9": pl 1. n ‘ 4,3 “a“? a -u it." ..,. v M... . .‘m” "We’- ' ~uk’ "" . LIBRARY m Michigan State University This is to certify that the dissertation entitled LOCAL AND REGIONAL EFFECTS ON PLANT DIVERSITY: THE INFLUENCE OF SPECIES POOLS, COLONIZER TRAITS AND PRODUCTIVITY presented by Gregory R. Houseman has been accepted towards fulfillment of the requirements for the Ph.D. degree in Plant Biology MM 1. 255” Major Professor’s Signature 01; r/ o 74 / / Date MSU is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE — 2/05 c:/ClRC/DateDue.indd-p.15 LOCAL AND REGIONAL EFFECTS ON PLANT DIVERSITY: THE INFLUENCE OF SPECIES POOLS, COLONIZER TRAITS AND PRODUCTIVITY By Gregory R. Houseman A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements For the degree of DOCTOR OF PHILOSOPHY Department of Plant Biology and Program in Ecology, Evolutionary Biology, and Behavior 2004 ABSTRACT LOCAL AND REGIONAL EFFECTS ON PLANT DIVERSITY: THE INFLUENCE OF SPECIES POOLS, COLONIZER TRAITS AND PRODUCTIVITY By Gregory R. Houseman Traditionally, ecologists have examined variation in species richness as a consequence of local and regional scale processes. At local scales, species diversity may be driven by a complex set of species interactions occurring within and among trophic levels. Conversely, regional scale process may influence local diversity through variation in immigration from the regional species pool. I used a combination of observational and experimental studies to examine the extent to which local and regional processes limit plant species richness in SW Michigan grasslands and if the relative importance of these factors is controlled by community productivity. I focused on four SW Michigan grasslands in which there are natural gradients in productivity and definable community types. Observational data from these sites show that unimodal productivity-diversity relationships across-communities is a general pattern given a sufficient range in productivity, a large number of communities, and identification of regional scales that match key regional processes. Because local scale patterns cannot account for these unimodal patterns, species richness is probably controlled by an interaction between local and regional factors. I experimentally tested this hypothesis with species pool augmentation experiments and found that local environmental conditions limit species richness in both low and high productivity communities. Recruitment was reduced in low productivity communities due to stressful abiotic conditions while strong competitive interactions limited recruitment in high productivity communities. I found little evidence for facilitation across these gradients. The species pool augmentation experiments also demonstrated that species pools limit local diversity across all sites but the magnitude of ' species pool-diversity relationships depend upon site productivity and resident species traits. An analysis of colonizer species traits suggested that the composition of the species pool can change species pool-diversity relationships. Hence, a large species pool with many species possessing unsuitable traits may have less effect on local richness than a small species pool with species possessing suitable traits. The results of these studies demonstrate that both local and regional effects appear to control local species richness in SW Michigan grasslands and that the importance of each varies with community productivity. For Fran iv ACKNOWLEDGMENTS I thank Kay Gross for her outstanding guidance and willingness to let me explore a variety of ideas. She consistently provided excellent feedback on all my written and oral work and provided invaluable logistical support. I also thank Gary Mittelbach, Pete Murphy and Phil Robertson for thoughtful feedback on my research. I would like to thank the many members of the Gross’ lab who have assisted in the field and laboratory. Special thanks to Carol Baker, Stacey Andres, Justin Rensch and Pam Moseley, for their hard work. I also thank the KBS community for your encouragement and camaraderie over the past six years. The outstanding faculty and graduate students make KBS a unique research environment. Special thanks to Nate Dorn, Eric Theboben Rich Smith and Sarah Emery. I am grateful for financial support from a NSF Dissertation Improvement Grant, Sigma Xi and George Lauff Research Grants and a NSF-Research Training Grant to the W. K. Kellogg biological Station. I also received support from the Plant Biology Department and the Ecology, Evolution and Behavioral Biology Program at Michigan State University. Most of all, I thank my wife Fran for her unwavering love and support. She has made this possible. PREFACE “...a farmer went out to sow his seed. As he was scattering the seed, some fell along the path, and the birds came and ate it up. Some fell on rocky places, where it did not have much soil. It sprang up quickly because the soil was shallow. But when the sun came up, the plants were scorched and they withered because they had no root. Other seed fell among thorns, which grew up and choked the plants. Still other seed fell on good soil where it produced a crop—a hundred, sixty or thirty times what was sown. He who has ears let him hear.” The Gospel of Matthew 13:3 vi TABLE OF CONTENTS LIST OF TABLES ............................................................................ LIST OF FIGURES ........................................................................... CHAPTER 1 INTRODUCTION ............................................................ Background ........................................................................... Organization of Dissertation ........................................................ Site Dscriptions ....................................................................... CHAPTER 2 SPECIES DIVERSITY WITHIN AND ACROSS- COMMUNITIES: TESTING SCALE-DEPENDENT HYPOTHESES ............... ABSTRACT ........................................................................... INTRODUCTION ................................................................... METHODS ........................................................................... RESULTS ............................................................................. DISCUSSION ........................................................................ Testing the PAH vs. CAH .................................................. Explaining deviations from the unimodal pattern ....................... Summary ..................................................................... CHAPTER 3 SPECIES POOL SIZE ALTERS PRODUCTIVITY-DIVERSITY PATTERNS: HOW ECOLOGICAL FILTERING CONTROLS LOCAL- REGIONAL DYNAMICS ................................................................... ABSTRACT ........................................................................... INTRODUCTION ................................................................... METHODS ........................................................................... RESULTS ............................................................................. DISCUSSION ......................................................................... CHAPTER 4 THE IMPORTANCE OF SPECIES POOLS AND COLONIZER TRAITS TO LOCAL DIVERSITY ACROSS MULTIPLE PRODUCTIVITY GRADIENTS: TESTING MACARTHUR’S PARADOX .............................. ABSTRACT ........................................................................... INTRODUCTION ................................................................... METHODS ........................................................................... Field data Species Traits ................................................................. Seed Traits .......................................................... Seedling Traits ...................................................... Statistical Analysis .......................................................... RESULTS ............................................................................. Light-richness patterns along the gradients .............................. Species sorting of successful colonists along the gradients ............ DISCUSSION ........................................................................ vii ix Hfl 17 17 18 21 23 26 26 28 29 3O 3O 31 35 37 41 45 45 46 48 48 50 51 51 54 55 55 61 67 Limits to species diversity along productivity gradients .............. Species trait and colonization along productivity gradients CHAPTER 5 SUMMARY AND FUTURE DIRECTIONS ............................ Summary of findings ................................................................. Future directions ..................................................................... REFERENCE LIST ........................................................................... APPENDIX A ................................................................................. viii 67 69 73 73 74 78 88 LIST OF TABLES Table 1.1. Productivity and resident species composition for species pool experiment (Chapter 3). Cover values are presented as means i 1 standard error for the four most abundant species in each community. ....................................... Table 1.2. Seed sources for the species pool experiment (Chapter 3). PM = Prairie Moon; MWF = Michigan wildflower farm; PR = Prairie Ridge. ........................ Table 1.3. Productivity and resident species composition for cross-site experiment (Chapter 4). Cover values are presented as means i 1 standard error for the four most abundant species in each community. ................................................. Table 1.4. Seed sources and functional group for species added in the cross-site experiment (Chapter 4). Seed sources: 1 = Prairie Moon;2 = Michigan wildflower farm; 3 = Oak Prairie; 4 = Agerecol; Reproduction: A = annual or short-lived perennial, B = biennial, P = perennial; Life form: G = grass, L = legume, F = dicots excluding legumes. All species are native to Michigan and originate from Midwestern seed providers. Taxonomy following Voss 1996 .......................... Table 1.5. Site and community (Com.) locations for the species pool (SP; Chapter 3) and cross-site (CS; Chapters 2 & 4) experiments. Coordinates are latitude/longitude and geographic coordinate systems. .................................... Table 4.1. Tests for an effect of seed addition on total richness. The AN OVA table represents mean slicing by site. ............................................................... Table 4.2. Loadings of environmental variables on the first three PCAE axes. Table 4.3. Loadings of species trait variables on the first four PCAT axes. Table 4.4. Results of the fourth-comer analysis. The r-value is the Pearson correlation coefficient. ......................................................................... APPENDIX A. Summary of environmental variables for ancillary plots measured at four grassland sites in SW Michigan. Values represent mean (1 standard error) for 4 replicates, n = 48. All N variables measured in ug/g dry soil. ANPP is estimated by aboveground plant biomass. %Light availability is measured in potential photosynthetic flux density as described in Chapter 4. ..................................... ix 10 12 16 60 64 65 66 88 LIST OF FIGURES Figure 1.1. Relationship between local and regional richness where the form of the relationship varies from linear to asymptotic depending on the strength of species interactions (after Cornell and Lawton 1992). The dashed line illustrates the case when every species in the region is found in a local community. ........................ 3 Figure 1.2. Relationship between aboveground biomass and light availability for the four sites examined. ............................................................................ 6 Figure 1.3. Map of the four study sites in relation to KBS (Kellogg Biological Station). Site 2 was used for the species pool (Chapter 3) and the cross-site experiments (Chapter 4). ....................................................................... 15 Figure 2.1. The within-community (shaded bars) and among-community (dotted line) diversity patterns across productivity gradients (after (Scheiner et a1. 2000). Each bar represents a separate community within the larger region. Under the pattern accumulation hypothesis (A) the among-community PDR is a sum of within- community patterns so that local and regional factors operate the same way at both scales. Conversely, the community aggregation hypothesis (B) suggests that unimodal PDRs across-communities are unrelated to within-community PDRs so that the processes controlling diversity differ between the two scales. .................. 20 Figure. 2.2. Arrangement of plots and communities within a site. Actual arrangement of plots varied with community. .............................................. 22 Figure 2.3. Relationship between % full sun (surrogate for above-ground biomass) and species richness within and across three sites. Low, Med, and High represent communities along a natural productivity gradient at each site. Panels 2D, 2H, and 2L are the compiled data for each site with the best regression fit. ...................... 24 Figure 2.4. Relationship between light availability (a surrogate for above-ground biomass) and species richness. Plots from all communities were pooled across the three study sites. ................................................................................. 25 Figure 3.1. Ecological filtering of species pools. The total species pool includes all species found in region. The geographic species pool is comprised of species that are able to disperse to a local community. The habitat species pool are those species from the geographic species pool that can grow under the abiotic conditions found in a community in the absence of competitors. Species interactions further reduce the habitat species pool to those species found in the community (after Lancaster and Belyea 1999). .................................................................................... 33 Figure 3.2. Species pool-seedling richness relationship in three environments along a natural productivity gradient. Species were added to intact plant communities. Panel A is low, B is medium and C is high productivity (147, 338 and 536g/m2 mean above-ground biomass, respectively). Overlapping data points plotted with random jitter for clarity. ......................................................................... 38 Figure 3.3. Abiotic and competitive filtering of species pools along a productivity gradient. The total species pool is indicated by the top of the open bars (45 species) and represents all species added in the experiments. The hashed bars show the number of species established in vegetation-free plots (competitors removed; see text) and indicates the proportional reduction due to physiological filtering. The black bars are seedling richness in the vegetation-intact plots (vegetation intact; see text) and illustrate the proportional reduction attributed to competitive filters. The importance of abiotic filtering decreased (65, 29, 38%) while competitive filtering increased (56, 72, 97%) along the gradient. .................................................. 39 Figure 3.4. Species pools and productivity-diversity relationships. Total plot richness as a firnction of productivity when adding 0 (A) or 45 (B) species to the existing species pool. Productivity is based on estimates of above-ground biomass at the community level. Overlapping data points plotted with random jitter for visibility. .......................................................................................... 40 Figure 4.1. Regression of light on aboveground biomass. ................................. 56 Figure 4.2. Resident species richness as a function of light availability across four sites (12 communities). ........................................................................ 57 Figure 4.3. Number of species that colonized following seed addition as a function of light availability across four sites (12 communities). ................................... 58 Figure 4.4. Species richness as a function of light availability for each site for resident species (circles, solid lines) and total richness (resident + colonizers: triangles, dashed lines). ......................................................................... 