THE EFFECT OF BLACK LOCUST ON ASSOCIATED SPECIES WITH SPECIAL REFERENCETO FOREST TREES'1*

byARTHUR GLENN CHAPMAN,

Department of Botany, The Ohio State University.

Introduction

For a number of centuries, it has been known amongagriculturists that such crop plants as beans, peas, andclover lead to increased yields of certain other crop plantsfollowing in the rotation; but it remained for moderninvestigators to demonstrate that nodule formation onroots of leguminous species is due to infection by nitro-gen-fixing bacteria and that through the activity of thesebacteria the nitrogen content of the soil is increased. In1888 Beijerinck isolated and cultured the organism andnamed it Bacillus radicicola; Prazmowski, in 1889, chang-ed the name to Bacterium radicicola; and later the So-ciety of American Bacteriologists placed the Bacteriumradicicola in the genus, Rhizobium, that commonly in useat present. In 1932, Fred, Baldwin,(and McCoy listed allleguminous species, whose nodule bacteria have beenstudied, in sixteen groups on the basis of interinocula-tion; thus, each species within a group may successfullybe inoculated with bacteria from the nodules of anyother species within the category. The bacteria, inoculat-ing roots of species in any one group, are deemed dis-tinct enough from those inoculating other groups to beconsidered as a separate species. Black locust (Robinapieudoacacia L.) alone comprises "Group XII.

Although foresters have reported, during the lastdecade, their recognition of the importance of this treelegume as a benefactor to associated tree species, theirdata represent studies on camparatively few plantationswith associated species, primarily catalpa. In 1922, Fer-guson, from studies of adjacent black locust and catalpaplantings at College Farm, State College, Pennsylvania,showed small but consistent decreases in total nitrogencontent of the soil at increasing distances from thelocust planting. Also average heights and diameters ofthe catalpa trees were shown to decrease as distance fromthe locust increased. McIntyre and Jeffries, in 1932,reported recent studies of soil nitrogen in relation toheight and diameter growth on two catalpa plantingsadjacent to black locust at the Pennsylvania State Col-lege, State College, Pennsylvania. Their results in generalwere similar to those of Ferguson.

Black locust may become established under widelyvarying combinations of site factors. Cuno, 1930, sug-gests that the range of black locust may have been orig-inally restricted to the Appalachian Mountains fromPennsylvania to Georgia, and to parts of western centralArkansas and eastern central Oklahoma. He further statesthat the best development of the species occurs on thewestern slopes of the Appalachians in West Virginia.This tree has been introduced into practically every statefor one or another of its many uses. In the westernstates, eminent success with plantations has been attainedin the valleys of the northern Roeky Mountain region,particularly, in Idaho, eastern Oregon and Washington.Black locust was introduced into Europe in 1601 and hasbeen considered the most successful of the tree species

(1) An abstract of the original paper which will appear in theJanuary number of Ecological Monographs, 1985.

introduced from America. Although establishment oflocust is often found on badly eroded, rocky slopes, thehigh rates of growth usually occur only in plantings ornatural stands on well drained silty loams.

The writer found from observations throughout thestates of Ohio and Indiana, during the summers of 1931and 1932, that the black locust-catalpa combination isthe usual one from which comparative measurementsmay be obtained. Suitable plantations were also found,however, where data on the relation of locust to thegrowth of white ash, tulip poplar, black and chestnutoaks were collected.

The purpose of the present investigation was toobtain more critical data to demonstrate to what extentblack locust affects the soil nitrogen content and thegrowth rate of associated species. Studies have beenconfined' to plantings of species adjacent to locust in thestates of Ohio and Indiana and recently established plotsin the Botanic Garden at the Ohio State University.

Plantations

Six plantings have been included in the major partof this investigation, one in Highland County, Ohio; onein Clermont County, Ohio; and four in Clark County,Indiana. Several other plantings in various counties ofOhio have been supplemented for less detailed study. Allof these have been established twenty years or more.

