Tree Roots: Facts and Fallacies - Arborcare Tree Solutionsarborcaresolutions.com.au/treerootfacts.pdf · Tree Roots: Facts and Fallacies Thomas O. Perry A proper understanding of

Tree Roots: Facts and Fallacies

Thomas O. Perry

A proper understanding of the structure and function of roots can helppeople become better gardeners.

Plant roots can grow anywhere-in the soil,on the surface of the soil, in the water, andeven in the air. Except for the first formedroots that respond positively to gravity, mostroots do not grow toward anything or in anyparticular direction. Root growth is essentiallyopportunistic in its timing and its orientation.It takes place whenever and wherever theenvironment provides the water, oxygen,minerals, support, and warmth necessary forgrowth.Human activities, such as construction,

excavation, and gardening, often result in seri-ous damage to trees. In some cases, trees canbe inadvertently injured by people who aretrying to protect them. Indeed, people can killtrees in hundreds of ways, usually because ofmisconceptions about root-soil relationships,or because of a disregard of the basic functionsthat roots perform.

In order to maintain the health of cultivatedtrees and shrubs, it is necessary to understandthe morphology and physiology of tree rootsin relation to the aerial portions of the plant.For those who are responsible for maintain-ing the health of woody plants, this articleexamines some widely held misconceptionsabout roots. It describes the typical patternsof root growth as well as their locations anddimensions underground. It also describes therelationship of healthy roots to typical forestsoils as well as the behavior of roots adaptedto atypical circumstances-growing through

deep sands, under pavements, down crevices,inside shopping malls, and in sewer lines.

The Relationship Between Roots andOther Parts of the Plant

The growth of a plant is an integrated phe-nomenon that depends on a proper balanceand functioning of all parts. If a large portionof the root system is destroyed, a correspond-ing portion of the leaves and branches will die.Contrariwise, if a tree is repeatedly defoliated,some of its roots will die back. Proper func-tioning of roots is as essential to the processesof photosynthesis as are the leaves and otherchlorophyll-bearing parts of the plant. Typicalroots are the sites of production of essentialnitrogenous compounds that are transportedup through the woody tissues of the plant,along with water and mineral nutrients.The fine feeder roots of a tree are connected

to the leaves by an elaborate plumbing systemconsisting of larger transport roots, trunk,branches, and twigs. Many researchers haveweighed and estimated the proportions of var-ious plant parts. Weighing and counting everyroot tip and every leaf is a heroic if not impos-sible task, and careful sampling is essentialto making accurate estimates. Sampling errorsand variation among species produce variableresults, but the biological engineering require-ments of plants are apparently similar, andthe relative proportions of both mature herbsand mature trees are of the same order of mag-

4

nitude: 5 percent fine or feeder roots, 15 per-cent larger or transport roots, 60 percent trunkor main stem, 15 percent branches and twigs,and 5 percent leaves (Bray, 1963; White et al.,1971; Meyer and Gottsche, 1971).A tree possesses thousands of leaves and

hundreds of kilometers of roots with hundredsof thousands of root tips. The numbers,lengths, and surface areas of roots per tree andper hectare are huge. Plant scientists try tomake the numbers comprehensible by talkingabout square units of leaf surface per unit ofland surface-the "leaf area index." If bothsides of the leaf are included, the leaf areaindex of a typical forest or typical crop isabout 12 during the height of the growing sea-son (Moller, 1945; Watson, 1947; and manymodern texts on crop physiology).The number of square units of root surface

per unit of land surface, the "root area index,"can be calculated from studies that report thenumber of grams of roots present in a verti-cal column of soil. Such data are determined,first, by taking core samples or digging outsuccessive layers of soil and screening andsorting the roots and, second, by determiningtheir average lengths and diameters as well astheir oven-dry weights. The quantity of rootsdecreases rapidly with increasing depth innormal soils, so that 99 percent of the rootsare usually included in the top meter (3 ft) ofsoil (Coile, 1937). A reasonable approximationfor non-woody tissues is that the oven-dryweight is one-tenth of the fresh weight andthat the density of fresh roots is very close toone. If one makes these assumptions for Lel-bank’s data (1974) for winter wheat (Tiiticumaestivum) and for Braekke and Kozlowski’sdata (1977) for red pine (Pinus resinosa) andpaper birch (Betula papyrifera), the calcula-tions indicate a root area index between 15and 28. E. W Russell’s data (1973) are of thesame magnitude, clearly indicating that thesurface of the root system concealed in thesoil can be greater than the surface of theleaves! Amazingly, this conclusion does nottake into account the fact that nearly all treeroots are associated with symbiotic fungi

(mycorrhizae), which functionally amplify theeffective absorptive surface of the finer rootsa hundred times or more.The pattern of conduction between the

roots and leaves of a tree varies between andwithin species. Injection of dyes and observa-tion of their movement indicate that, in oaksand other ring-porous species with largediameter xylem vessels, a given root is directlyconnected to a particular set of branches,usually on the same side of the tree as the root(Zimmerman and Brown, 1971; Kozlowski andWinget, 1963). Death or damage to the rootsof trees with such restricted, one-sided plumb-ing systems usually results in the death of thecorresponding branches. Other tree speciespossess different anatomies in which dyesascend in zigzag or spiral patterns, indicatingthat the roots of the tree serve all of thebranches and leaves (Figure 1). Death or injuryto the roots of such trees does not lead to aone-sided death in the crown of the tree. Theanatomy of trees can vary within species, andthe patterns of connection between the rootsof most species are unknown. Sometimes thepattern can be detected by examining the pat-tern of bark fissures, which usually reflectsa corresponding pattern in the woody tissuesconcealed beneath the bark. Knowledge of thepattern of conduction between roots andleaves is of practical importance in predictingthe results of treating trees with fertilizers,insecticides, and herbicides, or in predictingthe results of one-sided injuries to trees dur-ing construction.