59 Figure 4.5. Indicator Species Analysis for colonizing species. Species Codes are as follows: PETPUR = Petalostemum purpureum, ANDGER = Andropogon gerardii, SCHSCO = Schizachyrium scoparium, KUHEUP = Kuhnia eupatoriodes, ASCTUB = Asclepias tuberosa, LESCAP = Lespedeza capitata, AMOCAN = Amorpha canescens, DESSPP = Desmodium spp., HELMAX = Helianthus maximilliani, ASCSYR = Asclepias syriaca, MONFIS = Monardafistulosa. .......................... 62 xi CHAPTER 1 INTRODUCTION Background Understanding why species diversity varies in space and time continues to be a critical goal for ecology for two reasons. First, identifying the mechanisms that control species diversity addresses the fundamental ecological questions of how species coexist and what controls patterns of distribution and abundance (Palmer 1994, Grace 1995, 1999). Second, understanding the mechanisms that control species diversity is needed to ameliorate the well-documented loss of species diversity world-wide (Bazzaz et a1. 1998, Dirzo and Raven 2003). Despite these two compelling motivations, large gaps in my understanding of the controls on species diversity remain. Ecologists have taken two approaches to explain variation in species diversity that emphasize either local or regional control. The local view has focused on how species interactions control species diversity (MacArthur and Levins 1967, Tilman 1982, Leibold and Chase 2003). Theoretical models and manipulative studies indicate that species have negative effects through direct and indirect interactions that occur within and among trophic levels (Vandermeer 1969, Holt 1977, Abrams 1987, Leibold 1989, Miller 1994). Consequently, diversity might be controlled by a complex set of local interactions. Conversely, island biogeographers have focused on the importance of immigration for controlling species diversity based on species pools, dispersal distances and habitat size (MacArthur and Wilson 1967, Brown 1971, Diamond 1974). Hence the current debate is whether species diversity is primarily controlled by local or regional factors (Ricklefs 1987, Cornell and Lawton 1992, Zobel 1997) and if the relative importance of each varies with secondary factors (6. g. productivity; (Huston 1999, Foster et al. 2004). These potential interactions and feedbacks between different ecological scales is part of the emerging topic of metacommunities (Leibold et al. 2004) Cornell and Lawton (1992) suggested that the relative importance of local or regional factors is evident from the relationship between local and regional richness. They hypothesized that an asymptotic (saturating) relationship occurs in communities with strong species interactions while a linear relationship occurs when species interact weakly (Figure 1.1). Huston expanded this idea by suggesting that the strength of interactions is controlled by system productivity. In low productivity systems, species interactions are weak or occur slowly while in high productivity systems species interactions are strong or occur quickly. Hence, as species pools (regional diversity) increase local diversity increases linearly in low and asymptotically in high productivity systems. This approach provides a potentially important framework for community ecology by indicating the relative importance of local and regional controls on local species diversity and predicting how these relationships change under different resource conditions. However, few experimental tests of these hypotheses exist (but see Foster 2001, 2004). Limit non-interactive Local Richness strong interactions Regional Richness Figure 1.1. Relationship between local and regional richness where the form of the relationship varies from linear to asymptotic depending on the strength of species interactions (after Cornell and Lawton 1992). The dashed line illustrates the case when every species in the region is found in a local community. Organization of dissertation In this thesis, I focus on the patterns of plant species diversity within and across- communities and the potential importance of local and regional factors to explain these patterns. I concentrate on species richness at small-scales (1m2) because herbaceous plants should compete for resources at this scale (Huston and DeAngelis 1994) and within-plot heterogeneity should be lower than at larger scales. Although plant community boundaries may be artificial constructs, they are necessary to examine within and across—community patterns (Palmer and White 1994, Morin 1999). I delineate a community as a contiguous area dominated by 1-2 species with a relatively homogeneous species composition. I refer to these communities as grasslands because the species composition is a mix of grasses and forbs with few woody plants. However, these communities could also be called oldfields because many communities have been cultivated and subsequently abandoned. In addition to these local patterns, I also address regional factors that may limit local species diversity. Defining a region is also a difficult problem because the size of the region changes with the process of interest. For this thesis, the minimum region size is the area that includes all four of my study sites (7km radius) but the region could be much larger depending upon the geographic ranges of species and potential dispersal distances. For the studies reported, defining the region as >7km has little bearing on the results. The importance of productivity to local—regional interactions is also a central theme of this work and I will refer to productivity—diversity relationships as PDRs throughout the dissertation. In some cases, I use light availability below the plant canopy as a surrogate for above-ground biomass. The light availability-aboveground biomass relationship is linear for the four study sites (productivity gradients; Figure 1.2) In chapter 2, “Species diversity within and across—communities: Testing scale-dependent hypotheses,” I compare species richness within and across-communities to test two hypotheses that have been proposed to explain the unimodal PDRs found across- communities. According to the pattern accumulation hypothesis, unimodal cross- community PDRs are the sum of local scale (within-community) patterns (Scheiner et al. 2000). Conversely, the community aggregation hypothesis suggests that the unimodal cross-community PDRs occur regardless of the within-community PDRs. While these two hypotheses do not directly address the mechanism behind the unimodal pattern, they isolate the scale at which the mechanisms potentially operate. The results from this observational study indicate that unimodal PDRs occur across-communities regardless of PDRs found within individual communities. Consequently, local factors alone cannot explain the unimodal pattern suggesting that regional factors may also be important to local species richness. In chapter 3, “Species pool size alters productivity-diversity patterns: How ecological filtering controls local-regional dynamics,” I tested whether species pools (an important regional factor) are important determinants of local richness in grasslands and whether the relative importance of local and regional factors vary with community productivity. To test this hypothesis, I experimentally augmented the species pool of potential colonists in three intact grassland communities (Table 1.1) that occurred along a natural 90 I 80 — A 70 — E n_ 60 - 9.: E 50 T g) 40 SITE == 0 1 ............ LE 30 - A 2 °\° 20 - * 3 -------- 10 D 4 ....... 0 l l l l *kx \xin 100 200 300 400 500 600 700 800 900 Aboveground Biomass (g/m2/yr) Figure 1.2. Relationship between aboveground biomass and light availability for the four sites examined. Table 1.1. Productivity and resident species composition for species pool experiment (Chapter 3). Cover values are presented as means i 1 standard error for the four most abundant species in each community. Productivity Standing Class Biomass Dominant Species % Cover LOW l46.6+8.l Centurea maculosa 7.80:0.40 Hieacium spp. 4.16:0.56 Panicum praecocius 0.68:0.11 Paspalum Ciliatifolium 0.47:0.19 MED 338.5+24.l Bromus inermus 10.7810.27 Saponaria officinalis 1.46:0.96 Hieacium spp. 0.94:0.50 Centurea maculosa 0.55:0.20 HIGH 536.2+4l.3 Phalaris arundinacea 63.27:4.4l Bromus inermus 11-O7i3-47 Vitus spp. 0.76:0.76 Solidago canadensis 0.36:0.36 productivity gradient in a single site in the vicinity of KBS (Figure 1.2). To determine how competitive effects (local factor) influenced colonization along the gradient, I also added seeds (Table 1.2) to plots with competitors removed. The results of this experiment demonstrated that the relative importance of local (abiotic & competitive) conditions varied with community productivity. The importance of species pools also had important effects on productivity-diversity relationships and influenced the cross- community PDR. In chapter 4, “The importance of species pools and colonizer traits to local diversity across multiple productivity gradients: Testing MacArthur’s Paradox,” I test the generality of the results found in Chapter 3 by conducting a species pools augmentation experiment across multiple productivity gradients. By conducting seed additions across four sites (Table 1.3), I determine if species pool augmentation had consistent effects across different sites that varied in resident species composition. In addition, I examined how the composition of the species pool might influence local diversity by analyzing how species traits related to colonization success in different environments. The latter analysis was based on a greenhouse assay of seed and growth-related traits for 28 of the 40 species in the augmented species pool (Table 1.4). The species used in the field and greenhouse experiments were similar to those in the single site experiment (Chapter 3). Seed additions in two different years at site 2 indicate that the results from Chapter 3 are consistent and not dependent upon anomalous conditions found in a particular year. Table 1.2. Seed sources for the species pool experiment (Chapter 3). PM 2 Prairie Moon; MWF = Michigan wildflower farm; PR = Prairie Ridge. Scientific Name Agropyron trachycaulum Allium cemuum Amorpha canascens Andropogon gerardii Anemone cylindrica Asclepias syriaca Asclepias tuberosa Astragalus canadensis Aster novae-angliae Aster pilosus Baptisia Ieucantha Bromus kalmii Carex bre vior Coreopsis tripteris Desmodium canadense Echinacea purpurea Elymus canadesis Eryngium yuccifolium Galium boreale Gaura biennis Gentiana fla Vida (alba) Helianthus grosseserratus Heliopsis helianthoides Heuchera richardsonii Juncus tenuis Lespedeza capitata Monarda fistulosa Oenothera biennis Panicum virgatum Pedicularis canadensis Petalostemum purpureum Pycnanthemum virginianum Flatibida columnifera Hatibida pinnata Rudbeckia hirta Rudbeckia subtomentosa Schizachyn’um scopan'um Silphium integrifolium Silphium Iaciniatum Silene stella ta Solidago n’gida Sorghastrum nutans Sporobolus hetero/epsis Tradescantia ohiensis Vemonia missurica FAMILY Poaceae Liliaceae Fabaceae Poaceae Ranunculaceae Asclepiadaceae Asclepiadaceae Fabaceae Asteraceae Asteraceae Fabaceae Poaceae Cyperaceae Asteraceae Fabaceae Asteraceae Poaceae Apiaceae Rubiaceae Onagraceae Gentiaceae Asteraceae Asteraceae Saxifragaceae Juncaceae Fabaceae Lamiaceae Onagraceae Poaceae Scrophulariaceae Fabaceae Lamiaceae Asteraceae Asteraceae Asteraceae Asteraceae Poaceae Asteraceae Asteraceae Caryophyllaceae Asteraceae Poaceae Poaceae Commelinaceae Asteraceae Common Name slender wheat grass nodding wild onion leadplant big bluestem thimbleweed common milkweed butterflyweed Canadian milk vetch New England aster frost aster white wild indigo prairie brome plains oval sedge tall coreopsis showy tick-trefoil purple conetlower Canada wild rye rattlesnake master northern bedstraw biennial gaura cream gentian saw-toothed sunflower false sunflower prairie alumroot path rush round-headed bush clover wild bergamont common evening primrose switch grass wood betony purple prairie clover common mountain mint long-headed conetlower yellow conetlower black-eyed Susan sweet black-eyed Susan little bluestem rosin weed compass plant starry cam pion stiff goldenrod Indian grass northern dropseed common spiderwort Missouri ironweed Spnng 2001 PM PM PM PM PM MWF PM PM PM PM PM PM PM PM MWF PM PM MWF PM PM PM PM PM PM PM PM PM MWF PM PM PM PM PM MWF MWF PM MWF MWF PM PM MWF MWF PM PR/PM PM Table 1.3. Productivity and resident species composition for cross-site experiment (Chapter 4). Cover values are presented as means i 1 standard error for the four most abundant species in each community. Site Productivi ty Number Class Standing Biomass Dominant Species % Cover SITE 1 LOW 235.6:14.0 Centurea maculosa 19.1:1. Medicago spp. 12 . 3:4 . Berteroa incana 1 . 2:0 . Hieracium spp. 0.8:0. SITE 1 MEDIUM 400 . 6:12 . 1 Solidago canadensis 12 . 0:1 . Centurea maculosa 8 . 5:1 . Aster pilosus 7 .4:0. Trifolium pratense 2 . 7:0 . SITE 1 HIGH 529 . 5:41 . 2 Solidago canadensis 22 . 1:5 . Solidago graminifolia 8 . 0:1 . Agrepyron repens 4 . 8:1 . Vinca spp. 2.9:2. SITE 2 LOW 203 .3:39.4 Centurea maculosa 8.8:0. Hieracium spp. 5.3:1. Rumex acetosella 3 . 8:0 . Panicum capillare 1 . 8:0 . SITE 2 MEDIUM 406 . 4:25 . l Bromus inermus 47 . 7:2 . Saponaria officinalis 2 . 8+1 . Asplenium spp. 1.9:0. Centurea maculosa 0 . 4:0 . SITE 2 HIGH 553 . 5 84 . 9 Phalaris arundinacea 84 . 5:4 . Solidago canadensis 3 . 9:2 . Solidago graminifolia 1 . 2:0 . \JCDWCDKOKOI-‘QHGUJOIPKOanl-‘ibflkommU'IKOID-LOQUI Onoclea sensibilis 0 . 9:0 . SITE 3 LOW 335-5i33.2 Centurea maculosa 18.0:1. Hieracium spp. 12.0:3. Achillea millefolium 8.6:0. Panicum capillare 2 . 8+0 . 10 Table 1.3. (Cont’d) SITE 3 MEDIUM SITE 3 HIGH SITE 4 LOW SITE 4 MEDIUM SITE 4 HIGH 467 554. 282 406 675 .8:35. 6:11. .4:l7. .9:30. .4:58. Solidago canadensis Hieracium spp. Bromus inermus Poa pratense Solidago canadensis Carex spp. Bromus inermus Achillea millefolium Hieracium spp. Centurea maculosa Rumex acetosella Panicum capillare Hieracium spp. Solidago canadensis Rubus spp. Bromus inermus Solidago canadensis Poa pratense Hieracium spp. Solidago graminifolia ll 20. 14. 13. 12 12 46. 17 47. 19. 12. 10. 35. 13 8:2. 8+4. .1:2. .7:1. 8+2. .4:2. .4+2. .4:0. 6:4. .3+l. .6:O. .5+O. 1:4. 6:2. 1:3. 7:2. 8+2. .3:1. .9:3. .O:1. OOWU‘IKOIbCDQWIbKOQNOI-‘UII—‘OKOCD Table 1.4. Seed sources and functional group for species added in the cross-site experiment (Chapter 4). Seed sources: 1 = Prairie Moon;2 = Michigan wildflower farm; 3 = Oak Prairie; 4 = Agerecol; Reproduction: A = annual or short-lived perennial, B = biennial, P = perennial; Life form: G = grass, L = legume, F = dicots excluding legumes. All species are native to Michigan and originate from Midwestern seed providers. Taxonomy following Voss 1996. Scientific Name Amorpha canescens Andropogon gerardii Aquilegia canadensis Asclepias syriaca Asclepias tuberosa Aster azureus Astragalus canadensis Baptisia Iactea (Ieucantha) Bromus kalmii Coreopsis Ianceolata Desmodium canadense Desmodium illinoense Echinacea pallida Echinacea purpurea Galium boreale Gaura biennis Helianthus grosseserratus Helianthus maximiliani Helianthus occidentalis Family FABACEAE POACEAE RANUNCULACEAE ASCLEPIADACEAE ASCLEPIADACEAE ASTERACEAE FABACEAE FABACEAE POACEAE ASTERACEAE FABACEAE FABACEAE ASTERACEAE ASTERACEAE RUBIACEAE ONAG RACEAE ASTERACEAE ASTERACEAE ASTERACEAE 12 Common Name lead plant big bluestem columbine Common milkweed butterfly weed Sky-blue aster Canada Milk-vetch white wild indigo arctic brome lanceleaf coreopsis showy tick-trefoil Illinois tick trefoil pale purple conetlower Eastern purple conetlower Northern bedstraw biennial beeblossom Sawtooth sunflower Maximilian sunflower western sunflower Seed Mass 0.0027 0.0022 0.0008 0.0061 0.0044 0.0004 0.0016 0.0128 0.0021 0.0010 0.0045 0.0072 0.0049 0.0032 0.0021 0.0078 0.0017 0.0021 0.0012 Source 3, P, L 2, P,G 3, P,F 1, P,F 3, P,F 4, P,F 1, P,L 4, P,L 1, P,G 1, P,F 4, P,L 3, P,L 3, P,L 1, P,F 1, P,F 1, B,F 1, P,F 3, P,F 3, P,F Table 1.4 (Con’t) Heliopsis helianthoides Juncus tenuis Kuhnia eupatorioides Lespedeza capitata Liatris aspera Monarda fistulosa Oenothera biennis Panicum virgatum Penstemon digitalis ASTERACEAE JUNCACEAE ASTERACEAE FABACEAE ASTERACEAE LAMIACEAE ONAG RAC EAE POAC EAE SCROPHULARIACEAE Petalostemum purpureum FABACEAE Pycnanthemum virginianum LAMIACEAE Hatibida columifera Hatibida pinnata Fiudbeckia hirta Rudbeckia subtomentosa Schizachyrium scoparium Silphium integrifolium Silphium Iaciniatum Solidago n'gida Sorghastrum nutans Sporobolus hetero/epis Tradescantia ohiensis ASTERACEAE ASTERACEAE ASTERACEAE ASTERACEAE POAC EAE ASTERACEAE ASTERACEAE ASTERACEAE POAC EAE POAC EAE COMMELINACEAE 13 false sunflower, ox eye poverty rush false boneset 0.0048 <0.0001 0.0011 round headed bush-clover 0.0025 rough blazing star Bee balm evening primrose switchgrass Foxglove beardtongue Purple Prairie-clover mountain mint upright prairie conetlower gray-headed conetlower black eyed susan sweet conetlower little bluestem wholeleaf rosinweed compass plant stiff goldenrod lndiangrass sand dropseed spiderwort 0.0064 0.0003 0.0006 0.0015 0.0005 0.0013 0.0002 0.0007 0.0028 0.0004 0.0007 0.0041 0.0184 0.0331 0.0006 0.0022 0.0015 0.0048 3, P,F 1, P,G 3, P,F 3, P,L 4, P,F 2,P,F 2,B,F 2, P,G 3, P,F 1, P,L 1, P,F 1, P,F 2,P,F 2,S,F 1, P,F 2,P,G 1, P,F 1, P,F 2,P,F 2,P,G 1, P,G 3, P,G Finally, in Chapter 5, I provide a general summary from each of the studies and explore the important next steps in the debate over local vs. regional control of species diversity. Site Descriptions All the field studies were conducted on sites in southwestern Michigan near the W. K. Kellogg Biological Station (Kalamazoo County; 42° 24’ N, 85° 24’ W). The four sites occurred within a 7km radius (Figure 1.3; Table 1.5) and varied strongly in plant community composition and above-ground biomass (Table 1.1). Sites were previously farmed or grazed and had been abandoned 8-50 years prior to the experiments. The soils for all sites were Oshtemo and Spinks sandy loams (Austin 1979). Each site represented a gradient in plant productivity that was apparently driven by variation in topography. Generally, low productivity communities occurred on ridgetops and high productivity communities in low-lying depressions. Although both field experiments (Chapters 3 & 4) shared Site 2, plots were established in separate locations. 