THE EFFECTS 8F PATTERN 0N TKE RELATIVE VIEWS OF FOUR FIRST YEAR FELLOW FIELD SPECEES A Disscflahon Eur H19 Degree of DH. D. MICHIGAN STATE UMVERSITY Glenn Clinton Kroh 1975 mnuwmgmnwmmmmmM«~21; « 3 1291005 4124 v. €5.41“ (u A?) (it / ABSTRACT (”I (::/' THE EFFECTS OF PATTERN ON THE RELATIVE YIELDS OF FOUR FIRST YEAR FALLOW FIELD SPECIES By Glenn Clinton Kroh Two hypotheses were tested: first, that a diverse array of plants will out-yield a less diverse array of plants, on the same site with plant density fixed, and second, that when the pattern of positions of individuals of different species, in an array with fixed density and diversity, is changed, overall yield may be altered as a result of change in competitive stress among individual plants. The four Species used were Amaranthus retroflexus, Chenopodium album, Panicum capillare, and Setaria viridis. Mixtures containing four species each and pure stands of each species were grown. The mixture plots had fixed distri- butions of species in equal proportions but in different patterns. Interplant distance was 15 cm, giving an effective density of Sl plants/m2. At the end of the growing season, the plants were harvested, dried and weighed. Pure stands of Amaranthus produced higher yields than any of the mixtures. Yields among mixture plots with different patterns did not differ significantly. A second study was run con- currently with the above study to determine the feasibility of using results of a paired-species competitive ability experiment as a Glenn Clinton Kroh predictor of the outcome of different combinations of species. The five treatments consisted of competition from each of the four species plus a control. Each treatment was composed of a "target" plant surrounded by six competitor plants. Controls were single plants without competitors. The outcome of different combinations could not be predicted from the results of this experiment. Below ground biomass was sampled and estimations of root yields were lower than expected. However, data collected indicated that dicot roots suppress root growth of neighboring plants more than the monocots. THE EFFECTS OF PATTERN ON THE RELATIVE YIELDS OF FOUR FIRST YEAR FALLOH FIELD SPECIES BY Glenn Clinton Kroh AN ABSTRACT OF A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1975 ACKNOWLEDGEMENTS I gratefully acknowledge the guidance, time and support of my major professor, Dr. Stephen N. Stephenson. I also wish to thank Dr. William E. Cooper, Dr. William B. Drew, Dr. Erik D. Goodman and Dr. Peter G. Murphy for helpful suggestions concerning this research and manuscript. Others who contributed their time and effort in support of this research include John Barko, Frank Reed, Seth Reice, Teme Reice, Darlene Valasek, John Van Sickle, Steven Weiss and Jim Zimmer. Special thanks go to Patricia Paulus and Lynn Murry for their help in setting the experiment up in the field. Finally, I very gratefully acknowledge the enormous help given by Linda Hansen and my wife Karen in weighing the thousands of plant samples in the experiment. This research was financially supported by National Science Foundation grant GI-ZD. ii TABLE OF CONTENTS INTRODUCTION Diversity versus Net Primary Productivity (NPP) ....... Mixture-Yield Studies .................... Crap Mixtures versus Yield ................. Pattern versus Yield .................... Allocation of Site Resources Among Species ......... Intra- versus Interspecific Competition ........... Approaches to the Study of Plant Communities ........ MATERIALS AND METHODS Selection of EXperimental Species .............. Diversity-Yield Experiment ................. Competitive Ability Experiment ............... RESULTS EXperiment I - The Diversity-Yield Plots .......... Experiment II - Competitive Ability ............. DISCUSSION Diversity versus Yield ................... Allocation of Biomass Among Species in the Mixture Plots . . Pattern versus Yield .................... Competition Indices ..................... Experiment II - Pair-wise Competitive Ability Experiment . . Below Ground Biomass .................... CONCLUSIONS ........................... LITERATURE CITED ......................... APPENDICES ............................ iii Page mNm-b-P-N-H 10 TO 15 T7 21 TABLE Al A2 A3 A4 A5 LIST OF TABLES Mean yield per plot of treatments in eXperiment I, the diversity-yield experiment .............. Mean shoot weight of target plants in experiment II when surrounded by intra- and interspecific competitors Mean root weights of "target“ plants in experiment II when surrounded by intra- and interspecific competitors Importance values of each species within the mixture plots . Mean row weight of each species within the mixture plots . . Mean yields of target plants in grams when subjected to intra- and interspecific competition ......... Mean root yields in grams per row of each species when flanked by two rows of each of the mixture plot components, along with competition indices and competitive ability indices .............. Mean root yields of target plants for each species, along with corresponding competition indices and relative competitive ability indices .............. Analysis of variance and least significance range tables for comparisons of treatment plot yields in the diversity-yield experiment .............. Analysis of variance tables on competitive interactions within mixture plots ................. Analysis of variance tables on competitive interactions within mixture plots and their effect on root yields Analysis of variance and least significant range tables for the competitive ability experiment to measure the effect of competitive interactions on shoot yields ........................ Analysis of variance and least significant range tables for the competitive ability experiment to measure the effect of competitive interactions on root yields . . . iv Page 45 54 62 67 70 73 74 80 82 94 99 103 LIST OF TABLES (Continued) Table Page A6 Calculations of Chi-square values to determine goodness of fit of experimental data with data estimated to fit a geometric progression series ......... l07 Figure 10 ll 12 LIST OF FIGURES Four basic patterns of plants in the mixture plots . . . . Mean yields of total above ground biomass per plot in the diversity-yield experiment ........... Relative contributions to total plot yields by component Species of the mixture plots in the diversity-yield experiment ..................... Mean yield per row of Amaranthus (‘) and Chenopodium (.) when flanked by two rows of each of the mixture plot components ................... Mean yield per row of Panicum (‘) and Setaria (.) when flanked by two rows of each of the mixture plot components ................... Diagrammatic representation of statistical differences in mean shoot yields per row of diversityzyield plot species when interfaced by rows of different competitors ..................... Estimated mean root yield/row of Amaranthus and Cheno- odium when between two rows of eaCh of the otfier mixture plot species ................ Estimated mean root yield/row of Panicum and Setaria when between two rows each of tfie other mixture plot species .................... Diagrammatic representation of statistical differences in mean root yields of diversity-yield plot species when interfaced with different competitors ..... Mean shoot yield of Amaranthus target plants when encircled by intra- and interspecific competitors ....... Mean shoot yield of Chenopodium target plants when encircled by intra- andinterspecific competitors . . Mean shoot yield of Panicum target plants when encircled by inter- and intraspecific competitors ....... vi Page l3 19 23 25 27 29 31 33 35 37 39 41 Figure 13 14 15 16 17 18 19 20 21 Mean Mean Mean Mean Mean LIST OF FIGURES (Continued) shoot yield of Setaria target plants when encircled by intra- and interSpecific competitors . . root yield of Amaranthus target plants when encircled by intra- and interspecific competitors . . root yield of Chenopodium target plants when encircled by intra- andinterspecific competitors . . root yield of Panicum target plants when encircled by intra- and interspecific competitors . . root yield of Setaria target plants when encircled by intra- and interspecific competitors ... Distribution of importance values of the component species of mixture plot (CAPS) ........... Distribution of importance values of the component species of mixture plot (SACP) ....... w. . . . Distribution of importance values of the component Species of mixture plot (PACS) ........... Distribution of importance values of the component species of mixture plot (RANDOM) .......... vii Page 43 47 49 51 53 58 59 60 61 INTRODUCTION Diversity Versus Net Primary Productivity_(NPP) One of the most important parameters of a community iS the net primary productivity (NPP) (Woodwell and Whittaker, 1968). This is the amount of biomass produced by the plant component of the community, which will be the energy source for the heterotrOphic organisms. Robert McIntosh (l970) commented, "The conventional wisdom of ecologists is that diversified utilization of site resources should result in greater productivity and efficiency." In other words, niche specialization would not only allow more species to coexist in an area, but also enable them to physiologically and morphologi- cally eXploit a resource base more thoroughly. Depending on the type of community present, one could judge its success at utilizing the resources of a particular site by resultant productivity, yield, viability of seeds, or persistance of the community through time. As yet, only one study (Werner, 1972) has indicated that increase in Species number in a particular plant array increases the produc- tivity or yield of a particular Site. In an ecological sense, diversity of a community relates to the richness and evenness of species within that system. Richness refers to the number of species or functional groups while evenness is a measure of the relative proportions of those species or groups 2 within the community. Evidence strongly supports the hypothesis that diversity is inversely related to NPP within a given community (Margalef, 1963; McNaughton, 1967, 1968; Stephenson, 1973). When Site potential is increased, planktonic communities respond with increased NPP and decreased diversity (Margalef, 1963; Patrick, 1949). McNaughton (1968) demonstrated that the most productive sites in annual California grasslands have lower diversities than less produc- tive sites. In a study on the community dynamics of a mid-Michigan oldfield, Stephenson (1973) utilized NPP as an experimental variable by enhancing site potential through enrichment of the soil with fertilizer treatments. The most productive arrays of plants exhibited the lowest diversities. Mixture-Yield Studies One approach to studying the relationship of diversity to productivity in plant arrays is to conduct mixtureayield experiments. Literally hundreds of pair—wise competitive ability experiments have been completed with plant populations (see reviews by Donald, 1963; Harper, 1961; McIntosh, 1970). Watt (1964) suggested it was possible to rank a group of plants by their respective competitive abilities and consequently predict the outcome of different combinations. Paired-species experiments have shown that mixture yields fall in between those of the highest and lowest yielding pure stands of the component species (Donald, 1963). A clear exception to these results was demonstrated by Whittington and O'Brien (1968). They stated that failure of other studies to demonstrate mixtures that were higher yielding than pure stands of the component species was probably due 3 (1) to use of species that were not suppressed a great deal by intra- specific competition, (2) use of Species not differing significantly in their patterns of root growth, (3) use of shallow pots or boxes that restricted root growth and finally (4) use of Species in which one or both were suppressed greatly through interspecific competition. Their study explored the competitive interactions of rye grass, meadow fescue and a triploid of rye grass. Under certain clipping treatments, mixtures yielded more biomass than pure stands. Plant mixture studies involving more than two species have shown that mixtures do not out-yield pure stands of the most productive Species in the mixtures. Bornkhamm (1961) ran a mixture study using white mustard (Sinapsis alba), corn cockle (Agrostema githago), cheatgrass (Bromus secalinas) and pimpernel (Anagallis arvensis). He grew these Species in pots as pure stands, two-species mixtures, three-Species mixtures and four-Species mixtures. All mixtures were equi-prOportionate, with regard to the separate species. Pure stands of white mustard out-yielded all of the mixture plots, while pure stands of pimpernel had the lowest yield in the experiment. Another diversity-yield experiment (Haizel, 1972), using three species, had similar results to those of Bornkhamm (1961). Haizel planted barley (Hordeum vulgare), white mustard (Sinapsis alba) and pappy (Papaver rhoeas) in pure stands, two-species mixtures and three-Species mixtures. He repeated the experiment, using wild oats (Avena fatua) in place of p0ppy. In his summary of the experiment, he says “On no occasion did the yield of any of the mixtures exceed the highest yielding Species in pure stand." These studies have not demonstrated that NPP increases with diversity. Crop Mixtures versus Yield There is very sparse evidence concerning greater yields in mixtures of craps than yields in monoculture (Loomis, 1971; Rhodes, 1970). Loomis feels a synergistic relationship among two species of crops seems unlikely, with regard to use of available light, the principal limiting factor in intensive agriculture. He argues that since a single species can be grown at a sufficient density to com- pletely intercept available light, the resource benefit cannot be increased by establishing a mixture of plants with "diverse stature and leaf display". It would seem if mixtures of crops did, in fact, commonly outeyield pure stands, that the practice of multiple-crop farming would be in frequent use today. In regions where hand implements are used at sowing and harvest, man has developed mixed cropping (Donald, 1963). Examples can be found in Ceylon where pastures or bananas are grown beneath coconut palms; in the Mediterranean where wheat is planted among olive plants or cork oaks and in Greece where alternate rows of cotton and corn are sewn, under irrigation. There does not seem to be data available concerning yields of these mixtures versus those of the component species in pure stands. Pattern versus Yield Altering the arrangement (pattern) of individuals within a given plant array may affect the NPP and resultant yield of that array. When studying the inter-varietal competitive abilities of barley and rice, Sakai (1957) develOped and used an experimental design incor- porating the use of a six-member competitor ring. One plant variety 5 was individually planted and encircled by (1) six plants of the same variety, (2) five of the same and one of the other, (3) four of the same and two of the other, and so on, until all six plants were of the other variety. He found a linear relationship between the yield of the target (center) plant and the number of plants in the ring of the variety with