a bulge appeared in the NO3 level at a depth which
varied according to the location of the clay horizon.
The level of Bray2 extractable P is shown in Fig. 7. The largest
variation in soil P occurred at the surface with the content ranging
from 40 to 85 ppm. P content decreased below the surface to 10 ppm,
and continued to decrease slightly with depth except in the soil samples
from the trees on rough lemon where P increased below 190 cm. It is
W 4- 0
O -H 4J
> 4 1
a a ^
1- i <
(1 3 .
t- l > e
xl VI) H'd'
1-01 x V-) H I d I (I
H .H 4
4 4- w1
CA ~L^ Lond
0 / 0
C:-) C cn 12 S-0 C~
C-' C~ -j
01xV)- FlIdciI (
M 0 a
o o"n C
1-01 x 17)~- Hid I
SOIL K pp x 0o1
2 3 4 5 6 7
DO 0. 7
-r 0 0-
3 V 0
30- I 1 RL
0 0V Rusk
Fig. 8. The relationship of rootstock and soil depth to the level of
extractable K in Astatula fine sand and the subsoil clay hori-
zon from the experimental site. Mean values are shown as a
dashed line. Depth to the clay horizon for each rootstock is
indicated by an arrow.
- 0 U1
*H 0 N
0 @ 0
C% CLU ,4C
Li.) I. 7 a
0 .-- > O--o
p -1 Ir I-H dC
4 0 #
W r 44C
Co C-4 -- H-d)
10L~ x W HidBG
possible that some lenses of native phosphatic material were present at
the north end of the planting where the trees on rough lemon were located.
The extractable bases, K, Ca and Mg, showed a similar distribution
with depth (Figs. 8, 9 and 10 respectively). The mean surface values
were 320, 75 and 60 ppm for Ca, Mg and K, respectively. All 3 nutrients
showed a second peak value at a point near the sand-clay horizon inter-
The extractable Fe content of the soil (Fig. 11), unlike that of
the macronutrients, decreased gradually from 90 ppm at the surface to
approximately 25 ppm at 165 cm and then increased in the clay horizon.
Values for Zn, Mn and Cu varied so little between samples that only
means were plotted (Fig. 12). These nutrients appeared to be relatively
immobile, with high values at the surface. The concentration of each
nutrient was less than 2 ppm for every depth below 30 cm.
The solution containing the extracted nutrients was also analyzed
for Mo. If Mo was present, it was not detectable by the procedures used.
The level of exchangable Al, extracted with 1.0 N KC1, was very
small in the sandy soil (Fig. 13). Appreciable quantities of Al were not
detected until samples from the clay horizon were analyzed. The level
of Al rose sharply with depth from 2 to 28 ppm in the clay horizon. The
values shown in Fig. 13 are means for all rootstocks but do not include
those values for rough lemon samples for the depths 180 to 300 cm.
Thus, the sharp increase in Al content at 180 cm can be considered to
represent the change that occurs when the clay horizon is encountered.
m S "
.r -i 4
0 4 r-
r 4C 0
co c7 Z- 7- c
c\ cq c- c> ')
o V Hd
C'3 CO t7 '4) ON 1t I- a 5*t)Co
10L x 11) H.Id~G
C6ClJ 1 4 l
--.. C14 CD. cn C'a
10Lx 111) -Hid 31E1
/ c r o
Milliequivalent values for the nutrients presented are given in
Rootstock Water Use
Fluctuations in Soil Moisture Content
Two months during the year of study, August and December, were
selected to illustrate the general trend of soil water content with
depth and with the seasonal variation in rainfall. Table 8 shows the
general decrease in water content with depth except for a small peak
at 60 cm. The highest water content for the sandy soil was usually in
the surface 15 cm. The water content of this depth interval also showed
the greatest fluctuation.