14 \ F Figure 1.3. Map of the four study sites in relation to KBS (Kellogg Biological Station). Site 2 was used for the species pool (Chapter 3) and the cross-site experiments (Chapter 4). 15 Table 1.5. Site and community (Com.) locations for the species pool (SP; Chapter 3) and cross-site (CS; Chapters 2 & 4) experiments. Coordinates are latitude/longitude and geographic coordinate systems. Site Com Study Chapter Coordinates 2 L SP Ch 3 42.3942°N, 85.3510°W (WGS84/NAD83) 2 M SP Ch 3 42.3937°N, 85.3506°W (WGSS4/NAD83) H SP Ch 3 42.3925°N, 85.3502°W (WGSB4/NAD83) 1 L CS Ch 4 42.3873°N, 85.3566°W (WGSB4/NAD83) 1 M CS Ch 4 42.3877°N, 85.3566°W (WGS84/NAD83) 1 H CS Ch 4 42.3880°N, 85.3565°W (WGSB4/NAD83) 2 L CS Ch 4 42.3943°N, 85.3511°W (WG584/NAD83) 2 M CS Ch 4 42.3939°N, 85.3504°W (WGSS4/NA083) 2 H CS Ch 4 42.3923°N, 85.3502°W (WGS84/NA083) 3 L CS Ch 4 42.4891°N, 85.4524°W (WGSB4/NAD83) 3 M CS Ch 4 42.4888°N, 85.4523°W (WGS84/NAD83) H CS Ch 4 42.4887°N, 85.4523°W (WG584/NAD83) 4 L CS Ch 4 42.4837°N, 85.4523°W (WGSS4/NAD83) 4 M CS Ch 4 42.4838°N, 85.4520°W (WGSB4/NAD83) 4 H CS Ch 4 42.4842°N, 85.4513°W (WG584/NAD83) 16 CHAPTER 2 SPECIES DIVERSITY WITHIN AND ACROSS-CONHVIUNITIES: TESTING SCALE— DEPENDENT HYPOTHESES ABSTRACT Unimodal productivity-diversity relationships (PDRs) are commonly reported when examined across plant communities; however, it is not known if such patterns are driven by mechanisms that operate at within- vs. across-community scales. Recently two hypotheses have been proposed that link patterns at these two scales. According to the pattern accumulation hypothesis (PAH), unimodal PDRs found across-communities are a consequence of the accumulation of within-community PDRs. Alternatively, the community aggregation hypothesis (CAH) suggests that when many communities are aggregated a unimodal PDR emerges regardless of within-community patterns. I tested these two hypotheses by sampling species richness within and among-communities in low, medium and high productivity communities along each of three independent, natural productivity gradients in SW Michigan. I found only one significant (but weak) within- community PDR for the nine communities. Conversely, I found a strong, unimodal PDR when all communities were combined providing strong support for the CAH. These results suggest that unimodal patterns may commonly occur given 1) a sufficient range in productivity, 2) a large number of communities, and 3) regional scales that match regional processes. 17 INTRODUCTION Explaining why species diversity patterns vary with scale is a central but unresolved question in community ecology. Productivity-diversity relationships (PDRs) have emerged as a powerful way to examine scale-dependent patterns because system productivity is relevant at nearly all spatial scales (from 1m2 to continents). Furthermore, productivity provides a common currency that potentially allows generalization across disparate systems and taxa. Recent reviews report that unimodal PDRs are commonly observed in a broad range of community types including aquatic and terrestrial for both plants and animals (Mittelbach et al. 2001, Comwell and Grubb 2003). Although several hypotheses have been proposed to explain the determinants of the unimodal pattern (reviewed in Rosenzweig and Abrarnsky 1993, Leibold 1999, Waide et al. 1999, Rajaniemi 2003) most do not distinguish between processes that operate at within or across—community scales. Yet the scale-dependency of PDRs may be critical to understanding the mechanism driving the observed pattern (Scheiner et a1. 2000, Chase and Leibold 2002, Gamarra and Sole 2002, Scheiner and Jones 2002, Weitz and Rothman 2003). Recently two hypotheses have been proposed to explain unimodal PDRs across communities. The pattern accumulation hypothesis (PAH; Scheiner et al. 2000) says that unimodal across-community PDRs are the sum of local scale (within-community) patterns (Figure 2.1A). Low productivity communities have positive PDRs while high productivity communities have negative PDRs. Hence a unimodal across-community PDR occurs when a range of low to high productivity communities are combined. 18 Consequently, the same processes control diversity at both the local (within-community) and regional (across-community) scales. A second hypothesis suggests that unimodal across-community PDRs occur regardless of the within-community PDRs. For example, several low productivity communities may have positive, negative or no within-community PDR, but when combined with many medium and high productivity communities, a unimodal across-community PDR emerges (Figure 2.1B.). In this case, different processes control diversity patterns at local vs. regional scales (Scheiner et al. 2000). I call this the community aggregation hypothesis (CAH). There have been few tests of the PAH or the CAH (but see Gross et al. 2000) and the available data on PDRs are often confounded by scale-dependent processes. For example, regional factors such as climate, species pools, large-scale disturbance and herbivory may vary with spatial scale (Mittelbach et al. 2001). Consequently, such regional factors may differentially affect communities found in the region. I was able to overcome these limitations by locating three sites in SW Michigan that have broad ranges in productivity (mean range = 274-594g/m2) over small spatial scales (<200m). Consequently, regional factors are likely to be similar at each site. I tested these two hypotheses by comparing species richness patterns within and across-communities for three replicate productivity gradients. l9 Pattern Accumulation Hypothesis (PAH) a) A 8 E .—-..O 00...... ax Productivity Richness Community Aggregation Hypothesis (CAH) Productivity Figure 2.1. The within-community (shaded bars) and among-community (dotted line) diversity patterns across productivity gradients (after Scheiner et al. 2000). Each bar represents a separate community within the larger region. Under the pattern accumulation hypothesis (A) the among-community PDR is a sum of within-community patterns so that local and regional factors operate the same way at both scales. Conversely, the community aggregation hypothesis (B) suggests that unimodal PDRs across-communities are unrelated to within-community PDRs so that the processes controlling diversity differ between the two scales. 20 METHODS I selected three grassland sites in the vicinity of KBS (7km radius) in SW Michigan. All sites had been farmed or grazed and left undisturbed for at least 10-40 years. Each site included a broad gradient of ANPP located along slopes with similar topography. At each site, I selected low, medium and high productivity plant communities based on aboveground biomass. Each community was defined as a contiguous area dominated by 1-2 species with a relatively homogeneous species composition. Across the three sites, communities varied in relative dominance of grasses and forbs (Chapter 1, Table 3). I established 16 replicate plots in the central portion of each community (1.2m x 1.2m) spaced approximately 0.5-1m apart (Figure. 2.2). Plots were located systematically rather than randomly for the purposes of a seed addition experiment (presented in chapter 4). Because most communities were small, locating plots in the center of each community minimized edge effects that occurred at the transitions between communities while providing a good characterization of within-community variability. Species richness was quantified in the central 1m2 of each plot during June and August of 2003. I used light interception by the plant canopy as a surrogate for productivity to minimize disturbance and because light is a key resource for plant growth and species interactions (Grace and Pugesek 1997, Kull and Aan 1997, Grace et al. 2000, Liira and Zobel 2000). In late August and early September, I measured light availability below the plant canopy (at ground level but above the litter) using a Decagon, sunfleck ceptometer (SF-80) i 2 hours of solar noon. Light readings were taken every 10 cm across each plot. I 21 Productivity Gradient I Plot 0 Community Figure. 2.2. Arrangement of plots and communities within a site. Actual arrangement of plots varied with community. 22 light availability for each plot as a percentage of light measured above the plant canopy (Photosynthetic Photon Flux Density; PPFD) for 800 readings (80 sensors x 10 sample points I used linear regression to test for linear or polynomial relationships between species richness and light within and among communities and the MOS test (Mitchell-Olds & Shaw 1987) to test for unimodal relationships. All analyses were conducted with SYSTAT v9.0. RESULTS At the within-community scale there was no significant light-richness relationship in the low or medium productivity communities. In the high productivity communities, I found one significant, but weak, negative 1i ght-richness relationship (Site 4, p=0.03, R2=0.28) that was consistent with the PAH. When data from low, medium and high productivity communities were aggregated within a site, all three sites had significant PDRs. Site 2 (Figure 2.3D) and 4 (Figure 2.3L) had significant quadratic terms (p<0.001, R2=0.77 and p<0.001, R2=0.44 respectively) and both were unimodal (MOS test p<0.001). However site 3 (Figure 2.3H) had a negative linear pattern (p<0.001, R2=0.79). When data from all 3 sites were combined, I found a strong unimodal relationship (quadratic p<0.001, R2=O.46, MOS Test p<0.001; Figure 2.4.). 23 .5 ”~282on “won 65 £3, 38 some new See cozaaoo 05 one AN use in .ON flash .8? none 8 323% 338355 Ease a mac? 835888 60858 :wi use .82 .33 .85 8:: awoken can 553, mmocnoc 36on use @8803 950%-??? c8 8&8..sz :3 :3 exe :8an Samoan—om .mN Semi cam ES 68 cm on 00 o — 446an w OWOWOWOWOWO H I—‘F-t zrn/ssouqom 'dds zIn/sseuqom “dds zur/ssouqom 'dds '1‘: v—1 0 N cam ES 88 cam 35 68 gm. ==m axe gees ease ease no :2 zm/ssouqom 'dds zuI/sseuqom 'dds zrn/sseuqom 'dds 4 25 m cum N as 24 20 I I I I N E (D a) OJ C .C .9 II a) .‘2 8 0. SITE (D Q 2 A 3 D 4 l l l l o 100 80 60 4o 20 o % Full Sun (PPFD) Figure 2.4. Relationship between light availability (a surrogate for above-ground biomass) and species richness. Plots from all communities were pooled across the three study sites. 25 DISCUSSION Testing the PAH vs. CAH Across these three sites, I found little evidence for unimodal or any other PDRs within- communities. In contrast, the strong unimodal pattern found‘when sites and communities were aggregated was strong evidence for the CAH. Despite the clear unimodal pattern across all sites, not all single sites had a unimodal pattern. One possible reason for differences between single and combined site PDRs is that a sufficient range in productivity was not always found on a single site. For example, in site 4 (Figure 2.3H) the low productivity community was more intermediate than those found on sites 2 (Figure 2.3D) and 4 (Figure 2.3L) and may explain why the PDR was negative rather than unimodal. The site-specific patterns (Figures 2.3D, 2.3H, 2.3L) diminished as the number of communities and the range in productivity increased (Figure 2.4) suggesting that a large sample of communities may be needed to detect unimodal PDRs across communities (Guo and Berry 1998). Other studies that have examined PDRs within and across-communities report results similar to those presented here (Moore and Keddy 1989, Guo and Berry 1998, Gross et al. 2000); however, the unimodal curves presented in those studies were less clear because of either small peaks in or high variation around the curves. The system reported here likely had a stronger unimodal pattern because of fewer confounding effects. For example, species pools were probably very similar within each site because low, medium and high productivity communities were close in space (<200m). In contrast, the regions defined by Gross et al. (2000) were extremely large so that species pools and other 26 “regional” factors were not shared by all communities included in the study. Consequently, the unimodal pattern was weaker than in my system. The implication is that the strength of unimodal PDRs across communities decreased as the scale (e. g. extent sensu (Scheiner et al. 2000) of the study increased. Eventually, the region becomes so large that it is equivalent to continental scales where positive linear PDRs are reported (Currie and Paquin 1987). Such shifts across-scales suggest that diversity is driven by scale-dependent mechanisms (Shmida and Wilson 1985, Waide et al. 1999, Scheiner et al. 2000, Mittelbach et al. 2001, Chase and Leibold 2002). Given the unimodal cross-community pattern, new hypotheses are needed to address the potential interaction between local and regional scale factors. One recent explanation is the shifting-limitations hypothesis (SLH; Huston 1999, Foster et al. 2004). Under the SLH, productivity becomes the key driver of diversity by controlling the relative importance of local and regional factors for species diversity. The SLH predicts that when productivity is high, strong species interactions reduce the importance of immigration processes from regional species pools. Consequently, local processes/interactions control diversity. However, at intermediate levels of productivity, species interactions are less important and colonization potential is more important than at high productivity. Hence productivity (resource supply rates) determines the relative importance of local vs. regional effects (Comwell and Grubb 2003). However, few data are available to directly test this idea (but see Foster et al. 2004; Houseman chapters 3 & 4). 