the best competitive ability. Harper (1961) did an experiment using Bromus rigidus and §;_madritensis at equal densities. Five plant patterns of the two Species were used. Individual plants were sown into hexagonal patterns that allowed (1) each plant to be surrounded by three of its own species and three of the other, (2) two of its own Species and four of the other, (3) four of its own species and two of the other, (4) random placement of each species on the hexagonal pattern and finally (5) a completely random pattern in the pot.' Total yield was not significantly affected by changes-in the pattern. However, there were definite differences in the preportional contributions made by the two species to the total biomass. Further, contrary to the findings of Sakai (1955), there was not a linear relationship between the suppression of growth of §;_madritensis and the number of §;_rigidus plants in the surrounding hexagon. Allocation of Site Resources among Species In natural plant communities, the allocation of available site resources is rarely equally distributed among the different species populations. Relative success of a species at utilizing the available resources of a particular site can be quantitatively expressed by an importance value (Cox, 1967). Importance value refers to the sum of the relative frequency, relative density, and relative dominance of 6 a Species within a particular array of plants. Whittaker (1969) suggests that if it is assumed that there is some correspondence between the share of the community's resources a species utilizes, the share of the niche space it occupies, and the share of the community's productivity it realizes, then relative importance values can be expressed as relative niche sizes. There are three hypotheses on how niche space is divided to produce the distribution of produc- tivity among species within an array. The random niche boundary hypothesis, as pr0posed by MacArthur (1957, 1960), states that the boundaries of niches are located at random in niche hyperspace. This type of distribution is found in some animal communities (King, 1964; Hairston, 1964) and in particular with territorial birds (MacArthur, 1960). The second is the 109 normal hypothesis put forth by Preston (1948, 1962). Basically, it says that Species importance values are determined by independent variables and that Species importance will then approach a normal frequency distribution for which a logarithmic scale of importance is apprOpriate. Communities that are rich in species generally exhibit this type of distribution. The niche pre-emption hypothesis (Whittaker, 1965, 1969) states that the most dominant Species pre-empts a given fraction of the total niche Space. The second most dominant species then takes a similar fraction of the remaining space and the third most dominant Species takes a similar fraction of the remaining niche space not utilized by the first two species and so on down to the last Species in the community. This distribution seems to be characteristic of communities with low species richness. 7 Intra- versus Interspecific Competition In this study, the term competition is used in the manner that Harper (1961) uses the term interference to mean "...those hardships which are caused by the proximity of neighbors..." In this context, "competition" will include both the effects of competition for limited resources and allelOpathy among species populations. All plants utilize light, water, nutrients and carbon dioxide throughout their life cycles. Individual plants within each papula- tion have genetic potentials for growth and reproduction. This potential may or may not be realized, depending on the site resources and biological interference from neighboring plants. Plants of the same species tend to utilize, both spatially and physiologically, Site resources in the same way. In other words, plants of the same species acquire the same nutrients in the same amounts from the same regions of a particular site. On the other hand, individuals of different species presumably utilize a site in different ways. The term annidation has been used by Ludwig (Ludwig, 1n_Harper, 1967) to describe the evolutionary process of direct selection for some difference in niche occupancy of Species within a community. Since individuals within a species are thought to occupy the same niche, it is thought that intraspecific competition is more severe than interspecific competition. It is thought that this may be one of the most important underlying mechanisms for control of population sizes in a community (McIntosh, 1970). Several studies, run with two species mixtures, have shown that intraspecific competi- tion exerts a disproportionate mortality rate on the population with the highest number of individuals (Population I). Resultant reduction 8 in size of Papulation I reduces interspecific competition stress and allows the previously recessive population (Population II) to become the largest in the mixture. When this occurs, P0pulation II now experiences diSproportionate mortality from intraSpecific competition with a resultant decrease in numbers and a corresponding increase, once again, in the Size of Population I. This self-stabilizing process has been shown to occur from year to year in a grass-clover association (Leith, 1960). Approaches to the Study of Plant Communities Many investigators feel that studies of communities, as a whole, are essential to understand the inherent dynamics of the system (Waddington, 1961, 1965; Slobodkin, 1962; Orians, 1962). Rather than the "holistic" approach, Lewontin (1968) prefers an analytical approach to get at the mechanisms of a system. Complex interactions are broken down into simpler components that lend themselves more readily to experimentation than does the whole. As most plant communities are difficult, if not impossible to manipulate experimentally, the latter approach to probing complex interactions of a community is the most pragmatic. Community ecologists generally believe that competition is an important factor in community dynamics (Poore, 1964; Watt, 1964; Major, 1958). Watt (1964) suggested that it was possible to rank a group of plants by their reSpective competitive abilities and consequently predict the outcome of different combinations. Harper (1967) felt that "ecological combining prOperties" of Species might be examined in all possible combinations of pairs and their yields 9 compared with that in pure stands. He felt that in this manner, dominance could be detected and bring "... to experimental synecology a refinement and subtlety appr0priate for a science which has outgrown its qualitative and descriptive youth.“ Although the analytic method will not solve all community level questions, it very well could lead to important insights into those problems. The analytical approach is used to investigate the relationships discussed in the preceding sections. In general, I have studied the effects of intra- and interSpecific competition on above and below ground plant yield. More specifically, I investigated: 1. The hypothesis that productivity and resultant yield are positively related to the diversity of a given system. 2. The hypothesis that altering the arrangement (pattern) of individuals, within a given plant array, will afféct the NPP of that array. 3. The expression of dominance and resultant allocation of yield among species within a given plant array. 