Patterns of Growth and Development inTypical SoilsEarly observations of tree roots were limitedto examining the taproot and the larger rootsclose to the trunk of the tree or to examiningthe vertical distribution of severed roots

exposed by digging trenches and pits (Busgenand Munsch, 1929; Coile, 1952; Garin, 1942;Bohm, 1979). Attempts to examine the depthand extent of the larger roots of an entire treewere not really possible until bulldozers, back-hoes, front-end loaders, and fire pumps

5

became available (Stout, 1956; Berndt andGibbons, 1958; and Kostler et al., 1968). Unfor-tunately, most tree roots are less than one mil-limeter in diameter and are destroyed by therough action of such heavy equipment.Examination of the small non-woody roots

of trees and their relationship to the largerroots requires years of study, infinite patience,and the gentle use of heavy equipment. WalterLyford and his colleagues at the Harvard

Figure 1. Five types of water-conducting systems mvarious comfers as shown by the tracheidal channelsdyed by trunk infection. The numbers give the heightin centimeters of the transverse section aboveinjection. A Spiral ascent, turning right Abies, Picea,Larix and Pinus (Rehder’s section 3, Taeda). B. Spiralascent, turning left Pinus (Rehder’s section 2,Cembraj. C. Interlocked ascent: Sequoia, Libocedrusand Jumperus. D. Sectonal, winding ascent: Tsugaand Pseudotsuga. E. Sectonal, straight ascent. Thujaand Chamaecypans. Oaks and many ring-porousspecies have a pattern similar to E. From Rudinskiand Vite, 1959. Reprinted courtesy of the BoyceThompson Institute for Plant Research.

Forest in Petersham, Massachusetts, wereamong the first to combine tweezers andpatience with bulldozers and haste to developa comprehensive picture of the normal pat-terns of root development for trees growing innatural situations. The following descriptionof the growth of tree roots is a synthesis ofLyford’s published descriptions, the author’spersonal observations, and recent books onthe subject (Kostler et al., 1968, Bohm, 1979;Torrey and Clarkson, 1975; R. S. Russell, 1977;E. W. Russell, 1973).

Tree roots vary in size from large woodyroots 30 centimeters (12 in) or more indiameter to fine, non-woody roots less than0.2 millimeters (0.008 in) in diameter. Thevariation in size from large to small, and thevariation in categories from woody to non-woody, perennial to ephemeral, and absorbingto non-absorbing, is continuous. This continu-ous variation makes the sorting of roots intovarious categories arbitrary. Nonetheless, clas-sification and sorting are essential to compre-hending the pattern and integrated functionof the total root system.The first root, the radicle, to emerge from

the germinating seed of some species, such aspines, oaks, and walnuts, sometimes persistsand grows straight down into the soil to

depths of 1 to 2 meters (3 to 6 ft) or more, untilsupplies of oxygen become limiting. If this"taproot" persists, it is usually largest justbeneath the tree trunk and decreases rapidlyin diameter as secondary roots branch fromit and grow radially and horizontally throughthe soil. The primary root of other species,such as spruces, willows, and poplars, doesnot usually persist. Instead, a system offibrous roots dominates early growth anddevelopment.Between four and eleven major woody roots

originate from the "root collar" of most treesand grow horizontally through the soil. Theirpoints of attachment to the tree trunk areusually at or near ground level and are

associated with a marked swelling of the treetrunk (Figure 2). These major roots branch anddecrease in diameter over a distance of one to

6 1

Figure 2. Plan-view diagram of the horizontal woody root system developed from a single lateral root of a redmaple about 60 years old. Sohd circles show the location of other trees m the stand. Arrows indicate that theroot tips were not found; therefore these roots continued somewhat farther than is shown. From Lyford andWilson, 1964.

four meters (3 to 15 ft) from the trunk to forman extensive network of long, rope-like roots10 to 25 millimeters (.25 to 1 in) in diameter.The major roots and their primary branches

are woody and perennial, usually with annualgrowth rings, and constitute the framework ofa tree’s root system. The general direction ofthe framework system of roots is radial andhorizontal. In typical clay-loam soils, theseroots are usually located less than 20 to 30centimeters (8 to 12 in) below the surface andgrow outward far beyond the branch tips of thetree. This system of framework roots, oftencalled "transport" roots, frequently extends toencompass a roughly circular area four toseven times the area delineated by an imagi-nary downward projection of the branch tips(the so-called drip line).