27 Taken together, diversity patterns and presumably the underlying mechanisms vary with scale (Shrnida and Wilson 1985, Waide et al. 1999, Scheiner et al. 2000, Mittelbach et al. 2001, Chase and Leibold 2002). Within communities, diversity is likely controlled by factors such as competition (Tilman 1982, Leibold and Chase 2003), disturbance (Grime 1973, Connell 1979) and resource heterogeneity (Palmer and Dixon 1990, Vivian-Smith 1997, Lundholm and Larson 2003, Pausas et al. 2003). At large (continental) scales, diversity is linearly related to climatic gradients although the specific underlying mechanisms remains controversial (Currie and Paquin 1987, Scheiner and Reybenayas 1994, Sax 2001, Whittaker et al. 2001, Bromham and Cardillo 2003). Between these two extremes, species diversity may be unimodally related to productivity as resource conditions regulate the relative importance of local and regional factors (Huston 1999, Foster et al. 2004) Explaining deviations from the unimodal pattern I found that the unimodal PDRs became more apparent as the number of communities sampled increased (Figures 2.3D, 2.3H, 2.3L vs. 2.3A). Like all inferences based on sampling, larger sample sizes improve the estimate and the shape of the relationship. Consequently, sampling a few communities along a single gradient revealed both linear and unimodal patterns while the aggregation of many communities revealed a general unimodal pattern. This result is consistent with reviews of single studies that found positive, negative, unimodal and U-shaped relationships for individual studies (Mittelbach et al. 2001) while synthesis studies revealed a unimodal when communities were aggregated (Gross et al. 2000). Hence, the unimodal PDR across-communities may 28 be the general underlying pattern but this pattern may be difficult to detect in studies that sample only a few communities. Assuming that the unimodal pattern is the expected pattern, variation around the unimodal function may offer great insight into how secondary factors influence local diversity. A community that deviates from the unimodal pattern may indicate either shifts in the relative strengths of local and regional factors or increased importance of previously insignificant factors. For example, higher richness (compared to the unimodal trend) suggests that the constraints on diversity may be relaxed because of larger species pools, fewer strong competitors or disturbances that reduce competitive exclusion, etc. Conversely, decreased richness from the unimodal pattern indicates additional suppressive effects from those found in communities close to the unimodal curve. Additionally, the magnitude of these deviations suggests whether these secondary factors are relatively strong or weak. Summary This study demonstrates that unimodal PDRs can occur across-communities regardless of the PDR patterns found within communities. Such unimodal patterns may be a general rule given 1) sufficient range in productivity, 2) a large number of communities, and 3) regional scales that match regional processes. Because many studies do not meet these criteria, reports of other PDRs should be expected. A next step might be to test whether the unimodal pattern found across-communities is controlled by the interaction between local and regional processes. 29 CHAPTER 3 SPECIES POOL SIZE ALTERS PRODUCTIVITY-DIVERSITY PATTERNS: HOW ECOLOGICAL FILTERIN G CONTROLS LOCAL-REGIONAL DYNAMICS ABSTRACT Ecologists are currently debating whether local diversity is limited by the number of species capable of dispersing to a community (species pool) or by local species interactions. Understanding the importance of species pools to diversity is critical given the alteration of species pools as a consequence of habitat fragmentation and introduction of non-native species. In a plant species augmentation experiment, I demonstrated that the relationship of species pools to local richness varied from linear to asymptotic depending on site productivity. Colonization by added species was higher at intermediate productivity than at the extremes of the productivity gradient. At low productivity, species pools were limited (filtered) by unsuitable abiotic conditions while strong competitive interactions limited colonization at high productivity. The augmented species pool also shifted a negative linear productivity-diversity pattern to unimodal and offered a potential explanation for variation in productivity-diversity patterns across communities. These results suggest that alteration of species pools by human activities may have important consequences for local, native species diversity. 30 INTRODUCTION Ecologists have traditionally sought to explain local patterns of species diversity by focusing on species interactions and other local-scale processes (Ricklefs 1987) such as competition or disturbance but are increasingly considering the additional constraint of processes that occur outside individual communities (Cornell and Lawton 1992, Caley and Schluter 1997, Cornell 1999, Shurin et al. 2000). For example, the pool of species available to local communities sets an upper limit to local diversity and determines the collection of species that will interact locally. Consequently, local species diversity may be driven by differences in species pools rather than local species interactions. The size of regional species pools is determined by a combination of ecological, evolutionary and historical factors. Although speciation and extinction are the ultimate causes of regional species pool size, within a region physical barriers to dispersal or distances between suitable habitats may limit the actual pool of species able to disperse to a particular site (MacArthur and Wilson 1967, Yeakley and Weishampel 2000). For example, within the distributional range of a single species some suitable habitats may be unoccupied if the suitable patch is sufficiently isolated or barriers (forests, water, human settlements) impede dispersal. Human activity has dramatically altered dispersible species pools through habitat destruction, fragmentation, and the introduction of exotic species. These anthropogenic effects on species pools may have strong consequences for local diversity depending upon the importance of species pools to local diversity, but such relationships are poorly understood. 31 Theoretical arguments regarding the effects of regional pools on local diversity have primarily considered how the strength of local interactions may limit the number of coexisting species (Cornell and Lawton 1992, Loreau 2000). Asymptotic relationships between species pool size and local richness are expected when local interactions such as competition and predation strongly limit the number of species capable of coexisting. Conversely, a linear relationship is expected if local species interactions are relatively weak or competitive exclusion is slow (Chapter 1: Figure 1.1). Huston (1999) has recently suggested that the form of the relationship between species pools and local diversity may be mediated by abiotic conditions. Specifically, Huston proposes that ecosystem productivity dictates whether the local-regional relationship is linear under conditions of low or intermediate productivity or asymptotic under high productivity conditions. The species pool-richness relationship may vary with productivity because the relative magnitude of abiotic and competitive filters differs across the productivity gradient. Ecological filtering is the sequential reduction of species pools from a large regional pool to the species found in a local community (Figure 3.1; Keddy 1992, Belyea and Lancaster 1999, Booth and Swanton 2002). Under this model, the regional species pool is initially reduced by dispersal barriers (distance and landscape arrangement) among suitable habitats (dispersal filters). Species are subsequently filtered from the remaining pool because they lack the physiological traits to tolerate abiotic conditions (abiotic filters). Interactions between resident species and potential colonists further reduce the pool of species to those found in a local community (competitive filters). However, species 32 Total Species Pool . . . "‘=-...-.;.;;:;;;5:SI:;:.;. ....... . . Dispersa' filter Geographic Species Pool ....... .............. - o o o 0 0 0 Abiotic filter Habltat Specles Pool o o c o o o SPECIES Interactions filter Figure 3.1. Ecological filtering of species pools. The total species pool includes all species found in region. The geographic species pool is comprised of species that are able to disperse to a local community. The habitat species pool are those species from the geographic species pool that can grow under the abiotic conditions found in a community in the absence of competitors. Species interactions further reduce the habitat species pool to those species found in the community (after Belyea and Lancaster 1999). 33 interactions can ameliorate competitive filtering if resident species facilitate colonization (Callaway et al. 2002). If diversity is related to species pool size and productivity, then reported productivity- diversity relationships (PDRs) may vary because of differences in available species pools. Interest in the mechanisms accounting for PDRs has increased dramatically over the past few decades as vigorous debate about the ‘true’ relationship continues (Huston 1994, Mittelbach et al. 2001). In plant communities, negative linear or unimodal PDRs are most commonly observed, but other patterns are also reported (Mittelbach et al. 2001). Explanations for the observed variation in PDR in plant communities primarily focus on how competitive interactions vary under different resource conditions (Abrams 1995, Rajaniemi 2003) with little consideration of how regional species pools might influence these patterns (but see Huston 1999). Here I report the results of a test of the relative importance of regional vs. local interactions on diversity by experimentally augmenting species pools across a plant productivity gradient in a successional grassland. This experiment allows me to determine 1) if the magnitude of regional (species pool size) and local (competition) factors are determined by system productivity, 2) the influence of abiotic vs. biotic factors on these dynamics, and 3) whether species pool sizes can explain different reported productivity-diversity relationships (Gross et al. 2000). 34 METHODS To test how species pool size affects local diversity under different productivity conditions, I augmented species pool sizes in three sites that differed in productivity in a mid-successional old-field. Standing biomass varied from 170-550g/m2/yr in these environments and was associated with variation in soil N and moisture across the field. In the Spring of 2000, I added seeds of 0, 5, 15, 30 or 45 species not already present in the field (with two exceptions) to 1m2 plots within each site. All species are native to SW Michigan and represent a range of life forms (grasses, forbs, N -fixers). Each treatment was replicated 8-10 times in each site. Because there may be a mass-related tradeoff between germination success and dispersal distance, I added 1g/m2/spp as a way to standardize for differences in seed mass. The species pool added to each plot was a unique random draw of the maximum pool of 45 species. Seeds were added to undisturbed vegetation in early spring and established species were quantified at the end of the growing season. Species added were random draws from a general pool of 45 species that could occur in SW Michigan grasslands (Voss 1996), but were nearly all absent from the site. Seed was purchased from native seed suppliers and represented a range of life forms (grasses, legumes, forbs) and species traits (seed size, plant size, etc; Houseman and Gross in prep). I estimated above-ground net primary productivity (ANPP) by harvesting all aboveground biomass near peak biomass (August) in eight 0.5x1m plots that were randomly located between the seed addition plots. Biomass was dried for a minimum of 48 hours at 60°C. I interpreted the change in richness as the potential diversity response to increased species pool size. If communities were primarily controlled by local factors, I expected a non-linear (asymptotic) relationship or 35 no increase in richness if the community was already saturated. Conversely, if diversity is limited by regional factors, I expected a linear relationship between species pool size and richness and the slope of this relationship would indicate the relative magnitude of the species pool effect. I tested the relative importance of abiotic and competitive filters along the gradient by comparing colonization in vegetation-free plots to plots with vegetation-intact. In the Spring of 2000, I created two vegetation-free plots in each site (Low, Medium, High productivity) by applying a 2% solution (~2.4IJha) of Round-up herbicide in mid-April 2001 to kill all existing live vegetation. A second application was required ten days later. Following plant senescence, I clipped and removed dead vegetation and randomly assigned a single species into each O.5x0.5m subplot. Seed were sown (1 g/mzlspp) two- weeks following the second herbicide application. All non-target species were removed by hand weeding at regular intervals throughout the growing season. At the end of the growing season, I determined the number of species that had established in the vegetation-free and vegetation-intact plots in each environment and used this measure to characterize the magnitude of abiotic and biotic filters in each site. Statistical tests were calculated with SAS 8.02 and SYSTAT 8.0. I used regression to test for significant relationships of seedling richness on species pool size. For significant regressions, I used a Lack of Fit test to determine linearity. Regression slopes were tested by GLM with planned contrasts. I tested the effect of species pool augmentation (0 vs. 45) on total plot richness with multiple regression to determine whether a linear or 36 quadratic relationship best fit the data (Zar 1996). I used the Mitchell-Olds and Shaw test to detennine if a significant peak in the quadratic function occurred within the data range (Mitchell-Olds and Shaw 1987). Data were log-transformed were necessary to meet statistical assumptions. RESULTS I found that plot-scale species richness increased linearly with increasing species pool size in low and medium productivity sites (lack of fit test p>0.05). In contrast, at the high productivity site there was no increase in species richness in response to the seed pool addition indicating the community was saturated (regression p=0.45; Figure 3.2A). The slope of the species pool-richness relationship was higher at intermediate than at low productivity (p=0.035), suggesting that sites of intermediate productivity may have a greater potential diversity response to augmentation of species pools size than low or high productivity sites. Of the 45 species added to vegetation-free plots, 65% failed to establish in the low productivity site while 29% were absent from medium and 38% from high productivity sites, respectively (Figure 3.3.). These results suggest that the abiotic filter is much stronger at low than medium or high productivity. In contrast, the reduction of species pools from vegetation—free to vegetation—intact in Low (56%), Medium (72%) and High (97 %) productivity sites suggest that competitive filtering increases markedly along the gradient. 37 Seedling Richness Seedling Richness Seedling Richness O-bNOD-kUIOJVO-INODAWCDVO—LNOO-bmmfl o 5 15 30 45 Species Added Figure 3.2. Species pool-seedling richness relationship in three environments along a natural productivity gradient. Species were added to intact plant communities. Panel A is low, B is medium and C is high productivity (147, 338 and 536g/m2 mean above- ground biomass, respectively). Overlapping data points plotted with random jitter for clarity. 38 Flltering of Species Pools U) 45 I I I I 1; I I\. III g g § Abiotic Filter 5 36' E a E I ‘ Q 47/ I I/ a: .. .././ "i 7 _. (29 27 42 \ g 13 7 ‘ Competitive w Filter (I) 9 /n LOW MED HIGH PRODUCTIVITY Figure 3.3. Abiotic and competitive filtering of species pools along a productivity gradient. The total species pool is indicated by the top of the open bars (45 species) and represents all species added in the experiments. The hashed bars show the number of species established in vegetation-free plots (competitors removed; see text) and indicates the proportional reduction due to physiological filtering. The black bars are seedling richness in the vegetation-intact plots (vegetation intact; see text) and illustrate the proportional reduction attributed to competitive filters. The importance of abiotic filtering decreased (65, 29, 38%) while competitive filtering increased (56, 72, 97%) along the gradient. 39 TOTAL RICHNESS/m2 0 I I I I 0 n I I I 100 200 300 400 500 100 200 300 400 500 600 PRODUCTIVITY (ANPP g/mz) PRODUCTIVITY (ANPP 9/m2) Figure 3.4. Species pools and productivity-diversity relationships. Total plot richness as a function of productivity when adding 0 (A) or 45 (B) species to the existing species pool. Productivity is based on estimates of above-ground biomass at the community level. Overlapping data points plotted with random jitter for visibility. 