TlIE INFLUEXCIE OF EARTHWORMS ON SOIL REACTION. 41.5

The Influence of E:irthworins on Soil Reaction and the Stratificalion of Undisturbed Soils. By E. J. SALISBURY, D.Sc., F.L.S. (Reader in Ecology, University of London, university College).

(With 3 Texbfigurea.)

[Read 16th November, 1933.1

TO appreci:ite the significance of Earthworm action on the acidity of undis- turbed soils, it is essential to rwognize the inarked gradient with respect to hydrogen-ion concentfiition which such soils exhibit. The :iuthor his ;ilre:idy shown that just as there is ii fairly rapid incronse in the organic contenk of tlie soil when we pass from the subsoil to the surface, so too there is u.su:illy a inarked rise in the real acidity in the sainct direction ( e j . Salisbury, E. J., " Stratific:ition and Hydrogen-ion Conccmtration of the Soil in relation to Leaching and Plant-succession, with special reference to Woodlands," Journal of Ecology, vol. ir. pp. 220-240, 1922 ; c$ also Discussion on Soil Problems, Trans. Faraday Society, 1921). The following data for various t y p of plant community illustrate this gradient of reaction and organic content in undisturbed soils :-

TABLE I. Organic Content and pH a t Varying Depths.

(Organic= Loss on ignition corrected for CO, erolved from cnrbonntes.) . .. _. _ _ ~~ . .

~ ~

I

Beech Wood. Clinlk 1)omn. Pine Wood I in inches. ! Oak Wood. i

Evidence has been adduced elsewhere to show that for a gircm ininc~ral snbstr;ituni and with ;I uniform vegetation there is, up to a point tl(~terinined by the origin of the organic inaterial, an increase of acidity :icconipnnying increase in the organic content of the soil (cf. S:ilisbury, lor. cit., and Salisbury, E. J., " The Soils of Blakeney Point : A Study in Edaphic Sue- cession," Ann. Bot. vol. xxxvi. pp. 391-432, 1912) . This, Iio\ve\-er, only holds where the conditions are approrimntely uniform, since the rate of dec:ty influenoes reaction by reason of the earlier stages being more acid than the later stages of decomposition.

LINN. J0UBN.-BOTANY. VOL. XLVI. 2 1

416 DH. E. J. SALISBURY ON THE

The action of the plough in arable land naturally tends to obscure, although it does not completely destroy, the soil gradient, and earthworms in il ininor degree perform for uncultivated soils the same f'uuction.

It is now over forty years since Darwin called attention to the importance of earthworms affecting a natural cultivation of tho soil (cf. C. Darwin, ' Vegetable Mould and Earthworms,' London, 1881). H e found from observations in four separate locations that the amount of soil brought to the surface as wormcasts ranged in weight from 7.56 tons to 18.12 tons per acrr per year. These results were based on continuous collection of the worm- casts from definite areas, and are open to the objection that removal may have stimulated a inore than norinal deposition a t the surface. There is nt all events some reason to believe that the wormcast on the surface serves as a protection to the orifice of the worm burrow. The values which Darwin obtained, however, represent an even layer of soil of from -09 in. to about 010 in. depth per acre per year, which is considerably less than the corre- sponding value deduced by Darwin from the depth of burial of objects after a period of years (0.19-0.83 in. per ann.).

During adverse seasons, namely the cold of winter and the heat of summer, earthworms have been known to descend to a depth of as much as eight feet, though they are not usually met with below three feet. The cultivating action thus mainly iifects the upper layers of theesoil, and in natural soils it would seem that worms feed mainly quite close to the surface. I n the great majority of instances where the writer has dug up worms in woodl:~nds, chalk downs, and heaths, they have been found feeding within the top few inches.