METHODS:— Rectangular plots were laid out inrepresentative parts of the plantations adjoining theblack locust. In every case, the trees were spaced atregular intervals in the rows paralleling those of thelocust, and the widths of the plots were sufficient toinclude from six to twenty trees in each row. The meanheight and diameter of the trees were calculated for eachrow.

Total nitrogen determinations were made by theboric acid modification of the Kjeldahl method on soilsamples collected at regular intervals of distance fromthe locust in the plots. Also hydrogen-ion concentrationswere obtained1 from the several soil samples by means ofa quinhydrone potentiometer.

Comparisons were made from increment boringsof the growth rates for the first ten-year and the lastten-year periods for trees near to and at a distance fromthe locust in six Ohio catalpa plantings.

During the summer of 1931, observations were madeon relative degrees of nodulation on grazed and ungrazedlocust plantings and also in plantings on various soiltypes.

RESULTS AND DISCUSSION:— The greater por-tion of the results from the study of the plantationshas been represented in fourteen figures and two tables,The following table (Table 1) gives the ranges of averageheights and diameters for rows adjacent to and thosefarthest from the locust in the seven plots, data takenfrom Figures 1, 2, 4, 6, 7, 8, and 9. Very noticeable

Page 39

Table 1. — Ranges of average heights and diameters forrows adjacent to and those .farthest from the locust.

Range of Avg. Kange of Avp.Plot No. Species height—ft. diam.—in.

1 Catalpa 40-32 6.0-4.42 Catalpa 48-27 7.5-5.53 Catalpa 37-17 6.9-3.24 White ash 49-20 6.4-2.65 Tulip poplar, black oak 53-34 7.5-3.26 White ash 53-20 4.7-2.17 Tulip poplar, black oak,

chestnut oak 54-23 5.2-2.5

decreases in heights and diameters of the various speciesin the plots were evident with increasing distance fromthe locust. Recognizing that situations upon which in-vestigations of this nature may be made are infrequent,the writer was careful to select for study those in whichdifferences in topography, soil types, degree of pasturing,and age of locust and adjacent species, played no signi-ficant role.

In Table 2 are represented the results from totalnitrogen analyses on the series of soil samples collectedat different intervals from the locust in each plot.The results clearly indicate the rapid decrease in the

Table 2. — Percents of total nitrogen in air dry soil sam-ples taken in plots at different distances away

from locust plantings.

Plot No. Distances in feet away from locust12 24 DC 4S 60

1.2.3.

4.5.6.7.

0.1960.1900.216

00.2000.2080.2050.209

0.1850.1740.194

200.1640.1720.1990.173

0.1670.1500.177

400.1520.1560.1840.163

0.1400.1380.150

600.1340.1410.1630.136

0.1250.1340.139

so0.1280.1330.1500.132

0.1100.1160.127

1000.1250.1180.1370.119

0.0900.0960.112

percent of total nitrogen content of soil, with increas-ing distance from the locust. A range from 0.090 percentto 0.195 percent total nitrogen may appear to be small;but, quantitatively, it is equivalent to a range from 1800pounds to 3900 pounds of nitrogen per acre when basedupon two million pounds v:eight of the top six acre-inches of soil.

While the response of the subordinate vegetation tothe presence of black locust was of minor importance inthe study, orchard grass (Dactylis glomerata L.), undercatalpa in plot 1 and under adjacent locust, illustratedone of many similar contrasts in growth rates of groundcover species observed by the writer. The grass becamesparser under the catalpa as distance from the locustincreased. The difference in light intensity under thetwo species, due to the thinner foliage of the locust thanthat of the catalpa, is not thought to be a limiting factorhere for the growth of orchard grass since the grassdevelopment was only slightly better beneath large open-ings in the catalpa canopy than elsewhere beneath thesame species.

Mean increments for four catalpa plots in Ohioshowed that the radial increase for the first ten yearsfor the trees near the locust was greater than that forthe same period for those trees at a distance. Thisrelationship held also for the last ten-year period forthe same trees.