40 The addition of 45 species as seed resulted in a shift from a negative linear (p<0.0001) to a unimodal productivity-diversity pattern (quadratic p=0.0037; Figure 3.4) with an internal peak (p<0.01). DISCUSSION This is the first experimental evidence showing that productivity determines whether the species pool-richness relationship is linear or asymptotic. Earlier correlational studies (Partel et al. 1996, Caley and Schluter 1997) have been inconclusive because of methodological problems (Cresswell et al. 1995, Srivastava 1999, Herben 2000). My result demonstrates experimentally that species pools are important to diversity and that different species pool-richness patterns reported elsewhere may be linked to differences in site productivity. Additionally, I show that the importance of the species pool effect on richness is greater at a site with intermediate productivity (as measured by the slope of the species pool-local richness regression; Figure 3.2). This suggests that the reduction in species pool size through habitat loss and fragmentation may be strongest in communities with intermediate levels of productivity. My results also show that physiological filtering of the species pool controls local diversity in the low productivity sites in this grassland community. Although I added species with a wide range of traits, few of them established, suggesting that they lacked the physiological or morphological traits to establish in the low resource/high abiotic stress conditions (6. g. low water and N; high temperature) found in this environment even in the absence of competitors. While I detected some competitive filtering at the low 41 productivity site, abiotic constraints were far stronger than competitive interactions. Conversely, in the high productivity environment, competitive interaction filters were stronger controllers of local diversity than abiotic/physiological factors. Although the importance of competition along productivity gradients has been strongly debated (Welden and Slauson 1986, Goldberg et al. 1999) few data on colonization have been available to evaluate how this may control diversity patterns (but see Foster et al 2004) despite the fact that establishment is generally the critical life stage for plants (Harper 1977, Howard and Goldberg 2001). Several recent studies have demonstrated the potential for facilitative interactions to enhance establishment or growth of species in abiotically stressful environments. I found little evidence for facilitation at my low productivity site. For species that were unable to colonize the vegetation-free plots, only a few individuals were found in the vegetation- intact plots. The absence of facilitation in my study may reflect differences in the magnitude of environmental stress between my system and other studies in which facilitation has been demonstrated such as alpine, desert, or salt flat systems (Callaway et al. 2002). My data also show that variation in species pool size can account for different productivity-diversity patterns. Although the magnitude of the peak was not large, the form of the relationship is consistent with many reported unimodal patterns. Thus, alteration in species pool size alone can lead to different productivity-diversity relationships even when examined at small spatial scales (lmz). The shift from negative 42 linear to unimodal is important because these are the two most common patterns reported for plant communities within geographic regions (Mittelbach et al. 2001). Therefore, variation in species pools among communities may account for the various reported PDRs. Taken together, my experiments show that local diversity depends upon both species pool size and community productivity as abiotic and competitive interactions filter available species pools. Although these results are from a short-term experiment, the relevance to longer term patterns is expected because seedling establishment is the critical stage for plant population establishment (Harper 1977). The next step is to determine whether removing seed limitation leads to successful reproduction by new colonists and increased coexistence or whether increased immigration leads to short-term increases in diversity through mass effects (Shmida and Wilson 1985). My results also suggest that management of biodiversity in grasslands requires augmentation of species pools. Historical species pools for many Midwestern (and other) grasslands have been dramatically altered by habitat loss, landscape fragmentation and the introduction of exotic species. Consequently, native species are lost from remnant communities because native species are unable to immigrate from other communities (Leach and Givnish 1996). Hence, maintenance of native diversity in highly fragmented systems will require species pool augmentation particularly when dispersal from other sites is unlikely. My results also indicate that abiotic factors limit species richness in low productivity grasslands. Hence, adding species capable of tolerating conditions found in 43 these low productivity systems will increase local diversity. Conversely, competitive interactions strongly limit diversity in high productivity systems suggesting that periodic disturbance will be required to maximize diversity in these systems. CHAPTER 4 THE IMPORTANCE OF SPECIES POOLS AND COLONIZER TRAITS TO LOCAL DIVERSITY ACROSS MULTIPLE PRODUCTIVITY GRADIENT S: TESTING MACARTHUR’S PARADOX ABSTRACT The debate between local vs. regional control of local species richness has been called MacArthur’s paradox because he was instrumental in the simultaneous development of both local and regional models of species richness. Despite recent advancements in theory and empirical evidence, resolution of the paradox remains unclear. I used seed addition experiments across replicated productivity gradients to test the importance of local vs. regional factors for local plant richness. My results showed that local environmental conditions limit species richness in low and high productivity communities. However, augmented species pools increased local richness in many communities suggesting that regional processes that control regional richness may also limit local diversity. Furthermore, I show that successful colonization along the productivity gradients is correlated with species traits measured in a greenhouse experiment. Consequently, the composition of species pools may also limit local richness when available species pools are a non-random collection of species from a regional species pool. Resolution of MacArthur’s paradox may be that both local and regional factors limit local species richness and the relative importance of each is related to community productivity. 45 INTRODUCTION Understanding the factors that limit local species diversity remains one of the most fundamental and unresolved issues in community ecology (Tilman and Pacala 1993, Palmer 1994, Grace 1999). Over the past few decades, two contrasting explanations have emerged to explain limits on species diversity at a site. The local view argues that species compete for a small number of resources so that diversity is constrained by niche differentiation and resource heterogeneity (Pianka 1966, MacArthur and Levins 1967, Schoener 1974). In contrast, the regional view contends that diversity is limited by immigration rates from a regional species pool (MacArthur and Levins 1967, Cornell and Lawton 1992, Zobel 1992). The debate between the local and regional control of diversity has been called “MacArthur’s Paradox” because MacArthur was instrumental in the simultaneous development of local and regional models of species diversity (Schoener 1983, Loreau and Mouquet 1999). One possible resolution of the paradox is that both local and regional factors limit species richness and that the relative importance of each factor varies in relation to resource supply (Huston 1999). For example, in low productivity systems, large species pools (regional factor) may have large effects on diversity because species interactions (local factors) are weak or occur slowly (Cornell and Lawton 1992). Conversely, in high productivity environments, species pools may have little effect on local diversity because species interactions are strong or competitive displacement is rapid (Grime 1979, Huston and Deangelis 1994, Keddy et al. 1997). The mediation of local-regional interactions by productivity has been called the shifting limitations hypothesis (SLH; Foster et al. 2004). 46 Two recent experiments have found support for the SLH (Foster et al. 2004); Houseman & Gross, chapter 3) on single grassland sites, but it is not known if the SLH is a general pattern across sites. The composition of the regional species pool may also influence local diversity. For example, the regional species pool may not be a random collection of species but rather a group of species adapted to specific environments (e. g. productivity ranges or disturbance conditions). Non-random species pools might occur as a consequence of evolutionary forces (Scharnp et al. 2002), the arrangement of communities on the landscape (metacommunities; Leibold et al. 2004) or through introduction of exotic species (Smith and Knapp 2001). Consequently, a species pool may have strong or weak effects on local diversity depending upon how the traits of species in the pool match local environmental conditions. For example, a species pool with many weedy species may have little effect on local diversity in an undisturbed, high productivity community because many available species have unsuitable traits and are sorted (filtered) by the environmental conditions (Keddy 1992, Belyea and Lancaster 1999, Booth and Swanton 2002); Chapter 2). Here I address the importance of local vs. regional control on species richness by asking three questions: 1) Is diversity limited by species pools (local vs. regional control) across replicated local productivity gradients? 2) Is the species pool-richness relationship related to community productivity? and 3) Do the traits of potential colonists limit local diversity? I tested the first two questions by augmenting species pools on four sites that 47 encompass broad gradients in productivity. I tested the third question by relating species traits measured in a greenhouse experiment to colonization success in the augmentation experiment. METHODS Field Data I selected four old-field sites in SW Michigan located in and around the Kellogg Biological Station (7km radius; Chapter 1) for the assessment of local vs. regional control on species richness. All of the sites had broad ranges in productivity (Chapter 1: Table 1.3) over relatively small scales (<200m). These sites had been farmed or managed and subsequently abandoned 10-40 years prior to my experiment. Within each site, I identified a low, medium and high aboveground productivity community that occurred along topographic gradients. Communities were delineated as a contiguous area with a relatively homogeneous species composition. In each community, I established a grid of sixteen 1.2x1.2m plots with 0.5-1m buffer strips. I augmented species pools by adding seeds of 40 species to eight randomly selected plots in each community. The species added were native to Michigan, represented a range of plant traits (Chapter 1: Table 1.4), and all but four were absent from all communities in the study. Seeds were obtained from commercial growers from Minnesota and Wisconsin and included source populations collected throughout the Midwest, USA. I sowed lglm2 of each species into seed addition plots in April of 2003. A consistent seed mass was used rather than number because of the potential tradeoffs between seed number, seedling survival and dispersal distance (W estoby et al. 2002). A high rate of seed addition was chosen to remove seed 48 limitation at all microsites within each plot. At the time of planting, I agitated the litter of all plots with a stiff rake to assure that the introduced seeds reached the soil surface. To promote seed germination in what was an unusually dry spring, all plots were watered once in late April equivalent to a single rainfall event of 2mm. Rainfall patterns returned to average levels shortly thereafter and no supplemental water was added thereafter. I sampled all vegetation in June and August of 2003, quantifying presence and percent cover of all resident vegetation in the central 1m2 of each plot. Percent cover was quantified by comparing plant cover of each species to reference squares of known area. Seedling germination and establishment were quantified in 0.25m2 (counts by species) and 1m2 (presence/absence) quadrats. To relate diversity patterns to local environmental conditions, I established four ancillary plots in each community that I sampled for light availability, biomass and soil resources. I measured light interception by the plant canopies in late August and early September with a Decagon sunfleck ceptometer (SF-80) in the central 0.5x1m of the plot. At each sample point, I inserted the ceptometer below the plant canopy (but above the litter) and recorded the mean of 80 light sensors located at 1cm intervals along the probe. A similar reading was taken every 10 cm along the short side of each plot. Hence, plot level light availability was the mean of 400 sensor readings (5 sample points x 80 sensors). All readings were taken i 2hrs of solar noon and were expressed as a percentage of light measured above the plant canopy. Following light measurements, I harvested aboveground biomass in the central 0.5x1m portion of the plot and separated live biomass 49 and litter. All biomass was dried for at least 48 hours at 60°C. Hence, aboveground biomass was a surrogate for aboveground net primary productivity. I sampled soils prior to harvest by collecting five 2.5—cm-wide by 10-cm-deep soil cores haphazardly located throughout the plot. Soils were placed in plastic bags and kept on ice until processed. In the lab, I sieved soils through a 2mm soil sieve. Soil moisture was obtained gravimetrically by measuring moisture lost from 15-20g of soil that was dried at 105°C for 48hrs. Soil N-pools were determined by extracting 20g wet soil with 50mL KCI for 24hrs. After samples were shaken for 1 minute and allowed to settle, samples were filtered through a l-Itm Gelman glass filter. The NOg' and NH] concentrations of the extract were determined by continuous flow analysis (Alpkem 1992). Soil N- mineralization potential was assayed by incubation. Two replicate 20g soil subsamples were placed into 150mL polyurethane cups and incubated for 35 days in an environmental chamber at 25°C. Samples were checked periodically for changes in moisture conditions by monitoring the mass of the sample and container. When necessary, samples were rehydrated with deionized water applied with a mister to return them to initial soil moisture levels. At the end of the incubation, soils were extracted and analyzed for N as described above. Species Traits To relate species traits to colonization success, I assayed a suite of morphological and physiological traits in laboratory and greenhouse experiments for 28 of the 40 species added in the field. Seed traits were quantified by germination trials conducted in an 50 environmental chamber while seedling growth and allocation patterns were assayed in a greenhouse experiment. Seed Traits. The seed of each species was thoroughly mixed and five sub-samples that contained at least 50 seeds were drawn to assess seed mass, total %germination and germination rate. After each sub-sample was counted and weighed, I placed the seeds in a Petri dish (8.6cm diameter) filled with moistened sand. I covered the Petri dishes with a ventilated lid to maintain high humidity and some air flow. Petri dishes were placed on trays in an environmental chamber maintained at a constant 25°C with a 12/12 hr. light/dark cycle. Trays were rotated daily within the chamber to minimize spatial variation. Each Petri dish was checked daily for germination and the sand was rewetted when necessary with deionized water applied with a mister. I used the day of radicle emergence as my assay of germination time and rate. The germination trial continued for 63 days. I calculated total germination percentage based on the percentage of seeds in each subsample that germinated. Because species differed in the timing of germination, I calculated germination rate as the number of days required for 50% of the seeds in a subsample to germinate (Grime et al. 1981). Seedling Traits. Following germination, seedlings were transplanted into plug trays filled with soil and placed in a heated greenhouse. This transitional stage was a way to insure that plants were of similar size and development for the greenhouse traits analysis. Each plugtray “cell” was 1.4 x 1.4cm and was filled with a sandy soil collected from a local field site. The drainage hole of the plug tray was filled with a small amount of peat moss 51 to prevent soil washout. The soil was similar to the low productivity soil found in the field experiments (90% sand, 3.6% silt, 6.4% clay; MSU soil lab) to allow manipulation of soil N. The soil was homogenized and sieved through a 4mm sieve to remove stones prior to adding it to the plug trays. To reduce desiccation from high light and temperature in the greenhouse, I enclosed the trays in a plastic frame covered with a 55% shade cloth. I also watered the trays daily as necessary and each tray was periodically rotated around the greenhouse bench to reduce spatial variation. Generally, each species was kept in this transitional stage for a few weeks until at least 20 individuals had matured to the point of having one true leaf. Following establishment in the plug trays, individual seedlings were transplanted into plastic “conetainers” (Stuewe & Sons, SC-lO Super Cell; 3.8cm width x 21cm depth; 164cm3) that were placed in racks capable of holding 98 ‘conetainers’. Individual conetainers were initially lined with #1 Whatman filter paper to prevent soil loss and then filled with the same field soil used in the transitional plantings. I left an empty conetainer between those filled with soil to minimize light competition between seedlings. The conetainer racks were placed in a foam box filled with medium-grain vermiculite that was periodically watered to maintain high humidity and reduce temperature fluctuations in the soil. A pair of these boxes (replicate) was required to accommodate one individual from each of the 40 species. Seedlings were randomly assigned to each of five replicates and the positions within the replicate were randomized. The paired boxes were rotated three times a week to new positions on the greenhouse benches to minimize spatial variation in light. 