ISLAND- MAlNLAND AND WITHIN -._SEASON COMPARISONS 0F COMMUMTYLEVEL 7 PARAMETERS AND A MODEL OF I ‘ ' THEIR INTER-RELATIONSHIPS A i. Dissertation for the Degree of Ph. D. .7 ,MiCHIGAN STATE UNIVERSITY JERRY DEXTER HALL 1974 A rflnat‘r“""‘ N rad L1,.ft’. ‘9. lflichifznn ;_, It ;‘ Y_ ‘ ... .'~ 1",“‘1 I . ; nit-W3" ,; ‘3 9"" I T' I . ‘ _ , . 25“.”; -_ A...“— UBR ARIES IIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIII II II 293 00078 6487 ,This is to certify that the P thesis entitled Island-Mainland and within Season Comparisons of Community Level Parameters and a Model of their Inter- Relationships ! presented by I Jerry Dexter Hall has been accepted towards fulfillment of the requirements for 1:th . degree in Zoology Major professor Date J/‘/'7‘/ 0-7639 IV. "(JAEm & SEIIS' ' BOOK BINDERY LIBRARY BINDERS . srnllmn. Incllm --I-I" ‘V. v -I‘ LL. fifference: Tzese data were parameters that deduzed from the IT‘I 2‘" \~/ U‘I'enneSS Of .“ f‘ L: 3‘ :Tvp.a.C LETS 3:;iected in ta 1 ABSTRACT ISLAND-MAINLAND AND WITHIN-SEASON COMPARISONS OF COMMUNITY LEVEL PARAMETERS AND A MODEL OF THEIR INTER—RELATIONSHIPS By Jerry Dexter Hall Differences in five ecological parameters were investigated between island and mainland communities and during the summer season, 1972. These data were also used to test a model of relationships among the parameters that was derived from a review of literature. The parameters, deduced from the structure of food-web diagrams, are (1) Number of Taxa, (2) Evenness of Taxa, (3) Resource Breadth of Animal Taxa, (A) Evenness of Trophic Levels, and (5) Distinctness of TrOphic Levels. Data were collected in two pairs of forest floor communities, one pair on North Bass Island in Lake Erie and nearby Marblehead Penninsula, Ohio, and the other pair on South Manitou Island in Lake Michigan and nearby Lelanau Penninsula, Michigan. Organisms studied were herbaceous vascular plants and herbivorous, detritivorous and carnivorous invertebrate animals. Non- parametric analyses of variance of island-mainland differences yielded the following results: Number of Taxa is lower in island communities, possibly due to absence of rare taxa. Resource Breadth may be more influenced by scale of sampling than in insularity or seasonality. Evenness of TrOphic Levels is higher in mainland communities than in island communities. Distinctness of TrOphic Levels is higher in mainland communities in Ohio, but lower in Michigan, and increases during the summer season. To test the model of relationships among the parameters, ten predictions were made of correlation among them. Predictions were rut tested by partial: predittions were Fear were refute: appears to clari structure of the n and the model i. Y a. ‘ ' I \ e e-a:;or.sn1ps instigation o: Jerry Dexter Hall tested by partial correlation analyses applied to the data. Six predictions were not refuted and were partially or completely verified. Four were refuted, two with little confidence. A third refuted prediction appears to clarify, rather than refute, the relationship predicted. The structure of the model and existence of the relationships are accepted, and the model is revised and awaits further test. These community-level relationships modeled and tested are complex and consistent and require investigation of mechanisms. ISLAND-MAINLAND AND WITHIN—SEASON COMPARISONS OF COMMUNITY LEVEL PARAMETERS AND A MODEL OF THEIR INTER—RELATIONSHIPS By Jerry Dexter Hall A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1974 a :3. ‘lniA .L"‘ a!- - I'V? ' IiL 'j Irish to e ling, Walter C01 they have pr vi: have rendered d funds for the f use the Bealgé an grateful to «1 0.“ ‘n. ‘+ . 73: “Isl u e Stat at; . ‘0 81m Stu: ‘uent. ACKNOWLEDGMENTS I wish to express sincere appreciation to Drs. Rollin Baker, John King, walter Conley, Peter Murphy and William COOper for the inspiration they have provided throughout my graduate career and for the advice they have rendered during this study. In addition, Dr. Rollin Baker obtained funds for the field work and Dr. William Cooper provided funds for use of the Michigan State University Computer Laboratory. I would also like to thank Dr. John Beaman, who gave permission to use the Beal-Darlington Herbarium facilities at Michigan State University, and Dr. Steve Stephenson, who helped identify monocot plant species. I am grateful to Jesse Saylor, who spent many hours helping identify a wide variety of plant species. Dr. Roland Fischer provided advice, equipment and space for the identification of invertebrate animal specimens. The Michigan State University Museum provided some field equipment and facilities, while the Ohio State University Instruction and Research Computer Center provided computing facilities and services. The Society of Sigma Xi, Michigan State University Chapter, granted funds used to obtain identification of animal specimens. I would like to acknowledge and thank Terry Truax, who provided valuable statistical and computer programming advice; David Dennis, whose artistic skill provided most of the figures, and many friends and various students, who volunteered their help in various parts of the field work, identification and handling of specimens and drawing of figures. ii IL-‘l'W-b"~ Mould 3‘15, handed remiss: Emmiwi m ' a. me, and the I Stocking Land CC transportation Lastly and 1e field work: her tolerance 1 I would also like to acknowledge and thank Joseph Mahler, who provided permission to work in the woodlot studied as the Marblehead Community; Mr. and Mrs. Paul Stonerook, who provided encouragement, advice, and the permission to work in the woodlot studied as the North Bass Community; Graham Downer and the Sleeping Bear Dune Park of the Stocking Land Company, and John Brown, who furnished advice and transportation in the area studied as the South Manitou Community. Lastly and mostly, I want to thank my wife Jan for her help with the field work, handling of specimens and useful advice, but mostly for her tolerance and support throughout this tedious study. iii ”1»? I ‘ “Ii-'71) “JIL'; "' “4‘ 1‘ ". a"... hfivwfimfi' 1"“ . _ I: .. “thus/pub lLUA‘ I an "Pfi'r‘rfl'rrhf‘ am wh'uulilbD or Conn l;t;e Transects Organism Sampling V I':.;CFET " . . selection EXpected E INTRODUCTION . . . . . . COMMUNITIES STUDIED . . Communities . . . . Transects . . . . . Organisms Studied . Sampling Visits . . THEORETICS . . . . . . . Selection of Parameters TABLE OF CONTENTS Expected Effects of Insularity Number of Taxa . . . Evenness of Taxa . . . . . Resource Breadth . . . . Evenness of Trophic Levels Distinctness of Trophic Levels Seasonality Medels of Relationships Among Parameters Free Body Model Concept Number of Taxa . . . . . Evenness of Taxa . . . . Resource Breadth . . . . . Evenness of TrOphic Levels Trophic Level Distinctness System Model and Predictions . iv IEISES .. - - Estimation C Number c Erennes: Resourc' Evennes Distinc Probler Analysis 01 Ana-*sis c: l U 3:113 its :1. T ,1 - anSLlarlLT Number Resou; TTcph: Dist: 7‘ - rred;cted R7,. 1 {I m:- e Numbe aimhe IL, u “1an E‘Jem EVenz 1": P‘eSO' ANALXSES . . . Estimation Number of Parameters . . . . . of Taxa . . . . . . . . Evenness of Taxa . . . . . . . Resource Breadth . . . . . . . Evenness of Trophic Levels . . Distinctness of Trophic Levels Problem Taxa . . . . . . . . . . . . . Analysis of Insular and Seasonal Effects . Analysis of Predicted Correlations RESULTS AND DISCUSSION . . . . . . . . Insularity and Seasonality . . . . Number of Taxa . . . . . . . . Evenness of Taxa . . . . . . . Resource Breadth . . . . . . . Trophic Level Evenness . . . . Distinctness of TrOphic Levels Predicted Correlations . . . . . . Number Number Number Number of Taxa versus Evenness of Taxa versus Resource of Taxa Breadth of Taxa versus Evenness of TrOphic Levels . of Taxa versus Distinctness of Trophic Levels Evenness of Taxa versus Resource Breadth . . . . . . Evenness of Taxa versus Evenness of Trophic Levels . Evenness of Taxa versus Distinctness of Trophic Levels Resource Breadth versus Evenness of Trophic Levels . . Resource Breadth versus Distinctness of Trophic Levels 74 74 7A 76 77 79 81 83 83 83 88 88 88 94 95 97 98 100 105 106 107 108 108 109 109 110 Ill "L, ‘ v “MT'FE.)LK"'.'"' Evenness ”revs Predict . « Mrs. (‘1‘ W n f E CIV‘ F1 3.3.5.114 In“ .. . v-Ev-v‘vt“ ‘T O y -n- a.»- d‘ll r ‘ lulu; Evenness of Trophic Levels y§g§u§.Distinctness of Trophic Levels . . . . . . . . . . . . . . . . . . . . . . . . . 111 Predicted Correlations . . . . . . . . . . . . . . . . . . . 111 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . 115 LIST OF REFEREINCES O O O O O O O O O O O O O O O O O O O O O O O O 0 11? vi A Nina-rm ‘- . 2. List of P1. Visit List of Ir; Occurr Schedule c I‘nree Fcoc' Measu: Table of Z Free; Table of Free- Table of Free. 13. 1c. 15. LIST OF TABLES List of Plant Species and Their Occurrences by Community and ViSit O O O O O O O O O O O O O O O O C O O O O O O O O 0 List of Invertebrate Animal Taxa, Food Habits Ranking, and Occurrences by Community and Visit . . . . . . . . . . . . Schedule of Visits to the Four Communities Studied . . . . . . Three Food Web Components, Rowe, and Three Kinds of Measurements, Columns, and the Resulting Nine Measurements Table of References Relating Ecological Parameters that Lead to Free-Body Model of Number of Taxa . . . . . . . . . . . . Table of References Relating Ecological Parameters that Lead to Free-Body Medel of Evenness of Taxa . . . . . . . . . . . . Table of References Relating Ecological Parameters that Lead to Free-Body Model of Resource Breadth . . . . . . . . . . . . Table of References Relating Ecological Parameters that Lead to Free-Body Model of Evenness of Trophic Levels . . . . . . . Table of References Relating Ecological Parameters that Lead to Free—Body Medel of Distinctness of Trophic Levels . . . . . Correlations Between Parameter Pairs Predicted from the System M61 0 O O C O O O O C C O O O O O O O O O O O O O O O O 0 Table of Symbols Used to Abbreviate the Five Parameters and the Fourteen Variables Used to Estimate the Parameters . . . . Table of Critical t-Values and Critical Numbers of Coefficients Used to Validate Correlation Between Variables . . . . . . . Effects of Insularity and Seasonality on the Five System Parameters O O O O O O O C O O O O O O O O O O O O 0 O O 0 Results of Analyses of variance of Variables in Ohio, Plots scale Of sampling 0 O O O O O O O O O O O O O O O O O O 0 Results of Analyses of Variance of variables in Ohio, Transects Scale of Sampling . . . . . . . . . . . . . . . 16 23 27 30 38 5A 59 69 73 75 87 89 90 91 'V |.'I' t. i -.' '1'“ Results of Scale C Results Of TranseC Summary of System Frequency Value Variat 16. 17. l8. 19. Results of Analyses of variance of variables in Michigan, Plots scale or Smpling O O O O C C O O O O O O O O O C O O O I O 92 Results of Analyses of Variance of variables in Michigan, Transects Scale of Sampling . . . . . . . . . . . . . . . . 93 Summary of Results of Partial Correlation Analyses of the Five System Parameters . . . . . . . . . . . . . . . . . . . . . 101 Frequency of Correlation Coefficients of Minus, Zero, or Plus value Between each Variable Pair for Each State and Type of Variable, and Results of t-Test Validation . . . . . . . . . 103 viii ,. its-rs ' Yap of Sr western I Martlehea North Bas lkrthern Leelanau South Mar Diagram c Free-30d: Free-BM; Free-3 3 F. tee-Emir System IR LIST OF FIGURES Map of Great Lakes Region Showing All Communities Studied . ‘Western Lake Erie Showing the Two Ohio Communities Marblehead Community and Pattern of Transects . . . . . . . North Bass Community and Pattern of Transects . . . . . . . Northern Lake Mfichigan Showing the TWO Nfichigan Communities Leelanau Community and Pattern of Transects . . . South Manitou Community and Pattern of Transects Diagram of a Foodaweb . . . . . . . . . . . . Free-Body Model of Number of Taxa . . . . . . Free-Body Model of Evenness of Taxa . . . . . Free-Body'Model of Resource Breadth . . . . . Free-Body Model of Trophic Level Evenness . . Free-Body Model of Trophic Level Distinctness System Model of Interrelationships among the Five Parameters RBVised syStem MOdel O O O O O O O O O O O O O O O O O O O O \O'\‘I0\U‘t 10 13 1A 29 37 53 58 63 68 71 113 In this 5 com ities on during the so: as a sub-set c the herbaceous hit}. them. T1 Can be deducec‘ Parameters are I2) the evenm Cf resources 1 of the taxa a} the Various t labeled respe (‘2) Evel’lness INTRODUCTION In this study, foodaweb ecology was investigated in forest floor communities on islands and mainlands of Lake Erie and Lake Michigan during the summer season of 1972. A portion of each community was studied as a sub-set of the whole community. This portion, or sub-web, included the herbaceous green plants and selected invertebrate animals associated with them. The investigation focused on five ecological parameters that can be deduced from the structure of a foodaweb diagram. These five parameters are (l) the number of taxa composing a given food web, (2) the evenness of the abundances of those taxa, (3) the average breadth of resources used by consumer taxa, (4) the evenness of the distribution of the taxa among tr0phic levels, and (5) the distinctness with which the various trophic levels are determined. These parameters are herein labeled respectively with the following terms: (1) Number of Taxa, (2) Evenness of Taxa, (3) Resource Breadth, (A) Evenness of Trophic Levels, and (5) Distinctness of Trophic Levels. It is recognized that these parameters are sufficiently general that each can be approached in a variety of wayse For purposes of this study, each of these parameters was estimated quantitatively by two or more kinds of measurements in each community. These measurements, designed to estimate the parameters, are herein called variables to distinguish them from the parameters they estimate. The parameters and the variables are described in more detail in later sections of this report. This 5:111 excerilt‘efital' rtC'I'e for at? 5 «1:13 the 51:: Tara and Peso differences . The secc: , -east 1n Cer‘ “e“ (1971). This study has two goals; one is primarily descriptive, the other experimental. The first goal is to examine the five parameters listed above for any pattern of differences between mainland and island and during the summer season. Numerous authors deal with such insular and seasonal patterns of differences, especially with respect to Number of Taxa and Resource Breadth, and predictions can be made of expected differences. These expected differences are explored in a later section. The second goal of this study is to develop from the literature and test experimentally a model of relationships among the five ecological parameters presented in the opening paragraph. Predictions of relationships between pairs of these parameters are developed from the model and tested by partial correlation analyses. Additionally, the question is asked whether relationships indicated by these correlation analyses can be explained by influences of insularity or seasonality. Studies by previous authors have dealt with one or perhaps several of the five parameters considered here, but investigations of interactions among them, at the level of community function, are few. Examples of such studies are those by Paine (1966), who has shown that a top level predator can influence the number of species in lower tr0phic levels, at least in certain intertidal invertebrate communities, and by Wiegert and Owen (1971), who suggest a modified tr0phic model to explain differences in density levels and in regulatory mechanisms of species of different tr0phic levels. Other studies of interactions among community parameters, dealing with various parameters, are those by Leigh (1965) and MacArthur (1970), who have mathematically treated the relationships among diversity, productivity, stability, and other parameters, and by Connell and Orias a conceptual m (1964) and E. P. Odum (1969), who have treated these same parameters in a conceptual manner rather than mathematically. ,. , :4»: L/Omuro‘ 'v..es P i ‘I COMMIT. floor coma: Eric; the 0' “Frvr- : , 69:. Jam-rte .." 