In view of wltat has already been said with regard to the reaction gradient in undisturbed soils, i t is ittiportant to know whence the soil is derived which passes through the woriii and is deposited at the surface. If the wormcasts show a different reaction to the underlying surface soil, the question at once arises as to whether the wormcasts represent soil from the more alkaline (or less acid) layers of the subsurf:ice or subsoil transported from below, or are they derived from the regions of maximum acidity near the surface?

The obvious test to employ for this purpose is the organic content which, 85 we have seen, diminishes with iiicre:lsing depth. Some of the organic inaterial is doubtless decomposed in the passage of the soil through the digestive tract ; but if it can be shown that tlie organic content of wormcasts is higher than that of the subsurEace, then any change of reaction as com- pared with that of the surface must clearly be attributed to direct action of the Earthworm, and not to niere transport of soil froin a stratum having a different reaction in the sense observed.

The appended data (Table 11. & fig. l), wliich represent losses en ignition, corrected for the carbon dioxide evolved from the carbonates, show clearly that in the eight locations investigated the wornicasts were derived from surface soil. The average values for the organic content of woriiicasts in

sis of the locations are Iiighcr, nud in three consitlcrably so, than tliosc of the surface soil (0-2 in.). I:! the two locatioiis where the reverse was the

FIG. 1 .

!z 2c

15

10

9 8 7

6 5

4 8 2 I

0

Organic content 05

h

S 1 U

caw the difference is not so marked. I t would appear, then, that eartliworms feed chiefly very near the surface, or, if in the subsurface, in regions of high organic content.

2 1 2

418 I)R. E. d . SALISBURY OX THE

TABLE 11. Average Organic Content (=Corrected Loss on 'Ignition) of Soils

and Worincnsts. . ~. ~. -

I 1 Location.

I------ 1 ~es tuca-~gras t i s . Society . . . . . . . .

Polytrichum-Agrostis. ll ........ Calluna- Triodin. 1 , . . . . . . . . Colney Heath. LOC. A . . . . . . . . . .

Lot. B .......... Hertford Heath. Loc. A ..........

19 7, LOC. n .......... Quercw Rohcr-Coiylus. Wood . , . .

11 1 9

I.

..... -. .

Organic Content of Soil (0-2 in.)

13.13 per cent.

1 7 6 B l l ,l

8.82 ,, ,,

~ - -

m50 1, ,l

12-17 )) ,, 9.61 11 I,

11.42 ,) ,, i .70 l l ,,

__

-. ._

Orgauic Content of Wormcasts.

Differencc w-s. --

+4.11 -?53 +490 i-6'08 +3%l +0*29 - 1 *70 +OGO

TABLE 111. Effect of Earthworms on the Hydrogen-Ion Concentration

of the Soil.

Locnlity.

Ilarpenden.

Kadlett. Harpenden. Rndlett. Colney Heath. Harpenden. Surrey. Colney Heath. IIertford Heath. TIolniwood Common. Ilarpenden. rtadlett. Alarkyate. Radlett. Harpenden. Blunden. Burton's Down, Surrey.

11

Vegetation.

Ebstzic*a-A!yrostis (LOC. A) . . . . . Crcllunn- Triodin ............. Quercuo Robur-Corylus . . . . . . . . Festuca-Agro8ti.q (LOC. B) . . . . . .

Nardetuui .................... Polytric.Lioitr-Ayrastis (Loc. B) . . Q. sessi[t~ora . . . . . . . . . . . . . . . . . Pasture ......................

Q. Robu9.- C01:ylus. . . . . . . . . . . . . .

,) ...................... ,, ......................

Piilyti~ichici1i-~4grostis (LOC. A) . . Q. Robtlr- Corylus. ............. Scrub ........................ Q. Robio.-Coiylus. . . . . . . . . . . . . . 3halk pasture . . . . . . . . . . . . . . . . Slluvial meadow . . . . . . . . . . . . . . 21IRlk pasture ................

Av. pH of

Soil.