52 Following transplanting, each seedling received an initial dose of N (ammonium nitrate) dissolved in water and equivalent to 1.1g/m2. Thereafter N was added at this same level three times per week for a total of 38g/m:2 applied over the entire experiment for all species except Asclepias syriaca, which mistakenly received 39g/m2. These rates were chosen so that the potential response of each species under ‘ideal’ growing conditions (high light and nutrients) could be compared to field responses. These addition rates were similar to other fertilization experiments conducted in low fertility soils (Gross et al. in press). Conetainers were weeded and watered as necessary to prevent moisture stress. I harvested plants after about 14 weeks of growth in the ‘ideal environment” and measured the following traits: specific leaf area (SLA), available soil N as a surrogate for N uptake, rootzshoot ratio and root and shoot growth rates (Weiher et al. 1999, Cornelissen et al. 2003). To measure SLA, I harvested 1-3 fully expanded apical leaves from each plant. These leaves were scanned with an Epson 1680 scanner and converted to area measures with Scion Image v4.02 software. Scanned leaves and remaining shoots were dried separately for at least 48hrs at 60°C. Soil was carefully removed from the root system and processed for soil N as described for the field soils. Roots were carefully washed, scanned and dried (as above). Initial root and shoot biomass was based on mean root and shoot biomass of 5-10 randomly selected individuals for each species at the time of transplanting into conetainers. Rootzshoot allocation was based on final masses while root and shoot growth rates were calculated as 53 Growth rate = Ln [(Final mass-initial mass)/time] Statistical Analysis I compared plant species richness among the 4 sites and 12 communities with a three-way ANOVA (SAS v9.0). Relationships between species richness and light availability were tested with regression (SYSTAT v9.0). When a significant quadratic relationship was found, I used the Mitchell-Olds (MOS) test to determine if the peak of the curve occurred within the data range (Mitchell-Olds and Shaw 1987). To test for species sorting along the gradients, I used Indicator Species Analysis (ISA; Dufrene and Legendre 1997, McCune and Grace 2002). The ISA calculated an importance value based on occurrence and abundance in the 0.5m2 plots and tested for significant association with low, medium or high productivity environments with a permutation test. I used a fourth-corner analysis (FCA) to relate species traits from the greenhouse and lab assays to environmental conditions found in the field (Legendre et al. 1997). Because of intercorrelation between variables for the trait and environmental matrices, I reduced the number of variables for both species traits and environmental conditions with PCA after standardizing all variables (McCune and Grace 2002). I then performed a FCA by creating three matrices. The first matrix was a species x plot matrix based on presence-absence measures in the 1m2 plots. The second matrix was a plot x environment matrix using the first two PCA axis scores for the environmental variables. The final matrix was a species x trait matrix using the first four PCA axis scores for traits. From these three matrices, a fourth matrix was derived where species traits were 54 related directly to environmental conditions. Significant association was tested with permutation tests (10,000 runs) and p-values were adjusted for multiple tests with the Hope correction. All ordinations were performed with PC-ORD v4.0. RESULTS Light-richness patterns along the gradients I examined species richness patterns across the sites as a function of light availability as a surrogate for aboveground biomass because the two were highly correlated (R2 = 0.80, n = 48; Figure 4.1.). When all four sites were combined, I detected a unimodal relationship between resident species richness and light availability (quadratic term: p<0.0001, Figure 4.2). Similarly, species richness of successful colonists was unimodally related to light availability (quadratic term: p<0.0001, Figure 4.3). The MOS test indicated that both resident and colonist quadratic functions were unimodal (p<0.0001 and p<0.0001, respectively) with peak richness occurring near 45% of full light. However, less variance was explained by the light-richness relationship for colonists (R2 = 0.17) compared to resident species (R2 = 0.36). When sites were examined separately, I found similar unimodal light-richness patterns for resident species on three sites (Figure 4.4; Table 4.1). Seed addition plots had similar light-richness patterns on all four sites. Because I had a three-way interaction between site, environment and seed addition (F = 5.69, p<0.0001), I used mean slicing to compare the effect of seed addition at each site. Species number increased (Table 4.1) at all sites 55 % Light Availability (PPFD) <30 0 100 200 300 400 500 600 700 800 900 Aboveground Biomass (g/m2) Figure 4.1. Regression of light on aboveground biomass. 56 Resident Species Richness 20 I I r r r r T O O O O O O O 15- O O O O O - (:1 g g g 3 3 ° Quadradrc term: p<0.0001 75 o o o 0 R2: 0.358 .9 o o o 8 10- o o o o o o 1 O O c?)- o o o o o o O O O O O O 5 O O O O 0 I I I I I I °I 7o 60 50 4o 30 20 10 % of Full Light (PPFD) Figure 4.2. Resident species richness as a function of light availability across four sites (12 communities). 57 Colonist Species Richness 20 I I I I I I I o 15" ‘ N o E 0 o 8’ 3 o 3 o o Quadratic term p<0.0001 '310- ° o o o o o - R2=0.173 o. o o o (D o o I I I I I l .I- 7o 60 50 4o 30 20 i6 0 % of Full Light (PPFD) Figure 4.3. Number of species that colonized following seed addition as a function of light availability across four sites (12 communities). 58 Site 1 Site 2 Site 3 Site 4 GD 0 N O .I. C Species Richness/m2 O \ I" I I I I'l 80604020 0 604020 0 604020 0 604020 0 % Light Availability % Light Availability % Light Availability % Light Availability I \ 1" ’u I I I Figure 4.4. Species richness as a function of light availability for each site for resident species (circles, solid lines) and total richness (resident + colonizers: triangles, dashed lines). 59 Table 4.1. Tests for an effect of seed addition on total richness. The ANOVA table represents mean slicing by site. Site DF MS F-value p-value 1 5 331.3 49.8 <.0001 2 5 188.8 28.4 <.0001 ' 4 5 367.2 55.2 <.0001 5 5 239.9 36.1 <.0001 6O in response to seed addition suggesting that observed species richness is partially limited by species pools. When smoothed unimodal curves were fit to patterns found in individual sites, both the peak and shapes of the light-richness curves suggested similar patterns among the four sites. Although richness increased in most communities with augmented species pools, richness in two of the high productivity communities (<15% of full light) did not change (Site 4: t < 0.001, df = 14, p = 1.000; Site 4: t = -1.580, df = 14, p = 0.136). Species sorting of successfitl colonists along the gradients I found evidence for species sorting along the productivity gradients. Of the twenty-two species that colonized at least one plot in the experiment, nearly half had a significant association with a particular portion of the gradient (low, medium or high; Figure 4.5). Differences in importance values for the three community types suggested that several species responded strongly at low productivity while a larger group of species had peak values at medium productivity. One species had a peak importance value at high productivity. I used PCA to reduce the intercorrelation among environmental (Appendix 1) and species traits data before using the fourth-comer analysis (FCA) to determine the traits associated with colonization success. The PCA of environmental variables (PCAE) indicated that the first two axes explained 80% of the variation in plots among all sites. Axis I explained 69% of the variation with plant live biomass, litter, soil moisture and N- mineralization potentials loading positively and light availability loading negatively 61 60 50 r 40 r 30 r 20: Importance Value From ISA -+-— PETPUR + ANDGER + SCHSCO + KUHEUP —)l(— ASCTUB —e— LESCAP —+— SENSPL —B- DESCAN —-— HELMAX + ASCSYR —A— MONFIS Lo Medium Environment Figure 4.5. Indicator Species Analysis for colonizing species. Species Codes are as follows: PETPUR = Petalostemum purpureum, ANDGER = Andropogon gerardii, SCHSCO = Schizachyrium scoparium, KUHEUP = Kuhnia eupatoriodes, ASCTUB = Asclepias tuberosa, LESCAP = Lespedeza capitata, AMOCAN = Amorpha canescens, DESSPP = Desmodium spp., HELMAX = Helianthus maximiliani, ASCSYR = Asclepias syriaca, MONFIS = Monarda fistulosa. 62 (Table 4.2). Axis 2 explained an additional 11% of the variation and was associated with increasing soil inorganic N and litter. The PCA of species traits (PCAT) revealed that the first four axes explained 91% of the variation among species. Axis 1 explained 48% of the variation and was positively related to root and shoot growth rates (Table 4.3). Axis 2 explained 19% of the variation and was positively related to rootzshoot allocation. Axis 3 explained 15% of the variation and was positively related to seed mass and germination rate, but negatively related to total germination percentage. Axis 4 explained 8% of the variation and was positively, but weakly, related to SLA. Consequently, four types of traits differentiated species: growth rates, allocation, seed traits and SLA. The FCA indicated that the four PCA—r axes were associated with successful colonization under specific environmental conditions. Species that successfully colonized low productivity environments had either low growth rates, high allocation to roots, large/fast germinating seeds (but low total germination) or low SLA (PCAT1-4; Table 4.4). Species with the opposite set of traits were more successful in high productivity environments. Results were nearly identical when productivity or the inverse of light availability (data not shown) was used in place of PCAE Axis I emphasizing the close association between PCAE Axis I and productivity. Species with high allocation to shoots (PCAT Axis 2) and low SLA were more successful in sites with high nitrate pools and N-mineralization potentials (PCAE 2; Table 4.4). However, correlations with PCAE 2 were very small despite statistical significance. 63 Table 4.2. Loadings of environmental variables on the first three PCAE axes. Total N-Pool (NO3- + NH4+) N-mineralization Live Biomass/M2 Litter Biomass/M2 %SOIL Moisture PAR Eigenvector Axis I Axis II -0.343 -0.679 -O.412 -0.291 -0.445 0.172 -0.351 0.647 -0.425 0.036 0.459 -0.075 64 Table 4.3. Loadings of species trait variables on the first four PCAT axes. Seed Mass Total Germination Germination Rate NO3-N Pool NHa-N Pool SLA Root/Shoot Shoot Growth Rate Root Growth Rate Axis I 0.244 0.287 -0.289 -0.388 —0.374 -0.236 0.089 0.460 0.457 Ei genvector Axis 2 0.050 0.364 -0.237 0.407 0.395 -0.156 0.680 0.025 0.025 65 Axis 3 0.653 -0.422 0.519 0.092 0.103 -0.175 0.193 0.142 0.136 Axis 4 0.211 0.003 -0.072 —0.042 -0.022 0.938 0.203 0.104 0.125 Table 4.4. Results of the fourth—comer analysis. The r-value is the Pearson correlation coefficient. ENV PCA 1 r(i,j) p-value ENV PCA2 r(i,j) p-value Growth PCAl 0.200 0.0001 0.016 0.0661 Allocation 66 PCA2 -0.111 0.0002 -0.075 0.0480 Seed PCA3 -0.167 0.0001 —0.025 0.0765 SLA PCA4 0.063 0.0095 -0.069 0.4913 DISCUSSION Limits to species diversity along productivity gradients My results show that local environmental conditions limit species richness in low and high productivity grassland communities. This conclusion is substantiated by the remarkable similarity in the unimodal productivity-richness relationship across four sites with similar productivity gradients. Limitation of species richness in low and high resource environments is likely due to abiotic conditions (stresses) at low productivity and competitive interactions at high productivity (Grime 1979, Huston 1994). Although the relationship of competitive intensity to productivity remains highly contested (Grace 1993, Goldberg et al. 1999, Arii and Turkington 2001), few studies have examined how colonization varies along productivity gradients. Most empirical and theoretical studies of competition along productivity gradients have focused on changes in growth rates but have not addressed survival rates particularly at the seedling stage (Goldberg and Novoplansky 1997). My experiment and those in chapter 3 suggest that competitive effects on colonization vary strongly across productivity gradients and demonstrate how local factors constrain species richness. My data also reveal that species pools limit local richness across productivity gradients. Following seed addition, richness increases in nearly all communities. The similar responses across multiple gradients suggest that communities are commonly limited by immigration. Other species addition experiments in single systems report similar results (T ilman 1997, Tumbull et al. 2000, Zobel et al. 2000, Foster and Tilman 2003, Foster et al. 2004, Gross et al. in press). Furthermore, my experiment shows that large species 67 pools have greater effects on richness at intermediate productivity than at low or high productivity. Thus, the relative importance of species pools to local diversity appears to be mediated by productivity (Huston 1999, Foster et al. 2004). However, it is not clear if productivity or factors associated with productivity control the species pool effect. Soil N and water are positively and light negatively related to productivity, but colonization may also depend upon light quality (Bell et al. 1999) or interference from litter (Foster and Gross 1998). Further work is needed to determine whether resource supply rates or other environmental variables drive differences in colonization along productivity gradients. In addition to species pool size, the composition of species pools also affects local richness. Because species sort along productivity gradients based on colonizer traits, local diversity is also constrained by the traits of the species in the available species pool. Consequently, richness might deviate from the expected unimodal pattern if the available species pool is a non-random collection of species. For example, in a low productivity community, species richness may be lower than expected because the available species pool has relatively few species with suitable traits to colonize at low productivity. Hence, a large species pool with many species possessing unsuitable traits may have less effect on local richness than a small species pool with species possessing suitable traits. Non- random species pools are likely to occur because of differences natural landscape variation, habitat fragmentation and differences in dispersal associated with seed characteristics and vectors (Munoz et al. 2004). 