The degree of nodulation on the roots of black

locust and hence the amount of atmospheric nitrogenfixed vary with certain physical conditions in the soil.Examination of many roots in numerous plantings inOhio and Indiana showed that minimum nodule forma-tion occurred in those areas which were heavily grazedor water-logged for long periods in the year. In thegrazed plantings, the few nodules present were smalland occurred at the base of trees near the soil surface.Both the old and recently developed tubercles were small.In the water-logged plantings, the few nodules presentwere mainly in the upper one or two inches of soil butwere distributed much more widely than in the pasturedareas. A characteristic common to both the heavily grazedand water-logged soils was poor aeration and, therefore,an insufficient amount of free oxygen for the aerobicnitrogen-fixing bacteria. Clearly, this was the limitingfactor in nodulation in these plantings.

In the ungrazed plantings, nodulation was at itsmaximum in well drained, porous, silt loams over whichhad accumulated a thick duff and leaf litter. The positionof most of the nodules in the soil profile was betweenthe mineral soil and the duff or in the upper inch of themineral soil. In such an environment, the moisture supplywas adequate for nodule development even during thedry season of 1931. Where accumulation of duff andlitter was scant and periodic drying of the surface soiloccurred, the greatest abundance of tubercles was foundsomewhat deeper, but in no place were many nodulesfound below a depth of six inches.

There is little discrimination to be made on therelative amounts of nodulation by black locust in soiltypes whose parent materials are of different geologicalorigin, provided such factors as soil moisture, soil aera-tion, and organic matter content of soil are approximatelyequivalent. No definite range of pH tolerance can becited for the locust bacteria, but it is known to be widerthan for the bacteria of some of the field crop legumes.Since Albrecht in 1932 showed that increased nodulationon the roots of soy beans accompanied increased avail-able calcium in the "growth media" but did not accom-pany or follow increased pH, it seems that a correlationbetween a pH range and the amount of nodulation forblack locust would be of questionable significance. Oneof the best examples of a high degree of nodulation wasfound by the writer in an immature glacial sandstone andshale soil with a pH of 4.4. There is no indication thatthe acidity range found in numerous areas has set anylimitations on nodule formation.

Garden Plots

As preliminary work in the field had clearly demon-strated that black locust is effective in increasing thegrowth rate of certain associated species, the garden plotstudy was begun to determine how soon the effects wouldbe perceptible.

METHODS:— Soil, underlain by limestone glacialoutwash gravel, was excavated to a depth of twelveinches along a terrace contour for eleven plots four feetsquare and two feet apart. Frames constructed fromboards one inch thick by fourteen inches wide werefitted into the excavations so that about two inches pro-jected above the general ground level. The excavatedsoil material was returned to the frames by passingthrough a sieve with one-quarter inch meshes. Care wasobserved to give all the plots uniform horizontal surfaces.

Page 40

The seedlings of black locust, grown in sterilized soilmedia and free from nodules, and Chinese elm (Ulmuspumila L.) were used for planting, the locust plantsbeing inoculated at the time of setting. The stock wasplaced six inches apart in rows at the same intervals onJune 1, 1931. Only elm was planted in plots 2, 4, 6, 8,and 11, while locust was alternated with the elm in therows of the remaining six plots.

Well distributed soil samples were collected fromthe surface four inches in each plot on June 23, July13, August 9, September 2, September 22, October 12,and November 10 in the second season for total nitrogendeterminations.

Also, samples of elm leaves were gathered fromeach plot on the dates of soil collection. Each samplewas composed of all ages of leaves from the entire elmpopulation of the individual plots. The last collectionwas made when the leaves were turning brown andabscissing. Total nitrogen analyses were made on ovendry samples.

RESULTS AND DISCUSSION:— Table 3 containsdata taken from Figures 15, 16, 17, and 18. The condi-tions under which the plants of the various plots de-veloped were deemed sufficiently similar to justifyplotting an average of the amounts of nitrogen for eachsampling date in order to present more clearly the generaltrend of the curves. The following table presents suchresults:Table 3. Averages of total nitrogen percents for soiland elm leaves for all plots in each of the two groups,pure elm and elm-locust plots.

