sign in sign up
<<
Home , Andropogon gerardi, Biomass (ecology), Productivity (ecology), Sclerophyll

Plant and of , , Oakwood, and Field in Central Minnesota Author(s): J. D. Ovington, Dale Haitkamp, Donald B. Lawrence Reviewed work(s): Source: , Vol. 44, No. 1 (Jan., 1963), pp. 52-63 Published by: Ecological Society of America Stable URL: http://www.jstor.org/stable/1933180 . Accessed: 01/02/2012 13:44

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Ecology.

http://www.jstor.org 52 J. D. OVINGTON AND OTHERS Ecology,Vol. 44, No. 1 slowlythrough the moreopen savanna . Botting,G. 1956. "Black Sunday" and its effectson The dense exotic scrub communities,however, Adelaide Hills orchards. J. Agric. South 57: 256-61. constitutethe greatestfire hazard. Cochrane,G. Ross, S. M. Burnard,and J. M. Philpott. Cover stratificationdiagrams show the general 1962. Land use and -firesin the Mount Lofty lack of a dense, completeground cover of low Ranges, South Australia. Australian Geographer8: vegetationunder savanna woodlandsand also in- 143-160. dicate the much denser tree canopy and well de- Glaessner,M. F., and L. W. Parkin (eds.). 1958. The geologyof South Australia. MelbourneUniv. Press, finedshrub understory of sclerophyllforests. Melbourne,Australia. 163 p. The "stringybark"tree stratumof sclerophyll Marshall,Ann. 1958. Climate,p. 76-83. In R. J. Best forestsis defoliatedbut not greatlyaffected by (ed.), IntroducingSouth Australia. MelbourneUniv. fireand recoversrapidly by growthfrom adventi- Press, Melbourne,Australia. tious lateral shoots. In contradistinction,the in- Prescott,J. A., and J. A. Thomas. 1948-49. The length of the growingseason in Australia as determinedby digenous shrub stratumof these is de- the effectivenessof the rainfall. Proc. Roy. Geogr. stroyedby fireand undergoesa well-definedthough Soc. Australasia (South Australia) 50: 42-46. rapid seral developmentfrom fire-razed condition Specht,R. L., and R. A. Perry. 1948. ecologyof to climax vegetationin 7-10 years. Five definite the Mount Lofty Ranges. Trans. Roy. Soc. South stages can be recognizedin the sclerophyllforest Australia 73: 91-132. pyric sere, limited,however, to the understory Sprigg,R. C. 1942. The geology of the Eden-Moana Fault Block. Trans. Roy. Soc. South Australia. 66: only. Exotic is generally killed by 185-214. bushfires,but exotic scrub regrowth,particularly . 1946. Reconnaissancegeological surveyof por- of broomand gorse, is vigorous,dense, and very tion of the westernescarpment of the Mount Lofty rapid. Ranges. Trans. Roy. Soc. South Australia. 70: 313- The climaxcommunities are not changedby fire, 47. Trumble,H. C. 1939. Climatic factorsin relationto successionbeing a rapidprocess. However,if the the agriculturalregions of South Australia. Trans. indigenousvegetation is disturbedby cultivation, Roy. Soc. South Australia.63: 36-43. vigorous, exotic, scrub growth can replace the Wood, J. G. 1937. Vegetation of South Australia. slower growing, less dense, indigenous scrub GovernmentPrinter, Adelaide, S.A. 164 p. forms. . 1958. The vegetationof South Australia,p. 84- 95. In R. J. Best (ed.), IntroducingSouth Australia. MelbourneUniv. Press, Melbourne,Australia. LITERATURE CITED Wood, J. G., and R. J. Williams. 1960. Vegetationof Black, J. M. 1943-57. Flora of South Australia. 4 Australia,p. 67-84. In C.S.I.R.O., The Australianen- parts. 2nd edition. GovernmentPrinter, Adelaide, vironment,3rd ed., Melbourne Univ. Press, Mel- S.A. 1008%. bourne,Australia.

PLANT BIOMASS AND PRODUCTIVITY OF PRAIRIE, SAVANNA, OAKWOOD, AND MAIZE FIELD ECOSYSTEMS IN CENTRAL MINNESOTA J.D. OVINGTON The Conservancy,London, England DALE HEITKAMP AND DONALD B. LAWRENCE Departmentof ,University of Minnesota,Minneapolis, Minnesota