68 «I While I show that species pools have strong effects on richness, it is not clear whether this result is an example of mass effects (Shmida and Wilson 1985) or increased species coexistence. Here, I have focused on the establishment phase of colonization because it is considered the critical stage of population growth (Grubb 1977, Harper 1977). Other experiments conducted over several growing seasons indicate that increased species pools have lead to long-term coexistence (Tilman 1997, Foster et al. 2004) but short-term mass effects are also found (Thompson et al. 2001, Gross et al. in press). Regardless, reported patterns of species richness do not discriminate between reproductive and non- reproductive individuals. Hence, regular immigration from species pools likely contributes to diversity patterns through both species coexistence and mass effects. My results suggest that, rather than a paradox, MacArthur’s local and regional approaches are both required to explain local species richness patterns (Ricklefs 1987). However, the relative importance of local and regional factors may be related to community productivity as predicted by the SLH (Huston 1999, Foster et a1. 2004). Consequently, local richness patterns may be explained by the integration of species pool size and composition along with environmental conditions. Species trait and colonization along productivity gradients My results show that colonization success across productivity gradients is related to a four types of species traits. First, species with high growth rates had greater colonization in high than low productivity environments. Reviews of experimental and observation 69 studies also indicate that plants with high growth rates are found where soil resources are high (Grime and Hunt 1975, Gaudet and Keddy 1988, Poorter and Gamier 1998). Presumably, high growth rates increase resource preemption thereby enhancing competitive ability (Lambers et al. 1998, Poorter and Gamier 1998). Preemption may be important at the colonization stage if there is competition between seedlings for a limited number of available patches. Additionally, in high productivity environments where light is the critical limiting factor, high growth rates may allow species to capitalize on high light availability early in the growing season. Second, high allocation to roots was negatively related to productivity. Other physiological studies indicate that plants invest more resources in root than shoot biomass in low productivity or stressful environments (Lambers et al. 1998). Such responses are likely linked to greater below- than above-ground resource limitation. Because growth rates and allocation patterns were orthogonal axes in the PCA, it appears that the importance of allocation patterns to colonization is independent of growth rates. Third, species with large or fast germinating seeds were more successful in low productivity environments. Studies of distributional patterns often report that seed mass and shade are positively correlated (Salisbury 1974, Mazer 1989, Hewitt 1998) but this pattern may depend upon whether low shade environments are drought-prone (Baker 1972). A recent review of seed mass—environment relationships, suggests that larger seed mass is an advantage in “hazardous” environments (e. g. high shade, nutrient shortage or drought; (Moles and Westoby 2004). Hence, high seed mass may be an advantage in my 70 low productivity communities because of droughty conditions. Additionally, the negative correlation between seed mass and productivity reported here may be dependent upon germination rate because both seed mass and gemrination rate were related to PCA axis III. In low productivity environments, high germination rates may allow species to maximize growth early in the growing season before high soil temperature and low soil moisture develop. Fourth, species with high SLA were more successful in high productivity environments although the amount of variance explained was small. The success of high SLA species with increased productivity is consistent with both field studies (Ellsworth and Reich 1992, Fonseca et al. 2000) and physiological models (Lambers et al. 1998). In this analysis, SLA may explain relatively little variation in colonization because all individuals were grown under high light conditions. Further work is currently underway to examine whether the importance of this trait changes when measured under a range of light and nutrient conditions. My analysis indicates that a suite of traits can be used to predict colonization along productivity gradients. The strength of my approach is that I removed colonization limitation and examined the traits of species that failed to establish as well as those that established. Most studies of plant traits have only examined extant species. Consequently, the strength of a trait-environment correlation may change when all species present--including the seed stage--are factored into the analysis. The traits identified in this analysis focus on the establishment stage as opposed to adult or 71 reproductive stages. While the establishment stage is often the critical to population growth (Harper 1977; Howard and Goldberg 2001), other traits may become more important at different life stages. 72 CHAPTER 5 SUMMARY AND FUTURE DIRECTIONS SUMMARY OF FINDINGS Results presented in this dissertation suggest that both local and regional factors are important determinants of local patterns of diversity in grasslands. The observational data (chapter 2) indicated that the unimodal productivity-diversity relationship (PDR) across communities is a general pattern given a sufficient range in productivity, a large number of communities, and identification of regional scales that match key region processes. Because local scale patterns could not account for these unimodal patterns, species richness is probably controlled by an interaction between local and regional factors. Data from the experimental studies involving seed augmentation (chapters 3 & 4) showed ' that local environmental conditions limited species richness at low and high productivity communities. Recruitment was reduced in low productivity communities due to abiotic (stressful) conditions, whereas in the high productivity sites competitive species interactions limited colonization. The importance of these effects on recruitment and richness was supported by unimodal PDRs for resident species across the four natural productivity gradients. Species richness increased with species pool augmentation across all sites, but larger increases were generally found in intermediate productivity communities suggesting that 73 regional factors also limit local diversity. Consequently, the failure to detect unimodal PDRs may in part be due to reductions in the available species pool. The analysis of species traits also shows that it is not simply the number of species that will be critical to local diversity. In other words, a large species pool with many species possessing unsuitable traits may have less effect on local richness than a small species pool with species possessing suitable traits. The results of these studies demonstrate that both local and regional effects appear to control local species richness in these grasslands. While other studies have reported similar findings (Ricklefs 1987, Tilman 1997, Zobel 1997, Foster 2001, Foster et al. 2004), my results demonstrate that the relative importance of local and regional effects might be controlled by community productivity. Community productivity is a composite variable that represents resource availability and stress conditions. Consequently, productivity provides a link between abiotic and competitive conditions that occur at local scales and potential colonization from available species pools. FUTURE DIRECTIONS The results reported in this dissertation demonstrate that integrating local and regional processes across productivity gradients can offer insight into complex patterns of species richness in grassland systems. This is also one of the first studies to show how specific traits lead to successful colonization across replicated productivity gradients. While these are potentially important insights for ecology, several ideas need further examination. 74 One of the most pressing questions is whether the results from the seed addition experiments occurred because of species coexistence or mass effects. The increases in richness may be transient if few species survive to reproduction, or if colonizers replace resident species. Some studies report that increased diversity from species pool augmentation is maintained for several growing seasons (Tilman 1997, Foster et al. 2004), while this effect is transient in others (Thompson et al. 2001). I plan to follow patterns of persistence on my study sites over the next few years to determine how species pools affect longer-term patterns of coexistence. In addition to species pool effects on richness, it is not known how species pools and community productivity affect species evenness. There is some evidence that evenness in extant communities may relate to productivity and a preliminary analysis of my observational data suggests that species evenness decreases with increasing productivity in these sites. This pattern raises the possibility that decreased evenness may influence colonization (invisibility) under different resource conditions (Wardle 2001). The results of the seed addition experiments reported here may be driven by seed addition rates or the growing conditions found in specific years. Environmental variability among years has been shown theoretically to be an important mechanism of coexistence (Chesson 1994); yet few seed addition experiments have tested this idea. My experiments showed similar responses to seed addition on Site 2 in two different years suggesting that my results are unlikely related to a single year with unusually good conditions for seedling growth. However, more work is needed to determine the 75 importance of temporal variability to colonization. In addition to environmental variation among years, immigration rates may vary spatially or temporally. Most seed addition experiments, including those presented here, have chosen an arbitrary seed addition rate because natural immigration rates are difficult to obtain. Consequently, future studies should vary seed addition rates to determine if the species pool effect is driven by artificial seed addition rates. While results over multiple natural productivity gradients suggest that community productivity is a key driver of local-regional dynamics, the experiments in this dissertation cannot separate the effect of resident species composition and productivity. Consequently, I have initiated an experiment in which I test how resource availability (productivity) and resident species traits may influence colonization. In this experiment, I have manipulated fertility levels in mesocosms by adding N or C to communities that are dominated by clonal or non-clonal species. This experiment will allow me to examine how variation in productivity affects colonization in communities starting with the same species composition. Although I have shown that colonizer traits influence potential colonization, the analysis presented here may be biased by growth under “ideal” conditions of high nutrients and light. These conditions may not reveal difference between the response of generalist and specialist species to environmental variation. For example, a species that is adapted to conditions found at low productivity may not respond in a meaningful way to the artificial conditions of the “ideal” treatment. Consequently, I have also conducted a 76 similar assay of plant traits in three environments that represent a simulated productivity gradient (high light/low nutrient to low light/high nutrient). This additional experiment will allow me to relate seedling performance under a range of conditions to success in the field and to directly test models of species richness linked to resource use (Herbert et al. 2004). The future directions described above illustrate that much more work is needed to understand the factors that control local species diversity. Yet, integrating species pools, productivity and species traits appears to be a promising approach to the complex question of what controls species diversity. 77 LITERATURE CITED Abrams, P. A. 1987. Indirect interactions between species that share a predator: varieties of indirect effects. in W. C. Kerfoot and A. Sih, editors. Predation: direct and indirect impacts on aquatic communities. University Press of New England, Hanover, NJ. Abrams, P. A. 1995. Monotonic or unimodal diversity productivity gradients - What does competition theory predict? Ecology 76:2019-2027. Alpkem. 1992. The flow solution operation manual. Alpkem, Wilsonville, Oregon. Arii, K., and R. Turkington. 2001. Assessing competition intensity along productivity gradients using a simple model. Canadian Journal of Botany-Revue Canadienne De Botanique 79:1486-1491. Austin, F. R. 1979. Soil survey of Kalamazoo County, Michigan. U. S. Department of agriculture, Washington, D. C. Baker, H. G. 1972. Seed weight in relation to environmental conditions in California. Ecology 53:997-1010. Bazzaz, F., G. Ceballos, M. Davis, R. Dirzo, P. R. Ehrlich, T. Eisner, S. Levin, J. H. Lawton, J. Lubchenco, P. A. Matson, H. A. Mooney, P. H. Raven, J. E. Roughgarden, J. Sarukhan, G. D. Tilman, P. Vitousek, D. H. Wall, E. O. Wilson, and G. M. Woodwell. 1998. Ecological science and the human predicament. Science 282:879-879. Bell, D. T., L. A. King, and J. A. Plummer. 1999. Ecophysiological effects of light quality and nitrate on seed germination in species from Western Australia. Australian Journal of Ecology 24:2-10. Belyea, L. R., and J. Lancaster. 1999. Assembly rules within a contingent ecology. Oikos 86:402-416. Booth, B. D., and C. J. Swanton. 2002. Assembly theory applied to weed communities. Weed Science 50:2-13. Bromham, L., and M. Cardillo. 2003. Testing the link between the latitudinal gradient in species richness and rates of molecular evolution. Journal of Evolutionary Biology 16:200-207. Brown, J. H. 1971. Mammals on mountaintops - Nonequilibrium insular biogeography. American Naturalist 105:467-&. Caley, M. J ., and D. Schluter. 1997. The relationship between local and regional diversity. Ecology 78:70-80. 78 Callaway, R. M., R. W. Brooker, P. Choler, Z. Kikvidze, C. J. Lortie, R. Michalet, L. Paolini, F. L. Pugnaire, B. Newingham, E. T. Aschehoug, C. Arrnas, D. Kikodze, and B. J. Cook. 2002. Positive interactions among alpine plants increase with stress. Nature 417:844-848. Chase, J. M., and M. A. Leibold. 2002. Spatial scale dictates the productivity-biodiversity relationship. Nature 416:427-430. Chesson, P. 1994. Multispecies competition in variable environments. Theoretical Population Biology 45:227-276. Connell, J. H. 1979. Interrnediate-disturbance hypothesis. Science 204:1345-1345. Comelissen, J. H. C., S. Lavorel, E. Gamier, S. Diaz, N. Buchmann, D. E. Gurvich, P. B. Reich, H. ter Steege, H. D. Morgan, M. G. A. van der Heijden, J. G. Pausas, and H. Poorter. 2003. A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Australian Journal of Botany 51:335-380. Cornell, H. V. 1999. Unsaturation and regional influences on species richness in ecological communities: A review of the evidence. Ecoscience 61303-315. Cornell, H. V., and J. H. Lawton. 1992. Species interactions, local and regional processes, and limits to the richness of ecological communities - a theoretical perspective. Journal of Animal Ecology 61: 1-12. Comwell, W. K., and P. J. Grubb. 2003. Regional and local patterns in plant species richness with respect to resource availability. Oikos 100:417-428. Cresswell, J. E., V. M. Vidalmartinez, and N. J. Crichton. 1995. The Investigation of saturation in the species richness of communities - some comments on methodology. Oikos 72:301-304. Currie, D. J ., and V. Paquin. 1987. Large-scale biogeographical patterns of species richness of trees. Nature 329:326-327. Diamond, J. M. 1974. Colonization of exploded volcanic islands by birds - supertramp strategy. Science 184:803-806. Dirzo, R., and P. H. Raven. 2003. Global state of biodiversity and loss. Annual Review of Environment and Resources 28: 137-167. Dufrene, M., and P. Legendre. 1997. Species assemblages and indicator species: The need for a flexible asymmetrical approach. Ecological Monographs 67:345-366. Ellsworth, D. S., and P. B. Reich. 1992. Leaf mass per area, nitrogen-content and photosynthetic carbon gain in Acer saccharum seedlings in contrasting forest light environments. Functional Ecology 6:423-435. 