It is not uncommon to find trees with root

systems having an area with a diameter one,two, or more times the height of the tree(Stout, 1956; Lyford and Wilson, 1964). In driersoils, pines and some other species can form"striker roots" at intervals along the frame-work system. These striker roots grow down-ward vertically until they encounter obstaclesor layers of soil with insufficient oxygen.Striker roots and taproots often branch to forma second, deeper layer of roots that growhorizontally just above the soil layers whereoxygen supplies are insufficient to supportgrowth (Figures 3 and 4).The zone of transition between sufficient

and insufficient oxygen supply is usuallyassociated with changes in the oxidation-reduction state and color of the iron in the soil

7

Figure 3. Drawing, not to scale, of framework system of longleaf pme tree grown in well-dramed soil with asecond layer of roots running in the soil layers where oxygen supphes become limiting.

Figure 4. Photograph of framework roots of longleaf pme including striker roots, 90 percent of the surface rootsystem has rotted and washed away, Kerr Lake, North Carolina.

8

Figure 5. Mat of roots above the permanent water table exposed by digging a drainage canal, Green Swamp,North Carolina. A few species have specialized tissues contammg air passages and specialized metabolismsthat permit their roots to penetrate several feet below the permanent water table where httle or no oxygen isavailable. Iron oxide deposits are typically associated with such roots

(from reddish-yellow to gray for example).Water can hold less than 1/10,000 the oxygenthat air can hold, and limited supplies of oxy-gen are usually associated with wet soils.Drainage ditches in swamps reveal an impres-sive concentration of matted roots just abovethe permanent water table (Figure 5).

Feeder Roots

A complex system of smaller roots grows out-ward and predominantly upward from the sys-tem of framework roots. These smaller rootsbranch four or more times to form fans ormats of thousands of fine, short, non-woodytips (see Figures 6, 7, 8, and 9). Many of thesesmaller roots and their multiple tips are 0.2to 1 millimeter or less in diameter and less

than 1 to 2 millimeters long. These fine, non-woody roots constitute the major fraction ofthe surface of a tree’s root system. Their mul-

tiple tips are the primary sites of absorptionof water and minerals. Hence they are oftencalled feeder roots.Root hairs may or may not be formed on the

root tips of trees. They are often shriveled andnon-functional. Symbiotic fungi are normallyassociated with the fine roots of forest trees,and their hyphae grow outward into the soilto expand greatly the effective surface area ofthe root system (Figure 10).The surface layers of soil frequently dry out

and are subject to extremes of temperatureand frost heaving. The delicate, non-woodyroot system is killed frequently by these fluc-

9

tuations in the soil environment. Nematodes,springtails, and other members of the soilmicrofauna are constantly nibbling away atthese succulent, non-woody tree roots (Lyford,1975). Injury to and death of roots are frequentand are caused by many agents. New rootsform rapidly after injuries, so the populationand concentration of roots in the soil are asdynamic as the population of leaves in the airabove, if not more so.The crowns of trees in the forest are frayed

away as branches rub against one another inthe wind. One can easily observe the frayedperimeter of each tree crown by gazingskyward through the canopy of a mature

forest. Such "shyness" is not seen below theground. Roots normally extend far beyond thebranch tips, and the framework root systemsof various trees cross one another in a com-

plex pattern. The non-woody root systems ofdifferent trees often intermingle with oneanother so that the roots of four to sevendifferent trees can occupy the same squaremeter of soil surface (Figure 9). Injuries, rocks,or other obstacles can induce roots to deviate90 degrees or more from their normal patternof radial growth. These turnings and intermin-glings of roots make the determination ofwhich roots belong to which tree extremelydifficult. Furthermore, natural root grafts

Figure 6. Schematic diagram showing reoccupation of soil area near the base of a mature tree by the growthof adventitious roots. 1) Root fans, growing from the younger portions of the woody roots, have extended to adistance of several meters from the tree 2) Root fans on adventitious roots have only recently emerged fromthe zone of rapid taper or root collar and now occupy the area near the base of the tree. 3) Vertical roots. FromLyford and Wilson, 1964.

10

Figure 7. Schematic diagram showing woody and non-woody root relationships. 1) Stem. 2) Adventitious rootsin the zone of rapid taper. 3) Lateral root. 4) Non-woody root fans growing from opposite sides of the rope-likewoody root. 5) Tip of woody root and emergmg first order non-woody roots. 6) Second and higher order non-woody roots growing from the first order non-woody root. 7) Umnfected tip of second order non-woody rootwith root hairs. 8) Third order non-woody root with single bead-shaped mycorrhizae. 9) Fourth order non-woodyroot with smgle and necklace-beaded mycorrhizae. The horizontal bar beneath each root section represents adistance of about 1 centimeter. From Lyford and Wilson, 1964.

commonly occur when many trees of thesame species grow together in the same stand.

In summary, large woody tree roots growhorizontally through the soil and are peren-nial. They are predominantly located in thetop 30 centimeters (12 in) of soil and do notnormally extend to depths greater than 1 to2 meters (3 to 7 ft). They often extend outwardfrom the trunk of the tree to occupy an irregu-larly shaped area four to seven times largerthan the projected crown area. Typically, thefine, non-woody tree roots grow upward intothe litter and into the top few millimeters of

the soil, are multiple-branched, and may ormay not be ephemeral.

Why Roots Grow Where They DoRoots grow where the resources of life areavailable. They do not grow toward anything.Generally they cannot grow where there is nooxygen or where the soil is compacted andhard to penetrate. In most soils, the numberof soil pores, and the consequent availabilityof oxygen, decreases exponentially with depthbelow the surface, the amount of clay, and theresistance to penetration (hardness).