-- 6.1 5.1 8.2

4.27 5.6 6.7 6.73 6.8 G . 0 G.2 G.2 6.3 6 5 6.9 6.9 7.24 i -45

7.45

Range of pH of Soils pH 5.1-pH 7.45. Range of pH of Woriucasts pH 6.4-pH 7.35. Range of difference - 0.12 to + 075.

Av. pH of

Worm- casts.

--- 5.7 5.4 5.6 5.84

6.35 6.4 5.93 6.5

G,3 6.4 6.43

_ _ _ I

~

+03 I + 0 4 I

Differ- I

ence w--s. '

+0.6 i

+0.57 +OT5 +0.7

INFLUENCE OF EARTHWOKBIS ON SOIL ItEACTION. 419

Hence any difference of reaction between surface soil and wormcasts iiiay safely be attributed to eartliworm :rction. Tlie data respecting the hydrogen- ion concentration given in T:il,le TIT. and fig. 2 show that in the majority of localities the woriiicasts exhibit a marked reduction in acidity (increased pH) as coinpared with tlie soils from which they are derived.

The detailed determinations need not be given in extenso, but two er:lmples, the one froin an acid location, the other froni an :illdine, will suffice to sliow the kind of' range to be ex1)ectetl in a given arex, the soil siimplcs in all cases being taken froin the actual levcl at wliicli the worm weru found feeding.

TABLE IV. Details of licaction in Two Locations.

A. Acid Soil with Fucstzccci rubrc6 dominaiit and Agroslis ectniitrc.

No. of KO. of p l l Values. Soil Samples. Wormcttst Samples.

6.1 .................... 1 5.2 .................... 6 5.3 .................... 1 5.4 .................... 0 6.5 .................... 1 6.6 .................... 1 3 5.7 .................... - 3 6.8 .................... 5.9 .................... - 2 6.0 .................... - 2 6 1 .................... - 2

1 6.2 -

- -

.................... Total Samples ...... 10 (Av. pII 5.27). 13 ( 9 T . pH 5.84).

B. Alkaline Alluvial Soil under mixed herbage. No. of xo. of

pH Values. Soil Samples. Wotnicnst Samples. 7.2 . . . . . . . . . . . . . . . . . . . - 1 7.3 .................... - 3 7 4 .................... 4 4 7.5 .................... 3 -

- __-________-- - Total Samples ...... 6 (AT. pII 7'45). 8 (Av.pH 7-33).

---__-__- ------- r i i h e first and most important fact which these data chb l i sh is tlie con-

siderable change in the reaction of acid soils effected by the e;irtliworms, amounting to 0.75 pH in one instance. The change in reaction will be more readily a1)preciatcd if the differences in tlie negative logarithniic n lues given in T;rl)le 111. are presented in the form of specific :iciclities ((;$ T:ible Jr.). lhese show that the effect is niorc inarlwl tlie more :wid the origind soil, and that whereas the acidity of thc soil is alniost inwriably decreased, the effect 011 alkaline soils may be to diiiiinish tlie alknlinity (Table I V . U QE fig. 2).

r i

420 DR. E. f . SALISBCWY ON THE

I

In the case of soil (1) the specific acidity was 79.4, so that in this case the earthmorin action has result,ed in a diminution of nearly 75 per cent.

FIG. 2. t

V. Differences in Specific Acidity betwccn Soil and Wormcasts.

(Neutral Wnter pH 7*03=1.) 1 . Ebtuca A y o s t i s . LOC. A.