5' - C~U fated “n . a «.9 SchaCE .» h. .w B: COMMUNITIES STUDIED Communities Communities studied include two mainland-island pairs of forest floor communities. One pair of these communities is located in northern Ohio; the other is located in west-central Michigan (Fig. l). The Marblehead Community is the mainland community of the Ohio replicate. It is located on Marblehead Peninsula, Ottawa County, Ohio, approximately one and one-quarter kilometers north of the Sandusky Bay Bridge (U.S. 2), and six and two-thirds kilometers east of Port Clinton, Ohio (Fig. 2). This community is an irregularly shaped wood-lot approximately 9.3 hectares (23 acres) in area and is surrounded by cultivated fields (see Figure 3). The topography is regular, and two small areas are somewhat marshy, possibly indicating a water table near the surface. The canOpy layer of the woodlot is dominated by pin oak (Quercus palustris), sugar maple (Aggg_saccharum), white oak (Quercus Elba), and hackberry (9913i; occidentalis). At the time of the study, the ground cover was frequently dense and dominated by poison ivy (Rhug toxicodendron). The woodlot had not been lumbered or grazed for approximately 35 years (Mahler, 1972). The North Bass Community is the island community of the Ohio replicate. North Bass Island, assigned to Ottawa County, is located (Fig. 2) in Lake Erie about ten and two-thirds kilometers north of the Catawba-Marblehead peninsula and has an approximately two kilometers average diameter (Fig. 2). It is third in a series of three islands extending north from the mainland. The North Bass Community is a woodlot about 13.5 hectares (33.3 acres) in area which extends inland from the western shore of the island and is bordered on three sides by cultivated 4 W Lake Michigan 0 O / . lake South Huron Manitou Island North Lake Bass Erie Figure 1. Map of Great Lakes Region Showing All Communities Studied. North Bass Community a ‘9 0 Lake Erie , V, ’5. f I \\ I ,4 Marblehead “1Ii‘ Community a Figure 2. western Lake Erie Showing the Two Ohio Communities. 75*m Vim | l \ | l I\\{ 1E4: 53 I\ | N | \\ I I T"3\ ' I T-2I \ I A I ' ‘ §l"'"‘I———‘T-_; ————— Figure 3. Marblehead Community and Pattern of Transects. land or mowed grass air-strip (see Figure A). The topography is as regular as that of the Marblehead Community. However, the water table appears to be lower, despite the closer proximity to the lake. Also, the substrate appears more rocky than in the Marblehead Community. The canopy layer is dominated by.American basswood (Tilig_americana), sugar maple (A223 saccharum), and hackberry (Cgltig occidentalis). At the time of the study, the ground cover was frequently sparse, and was not clearly dominated by any specific species of plants. This wood—lot had net been lumbered or grazed for approximately fifty years (Stonerook, 1972). The Leelanau Community is the mainland community of the Michigan replicate. This community is located in the southwestern portion of Leelanau Peninsula, Leelanau County, Michigan (Fig. 5). It is approximately five kilometers north of Empire, Michigan, one and one-third kilometers southwest of Glen Lake, and one and one-fourth kilometers east of Lake Michigan. The site studied is part of a forest system that extends north and south for several kilometers, and is about two-thirds kilometer inland from an active front of Sleeping Bear Dune. The topography consists of a series of low parallel ridges running approximately northawest to south-east. The soil is sandy beneath the organic laden top layer. The forest canopy is dominated primarily by sugar maple (Age; saccharum) and beech (Eggug grandifolia). At the time of the study, the ground cover was consistently dense and was quite diverse, although dominated in some areas by maple seedlings. The site of the study has not been lumbered or grazed for over twenty-five years (Downer, 1972). The South Manitou Community is the island community of the Michigan replicate of the study. South Manitou Island, assigned to Leelanau ll.— Loke Erie T-l Figure 4. <- Air Strip North Bass Community and Pattern of Transects. 10 .7? South Manitou Community I X . Leelanau '\ Community Figure 5. Northern Lake Michigan Showing the Two Michigan Communities. 11 County, is located in Lake Michigan approximately ten and a half kilometers west of Leelanau Peninsula, slightly north of the Leelanau Community study site (Fig. 5). The study site on South Manitou is about two-thirds kilometers west of the southern end of Lake Florence, one and two-thirds kilometers east of the island's western shore and about two- thirds kilometer north of its southern shore. The site is about one and one-third kilometers east of the high stable dunes forming the western shore of the island. The topography consists of parallel ridges, as on Leelanau Peninsula, but with slightly more exaggerated:relief and with a slight general lepe south-eastward, parallel with the ridges. The soil, as on Leelanau Peninsula is sandy beneath the organic top layer. The forest canOpy is dominated by sugar maple (A225 saccharum), beech (Eggug grandifolia), and some yellow birch (Bgtglg_lutg§). At the time of the study the ground cover, as on Leelanau Peninsula, was consistently dense and diverse. This portion had not been logged for approximately forty- five years (Brown, 1972). Transects In each of the communities studied, five belt transects were established. Each transect was 250 meters long and one-half meter wide. Where possible, these transects were established in a pattern parallel to one another. In the Marblehead Community, the five transects were laid out along four different directions to best fit the woodlot (Fig. 3). Transects were placed so that all parts of all transects were at least fifteen meters into the woodlot from its edge. In this community only, some transects intersected. 12 In the North Bass Community, the five transects were established parallel to one another and to the western lake shore (Fig. 4). The transects were separated from one another by fifty meters or more and were at least 150 meters from the lake-shore. In the Leelanau Community, the five transects were established parallel to the active dune front and parallel to each other, and were perpendicular to the low parallel ridges (Fig. 6). The transects were separated by fifty meters or more. Two transects extended near a paved road where the tree canOpy had been cut. In the South Manitou Community the transects were established parallel to the dunes which form the western shore of the island (Fig. 7). This pattern also placed them perpendicular to the parallel ridges and to the slight elevation gradient. From a central eastawest axis, two transects extended northward in a parallel fashion while three transects extended southward in a parallel fashion. Parallel transects were seventy-five meters apart, and the near ends of transects either side of the central axis were separated in a north-south direction by 100 meters. For purposes of data collection, each transect was divided into twenty-five sections, ten meters each, called plots. Fifteen of these plots constituted the basic units of data collection for each transect. Each of these plots was further sub-divided into ten one-meter sub-plots. Of these ten sub-plots, five were randomly chosen in each plot for collection of plant data. Organisms Studied Organisms investigated in this study include the vascular green plants 13 .mpoomnmne mo snapped one hpflsgoo seawaooq .0 onfimfim 1 ''''' J ”a . . It . . he. - n n lllllllllllllll L .0 I may 30. “— a... 0 % ea lllll 61:? ¢ ........... s o. It .4 I‘O’.d 0’ to Iv .67 I III-3 8% a... .3 .. 0Q 00“... 1... 9"... 14 Fkfld r I I I I I I I L--------------------. South Manitou Community and Pattern of Transects. Figure 7. 15 of the herbaceous layer of vegetation, and invertebrate animals collected in association with this layer. Plants with a one and one-quarter centimeter or less basal diameter were defined for the sake of study as herbaceous plants. Plant data were collected in the form of frequency data. The frequency of sub-plots per plot in which a given species occurred was recorded for all plant species present. Actual numbers of individual plants per plot were not used as data because of the difficulty of defining an individual in some plants, such as vines and cluster-forming plants. The sub-plot dimensions were selected on the basis of arfiJxHLexperiment in which the number of species was determined cumulatively in nested quadrats of increasing size, from twelve centimeters by six centimeters to four meters by two meters. The quadrat size at which the variance in cumulative species number, among five sets of nested quadrats, stapped increasing Was one meter by one-half meter. These dimensions were those used for the sub-plot sizes in this study. Table l is a list by community of plant species found. Invertebrate animals investigated in this study were those collected in pit-fall traps placed on the forest floor beneath the herbaceous vegetation. One pitfall trap was placed along each plot of each transect, approximately near the center of the plot. These traps were ten centimeter deep waxed cardboard cups with a ten centimeter bottom diameter. They were placed such that the top edge of each trap was flush with the level of the surrounding soil, and they were filled to a depth of one and oneqhalf centimeter with industrial grade ethylene glycol as a killing agent and a temporary preservative. Traps were left Operative in the forest floor for seventy-two hours. Specimens were later transferred to seventy percent ethanol and then identified to family (in the cases of l6 Table 1. List of Plant Species and Their Occurrences by Community and Visit. SPECIES COMMON NAME M NB L SM §h_u§_ toxicodendron poison ivy X X 0 O _I_fl1_u§_ typhina staghorn sumac X X 0 O gigging aparinus cleavers X X 0 O m boreale northern bedstraw X 0 O O _G_ali_u_m_ triflorum fragrant bedstraw O O X X Elli—“l“. lanceolatum yellow wild licorice O O X 0 Mitchella m partridge berry X X 0 O Lonicera villosa fly honeysuckle X 0 O O Lonicera canadense Canada honeysuckle O O X X m alternifolia alternate leaf dogwood X 0 X O M drummondi rough-leaf dogwood O X 0 0 99313.13 _p. dogwood X 0 O O .U_l_I_n_1_1_s_ £1213 slippery elm X X 0 0 m Virginians ironwood O O X X Crataegus spp. Hawthorne X 0 O O M occidentalis hackberry X X 0 O fill-1E coronaria pear O X X X m serotina black cherry O X 0 O m virginiana choke cherry O X X X M occidentalis black raspberry O X 0 O M31135. strigosus red raspberry O O X 0 Mug 5 . black berry X 0 O O Rps_a_ setigera prairie rose X 0 O X Ro_s_a_ palustris swamp rose X 0 O X _R_o§§ carolina pasture rose X 0 O O Table 1. Continued. l7 SPECIES COMMON NAME M L SM Role. _2. rose 0 O O Duchesnea ipdigg Indian strawberry X 0 O Potentilla implex cinquefoil X 0 O Fragaria virginiana canadensis common strawberry O X 0 923m canadense white avens X 0 O Sanicula canadensis black snake root X X 0 .Ribg§,americanum black current X 0 O Ribgg'cygosbati pasture goosberry O X X Juglans.nig3g black walnut X 0 O Qggyg_cordiformis bitternut hickory X 0 O ngyg'gyglig pignut hickory X 0 O Qggyg.gy§t§ shag-bark O O O Fagus grandifolia beech O X X Quercus £123 white oak X 0 O Quercus palustris pin oak X 0 O Quercus velutina black oak O X 0 Fraxinus pennsylvanicus green ash X X X subintegerrina Xanthophyllum americanum northern prickly ash X 0 O ‘Aggg saccharum sugar maple O X X A_ce_I_'_ W black maple O X 0 A22; spicatum mountain maple O O X Bgtulg_lutgg yellow birch O X X Bgtulg papyrifera white birch O X 0 Carpinus caroliniana blue beech, ironwood X 0 O 18 Table 1. Continued. SPECIES COMMON NAME- M. NB L SM lili§,americana American basswood X X X 0 Sambucus canadensis common elder X X X X Viburnum acerifolium maple-leaf Viburnum O O X X , Ailanthus altissima tree of heaven 0 X 0 O 'Vitigrpalmata cat grape X X X 0 Campsis radicans trumpet creeper X X 0 O Menispermum canadense canada moonseed X 0 O O Solanum dulcamara bitter nightshade O X 0 O Parthenocissus guinguefolia Virginia creeper X X 0 O Solidagoigp. goldenrod X X X X Hydrophyllum virginianum Virginia waterleaf X X X X Arisaema triphyllum jack-in-the-pulpit X 0 X X Arisaema dracontium green dragon X 0 O O Smil§x_ecirrhata carrion flower X X 0 O Lysimachia ciliata fringed loosestrife X 0 O O Boehmeria cylindrica bog-hemp X 0 O O Leonurus cardiaca motherwort O X 0 O Arctium mingg common burdock O ' X X X Gratiola guggg hedge hySSOp X X 0 O Osmorhiza longistylis sweet cicily O X X X Geranium robertianum Herb Robert X X 0 X Mitella diphylla mitrewort X X X X lAggbi§,perstellata Va. shortti rock cress O X 0 O ngpanula americana tall bellflower O X 0 O Agglig nudicaulis sarsaparilla O O X X l9 Table 1. Continued. SPECIES COMMON NAME M NB L SM Caulthyllum thalictroides blue cohosh O O X X Thalictrum dioicum early meadow-rue O O X X £939.29: 113 white baneberry O O X X Po onum virginianum Virginia knotweed X 0 O O Impatiens biflora jewel weed X 0 O O Impatiens gp. touch-me-not X 0 O 0 PM leptostachya lopseed X X 0 O Chenopodium gm lamb's quarters O X 0 O Solanum m black nightshade 0 X 0 O Heracleum lanatum cow parsnip O O O X m canadensis Canada violet X X X X [Lola pubescens downy yellow violet X X X X lie—13 renifolia kidney-leaved violet X 0 O O 1311;; incogpita large-leaved violet X X 0 0 Ma eriocarpa smooth yellow violet X X X X _Vi_ol§_ conspersa American dog-violet O O X X M selkirkii great-spurred Violet O O O X ME. _p. violet O O O X li_oLa_ pgilionacea common blue violet X X 0 O m trifolia gold thread X X 0 O M _p. sorrel X X 0 O Dentaria dyphylla toothwort O 0 X X Sanguinaria canadensis bloodroot O O X X Hepatica acutiloba sharp lobed hepatica O O X X m s . shinleaf O O X 0 20 Table 1. Continued. SPECIES COMMON NAME M NB L SM Habenaria orbiculata round-leaved orchid O O X 0 Epipactus latifolia helleborrine O O X 0 Trillium grandiflorum white trillium X 0 X X Trillium erectum purple trillium O O X 0 Maianthemum canadense Canada mayflower X 0 X X Polygonatum pubescens Solomon's seal 0 O O X Polygonatum caniliculatum Great Solomon's Seal 0 O X 0 Polygonatum biflorum Solomon's-seal O O X X Polygonatum multiflorum Great Solomon's-seal O O X X Smilacina racemosa False Solomon's seal 0 O X X Streptoius w rose twisted-stalk O O X X Uvularia grandiflora large-flowered bellwort O O X 0 _A_l.l_iu_m_ tricoccum wild leek O X X X _A_l_'_L_i_lim_ canadense wild onion 0 X 0 O m _p. cattail X 0 O O m plantiqinea broad-leaved sedge O O X X Car—ex _p. sedge X X 0 0 Egg; convolute sedge X X X X §a_regc_ _p. sedge O O X X M £2. sedge X 0 O O gag; g2. sedge X 0 O O M s . rush X 0 O O Leerzia ogyzoides rice cut grass X X 0 O 20 sis vacemosa rice grass 0 O O X m virginicus virginal wild rye O X 0 O 21 Table 1. Continued. SPECIES COMMON NAME M NB L SM _Mi_.1_:_'|._um effusum grass 0 O X 0 ‘Qign§_arundinacea grass X 0 O O Taxgg_canadensis American yew O O O X Dryopteris g2. Austriaca* sword fern O O X X Adianturn pgdatum maiden hair fern O O X X Botgychium.virginiana rattlesnake fern O O X X plus distinct but unidentified species X X X X Columns M, NB, L, and SM indicate the communities: Ohio mainland, Ohio island,‘Michigan mainland, Michigan island, respectively._ Under thesef~ columns is placed an X if the taxon was encountered, an Osif not. 22 insects, spiders, harvestmen, idopods and snails) and to order (pseudo- scorpions, myriapods and annelids). Table 2 is a list by community of taxa of invertebrate animals found. Since plants and animals are studied here at different taxonomic levels, certain assumptions are made in order to compare their insular or seasonal differences and to compare correlations between them and other parameters. These assumptions are discussed in the Analyses section of this report. If these assumptions hold, comparative results are not altered by this difference in taxonomic levels. Sampling'V;§it§ Data were collected from each community'three times during the summer of 1972. Each visit lasted about a week and was separated from the previous or next visit by about a month. The schedule for these visits is shown in Table 3. Invertebrates were collected during Visit I and Visit II. Plant data were collected during all three visits, but were complete only for Visit III. Incomplete plant data of Visits I and II, compared sub-plot by sub-plot, were highly consistent between these visits and highly consistent with analogous data of Visit III, indicating that the sampled plant communities were constant throughout the period of study. Spring ephemerals had apparently disappeared before the beginning of the study. All subsequent analyses included animal data from Visits I and II separately and assumed plant data of Vitis III to be identical for the first two visits. 