11

Figure 8. Scale diagrams of horizontal, woody, third order lateral roots of red oak, Quercus rubra. Emphasis ison the roots that return to the surface and elaborate into many small-diameter non-woody roots m the forestfloor. Tbp view (above), side view (below). The squares are 1 meter on a side. From Lyford, 1980.

Frost action and alternate swelling andshrinking of soils between wet and dry con-ditions tend to heave and break up the soil’ssurface layers. Organic matter from the

decomposing leaf litter acts as an energy sup-ply for nature’s plowmen-the millions ofinsects, worms, nematodes, and other crea-tures that tunnel about in the surface layers.The combined effect of climate and tunnel-ing by animals is to fluff the surface layers ofan undisturbed forest soil so that more than50 percent of its volume is pore space. Air,water, minerals, and roots can penetrate thisfluffy surface layer with ease. The decompos-ing leaf litter also binds positively chargedcations (e.g., Ca++, K +, Mg++) and func-

tions to trap plant nutrients and prevent theirleaching into the deeper layers of soil. Soilanalyses show that the greatest supplies ofmaterials essential to plant life are located inthe very surface layers of the soil, and, predic-tably, this is where most of the roots arelocated (Woods, 1957; Hoyle, 1965).

Variations in Soil Conditions

Roots are most abundant and trees grow bestin light, clay-loam soils about 80 centimetersdeep (3 ft) (Coile, 1937, 1952). Conversely, rootgrowth and tree growth are restricted in shal-low or wet soils, or in soils that are excessivelydrained. Roots can and do grow to greatdepths-10 meters (33 ft) or more-when oxy-

12 ’:

Figure 9. Photograph of roots mtermmghng m the soil. Mixed hardwood stand, Harvard Forest, Petersham,Massachusetts. The roots m front of the trowel were exposed by careful brushmg and pulling away of the litter.The roots m the background were exposed by digging down and destroying the fme surface roots in the process.The roots have been sprayed with whitewash to make them stand out. Photo by T. O. Perry.

gen, water, and nutrients are available at these

depths. Tree roots can grow down severalmeters in deep, coarse, well-drained sands.However, in these cases, overall plant growthis slow, and trees tend to be replaced by shrubson topographies and soils that are drainedexcessively.Adapting to their situation, pines and other

trees tend to develop a two-layered root sys-tem in the deep sands of the Southeast andother similar sandy locations. They form a

surface layer of roots that absorbs water andnutrients made available by the intermittentsummer rains, and a deep, second layer ofroots that allows survival under drought con-ditions.Some soils of the western United States are

geologically young and unstructured, originat-ing primarily from the downward movementof eroded particles of rock. Such deposits canform a layer 10 meters (33 ft) or more deep andare extremely dry, especially on the western

13

Figure 10. Photograph of root tips growmg m the litter of a mixed hardwood forest. The mycorrhizae extendout from the root tips to expand greatly the functional absorptive surface area of the roots they are attachedto. Root diameters about 0.5 mm. Photo by Ted Shear, North Carolma State University.

slope of the Sierras where summer rains arelight and infrequent. Most water in the soilsof this region originates from winter rains andsnowmelt that travel along the surface of theunbroken bedrock that lies below the soillayer. Seedling mortality in such climates isextremely high, and years with sufficientmoisture to permit initial survival are infre-quent. Growth takes place predominantly inthe early spring, and those trees that manageto survive and grow in the area are character-ized by a taproot system that plunges downand runs along the soil-rock interface. Deepcuts for superhighways sometimes revealthese roots 15 meters (50 ft) or more below thesurface.Some trees, like longleaf pine (Pinus palus-

tiis), have made special adaptations to insure

survival and growth on sands and other deepsoils. During the initial stage of establish-ment, the tops of longleaf pine seedlingsremain sessile and grass-like for four or moreyears while the root system expands and estab-lishes a reliable supply of water. Only thendoes the tree come out of the "grass stage" andinitiate height growth.

Spruces, willows, and other species growcharacteristically on wet sites where oxygensupplies are very limited. Their root systemstend to be shallow and multi-branched.Tupelo, cypress, and other species of theswamps and flood plains have evolved special-ized anatomies that permit conduction of oxy-gen 30 centimeters (12 in) or more below thesurface of the water and special metabolismsthat eliminate alcohols, aldehydes, and other

14

toxic substances produced when fermentationreplaces normal respiratory metabolism.

Many such flood-plain species can survive theconditions of low soil oxygen that result fromseveral months of flooding (Hook et al., 1972).Other species, particularly cherries and

other members of the rose family, are espe-cially sensitive to conditions where oxygensupplies limit growth. Cherry roots containcyanophoric glucosides, which are hydrolizedto form toxic cyanide gas when oxygen sup-plies are limited (Rowe and Catlin, 1971).Flooding that lasted less than 24 hours killedmost of the Japanese cherry trees aroundHains Point in Washington, D.C., followingHurricane Agnes in 1973. Sediment buildup,which in some locations exceeded 20 cen-

timeters (8 in), also contributed to this mor-tality.There are important genetic differences in

the capacity of tree species and varieties totolerate variations in soil chemistry, soil struc-ture, or oxygen supply (Perry, 1978). The dis-tribution of trees in the landscape is notrandom. There is no such thing as a "shallow-rooted" or a "deep-rooted" species of tree. Onthe one hand, the roots of flood-plain speciessuch as cypress, tupelo, maple, and willow,which are generally thought of as "shallow,"will grow deep into the soil and down sewerlines if oxygen and water supplies are ade-quate. On the other hand, the roots of pines,hickories, and other upland species, which aregenerally thought of as "deep," will stay close

Figure 11. Roots growing in the crevices between bncks. There was no oxygen below the bncks that overlaida compacted clay soil on the North Carolina State Umversity campus. Tree roots commonly follow cracks, crevices,and other air passages underneath pavement. Photo by T. O. Perry.