3. Quercus Rohur-Coryhis. Wood. -38.0 , Difference in sp. Acidity . . . . -59.4 (s-w)

2, Callttna- Triodia. -- 396 4. E'eslncu-Ayrostis. Loc. u. -26.1

6. Nardetum. - 16.0

- 5.0 - 2.3 - 3.3

- 1.68

1 , ?, 7,

7. ,, I 1

1 , 1 ,9

9 ) 11 I 9

I , ,f 17

,? 1,

,! 9 ,

1 , 7, 19

t 9 , ,Y

6. Q. Robr-Cory1u.s. Wood. - 1 4 0

T. Pol~trichu~n,jicnil)eriiluni-Agrostis. LOC. 13. ,, - 7.4 8. Q. sessil$orn. 11 - 9.5 9. l'asture, Colney tlentli.

10. ,, Hertford Heath. 11. ,, Holmwood. ,t 99 7, 12. P~'o!t/trir~pcm-,4~rostis. Lor. A. - 3.0 1;s. Q. H o h ~ r - CO~.IJLUS. 14. Srriib, N r r k p t e . I ? 7 + 1.2 1 3 . Q. Rohur-Corykis. 1, 91 + 0 3 2 16. Ulinlk pasture. IT. Alliivial meadow. 1 , > l

18. Chalk pasture. 9 , 1 ) 9 ,

9 1, 7.

9 9 17

Ditlbknce iu sp. Alkalinity . . f

0 0 - 0.4 - 0 4

77

Range *00-60*4. IIlenn L)iference = 12.9.

INFLUENCE OF EARTHWORMS ON SOIL REACTION. 42 1

- . - - - . -

: I'er cent. 1 Carbouates , of Soil.

---- Festiica- Agrostia .................. 0.022-003

Nnrkyate Scrub .................. 0.20 nv. Pdytrichum-Agrostis .............. 002 HV.

Colney Heath. LOC. A . . . . . . . . . . . . Ot)5-0.08 ............ 1 0'04 av*

Qriercus Robur-Coiylus. Wood ... .I 002-010 ,, ,, LOC. €3

Hertford Heath .................. 002 av. ----__-- --

Nean vnlues .......... 0.060 'I,,

I n the course of these estimations no attempt was made to distinguish between the v:irioiir British species o f Earth worm, and it is not improbabls that these :ire capable of inoclifying the reaction in different degrees. Even, however, were. the same species involved, the cffect on different soils of the same original acidity might be expected to vary. For, though many soils are strongly buffered, the degree of the buffering is often very different in soils of the same original reaction.

-As to the nianner in wliicli earthworm effect the cliiwge, H:iy Lankester showed in 1864 (Q J. M. S . vol. iv. p. 258) that the acsophagus of the earth- worm bears three paired diverticula whose epi tlieliuiii secretes calcarcous p:irticles (c$ also Beddard, Caiiib. Nat. Hist. p. 359), :md Darwin long ago suggested that these might serve to neutralize tlie so-called Humic acids.

That this suggestion was justified, the above results would appear to show fully. However, the final proof was obtained by placing worms i n a soil of known reaction and determitiing that of the wormcasts formed. The soil w a s carefully mixed, but iit Ihe time the worrnc:ists were collected tbc soil reaction showed a small range, naiiiely froni pH 6.6 to pH 6.8 with 6.66 as the inean value (specific acidity= 2.25). IIie wormcasts exhibited a range from pH 6.9 to pH 7.1 with a nieaii v:tlue of pH 7 (specific acidity =l). The experimental conditioiis h e precluded the possibility of the surfnee soil being iniscd in the digestive tract of t,hc worm with subsoil substance. The writer has elsewhere shown ( loc. (-it. Aiinals of Botany, vol. sxxri.) that the latcr stages of decay are less acid than the earlicr oiies, and it is quite possible that during the process of digestion the chemical chaoges which the soil undergoes may lead to a decrease ill acidity, but analyses of soil and wormcasts with respect to the carbonate content havc shown that the latter normally contain an appresiably higher percentage thin the soils from which they are respectively derived ((9; Table VI.). This is shown graphically in

I 1

fig. 3. TABLE VI.

Carbonate Content of Soils and Wornicasts. _.

Per cent. ~

Carbonntes of I

Wormcnsts. ,

0.047-0*07 , 0.08 av. 026 nv. 012-028

----

I

0.09 av. 1 0.06-017 1

i 0.07 av.