INTRODUCTION sand and soils, so thatmany fields staked out Cedar Creek Natural History Area is situated by the earlysettlers were soon abandoned,in some 50 km (30 mi) northof Minneapolisand St. Paul, cases after only one crop had been planted and Minnesota, and is about 1,620 hectares (4,000 failed. The wooded areas were selectivelylogged, acres) in area. In view of its nearness to these particularlyfor whitepine, Pinus strobus,and no urban centers,the influenceof the early settlers doubt burningand forest destructionfrequently and their successorshas been surprisinglysmall. followedlogging but usually natural regeneration The firstwhite settlers arrived in 1856 and found restored some form of tree cover. Eventually, a patchworkvegetation reflecting in part the ef- large blocks of land were purchased by private fectsof burningby Indians (Pierce 1954). Settle- landownerswho, because theyappreciated the soli- mentby the European pioneerswas neververy in- tude and wildernessvalues of the area, protected tensive,probably because of the infertilityof the it fromfurther development and terminatedvir- Winter 1963 BIOMASS AND PRODUCTIVITY OF ECOSYSTEMS 53 tually all agriculturaland forestryoperations so Oil whletherit was green or nlot) and non-living that the general landscape has probablyaltered plant material. From the same 20 quadrats all littleand the regionretains its wildernesscharac- the abovegroundplant litter was removed and teristics. Throughthe generosityof a numberof separated into amorphous matter and relatively farsightedpeople the area is now underthe guard- undecomposedplant material,which was further ianship of the Universityof Minnesota and the sortedaccording to whetherit originatedfrom the MinnesotaAcademy of Science, who are continu- herbaceousor woodyplant layers. ing the past policyof preservingits naturalstatus Samplingof the shrubswas not done by quadrat whilstencouraging suitable biological research. clippingbut was based on aerial stemlengths. In Since the land surfaceis gentlyundulating and the winterof 1958-59,prior to the main sampling the plant cover varies locallydepending upon dif- season of 1959, the heightsof all shrubs in the ferencesof soil, climate,and past land use, the type plot were measured. Subsequently,at each Cedar Creek Natural History Area provides a samplingperiod severalaerial stemsof each shrub uniqueopportunity to comparevery different types were collected,their heights (exclusive of of relativelyundisturbed ecosystems within a fairly new stem produced in the sample year) being compactarea, and thesecan be contrastedwith the within10 cm of the average heightmeasurement highlyartificial communities on adjacent agricul- recordedfor the species the previouswinter. The tural land. Some of the pioneerwork on the dy- numberof shrub stemstaken varied accordingto namics of natural ecosystemswas done at Cedar the of the species but was normally Creek, notablyby Lindeman (1942). The pres- about ten for each species at each samplingocca- entaccount is concernedwith comparisons of areas sion. These stems were collectedfrom the type of prairie,savanna, and oakwood,and a neighbor- examplebut outsideof thetype plot, since repeated ing fieldof maize, all of whichare on an upland, samplingsof shrubs in the type plot would have sandysoil type,and in close proximity. seriouslymodified the vegetation. The collected shrubstems were divided into theirvarious com- METHODS ponentsas given-in the tables and the weightsof A general surveyof the Cedar Creek Natural the parts per uniitlength of old stem were calcu- HistoryArea was made in autumn1958 in order lated. Asstumingthat the weights of the com- to identifythe main plantcommunity systems and ponentsper tinitlength of old stemwere the same to select single examples of typical prairie, sa- forthe collectedshrubs and forthose measuredin vanna, and oakwood for detailed study in 1959. the plot the previouswinter, the shrubweight for None of the selectedexamples showed evidence of the plot could be determinedon an area basis by recenthuman interference, burning, or grazingby multiplyingthe averagemeasured weights per unit domesticanimals. Toward the centerof each type lengthof old stemby the total lengthof old stem a plot 30 by 30 m square was pegged out for de- in the plot. tailed monthlysampling, hereafter designated as Tree samplingwas more difficultthan that of the typeplot. The same samplingmethods were the herbs or shrubs and throughoutmost of the used in all threetype examples, but differenttech- observation period monthly sampling was re- niqueshad to be adoptedto samplethe herb, shrub, strictedto the new shoots. Living brancheswere and tree layers. For comparisonwith the natural cut fromthe general tree canopy of the type ex- areas, a field of maize on land adjacent to the amplesoutside of the typeplots using an extension Natural HistoryArea was sampledusing different prunier. The average weightsof , fruit,and methods,and we are gratefulto the owner,Alvar new stemfor a large numberof shoots formedin Peterson,for permission to samplehis crop. the samplingyear were determined. In August To sample the herb layer (here all non-woody and Septemberof 1959 some treeswere felledjust ) each plot was subdividedinto quarters, outsidethe sample plots; size range approximated across each of which a 15- by 15-m grid was thatof the treesin the plots. Three trees (north- marked out to give 225 squares, each a square ern red oak) were harvestedin the oak meter. At each samplingdate, fivemeter squares and six in the savanna (three northernpin and were selectedrandomly in each quarter. Within three bur oak). The numbersof currentyear each meter square, the abovegroundvegetation stems per felledtree were countedso as to give froma centralsquare quadratof 20 by 20 cm was an estimateof the numbersof new stemsproduced clipped. No quadrat was sampledtwice and this in the plots. By multiplyingthe numberof new arrangementpermitted access and samplingwith- shoots by the average shoot weight the total out risk of serious tramplingof future sample weightof the shootsformed in 1959 could be esti- quadrats. Later, the clippedvegetation was sepa- matedon an area basis. The felledtrees were also rated in the laboratoryinto living (mainly based separated into bole, living branches older than 54 J. D. OVINGTON AND OTHERS Ecology,Vol. 44, No. 1 those of the sample year, dead branches,stems beginningin the second week of each monthand produced in the currentyear, and . The being completedwithin the following2 weeks. various tree parts were weighed and the results The samples were usually taken in sequence,viz. convertedto an area basis forthe sampleplots. herbaceouslayer and litter,shrubs, trees, and fi- To obtain informationon the subterranean nally . Each plant layer was sampledfor all plantparts, in each typeplot cylindricalsoil cores communitiesbefore starting on the next layer to approximately77.7 sq cm in cross section and make comparisons between communitiesmore 50 cm deep were extractedmonthly from ten of meaningful. the quadratsfrom which the herbaceouslayer had Detailed recordsand statisticalanalyses will be been removed. The soil so collectedwas washed providedon requestto the Departmentof Botany, with a jet of water througha finemesh sieve on Universityof Minnesota,or to the Nature Con- whichroots and subterraneanstems were retained servancy,London, England. forlater sorting by hand. Weaver (1959a, b) has describedin detail the developmentof the under- DESCRIPTION OF SAMPLE PLOTS groundparts of typicalprairie plants, and although Plant species seen in the sampleplots and their many roots and subterraneanstems go deeper immediatevicinity are listed in Table I; nomen- than 50 cm they tend to be concentratedin this clatureis thatof Fernald (1950) and the identifi- upper zone. cationswere checked by Dr. J. W. Moore to whom The maize was sampled as follows. Towards we are most grateful. Herbariumspecimens are the centerof the fielda plot was markedout 40 filed at the Cedar Creek field laboratoryand at rows wide and 60 m along the north-southrows. the Herbarium of the Universityof Minnesota. At each monthlysampling 20 maize plants were The prairie and oakwood vegetationswere quite taken, one from every alternaterow, the plants distinct,having only two species in common; the beingselected in the rows by randomnumbers. At savanna floracontained a numberof species pres- the same time the heightsof 400 maize plants in ent in the othertwo areas and can be regardedas the plot were measured. To determinethe weight intermediatein character,although the threeplots of maize plants per unit area, the average weight do not representa successional sequence from per unitlength of the 20 sampleplants was multi- prairieto forest. plied by the total lengthof maize plants per hec- The prairietype plot (Fig. 1, upper) was in the tare,based on a figurederived from the measure- same generalarea as stand 3 of Bray (1960) who mentof the 400 plants and countsof the number reportedan old residentas saying that the area of maize plants in the plot. Each sample maize was never completelyploughed or grubbed,the plantwas dividedinto stem, leaves, and ear (grain, crop being planted in single furrows. Soil pro- cob, and husk) which were weighed separately. filesover the area showed no evidenceof earlier The weeds were collectedby takinga distanceof ploughingand since the yield would have been 1 m along the row of maize northof each of the poor any desultoryattempts at cultivation,if made, 20 sample maize plants and collectingall weeds were probablysoon abandoned. The vegetation betweenthe sample row and the next row to the was tall-grassprairie and the mostcommon plants east over themeter length. The rootsof themaize were the threegrasses Stipa spartea,Poa praten- plants and the weeds were removedas carefully sis, and Andropogongerardi, which occurredin as possible by looseningand