UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

On the determination of local species richness and abundance Starzomski, Brian Martin 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-200995.pdf [ 4.35MB ]
Metadata
JSON: 831-1.0092973.json
JSON-LD: 831-1.0092973-ld.json
RDF/XML (Pretty): 831-1.0092973-rdf.xml
RDF/JSON: 831-1.0092973-rdf.json
Turtle: 831-1.0092973-turtle.txt
N-Triples: 831-1.0092973-rdf-ntriples.txt
Original Record: 831-1.0092973-source.json
Full Text
831-1.0092973-fulltext.txt
Citation
831-1.0092973.ris

Full Text

ON THE DETERMINATION OF LOCAL SPECIES RICHNESS AND ABUNDANCE  by BRIAN MARTIN STARZOMSKI B.Sc. Joint Adv. Maj., Saint Francis Xavier University, 1996 M.Sc, Acadia University, 2000  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Zoology)  THE UNIVERSITY OF BRITISH COLUMBIA August 2006 © Brian M. Starzomski, 2006  ABSTRACT A goal of c o m m u n i t y e c o l o g y is to u n d e r s t a n d t h e p r o c e s s e s controlling local s p e c i e s r i c h n e s s a n d a b u n d a n c e . I u s e d t w o e x p e r i m e n t a l s y s t e m s to e x a m i n e this: a m o s s - m i c r o a r t h r o p o d s y s t e m a n d the larval insect c o m m u n i t y of a water-filled b r o m e l i a d t a n k . In c h a p t e r 2, I t e s t t h e e f f e c t s of local d i s t u r b a n c e f r e q u e n c y a n d u n d i s t u r b e d habitat loss o n local s p e c i e s r i c h n e s s of m o s s m i c r o a r t h r o p o d s . Diversity d e c l i n e d linearly with d i s t u r b a n c e rate, a n d t h e d e c l i n e d e p e n d e d o n t h e s u r r o u n d i n g r e g i o n : it w a s s t r o n g e s t in local p a t c h e s d i s c o n n e c t e d f r o m the r e g i o n . F r a g m e n t a t i o n r e s u l t e d in loss of s y s t e m resilience, w i t h a t h r e s h o l d effect in t h e local p a t c h e s of the s m a l l e s t , u n c o n n e c t e d r e g i o n s i z e . All t r e a t m e n t s h a d similar g a m m a diversity: r e d u c e d a l p h a diversity w a s o f f s e t by i n c r e a s e d b e t a diversity. D i s p e r s a l f r o m u n d i s t u r b e d habitat m a i n t a i n s local r i c h n e s s a n d a b u n d a n c e , t h o u g h this c a n be o v e r r i d d e n by d e c r e a s e d r e g i o n s i z e b e l o w a t h r e s h o l d level, as well a s f r a g m e n t a t i o n , a n d e d g e effects. In C h a p t e r 3, I e x a m i n e d t h e i m p a c t of the n u m b e r of s p e c i e s p r e s e n t in r e g i o n s o n t h e c o m m u n i t y in local p a t c h e s . O v e r 16 m o n t h s , local r i c h n e s s w a s i n d e p e n d e n t of r e g i o n a l r i c h n e s s . T h e n u m b e r of local s p e c i e s v a r i e d w i t h t i m e , d u e to c o m p e t i t i o n a n d d i s p e r s a l f r o m t h e r e g i o n , but m o s t s t r o n g l y b e c a u s e of s e a s o n a l i t y . E v e n w h e n local r i c h n e s s did not c h a n g e , c o m m u n i t y c o m p o s i t i o n d i d . T h u s , regional c o m p o s i t i o n w a s m o r e i m p o r t a n t t h a n regional s p e c i e s r i c h n e s s , a n d t h e imprint of b i o g e o g r a p h y (the c o m p o s i t i o n of r e g i o n a l s p e c i e s p o o l s ) , is o n l y s e a s o n a l l y e v i d e n t . C h a p t e r 4 e x a m i n e d positive a n d n e g a t i v e interactions in d e t e r m i n i n g s p e c i e s ' a b u n d a n c e s in b r o m e l i a d p h y t o t e l m a t a . T h e top p r e d a t o r r e d u c e s the a b u n d a n c e of detritivorous insect larvae, but a n i n t e r m e d i a t e p r e d a t o r h a s little effect. W i t h o u t t h e t o p p r e d a t o r , a p r o c e s s i n g c h a i n t a k e s p l a c e b e t w e e n s e v e r a l detritivore s p e c i e s , a n d t h e s e positive interactions i n c r e a s e t h e e m e r g e n c e rate of c h i r o n o m i d s at t h e e n d of t h e c h a i n . T h e m e c h a n i s m of facilitation is likely the p r o c e s s i n g , by tipulids a n d scirtids, of detritus into s m a l l particles w h i c h the detritivore c h i r o n o m i d s c a n f e e d o n . T h i s a l s o i n c r e a s e s t h e s u r f a c e : v o l u m e ratio of detritus, a n d m a y indirectly benefit c h i r o n o m i d s by i n c r e a s i n g their f o o d , b a c t e r i a , a l g a e , a n d f u n g i .  