4. The feasibility of using results of a paired-species competitive ability experiment to predict the relative performances of more than two species, when grown together. MATERIALS AND METHODS Selection of Experimental Species An ideal system for investigation of the diversity-yield relationship is the fallow field community. Annual herbs that complete their life cycles in one growing season are dominant plants in this type of comnunity. Since there are no perennial organs, total above and below ground biomass can be detenmined. Mid-Michigan first year fallow field communities contain plant components that are easily manipulated. They are "r" selected colonizers that consequently have highly overlapping niches, providing for greater interference (Odum, 1969). Shoots of these plants exhibit a very flexible growth reSponse to interference from neighboring plants. Since the dominant Species are herbaceous annuals, all plants are discrete entities that complete their life cycles in one growing season. The use of biennials and perennials, characteristic of later successional stages, was prohibitive, as they are prone to vegetative reproduction through formation of rhizomes and stolons. Since two or more plants can be attached physiologically, perennial plants cannot be considered as discrete entities. It was with this in mind that annual herbs from the first year fallow field community were used. Diversity-Yield Experiment Two eXperiments were designed and implemented. In both experi- ments, two dicotyledonous plants, Amaranthus retroflexus and Chenopodium 10 11 album, as well as two monocotyledonous plants, Panicum capillare and Setaria viridis, were used. They were selected on the basis of their representativeness of typical plant forms in the first year fallow field communities as well as differences in structure and possibly function of above and below ground biomass. Experiment I involved growing plots of mixtures containing four species and pure stands of each species. Yields of the mixture plots and pure stands were compared and analyzed statistically using a randomized block experi- mental design, replicated three times. In each block, four plots were monocultures and four were mixtures with fixed distributions of species in equal proportions but in different patterns (Figure 1). Each mixture plot originally contained 128 plants in eight rows of sixteen plants each. The number of plants in the patterns changed throughout the experiment due to mortality of plants originally sown, as well as invasion of plants from the natural seed pool. Interplant distance was 15 cm, giving an effective density of 51 plants/m2. This density was determined by observation of natural fallow field communities which subsequently allowed a qualitative decision concerning the interplant distance to be used in this study. Two rows of each of the four species were in each of the mixture plots. Species were arranged to allow interfacing of all possible combinations of the four Species. In addition to the three possible spatial arrangements of rows within the mixture plots, a fourth plot, using a randomized pattern was used. Positions of individual plants of each Species within this plot were randomly assigned. As all plots had 128 plants each, a single replicate contained 1024 plants. 12 Figure 1. Four basic patterns of plants in the mixture plots. 13 zooz<¢ F mmame av magma . 8v 3% A com: zugc< “caps .ucmswgwaxm u—mwzuxqugm>Mu use .H ucmswcqum :0 mucmspmmea Mo popa emu u_m_a cum: .— m_nmh 21 to lowest, Amaranthus, ChenOpodium, Panicum and Setaria. The analysis of variance table and least Significant range table are in Appendix A-l. Relative contributions to mixture plot yield by the component species exhibit definite and repeatable trends. Heirarchy of relative contributions to yields of mixture plots was Amaranthus > ChenOpodium.> Setaria > Panicum, except in the mixture plot (SACP) where Panicum contributed more above ground yield to the array than Setaria (Figure 3). Within the plots, Amaranthus was generally the best competitor while Setaria was the poorest (Figures 4-5). Lowest Shoot yields per row occurred when the species were between two rows of Amaranthus. Significantly higher Shoot yields per row were attained when they were interfaced with the grasses. ChenOpodium had a significantly greater effect than Panicum on repressing shoot yields of all species except Amaranthus. There was no difference between effects of Panicum or Chenopodium on the mean yield of Amaranthus (Figure 6). Analysis of variance tables are in Appendix A-2. Mean root yields of species within plots, except for Amaranthus were significantly lower when subjected to competition from the dicots as compared to the effect of the grasses. Relative root yields are shown in Figures 7 and 8 and statistical relationships are represented in Figure 9. Analysis of variance tables are in Appendix A-3. Experiment II - Competitive Ability In all cases, the dicots, Chenopodium and Amaranthus, were the most effective competitors (Figures 10 - 13). Target plants of all Species attained the least shoot yield when surrounded by either of the dicots. In all cases, the dicots were stronger competitors than 22 Figure 3. Relative contributions to total plot yields by component species of the mixture plots in the diversity-yield experiment. m $5.8M 3D§0§§fiag 9%.Z 23 .0 q 0 / . 00 30000 W - 00 00 -. 00 c a M i .. 00 anES ” n o: 7 too 0131* 101d 1v101 JO Iwaouad 24 Figure 4. Mean yield per row of Amaranthus (A) and Chenopodium (.) when flanked by two rows of each of the mixture plot components. 25 A O .. Av J A. g. . A O L A o . P n n b b b n m m m m m m m 1; 1; 1; M02<000 30¢ «00 o000> z<0z zonz Poem z boom z<0z ame---l d : d 4. - lava ‘ L l I 1 $3 8% 53 52 F1 (SNV89) lNVWd 83d 0131A lOOHS NVEN ioutwog VIUVISS WflDINVd NflIGOdONBHj SOHINVUVWV COMPETITORS FIGURE 11 40 Figure 12. Mean shoot yield of Panicum target plants when encircled by inter- and intraSpecific competitors. Vertical bars represent i two standard errors. 41 I—o—l - :— 4 4 - l—O—l - l-o-l- 1.1.. a_ a I l L O O O O O C) Ln :1- M N H (swvuo) 1NV1d 33d 0131) IOOHS wvaw 1081N03 VIBVIBS WflDINVd wnlooaowawg snwtwvuvwv COMPETITORS FIGURE 12 42 Figure 13. Mean Shoot yield of Setaria target plants when encircled by intra- and interSpecific competitors. Vertical bars represent 3 two Standard errors. 43 (swvuo) LNVid aad 0131) iOOHS wvaw l---4D---1 1—-—o-——1 4 l---Cr---1 - |-—-o——-l- r---4l---4 '- MIIILJIIA 22. :3 .91 °° =' 0 1031N03 VIUVLBS wnoINvd wn1ooaow3H3 SflHlNVUVWV COMPETITORS FIGURE 13 44 the grasses. The statistical relationships of "target" plant shoot yields under different treatments are shown in Table 2. AS with the shoots, the root yields of "target" plants of all species were repressed more by the dicots, Amaranthus and Chen0podium, than by the grasses (Figures 14 - 17). In each case, root yields of "target“ plants surrounded by either of the dicots, were not signifi- cantly different (Table 3). Similarly, the effects of the competitor rings of either Panicum or Setaria on root yields of all Species were not Significantly different. Analysis of variance tables and least significant range tables are in Appendix A-S. 45 ..0000. ..050 0500 us» an 00000006550 umzo—.om men 00:» 00 “00.00000 ».u=0uw$_:m00 no: 020 0:00: a 00.0. 0 00.00 00.0 0 00... ...0 00.00 00.0 00 00.00 .000000 00.. 00 00.0. 0.0 0 ...00 00.0. 00 00.00 00.0 00 00.00 0.00000 0... 0 ...0. 00.0 0 00... 00.0. 0 00.00 00... 00 0..00 000.000 00.. 0 0..0 00.. 0 00.0 00.0 0 00.0. 00.0 00 00.0. 00.00000000 00.. 00 00.0 00.0 0 00.0 00.0 0 0.... 0..0. 00 00.0. 0000000000 00 .0: .0>0 00 .0: .0>0 00 .0: .0>< 00 .0: .000 0.0000m 5:00:00 sawuomo=mgu masucmems< 00:0.0 ammcmp meou.0masoo .meouwumaeou owmwu00000000 0:0 1000:. 00 umcczoeeam 00:3 W. ucmswgqum =0 00:000 pomemp mo usmww: 00600 :00: .0 0.000 46 Figure 14. Mean root yield of Amaranthus target plants when encircled by intra- and interspecific competitors. Vertical bars represent i two standard errors. 