Samples

June23

Dates of sampling

July13

Aug. Sept.9 2

Sept.22

Oct.12

Nov.10

Soil fromelm plots 0.165 0.167 0.140 0.132 0.143 0.150 0.145Soil from elm-locust plots 0.166 0.167 0.147 0.139 0.173 0.167 0.172Leaves fromelm plots 0.38 2.40 2.28 2.27 2.03 2.06 2.69Leaves from elm-locust plots 2.25 2.34 2.42 2.47 2.47 2.46 1.71

It is evident from inspection of the table that ingeneral there was a decrease in the total nitrogen in thesoil of the elm plots until a minimum was reached aboutSeptember 2. Although there is an increase in nitrogencontent in the plots after the low points, it is evident thaton the dates of sampling subsequent to July 13 the nitro-gen present did not equal that of June 23.

The minimum nitrogen content of soil from the elm-locust plots in general was reached about September2. During the time from September 2 to September 22the nitrogen increment for these plots was much greaterthan that for the elm plots. On November 10, all plotsexcept number 10 were found to have a nitrogen contentequivalent to or greater than that on June 23.

There is a fairly consistent reduction of total nitro-gen in the leaves from the elm plots throughout theseries. The average plot data for the elm-locust plotsshow: first, a gradual percent increase of leaf nitrogenuntil September 2, in contrast to that of the elm plots;second, a constant nitrogen content until October 12;and, third, a sudden drop in total nitrogen on November10. The sharp decrease in leaf nitrogen for nearly all of

the plots on November 10 is noteworthy. The percentagesof nitrogen are based on yellow, abscissing leaves. Al-ready existing soluble nitrogen compounds in the leaves,and the soluble forms resulting from processes of de-composition during the yellowing and hence the browningperiod, apparently move into the stems. However, someof the nitrogen loss to the leaves may be due to leachingby rain.

Summary

1. An investigation has been made of the effects ofblack locust on soil nitrogen and on the growth rates ofassociated species in forest plantings in Ohio and Indiana.Studies of total nitrogen changes were also conductedduring the 1932 growing season on the soil material andleaves of elm seedlings of eleven garden plots containingelm in pure stand and in mixture with black locust.

2. In plantings of catalpa; white ash; tulip poplarand black oak; mixed tulip poplar, black oak, and chest-r.ut oak adjacent to black locust, where such site factorsas differences in topography, soil types, grazing, and ageof locust and adjacent species were irrelevant, appreciabledecreases in the heights and diameters of trees with in-creasing distance from the locust were found.

3. Analyses of soil samples, collected at regularintervals from the locust in each of six Ohio andIndiana plantations, indicated reductions in total nitrogenas distance increased. These reductions in each plot show-ed definite correlations with reductions in mean heightsand diameters for adjacent trees.

4. Definite decreases in the number and in thedegree of development of ground cover species wereobserved as distance from the locust increased.

5. Increment borings have shown that the differ-ences between the radial increases for the first and, lastten-year periods for catalpa adjoining black locust wasless than that for the same periods for catalpa at dis-tances from the locust; and that the difference betweenthe radial increases for the first ten-year period foradjacent catalpa and catalpa at distances from the locustwas decidedly less than the difference between radialincreases for the last ten-year period, for the same trees.

6. Nodulation was at a maximum in the upper fourinches of soil where a moderately constant moisture con-tent, a high percentage of organic matter, and goodaeration existed.

7. Although poor nodulation was found in manyplantings and natural reproduction stands of black locust,evidence from field and garden plot studies indicates thatnot absence of the locust bacteria but poor aeration ormoisture conditions or both were the primary causes.

8. No correlation between soil reaction and degreeof nodulation was evident from studies in locust planta-tions growing on different soil types.

9. Significant increases in amounts of total nitrogenwere detected during the second growing season in thesoil material and leaves of elm and of the elm-locustgarden plots.

10. An appreciable decrease in the total nitrogencontent of the soil material of the elm plots was foundat the end of the 1932 growing season.

11. During the summer of 1932, progressive de-creases occurred in the total nitrogen content of theleaves of elm.

Page 41