15

to the surface if the soil is too compact, or ifoxygen supplies below the surface are limited.

Roots grow parallel to the surface of the soilso that trees on slopes have sloping root sys-tems that actually grow uphill. In search ofnutrients, roots often grow along cracks,crevices, and through air spaces for unbeliev-able distances under the most impermeablepavements and inpenetrable soils (Figure 11).Roots commonly grow down the cracksbetween fracture columns ("peds") in heavyclay soils they could not otherwise penetrate.

Tbmperatures and Tree RootsThe roots of trees from temperate climates,unlike their shoots, have not developedextreme cold tolerance. Whereas the tops ofmany trees can survive winter temperaturesas low as -40 to -50 degrees C (-40 to -60 F),their roots are killed by temperatures lowerthan -4 to -7 degrees C (20 to 25 F) (Beattie,1986). In areas that experience severe cold,such as northern Europe or Minnesota, a goodsnow cover or a layer of mulch can often pre-vent the ground from freezing completely dur-ing the winter (Hart, Leonard, and Pierce,1962). By repeatedly digging up, measuring,and then reburying them, researchers haveobserved that roots can grow throughout thewinter-whenever soil temperatures are above5 degrees C (40 F) (Hammerle, 1901; Crider,1928; Ladefoged, 1939).One of the subtle impacts of raking leaves

in the fall is that it exposes roots to destruc-tive winter air temperatures that they wouldordinarily be insulated from by the layer ofleaves. Similarly, the potted trees so commonin the central business districts of northerncities seldom survive more than a few yearsbecause their roots are exposed to air temper-atures that are substantially lower than thoseof the soil. Skilled horticulturists are carefulto move potted perennials to sheltered loca-tions where they will be insulated from thefull blast of winter.

Contrariwise, soil surface temperatures insummer are often hot enough to "fry an egg,"as newspapers boastingly report. Such temper-

atures, which can be as high as 77 degrees C(170 F), also fry plant roots. Fortunately, mostsoil temperatures decrease rapidly with depth,and roots only a few millimeters below thesurface generally survive, particularly if an

msulating layer of mulch is present. As in thecase of freezing temperatures, plants growingin containers are more susceptible to heatdamage because of the lack of insulation.Roots, like shoots, grow most rapidly whentemperatures are moderate-between 20 and30 degrees C (68 and 85 F) (Russell, 1977).

Misconceptions about Tree Roots and thePractical ConsequencesThe rope-like roots at or near the surface ofthe soil have been obvious to diggers of holesfor fence posts and ditches for thousands of

years, as obvious as Galileo’s "shadow of theearth on the moon." However, trees can

become huge-larger than the largestwhale-and very few human beings have hadthe privilege of actually seeing even a smallfraction of the root system of an entire tree.Illustrations in textbooks, in natural historybooks, and in manuals of landscape architec-ture or of tree care are usually the creationsof artistic imaginations and highly inaccurate(Figure 12).An insurance company, hearing of Walter

Lyford’s work on tree roots, wanted to developan idealized picture of tree roots, penetratingthe depths of the soil and securely anchoringthe tree in an upright position, as the symbolof the security its customers would achieveby purchasing its insurance. The companycommissioned an artist to visit Lyford andexamine his findings in order to prepare a logoof tree roots for its advertising campaigns. Theprojected logo and advertising scheme werenever started because it is impossible to por-tray an entire tree with its roots accurately onthe page of a typical textbook.As an example, take a healthy, open-grown

oak tree, 40 years old, with a trunk 21 meters(70 ft) tall and 0.6 meters (2 ft) in diameter.The spread of the branches of such an open-grown tree is rarely less than two-thirds of the

16

Figure 12. Roots do not grow as this artist’s conceptionmdicates. Inaccurate illustrations hke this one haveled to harm ful practices m the management of treesin both forest and urban situations. Illustration froma brochure produced by the Society of AmencanForesters.

height of the tree and is often equal to orgreater than the height. The root system ofsuch a tree usually extends more than 9meters (30 ft) beyond the tips of the branches,generally forming a circle with a diameter twoor more times the height of the tree. Theproblems of scale are overwhelming and canbe appreciated by examining Figures 13 and14.A significant portion of the root system of

all trees in all soils is concentrated in the topfew centimeters of soil. Tree roots grow rightinto the litter layer of the forest, in among thegrass roots of suburban lawns, and in thecrevices of the bricks, concrete, and asphaltof the urban landscape (Figures 11 and 15). For

this reason, fertilizer broadcast on the surfaceof the soil is immediately available to treeroots. It does not have to move "down" intothe soil. Even the reportedly immobile phos-phates are readily available to tree roots. Care-ful research has failed to show any differencesin the response of trees to fertilizer placed inholes versus that broadcast on the soil surface