23 Table 2. List of Invertebrate Animal Taxa, Food Habits Ranking, and Occurrences by Community and Visit. TAXA B T D MI MII NBI NBII LI LII SMI SMII INSECTA Thysanura Machilidae 2 D 3 O O X 0 O O O O Ephemeroptera Caenidae 3 H O O O X 0 O O O Orthoptera Tettigoniidae 3 H 2 X 0 X 0 O O O O Gryllacrididae 3 H 2 X X X X X X X X Gryllidae 3 H 2 X X X X X X 0 X Blatidae 3 D l O X 0 O O O O O ThysanOptera Thripidae l H 3 O O O O O X X 0 Ploethripidae 1 D 3 O O O O X X 0 X Hemiptera Miridae 1 H 3 X X X X 0 X X X Nabidae l C 3 X X 0 X 0 O O O Reduviidae 2 C 3 O X X 0 O O O O Tingidae 2 H 3 X 0 X X 0 O O O Aradidae 2 D 3 O O O O O X 0 O Pentatomidae 2 C 2 O X X 0 X X 0 O Homoptera Membracidae l H 3 O X 0 X X 0 O O Cicadellidae l H 3 X X X X X X X X CerCOpidae 1 H 3 X X 0 X 0 O O O Fulgoridae l H 3 O X 0 O O O O O Aleyrodidae 2 H 3 O X 0 O O X 0 O Aphidae 1 H 3 X X X X 0 X 0 O NeurOptera Chrysopidae 2 C 3 O O O O O O O X ColeOptera Carabidae 2 C 3 X X X X X X X X Histeridae l C 3 X 0 O O O O O O Leptinidae l C 3 O O O O O O X 0 Ptiliidae 2 H 3 O O O O O O X 0 Leiodidae 2 D 3 O X 0 O O O O O Leptodiridae 3 D 3 O O O O O O O X Silphidae 2 D 2 O O O O O O X X Scaphidiidae 3 D 3 O O O O O O X 0 Staphilinidae 3 C l X X X X X X X X OrthOperidae 2 D 3 O O O O O O X 0 Cantharidae 3 C 3 O O O O X 0 O X Lampyridae 2 C 3 O X 0 O X 0 O O Lycidae 2 D 2 O O O O O O X 0 Cisidae 3 D 3 O O X 0 O O O O Cleridae l C 3 O O O O O O X 0 Elateridae 2 H 2 X X 0 X 0 X 0 O Eucnemidae 2 C 3 O O O O O X 0 O Buprestidae 2 D 3 O O X 0 O O O 0 24 Table 2. Continued TAXA B T D MI MII NBI NBII LI LII SMI SMII Coleoptera, cont. Dascillidae Byrrhidae Cucujidae Nitidulidae Lathridiidae Endomychidae Anthicidae Pyrochroidae Tenebrionidae Lagriidae Anobiidae Scarabaeidae Chrysomelidae Bruchidae- Curculionidae Scolytidae Mecoptera Bittacidae Trichoptera omit Lepidoptera Noctuidae Liparidae Pyralidae Tortricidae Gelechiidae Diptera Tipulidae Psychodidae Culicidae CeratOpogonidae Chironomidae Anisopodidae Mycetophylidae Sciaridae Cecidomyiidae Xylomyidae Stratiomyiidae Rhagionidae Asilidae Empididae Dolychopodidae Phoridae Syrphidae ConOpidae Otididae Sciomyzidae Lauxaniidae Piophilidae O B c.- b.) I—‘NHI—JUNUUNNNHI—‘N I-' DmmmuuumUmUUmOH-m N UUUUUUNWNWUUUU b) ONONNOONONOONOOO ONONNOOOOOOONNOO NNNNOOONOONONOOO ONONNNNOONNONOON ONOONOOONNOONOOO ONNONOONONOOfiOOO OOOONOOOOOOONONO OOOONOOOONONNOOO O O>< O NO >40 00 CO CO eJnoedkora OOOON OOOOO oxoxo NOOOO OOOOO OONOO NOOOO O 5 WUUUUNUUUUUN UUUUUNNUN UUUUU ONNOOONNNONOONNNONNOON NNOOO NNI—‘NNI—‘NNNUUW HquUI—IUNN OOONOONNNONOOONNONONON OOOONONNOOONONNOONNOOO xxoxooxooooxxoxooxxoox OOOOONNOONOOONNOONNNNO NNNOOONNNOOOOONNONOOONI ONNOOONNNOOOONNONNNNNN NNNOOONNNOONONNNONNOOO 25 Table 2. Continued. TAXA B T D MI MII NBI NBII LI LII SMI SMII Diptera cont. Lonchaeidae 3 D 2 O O X 0 O O O O Sphaeroceridae 3 D 3 X X X X 0 X X X Drosophilidae 2 D 3 O O X X X X X X ChlorOpidae l H 3 X X X X X X X X Agromyzidae l H 3 O O O O O X X 0 Heliomyzidae 3 D 3 O O X 0 O O X X Anthomyiidae 2 H 3 O O O O O X 0 O Muscidae 3 D 3 X X 0 X 0 X X X Calliphoridae 3 D 3 X 0 O O O O O O Tachinidae l C 3 O X 0 X X 0 X 0 Siphonaptera LeptOpsyllidae l C 2 O O X 0 O O 0 O HymenOptera Braconidoidea l C 3 X X X X X X X X Ichneumonidae l C 3 X X X 0 O X 0 O Mymaridae l C 3 X 0 O O O O X 0 Eulophidae 2 C 3 O O O X X X 0 O Encritidae l C 3 O O O O O O X 0 Eupelmidae 3 C 3 O O X X 0 O O O Euryomidae 2 H 3 O O X 0 O O O O Chalcidae l C 3 O O O O O X X 0 Cynipidae l H 3 O O O X 0 X X X Roproniidae l C 3 O X 0 X 0 O O O Proctotrupidae l C 3 O O X 0 X 0 O O Ceraphronidae 2 C 3 X X X X X X 0 X Diapriidae 1 C 3 X X X X X X 0 X Scelionidae 1 C 3 X X X X X X 0 O Platygasteridae l C 3 O O X X 0 O O O Dryinidae 1 C 3 X X 0 X X 0 O O Formicidae 3 C 1 X X X X X X X X Sphecidae 1 C 3 X 0 X 0 O O O O Halictidae 2 H 3 O O O X 0 O O O ARACHNIDA Chelonethida 3 C 3 X 0 X 0 O O O O Opiliones Phalangiidae 3 C l X X X X X X X X Araneae Dyctinidae 2 C 3 O O O O O O O X Theridiidae 2 C 3 X 0 O O O O O O Lyniphiidae 2 C 3 X 0 X 0 O X X X Micryphantidae 2 C 3 O O X X X X X 0 Araneidae 2 C 3 X X X X X X 0 O Agelenidae 3 C 3 O O O O O O X X Hahniidae 3 C 3 X 0 X X X X X X Lycosidae 3 C 3 X X X X X X X 0 Gnaphosidae 3 C 3 X X X X X X 0 X Clubionidae 3 C 3 X X 0 O O O O O ThomiSidae 3 C 3 X X X 0 X X 0 X Salticidae 3 C 3 X X X 0 X 0 X 0 26 Table 2. Continued. TAXA B T D MI MII NBI NBII LI LII SMI SMII CRUSTACEA Isopoda Armadillidiidae Porcelionidae Trichoniscidae DIPLOPODA Polydesmida Julida CHILOPODA watobiida Geophilida OLIGOCHAETA OpisthOpora Hirudinea PULMONATA Stylomatophora Cionellidae Pupillidae Succineidae Endodontidae Limacidae Zonitidae Polygiridae UU Ubtt: ox o><><>< o>< o>c>c> >400 ON NO 00 NN ON NO 00 NO ON NN ON NO ON NO NU Lou) bob.) MUM) 00 ON NN NN NNO GO ON ON NN ON ON UK.) Uh) Uh.) UUN OO 00 \okotokokozo\o awesomeness totoLototototo ><><><><>c>>c>c>c> <>><><3> c>><><><><> ><><><><><>4><>oA Oflnmone mofioomm mom mQOfipoonzoo cannonp m macapoosnoo cannonp pncmoamoa mcnfiq newcomm 3H mmfiommm pzomonmon mmHoaflo "V A“ 30 Table #. Three Food4Web Components, Rows, and Three Kinds of Measurements, Columns, and the Resulting Nine Measurements. RELATIVE NUMBER SIZES DISTINCTNESS TAXA NUMBER OF EVENNESS RESOURCE TAXA OF TAXA OVERLAP _ TROPHIC RESOURCE RESOURCE SELECTIVITY CONNECTIONS BREAD'IH BREADTH OF FEEDING TROPHIC NUMBER OF EVENNESS OF DISTINCTNESS OF LEVELS TROPHIC LEVELS TROPHIC LEVELS TROPHIC LEVELS this 31 Resource Breadth (the relative prOportions of those food sources used by the consumer taxa) Selectivity of Feeding (deviation of feeding by consumer taxa from random strategy is a measure of distinguishability of tr0phic connections per taxon) Number of Trophic Levels Evenness of TrOphic Levels (relative number of taxa, or of individual organisms, per trophic level) Distinctness of Trophic Levels (the degree to which consumer taxa selectively distinguish lower trophic levels in feeding) Five of the above ecological parameters investigated more fully in study, and already mentioned in the Introduction of this report, are Number of Taxa Evenness of Taxa Resource Breadth Evenness of Trophic Levels Distinctness of Trophic Levels The two parts of Resource Breadth, above, have been treated in combined fashion by Colwell and Futuyma (1971) and by Pielou (1972) and are also combined in the present study. Of the three parameters not investigated in this study, Number of Trophic Levels does not differ among the communities studied, and the logistics and time involved in measuring Resource Overlap and Selectivity of Feeding were incompatible with the resources of this study. , '- .- nrv'm ‘N'l‘ u '1 . ‘ —" 1:. ggpected E In ac explores, of insula: study. T Nrner of eXp-ain 1 of islan< supporte: factors elevatio source p resource of Speci Taylor, taXa the in the 1 Fex the Sum: nube r s 32 Expected Effects of Insularity and Seasonality In accordance with the first goal of this study, this section explores, by review of the pertinent literature, the expected influences of insularity and seasonality on the five parameters investigated in this study. The five parameters are treated individually below. Number of Taxa -- MacArthur and Wilson (1963) have proposed a model to explain the paucity of species on islands relative to mainlands in terms of island area and distance from mainland. This model has been strongly supported by experimental test (Wilson and Simberloff, 1969). Other factors also fOund to affect Species numbers on islands are island elevation (birds, Hamilton and Armstrong, 1965), densities of mainland source populations (small mammals, MCPherson and Krull, 1972), number of resource (plant) species (birds, Power, 1972) and evolutionary adaptation of species to the island environment and to each other (ants, Wilson and Taylor, 1967). In all cases, however, islands are expected to have fewer taxa than adjacent and similar mainland areas, and the same is predicted in the present study. Few studies indicate a consistent change in number of taxa during the summer growing season. A study by Hurd‘gtugl. (1971) suggests that numbers of herbivore insect taxa increase during this season in unused hay fields. Pulliam gt_§l, (1968) indicate that number of spider species increases from five to twelve in a field of millet between July 9 and September 2, 1966. Thus it may be predicted, although with little confidence, that in this study Number of Taxa will increase during the period of study. '1‘ ' I'~ I Evenness of / selective lo: Evenness of '. there are ma: approach equ predicted on will be larg Pulliar various art} Although the homcpteran : v61' time. increased 1;: increase in Predicted, . Study, will W Wider Varie 1962; Grant MOI'Se, 197] termed em] Rlelefs ar is an earls mine'Cion 33 Evenness of Taxa - The analyses of Preston (1962, a, b) suggest the selective loss of rare species on islands, which would result in increased Evenness of Taxa on islands. MacArthur (1969a) has suggested that where there are many species, as in the trepics, relative abundances would approach equality, and therefore evenness would be high. It can be predicted only'tentatively in the present study that Evenness of Taxa will be larger on islands than on mainlands. Pulliam, 0dum.and Barrett (1968) measured Evenness of Taxa of various arthrOpods during the growing season in a field of millet. Although the evenness of spider species tended to increase, that of homopteran species and carnivorous hemipteran species tended to decrease over time. As discussed above, Evenness of Taxa may decrease with increased Number of Taxa on the mainland, and we have predicted an increase in Number of Taxa during the period of study. It may again be predicted, with little confidence, that Evenness of Taxa, in the present study, will decrease during the summer growing season. Resource Breadth —- Several investigators have shown that birds utilize a wider variety of resources on islands than on mainlands (Crowell, 1961, 1962; Grant, 1966, 1968; Sheppard, K10pfer and Oelke, 1968; Keast, 1970; Mbrse, 1971; MacArthur, Diamond and Karr, 1972). This phenomenon has been termed evolutionary (or ecological) release (MacArthur and Wilson, 1967). Ricklefs and Cox (1972) suggest that this expansion of Resource Breadth is an earLy part of a more general cycle of invasion, adaptation, and extinction of taxa on islands. In addition, Williams (1969) has shown that colonizing anoline lizards tend to be of "versatile" species and may undergo ecological release. These reports, although dealing primarily 3# with birds, suggest the prediction that Resource Breadth of invertebrates in the present study will be higher on islands than on mainlands. The literature reviewed provides no evidence regarding any patterns of differences in Resource Breadth during any season. Evenness of Trophic Levels -- Again, the literature reviewed provides no evidence regarding any patterns of differences in Evenness of Trophic Levels between islands and mainlands. In a study of insect tr0phic diversity in salt marsh communities, Cameron (1972) has shown that, during a Single year, the diversities of herbivores, saprovores and predators are more nearly equal in June and July than in August and September. These results suggest the prediction, in the present study, that Evenness of Trophic Levels will decrease during the period of study. Distinctness of Trophic Levels -- Due to the total inadequacy of the reviewed literature regarding Distinctness of Trophic Levels, no §_priori predictions of insular or seasonal effects on this parameter may be made, but must await the outcome of the present study. Models g Relationships among Parameters Free-Bodngbdel Concept -- As a first step in inferring expected relationships among these parameters, each of them was investigated, independently, by survey of the ecological literature. The infOrmation from.this survey was collated into a diagramatic compartment model for each parameter independently. Each such model describes expected cause- effect relationships between one of the five parameters and any other ecological parameters that presumably'influences it. These individual 35 models are analogous to free-body'models of'systems science (see Caswell, Koenig, Resh and Ross, 1972), and are termed free-body models in the present study. Development of each free-body model independently of the others avoids the pitfall of defining relationships among those parameters studied as a necessary and sufficient set of relationships, even though parameters external to the system may exert an important influence on one or more of the parameters of the system. These five free—body models can be combined in a diagramatic compartment model of expected relationships among the parameters as a system, which is herein called a system model. The information obtained from the literature survey was sometimes clearly and concisely presented in the original source. At other times it was obtained by examination of data or conclusions of the source article from the viewpoint of the present study. This information was used in this report if it logically led to a hypothesis of relationship between any two ecological parameters and linked either of them directly or indirectly to any of the five parameters primarily