47 F—-—-—< mm .03 .m>< mm .0: .m>< mm .0: .m>< 0.00umm 000.000 03.:oao:m:u 0::0:0000< 00:0.0 women» 0000.000000 .00o0.0masou u...umamemu:. 0:0 1000:. 00 000:0oee00 00:: .. 0:00.:maxu :. 0000.0 0000.000 .0 0000.0: poo: :00: .m 0.00. DISCUSSION Diversity versus Yield One of the primary objectives of this study was to determine whether a mixture of different species can better exploit the resources of a particular site than pure stands of the component species of the mixture; in other words, whether diversity, in a given system, is related to NPP in a positive way. The experimental variables were diversity and plant array patterns. Site potential (nutrients, water, etc.) was held constant and treatment effect was gauged by the magnitude of the resultant yield at the end of the growing season. Plant species incorporated in this study normally coexist in natural plant communities. Further, the experimental plants were grown under natural field conditions, as Opposed to other mixture-yield studies (Bornkhamm, 1961; Haizel, l972) in which potted plants, grown in greenhouse conditions, were used. It is for the above reasons that I feel results of this study can be extrapolated to actual field dynamics of the species used. Although results of experiment I showed that mean yield per plot of pure stands of Amaranthus (4090.65 grams) was consistently greater than mean yields of any of the mixture treatments, statistically, there was no significant difference. Lowest mean yield per plot was that of Setaria (2368.25 grams) in monoculture. Thus mixture yields were comparatively lower than that of the highest yielding monoculture 55 56 (Amaranthus) and comparatively higher than the mean yield of the lowest yielding monoculture (Setaria). These results concur with those of Bornkhamm (196l) and Haizel (1972), with regard to the inability of mixtures to out-yield the highest yielding monoculture. It should be noted that although the mean yield per plot of pure stands of Amaranthus was higher than those of all other treatments, in a statistical sense, it did not have a significantly higher (a = 0.05) yield than any of the other treatments except for pure stands of Setaria. 'Perhaps life fonm differential among the component species used in this study was not sufficient to allow them to exploit site resources in different ways. Even though the root growth patterns of the dicots and monocots differed significantly, they still pene- trated to the same depth of the rhizosphere and presumably, therefore, acquired soil nutrients in the same area. In this respect, biennial or perennial species, representative of later successional stages might be better organisms fbr this type of study because they exhibit root systems that range from fibrous sub-surface roots in some grasses to deeply penetrating tap roots of some biennial and perennial dicots. Whittington and O'Brien (1968) feel that lack of diversity in root growth patterns of Species in mixtures is the major reason that those mixtures fail to attain yields higher than the highest yielding monoculture of the component species. Another possible reason for failure of the mixture plots to out- yield the highest yielding pure stand was due to the use of an arbitrarily chosen fixed-density. Presunably, the lowest density at which a species produces the maximum yield per unit area is species 57 specific. Determination of this density for each species could be accomplished experimentally and then used as a base line for the density of each species in mixture plots. In this way, mixture plots could be designed that would exhibit minimal intraspecific competition stress. Of course, utilization of Species with different growth patterns would enhance the "combining ability" of the species through minimizing interSpecific competition among the Species. Allocation of Biomass Among Species in the Mixture Plots A good measure of how species utilize available resources is how the resultant yield is allocated among them. Since relative densities and frequencies of Species in experiment I are all equal, the relative importance values among the Species are equal to their relative yields. Distribution of their yields generally follows a geometric series pattern, indicating a strong expression of dominance within the arrays. In all mixture plots, except (CAPS), the distribution of total yield among the Species is not significantly different than that predicted by a geometric series curve (Figures l8 - 2l). Distributions were compared to that predicted by a geometric series with a Chi-square test, using an a of 0.05 (Appendix A-6). In the (CAPS) plot, gthg; pggjgm_was directly interfaced with Amaranthus and therefore was subject to relatively high competition stress. Although the distri- bution was not a geometric series, the hierarchy of relative yields was the same as in other plots. Overall, Amaranthus, the most dominant Species, yielded about 55% of the total biomass, based on dry weight yield. Chenopodium, the second most dominant Species, produced about half the remaining biomass or about 25% of the total. Finally, the two grasses each produced about 12% of the total yield (Table 4). 58 60 " A 50 \\ "'""' PREDICTED 40 _ \ GEOMETRIC \ SERIES CURVE g \\ .1 g 30 ' \ 3 \ E \ S 20 r \ m \ :5 A \ \A ‘\ \ A ‘\ 10 - \ g . J l l + l 2 3 ll SPECIES RANK Figure 18. Distribution of importance values of the component species of mixture plot (CAPS). 59 60 - O 50 b \ -"'"" PREDICTED \\ GEOMETRIC 40 " \ SERIES CURVE \ 30 - \ \. \. .\ \. $.20 " \ £5 ‘\ “J ‘\ g; ‘\ .... E ‘\ £5 ‘\ 10 - \ 9 - \. 8 P *\ 7 \ __ \ 6 .\ 5 [ Q l I J L l 2 3 ll SPECIES RANK Figure 19. Distribution of importance values of the component species of mixture plot (SACP). 60 60 " . 50 P \ """ PREDICTED \ GEOMETRIC “0 " \ SERIES CURVE S \\‘ g 30 T \ Z \ :2 '\ .1 '\ E 20 - \ S \ ‘q \. ‘\ 10 - \ C 9 - \ 1 l l I l 2 3 LI SPECIES RANK Figure 20. Distribution of importance values of the component species of mixture plot (PACS). 61 ll 50 - ‘\ \\ """PREDICTED 140 " \ . GEOMETRIC Lg \ SERIES CURVE 2’ 30 __ \ i: \ U f: a‘\ E \ g 20 r- \ :0 ‘\ \ \ ‘\ \3\ 10 .. I \ 9 h- \ J I l \ '! l 2 3 A SPECIES RANK Figure 21. Distribution of importance values of the component species of mixture plot (RANDOM). 62 oo— x Auopm L00 vpmw> P0u0h\uopm can v~0P> mowumamv u u:4<> muz 00:00:005H .0 0—000 63 Allocation of site resources in a natural community can be expressed by the relative importance values of the component Species. Distribution of the importance values is thought to be a direct indication of how they divide up the niche hyperspace of a community (Whittaker, l970). The niche pre-emption hypothesis (Whittaker, l965, 1969) states that the distribution will approach a geometric series. Communities in which dominance is strongly develOped and the number of species is small usually exhibit this type of distribution. One example is the mid-Michigan first year old field community (Stephenson, 1973) in which Amaranthus, Setaria, Panicum and Chenopodium were strong dominants, in that order, based on their NPP. It is interesting that a geometric series curve describes the allocation of site resources among the species in experiment I as well as among the Species in the type of community where they occur in nature. The two grasses, Panicum and Setaria were both more dominant than Chenopodium in the natural community (Stephenson, 1973). Con- versely, in the mixture plots of this study, Chenopodium was clearly dominant over the grasses. Constraints imposed by the design of experiment I such as holding the density and plant number per Species constant as well as watering