diggingthe soil to a patches. Althoughthe two shrubsRosa arkansana depth of about 50 cm and attemptingto remove var. suffultaand sand cherryPrunus pumilawere completeroot systems. abundant (101 stems of rose and 115 of sand Samplingwas always done when the vegetation cherryin the plot of 900 sq m), theydid not form was dry (free of dew and rain), and smallersam- a dominantfeature of the vegetationsince they ples such as the herbaceouslayer and soil cores were not bushy and their average heightswere wereplaced in plasticbags to avoid excessivewater only 18 and 36 cm, respectively.The vegetation loss and for transportto the laboratory. Usually was fairlyopen and burrowingrodents and low samples were taken to the laboratorywithin an content of soil colloids probably preventedthe hour of collectionwhere they were quicklysepa- formationof a close cover. Deer were seen in the rated out for weighingfresh. All or a large por- area but no deer-browsedor -grazed plants were tion of each typeof freshplant materialwas then observed. Tree seedlingswere completelyabsent cut up and thoroughlymixed, and three sub- even thoughthe prairie area was surroundedby samples of each were dried at 80?C to determine savannahaving a fairdensity of seed-bearingoaks. the ovendryweight. The savanna type plot (Fig. 1, middle) con- From April to November1959, a completese- tained 17 treesand 8 shrubbyclumps (average of ries of samples was taken everymonth, sampling 10 stemsper clump) of bur oak, Quercus macro- Winter 1963 BIOMASS AND PRODUCTIVITY OF ECOSYSTEMS 55 TABLE I. Plant speciespresent (P) in prairie,savanna, Table I (continued) and oakwoodecosystems in centralMinnesota Typeof Species plant Prairie Savanna Oakwood Typeof body' Species plant Prairie Savanna Oakwood bodyloy ______~____ ~ ~ ~ ~ ~ ~ Quercusrubra var. borealisMichx..... T P Bromuskalmii Gray ...... H P Arenarialateriflora L...... H P Boueelouahirsuta Lag ...... H P Anemonequinquefolia var. interior Panicum capillareL ...... ,. H P Fern...... H P Cyperusfiliculmis Vahl ...... H P Aquilegiacanadensis L ...... H P Mollugoverticillata L ...... H P Fragaria virginianaDuchesne ...... H P Anemonepatensvar. cvolfgangianaBess. H P Rubus idaeus var. strigosusMichx .. S P Delphiniumvirescens Nutt ...... H P Rosa blanda Ait...... S P Helianthuslaetifiorus var. rigidutsCass. H P Amiphicarpabracteata L ...... H P Achillealanulosa Nutt...... H P Acer negundoL ...... T P Senecio plattensisNutt ...... H P Parthenocissusinserta L ...... S P Equisetumhyemalevar. affine Engelm.. H P P Vitis riparia Michx...... L P Poa pratensisL ...... H P P Vacciniumangustifolium Ait ...... S P Stipe sparteaTrin ...... H P P Galium borealeL ...... H P Panicum virgatumL ...... H P P Asterazureus Lindl ...... H P Andropogongerardi Vitman ...... H P P Astersagittifolius Wedemeyer ...... H P Carex muhlenbergiiSchkuhr ...... H P P Taraxacum officinaleWeber ...... H P Tradescantiaoccidentalis Britt ...... H P P Agropyronrepens L.4 ...... H P Sisyrinchiumcampestre Bickn ...... H P P Zea maysL.4 ...... H Chenopodiumleptophyllum Nutt ...... H P P Setaria glauca L.4...... H Mirabilis hirsutaPursh ...... H P P Ranunculusrhomboideus Goldie ...... H P P 1 H=herb;S=shrub; L=liana; T=tree. Potentillaarguta Pursh ...... H P P 2 Has someaerial woody stems. Rosa arkansanavar. suffultaGreene.. S P P 3 Consideredan herbin thepresent study; only about six shoots were present. 4 Presentin themaize field. Prunuspumila L ...... S P P Lathyrusvenosus var. intonsusButt.. . H P P Euphorbiageyeri Engelm ...... H P P ter of the plot. FourteenQuercus seedlingswere Viola pedatifidaoG. Don ...... H P P Oenotherarhombipetala Nutt ...... H P P present,and the wide range of ages of the treesin Asclepias tuberosaL ...... H P P the area indicatedno rapid ecologicalchange and Asclepiasovalifolia Dene...... H P P Lithospermumcanescens Michx ...... H P P no heavy burningfor some time. The frequency Scutellariaparvula var. leonardiEpling H P P and scattereddistribution of old treessuggests that Monarda fistulosaL ...... H P P the area had not been cultivated;almost certainly Physalis virginianaMill ...... H P P Penstemongrandifiorus Nutt ...... H P P it was grazed,though probably not heavily. Shrubs Campanula rotundifoliaL ...... H P P were a muchmore significantfeature of the vege- Liatris aspera Michx...... H P P Solidago nemoralisvar. decemfiora tation in the savanna than in the prairie,locally (DC.) Fern...... H P P dominatingthe herbaceouslayer. Withinthe plot Coreopsispalmata Nutt ...... H P P therewere 132 stemsof hazel, Corylusamericana, Artemisialudoviciana var. gnaphalodes Nutt...... H P P withan average heightof 72 cm, 365 stemsof the Smilacina stellataL ...... H P P P same species of rose as in the prairie with an Amorphacanescens Pursh ...... H2 P P P Elymus canadensisL ...... H P average heightof 32 cm, and 52 stems of choke Quercusmacrocarpa Michx...... T P cherry,Prunus virginiana,with an average height Quercusellipsoidalis E. J. Hill...... T P of 57 cm. The grasses formedthe dominantfea- Petalostemumpurpureum Vent ...... H P Rhus glabraL ...... S P ture of the herbaceous layer which was much Rhus radicansL ...... S3 P denser and more continuousthan in the prairie. Corylusamericana Walt ...... S P P in the Prunus virginianaL ...... S P P were very abundant savanna, many Pteridiumaquilinum L ...... H P more being presentthan in any otherarea inves- Carex pensylvanicaLam ...... H P tigated. Allium stellatumFraser ...... H P Maianthemumcanadense var. interius The oakwood type plot (Fig. 1, lower), con- Fern...... H P taining72 trees of Quercus borealis,was part of a ratheruniform stand withina mixed conifer- angiospermforest complex which also included carpa,and 3 treesand 6 shrubbyclumps (average Pinus banksiana,P. strobus,Quercus ellipsoidalis, of 6 stemsper clump) of northernpin oak, Quer- and Q. mtcrocarpa. The foresthad been heavily cus ellipsoidalis. The largerbur oaks in the plot logged over, or burned,or both,about 60 years were about 90 years old and 10 m high,while the previously. The treesdid not varygreatly in age; largerpin oaks were about 17 years old and 9 m ringcounts of the threefelled trees which covered high. The average diameterover bark at breast the size rangein the plot,gave ages of 45, 56, and heightof the bur oak trees was 22 cm and of the 58 years. The treesaveraged 20 cm dbh and had pin oak 6 cm. When the trees were in full leaf, a maximumheight of just over 17 m, theircrowns thetree crowns appeared to coverjust overa quar- forminga dense and continuous canopy. The 56 J. D. OVINGTON AND OTHERS Ecology,Vol. 44, No. 1 stems Acer negundo,average height95 cm. In additionblueberry, Vacciniurm angustifolium, was fairlyabundanit, being recordedin about 70% of the quadrats. Compared with the prairie and savanna plots, the herbaceous layer was poorly developedand was absentfrom about a quarterof the 20- by 20-cm quadrats, leaving exposed the surface organic layers which completelycovered the sandymineral soil. The fieldof maize was plantedon May 19, 1959, using Kings Crost Hybrid K-5-3 seed planted singly at a spacing of approximately20 cm in north-southrows about 0.9 m apart. In labora- torytests 2%s of the seed failedto germinatebut, accordingto field counts of the plants 1 montlh afterplantinig, 32% of the seed failed to produce plants. The increased mortalityis tentatively attributedto birds, insects,and mammalseating JEI' seed and youngseedlings. Afterthe firstmonth, mortalitywas negligibleand counts in the type plot consistenitlygave about 31,220 plantsper hec- tare. The soil althoughsandy and inherenitlylow in nutrients,except potassium, has been reasonably well managed with annual additionsof cow ma- nure. The land was also well fertilizedwith am- moniumiinitrate containinig33%s nitrate broadcast .4. at a rate of 200 lb/acre on May 15 when the area .1 ., was plouglhed. At the time of planting on May 19 fertilizercontaininig nitrogeni, phosphorus, and po- tassitumiiin the proportion 4:12:36 was applied at 190 lb/acre and on June 3 a further 167 lb/acre of 33% nitrate fertilizerwas added as a side dress- ing. Weed growth, includinig Agropyron repens andl Setar-ia glauca, although fairly luxuriant, was FIG.~~~~Prir e upe) * svna(ide, an oak not regarded as excessive by the farmers of the district. The greatest average height of maize was recorded in August wheni the average from soil level to the top of the tassel (male inflores- cence') wvas 208 cim. The maize crop was harvested for silage on September 10, but the type plot with a surrounidinigprotective strip of three to four rows was left to permit a final sampling in Oc- tober. The moisture contents of the top 10 cm of min- wroodlanid(lower) typeplots. Views northward.Camera is 5 m southof southedge of plot. The twovertical scales eral soil in the differentareas were determined as 10 fttall are placed 10 m and 20 m northof camera. The percentages of the ovendry weight, and the trend more distantscale is at centerof plot which is 30 m was for higlhermoisture contenit (max 41% ) in square. Lawrencephotos: prairieand savanna,Aug. 7, 1962; oak woodland,Sept. 20, 1962. spring, a consistently low moisture contenit (mim 3% ) fromiiJune to September, and increasing shrub layer was remarkably well developed, the moisture content in October. In general, the soil type plot conitaining1,344 stems of Corylus arneri- of the oakwood was wettest and the soils of the cana, average height 87 cm; 23 1 stems Prtmnus prairie and maize field were driest. These dif- v'irginiana,average height 119 cm; 86 stems Rubus ferences became less marked during the summer idaeiis, var. strigosus, average height 53 cm; 21 so that by midsummer there were no significant stems Rosa blanda.,average hei.ght90 cm; 17 seed- differencesbetweeni the four areas in soil moisture lings Pinus strobus, average height 47 cm; and 7 content. Winter 1963 BIOMASS AND PRODUCTIVITY OF ECOSYSTEMS 57