0.124 O//,

122 DR. E. J. SALISBURY ON THE

Apart, then, from the cultivating action of worms and the attrition of the soil p:trticles due to th& very muscnlar gizzard, tlic change which they erect on the reaction tends towards a reversal of the natural reaction gradient in uncu1tiv:itetl soils. Consitlerable evidence has accuinulated i n recent years showing the importance of both the direct and indirect effects of renction in determining the distribution of plants, and a siinilar relation would appear to mist between reaction and some animals.

FIG. 3.

% 0.3

0 *2

0.1

005

W

S

(Aver "445)

Several years ago the writer drew attention t)o the poor soil Fauna of acid woodlnnds as compared with those in which the soil was neutral or alkaline,

INFLUENCE OF EAJ(THW0RDIS ON SOIL REACTION. 423

a feature especially marked with respect to the Molluscan fauna. Atkins * has recently 1)ublished data empliaaizing the importance of this factor in the distribution of snails.

I t is tlien clearly essential to know lrow far earthworms a re affected by the reaction of the soil. S. H. Hunvitz performed experiments \\itli Allo- ~ohOpho7*U,fLctii&Z (“ The lieaction ot‘ E;irthworins to Acids,” Proc. Amer. Acad. vol. xlvi. pp. 67-81, 1910), in wliicli the tips of suspended worms were dipl’ed into dilute solutions & of Hytlrocliloric, Nitric, Sulphuric, and Acetic acids. The time which elapsed, hefore contraction resulted in the withdrawal of the worm from the solution, was determined for a number ot’ individuals, and i t was foiintl that the reaction time diminished with the increasing dis- sociation, thus indicating a con~it1erat)le sensitiveness to the hydrogen-ion concentration. Similar expcrinients performed by A. 1’. Soh1 (“ h i c t i o n s of Eartliworins to Hytlroxylions,” Jorirn. Amer. Physiology, vol. xxxiv. pp. 384- 401, 1914) on the saine species showed IL correspondingly diminislling ‘‘ reaction time ” with increasing concentration of Hyctrosyl ions. ‘l’hese cxperiinents taken together appear to wtirrant the suggestion t h d an approxi- mation to equality in the concentration of Hydrogen and Hydroxyl ions is the optimum condition for these aninrals. Arrhenius has performed experiinents with both Periclmta iiidica and Lzoiibriciis tewest r is , in which soils were r e i k h ~ l artificially acid or :tlk:iline. Both species were subjected to a range from p H 3 to pH 10, :~nd at the ond of a few days live individuals remained only in the soils wliich were very slightly acid or neutral (pH 6- pH 7).

The conditions in all these experiments, however, were clenrly very artificial, so that observations were made by the writer on natural areas of varying reaction, tlie nuniber of wornis being determined both by actual counts of the eartliworins in a cuhit foot of soil, and by enumerating the freshly-formed worincastr on a square yard. The average values and tlic observed range on each of the areas examined is shown in Table VII. It will be noted that broadly there is a dimiuution in nuinbcra 011 either side of the neutral region (with respect to the diminution a t pH 7.0 this is prolxtbly accidentnl ; q j : below).

TABLE VII. Enrthworip Frequency in Natural Soils of varying pH.

Real acidity ($1) . . . . . , . . . ,

I Obserred range. . . . . . . . . . . . . . Av. per acre , . . . ~. . . . . . . .

~. .~ . -~ ~

* Atlii l la, \ir, E. G. , 6 31, 1’. Lebour, in Scieut. i’roc. 1:oJ. Dublin SOC. nvii. (1923) pp. 233-2410.

424 DR. E. J. SALISBURY ON THE

This relation to reaction probably accounts for the rarity of earthworms in sonie siliceous areas. According to Darwin, for instance, earthworms are practically absent from the Welsh hills, where the soils most coinnionly exhibit an acid reaction. Similarly, earthworms are apparently infrequent in the very alkaline peat of true fens, though they may be present in considerable numbers in fen peat from which the alkaline salts have been partially leached, or in fen peat of naturally mild reaction.