Water content increased at the bottom of the profile due to the
increase in clay content of the soil. A noticeable increase often
occurred 30 to 40 cm above the clay horizon. It was apparent in this
case that the water from a preceding rain of 2.5 cm or more had caused
the water content to increase in the soil profile to the clay horizon
where further rapid drainage was impeded. Lighter rains of approximately
.50 to .75 cm or less, appeared to increase the water content only to
60 to 90 cm. Depth of penetration in every case was dependent on the
profile water content at the time of the rainfall.
December was a month of moderate rainfall, 12.19 cm (Table 9); how-
ever there was very little rain in the preceding 2 months. The lowest
soil water contents for the year were reached as a result, on December 5,
197S These values correspond to the low end of the range shown in Table 8.
The majority of the rain for the month fell after that date. Nevertheless,
Table 7. Mean soil level of selected nutrients in milliequivalents.
Depthz Nutrient level, meq/100 gY
cm P K Mg Ca Cu Fe Zn Mn Al Total
15 1.01 0.15 0.61 1.60 0.04x 0.52 0.05x 0.14X 0.01 4.13
45 0.21 0.11 0.22 0.35 0.39 0.01 1.32
75 0.13 0.09 0.13 0.31 0.34 0.03 1.06
105 0.13 0.07 0.10 0.23 0.26 0.03 0.85
135 0.13 0.07 0.09 0.21 0.20 0.03 0.76
165 0.12 0.07 0.11 0.20 0.14 0.03 0.70
195 0.11 0.13 0.42 0.34 0.18 0.17 1.28
225 0.16 0.11 0.28 0.34 0.20 0.16 1.28
255 0.15 0.10 0.21 0.29 0.18 0.20 1.16
285 0.12 0.11 0.24 0.33 0.16 0.22 1.21
315 0.11 0.10 0.25 0.33 0.18 0.26 1.26
345 0.11 0.08 0.22 0.31 0.15 0.31 1.21
ZEach depth value represents the center of a 30 cm interval.
YSee Table 1 for methods of analysis.
XAll remaining values are approximately 0.01 meq.
q o o e :i
0N 1 1 v- v E0
41 (D CD 00 -4 C'i C9
0 0 m LO 00 0 0 0
n O N 1N a O- N m 0
.a 0 0- ,-i t
4. to () D to ao o a a i-4 ') t- r(D .
0 .II ...
o ccI I I I I a
ft .-4 C) C) C C) C) 1/ .-4 a) 0 0 ,-4 S
0) cc n lm
u -o 1 a in o o C ) 0
a LU .. . . . a a)M NN 0
LU ~0 OO Cl il I) if a O a) -4 -4 -4 .
C I I I I I I I I I I I I I -
3 C' M C i- C') C) a) 0 a) a N tN i-l
- a 0 a) a iN) w C) O a) C' a)" '
Q) 40a 0
o. . . . . . . g
[3 ci 9 o o o t t c +
a4-4 Cf IN C) C' C) a L U
-4 v n Cv 0 t-
(D cc 0 C. . -4 0 0 w
0 In o o - - C)
< m . . .qn 1* o to -o 0- --
o m I o i i 0 a
l I I I I I I I I I I I I
. 0 m 0 M 0 C
o 4 0 0 U -- r C) m -
U +4 4 4-' 1 0
o I 1. 0 C a a .0 C >
' a) ou ta ) ) 0 ) <- C
o4 3 ) C a) a) a) a) a) N C') oC o a a N U
0l ) a) C a) a) a) a) C') O O I0 0 L
w t i oC a) a ) i La a) a ra N o a o Ia to rN a
U] I 2 I I I I I I I I I I I U 4.4
a a to ao c C) c) a 0 ') n t- .-' C) I. e
c) C' N a C ) n m C) a) ) a) a) a a 4-'
0 a.)0 C) ) C) C ) O a) a) f O N O I C a
a la .4 r- C' C' C] C C ) o C I]
Table 9. Precipitation or irrigation water additions to the soil
profile during the year of study.