investigated in this report. Information surveyed and found unsuitable is not referenced. The following paragraphs are descriptions of the free-body models and the system model. References are not included in these descriptions for the sake of clarity. Rather, they are listed in Tables 5-9. In each of these tables, the parameters listed on the left, heading the rows, are hypothesized to directly affect those parameters at the top of the table, heading the columns. The cell entries list the references used to hypothesize this effect. Statements in these descriptions must be viewed as reasonable hypotheses, not as clearly shown facts. Even though‘ the writing appears factual, it does so for the sake of brevity only. Although parameters of 36 importance may have been omitted from these free-body models, and they may contain some duplication, they represent the most parsimonious and complete models that this author has been able to infer from the literature reviewed. Number of Taxa -- Figure 9 illustrates the schematic compartment model that has been developed as the free-body model for Number of Taxa. References are listed in Table 5. This model is relatively complex: It consists of five hierarchic levels of a total of thirtyasix parameters and includes four instances of feed-back. This complexity'mey'be in part due to the large volume of literature reviewed: seventy-two published articles are used to generate this model. The four parameters most directly related to Number of Taxa (see Figure 9) each represent some general effect achieved by any of a variety of mechanisms. These mechanisms plus other "general-effect" parameters constitute the parameters Of the next level removed in this hierarchy from Number of Taxa. This pattern continues until the five levels of the hierarchy are completed. Beginning with the first parameter in Figure 9 affecting Number of Taxa, the NUmber of Taxa estimated by a sample is directly proportional to the size of the sample, is increased by inclusion of ecotones in the sample or by lack of discreteness of communities sampled and by relatively great differences among communities sampled. Number of Taxa will be decreased by Extinction and increased by speciation and by Immigration. The probability of extinction of species in a community is increased if the Minimum Density needed for reproduction by the constituent populations is increased, is increased by periodic small scale perturbations of the environment, is increased by Competition, but is decreased by Adaptation. 37 dance .Ho 3352 me Home: hvomlooam .m 0.3.me imatnefi 20:30.... U 20.25.22. \EESuawE 13.2mm vazuzzmzsamL ;\ \.\ 20:59.. 95.. :25 2523 20:58am \ \ \Eizomfim 20 3 _/ \\ 20:35: \\ \., . .\ . >335 02.8% C. mmmzzmb mg 92.6.: 4/ / ./ 3:232. 20:53; 6.523 593303.. 662.5on / x. 329229.35 <5: 6 E2530... 6 >555 \. :56... 839m. 4: I III 20:5..28 f 32:5 5:33.42 3.59.: a it 20.5263 ceasefire, sees. \ \ 20:335.. CELEBS! 2:1.» 32:03.. “.0 E5920 \\ a... \53 3 $932 W ¢O_>(Iwn 4(_U0m 38 moma .eomaaz one apnoeaoaz mama .enea ace BORE .Efitfiez omma zebmm use nomznoh soapmamaEEH mmma .eoaesoaaasm coma .macmam mmcma .eseaeaoaz omma .:c>em pom zomcgom wwma .>oaopom eoma .mmaao new flamenco noapmaocom moma .maaaaae; mmma .comaaz new aseptaoez eema .mmaao one HHonnoo nowponapxm moma .memaao mead .aoseoSOm one seesaw Nmma .mmamoao.z omma co>mm use zomonow mama .aoaameaz «convex coma .eeepaae meaamaem 28252: A . zoaaaaemmm zoaaezaaam 33 mo mmmzez made me aonapz Ho Hopes abomuooam op puma page maoposmaem Heoamoaoom msapwacm meoeoaemom Ho magma .n capes 39 oama .cmcma .wmcma .mcma .eacceaoaa oama .cmeoa coma .cseao coma .>oaovom mcma .aamsomo mama .zoaosmo aama .mmaem mama .mmow one aamHod< mama .oaozme mama .ogoamz oama .mmosoa oama .gOmgeoe mama .comaaazoCe eaaaz .camoaasm aama .aaoe one mamaosa cmoma .eocmotm acma .aaaaaom mama .aaama draw Somhmflmoz oama .ce>mm one somehow soapapomsoo zoapmnaopacm hpamsom Edgazaz mama .xoo one mammaaoam woma .ppcaamm Abodaapnoov one ezoo .emaaasm coaumamasea zo Hmamo HZZH ZOHB aama .emeocmca moma .aoooozom mama .xoe one mammaxoam moma .ccomsem one aseo .eeaaaam a.mama .mopmom coma .eoaoam coma .oeaam mcma .sseo acma .eomaaz one adgpamm use some: anemonow moma .msoapmea< pom sopaaeem acma .gomaaz one adopaoacpom ammaommam ccoscmaaoecmm cmaam oacmemo zoaemmwazza zoae mwma anonoonom mama .nou one maaeaaoam coma .onaem mcma .aseo emoma .ocma .Sfiefiez coma .ceame moma .aamzoao aama .moxam axes mama .oaocaa mo monasz mama .asccaaomz one neocaaomz agitate: £3243: mama .namm use macadam .3533: coma .emaoa mcma .aaosomo omam meom noabmsom acma .maaaaom aeaoom azeamaamaamm Ema oaamzme zoaaamama zoaaaammaoo Eamzma ESE / Ill U05£H¢COU . W QHQNH 43 omma .aooaaovzm> Ncma .amoe acma .ccma .mggmam amma mnemamom coma .mgamm moma .mamm mam mmgomz oama .mmggmmmga coma .gmama mcma .comgagogsm mama .mgaSma one oaozpmem gmama .mgaSmam acma .mmaao pom aaonooo mama .zoaoamo coma .ggmgmmm mam memo .amaaamm coma .mmgmmm mam gaggmmmmz coma .amggmammz acma .mmggmmmmz one aoaooam coma .mmgmmm mam gaggammmz coma .nsggmmmga acma .mmggmammz com aommoam mcma .eomgagmgsm mgaaagmaammm condomom mmaao>o meadomom Apossapcoov moopdomom Ho mgaaagamasag Hzmzmmaamdymm shame Daemzmo ZOHHm ao>oa cannons mooafiomom mo honmpmsoo npwmoam connomom Bzmzmmaammsmm emamo Daemzmo ZOHB¢HAoaovom moma .mmmmmmm one saxoonoam coma .mxamam acma .acma .mcaboa one asnpam aoboa canaoae 92mzmmaam¢Hmm shame Daemzmo zoae¢HmHD MBHaHm< mumbommm MNHm Mmom coscaacoo .c magma 47 acma mapaozmaamm mam meammgagz aama .gcgogama ccma .gmmamgc gmama .mmgaom mcma .acma .ccma .mgmmam aama anomacom mcma .memamo mama .somaopom one mnmbm agoovnaz mcma .aamc mam amgmmz ccma .mcma .Qfimamaa oama .mmmsoa acma .egoa oama .emsmc one songow coma .nmsmc cam comm: anomnsoc oama .eomggom mcma .eomgagmgsc coma .mcoaamea¢ cam mmgaaemc gmama .cmaSmaa amaaagmgmaemam mama .gmazc amagmmc cccceocmc ao cagmmcc aaaaacaaaaam cauccocmc ac cmac acoc aczaaczoc coccoccc coccoccc aaaaaca caaac emanacgoe .c magma coma .mgemaa ccma .mmamm mcma .mgmaao acma .amamcmmz mama .mamm one ccoacam .msgtammz oama .asggaaomz gmama .cnaamaa mama .goamamc acma .mgemam mama .meoamz mama .amzom coma acobmm one comm: .som:£0h mama .mmzmm mcma .mgmamo mama .comaopom one msm>m .nooonsz gmcma .mmcma .a5hpamemz coma . mad Mm .gogmmamc aama .mmgmc mama .mamgaa amasagoecogm am mgaaagmgc mmsgagma coapmpamaooam moundomom Mo nebepz mmumbommm mo MUZdezco HBHAHm¢AH<>¢ mumbommm mamdmmm mumbommm mmumbommm ho MBHAHmHmH>HQ mNam Noam omegagnoe .c magma 49 acma .Gomaaz ocmaasspaoaovom Aa canoe oomv Aa oaflme oomv Aa canes oomv acma .spaosmaacm com opamoaagz com asp acma .ccma .mgmmam coma .mmamm mama .aama com vooamaa .msggmammz oama .ccma .msggmmomz aama .noanwdm202_vcm aomsaaaoz .Uadm haaamnommcm agaaagmaamsm condemem awoaeaam mnavooh soapmaam> caaozou mcma .mmamo mam aammmoc agasagmmcmam cccccoccc mo cacmccc aaaaacmaaaam cceceoccc ao cmac acoc Bzmaczoo compo ccc cacao cam aaaaaca caaac mmmmameoo .c magma 5O mcma .mmmcgmc egm camcocoac ccma .mgmmam acma .mcma .mmasma p:c.asnaa< mumbommm mmumbommm mo MBHAHmHmH>HQ MNam Nmom vodnapsoo .c magma 51 Adaptation may also lead to Speciation, or Speciation may come about by Genetic Drift. Establishment may be considered a necessary component of Immigration, while enhanced dispersal capabilities of species may increase Immigration. Minimum Density needed for reproduction of a pOpulation is decreased by gregarious Social Behavior and, for widely ranging rare, large, or competing species, may not be reached if area is restricted. Review of the literature suggests that Competition is most often used as a "general-effect" parameter, rather than a clearly defined biological parameter. It is used to categorize a variety of non-predaceous inter-species or intra-species interactions and to generalize the effects of these interactions on Extinction or Adaptation of species. The literature indicates that Competition is influenced by a large number of other parameters, (This report categorizes nine of them.) but whether all mechanisms of competition are affected by all these parameters is unclear. Competition is presumably increased by an increase in Number of Taxa. This effect represents part of a negative feedaback lOOp since increase in Competition in this model leads to an increase in probability of Extinction, which leads to a decrease in Number of Taxa. Increase in Number of Taxa may also lead to a decrease in Body Size, which may act as a positive feed-back lOOp, ultimately permitting the larger Number of Taxa. A finer Divisibility of Resources may reduce Competition, as would increased Resource Overlap and increased Resource Availability. An increase in Resource Breadth may increase Competition, while an increase in Competition may decrease Resource Breadth, forming a negative feedaback loop. As Constancy of Resource supply decreases, Competition may also be expected to decrease. Under conditions of low Trophic Level Evenness 52 (proportion of predator taxa is small), or when Predation in general is reduced, then Competition among the taxa preyed upon may be high. This increase in Competition is likely to reduce the Number Of Taxa preyed upon, thus increasing Evenness of Trophic Levels and, relatively, increasing Predation. This set of relationships forms another negative feed-back loop. Adaptation, somewhat like Competition, is used to categorize a variety of evolutionary processes which may reduce Competition, directly reduce the probability of Extinction, or lead to Speciation. These changes may be guided by the selective forces of Competition, Predation or by Temporal Predictability. Genetic Drift may contribute significantly to Speciation in the absence of Seasonal rythm coupled with strong Biotic Isolation. Divisibility of Resources may be limited by available Area, by Spatial Predictability (heterogeneity) or by Number of Resources. Influences on Resource Breadth are discussed in the sub-section dealing with that parameter. Evenness of Taxa -- The free-body model for Evenness of Taxa (Figure 10) is simpler than that for Number of Taxa. It consists of only three hierarchic levels of a total of fourteen parameters and includes no Obvious cases of feed-back. References are listed in Table 6. Evenness of Taxa can be under-estimated by samples so large that they compound different habitats or over-estimated by small random samples. Evenness of Taxa per unit area is decreased by any tendency of organisms to be distributed spatially in a Clumped manner, and Clumping can apparently be determined by Resource Patch Size and by gregarious 53 .389 me mccnzokm .ao ammo: hpomlooam .oa magmam >2>:.UDQO~E meZDSSOU “.0 wow Z._.UZ_._.m_o >._..>_.—U:00¢d “—0 iguana >.:.=m<.=<>< mU¢DOmm¢ months ZO_h<._._n:Um~E 20:.(Owam A96. cc .5353: Low .0108 .33 (XS. “—0 «32:2 mmUaDOmma m0 mmmZZm>w ¢O_><_._wm ._<. U0m1~t t 025230 mNB :95». mUmDOmmcla menu-+52 023mg IV wmmZZm>w m .gcocasz oama mao>om vac mommzoc aama .aaoc mam mamamsm mmcma .3523: mcma .eogmmamc nmcma .noamoam mama ac aonasz mama .momaopom new mnc>m .ooovasz moOAdOmom nmama .wcaSoam mo mmozno>m moma .aoemaa mcma .gogmaamc .cmamasac oama .no>cm cam oomogoc avenue: coma .mmggmmmmz cmaammmc hpaaancaae>< meagasao mmozco>m mama condomom .cxma Ho muonoobm mo ammo: hoomlooam op cmoa acma maoaoemamm amoamoaoom weapcaom moocoaoamm Ho mange .o canoe 55 Am canoe oomv Ac canoe oomv Am magma oomv Am mange oomv mama .mmgmaac mama .amgoaam cmcma .zopmoam coma .cgmgamc am 58 .amaaasa mmama .cmaamaa mmcma .amgtammz 33308on agasagmsmoma am acaaagmac coopaama :oaamaamaooam aoabmSom amaoom omam soamm coasomom moapaodeaoo Mo mmozpcnapcam agaaagmaamsm condomom mgaaagmaami meadomom mcaossao mmozoo>m mama cmemaggoo .c magma 56 Social Behavior and Opposed by Territoriality. Two studies show or suggest a correlation between Evenness of Taxa and the Evenness of their Resources. AS Number of Taxa increases, new taxa are apparently more likely to be rare than common, decreasing Evenness of Taxa. On the other hand, the large Number of Taxa in the tropics may require that they all be about equally distributed, increasing Evenness of Taxa. Several authors indicate a positive correlation between Evenness of Taxa and either Number of Taxa or diversity. Another finds no such correlation, and still another claims that published data only indicate, biologically, a decrease in the variance of Evenness Of Taxa as Number of Taxa increases. It appears logical, at this point, to hypothesize a second order relationship between Evenness of Taxa and Number of Taxa, such that, beginning with no taxa, the first few taxa added will be the more common ones, and Evenness will be high. Further taxa added will increasingly be the more rare ones, and Evenness will decrease. But at some point, addition of further taxa will require decreased abundance of the more common taxa already present, and from this point forward addition of taxa will be accompanied by increase in Evenness. For further influences on Number of Taxa, see the sub-section and model for that parameter. Predation by the starfish, Acanthaster, can markedly increase the Evenness of prey Taxa, coral species, presumably by preventing competitive interactions among those taxa. However in another study, Evenness of plant taxa is higher at low grazing pressure than at high grazing pressure. It appears reasonable to suggest that as intensity of predation increases 57 from zero Evenness also increases, until a threshold tolerance of some taxa is reached, and Evenness will then begin to decline. Evenness of Taxa may be high in the tropics where Resource Availability is great, and Evenness of Costa Rican rodent species was low where food was limited. Others suggest that high Evenness indicates food limitation of the taxa involved. It is possible that Resource Availability influences Evenness of Taxa indirectly only by first influencing Number of Taxa. Evenness of Taxa, computed from communities which are not Distinct, but rather continuous, may be relatively low. Resource Breadth -- The free-body model for Resource Breadth (Fig. 11) is also relatively simple, consisting of three hierarchic levels and a total of thirteen parameters. It does, however, include two feedback loops. References are listed in Table 7. In general, Resource Breadth should increase as Constancy of Resources decreases, and the latter may be determined by Temporal Predictability and Seasonality. An increase in Competition, as discussed previously, may decrease Resource Breadth, while an increase in Resource Breadth may increase Competition, forming a negative feedback 100p. For discussion of other influences on Competition, see discussion of free-body model for Number of Taxa. Feeding Strategy can apparently influence Resource Breadth in various ways. MacArthur has hypothesized that searchers should be relatively more generalized than pursuers, and thus Should exhibit greater Resource Breadth. 58 .aficcoam @93ch mo ammo: afiemuooam ZOF U_hw2m0 ZOFUDOOmm IN 20.53005. ”.0 t...=m<._.m >._._.=o(.=<>< uU¢30mm¢ MOP—23.. 209—3228“; Swag 02.0mm“— flra A93. we 35:: a3 .0305 03V 53 “.0 «3332 v 29:5..on \ rpm—330338.— 230..sz V . Tag—389. “.0 62(th >2..