to augment natural precipitation, may have released Chenopodium from competition stress normally experienced in the fallow field community allowing it to increase its importance as a dominant Species. One constraint, with regard to my study, was the synchronous germination of the component species. I feel this germination synchrony was justified, with regard to the real fallow-field community, as the four species studied herein also germinated synchronously in the 64 fallow field on the periphery of my experimental plot. However, it would have been interesting to see how staggered sowing of the component species would have affected the relative contributions of the component species to total mixture plot yield. Harper (l961), in a study with Bromus rigidus and §;_madritensis, measured their relative yields in mixtures when each species was sown at different times. He found there was no detectable difference in total yield per pot of differently timed mixtures. However, relative contributions of each Species to total yield was greatly altered. It is possible, therefore, that some other species, in my experiment, would have been dominant had the germination times of the component species not been synchronous. Whether the total yields of the mixtures would have been different, can only be answered through further studies. In natural fallow-field communities, productivity of the whole system is allocated among component Species in a temporal sense. Reed (personal communication), in a study on the productivity of a first year fallow-field community in mid-Michigan, demonstrated that the plant array contains functional groups, based on the type of photosynthetic pathways the plants possess and whether they are monocotyledonous or dicotyledonous plants that contribute differentially to the total array NPP throughout the season. Monocotyledonous plants possessing the C3-Calvin cycle photosynthetic pathway (cool season grasses) and dicotyledonous plants with the C3 pathway are the major contributors to the productivity of the system during June and July. In late summer, dicots with the C4 - dicarboxylic acid pathway of carbon fixation, are the dominant producers, with NPP rates five times greater than any other functional group. Finally, at season's end, exhi rela tOS pro con inc dii is ”m SD fo nu ne dc ti /5' ha be 65 end, between September and October, the perennial C3 grasses exhibited a resurge in NPP and subsequently produced biomass at a relatively high rate until late September when cool weather started to slow down the photosynthetic machinery. Pattern versus Yield Results of experiment I do not indicate that the net primary productivities of the mixture plots, with diversity and density held constant, are significantly changed when the arrangement (pattern) of individuals within that array is altered. However, mean yields of different mixture plots range from 2683.7 grams to 3701.9 grams. It is possible that with more replication significant differences may have been indicated. The proportional contributions of the separate species also remained unchanged. Also, it is possible that the life forms or physiological processes (i.e. rates of photosynthesis, nutrient uptake, proportions and absolute quantities of nutrients needed per unit of biomass produced, etc.) among the component species do not differ enough to give a gradient of combining ability among them. It Should be noted that it is beyond the scape of this study to determine what, if any, are the physiological differences among the Species used. Competition Indices In my diversity-yield experiment, the dominant component, Amaranthus, was suppressed most by intraspecific competition. Domi- nance is not necessarily directly related to competitive ability. Species in a community could be dominant and at the same time only be realizing a small pr0portion of their genetic potential for 66 productivity. Conversely, species could conceivably be realizing their full potential for productivity and still be a minor contributor to community NPP. In other words, if Species A had a genetic potential to produce 100 grams, but only yielded 40 grams and Species B could potentially produce 20 grams and, in fact, did yield 20 grams, then species A, as a producer, would be dominant to species B but at the same time, species B would be the better competitor. In order to more Clearly demonstrate the intra- and interspecific relationships of the species in mixture plots, and to "equalize" the inherent weight differential among Species, a competition index (CI) was developed (Table 5). The mean row weight of each species in pure stands was used as a base line to determine whether a species was affected (suppressed) more by intra- or interSpecific competition. The CI is calculated by dividing this base-line mean row weight for each species, into mean row weights attained when those species grew between two rows of each of the other Species. Consequently, the intraspecific CI is always one, by definition. Any situation in which the index is greater than one indicates that interspecific competitive stress, from that particular competitor, is less than stress from intraspecific competition. Conversely, any Situation where the CI is less than one, indicates that interspecific competition stress is greater than that exerted through intraspecific competition. Amaranthus rows yielded more in mixture than in pure stands. Chenopodium produced more yield per row in mixture plots than in pure stands when it was in between rows of grasses. However, when Chenopodium was in mixture plots where it was between rows of Amaranthus, it yielded less per row than when in pure stands. Both Panicum and Setaria were more suppressed by 67 0000000 0003000000 00:0 :00 0.Hu 0:0 00 53m u x00:H :0_00:< 0>000000500 030300t0 00:0500000 0 :w.00ww00m 0:».00 0:000: 300 :00: 0:00m 00:0 :0 00000mw 0 mo psm003 300 :00: 0 u :000 000:0 :00000000000 00.0 00._ 00.0 00.0 00.0 00.: 000:: :0000000000 00.000 00.000 _0.00_ 00._00 00.0__ 0:0_0: 30: :00: 00:000m 00.0 00.0 00.0 00.0 00.0 00.0 x00:0 :0000000000 00.000 00.00_ 00.000 00.000 00.000 0:0_0: 30: :00: 5:00:00 00.0 00._ _o._ 00.0 00.0 0_._ x00:0 :0000000000 00.000 00.000 0_.000 00.000 00.000 0:000: :00 :00: Eswuo0oc0:u 00.0 :0._ 00.0 _0._ 00._ 00.0 0x00:0 :0000000000 00.__0 00.000 00.000 00.000 00.000 0:000: :00 :00: 0::0:0:0E< 00:0 0000 :00::: 0000 0000 A0000 x00:0 :0__0:< ««.anu m>wpmpmm 00000 000000: 00000 0::0x02 :0:0_z 00_000m .00000 00:0x05 0:0 :wzpwz 0000000 :000 00 0:000: 30: :00: .m 00:00 68 interspecific competition than by intraspecific competition. Relative competitive abilities of each species in the mixture plots was determined by summing the competition indices of each Species to give values that describe overall performance of each species. This value is the relative competitive ability index (RCA). A high index indicates a species is a good competitor. Based on their respective RCA's, the ranking of the component species, with regard to their competitive abilities, is Amaranthus > Chenopodium > Setaria > Panicum. Amaranthus and Chenopodium were the best competitors of the four component species. Obviously these dominant species exhibited a greater rate of growth which is a function of many mechanisms (photo- synthesis, root uptake rates, etc.). It is beyond the scope of this study to determine which mechanisms were most important in allowing Amaranthus and Chenopodium to be the dominants. Walter (197l) states that, "Where the growth eycles of individuals in a population are synchronous and light is nearly fully intercepted, victory goes to the individuals which achieve height quickest whether by producing (l) taller leaves (as in grasses), (2) longer petioles, (3) taller stems or (4) more perennial stems." All components were sown at the same time and germination was close to being synchronous. Chenopodium germinated about f0ur to five days after the other species. It appeared that the competitive advantage was acquired by the two dicots by virtue of their fast growing vertical stems. 