79 Fonseca, C. R., J. M. Overton, B. Collins, and M. Westoby. 2000. Shifts in trait- combinations along rainfall and phosphorus gradients. Journal of Ecology 88:964- 977. Foster, B. L. 2001. Constraints on colonization and species richness along a grassland productivity gradient: The role of propagule availability. Ecology Letters 4:530- 535. Foster, B. L., T. L. Dickson, C. A. Murphy, 1. S. Karel, and V. H. Smith. 2004. Propagule pools mediate community assembly and diversity-ecosystem regulation along a grassland productivity gradient. Journal of Ecology 92:435-449. Foster, B. L., and K. L. Gross. 1998. Species richness in a successional grassland: Effects of nitrogen enrichment and plant litter. Ecology 79:2593-2602. Foster, B. L., and D. Tilman. 2003. Seed limitation and the regulation of community structure in oak savanna grassland. Journal of Ecology 91:999-1007. Gamarra, J. G. P., and R. V. Sole. 2002. Biomass-diversity responses and spatial dependencies in disturbed tallgrass prairies. Journal of Theoretical Biology 215:469-480. Gaudet, C. L., and P. A. Keddy. 1988. A comparative approach to predicting competitive ability from plant traits. Nature 334:242-243. Goldberg, D., and A. N ovoplansky. 1997. On the relative importance of competition in unproductive environments. Journal of Ecology 85:409-418. Goldberg, D. E., T. Rajaniemi, J. Gurevitch, and A. Stewart-Oaten. 1999. Empirical approaches to quantifying interaction intensity: Competition and facilitation along productivity gradients. Ecology 80:11 18-1 131. Grace, J. B. 1993. The effects of habitat productivity on competition intensity. Trends in Ecology & Evolution 8:229-230. Grace, J. B. 1995. In search of the holy-grail - explanations for the coexistence of plant- species. Trends in Ecology & Evolution 10:263-264. Grace, J. B. 1999. The factors controlling species density in herbaceous plant communities. Perspectives in Plant Ecology Evolution and Systematics 2: 1-28. Grace, J. B., L. Allain, and C. Allen. 2000. Factors associated with plant species richness in a coastal tall-grass prairie. Journal of Vegetation Science 11:443-452. Grace, J. B., and B. H. Pugesek. 1997. A structural equation model of plant species richness and its application to a coastal wetland. American Naturalist 1492436- 460. 80 Grime, J. P. 1973. Competition and diversity in herbaceous vegetation - Reply. Nature 244:31 1-31 1. Grime, J. P. 1979. Plant strategies and vegetation process. John Wiley, New York, NY. Grime, J. P., and R. Hunt. 1975. Relative growth-rate - its range and adaptive significance in a local flora. Journal of Ecology 63:393-422. Grime, J. P., G. Mason, A. V. Curtis, J. Rodman, S. R. Band, M. A. G. Mowforth, A. M. Neal, and S. Shaw. 1981. A comparative-study of germination characteristics in a local flora. Journal of Ecology 69: 1017-1059. Gross, K. L., G. Mittelbach, and H. L. Reynolds. in press. Grassland invasibility and diversity: Responses to nutrients, seed input, and disturbance. Ecology. Gross, K. L., M. R. Willig, L. Gough, R. Inouye, and S. B. Cox. 2000. Patterns of species density and productivity at different spatial scales in herbaceous plant communities. Oikos 89:417—427. Grubb, P. J. 1977. Maintenance of species—richness in plant communities - importance of regeneration niche. Biological Reviews of the Cambridge Philosophical Society 52: 107-145. Guo, Q. E, and W. L. Berry. 1998. Species richness and biomass: Dissection of the hump-shaped relationships. Ecology 79:2555-2559. Harper, J. L. 1977. The population biology of plants. Academic Press, New York, New York. Herben, T. 2000. Correlation between richness per unit area and the species pool cannot be used to demonstrate the species pool effect. Journal of Vegetation Science 11:123-126. Herbert, D. A., E. B. Rastetter, L. Gough, and G. R. Shaver. 2004. Species diversity across nutrient gradients: An analysis of resource competition in model ecosystems. Ecosystems 7:296-310. Hewitt, N. 1998. Seed size and shade-tolerance: a comparative analysis of North American temperate trees. Oecologia 114:432-440. Holt, R. D. 1977. Predation, apparent competition, and structure of prey communities. Theoretical Population Biology 12: 197-229. Howard, T. G., and D. E. Goldberg. 2001. Competitive response hierarchies for germination, growth, and survival and their influence on abundance. Ecology 82:979-990. 81 Huston, M. A. 1994. Biological diversity: The coexistence of species on changing landscapes. Cambridge University Press, Cambridge. Huston, M. A. 1999. Local processes and regional patterns: Appropriate scales for understanding variation in the diversity of plants and animals. Oikos 86:393-401. Huston, M. A., and D. L. Deangelis. 1994. Competition and coexistence - the effects of resource transport and supply rates. American Naturalist 144:954-977. Keddy, P., L. TwolanStrutt, and B. Shipley. 1997. Experimental evidence that interspecific competitive asymmetry increases with soil productivity. Oikos 80:253-256. Keddy, P. A. 1992. Assembly and response rules - 2 goals for predictive community ecology. Journal of Vegetation Science 3: 157-164. Kull, O., and A. Aan. 1997. The relative share of graminoid and forb life-forrns in a natural gradient of herb layer productivity. Ecography 20: 146-154. Lambers, H., F. S. Chapin, and T. L. Pons. 1998. Plant physiological ecology. Springer- Verlag, New York, New York. Leach, M. K., and T. J. Givnish. 1996. Ecological determinants of species loss in remnant prairies. Science 273: 1555-1558. Legendre, P., R. Galzin, and M. L. HarmelinVivien. 1997. Relating behavior to habitat: Solutions to the fourth-comer problem. Ecology 78:547-562. Leibold, M. A. 1989. Resource edibility and the effects of predators and productivity on the outcome of trophic interactions. American Naturalist 134:922-949. Leibold, M. A. 1999. Biodiversity and nutrient enrichment in pond plankton communities. Evolutionary Ecology Research 1:73-95. Leibold, M. A., and J. M. Chase. 2003. Ecological niches: Linking classical and contemporary approaches. University of Chicago Press, Chicago, IL. Leibold, M. A., M. Holyoak, N. Mouquet, P. Amarasekare, J. M. Chase, M. F. Hoopes, R. D. Holt, J. B. Shurin, R. Law, D. Tilman, M. Loreau, and A. Gonzalez. 2004. The metacommunity concept: A framework for multi-scale community ecology. Ecology Letters 72601-613. Liira, J ., and K. Zobel. 2000. Vertical structure of a species-rich grassland canopy, treated with additional illumination, fertilization and mowing. Plant Ecology 146: 185- 195. Loreau, M. 2000. Are communities saturated? On the relationship between alpha, beta and gamma diversity. Ecology Letters 3:73-76. 82 Loreau, M., and N. Mouquet. 1999. Immigration and the maintenance of local species diversity. American Naturalist 154:427-440. Lundholm, J. T., and D. W. Larson. 2003. Relationships between spatial environmental heterogeneity and plant species diversity on a limestone pavement. Ecography 26:715-722. MacArthur, R. H., and R. Levins. 1967. The limiting similarity, convergence, and divergence of coexisting species. American Naturalist 101:377-385. MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, N. J. Mazer, S. J. 1989. Ecological, taxonomic, and life-history correlates of seed mass among Indiana dune Angiosperms. Ecological Monographs 59:153-175. McCune, B., and J. B. Grace. 2002. Analysis of ecological communities. MjM software design, Gleneden Beach, OR. Miller, T. E. 1994. Direct and indirect species interactions in an early old-field plant community. American Naturalist 143:1007-1025. Mitchell-Olds, T., and R. G. Shaw. 1987. Regression-analysis of natural-selection - statistical-inference and biological interpretation. Evolution 41:1149-1161. Mittelbach, G. G., C. F. Steiner, S. M. Scheiner, K. L. Gross, H. L. Reynolds, R. B. Waide, M. R. Willig, S. I. Dodson, and L. Gough. 2001. What is the observed relationship between species richness and productivity? Ecology 82:2381-2396. Moles, A. T., and M. Westoby. 2004. Seedling survival and seed size: A synthesis of the literature. Journal of Ecology 92:372-383. Moore, D. R. J., and P. A. Keddy. 1989. The relationship between species richness and standing crop in wetlands - the importance of scale. Vegetatio 79:99-106. Morin, P. J. 1999. Community ecology. Blackwell, Malden, Massachusetts. Munoz, J., A. M. Felicisimo, F. Cabezas, A. R. Burgaz, and I. Martinez. 2004. Wind as a long-distance dispersal vehicle in the Southern Hemisphere. Science 304: 1 144- 1 147. Palmer, M. W. 1994. Variation in species richness: Towards a unification of hypotheses. Folia Geobotanica and Phytotaxonomica 29:511-530. Palmer, M. W., and P. M. Dixon. 1990. Small-scale environmental heterogeneity and the analysis of species distributions along gradients. Journal of Vegetation Science 1:57-65. 83 Palmer, M. W., and P. S. White. 1994. Scale dependence and the species-area relationship. American Naturalist 144:717-740. Partel, M., M. Zobel, K. Zobel, and E. vanderMaarel. 1996. The species pool and its relation to species richness: Evidence from Estonian plant communities. Oikos 75:1 1 1-1 17. Pausas, J. G., J. Carreras, A. Ferre, and X. Font. 2003. Coarse-scale plant species richness in relation to environmental heterogeneity. Journal of Vegetation Science 14:661- 668. Pianka, E. R. 1966. Latitudinal gradients in species diversity: A reveiw of the concepts. American Naturalist 100:33-46. Poorter, H., and E. Garnier. 1998. Ecological significance of inherent variation in relative growth rate and its components. Pages 901 in Handbook of functional plant ecology. M. Dekker, New York, New York. Rajaniemi, T. K. 2003. Explaining productivity-diversity relationships in plants. Oikos 101:449-457. Ricklefs, R. E. 1987. Community diversity - relative roles of local and regional processes. Science 235: 167-171. Rosenzweig, M. L., and Z. Abramsky. 1993. How are diversity and productivity related? Pages 52-65 in R. E. Ricklefs and D. Schluter, editors. Species diversity in ecological communities. University of Chicago Press, Chicago, IL. Salisbury, F. R. S. 1974. Seed size and mass in relation to environment. Proceedings of the Royal Society of London Series B 186:83-88. Sax, D. F. 2001. Latitudinal gradients and geographic ranges of exotic species: implications for biogeography. Journal of Biogeography 28:139-150. Schamp, B. S., R. A. Laird, and L. W. Aarssen. 2002. Fewer species because of uncommon habitat? Testing the species pool hypothesis for low plant species richness in highly productive habitats. Oikos 97: 145-152. Scheiner, S. M., S. B. Cox, M. Willig, G. G. Mittelbach, C. Osenberg, and M. Kaspari. 2000. Species richness, species-area curves and Simpson's paradox. Evolutionary Ecology Research 2:791-802. Scheiner, S. M., and S. Jones. 2002. Diversity, productivity and scale in Wisconsin vegetation. Evolutionary Ecology Research 4:1097—1117. Scheiner, S. M., and J. M. Reybenayas. 1994. Global patterns of plant diversity. Evolutionary Ecology 8:331-347. 84 Schoener, T. W. 1974. Resource partitioning in ecological communities. Science 185127- 39. Schoener, T. W. 1983. Rate of species turnover decreases from lower to higher organisms - a review of the data. Oikos 41:372-377. Shrrrida, A., and M. V. Wilson. 1985. Biological determinants of species-diversity. Journal of Biogeography 12:1-20. Shurin, J. B., J. E. Havel, M. A. Leibold, and B. Pinel-Alloul. 2000. Local and regional zooplankton species richness: A scale-independent test for saturation. Ecology 81:3062-3073. Smith, M. D., and A. K. Knapp. 2001. Size of the local species pool determines invasibility of a C—4-dominated grassland. Oikos 92:55-61. Srivastava, D. S. 1999. Using local-regional richness plots to test for species saturation: pitfalls and potentials. Journal of Animal Ecology 68: 1-16. Thompson, K., J. G. Hodgson, J. P. Grime, and M. J. W. Burke. 2001. Plant traits and temporal scale: Evidence from a 5-year invasion experiment using native species. Journal of Ecology 89: 1054-1060. Tilman, D. 1982. Resource competition and community structure. Princeton University Press, Princeton, NJ. Tilman, D. 1997. Community invasibility, recruitment limitation, and grassland biodiversity. Ecology 78:81-92. Tilman, D., and S. Pacala. 1993. The maintenance of species richness in plant communities. Pages 13-25 in R. E. Ricklefs and D. Schluter, editors. Species diversity in ecological communities. University of Chicago Press, Chicago, IL. Tumbull, L. A., M. J. Crawley, and M. Rees. 2000. Are plant populations seed-limited? A review of seed sowing experiments. Oikos 88:225-238. Vandermeer, J. H. 1969. The competitive structure of communities: an experimental approach with protozoa. Ecology 50:362-371. Vivian-Smith, G. 1997. Microtopographic heterogeneity and floristic diversity in experimental wetland communities. Journal of Ecology 85:71-82. Voss, E. G. 1996. Michigan flora; a guide to the identification and occurrence of the native and naturalized seed-plants of the state. Cranbrook Institute of Science, Bloomfield Hills, MI. 85 Waide, R. B., M. R. Willig, C. F. Steiner, G. Mittelbach, L. Gough, S. I. Dodson, G. P. Juday, and R. Parmenter. 1999. The relationship between productivity and species richness. Annual Review of Ecology and Systematics 30:257-300. Wardle, D. A. 2001. Experimental demonstration that plant diversity reduces invasibility - evidence of a biological mechanism or a consequence of sampling effect? Oikos 95:161-170. Weiher, E., A. van der Werf, K. Thompson, M. Roderick, E. Gamier, and O. Eriksson. 1999. Challenging Theophrastus: A common core list of plant traits for functional ecology. Journal of Vegetation Science 10:609-620. Weitz, J. S., and D. H. Rothman. 2003. Scale-dependence of resource-biodiversity relationships. Journal of Theoretical Biology 225:205-214. Welden, C. W., and W. L. Slauson. 1986. The intensity of competition versus its importance - an overlooked distinction and some implications. Quarterly Review of Biology 61:23-44. Westoby, M., D. S. Falster, A. T. Moles, P. A. Vesk, and I. J. Wright. 2002. Plant ecological strategies: Some leading dimensions of variation between species. Annual Review of Ecology and Systematics 33: 125-159. Whittaker, R. J ., K. J. Willis, and R. Field. 2001. Scale and species richness: Towards a general, hierarchical theory of species diversity. Journal of Biogeography 28:453- 470. Yeakley, J. A., and J. F. Weishampel. 2000. Multiple source pools and dispersal barriers for Galapagos plant species distribution. Ecology 81:893-898. Zar, H. 1996. Biostatistical analysis. Prentice-Hall, Upper Saddle River, NJ. Zobel, M. 1992. Plant-species coexistence - the role of historical, evolutionary and ecological factors. Oikos 65:314-320. Zobel, M. 1997. The relative role of species pools in determining plant species richness. An alternative explanation of species coexistence? Trends in Ecology & Evolution 12:266-269. Zobel, M., M. Otsus, J. Liira, M. Moora, and T. M018. 2000. Is small-scale species richness limited by seed availability or microsite availability? Ecology 81:3274- 3282. 86 APPENDIX 87 APPENDIX A. Summary of environmental variables for ancillary plots measured at four grassland sites in SW Michigan. Values represent mean (1 standard error) for 4 replicates, n = 48. All N variables measured in ug/g dry soil. ANPP is estimated by aboveground plant biomass. %Light availability is measured in potential photosynthetic flux density as described in Chapter 4. Site Community N-Pool Site Site Site Site Low Med High Low Med High Low Med High Low Med High N03' 029 (014) 007 (001) 070 (059) 042 (009) 026 (008) 025 (017) 011 (006) 023 (006) 010 (0£M) 006 (000) 006 (000) 003 (008) N-Pool NH4+ 517 (112) 556 (104) 794 (153) 370 (020) 757 (052) 1057 (145) 709 (050) 1054 (07mg 972 (059) 318 (054) 554 (072) 552 (058) Total N-Pool 546 (116) 563 (105) 855 (Len 412 (028) 813 (048) 1092 015) 720 (079) 1057 (080 953 (00» 324 (050 590 (072) 592 (060 88 N-Min 054 (000 054 (00m 105 (050) 045 (00m 056 (00m 143 (017) 093 (00m 107 (010 128 012) 051 @16) 058 @14) 154 (04a o/oSOII 95 (12) 131 (02) 185 (16) 75 (04) 120 (05) 187 (16) 155 (02) 135 (05) 144 (11) 127 (05) 169 (07) 208 (17) AN PP/ Moisture mz/yr 236 (14) 401 (12) 530 (41) 203 (39) 406 (25) 554 (85) 336 (33) 468 (35) 555 (1 1) 282 (17) 407 (so) 675 (58) Litter/ m2 69 (1 4) 1 47 (35) 472 (1 00) 36 (8) 280 (34) 528 (1 43) 55 (1 1) 1 41 (22) 336 (25) 32 (2) 1 53 (43) 355 (50) o/onghI availability 71 (5) 37 (4) 22 (3) 64 (4) 31 (4) 12 (4) 44 (7) 22 (3) 8 (2) 57 (3) 31 (5) 11 (3) TE 54 liliiiizijiiiii”)iiiiiiil l 1‘ 3 3 0273 070