(20m< coapaaoflfioo macadOmom Ho gpvmoam commemom condomom hccmamnoo .gpcmoam condomom Ho ammo: hoomlooam op coca poop maoaoEmamm amoawoaoom wsapmaom moozoaomom Ho manna .a canoe 60 mcma .mmmaaaaz mama .Noo one maaoaxoam mcma .magmam aama .mcma .mmasma cam aeggmmmmz coma .ggmmc aama .gmgmc Am canoe oomv Am magma oomv mcma .gomaaaz oama .aamzopm one cafiom oama moasoc "00m p.53 oama .ggmgc fiflm SOHm> Qm> coma .mmam> mm> mama .gmoaamac vac aamwsnm coma .asggmmmmz coma .ggmgc acma .mcma .mmamma mom adgpa caaocom Avodzaazoov agaaagmaammm coasomom a«toaagmaamza coaSOccm goagagmmaoc mooasommm Ho npvmoam meadomom hcsmamnoo ommmagmoe .a magma 61 Am canoe oomv hpa>aaodooam agaeagmemomm Ac magma mmmv am mgaaagmgc Ac magma mmmv mmmgagma Ac canca oomv moapmaamaooam Am canoe occv mace a0 acnsoz haaaanmaac>¢ coaaaaooeoo mccadomom Ho gacccam enamomom ooazomom mommamzoo omegaggoc .a magma 62 Resource Breadth is apparently narrowed by increased Resource Availability and limited by sufficient Precipitation, Latitude (determining length of growing season), Stability of Productivity, and amount of Productivity. Finally, Resource Breadth appears to correlate with Genetic Variation, though this correlation is disputed. Evenness of Trophic Levels -- The free-body model for Evenness of Trophic Levels (Figure 12) is also relatively simple. Although it consists of five hierarchic levels, it contains a total of only eleven parameters, but with three feed-back lOOps. References are listed in Table 8. Evenness of Trophic Levels will be reduced where Seasonality is more marked, because different tr0phic levels (herbivore, detritivore) will reach peak abundances and diversities at different seasons. Where intensity of Predation is highest, Trophic Level Evenness will also be highest, at least in terrestrial systems where the Eltonian pyramid of numbers or biomass is rarely inverted, because Predation will reduce Competition at the lower tr0phic levels. If intensity of predation is decreased, Competition will reduce diversity at lower trophic levels and the degree of Evenness of Trophic Levels is restored and Predation may become relatively higher. These interactions thus form a negative feed— back loop. Predation can be made more intense by addition of Trophic Levels, if predators don't completely distinguish the trophic level of their prey, or addition of Trophic Levels can release a lower tr0phic level from control (compare Hurlbert, Mulla and Willson, 1972, with Hurlbert, Zedler and Fairbanks, 1972). Number of TrOphic Levels appears limited by Resource Availability. Influences on Resource Availability are discussed under Number of Taxa. Competition within a trophic level 63 .mmmnngm ae>oa oasmoaa. .ao ammo: Bomlooaaa .ma magma 36. t .35.. 8. .32. 8: 33 3 «31:2 Illv 29:52.20“. l, I I / b.2530! / / / E2530! 8 >535 E33332 . magma ciao: . 2 o. 23.: /ccww.>%>m wUflaomwa “.0 cwnZDZ Alli: 2130:. 32:3 20.2565: >._.3(vam(mmW 64 mmama .gmggom oama .maamc vac nmaaamm coma .mmamm oama .gmmcmc mama .mxgcnaamm one acacom .aaopaasm mama anomaaaz.mmc maasz .mmmgammc acma .gmmc mufiw Hmv—QHHHNU mama .mxscnaamm com acaccm .aaonaasm mama .QOmaaa3.ccm maadz .paonaasm mmama .gmgmom coma .mmamm mmama .amgmom oama .maamc mom gmaaacm ccma .mmamm oama anomzww mama .mxnmnaamm one aoaoom .aaopaasm mama .nocaaa3.com maadz .aaonaadm mcma .gmmc mom acxsamamw mama .ooaoemo mac>ca oasmoaa ac acflggz soapaaoQEoo soapmooam hpaamsomwom mae>ea cagmoae ac aoossz agaaagmaammm moaSOcom QoaaaaoQEoo zoapmocam mmoomo>m ao>oa cannons .cao>Qa cannons mo mmozso>m mo ammo: hoomlooam op coca acma cacaoewamm accamoaoom mmaamaom moocoaomom mo canoe .w oanme 65 Ac canoe occv Ac canoe oomv mcma .aaoaamgsac Gram Gog—H3 mcma .momaaz mcma .gOmaaz mam aamammgsac gmcma .aaoammgsac coma .mgamm mama .mama .mmasma USN 046.3%QO mcmaa.gmmc mam amgmaammc Am oaflme oomv mmama .mmggom coma .ozamm mama .mxomnaacm one amaoom apaonaadm mama anomaaaz mom maamz..gmmgamsc mmama .mmgmoa coma .mmamm mmmgagma :oaampamaooam mgaaagmaamsm commemom mxca mo amnesz aoboa canmoaa agaaagmaamsm ooazocmm mao>oa cannons moaaapoosoo mo acnemz soaacccam mmomso>m aoboa cannons omegamgoo .c magma 66 Ac canoe oomv Ac canoe oomv mmamagmeomam mgasagmscmgm am agaaagmgc agaaagmaamsa ooafiomom mac>oa cagooaa ac acnesz coaaaacmsoo noaaccoam mcocoo>m ao>oa canmoaa mmamaggoe .c magma 67 can reduce the Number of Taxa in that level and thus alter Trophic Level Evenness, increasing it if lower trophic levels are affected and decreasing it if higher levels are affected. Since Evenness of Trophic Levels can affect relative Predation intensity, and Predation can reduce Competition, we have another feedback loop. Number of Taxa and other parameters influencing Predation are discussed under Number Of Taxa. Trophic Level Distinctness -- The free-body model for Distinctness of Trophic Levels (Figure 13) is highly simplified relative to the other four models. It consists of three hierarchic levels but contains a total of only four parameters and no apparent feed-back lOOps. The Simplicity of this model is almost certainly due to the paucity of published reports that can be in any way associated with TrOphic Level Distinctness. Though the distinctness, and indeed the reality, of trophic levels is Often debated (g,g,, Darnell, 1961), it has never been clearly analyzed. It seems reasonable and also of interest to retain the concept of trophic levels while acknowledging that they may be more distinct in some communities than in others and even developing a means of measuring the degree of their distinctness. References are listed in Table 9. In an estuarine system which depends largely on production coming unpredictably from outside the system, Trophic Levels appear very indistinct and most organisms feed opportunistically. Where food is unpredictable and variable, communities may not be trophically specialized and structured. These studies suggest that Distinctness of Trophic Levels should be decreased by a decrease in Constancy of Resources. And simulation studies Show a two species model to be more stable with just competition or predation interactions between the two species, and not both kinds of interactions. 68 .mmmggmgagmag amsma magamaa am ammo: moocummaa .ma mascaa t...__m<._.U.0m~_m ._<~_Om<UZ<._.mZOUIIV am>m._ Elmo“.— >._._._Is I>II E Mn (Is I > II Mn < Is No Difference B Mn < Is ? Variable Variable T 7 I >II Mn >Is I>II D r ? Mn >Is I(II Ohio Mn{:Is Mich Mn: Mainland Community .I: Visit I Is: Island Community II: Visit II 9O Tablell4. RESULTS OF ANALYSES OF VARIANCE OF VARIABLES IN OHIO, PLOTS SCALE OF SAMPLING. Mg>Is VARIABLE MEDIAN EFFECT TOTAL INSULARITY SEASONALITI BLOCK INTERACTION NF 8.0 P<.001 N.S. 134.001 N.S. N.S. fi _ I>II _ NS 10.0 P<.01 P\<.001 N.S. P\<.Ol N.S. ijzls EF 0.837 P<.001 134.01 P\<.001 N.S. N.S. ‘ Mn>Is III TF 0.772 P<.001 N.S. P<.001 N.S. N.S. I>II T1 0.833 P\<.O5 P<.Ol N.S. N.S. N.S. Mp>Is DFl 0.187 P<.001 P<.001 N.S. N.S. N.S. Mnfls DF3 0.750 134.001 114.001 N.S. N.S. N.S. Lig>Is 011 0.087 P<.001 P<.001 N.S. N.S. P<.001 _ MnII NS 36.5 N.S. P<.01 N.S. N.S. Mn2>Is EF 0.292 N.S. N.S. N.S. N.S. """ES 0.671 N.S. Pg.01 N.S. N.S. Rhsgls BFl 0.207 N.S. N.S. N.S. N.S. BF3 0.093 N.S. N.S. N.S. N.S. BI1 0.033 N.S. N.S. N.S. 132.01 B13 0.621 P\<.05 P\<.Ol N.S. N.S. Mn3>Is TF 0.820 P\<.O5 N.S. P 4. 01 N.S. I>II TI 0.890 P\<.05 193.01 N.S. N.S. Mh2>Is DFl 0.092 P<.05 N.S. P<.Ol N.S. I>II DF3 0.783 N.S. N.S. N.S. N.S. 011 0.103 Rti 12.001 N.S. N.S. Mnstls DI3 0.838 P4.001 P<.001 N.S. N.S. MmI>IS 92 Table 16. RESULTS OF ANALYSES OF VARIANCE OF VARIABLES IN MICHIGAN, PLOTS SCALE OF SAMPLING. VARIABLE I MEDIAN EFFECT TOTAL INSULARITY SEAS ONALITY BLOCK [ INTERACTION NF 6.0 P<.001 Pg.001 N.S. P<.Ol P\<.05 Mn>Is NS 13.0 N.S. P\<.05 N.S. N.S. N.S. .5 Mn>Is EF 0.856 P(.001~ P<.001 N.S. P (.001 N.S. MnIs TI 0.8 7 N.S. N.S. N.S. P.<.05 N.S. DFl 0.0 N.S. N.S. N.S. N.S. N.S. DF3 0.833 N.S. P<.05 N.S. N.S. N.S. Mngs DIl 0.0 N.S. N.S. N.S. N.S. N.S. DI3 00903 NOS. N.S. P<005 NeSe NeSe IS II 93 Table 17. RESULTS OF ANALYSES OF VARIANCE 0F VARIABLES IN MICHIGAN, TRANSEC'I‘S SCALE OF SAMPLING. " " v'A'RIABLE MEDIAN , EFFECT TOTAL INSULARITY SEASONALITY INTERACTION NF 28.0 N.S. N.S. N.S. N.S. NS 36.0 N.S. N.S. N.S. N.S. EF 0.1167— N.S. N.S. N.S. N.S. ES 0.586 N.S. N.S. N.S. N.S. BF]. 0.200 134.05 N.S. N.S. 134.05 BF3 0.536 N.S. N.S. N.S. N.S. BIl 0.0118 N.S. N.S. N.S. N.S. 813 0.625 134.01 N.S. 134.001 N.S. IIs DFl 0.069 N.S. N.S. N.S. N.S. DF3 0.7732 134.05 N.S. 134.01 N.S. _ fl I Is DI3 0.86? N.S. N.S. N.S. N.S. 90 scale of sampling, or may reflect different scales of pattern of diversity in the community. Alternatively, the insular effect found on plots data in Michigan may be accidental or spurious. Lack of an insular effect on transects data of Number of Plant Species in Michigan suggests either of two possible conclusions, that resolution is lower at the transects scale of sampling or that the coarse scale of sampling represents a different scale of pattern of diversity in the __ community. It was predicted that Number of Taxa would increase during the summer season, and, as Table 13 shows, the reverse effect was actually found. Reference to Tables 14-17 shows that this effect occurs only in h Number of Families of invertebrate animals in the Ohio replicate of the ‘ study. But where found, this effect is very strong, with probability of less than 0.001 of no difference between visits, for both plots and transects data. A significant block effect and interaction effect for Michigan plots data, Number of Families of invertebrates, reflects an aberrantly low mean value for one of the mainland transects during the second visit. Evenness of Taxa -- Evenness of Taxa is expected to be lower on mainlands than islands. Table 13 shows that this difference is found. But reference to Tables 10-17 shows some inconsistency. This difference is clearly found for plant data, for both Ohio and Michigan at the plots scale of sampling, and for Ohio at the transects scale of sampling also. No difference was found for Evenness of Species of plants in Michigan for the transects scale of sampling. The animal data is less consistent. Island is found to be greater than mainland, as expected, in Michigan at the plots scale of sampling, but no difference between island and mainland 95 was found in either state at the transects scale of sampling, and island was actually lower than mainland in Ohio at the plots scale of sampling. Lack of an insular effect on Evenness of Taxa, for animals in Ohio and for both plants and animals in Michigan, at the transects scale of sampling, again may reflect either lower resolution or a different scale of pattern of diversity at that scale of sampling. In every comparison where an insular effect was found on Number of Taxa, a reverse effect was found on Evenness of Taxa. The one aberrant comparison, where island Evenness of Taxa was actually lower than that on mainland in Ohio for plots«data, is also the only comparison where an insular effect was found on Evenness of Taxa but not on Number of Taxa. It may be hypothesized that reduction of taxa on islands may differentially involve rare or uncommon taxa, increasing Evenness of Taxa on islands. It was predicted that Evenness of Thxa would decrease during the summer season. However, Table 13 indicates that no differences were found. As Tables 14-17 show, this is true of all data except for Ohio plots data. This one set of data showed Evenness of Families of invertebrate animals to increase during the summer, a direction of change apposite that expected. This one effect may again have been related to changes in weather factors such as precipitation. Resource Breadth -- Resource Breadth is expected to be greater on islands than on mainlands, but where significant differences are found, they are somewhat inconsistent (Table 13). Tables l#-l7 show that indeed, for the plots scale of sampling in Ohio, Resource Breadth of taxa (BBB) and of individuals (B13) are greater on the island than the mainland. However, still in Ohio, transects data for individuals (B13) is greater on the mainland, and no other comparisons in Ohio and none in Michigan showed 96 significant differences. These variables that do show differences (BE3 and BI3, plots, and BI3, transects) all deal with the percentages of animals assigned a rank of three for Resource Breadth, and the rank of three represents greatest Resource Breadth. Thus, the direction of differences in these variables reflects differences in the same direction in Resource Breadth. Insular and seasonal effects on Resource Breadth vary and may differ between states. However, a reversal of:insular effect between transects and plots data in Ohio suggests the possible conclusion that there is some difference between the two scales of sampling other than simply loss of resolution at the transects scale. Alternatively this reverse effect may merely reflect the variable and inconsistent effect of insularity on Resource Breadth. There also exists the possibility that the methods chosen for measuring Resource Breadth are inadequate. Significant insular or seasonal effects found on any of the variables estimating Resource Breadth are largely restricted to ”individuals” variables. "Individuals" variables may provide more resolution, as there are more individuals to work with than there are taxa. The alternative conclusion is that insularity and seasonality differentially influence the population sizes of taxa assigned different ranks but do not influence the number of those taxa. No prediction was made regarding any seasonal effects on Resource Breadth,.and as Table l3 indicates, results showed no definite pattern. Reference to TablesllN—l? shows that in Ohio, Resource Breadth, sampled by plots for individuals (BI3), decreased during the summer season, but that there were no other significant Ohio differences. In Michigan, in contrast, Resource Breadth increased during the summer season for 97 individuals sampled by plots and also for individuals sampled by transect. These variable results suggest that Resource Breadth is not 0directly influenced by either insularity or seasonality. However, Ohio data, at both scales of sampling, show several statistically significant interaction effects which generally indicate that the insular effect is reversed during the season. That is, the mainland has lower Resource Breadth than island during Visit I but higher during Visit II. An alternative way of saying the same thing is that the seasonal effect is reversed between mainland and island. That is, that 1 Resource Breadth increases during the summer season on the mainland but decreases on the island. Michigan plots data show a significant interaction effect for one of the "individuals" variables (BI3) which suggests that the seasonal effect is not exactly consistent across all blocks (transects). Evenness of TrOphic Levels -- No prediction was made regarding any insular effect on Evenness of Trophic Levels. As Table l3 shows, however, in all cases where significant differences were found, Evenness of Trophic Levels was higher on mainlands than on islands. Tables lfi-l? indicate these differences. In Ohio, "individuals" variables (TI) were greater on the mainland than on the island at both scales of sampling, while "taxa" variables did not differ between the two communities. In Michigan, Trophic Level Evenness by "individuals" was again greater on the mainland, but only for the transects scale of sampling. Also, TrOphic Level Evenness by "taxa" was higher on the mainland in Michigan, but only for plots scale of sampling in this case. Again, "individuals" data