69 Experiment II - Pair-wise Competitive Ability Experiment For experiment II, each species was subjected to intraspecific and interSpecific competition. Results indicate that the dicots were superior competitdrs to the grasses, under conditions dictated by the experiment and that particular growing season. In a statistical sense, ranking the four species as to their competitive abilities was not possible since the effects of Amaranthus and ChenOpodium on the target plants were not significantly different (a = 0.05) and a clear trend was not apparent. The same is true for the grasses. About all that can be said statistically is that the dicots were significantly better competitors than the monocots; as was the case in the diversity- yield experiment. Another way to look at the competitive abilities of the species used in this study is to calculate competition indices (CI) as was done in the diversity-yield experiment. Mean yields of target plants, when ringed by their own species, were used as base-line values. For each species, the mean yield of the target plants in each treatment was divided into the mean base-line yield to give the CI for each treatment. These values were then added to give a relative competitive ability (RCA) index for each species. Ranking was then accomplished, using the RCA for each species as the criterion. The greater the RCA, the higher the rank. Relative competitive abilities of the species were, from highest to lowest, Chenopodium > Amaranthus > Setaria > Panicum (Table 6). This is not the same sequence found with the relative competitive abilities of these species in the diversity- yield experiment (Table 5). It, therefore, does not support the suggestion of Matt (l964) concerning the predictability of pair-wise competitive ability experiments. 70 Table 6. Mean yields of target plants in grams when subjected to intra- and interSpecific competition. Target Plants Competitors RCA** Amaranthus Chenopodium Panicum Setaria Amaranthus Mean weight 19.48 14.83 39.17 39.76 CI* 1.00 0.76 2.01 2.04 5.81 Chenopodium Mean weight 11.18 10.90 66.06 53.33 CI 1.02 1.00 6.06 4.89 12.97 Panicum Mean weight 3.72 5.52 17.03 23.71 CI 0.21 0.32 1.00 1.39 2.92 Setaria Mean weight 6.87 6.13 15.11 12.99 CI 0.52 0.47 1.16 1.00 3.15 Mean yield of target plants when encircled by plants of its own Species ) Mean yield of target plants for a particular treatment **Relative Competitive Ability Index = (Sum of CI's for each species) *Competition Index = ( 71 Control plant yields were not used as base line data since it appears environmental variables such as wind and heat may have suppressed their growth and resultant yield quite severely. Whether these variables affected control yields of the four species equally or differentially was not determined. However, it was evident that with Chenogodium and Panicum the yields of control plants were lower than they were in some treatments in which they were subjected to competi- tion from other plants. Perhaps, even though the competitors were using the same resources from the same source as the target plant, their presence may have modified the physical microclimate (e.g. by providing a wind break, raising relative humidity, conserving heat at night, etc.) enough to make their presence more beneficial than harmful, in some instances. Undoubtedly, these Species have evolved to grow in communities with neighboring plants. To put them in the Open in a field, without neighboring plants, probably exposes them to conditions not conducive for optimal growth and reproduction. Harper (1964) followed this line of thought when he said, "It may be argued, therefore, that the essential qualities which determine the ecology of a Species may only be detected by studying the reaction of its individuals to their neighbors and that the behavior of the individuals of the species in isolation may largely be irrelevant to understanding their behavior in the community." Below Ground Biomass Yield of below ground biomass in experiments I and II were estimated from random samples. Root weight values were lower than similar data from other studies on the same Species in natural communities (Stephenson and Reed, personal communication). The lower 72 than expected root yields may have been partially due to watering the plots in the initial five weeks of the study. It has been shown that dry conditions are favorable to accumulation of organic matter below ground (Singh and Yadavah, 1974). Other investigators have shown that root/shoot ratios tend to increase with xeric conditions (Bray, 1963; Struik, 1965; Struik and Bray, 1970). Plants of the same species growing in drier habitats or in seasons of drought have a higher below ground dry matter production (Singh and Yadavah, 1974). Certainly watering the experimental site every three days for the first five weeks made the soil far more moist than in the surrounding natural communities. There is also a possibility that the sampling procedure may have been somewhat inadequate (especially fOr the fragile fibrous roots of the grasses), resulting in underestimation of the root weights. Data collected indicates that dicots suppressed root growth of neighboring plants more than the grasses did. Root yield of each Species was estimated for all competitive situations in both the diversity-yield experiment (Table 7) and the competitive ability experiment (Table 8). RCA indices, detennined for roots of Species in the diversity-yield experiment, indicated that ranking of competitive ability of roots was Amaranthus > ChenOpodium > Setaria > Panicum. For the pair-wise competitive ability experiment, the ranking was Chenopodium > Amaranthus > Panicum > Setaria. The root yields of the dicots were suppressed less by competition than those of the grasses. 73 00x0 0000000 :00 0000000 00000000500 000 00 5000 u A0300005 000000< 0>000000sou 0>00000000 000500000 0 :0 0000000 0:0 00 0:000: 300 :00: 0 00000 0:00 :0 0000000 0 0o 0:000: :00 :00: v n 0000 000:0 :00000000000 00.0 00.0 00.0 00.0 00.0 00.0 00 00.0 00.0 00.0 00.0 00.0 0:000: :00: 0000000 00.0 00.0 00.0 00.0 00.0 00.0 00 00.0 00.0 00.00 00.0 00.0 0:000: :00: 5:00:00 00.0 00.0 00.0 00.0 00.0 00.0 00 00.00 00.00 00.00 00.00 00.00 0:000: :00: 500000000:o 00.0 00.0 00.0 00.0 00.. 00.0 000 0.00 00.00 00.00 00.00 00.00 0:000: :00: 00:0:0:0E< 500000 0000000 5:00:00 530000000:o 0::0:0:0E< 00<00 00000000500 :00 :0 0000000 .0000000 0000000 0>000000soo 000 0000000 00000000500 :00: 0:000 .0000000500 0000 0::0x0s 0:0 00 :000 00 030: 030 »: 00:0000 00:3 0000000 :000 00 30: 000 05000 :0 00000» 0000 :00: .0 00:00 74 000500000 000800000 0 000 000000 009000 00° 30.; 000.07 0000000 030 000 00 000000 an 000000000 0003 000nm000m000 0o 0000x,000z 00000000 0000000000 0 00 0000000 000000005ou 0o Ezmv 00000 0000000 0>000000000 0000000000 0 u x0000 000000005000 00.0 00.0 00.0 00.0 00.0 00 00.0 00.0 00.0 00.0 00000; 0000 5300000 00.0 00.0 00.0 00.0 00.0 00 00.0 00.0 00.0 00.0 000003 000: 000000m 00.00 00.0 00.00 00.0 00.. 