TABLE II. Ovendry weight of vegetation in the prairie type-expressed in kilograms per hectare

SAMPLING DATE

Vegetationsample September October November April 13 May 12 June 10 July10 August 10 9 12 2

Living vegetation Herb layer...... 24 59 488 642 944 900 358 178 Shrub layer Flowersand fruit.. . . 9. < 8 < 1 < 1 < 1 Leaves ...... 0 1985 0 0 Stemsformed in 1959 1 J J 2 2 2 1 Olderstems ...... 6 4 4 7 6 6 6 5 Total forshrub layer. * 6 5 9 16 16 13 8 6 Subterraneanstems ...... 782 392 549 1,067 763 858 771 832 Roots...... 5,913 3,656 4,162 4,972 2,748 4,558 2,983 3,582 Roots and subterranean stems...... 6,695 4,049 4,711 6,039 3,511 5,416 3,754 4,414 Total forliving vegetation. 6,725 4,113 5,208 6,697 4,471 6,329 4,120 4,605 Dead vegetation Litter...... 2,871 2,044 2,453 3,047 2,687 2,374 3,023 3,805 Total weightof vegetation.... 9,596 6,157 7,661 9,744 7,158 8,703 7,143 8,410

TABLE III. Ovendry weight of vegetation in the savanna type-expressed in kilograms per hectare

SAMPLING DATE

Vegetationsample September October November April 14 May 12 June 11 July9 August 11 14 12 3

Living vegetation Herb layer...... 30 120 564 1,188 1,916 1,574 462 304 Shrub layer Flowers and fruit . 0 5 1 1 1 Leaves...... 0 3 325 341 28 22 4 0 Stems formedin 1959.. 0 7 4 4 4 Older stems...... 30 24 25 35 28 29 25 28 Total forshrub layer 30 27 50 76 68 56 34 33 Tree layer Flowers and fruit...... 0 0 15 21 1,090 0 0 0 Leaves ...... 0 76 1,570 1,277 1,480 1,469 86 88 Branchesformed in 1959 0 20 139 135 263 241 250 312 Older branches*...... (13,511) (13,511) (13,511) (13,511) (13,511) 13,511 (13,511) (13,511) Boles* ...... (16,645) (16,645) (16,645) (16,645) (16,645) 16,645 (16,645) (16,645) Total fortree layer .... (30,156) (30,252) (31,880) (31,589) (32,989) 31,866 (30,492) (30,556) Subterraneanstems ...... 1,061 915 1,083 1,563 1,807 1,023 1,004 1,627 Roots...... 12,010 12,048 11,785 10,335 6,317 13,878 7,807 10,049 Roots and subterranean stems...... 13,071 12,963 12,868 11,898 8,124 14,901 8,811 11,676 Total forliving vegetation. (43,287) (43,362) (45,362) (44,751) (43,097) 48,397 (39,799) (42,569) Dead vegetation Dead stemson trees*. ( 4,024) ( 4,024) ( 4,024) ( 4,024) ( 4,024) 4,024 ( 4,024) ( 4,024) Litter.7...... 7,060 8,169 6,884 8,848 10,280 10,767 12,527 12,469 Total fordead vegetation.. (11,084) (12,193) (10,908) (12,872) (14,304) 14,791 (16,551) (16,493)

Total weightof vegetation.... (54,371) (55,555) (56,270) (57,623) (57,401) 63,188 (56,350) (59,062)

* Sampledonly Aug. 31-Sept. 3 whentrees were felled. Data in parenthesesare based on thesevalues in theSept. 14 column.

RESULTS recognize the limitationsof the data. For ex- The resultsnot only provideinformation relat- ample, the weightsof the threemain plant strata ing to plant biomass, primaryproductivity, and are not given to the same degreeof accuracy; this the decompositionof organicmatter but also dem- situationis inevitablewith such enormousdiffer- onstratethe broad differencesin plant biomass ences in total weightsper unit area as exist be- betweenthe four ecosystemsand the changes oc- tween the tree and herbaceouslayers of the oak- curring in each ecosystemthroughout the year wood. It is notoriouslydifficult to estimatethe (Tables II-V). It is important,however, to weightof the mass accurately,and the data 58 J. D. OVINGTON AND OTHERS Ecology,Vol. 44, No. 1 TABLE IV. Ovendry weight of vegetation in the oakwood type-expressed in kilograms per hectare