may provide more resolution, or insularity may differentially influence pOpulation sizes of taxa but not numbers of 98 taxa at different trophic levels. An additional difference is that plant data are not included in computation of Evenness of TrOphic Levels by individuals. The one comparison where insularity affects Evenness of TrOphic Levels by taxa, the plots scale of sampling in Michigan, with an effect opposite to that in other comparisons where an insular effect was found, is also the only comparison Where insularity decreases the Number of Families of animals. It may be that, in general by taxa, decrease only in Number of Species of plants counterbalances a decrease in Evenness of Trophic Levels indicated by individuals data on islands. These results suggest a higher proportion of individual predatory animals, but not of predatory animal taxa, on islands. Any reasons why this should be so are obscure. It was tentatively predicted that TrOphic Level Evenness would decrease during the summer season, and as Table 13 shows, this was indeed the direction of difference for those significant differences found. As Tables lh—l? show, Trophic Level Evenness decreases during the summer only in Ohio and only for "taxa" variables (TF), but for both plots and transects scales of sampling. No other comparisons showed significant differences. This decrease during the summer season for taxa data in Ohio probably reflects a similar decrease in Number of Families of animals, also in Ohio, which would form the top of an Eltonian pyramid. This seasonal effect on Number of Families of animals in Ohio is discussed above. gistinctness of TrOphic Levels -- No predictions were made regarding either.insular or seasonal effects on Distinctness of TrOphic Levels, but. as Table 13 shows, effects were found. The effect of insularity was reversed in direction between the two states. In Ohio, Trophic Level 99 Distinctness was greater on the mainland than the island. Tables 14-17 show that this is true for taxa data at the plots scale of sampling (DEB greater on the mainland, DFl lower on the mainland) but that no difference is shown for the transects scale of sampling. It is also true for individuals data (again, DI3 greater on the mainland, DIl lower on the mainland) at both the plots and the transects scales of sampling. In Michigan, only one comparison shows a significant difference, and it is thf- 'Hv '5 reversed to that found in Ohio. Individuals data (DIl) is greater on the mainland for the transects scale of sampling. This indicates a larger percentage of animals with small values of Distinctness of Trophic Levels F”... on the mainland, so that the value of Distinctness of Trophic Levels is lower on the mainland. The two replicates of the study, in Ohio and in Michigan, differ in several ways. The dominant canOpy species differ, the Ohio communities are isolated woodlots while the Michigan communities are localities in extensive forests, latitude is lower in Ohio, topography is more regular in Ohio, and two additional islands can serve as stepping stones between the Ohio island community and the mainland. It is also possible that the one Michigan comparison showing an insular effect is spurious and aberrant. It is possible that immigration rate is highest on the Ohio island, that it therefore has a higher pr0portion of invading species than the other communities, and that these invading species may recognize less Distinctness of Trophic Levels in feeding. It is also possible that, in the Ohio island community, a less diverse resource base, represented by lower Number of Species of plants, selects for less tr0phically specialized organisms in the higher tr0phic levels. These results are not clarified by the variable effect of insularity on Resource Breadth in 100 Ohio. Increase in predatory animals on islands, suggested by the results of the insular effect on TrOphic Level Evenness (see above), may not include such animals as spiders, which clearly distinguish trophic levels. However, Drew (1967) found more individuals spiders on Beaver Island, Michigan, than on the nearby mainland. Only a few comparisons show a significant seasonal effect on Distinctness of Trophic Levels, but these all show the same effect, that Distinctness of Trophic Levels increases during the summer season (see Table 13). Tables lfi-l? Show that in Ohio for taxa data at the transects scale of sampling, the percentage of animals assigned a rank of one (DFl) decreases during the summer, indicating that Distinctness of Trophic Levels increases during the summer. In Michigan, for plots data, the percentage of individual animals given a rank of three (DI3) increases during the summer, and for transects data, the percentage of taxa of animals given a rank of three (DF3) also increases during the summer, both indicating such an increase in Distinctness of Trophic Levels during the summer season. This effect may be due to an increase in Number of Taxa of such animals as spiders late in the summer. Pulliam, Odum and Barret (1968) show that the number of spider taxa increases during the summer season in a field of millet more slowly than other arthrOpod species. Individuals data at the plots scale of sampling (DIl and DI3) show statistically significant interaction effects which suggest that the insular effect is not consistent across all blocks (transects). Predicted Correlations Correlations between pairs of the five system parameters are summarized in Table 18. The ten parameter pairs are listed in the first column and the correlations between them that are predicted from the Table 18. 101 FIVE SYSTEM PARAMETERS. SUMMARY OF RESULTS OF PARTIAL CORRELATION ANALYSES OF THE """I‘ Para- Correlation meter Ex- Found Comments Conclusion Pair pected. N-E - - Within invertebrate animals. Prediction partially 4 0 'Within plants and between verified. plants and animals. N-B - - ’Within invertebrate animals. Prediction partially 8 0 Between plants and animals. verified. N-T + + 'Within invertebrate animals. Prediction found 0 — With plants: taxa variables. invalid. 0 With plants: individuals var. I N-D 0 - Within invertebrate animals: Prediction found 8 individuals variables. invalid. 0 Between plants and animals. E-B 0 O (-) tendency in Ohio. Prediction verified. 8 0 In all cases in Michigan. E—T + + Within invertebrate animals: Prediction partially 4 individuals variables. verified. 0 In all other cases. E-D 0 - Within invertebrate animals. Prediction found 8 0 With plants. invalid. B-T + O In all cases. Prediction found 8 invalid. B-D 0 0 In all cases. Prediction verified. l6 T-D O 0 In virtually all cases. Prediction verified. '8 + For a single variant, a taxa variable in Michigan. See text for further explanation of table. 102 system model in the second column. In the third column correlations found in this study are tabulated, with slightly more detail. The "Comments” column characterizes this detail, and the last column concludes, for each parameter pair, whether the predicted correlation is verified or found invalid. This general conclusion is based on the following criteria. If no correlation is predicted between two parameters, the prediction is found invalid only if a correlation is found between .1 more than one pair of the variables estimating the parameter, and is otherwise verified. If a correlation is predicted between two parameters, the prediction is found invalid only if a correlation of sign opposite that predicted is found between more than one pair of the variables h estimating the parameters, or if no correlation is found between all, or all but one, pair of variables, and is otherwise verified or partially verified. These results are presented in more detail in Table 19. The sign of seventh order partial correlation between the pairs of variables with both the "taxa" set of variables and the "individuals” set of variables is indicated for both the Ohio and the Michigan replicates of the study. For each pair of variables within each set and for each state, the number of coefficients showing a negative correlation, no significant correlation and a positive correlation are recorded under the three columns, (-), (O) and (+), respectively. There are six of these coefficients for each set of variables (see Analyses -- Analysis of Predicted Correlations, above). A companion column to the (-), (0) and (+) columns within each set, labeled "Results”, lists the result of the t-test of the sign of the coefficients. 103 O + Him I+O CO 0000 I00 E 0010‘” I00 IOO+OO {ONOOOOOHOOWOH WanmHWmewdw: I\‘I n1 IOOOO mmm QIz IOOOOOOOOOOOOOOO Biz I + + c>c>c>c>+ c>c>c>c>+ + c>c>c:o WW 0 Od’tfim l-IN\OA\.O\O'KO\O\O\O'\O('1V3\0I-TVNNO IOO+OO I + mlz I+OOOOO+OOOOOOOO+OO I + de mm IOO |+ IOOO+OOO+ IOOOOOOOOOOOO IOO IOO IOOOOO+O+ IOOO+ HOOr—lOmOr—IHWOOOOOOOOOr-IVNOt-‘IOOOOOO O + $2 IHUNOHNOI-iouh1.:ONOOOOHOOOOOUNOOLfl'Or-l o m +OOOOOMHOOHHMOOOONr—IOOQ‘HOO omawmemeomddmdmmwddwbmme moanmfinoa amadde>fivan E 58 l0J0r—lr—‘IOOOOOWOHOOOOOOOOOOMOOWHO lOU‘VOOOONOOOOOJOOOOMOOOOOMOONON (“‘VfifiVO\O\O\n\d£T\O<3 c>c>r4<3 c>c>uwr4<:. +JOOOOd'OOOKOONOOr-IOOOOOOOOOOOUUU +£OOOOHOAYOWOOOOHOOOOOOOOOOOONO OHM-{\OUNKO mhfi' u xiv—{NOW O NNVJ LIV-INN + pgmom .H pgmom moanmflaah moanmfinm> moanefinm> amuse: :mamscw>fip:Ha1. annoy: 25302 8%. :oapmH loanoo “.3625 mnfiam oHnmwnw> .soapwufiam> pmoaup mo mpHdmcm one .memflnm> Ho omha use madam game now need oanmana> mode coozpom oon> wde no .onmm .mocfiz mo magoflOfimmooo :oapmflonaoo Ho zesosvonm mncpoa Iaawm .oa manna 100 .oHnwp mo nonpwanmxo nonpnsm now 9K0» mom .moHannm> gmHmspH>Hch= nom =H: noppoH one no moHannm> =mnme= non gm: nonpoH one moHSOHonH * o o w o + dIn o o o w,n o n m,m o LIPS o o m o o o m o o o w o o o 7w, n o *nIHIQ nun I o H m, I o H w I o o w I o H W‘ I m*QIH*Q Q o o m n o o m, n o 0 ex m o o e m o m*a-n*m o o m o o H w, o o H m o o N e o o H*mIm+m o o m o o o m, o 0 011m, 0 o o m o o n*mun*m o o m o o o o o o o o o o o e o o H*mun*m mum pHSmom + o I pHdmmm + o I pHsmom + o I pHsmmm + o I mwflflmflhw> MOHDmflHm> mOHDwHHd> mmflflmflhmgw £0.36.” :mHmdeprst gmama: :mHmSUH>szH: =wxme: Ionnoo mnme mnepofi zaunmunz onmo eoeomnxm onemnnm> Ianan eossnpgoo .mn manna 105 It must be remembered that any variable correlating with one of the variables involving the percentages of animals assigned a rank of one for Resource Breadth or Distinctness of Trophic Levels, BFl, BI1, DFl, DIl, is actually showing the reverse correlation with the parameter itself, since a rank of one indicates low values for either parameter. Correlation with a variable involving assignment of rank three (BF3, BIB, DF3, DI3), however, indicates correlation in the same direction with the parameter ran itself, because a rank of three indicates high value for both parameters. Any correlations found, those supporting predictions as well as L those refuting hypotheses, may represent direct causation, may arise from 0 common correlation with a third parameter, or may be spurious. Spurious L"- correlations are h0pefully eliminated by the statistical t-test criteria used above to validate correlations between pairs of variables. Direct causation cannot be investigated in this study. Although many possible sources of common correlation cannot be investigated here, two of the more likely sources can be evaluated by comparing the results of analyses of insular and seasonal effects on the five system parameters with the results of partial correlation analyses of relationships among them. Number of Taxa versus Evenness of Ta§§,-- A negative correlation is predicted between Number of Taxa and Evenness of Taxa and is only found (Table 18) between Number of Families of invertebrate animals (NF) and Evenness of Families of invertebrate animals (EF) for both states (Table 19). No significant correlation is found between Evenness and Number of Taxa either within plants or between plants and animals. The prediction in this case is partially verified. Analyses of variance results indicate that where Number of Families of invertebrates is greater on mainlands, Evenness of Families of 106 invertebrates is consistently smaller on mainlands. Factors of season- ality appear to have limited inverse effect on these two parameters. The most parsimonious explanation of the negative correlation found between Number of Taxa and Evenness of Taxa apparently is that they are correlated in common, although in inverse manner, with factors of insularity. Reports in the literature provide evidence of a relationship between Number of Taxa and Evenness of Taxa, and provide speculation at most of . ET the causality and mechanisms of that relationship. It is possible, and 5 parsimonious, that this reported relationship arises from common ? correlation of the two parameters with factors of insularity or other factors. 1 L- Number of Taxa versus Resource Bread§h_-- A negative correlation is also predicted between Number of Taxa and Resource Breadth and again is found (Table 18) between Number of Families of animals (NF) and individuals variables of Resource Breadth (BI1, BIB), in Ohio (Table 19). No significant correlation is found between Number of Species of plants (NS) and Resource Breadth by any variable. The prediction in this case is partially verified. Results of analyses of variance indicate that factors of insularity and also factors of seasonality have effects on Number of Taxa and Resource Breadth (BFl, BF3, BIl, BI3), but these effects do not show a relationship between the two parameters. These results suggest that animal taxa 2' readily gained or lost from a community, that is, with the least stable population dynamics, are also the animals with the least Resource Breadth, that is, the most specialized feeding behavior. This suggestion is contrary to the theories of fugitive species (Hutchinson, 1959) and of Opportunistic species undergoing "r-selection" (MacArthur and Wilson, 1967). 107 Number of Taxa versus Evenness of Trophic Levels -- A positive correlation is predicted between Number of Taxa and Evenness of Trophic Levels and is found between Number of Families of animals (NF) and Evenness of Trophic Levels (Table 18) for taxa variables (TF) in both states and for individuals variables (TI) in Ohio but not in Michigan (Table 19). However, a negative correlation is found between Number of Species of plants and taxa variables of Evenness of Trophic Levels (TF) in both fink states, No significant correlation occurs between Number of Species of plants and individuals variables (TI). In a terrestrial system, as in the present study, where an Eltonian pyramid is not likely to be reversed, an increase in the number of taxa of plants would widen the base of the pyramid and hence decrease the Evenness of Trophic Levels. An increase in the number of animal taxa, however, especially higher level consumers, would widen the apex of the pyramid and increase the Evenness of Trophic Levels. The correlations found here either reflect these considerations, or the assumption is invalid that animal families conservatively2estimate animal species. The prediction in this case is refutedfi but apparently clarified. Results of analyses of variance indicate that either the factors of insularity or the factors of seasonality have effects in a similar direction for all variables affected, both for Number of Taxa and Evenness of TrOphic Levels. These results are inconsistent with the opposing correlations of animal taxa and plant taxa with Evenness of TrOphic Levels, and it must be concluded that these two parameters are not commonly correlated either with factors of insularity or with factors of seasonality. 