Time Period Water Addition
May 12 June 2, 1973 11.20
June 2 June 30 11.48
June 30 July 14 (rain gauges inoperative)
July 14 August 4 10.74
August 4 September 8 20.35
September 8 October 3 11.10
October 3 November 3 3.00
November 3 December 5 5.33
December 5 January 12 12.19
January 12 February 8 0.58
February 8 March 8 5.08
March 8 April 26 4.22
these low initial soil water contents at the beginning of the month pre-
cluded any changes of more than 1 percent by volume from occurring at
depths below about 270 cm. Heavy rains in August, the wettest month
(Table 9), caused the soil water content to increase at all depths.
The months of August and December were also chosen because the
lowest moisture values of the year occurred in December and the highest
values in August. Applying these values to the moisture release curve
(Fig. 2) gives an indication of soil water availability assuming that
water becomes less available as the matric potential increases. Soil
moisture content had been depleted to its lowest point, less than 1
percent, in December by trees on rough lemon rootstock at depths of
150 and 180 cm.
Analysis of Rootstock Water Use
Soil water loss as influenced by rootstock and depth of rooting was
determined by selecting periods from the year of study in which the
confounding effects of rainfall and percolation were either absent or
minimal. This was accomplished by selecting approximately 16 periods
in which no rainfall had occurred but which had beem preceded by suffi-
cient precipitation so that the soil profile contained a reasonable
amount of available water. These 16 time periods were well distributed
over the year of study and the same periods were used for each root-
stock; however, there were 2 time periods in which the data for 1 root-
stock were not used. An increase in water content occurred at 1 depth,
apparently due to some peculiarity in the drainage of that profile,
which precluded using the data for that rootstock. The mean values for
these time periods are a measure of water use under a variety of condi-
tions due primarily to seasonal fluctuations in profile water content.
Results of the analysis described above are given in Table 10.
Water use for all rootstocks decreased consistently with depth with few
exceptions. Differences between rootstocks existed at nearly all depths.
Total water depletion from the profile ranged from 3.68 to 4.29 mm day-1
and rootstock differences were statistically significant. Approximately
50 percent of the total water losses occurred from the surface 90 cm of
The time period, October 10 to 20, 1973, one of the preselected
periods, was chosen for more detailed analysis. Water content was
measured 4 times during this period. Mean change in water content between
each measurement was used for statistical analysis. Data were analyzed
as a 4 x 12 factorial experiment. Moisture changes in the 0 to 15 cm
depth were not included because the 0 to 15 cm interval is subject to
considerable water loss by evaporation making it difficult to accurately
estimate the portion of water loss that resulted from transpiration.
Two sources of variation, rootstock and depth, were statistically
significant but their interaction was not (Table 11). Replicates, which
were represented as access tube locations, were included in the initial
analysis as blocks. They were not significant as a source of variation.
Mean change in water content was used for the analysis. The actual
change in percentage volumetric water content between measurements is
shown in Fig. 14. Each bar of the histogram for the time periods represents
a 30 cm depth interval beginning at 15 cm. The general trend was similar
for each rootstock. A large portion of the water loss took place at the
a a a u
N 0 0 0
.T Iv C
cM ci C9 NM
In 0 IO 0
Cq cq 03 t-
mI C0O 0( t
mn 0 0) 0
o C- I) C-
) 0 C-
w In O
0 0) 0 0
( 0 If I
CM CD Cto t
o0 00 C0 0
0 I7) C' 00
(0 In ) o
t n n r a
N 0 0 .
Un l C
) 'V CO 0
m') C1 U) 0)
UO (0 0)
00 0) 00 C-
U)o C V o0o C- -
0 0o 0 c V 0 .
S0 0 1 0 B
1 i) 02 2 3 O ;4 M
Table 11. Analysis of variance for mean change in water
content, October 10 to 20, 1973.
Source of Degrees of Meanz
Variation Freedom Square
Rootstock 3 217.24 *
Depth 11 1753.82**
Depth x Stock 33 112.16 n.s.
Error 528 84.19
ZSignificant at 5 percent level.
** 1 percent level.