SAMPLING DATE

Vegetationsample September October November April 15 May 13 June15 July10 August 12 15 13 4 Living vegetation Herb layer.25 37 91 81 159 207 66 41 Shrub layer Flowers and fruit . 67 1 l Leaves ...... 2 67 52 218 184 140 33 57 Stemsformed in 1959.. J ) 22 40 58 Older stems...... 319 244 368 371 407 458 464 490 Total forshrub layer. . . 321 311 520 589 613 638 555 547 Tree layer Flowers and fruit...... 0 0 3 7 11 20 0 0 Leaves ...... 0 853 2,389 2,626 2,848 3,543 975 1,274 Branchesformed in 1959 0 53 115 172 292 483 220 280 Older branches*...... ( 49,019) ( 49,019) ( 49,019) ( 49,019) ( 49,019) 49,019 ( 49,019) ( 49,019) Boles* ...... (111,888) (111,888) (111,888) (111,888) (111,888) 111,888 (111,888) (111,888) Total fortree layer .... (160,907) (161,813) (163,414) (163,712) (164,058) 164,953 (163,297) (162,461) Subterraneanstems ...... 74 51 48 114 126 122 90 192 Roots ...... 12,881 15,409 19,249 20,630 13,517 15,738 11,844 9,889 Roots and subterranean stems...... 12,955 15,460 19,297 20,744 13,643 15,860 11,934 10,081 Total forliving vegetation. (174,208) (177,621) (183,322) (185,126) (178,473) 181,658 (175,852) (173,130) Dead vegetation Dead stemson trees*. ( 21,838) ( 21,838) ( 21,838) ( 21,838) ( 21,838) 21,838 ( 21,838) ( 21,838) Litter...... 34,226 37,122 51,944 38,622 23,917 31,511 41,249 35,279 Total fordead vegetation.. ( 56,064) ( 58,960) ( 73,782) ( 60,460) ( 45,755) 53,349 ( 63,087) ( 57,117) Total weightof vegetation.... (230,272) (236,581) (257,104) (245,586) (224,228) 235,007 (238,939) (230,247)

* Sampledonly Sept, 23-24 when trees were felled. Data in parenthesesare based on these values in theSept. 14 column.