108 Number of Taxa versus Distinctness of Trophic Levels -- No correlation is predicted between Number of Taxa and Distinctness of Trophic Levels. However, a negative correlation is found (Table 18) between Number of Families of animals (NF) and individuals variables of Distinctness of Trophic Levels (DIl, DI3) (Table 19). No significant correlations were found between any of the other pairs of variables for this pair of parameters. The prediction of no relationship in this case is found fr invalid. Results of analyses of variance suggest that Number of animal Taxa is greater on mainlands than on islands in Michigan, while estimates of Distinctness of Trophic Levels are lower on the mainland. These results L1 suggest a negative correlation between the two parameters, and a parsimonious conclusion is that Number of Taxa and Distinctness of TrOphic Levels are commonly correlated with factors of insularity. Ohio analyses of variance provide no information to affect this conclusion. There is little or no apparent affect by the factors of seasonality on any of these variables in the analyses of variance and hence common correlation with these factors does not appear to influence the correlation between Number of Taxa and Distinctness of Trophic Levels. The simplest free-body model is that for Distinctness of TrOphic Levels, primarily because it is derived from the fewest sources. It is therefore probably the most incomplete, and the conclusion of common correlation of Number of Taxa and Distinctness of Trophic Levels with factors of insularity must be considered tentative until more is known of the parameter Distinctness of Trophic Levels. Evenness of Taxa versus Resource Breadth -- No correlation is predicted between Evenness of Taxa and Resource Breadth, and in fact no significant 109 correlation is found (Table 18), except between a single pair of parameters. A negative correlation shows up between Evenness of Families of animals (EF) and Resource Breadth by individuals (B13) in Ohio (Table 19). Analyses of variance indicate that in this comparison these two parameters may'both be influenced by factors of insularity in Ohio. Seasonality appears to have no influence on this result. It is concluded that the prediction of no relationship between Evenness of Taxa and Resource Breadth is verified. l Evenness of Taxa versus Evenness of Trophic Levels -- A positive correlation is predicted between Evenness of Taxa and Evenness of Trophic Levels and is in fact found (Table 18) between Evenness of Families of animals (EF) and Evenness of Trophic Levels by individuals (TI) in both states (Table 19). Other pairs of variables for these two parameters are not significantly correlated. The prediction in this case is partially verified. Analyses of variance indicate that Evenness of Families of animals (EF) and Evenness of Trophic Levels by individuals (TI) are apparently not commonly correlated with factors of insularity. Seasonality appears to have slight and inconsistent effect on these two parameters. Information in this report appears insufficient at this time to generate any hypotheses to explain the positive correlation found between these two parameters. Evenness of Taxa and Distinctness of TrOphic Levels -- Lack of any correlation is predicted between Evenness of Taxa and Distinctness of TrOphic Levels. However, a negative correlation appears (Table 18) between Evenness of Families of animals (EF) and Distinctness of Trophic Levels by individuals (D11, D13) but not by taxa in Ohio and by both 110 individuals and taxa (D11, DI3, DFl, DF3) in Michigan (Table 19). Evenness of Species of plants shows no correlation with Distinctness of Trophic Levels. The prediction in this case is found to be invalid. Results of analyses of variance show that the variables used to estimate Evenness of Taxa and those used to estimate Distinctness of TrOphic Levels generally are greater on mainland than on island in Ohio and smaller on mainland than island in Michigan. These results would Fa. suggest a positive correlation between Evenness of Taxa and Distinctness I of Trophic Levels in Ohio, but a negative correlation is found between them. Hence, this correlation has some other origin than common correlation with factors of insularity. Since both these parameters appear to be related to Number of Taxa by common correlation with factors of insularity, at least two factors must constitute insularity. There is little or no apparent effect by seasonality on any of these variables and hence common correlation with these factors does not appear to influence this correlation. These results suggest a relationship between taxa of animals that do not clearly distinguish the tr0phic levels from which they feed and taxa that tend to be either rare or dominant or both. Resource Breadth versugfiEvenness of_T§x§|-- A positive correlation is predicted between Resource Breadth and Evenness of TrOphic Levels. However, no significant correlation is found between any of the pairs of variables for these two parameters (Table 18, Table 19). The prediction in this case is found to be invalid. Results of analyses of variance provide no explanation for the lack of validity of this prediction. It may be conjectured that an inverse causal relationship exists between Number of Taxa and Resource Breadth and also between Number of Taxa and Evenness of TrOphic Levels, and that these relationships lll counteract any relationship between Resource Breadth and Evenness of Trophic Levels. From the information in this study, no other hypothesis can at present be formulated to explain the absence of the predicted negative correlation between Resource Breadth and Evenness of Trophic Levels. Resource Breadth versus Distinctness of Trophic Levels -- No correlation is predicted between Resource Breadth and Distinctness of Trophic Levels, :- and indeed no significant correlation is found (Table 18) between any pair of variables for these two parameters (Table 19). The prediction in this case is verified. Evenness o£_Trophic Levels versuEiDistinctness ongrophic Levels -- Lack Iv of any correlation is predicted between Evenness of Trophic Levels and Distinctness of TrOphic Levels, and no correlation is indeed found (Table 18) between all but a single pair of variables for these two parameters. AS a single exception, Evenness of Trophic Levels by taxa (TF) shows a positive correlation with Distinctness of Trophic Levels by taxa (DF3) in Michigan only (Table 19). The prediction in this case is verified. Predicted Correlations The system model developed from review of literature leads to ten predictions regarding correlation between pairs of parameters. Six of these predictions are either verified or partly verified, and have failed to be falsified. Four other predictions are found to be invalid. Two of these, however, involve Distinctness of Trophic Levels as one of the pair of parameters, and the confidence of the predictions was not high. In the case of a third prediction, falsification apparently clarified, rather than rejected, the relationship predicted. It must be concluded 112 that the general structure of this model is valid, though modification of certain relationships within it is required, leading to a revised system model (Figure 15). The hypothesized relationship between Number of Taxa and Evenness of Taxa is changed to one of common correlation with factors of insularity. (See Figures 14 and 15). The relationship between Number of Taxa and Resource Breadth is upheld. The relationship between Number of Taxa and Evenness of TrOphic Levels is clarified by breakdown of Number of Taxa . -mm; n-.'J.9 into number of animal families and number of plant species. A relation- ship is discovered between Number of Taxa and Distinctness of Trophic Levels; its source may or may not be common correlation with factors of insularity. The lack of relationship between Evenness of Taxa and Evenness of Trophic Levels is only partially upheld, possibly indicating that other related parameters may complicate this relationship. A relationship of unknown source is discovered between Evenness of Taxa and Distinctness of Trophic Levels. The relationship between Resource Breadth and Evenness of TrOphic Levels is lost for unknown reasons. Relationships between Evenness of Taxa and Resource Breadth, Resource Breadth and Distinctness of Trophic Levels, and Distinctness of Trophic Levels and Evenness of TrOphic Levels are all zero as expected, and no new information about them is provided. Now that the general structure of these models has withstood experimental test, and there is reasonable confidence in the reality of the relationships among these community-level parameters, it is reasonable to ask the nature of the relationships and the mechanisms that underlie them. The most productive test of the revised system model would 113 Distinctness o Trophic - - Consta of ,/Resources Resource Ros rce ' / \ 3mm Avoim’bility\ ' ./ \ \ / /\'22 \ / . \ \ ’ Competition Number Evenness ‘3‘ ....... .._rn._..._...... (3f Taxa I Taxa I ’z' Predation... 1% ./ Trophic Level Evenness Figure 15. Revised System Mbdel. 110 involve experimental manipulation of the parameters under controlled conditions. Controlled conditions are difficult to obtain with biological communities, and an alternative approach is to seek out just the "right” natural conditions where large numbers of parameters remain constant and few vary. As testing these models progresses, additional parameters can be included in the system.models, the free-body models of individual IL parameters can be refined, and the qualitative relationships discussed in E this report can be replaced by quantitative relationships. 5 Up to the present, community level studies appear to have concentrated 3 upon few parameters, their extent and mechanisms which may generate them. E; Discussion of relationships among community level parameters is rudimentary and generalized. The results of this study indicate that these relationships are complex and consistent and warrant further investigation. SUMMARY AND CONCLUSIONS Five parameters, Number of Taxa, Evenness of Taxa, Resource Breadth, Evenness of Trophic Levels, and Distinctness of Trophic Levels, are shown to be related to food-web structure. They are investigated from two points of view. First they are compared between islands and mainlands and through the summer growing season. Second, a model of expected relationships among the parameters is developed from review of literature, and ten predictions concerning correlation between parameters are drawn from this model and are tested experimentally. I. 1. Number of Taxa is lower in island than mainland communities, as would be expected from the theory of island biogeography. 2. Evenness of Taxa is higher on islands, possibly due to differential absence of rare taxa on depauperate islands. Evenness of Taxa is not influenced by summer season changes. 3. Resource Breadth may be influenced more by scale of sampling than by insularity or seasonality. 4. Evenness of TrOphic Levels is higher on mainlands than on islands. Insularity may differentially influence pOpulation sizes of taxa at different trophic levels but not the numbers of taxa at those levels. Evenness of Trophic Levels decreased during the summer season in Ohio, probably because of a similar decrease in Number of Families of animals in Ohio, which would narrow the apex of an Eltonian pyramid. 5. The effect of insularity on Distinctness of Trophic Levels is reversed between Ohio and Michigan. Several alternative hypotheses to explain this reversal are presented. Distinctness of TrOphic Levels increases during the summer season. 115 n... a'!’ ‘fih.= "p U" II. 116 l. The general structure of the system model has withstood experimental test. The relationships modeled are complex, consistent, and warrant further investigation and search for mechanisms. 2. It is parsimonious and not invalid to hypothesize that Number of Taxa and Evenness of Taxa are related only by common but inverse correlation with other parameters such as insularity. 3. A negative relationship between Number of Taxa and Resource i In. Breadth suggests the hypothesis that fugitive or opportunistic species are also the species with least Resource Breadth. 4. In communities where an Eltonian pyramid is not inverted, Evenness of TrOphic Levels is inversely related to Number of Taxa of plants and directly related to Number of Taxa of animals. 5. It is tentatively concluded that Number of Taxa and Distinctness of TrOphic Levels are related by common and direct correlation with factors cf insularity. 6. A negative correlation between Evenness of Taxa and Distinctness of Trophic Levels suggests the hypothesis that animal taxa that do not clearly distinguish tr0phic levels in feeding also tend to be either rare or dominant or both. I II'I. III! I! .111 LIST OF REFERENCES .Arnett, Ross H., Jr. (1960): The Beetles of the United States. Catholic University of America Press, wash., D. C. rs. .Ares, J. O. and R. J. C. Leon (1972): An Ecological Assessment of the Influence of Grazing on Plant Community Structure. J. Ecol., 60(2): 333-342- .Arnold, S. J. (1972): Species Densities of Predators and their Prey. Am. Nat., 106(948): 220-236. Auclair, A. N. and F. G. Goff (1971): Diversity Relationships of Upland - Forests in the western Great Lakes Area. Am. Nat., 105(946):499-528. "' Baker, R. H. (1971): Nutritional Strategies of Myomorph Rodents in North American Grasslands. J. Mamm., 52(4):800-805. Beer, J. R., P. Lukens and D. Olsen (1954): Small Mammal Populations on the Islands of Basswood Lake, Minnesota. Ecol., 35:437-445. Blatchley, W. S. (1920): Orthoptera 2;.Northeastern America. The Nature Publishing Co., Indianapolis. Blatchley, W. S. (1926): Heteroptera gr,True Bugs gf_Eastern North America. The Nature Publ. Co., Indianapolis. Borror, D. J., and D. M. DeLong (1971): An.1ntroduction tg_tthStudy‘g£ Insects, 3rd Edition. Holt, Rinehart, and Winston. Bristow, W. S. (1941): lhg_Comity 2; Spiders, vol. 11. London, Printed for the Ray Society. Bristow, W; S. (1958): [Thg_world gf_Spiders. Collins, St. James Place, London. Brown, J. (1972): Personal communication. Burch, J. B. (1962): How to Know the Eastern Land Snails. Wm. C. Brown Co., Dubuque, Iowa. Cameron, G. N. (1972): Analysis of Insect TrOphic Diversity in two Salt Marsh Communities. Ecol., 53(1): 58-73. 