(Himelick et al., 1965; van de Werken, 1981).Foresters broadcast fertilizers on millions of

acres of land and achieve rapid and largereturns on their investments. Except for whereslow-release fertilizers are used for specialeffects, there is no justification for "tree

spikes" or other formulations of fertilizer inholes bored in the ground or for fertilizerinjected into the soil. The root systems of one-year-old seedlings can take up nutrients tenor more feet from their trunks. The absorb-ing roots of larger trees commonly extendfrom their trunks to twenty feet beyond theirbranch tips. The tree will benefit from hav-ing fertilizer broadcast over this entire area.Herbicides and other chemicals should be

used only with extreme care near trees andshrubs since their roots extend far beyond thetips of the tree’s branches. When they grow ina lawn, trees can be thought of as "broad-leaved weeds" and application of the commonlawn herbicide dicamba (also called "Banvel®")by itself, in combination with other herbi-cides, or in combination with fertilizers caninjure trees. This chemical or its formula-tions, when improperly applied, can distortand discolor leaves and even defoliate and killtrees. Several tree and lawn-care companiesare selling these chemicals mixed with fer-tilizer at home garden centers or are applyingthe chemical on a contract basis. Improper useof dicamba will distort the leaves of oaks andsycamores and defoliate and kill more sensi-tive trees like yellow poplar."Roundup®" (glyphosate) herbicide and its

formulations are supposedly inactivated whenthey hit the soil or dirty water, but they donot have to actually penetrate the soil to inter-act with tree roots growing in a litter layer,lawn, or mulch. Dogwoods and other trees canshow extreme leaf distortion and crown die-

17

back even when herbicides do not strike thegreen portions of their trunks or their foliage.Since tree roots often grow in cracks andcrevices of pavement, applications ofsterilants and herbicides to kill weeds in thesesituations can inadvertently kill trees 20meters (60 ft) or more away from where theyare applied (Figure 15).Remember, natural root grafts are common

among trees of the same species, meaningthat herbicides applied to kill one tree can

flash back along root grafts to kill trees thatwere not treated. In addition, many trees, suchas poplars, sweet gum, and American beech,send up sprouts from their roots that can be

damaged when an herbicide is translocatedfrom a treated stem through the root systemto an untreated stem.

In larger residential lots, say roughly 32meters wide by 45 meters deep (105 ft by 150ft), the roots of a large tree will commonlyoccupy the entire front or back yard and

Figure 13. Scale drawmg of Memonal Oak Tree (Quercus alba), Schenck Forest, North Carolma State University.The onginal drawing was made by tracmg the projected image of the tree (Figure 14) onto a piece of paper witha pen that produced a line 0.2 millimeters thick, the thmnest line that can be reproduced in most publications.The ongmal drawmg was 24 centimeters wide (9.5 m) and represents a typical root spread of 65 meters (212ft). The Schenck Oak is about 33 meters tall (106 ft) and is represented on the vertical axis as 12 centimeters(4.7 m). The ongmal drawmg represented a 274-fold reduction in the actual height of the tree. Most branchesand 90 percent of the tree roots would not be visible if drawn to this scale. Indeed the width of a typical whiteoak leaf would be about the thickness of the lmes m the drawmg, and most of the roots would be located inthe soil layer represented by the thickness of the lme representing the soil surface. The dash-dot lme is located1.5 meters (5 ft) below the surface and very few if anyroots would penetrate beyond this depth m a representativesoil.

18

Figure 14. This photograph of the Schenck MemorialOak (Quercus alba) was projected and traced toproduce Figure 13. The Schenck Memorial Oak is 32.3meters tall (106 ft) and has a crown spread of 29meters (94 ft) and a diameter at breast height of 1.07meters (42 in).

trespass into the neighboring property. Nopart of an urban yard can be treated carelesslywith herbicides. Care must also be taken in

disposing of toxic chemicals, deicing salts, oldcrankcase oil, and high-strength detergents.Careless disposal of chemicals and improperuse of herbicides are among the most com-mon causes of tree death in urban areas.

Soil CompactionThe largest single killer of trees is soil

compaction-compaction from excessive useof city parks by people, from excessive graz-ing by livestock (including zoo animals)-andeven from the feeding activities of pigeons,whose small feet exert more pressure persquare centimeter than heavy machines. Trees

are also killed by compaction from construc-tion equipment and by compaction from carsin unpaved parking areas. Compaction closesthe pore spaces that are essential to the

absorption of water and oxygen and hardensall but the sandiest of soils so that roots can-not penetrate them, even when oxygen sup-plies are adequate (Patterson, 1965).

Excessive use of mulch can induce fermen-

tation, immobilize nutrients, and cut off theoxygen supply, thereby killing trees. Use ofbroad expanses of plastic, either as a surfacecovering or under a layer of organic mulch orstone, is a sure way to cut off oxygen and killtrees. As an alternative, porous landscapefabrics, which permit water and air to pene-trate the soil, are a vast improvement overplastic.The maximum leaf area index that a nor-

mal ecosystem can support is about 12, whenboth surfaces of the leaf are counted. The cor-

responding maximum root area index isbetween 15 and 30. A large planting of lawn,annuals, or shrubs underneath existing treesoften results in a reduction in the root and leafarea indexes of the trees. Gardening undertrees-planting lawns, daffodils, liriope, orazaleas and rhododendrons-tears up tree

roots and will produce a corresponding deathof twigs and branches in the crown of the tree.Surprisingly, turning over the soil when

gardening is another common cause of treedeath in urban situations. Gardeners shouldbe aware of the biological compromises thatneed to be made in order to achieve the properbalance between trees and garden plants.