TABLE V. Ovendry weight of vegetation in the field of tree trunksand main brancheswere not deter- maize 1-expressed in kilograms per hectare mined until near the end of the growing season and consequentlythe monthlychanges in the tree SAMPLINGDATE layer weights reflectdifferences only in stems, Vegetationsample September October leaves, flowers,and fruitformed in the sample June12 July14 August20 10 15 year. Finally,no attemptwas made to assess the Maize amountof photosynthateused in plant respiration Fruit(including grain,cob, and or the weight of living plant material eaten by husk).. 1,848 4,719 5,537 . Leaves .18 1,881 1,740 1,689 1, i44 Stem (includingI maleinflore- Differencesof plant biomass among scence)...... J 2,275 2,029 1,567 thefour ecosystems Roots.8 902 628 589 472 Total .26 2,873 6,491 9,026 8,720 On all occasionswhen the fourcommunity sys- Weeds tems were sampled,the amountsof the different Shoots.4 302 1,019 1,011 889 Roots and sub- types of organic matterpresent differedgreatly terraneanstemrs. 2 309 155 83 108 fromarea to area. For example, in September Total.6 611 1,174 1,094 997 the fieldof maize, the savanna, and the oakwood Totalweight of vegetation. 32 3 3,394 7 7,665 10,120 9,717 had, respectively,112, 8, and 28 times as much livingvegetation per hectareas the prairie. The 1 Fieldploughed May 15and planted May 19 with 8 kg,ovendry weight, seed/ weights of living vegetationin the savanna and hectare. oakwood were much greater than those of the for these weightsare probablythe least accurate prairie or maize fieldmainly because of the high of all the measurementsobtained. Errors result proportionof woody plants present,but the her- fromfailure to cut the large tree roots with the baceous layer of the savanna, even thoughpartly soil corerand the loss of fineroots in washingthe shaded,was double the weightof thatof the prai- soil away. Unexpectedlylarge differencesin the rie. The woodlandherbaceous layer was thepoor- total weightsof roots recordedfor the soil cores est developed. The distributionof living' plant fromeach plot occurred,and in view of thisvaria- 1 "Living"here refers to wholeplant bodies of theliving bilitymore soil coresper plotwould have increased plants,except for main roots and stumpsof the woody the precisionof the estimate. The weightsof the plants,which were not sampled. Winter 1963 BIOMASS AND PRODUCTIVITY OF ECOSYSTEMS 59 matter among the herbaceous, shrub, and tree it originatedfrom the plants of the herbaceousor layersvaried, but the overallaverages for the year woodylayers, but in the oakwoodthere was a con- expressedas percentagesof theaboveground living siderableaccumulation of blackamorphous organic vegetationwere for the prairie 98, 2, and 0; for matterover the mineralsoil, almostfour times the the savanna 2, 0.1, and 98; and for the oakwood weight of relativelyfresh litter overlyingthe 0.05, 0.4, and over 99; so thatin none of the three highlydecomposed (Table VII). communitysystems did the shrub layer represent Virtuallyall of the plant materialpresent had a largeproportion of thetotal living plant biomass. been producedwithin each type plot. The most Althoughthe values given for the weights of notableexception was the prairie into which oak roots and subterraneanstems are probably too small, particularlyfor the roots of the wooded TABLE VII. Ovendry weight of plant material in the ground litter-expressed in kilograms per hectare (aver- area, a significantpart of the plant biomass was ages of all observations in sample period, April-No- below groundlevel. The average weightsof roots vember) and subterraneanstems in the prairie, savanna, oakwood, and field of maize are 4,824, 11,789, 14,977, and 650 kg/ha respectively(Table VI), Vegetationsample Prairie Savanna Oakwood equal to 91%, 27%, 8%, and 1% of the living Litterfrom trees and shrubs TABLE VI. Plant biomass,ovendry weight, in foureco- Leaves ...... 13 1,337 3,550 systems-expressedin kilogramsper hectare (averages Acorns,twigs and bark. . . 0 1,447 3,873 of all observationsin sampleperiod) Litterfrom herbaceous layer. 2,775 6,841 97 Amorphousplant material. . 0 0 29,215 Total litter...... 2,788 9,625 36,735 ECOSYSTEMAND SAMPLE PERIOD Vegetationsample Prairie Savanna Oakwood Maize (Apr.- (Apr.- (Apr.- Nov.) Nov.) Nov.) (June-Oct.) leaves were blown fromthe surroundingsavanna Herbaccous layer...... 449 770 88 5,536 areas and held between prairie plants, but oak Shrub layer...... 10 47 512 0 leaves were nevermore than2%o by weightof the Tree layer...... 0 31,223 163,076 0 Roots and subterraneanstems. 4,824 11,789 14,997 650 prairielitter. Total livingvegetation ...... 5,283 43,829 1; 8,673 6,186 Litteron ground...... 2,788 9,625 36,735 0 Changes in the Total dead plant material..... 2,788 13,650 58,572 0 four ecosystems Total plant material...... 8,071 57,479 237,245 6,186 throughoutthe year The total dry weightsof all types of organic matter recorded in the four ecosystemsvaried plant biomass. The subterraneanstems collected considerablythroughout the samplingseason from in the soil cores were mainlyfrom plants of the 32 to 10,120 kg/ha in the fieldof maize, 6,157 to herbaceous layer and when separated from the 9,744 in the prairie, 54,371 to 63,188 in the roots gave average weights of 741, 1,208, and savanna, and 224,227 to 257,103 in the oakwood. 87 kg/harespectively for the prairie,savanna, and These differencespartly reflectinherent differ- oakwood, i.e., 165%, 155%, and 101% of the ences in the chosen samples,and partlychanges weightof the herbaceouslayer present. in amountsof accumulatedphotosynthate. Except Considerableamounts of dead plantmaterial had for the maize field,where there was a progressive accumulated aboveground in the three natural increasein the totalweight of the vegetationfrom plantcommunity systems; thus the average weights April to Septemberand a decrease in October of obviouslydead organic matterper hectarefor and November,the total weightof organicmatter the prairie, savanna, and oakwood were 2,788, in the ecosystemsvaried irregularlythroughout 13,650,and 58,572 kg, respectively. Such mate- the year, partlya resultof the varietyof factors rial as the heartwoodof the treeswas includedin influencingthe totalplant biomassvalue estimates the living biomass althoughthe cells are mainly of the naturalcommunities. dead. In the savanna and oakwood the dead or- Before the onset of spring growth,there was ganic matterwhich had not fallento the ground little significantdifference between the prairie, but remainedattached to the trees and shrubsas savanna,and oakwood in the weightsof the living dead branchesamounted to about one-thirdof the herbaceouslayers. In all threeareas the weight total dead plant matterin the savanna and just of the herbaceous layer increased from April under a half in the oakwood. In the prairieand to August or September,and there was a rapid savanna virtuallyall of the dead organic matter fall in dry weight in October and November; overlyingthe mineralsoil was undecomposedand sincethe Novembervalues weregreater than those could be readily separated accordingto whether of April therewould possiblybe furtherdecrease ()() J. D. OVINGTON AN'D OTHERS i y)1O04v, Vol. 44, No. 1 (luriingthe overwiniterinigperiod. Over the whole are only a small l)art of the total imiass. \Whentlhe growing perio(1 the greatest increase in dry weight April samples were collected, the tree buds hiad of the herbaceous layer occurred in the savanna not opened. Growth of the iieNw slhoots was rapid and the least in the oakwood. The herbaceous in the following months,and the maximumllweights layers of the inaturalcommunities had made about were reached in August or September, after which onie-thirdof their total growth in early spring be- the weight decreased rapidly to November Nhein fore the imaize wvasplanted. By June the herba- much of the fruit anid leaves had fallen. ILeaves ceous layer of the prairie had increased to about constituted the largest part of the shoot produc- 20 times the April Neight while the corresponding tion, but in the savanna large amounts of acorns inicrease in the savanna was 19 times and in the were produced wlhiclhincreased rapidly in veiglht oakwood 4 timies,hut dur-iingmost of this period at the end of July and( beginnin'11gof ANuglustand of active growth, the field of maize was bare and were shed before the Septemiber samiplinig. rhe tiii)rodtuctive. Although the production of or- samp)le year imay have beenI unusual in the large ganiicimiatter was delayed in the planted crop, wlhen amount of acornis produced oii the trees of the the imiaizeaiid its associated weeds started growing, savainina,but the abundance of o0l( acorn cuptnles the wveiglhtinicrease wvascomiparatively rapid, and oIn the grounid suggestedI that ligh acorni prodtlc- fromiijune to July dry weiglht increased over a tioIn was a fairlyregular evenitin tlhesavanniia area, hulnd(redfoldso that the weight of living vegetation possiblv because virtually the wh-ole of each tree in the iaize fiel(d in July greatly exceeded the crowniwas in fulll lighlt. weights of the herhaceous layers in the natural Thle weights of otlher comiponeintsof the eco- ConilIUinlity systems, (lespite their earlier start. systeimirecorded at each sampling period, namiely rhe w,veightsof all the shrub layers increased roots, subterranean stems, and litter, did niotshoW froImllow values in April, whein the buds of the marked progressive changes through the year. In woodland shrubs had just begun to open, to maxi- the case of root systems this lack of change may be iiitiiii values about August, after which leaf fall due to inadequacy of sampliing,but in the case ot occtirre(land(I weicht (lecreased rapidly to Novem- subterranean steimis chainge in gro'1ss. WneighIts her. Thlie recorded weights of the older shrubl) tlhrouglhoutthe year cotuldin fact Ic imir,a Since a-s stemiisin the prairie and savanna were nIot greatly old rlhizonmesare exhausted niew rhizomlles may 1e (lifferelit at the beginining and einl of the growiing forme(d aIndI the total rhizolllmemIaSs w\-otuldnlot se.asonlo, i.e., in April anid Novemilber,but in the chanige greatly. Greater weights of litter .ccnlmed oa),kwoodthere hadl been a large itncreasein weiglht to be present in the autumniiiiimoniths whl eni leaf fall of the older slhrul) stemiis. OIn the wlhole, the occurre(l, but inicreases wver-eniot ats imarl-kedas major clhanigesin the weiglhtsof the shrub layers mlight be expected. While bur oak leav-es fall throughout the year can be attributed to the pro- promptly in October, many of the leav es are re- dutctioniof new shoots, the magnitude of change tainled oIn northerinpin, nortlhernred, anld wvhite and the l)roportion of stemi, leaves, flower, and oaks througlhoutthe winiterand( they fall only a frtit v-aryingaccordiing to the species of shrub, week or tws-obefore bud expansioin in sprinig. In1- timie of year, and shrubl)size. \When the shrub creasing litter fall may hlave been compensate(l to layers attainiedtheir mlaximiiumiiweights, the weight some extent by increasiig in the of the new shloots formed a large proportion of warmil,wet autumin weatlher, and(I eveen if, in the thle g-ross weight, in the case of the prairie and oakwoo-o(lfor inistaince,all the leav-esand(I fi-ruit of the savainia mior-etlhani the wveightof the oldler stems. herbaceouts and shrul) and tree layers bad falleni I eaves constittited the largest part of the new froimiAugust to Nov-emb)er,the av erage mionthly shoot weight. After leaf fall in November the fall would have been only,about 1 .000 kgy'/haonito weight of the newvstems wvasotnly about 12%7cof a litter nmassweigg-hing about 40,000 kg 'ha. The that of the old stems so that if the weights of the increase in weight would be (liffictultto demiioni- shrub layers are constant from year to year, this strate more intensive sampling. 12% increase in stem xveightmust be counterbal- aniced by a correspondinigimortality of either old or Prodiction of organic msa-tter new stems in the winter period. Lawrence, who has visited these areas regularly for imany years, Aninual productiv,ityis Inot synonymous with believes that the shrub layer has gradually becomie plant biomass nor with gross changes in plant more lutxtrianltcltie to receintlprotection from btirn- biomass from year to year, for a plant com-iunity ing so that stemiiproduction aindIimiortality may not is normally composed of many differentspecies and be exactly equal. individuals of the same species which do not neces- The tree layers differfrom the shrub layers in sarily attain their greatest itndividualweights at the that the new shoots formed in the sample year time of maximum biomass. Further- Winter1963 BIOMASS AND PRODUCTIVITY OF ECOSYSTEMS 61 more some of the organic matterproduced, e.g., annualproductivity was greaterin thecommunities flowers,bud scales, and lower leaves, are shed be- wherethe proportion of woodyplants was greater. fore individual plants attain their maximum The shrubs and trees ratherthan the forbs and weights.The situationis furthercomplicated when grasses became the most productivemembers of the plantcommunity contains biennials and peren- the community,as the zone of photosynthetic nials. Finally,sampling can rarelybe so frequent activitybecame more complete and extendedhigher thatthere is no dangerof missingthe peak weights above the 'ssurface. ofthe plants. Odum (1960), workingwith the rela- Despite its short growing season the field of tivelysimple ecosystems characteristic of the early maize gave a high annual productionof organic successionalstages followingthe abandonmentof mattereven thoughit containedno woody plants, fields,has overcomesome of the difficultiesasso- but in thiscase the soil had been greatlymodified ciated with the determinationof organic produc- and the high level of organic matterproduction tionand turnoverby weighingtogether at monthly was dependentupon heavyapplication of fertilizers intervalsall plants of a given species, also litter to the soil. Bray,Lawrence, and Pearson (1959) of these species,separately. In the presentinves- have shown the large differencesin productivity tigation the maize, the shrubs, and to a more betweenmaize grownwith and withoutfertilizers. limitedextent the trees,have been sampled sepa- Provided soil conditionsare suitable,maize, al- ratelyby species. The herbaceouslayer, consisting thoughrelatively late in startingactive growth,is mainlyof perennials,was so complexthat it was able to grow rapidlyand, by virtueof its height consideredonly as a whole, and estimatesof its and density,carries a large weightof photosyn- annual productivityare thereforebased on the theticorgans, much greater than the prairievege- differencebetween the recordedmaximum weight tation. Most of the organic matteris produced and lowest overwinteringweight of the above- betweenmid-May and mid-Septemberover a pe- groundparts. Consequentlyproductivity estimates riod of about 125 days,which amounts to an aver- (Table VIII) will tendto be low. age daily productionof about 85 kg/ha for both maize and weed plantstogether. TABLE VIII. Annual primarynet productivity,ovendry weight-expressedin kilogramsper hectare DISCUSSION ECOSYSTEM All fourcommunity systems were very hetero- Vegetationsamp)le _._ _ geneous and the variabilitychanged considerably Oakwood Maize Prairie Savanna throughthe year (Tables II-V). The maize field Herbaceouslayer .920 1,886 182 9,456' was the simplestand mostuniform ecosystem, but Shrublayer .10 412 389 0 Treelayer in July the 400 measured corn plants varied in Shootsof sample year ... 0 2,833 4,046 0 heightfrom 73 to 153 cm, while in the,following Bolesand branches older thancurrent year 0 5033 3,5753 0 monththe heightrange was from128 to 276 cm. Total foraerial parts. 930 5,263 8,192 9,456 In additionto this spatialand timediversity, each Rootsand subterranean plant communitywas composed of plants with a 1,211 stems.- - - wide range of forms,e.g. frommosses to trees in 1 Peakbiomass of maize in Septemberplus peak of weeds in August. the savanna and oakwood,and the plant material 2 Peakof 1959stems, leaves, and flowersin July;increment of wood in older stemsconsidered to be zero(see Table III). exhibitedall stages of decomposition.With such 3 Thesevalues may seriously underestimate since they are mean values obtained bydividing the existing biomass by treeage. diverse organic matterso irregularlydistributed, multiplesampling is necessaryto characterizeeach The amounts of organic matterproduced an- ecosystemas a whole. From a practicalviewpoint, nually in the ecosystemsdiffer greatly (Table however,the intensityof samplingalso depends VIII). Site factorssuch as soil conditionswere uponthe available man-hours and theneed to avoid not exactly comparable, since the prairie and samplingso intensivelythat the vegetationis se- savanna areas were on dune sand fromwhich the riouslymodified for futurework. The sampling mineral colloids have been winnowed by wind, techniquesrepresent a compromisebetween these whereas the oakwood was on glacial outwash,a differentconsiderations but neverthelessserve to primaryphysiographic surface with a muchhigher demonstratebroad differencesin plant biomass colloidal content(Cooper 1935). In the case of and some of the seasonal changesoccurring in the the three natural communitiesit seems unlikely fourecosystems. that the site differenceswere sufficientlygreat to Over the wintermonths the maize fieldis bare accountfor the large differencesin annual produc- of vegetationand the soil surfaceis exposed to the tivity. Differencesof formand structureof the weather. In contrast,the threenatural communi- vegetationseem to be the more importantfactors ties always have a plantcover and withthe advent influencingproductivity, and it is significantthat of springare able to initiategrowth quickly and 62 J. D. OVINGTON AND OTHERS Ecology,Vol. 44, No. 1 effectively.It is clear, however,that duringthe primarilyto the presenceof woody perennialsca- period when the maize plantsare well established pable of makingfuller use of site conditions. and cover the ground,very efficientuse is made In contrast,the artificialcommunity of the ofthe site so thatthe monthly production temporar- maize fielddoes not containwoody perennials, but ily exceeds that of the naturalcommunities. The the recordedannual productivityis relativelyhigh higherefficiency of the maize field at this stage and appears to be about equal to thatof oakwood. comparedto the oakwood may be due to the high The recordedestimate of annual productivityof levels of soil nutrientsavailable as a resultof the the oakwood is certainlytoo low, however,for the fertilizerapplications and to the factthat a much annual productionof treeboles and main branches higherproportion of the livingcells are photosyn- is an averagefor the lifeof thetrees, during which theticin the maize than in the oakwood plants. growthhas increasedgreatly, and withinthe eco- The data have been analyzed statisticallyunder systemas a whole some treesand possiblyshrubs the guidance of M. D. Mountford,and the main and herbaceousplants will have been suppressed, differencesdescribed among the four ecosystems killed, and decomposed. Consequentlyit seems provedto be significant.To give some indication thatthe oakwood is the most efficientproducer of of the effectivenessof the samplingtechnique, data organic matter. Furthermore,the maize can be forthe herbaceouslayer and root mass were sum- sustainedon a long-termbasis onlyby considerable marizedin termsof the fourquarters of each type effortand the applicationof considerable plot (Table IX). Throughoutthe year the small- amounts of fertilizer. In contrast,the oakwood est weightof the herbaceouslayer recordedfor a has received no input of human energyand no applicationof fertilizers. TABLE IX. Ovendry weights of the aerial herbaceous layer and of the subterranean organs of the quarters of The organicmatter collected in thisinvestigation the type plots-expressed in grams has been analyzedfor , energy, and nu- trientcontent, and later papers will considerthe TYPE PLOT fourecosystems from these aspects.