117 118 Caswell, H., H. E. Koenig, J. A. Resh, and Q. E. Ross (1972): An Introduction to Systems Science for Ecologists. In: B. C. Patten,, ed., Systems Analysis apg_Simulation inDEcology. Academic Press, New York and London. Cloudsley-Thompson, J. L. (1939): Spiders, Scorpions, Centipedes and, Mites. Pergamon Press, London. Colwell, R. K. and D. J. Futuyma (1971): On the Measurement of Niche Breadth and Overlap. Ecel., 52(4):567-576. Connell, J. H. and E. Orias (1964): The Ecological Regulation of Species Diversity. Am. Nat., 98(903):399-4l4. n1. Crowell, K. L. (1961): The Effects of Reduced Competition in Birds. Proc. Nat. Acad. Sci., 47:240-243. Crowell, K. L. (1962): Reduced Interspecific Competition among Birds of Bermuda. Ecol., 43(1):?5-88. . Crowson. R. A. (1967): _T_h2 __Natural mm a 111.6. Earn—111.62 °_f g... ColeoEtera. E. W; Classey, Ltd., Middlesex, England. Curran, C. H. (1965): The Families and Genera of North American Diptera. Henry Tripp, W00dhaven, N.Y. Darnell, R. M. (1961): Trophic Spectrum of an Estuarine Community; Based on Studies of Lake Pontchartrain, Louisiana. Ecol., 42:553-568. DeBenedictis, P. A. (1973): On the Correlations Between Certain Diversity Indices. Am. Nat., 107(954):295-302. Downer, G. (1972): Personal Communication. Drew, L. C. (1967): Spiders of Beaver Island. Publications of the MMseum, Mich. State U., Biol. Series, 3(3):153-208. Dwyer, P. D. (1973): Feature, Patch and Refuge Area: Some Influences on Diversity of Bird Species. Emu, 72(4):l49-156. Ellis, A. R. (1969): British Snails. Oxford, Clarendon Press. Fedorov, A. A. (1966): The Structure of the Tropical Rain Forest and Speciation in the Humid Tropics. .J. Ecol., 54(1):1-ll. Fitch, H. S. (1963): Spiders g£_§hg.Universitylg§_Kansas Natural Histogy Reservation ang_Rockefeller Experimental Tract. U. of Kansas, Ewe. of Nat. Hist., Misc. Pub. No. 33, pp. 1-202. Fleming, T. H. (1973): Numbers of Mammal Species in North and Central American Forest Communities. Ecol., 54(3):555-563. Futuyma, D. J. (1973): Community Structure and Stability in Constant Environments. Am. Nat., 107(955):443-445. 119 Garfinkel, D. and R. Sack (1964): Digital Computer Simulation of an Ecological System, Based on a.Nbdified Mass Action Law. Ecol., 45(3)=502-507- Gertsch, W; J. (1949): American Spiders. New York. Graham, A. (1955): Molluscan Diets. Proc. Malac. Soc. London, 31:144-157. Grant, P. R. (1966): Ecological Compatibility of Bird Species on Islands. Am. Nat., 100(910):451.462. Grant, P. R. (1968): Bill Size, Body Size and the Ecological Adaptation of Bird Species to Competitive Situation on Islands. Sys. 2001., Ft 173319-333- Hairston, N. G. (1959): Species Abundance and Community Organization. Ecol., 40:404.416. Hairston, N. 0., J. D. Allan, R. K. Colwell, D. J. Futuyma, J. Howell, M. D. Lubin, J. Mathias and J. H. Vandermeer (1968): The Relationship Between Species Diversity and Stability: An . Experimental Approach with Protozoa and Bacteria. Ecol., 49(6): "' 1091-1100. Hamilton, T. H. and N. E. Armstrong (1965): Environmental Determination of Insular Variation in Bird Species Abundances in the Gulf of Guinea. Nature, 207:148-151. Hamilton, T. H., R. N. Barth and I. Rubinoff (1964): The Environmental Control of Insular Variation in Bird Species Numbers. Proc. U. S. Nat. Acad. Sci., 52:132-140. Heatwole, H. and R. Levins (1972): Trophic Structure Stability and Faunal Change During Recolonization. Ecol., 53(3):53l-534. Heatwole, H. and R. Levins (1973): Biogeography of the Puerto-Rican Bank: Species Turnover on a Small Cay, Cayo Ahogado. Ecol., 54(5): 1042-1055. Hill, M. O. (1973): Diversity and Evenness: A Unifying Notation and its Consequences. Ecol., 54(2):427-433. Hubbell, S. P. (1973): Populations and Simple Foodawebs as Energy Filters: 1. One-Species Systems. Am. Nat., 107(953):94-121. Hubbell, S. P. (1973): Populations and Simple Food webs as Energy Filters. 11. Two Species Systems. Am. Nat., 107(953):122-151. Hurd, L. E., M; V. Mellinger, L. L. Wolf and S. J. MCNaughton (1971): Stability and Diversity at Three Trophic Levels in Terrestrial Successional Ecosystems. Science, 173:1134-1136. 120 Hurlbert, s. H., M. S. Mulla and H. R.'Willson (1972): Effects of an OrganOphosphorus Insecticide on the Phytoplankton, Zooplankton, and Insect Populations of Fresh4Water Ponds. Ecol. Monogr., 42(3): 269—299. Hurlbert, S. H., J. Zedler, and D. Fairbanks (1972): Ecosystem Alteration by Mesquite Fish (Gambusia Affinis) predation. Science, 175:639-641. Hutchinson, G. E. (1959): Homage to Santa Rosalia, or Why Are There So Many Kinds of Animals? Am. Nat., XCIII:l45-l59. Janzen, D. (1970): Herbivores and the Number of Tree Species in Tropical Forests. Am. Nat., 104:501—528. Janzen, D. H. (1973): Sweep Samples of Tropical Foliage Insects: Effects of Seasons, Vegetation Types, Elevation, Time of Day, and Insularity. Ecol., 54(3):687-708. Janzen, D. H. and T. W. Schoener (1968): Differences in Insect Abundances and Diversity between Wetter and Drier Sites during a TrOpical Dry Season. Ecol., 49(1):96-109. Johnson, R. G. (1970): Variations in Diversity within Benthic Marine Communities. Am. Nat., 104:285-300. Johnson, M. P., L. G. Mason and P. H. Raven (1968): Ecological Parameters and Plant Species Diversity. Am. Nat., 102:297-306. Johnson, M. P. and P. Raven (1970): Natural Regulation of Plant Species Diversity. pp. 127-162 in: Evolutionary Biolo , ed. by Dobzhansky, Hecht and Steere, vol. 4. Appleton-Century-Crofts. Kaston, B. J. (1972): How 39 Know the Spiders. wn. C. Brown Co., Dubuque, Iowa. Keast, A. (1970): Adaptive Evolution and Shifts in Niche Occupation in Island Birds. Biotropica, 2(2):61-75. Kendall, M. G. and A. Stuart (1967): lhp_Advanced Theopy pf Statistics, vol. 2. Hafner Pub. Co., N.Y. Kilburn, P. D. (1966): Analysis of the Species-Area Relation. Ecol., 47(5):831-843. Klopfer, P. H. and R. H. MacArthur (1961): On the Causes of Tropical Species Diversity in Niche Overlap. Am. Nat., 95:223-226. Kohn, A. J. (1967): Environmental Complexity and Species Diversity in the Gastropod Genus Conus on Indoawest Pacific Reef Platforms. Am. Nat., 101:251-259. Kricher, J. C. (1972): Summer Bird Species Diversity in Relation to Secondary Succession on the New Jersey Piedmont. Am. Midl. Nat., 89(1):121—l37. 121 Leigh, E. G., Jr. (1965): On the Relation Between the Productivity, Biomass, Diversity, and Stability of a Community. Proc. N. A. S., 533777-783. Leonard, A. B. (1959): Handbook of Gastronods in Kansas. U. of Kansas, Mus. of Nat. Hist., Misc. Pub. No. 20, pp. 1-224. Levins, R. (1968): Evolution ;p_Changing Environments. Princeton University Press, Princeton, N. J. Levinton, J. (1972): Stability and Trophic Structure in De osit-Feeding and Suspension-Feeding Communities. Am. Nat., 106(950 :472-486. pm Loucks, O. L. (1970): Evolution of Diversity, Efficiency and Community Stability. Am. 2001., 10:17-25. MacArthur, R. H. (1964): Environmental Factors Affecting Bird Species Diversity. Am. Nat., 98:387-397. MacArthur, R. H. (1965): Patterns of Species Diversity. Biol. Rev., 40:510-533. iv MacArthur, R. H. (1969a): Patterns of Communities in the Tropics. Biol. J. Linn. Soc., 1:19-30. MacArthur, R. H. (1969b): Species Packing, and What Interspecies Competition Minimizes. Proc. of the Nat. Acad. of Sci., 64(4): 1369-1371. MacArthur, R. H. (1970): Species Packing and Competitive Equilibrium for Many Species. Theor. Pop. Biol., l(l):1-ll. MacArthur, R. H. (1972) Geographical Ecology: Patterns in the Distribution pf Species. Harper and Row, Publishers. MacArthur, R. H., J. M. Diamond and J. R. Karr (1972): Density Compensation in Island Faunas. Ecol., 53(2) :330-342. MacArthur, R. H. and R. Levins (1964): Competition, Habitat Selection and Character Displacement in a Patchy Environment. Proc. Nat. Acad. Sci., 1:1207-1210. MacArthur, R. H. and R. Levins (1967): The Limiting Similarity, Convergence and Divergence of Coexisting Species. Am. Nat., 101: 377-385. MacArthur, R. H., J. thArthur, D. MacArthur and A. MacArthur (1973): The Effect of Island Area on Population Densities. Ecol., 54(3): 657-658. MacArthur, R. H. and M. C. Recher (1966): On the Relation Between Habitat Selection and Species Diversity. Am. Nat., 100:319-332. 122 MacArthur, R. H. and E. O. Wilson (1963): An Equilibrium Theory of Insular Zoogeography. Evol., l7(4):373-387. MacArthur, R. H. and E. O. Wilson (1967): Theopy.pf Island Biogeography. Princeton, N. J., Press. Mackay, R. and J. Kalff (1969): Seasonal variation in Standing Crop and Species Diversity of Insect Communities in a Small Quebec Stream. Ecol., 50(1):101-109. Maguire, B., Jr. (1967): A Partial Analysis of the Niche. Am. Nat., 101:515-526. Mahler, J. (1972): Personal communication. Malone, C. R. (1972) Effects of a Non-Selective Arsenical Herbicide on Plant Biomass and Community Structure in a Fescue Meadow. Ecol., 53(3)=507-512. Margalef, R. (1967): Some Concepts Relative to the Organization of Plankton. Oceanogr. Mar. Biol. Ann. Rev., 5:257-289. M'Closkey, R. T. (1972): Temporal Changes in Populations and Species Diversity in a California Rodent Community. J. Mamm., 53(4):657-676. McPherson and J. N. Krull (1972): Island Populations of Small Mammals and Their Affinities with Vegetation Type, Island Size and Distance from Mainland. Am. Midl. Nat., 88:384. Morse, D. H. (1971): The Foraging of warblers Isolated on Small Islands. Ecol., 52(2):216-228. Murdoch, W. W., F. C. Evans and C. H. Peterson (1972): Diversity and Pattern in Plants and Insects. Ecol., 53(5):8l9-829. Nie, N. H., D. H. Bent and C. H. Hull (1970): Statistical Package for the Social Sciences. MCGrawaHill Book Co. Odum, E. P. (1969): The Strategy of Ecosystem Development. Science, 164:262-270. Oldroyd, H. (1964): 'Thp_Natural Histopy‘pf.Flies. 'Weidenfield and Nicolson, London. Orians, G. H. (1969): The Number of Bird Species in Some Tropical Forests. Ecol., 50(5):783-801. Paine, R. T. (1966): Food web Complexity and Species Diversity. Am. Nat., 100(910):65-76. Parrish, J. D. and S. B. Saila (1970): Interspecies Competition, Predation and Species Diversity. J. Theor. Biol., 27(2):207-220. ll It'lllnlli 123 Pearson, D. L. (1971): Vertical Stratification of Birds in a Trapical Dry Forest. The Condor, 73:46-55. Phillip, J. R. (1957): Sociality and Sparse Populations. Ecol., 38(1): 107-11.10 Pianka, E. R. (1966): Latitudinal Gradients in Species Diversity: A Review of Concepts. Am. Nat., 100(910):33-46. Pianka, E. R. (1967): On Lizard Species Diversity: North American Flatland Deserts. Ecol., 48(3):333-35l. Pianka, E. R. (1969): Habitat Specificity, Speciation and Species —: Density in Australian Desert Lizards. Ecol., 50:498-502. Pielou, E. C. (1966): Species-Diversity and Pattern-Diversity in the Study of Ecological Succession. J. Theor. Biol., 10:370-383. Pielou, E. C. (1969): An Introduction to Mathematical Ecology. Wiley- Interscience. Pielou, E. C. (1972): Niche Width and Overlap: A Method for Measuring "' Them. Ecol., 53(4):687-692. Porter, J. W. (1972a): Predation by.Acanthaster and its Effect on Coral Species Diversity. Am. Nat., 106(950):487-492. Porter, J. W. (1972b): Patterns of Species Diversity in Caribbean Reef Corals. Ecol., 53(4):?45-748. Power, D. (1972): Numbers of Bird Species on the California Islands. Evol., 26(3):451-563. Preston, F. W. (1962a): The Canonical Distribution of Commonness and Rarity, I. Ecol., 43:185-215. Preston, F. W. (1962b): The Canonical Distribution of Commonness and Rarity, II. Ecol., 43:410.432. Pulliam, H. R., E. P. Odum and G. W. Barrett (1968): Equitability and Resource Limitation. Ecol., 49:772-774. Purchon, R. D. (1968): 'I_'h_e_ Biology 'o_f_t_hp_ Mollusca. Pergamon Press. Reiners, W. A., I. WOoly and D. B. Lawrence (1971): Plant Diversity in a chronosequence at Glacier Bay, AlaSka. Ecol., 52:55-69. Rickleffa, R. E. and G. W. Cox (1972): Taxon Cycles in the West Indian Avifauna. Am. Nat., lO6(948):l95-219. Rohlf, F. J. and R. R. Sokal (1969): Statistical Tables. W. H. Freeman and Company, San Francisco. 124 Rosenzweig, M. L. (1973): Exploitation in Three TrOphic Levels. Am. Nat., 107(954):275-294. Schoener, T. W. (1968): The Anolis Lizards of Bimini: Resource Partitioning in a Complex Fauna. Ecol., 49(4):?04-724. Schoener, T. W. (1969): Size Patterns in West Indian Anolis lizards; 1. Size and Species Diversity. Sys. Zool., 18:386-401. Shannon, C. E. and W. weaver (1949): The Mathematical Theopy pf Communication. University of Illinois Press, Urbana. Sheldon, A. L. (1968): Species Diversity and Longitudinal Succession in ., Stream Fishes. Ecol., 49(2):l93-198. Sheldon, A. L. (1969): Equitability Indices: Dependence on the Species Count. Ecol., 50(3):466-467. Sheppard, D. H., P. H. Klopfer and H. Oelke (1968): Habitat Selection: Differences in Stereotypy Between Insular and Continental Birds. Wilson Bull., 80(4):452-457. Shugart, H. H. and B. G. Blaylock (1973): The Niche-Variation Hypothesis: An Experimental Study with Dros0phila pOpulations. Am. Nat., 107(956)=575-579- Simberloff, D. S. (1969a): Taxonomic Diversity of Island Biotas. Evol., 24:23-47. Simberloff, D. S. (1969b): Experimental Zoogeography of Islands, 111. Ecol., 50(2):296-314. Simberloff, D. S. and E. O. Wilson (1969): Experimental Zoogeography of Islands 11. Ecol., 50(2):278-295. Slobodkin, L. B. and H. L. Sanders (1969): On the Contribution of Environmental Predictability to Species Diversity. In Diversity and Stability ip Ecological Systems. Brookhaven Sym. in Biol., No. 22, pp- 82-95- Sokal, R. R. and F. J. Rbhlf (1969): Biometry. ‘W. H. Freeman and Co., San Francisco. Soule, M. (1970): Response to Van Valen, L. and P. R. Grant. Am. Nat., 104(940):590. Soule; M. and B. R. Stewart (1970): The "Niche-Variation" Hypothesis: A Test and Alternatives. Am. Nat., 104(935):85-97. Spight, T. G. (1967): Species Diversity: A Comment on the Role of the Predator. Am. Nat., 101:467-474. Stephenson, J. (1930): The Oligochaeta. Clarendon Press. 125 Stonerook, Mrs. P. (1972): Personal communication. Teal, J. M. (1962): Energy Flow in the Salt Marsh Ecosystem of Georgia. Ecol., 43(4):614-624. Terborgh, J. (1971): Distribution on Environmental Gradients: Theory and a Preliminary Interpretation of Distributional Patterns in the Avifauna of the Cordillera Vilcabamba, Peru. Ecol., 52(1):23-40. Tramer, E. J. (1969): Bird Species Diversity: Components of Shannon's Formula. Ecol., 50(5):927-929. Vandermeer, J. H. (1970): The Communit Matrix and the Number of Species in a Community. Am. Nat., 104(935 :73-83. Van Name, W. G. (1936): The American Land and Freshwater Isopod Crustacea. Bull. Amer. Mus. Nat. Hist., 71:1-535. Van Valen, L. (1965): Merphological Variation and Width of Ecological Niche. Am. Nat., XCIX(908):377-390. Van Valen, L. and P. R. Grant (1970): Variation and Niche Width Re-Examined. Am. Nat., lO4(940):589-590. Vuilleumier, F. (1972): Bird Species Diversity in Patagonia (Temporate South America). Am. Nat., 106(948):266-271. Whitehead, R. R. and C. E. Jones (1969): Small Islands and the Equilibrium Theory of Insular Biogeography. Evol., 23(1):17l-l80. Whiteside, M. C. and R. V. Harmsworth (1967): Species Diversity in Chydorid (Cladocera) Communities. Ecol., 48(4):664-668. Wiegert, R. G. and D. F. Owen (1971): TrOphic Structure, Available Resources and Population Density in Terrestrial versus Aquatic Ecosystems. J. Theor. Biol., 30:69-81. Williams, E. E. (1969): The Ecology of Colonizing Species as Seen in the Zoogeography of Anoline Lizards on Small Islands. Quart. Rev. of Biol., 44(4):345-389. Willson,M. F. (1969): .Avian Niche Size and Morphological Variation. Am. Nat., 103:531-542. Wilson, E. O. (1969): The Species Equilibrium. In: Diversity and Stability iprcological Systems. Brookhaven Symp. in Biol., No. 22, pp. 38- 7. Wilson, E. O. and D. S. Simberloff (1969): Experimental Zoogeography of Islands, 1. Ecol., 50(2):267-277. Wilson, E. O. and R. W. Taylor (1967): An Estimate of the Potential Evolutionary Increases in Species Density in the Polynesian Ant Fauna. Evol., 21(1):1-1l. 126 Wilson, K. V. (1956): A Distribution-Free Test of Analysis of Variance Hypotheses. Psychological Bull., 53(1):96-10l. HICHIGR U NISTATE NIV. LIIBRARIES I "MINIMUM”“WNWWWI“HI 078 3129300 6487