But acidity or alkalinity are by no means the only factors which govern the distribution of earthworms, and amongst other conditions the proportion of orgnnic iit:iteriaI :iod the water content of the soil would seem most important. Darwin regarded Henson’s figure (Zeitsch. fur Wiss. Zoo\. Bd. xxviii. 1877) of nearly forty thousand worms per acre ae very exceptional, but much h i g h freqnencias have been observed. Thus in a locality with nearly neutral reaction, moderately high water content, and a high organic content the writer found a frequency of nearly 700,000 earthworms per acre. Sonie little-known observations of Dr. Brett may be quoted in this connection (Trans. Herts. Nat. Hist. SOC. 1883). This observer counted the worms in an area of one square yard of soil, dug out to a depth of three feet, in four dicerent locations i n a garden. The figures obtained range froin sisty to one liundred and eighty earthworms per cubic yard. The latter nunher repi esents itboiit 8 i 0 , O O O earthworms per acre, and i t is pertinent to note that the loaition was a vine border where :in approximately neutral reactioii and a high orgiinic content are esscntid to successful production.

As already indicated, an increase of organic content is frequently accom- panied by an increased acidily :ind is usually correlated with ii high water content. Despite, however, the suitability of the habitat in two of these respects, we tind the I-eaction factor apparently dominating the frequency of the earthworms. Acidity (or alkalinity) woulcl then appear to be a “master factor ” in the clidribution of these aninials. It is significant in this con- nection that a very marked increase in the number of earthworms was observed to follow the application of lime to a garden soil well supplied with organic niatcrial but distinctly acid in reaction.

The writer has suggested that all natural soils tend to become more and more acid with increasing age, and this edaphic succession is accoinpanied by changes in the character of the vegetation (cf. Journal of Ecology, vol. ir . 1 9 2 ) . The fact that earthworins do not occur where the subsurface is appreciably acid shows that their effect on this edaphic succession is to cause a retardiitioii in the establishment of the reaction gradient, but it does not appear probahlo that the increasing acidity of the surface can be permanently ohecked by their influence. This conclusion seems to be indicated by the much more :icitl character of wormcasts on acid soils than on those which are less acid or neutral. Further work may, however, show that a given species produces worincasts of it inucli narrower range of reaction.

INFLUENCE OF EAHTH WORMS 0s SOIL IlEACTION. 425

But, even if the effect be but a retnrclation and not a complete check to the ed:iphic succession, tire iniportancc of earthworni action in relation to vegetation niay obviously be consitlerable.

To suinmnrize, then, the innin conclusions :- (u) Worinc:ists, as conipnrcd with the soil from which they are derived, are

usually less acid (or less :ilkali1icj, the reduction amounting in sonic cases to as iiiucli as 75 per cent.

(b) Worniatsts coinriiorily cont:iin a higher proportion of carbonates than the surfacca soil.

((a) The organic content oE worinc:tsts is very high, indicating that the etfects produced operate on th:rt 1:iyer of tlre soil which is norinally inost acid.

( t l ) The frequency of e:trth~voriiis in natural soils :ippears to be gre:itest in those w11ic:li are apl)roxiii::itely neutral in reaction, : i d to cliniinisli iis

tho :icidity or :ilkalinity incrcases. Other farourable factors are high water content and :L high organic content.

(e) E:irthworriis may thus have a innrked effect on vegetation and retard the " edapliic succcssion.~'

l h e conclusions (a) to ( e ) :ire unaffected by the fact that the different species OE eartliworins re not distinguisl~ed. I t is not improbable that the frequency in rel:itian to the f d o r s under (d ) may vary with the species, and that tlieir effects (e ) ni:iy be of varying niagnitude, though the data obtained \\~ould seem to inclicatc that such it difference, if it exists, is one of degree rather than of kind.

I 1