It should be obvious by now that any earthmoving or regrading that cuts or buries treeroots will result in the death of a correspond-ing portion of the branches in the tree. Unfor-tunately, this simple fact is often ignoredwhen utility lines, parking lots, or even irri-gation lines are being installed. Smearing sixinches of clay from the mineral soil layer overthe root system of an established tree or cover-ing its roots with pavement can be as lethalas cutting it down with a chain saw.When a new house is constructed, the yard

may have six different trench lines cut from

19

Figure 15. Many roots of trees grow closely intermingled with grass roots in the few top centimeters of a lawn.Therefore fertilizers and herbicides do not have to move down into the soil in order to affect trees.

the street to the house-for water, sewer, elec-tricity, telephone, gas, and cable television.Over 90 percent of the pre-existing tree rootsin the front yard are destroyed during con-struction and utility-line installation. In addi-tion, the soil structure of the entire lot isusually completely destroyed by heavy equip-ment and the spreading of excavated heavysoil on top of undisturbed soil. The proud newhomeowners are left to figure out for them-selves why all their trees have severe crowndieback and continue to decline (or die) for adecade or more after they have moved in.

Saving TreesPeople often try to save trees under impossi-ble circumstances. The root systems of a largetree often occupy the entire building site, andit is impossible to complete constructionwithout damaging some or all of its roots. By

tunneling or concentrating utility-line instal-lations in a single trench, this damage can beminimized. Careful watering and thinning ofthe tree crowns to compensate for root lossescan buy time until new roots can be produced.

It is often wiser and cheaper to accept a badsituation and cut down a tree before construc-tion begins rather than to try to preserve alarge specimen in the middle of a construc-tion site. Performing tree surgery after con-struction is complete-and crown dieback isobvious-will be more expensive and may betoo late to save the tree. Planting a young,vigorous sapling after construction is com-pleted not only may be more cost effective butalso may provide greater long-term satis-faction.

In urban situations, soil compaction andlimited oxygen supplies are the major res-traints to growing trees in city parks and in

20

highly paved areas. Inadequate supplies ofwater are usually secondary to these two fun-damental problems. In terms of survivingthese conditions, trees adapted to swamps andflood-prone areas, where soil oxygen tensionsare normally low, often perform the best.Indeed, most of our common street trees,including pin oak, willow oak, sycamore, sil-ver maple, and honey locust are flood-plainspecies that can thrive in compacted, urbansoils. Different trees grow on different sites innature, and it is unreasonable to expect spe-cies adapted to well-drained upland or slopingtopography to possess roots that would growwell in the compacted soils of a heavily usedrecreation area or in areas with extensive

pavement.There are hundreds of ways to kill or injure

trees. They range from zapping them withlaser beams (as in the Omni shopping mall ofAtlanta) to girdling them with the grindingtugs of dogs chained outside of college class-rooms. Many tree deaths are accidental andinvolve misconceptions about the structureand function of tree roots. Why else would theCity of New Orleans keep a rhinoceros cagedon the root system of its symbolic CentennialOak? Why else would the State of North Caro-lina use a ditch-witch in late June to installan irrigation system among the stately treesof the old Capitol building? Why else wouldthe National Capital Parks in Washington,D.C., allow rows of newly planted, eight-inch-caliper trees in front of the new AerospaceCenter to remain unwatered while the needfor irrigation was recognized and supplied totrees on the mall across the street?

People must know where tree roots arelocated and what they require if healthy treesare to become a gratifying part of the urbanenvironment.

Literature Cited

Baskerville, G L. 1965. Dry matter production m imma-ture balsam fir stands. Forest Science Monograph No 9

Baskerville, G L 1966. Dry matter production in imma-ture balsam fir stands roots, lesser vegetation, and totalstand. Forest Science 12(1): 49-53.

Beatie, D. J. 1986. Principles, practices and comparativecost of overwintering container grown landscape plants.Pennsylvania State University Agr Exp Sta , Southern

Cooperative Series Bull. No 313.

Berndt, H. W, and R D. Gibbons 1958 Root distributionof some native trees and understory plants growing onthree sites withm ponderosa pme watersheds m Colorado.Rocky Mountain Forest and Range Exp Sta Paper No 37.

Braekke, F. H., and T T Kozlowski 1977 Distributionand growth of roots in Pinus resmosa and Betula

papyrifera stands Norsk Institutt for Skogforskmng 33(10). 442-451.

Bohm, W 1979. Methods of Studying Root Systems. Ber-lin- Spnnger-Verlag.

Bray, J R. 1963. Root production and the estimation ofnet productivity. Canadian Tournal of Botany 41 65-72

Busgen, M , and Munsch, E. 1929. Structure and Life ofForest Trees Translated by T. Thomson. New York- Wiley.

Coile, T. S 1937. Distribution of forest tree roots in NorthCarolina Piedmont soils. journal of Forestry 36 247-257.

Coile, T S 1952. Soil and the growth of forests. Advancesin Agronomy 4: 329-398.

Cnder, F. J. 1928 Winter root growth in plants Science68- 403-404.