Prairie Savanna Oakwood Month _ _ SUMMARY Min Max Avg Min Max Avg Min Max Avg Sample plots were markedout in typicalareas Herbaceous layer per clip quadrat of prairie,savanna, and oakwood, and in a field April...... 0.1 0.1 0.1 0.1 0.2 0.1 0.1 0.2 0.1 of maize, at Cedar Creek Natural History Area, May ...... 0.2 0.3 0.2 0.4 0.5 0.5 0.1 0.3 0.1 JUne...... 1.1 2.9 2.0 1.8 2.6 2.3 0.1 1.0 0.3 Minnesota. Plant biomass determinationsfor July...... 1.9 2.9 2.6 3.9 5.5 4.8 0.2 0.6 0.3 monthlyintervals from April to Novembershow August...... 1.5 4.4 3.8 3.8 14.4 7.7 0.3 0.9 0.6 September... 2.8 4.2 3.6 3.2 8.9 6.3 0.4 1.5 0.8 the seasonal rhythmscharacteristic of the different October.... . 1.2 1.7 1.4 1.0 2.7 1.9 0.1 0.4 0.3 ecosystemsand the large differencesbetween eco- November... 0.6 0.9 0.7 0.5 1.9 1.2 0.1 0.3 0.1 systems in the amount of organic matter they Subterraneanorgars per ccre contain. 7.3 10.7 While it is recognizedthat thereare great dif- April...... 3.0 6.3 4.5 6.3 12.5 9.2 15.9 1.7 3.6 2.8 5.2 16.5 9.2 7.2 16.9 11.8 May ...... ferencesin the accuracyof the estimatedproduc- 14.8 June...... 1.9 4.7 3.1 5.6 13.9 9.0 9.2 19.5 tivityvalues, the maize estimatesbeing most accu- July...... 3.1 4.1 3.8 6.4 10.1 7.9 8.9 19.9 15.8 August...... 1.6 2.8 2.7 4.4 5.4 4.8 8.5 11.7 10.4 rate and the woodlands the least, primarynet September... 2.5 3.9 3.5 6.0 14.4 10.6 10.6 13.7 12.1 productivityprobably increases from prairie to October..... 1.8 2.9 2.3 4.3 6.9 6.0 7.3 11.0 9.1 November.. 2.1 3.9 2.7 4.0 9.0 7.7 5.7 11.4 7.6 savanlnato oakwood. The annual productivityof .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~the naturalwoodland is probablyhigher than that of the intensivelymanaged maize field. quarterin the savanna is consistentlygreater than It is suggestedthat in the upland communities themaximum for any quarterof the oakwoodplot. of the Cedar Creek region,the presenceof woody Similarly,for root mass, the smallestvalue for a perennialplants is associated with an increasein quarterin theoakwood is consistentlygreater than three of the dynamicprocesses in materialand the largestvalue for any quarterof the typeplot of ecosystems,i.e., the production, in the prairie. accumulation,and decompositionof organic matter. The living plant biomass decreases fromoak- wood to savanna to prairie,as do the amountof ACKNOWLEDGMENTS dead plant material,annual productivity,and or- This study owes much to a number of people who co- ganic matterturnover, if our assumptionis cor- operated in it, especially to Gerald Martin, Allan Bonde, rect that the weight of dead organic matteris Philip Neumann, and Joyce Ovington who assisted with the collection and sorting of the plant material, and to relativelyconstant from year to year. These dif- Elizabeth Lawrence who kindly checked the manuscript. ferencesin ecosystemdynamics seem to relate We are indebted to Alvar Peterson who planted and cul- Winter 1963 SYMPATRIC GASTROPODS 63 tivated the maize on his own farm and permittedsampling Time. Univ. Minnesota Press, Minneapolis, Minn. beyond the stage at which he would have harvested it for 116 p. silage. Gratefulacknowledgment is made to the Louis W. Fernald, M. L. 1950. Gray's manual of botany. Amer. and Maud Hill Family Foundation, the National Science Book Co., New York. 1632 p. Foundation, and the Graduate School of the Universityof Lindeman, R. L. 1942. The trophic-dynamic aspect of Minnesota for financial assistance. Permission to sample ecology. Ecology 23: 399-418. at the Cedar Creek Natural History Area was given by the University of Minnesota and the Minnesota Academy Odum, E. P. 1960. Organic production and turnover in of Science. old field succession. Ecology 41: 34-49. Pierce, R. L. 1954. Vegetation cover types and land use history of the Cedar Creek Natural History LITERATURE CITED Reservation, Anoka and Isanti Counties, Minnesota. Bray, J. R. 1960. The chlorophyll content of some M.Sc. Thesis, University of Minnesota, Minneapolis. native and managed plant communities in central 137 p. Minnesota. Canadian J. Bot. 38: 313-333. Weaver, J. E. 1958a. Summary and interpretation of Bray, J. R., D. B. Lawrence, and L. C. Pearson. 1959. underground development in natural com- in some Minnesota terrestrial munities. Ecol. Monographs 28: 55-78. communities for 1957. Oikos 10: 38-49. . 1958b. Classification of root systems of forbs Cooper, W. S. 1935. The history of the Upper Mis- of grassland and a consideration of their significance. sissippi River in Lake Wisconsin and Postglacial Ecology 39: 393-401.