00 00.0 00.0 00.0 00.0 000003 0000 00000000000 00.00 00.0 00.0 00.0 00.0 000 00.0 00.0 00.0 00.0 00000: 000: 00000000E< 000000m 0000000 53000000000 00000000E< 00000 m0o00000000 00000 000000 .0000000 0000000 0>00000asoo 0>000000 000 0000000 0o000000500 0000000000000 000: @0000 .0000000 0000 0cm 000000 000000 00 00000» 0oo0 000: .0 00000 CONCLUSIONS The following points have come to light during the course of this study: l. The diversity yield experiment demonstrated that mean yield of the highest yielding monoculture, Amaranthus, was comparatively higher than any of the mean yields of the mixture plots. However, it did not indicate that Amaranthus yields were significantly higher (a = 0.05) than those of the mixtures. This could be possibly due to (1) lack of sufficient differential among the root growth patterns of the separate species, (2) use of a fixed density that disregards maximal yield densities of the separate species, and (3) insufficient replica- tion. 2. Change of the arrangement of the pattern of the separate populations, within the mixtures, did not cause a significant difference in the yields of the mixtures. Once again, this could be due to lack of sufficient differential among the growth forms of the component species as well as the possibility of insufficient replication. 3. Distribution of the site biomass among the component Species follows a geometric pattern similar to that which is predicted by the niche pre-emption hypothesis (Whittaker, 1965, l969) for natural comunities low in species richness and high in expression of Species dominance. 75 76 4. Using yield as a criterion, dominance was clearly expressed in the diversity-yield plots. Amaranthus contributed greater than 50% of the yield in all the mixture plots. 5. The paired-Species competitive ability experiment did not predict the relative performances of the four Species when grown in mixture plots. It appears that this type of experiment does not allow one to rank a group of plants by their respective competitive abilities and consequently predict the outcome of different combinations. 6. In both the diversity yield experiment and the competitive ability experiment, the dicots were clearly superior to the monocots as competitors. 7. Root yields of the dicots were suppressed less by competition than those of the grasses. LITERATURE CITED LITERATURE CITED Bornkhamm, R. 1961. Competition for light in arable plants. (German text). Flora Jena, 151: 126-143. Bray, J. R. 1963. Root production and the estimation of net primary productivity. Can. J. Bot. 41: 65-72. Cox, G. w. 1967. Laboratory Manual of General Ecology. 195 p. Brown, Dubuque. ‘ Donald, C. M. 1963. Competition among crap and pasture plants. Adv. Agron. 15: l-118. Goodall, D. w. 1960. Quantitative effects of intraspecific competi- tion: An experiment with mangolds. Bull. Res. Coun. of Israel Sec. D. Botany, 8: 181-194. Hairston, N. G. 1964. Studies on the organization of animal communities. J. Ecol. (Suppl. ) 52: 227-239. Haizel, K. A. 1972. The productivity of mixtures of two and three species. J. Appl. Ecol. 9: 601-608. Harper, J. L. 1961. Approaches to the study of plant competition. In F. L. Milthorpe (ed. ), Mechanisms in Biological Competition. p. 1- 39. Oxford Univ. Press, London. . 1964. The individual in the population. J. Ecol. (Suppl.) 52: 149- 158. . 1967. Darwinian approach to plant ecology. J. Ecol. 55: 247-270. King, C. E. 1964. Relative abundance of Species and MacArthur's model. Ecology 45: 716-727. Leith, H. 1960. Patterns of change within grassland communities. In J. L. Harper (ed. ), The Biology of Needs, p. 27-39. Oxford Univ. Press, London. Lewontin, R. C. 1968. Introduction. In R. C. Lewontin (ed.), Population Biology and Evolution, BT'l-4. Syracuse Univ. Press, Syracuse, New York. 77 78 Loomis, R. S., w. A. Williams, and A. E. Hall. 1971. Agricultural productivity. Ann. Rev. Plant Physiol. 22: 431-468. MacArthur, R. H. 1957. On the relative abundance of bird species. Proc. Nat. Acad. Sci. U.S. 43: 293-295. 1960. On the relative abundance of Species. Am. Nat. 94: 25-36. Major, J. 1958. Plant ecology as a branch of botany. Ecology 39: 352-363. Margalef, R. 1963. Succession in marine populations. Advancing Frontiers Plant Sci. 2: 137-188. McIntosh, R. P. 1970. Community, competition and adaptation. Quart. Rev. of Biol. 45: 259-280. McNaughton, S. J. 1967. Relationship among functional prOperties of CalifOrnia grasslands. Nature 216: 168-169. . 1968. Structure and function in California grasslands. ‘EcOTogy 49: 962-972. ' Odum, E. P. 1969. The strategy of ecosystem development. Science 164: 262-270. Orians, G. 1962. Natural selection and ecological theory. Am. Natur. 96: 257-263. 1 Patrick, R. 1949. A proposed biological measure of stream conditions based on a survey of the Conestoga Basin, Lancaster County, Pennsylvania. Proc. Acad. Nat. Sci. Phila. 101: 277-341. Poore, M. E. D. 1964. Integration in the plant community. J. Ecol. (Suppl.) 52: 213-226. Preston, F. H. 1948. The commonness, and rarity of Species. Ecology 29: 254-283. . 1962. The cannonical distribution of commonness and rarity. Ecology 43: 185-215, 410-432. Rhodes, I. 1970. Competition between herbage grasses. Herb. Abstr. 40: 115-121. Sakai, K. 1955. Competition in plants and its relation to selection. Cold Spring Harbor Symp. Quant. Biol. 20: 137-157. Singh, J. S. and P. S. Yadavah. 1974. Seasonal composition, plant biomass and net primary productivity of a trapical grassland at Kurukshetra, India. Ecol..Mono. 44(3): 351-376. 79 Slobodkin, L. B. 1962. Growth and Regulation of Animal Papulations. 184 p. Holt, Rhinehart, and Winston, New York. Stephenson, S. N. 1973. A comparison of productivity and diversity in early and late season oldfield communities. Mich. Acad. 5(3): 325-334. Struik, G. J. 1965. Growth patterns of some native annual and perennial herbs in southern Wisconsin. Ecology 46: 401-420. . and J. R. Bray. 1970. Root-Shoot ratios of native farest herbs and gga_mays at different soil moixture levels. Ecology 51: 892-893. Waddington, C. H. 1961. The Nature of Life. 131 p. Allen and Unwin, London. . 1965. Introduction to the symposium. In H. C. Baker and G. L. Stebbins (eds.), The Genetics of ColoniEThg Species, p. 1-6. Academic Press, New York. Watt, A. S. 1964. The community and the individual. J. Ecol. (Suppl.) 52: 203-211. Walter, H. 1971. Ecology of Tropical and Subtropical Vegetation. 539 p. Van Nostrand Reinhold, New York. Werner, P. 1972. Effect of the invasion of Dipsacus s 1vestris on plant communities in early old-filed success1on. Pfi.U. Thesis. Michigan State Univ. 140 p. Whittaker, R. H. 1965. Dominance and diversity in land plant communities. SCience 147: 250-260. ' . 1969. "Evolution of diversity in plant communities" Divers1ty and Stability in Ecological Systems, Brookhaven Symposia in Biology, No. 22: 178-260. 1970. Communities and Ecosystems. 162 p. Macmillan, ‘London.. Whittington, W. J. and T. A. O'Brien. 1968. A comparison of yields from plots sown with a single species or a mixture of grass Species. J. Appl. Ecol. 5: 209-213. Woodwell, G. M. and R. H. Whittaker. 1968. Primary productivity in terrestrial ecosystems. Amer. Zool. 8: 19-30. Yates, F. 1933. The analysis of replicated experiments when the field results are incomplete. Emp. Jour. Exp. Agr. 1: 129-142. APPENDICES 80 APPENDIX A-l Analysis of variance table for comparison of treatment plot yields in the diversity-yield eXperiment. Source of Variation df 55 MS F Treatment 7 7272386.40 1038912.34 2.82* Error 16 5888965.92 368060.37 Total 23 13161352.32 572232.70 81 town—:o—mu och .ow.m~np mw wzpm> mm; .ummu wgowcmpmoaw mgsumuognuz zuxak asp Low mFQmp mace; pcwo_mwcmwm ammo; mm.omo¢ mm.mmm om.poum mp.oo~ cm.mm¢ Fn.Poom o¢.mm¢ mm.mmm em.N~m —m.mopm cm.mp mm.n¢¢ mm.m¢m m.mmm mm.¢mpm NN.mo¢ mp.nmv mm.o~m um.opo— mm.mmmp m~.—mom m¢.~ mm.om¢ Fm.¢m¢ po.wpm om.w~op mm.mo¢_ on.mmmm m¢.mpm mm.mwm op.mmm mo.oom me.mm~F mm.mmm— meqmwNH mm.womm mm.mom~ om.mmm~ mp._mm~ mm.¢m~m Pm.mo~m _~.~omm om.~oum mo.omo¢ mz