Duvigneaud, P, and S Denayer-DeSmet. 1970 Biologi-cal cycling of minerals m temperate deciduous forests,pp. 199-225, m Analysis of Temperate Forest Eco-systems,D. E. Reschle, ed. Berlin: Spnnger-Verlag

Garm, G. I. 1942. Distribution of roots of certain tree spe-cies in two Connecticut soils. Connecticut Agr Exp Sta.Bull. No. 454

Hammerle, J 1901 Uber die penodizitat des wurzelwach-stums bei Acer pseudoplatanus L. Beitrage zur Wiss.Bot 4: 105-155

Hart, G., R E. Leonard, and R. S Pierce. 1962. Leaf fall,humus depth, and soil frost m a northern hardwood forest.Forest Research, note 131.

Himelick, E. B., D. Neely, and W R. Crowley, Jr 1965.

Experimental field studies on shade tree fertilization.Illmois Natural History Survey, Biological Notes No. 53.

Hook, D. D., C. L. Brown, and R. H. Wetmore. 1972. Aer-ation in trees. Botanical Gazette 133 (4): 443-454.

Hoyle, M C 1965. Growth of yellow birch in a podzolsoil. Northeast Forest Exp. Sta Research Paper NE-38.

21

Kostler, J N., E. Bruckner, and H. Bibelnenthen. 1968Die Wuitzem der Waldbaume. Hamburg’ Paul Parey.

Kozlowski, T. T., and C. H. Wmget 1963 Patterns of watermovement m forest trees Botamcal Gazette 124- 301-311

Ladefoged, K. 1939 Untersuchungen uber die penodizitatim ausbruch und langenwuchstum der wurzeln Det

Foisthge Forsogsvaesen i Danmark No 16

Lyford, W. H., and B. F. Wilson 1964. Development of theroot system of Acer rubrum L. Harvard Forest Papers No10.

Lyford, W H. 1975 Rhizography of non-woody roots oftrees m the forest floor, pp. 179-196, in The Developmentand Function of Roots J. G. Torrey and D. T. Clarkson,eds. New York: Academic Press.

Lyford, W. H. 1980. Development of the root system ofnorthern red oak (Quercus rubra L.) Harvard Forest PapersNo. 21

Meyer, F H , and D. Gottsche. 1971 Distribution of roottips and tender roots of beech, pp. 48-52, in IntegratedExpenmental Ecology Berlin Spnnger-Verlag

Moller, C M 1945. Untersuchungen uber laubmenge,stoffverlust und stoffproduktion des waldes. Sonderadruckder Mittelilungen von Det Forstlige. Forsogsvaesen i Dan-mark Kandrup and Wunsch.

Ovmgton, J D. 1957. Dry matter production by Pmus syl-vestns L. Annals of Botany, N S 21 (82)- 277-314

Patterson, J C 1965. Soil compaction and its effects uponurban vegetation, in Better Trees for Metropolitan Land-scapes Forest Service Gen Tech. Rep. NE-22.

Perry, T. O. 1972 Dormancy of trees in winter. Science121: 29-36.

Perry, T. 0 1978 Physiology and genetics of root-soil mter-actions on adverse sites, pp 77-97, in Proceedmgs of the5th North Amencan Forest Biology Workshop, C. A.Hollis and A. E. Squillace, eds. Miami School of ForestResources and Conservation, University of Florida.

Pritchett, W. L , and W H. Lyford. 1977. Slash pine rootsystems. Soil and Crop Science Society of Florida37:126-131.

Rowe, R B., and P B Catlin 1971. Differential sensitiv-

ity to waterlogging and cyanogenesis by peach, apricot andplum roots. Horticultural Science 96(3). 305-308.

Rudmski, J A, and J P Vite 1959 Certain ecological andphylogenetic aspects of the pattern of water conductionin conifers Forest Science 5(3). 259-266

Russell, E W 1973 Soil Conditions and Plant Growth,10th ed London: Longman.

Russell, R. S 1977 Plant Root Systems Their Functionsand Interaction with the Soil New York McGraw-Hill.

Stout, B. A. 1956. Studies of the root systems of decidu-ous trees. Black Rock Forest Bull No 15

Torrey, J G , and D. T. Clarkson, eds. 1975. The Develop-ment and Function of Roots Third Cabot Symposium.New York Academic Press

Van de Werken, H 1981 Fertilization and other factors

enhancing the growth rate of young shade trees. Journalof Arboriculture 7(2) 33-37

Watson, D. J 1947 Comparative physiological studies onthe growth of field crops 1 Variation m net assimilationrate and leaf area between species and varieties, andwithin and between years Annals of Botany, N S. 11:41-76

White, E H , W L Pritchett, and W. K Robertson 1971.Slash pme biomass and nutrient concentrations, in ForestBiomass Studies, H. E. Young, ed. Symposium of Inter-national Union of Forest Research Organizations, No. 132.Orono University of Maine.

Woods, F. 1957 Factors limiting root penetration in deepsands of the southeastern coastal plain Ecology 38:357-359.

Zimmerman, M. H., and C L Brown 1971. Trees Struc-ture and Function. Berlin- Spnnger-Verlag.

Thomas O. Perry taught in the School of ForestResources at North Carolma State University for manyyears, and now operates his own consulting business,Natural Systems Associates An earlier version of thisarticle appeared in the Journal of Arboriculture 8 (8):197-211, 1982.