TROPHIC RELATIONSHIPS OF 8 SYMPATRIC PREDATORY GASTROPODS ROBERT T. PAINE Dcpartmnentof , Universityof Washingtont,Scattic

INTRODUCTION observerto concentratehis effortson a functional A commonapproach to ecologicalproblems in- unitwithin the community 1lexus in whichthe spe- volves the descriptionof food habitsof singlespe- cies membershipis limited. Where theseunits are cies or, withthe community in mind,of foodchains recognizablethey should serve as a naturalinter- and webs and their energeticimplications. In mediarybetween simple predator-prey interactions recentyears food chains have receivedrelatively in whichsome of the basic biologyfor each com- little quantitativeattention, except as an entity ponentis known,and analysesof whole communi- existing within and contributingto community ties whiclhtend to ignore specificdetails. The organization. Elton (1927) describedone type presentanlalysis has been concernedwith describ- of as originatingfrom of ing such a subdivisioninvolving 8 large (shell various sizes, radiatingout fromthese, and even- length5-35 cm) predatorygastropods and their tuallyculminating in some ultimatepredator with- prey,and with examiningthe associationfor any out consequential predators of its own. Food propertiescharacterizing the functioningof the chainstend to be shortbecause of continuedenergy whole assemblage. degradation at each transfer,and Hutchinson The authoris gratefulto Florida State Univer- (1959) suggeststhat a maximumof 5 linksexists. sityfor permission to reside at the AlligatorHar- Food chains thus are concise units of community bor i\'IarineLaboratory, Franklin County, Florida, structure. where the data were obtained,and to R. B. Root However, the ideally realized food chain is an and C. E. King forcriticism. abstractionsince it may be said thatall organismis either eat or are eaten by more than one other BIOTOPE AND BIOTA species (Allee et al. 1949). It is thus probably Shallow water marine of western difficultto describe a naturallyoccurring multi- Florida provideexceptional, and perhaps unique, linkedfood chain in whicheach successivepreda- opportunitiesto observepredatory gastropods. Not tor eats only a single prey species and when the only is the fauna diverse,but some of the gastro- food consumedat each trophiclevel is completely pods are amongthe largest, and hencemost readily known. It is possible, though,to describe the observable,in the world. Sandbars such as the communityfood web and, because a numberof one to be describedare inhabitednormally by 8 "top predators"exist in each community,to sub- species attaininga length greater than 5.0 cm: dividethe intomore or less discreteunits Pleuroplocagigantea Kiener, Fasciolaria tulipaL., each culminatingin its own "top predator." Such F. hunteriaPerry, Busycon contrariumConrad, a divisionis oftendesirable because it permitsthe B. spiraturnLamarck, Murex floriferReeve, Poli-