'The aim of science is not to open the door to infinite w i s d o m , but to set a limit to infinite error.' Bertolt Brecht  'Can one comprehend the ruin of natural systems without understanding how they are built?' Robert Ricklefs  T A B L E OF CONTENTS  '  Abstract  ''  Table of Contents  i v  List of Tables  vi  List of Figures  v  ''  Acknowledgements  x  Co-Authorship Statement  x  >  1  General Introduction 1.1 Introduction of concepts 1.1.1 Species Diversity 1.1.2 Local versus regional processes 1.1.3 Disturbance 1.1.4 Habitat loss 1.1.5 Positive versus negative interactions and processing chains 1.2 Study Systems 1.2.1 British Columbia moss and associated microarthropods 1.2.2 Costa Rican bromeliads and associated larval insects 1.3 Thesis objectives and overview 1.4 References  2  Landscape geometry determines community response to disturbance in a moss-micro-arthropod ecosystem 2.1 Introduction 2.2 Materials and Methods 2.2.1 Study site and system 2.2.2 Experimental design 2.2.3 Analyses 2.3 Results 2.4 Discussion 2.4.1 Conclusions 2.5 References •  14 15 18 18 18 20 21 23 26 34  Does regional species richness determine local species richness? An experimental test of saturation theory 3.1 Introduction 3.2 Materials and Methods 3.2.1 Study system 3.2.2 Experimental design 3.2.3 Moss collection 3.2.4 Patch construction... 3.2.5 Sampling 3.2.6 Analyses 3.3 Results 3.4 Discussion 3.5 References  38 39 42 42 42 43 43 44 44 45 46 59  3  1 2 2 2 3 4 5 6 6 6 7 9  •  •  4  Predation overwhelms a processing chain in a bromeliad-insect food web 4.1 Introduction 4.2 Materials and Methods 4.2.1 Study site and system 4.2.2 Experimental design 4.2.3 Experiment 1 4.2.4 Analysis 4.2.5 Experiment 2 4.2.6 Sampling 4.2.7 Analysis 4.3 Results 4.3.1 Experiment 1: Analysis of sediment produced in artificial wells 4.3.2 Experiment 2: Food web manipulation and chironomid emergences 4.4 Discussion 4.5 References  64 65 67 67 68 68 69 69 71 72 72 72 73 73 82  5  General Conclusions 5.1 Overview of thesis 5.2 General implications 5.3 Future directions 5.4 References  86 87 89 91 93  v  LIST OF TABLES Table 2.1.  Trait effects for body size, species identity, trophic level, and abundance, with  28  associated p-values from regressions (continuous variables: Body Size), A N O V A s (discrete  variables:  (Abundance).  Trophic  Position,  Taxonomic  Group)  and  an  ANCOVA  P-values >0.05 indicate there is no difference in mean abundance  for the various traits between disturbance level 0 and 3.  Table 3.1.  A . Initial estimated regional diversities from sorted microarthropods. Each of  ^1  the eight sample sites is referred to by its letter code, A through H. B. Locations of the moss collection sites in southwestern British Columbia.  Table 3.2. Table 4.1.  Summary of A N O V A results for local species richness.  Treatments  Mecistogaster Polypedilum  used in Experiment  modesta,  52  1 (A) and Experiment 2 (B): M M =  Tan = tanypodine, Tip = tipulids, Sc =  77  scirtids, PC =  chironomids, YC = Yellow chironomids.  vi  LIST OF FIGURES Figure 2.1.  29 Experimental design.  region, D = no region.  A = large region, B = Medium region, C = Small  0 = Not disturbed, 1 = Disturbed once/month for 1 month  (last 1 month), 2 = Disturbed once/month for 2 months (last 2 months), 3 = Disturbed once/month for 3 months.  Figure 2.2.  The relationship between disturbance and species richness (a), and  disturbance  and abundance  (b), as a function of region size.  30  Disturbance  measured from 0 (no disturbance) to 3 (highest disturbance), and region size from A (largest region) to D (no region).  For species vs. disturbance: region A - y = -  2.88x + 35.5, R = 0.89; B- y = -2.18x + 28.79, R = 0.90; C- y = -2.84x + 33.04, R 2  2  2  = 0.90; D- y = -6.5x = 41.85, R = 0.94. For individuals vs. disturbance: region A- y 2  = -32.32x + 428.50, R = 0.50; B- y = -34.78x + 333.26, R = 0.70; C- y = -36.66x + 2  2  446.83, R = 0.27; D- y = -139.66x = 727.77, R = 0.93. Error bars of slope values 2  2  are shown in Fig. 2-3.  Figure 2.3.  ; Mean (± SE) of the slopes of disturbance frequency and species richness  per region (a), and disturbance frequency and abundance per region (b).  31  For  region sizes, A (largest region) to D (no region).  Figure '2.4.  Ordination (NMDS ) plot of all treatment communities, showing mean  species richness per treatment.  32  The undisturbed patches are enclosed with a  stippled line, and arrows show increasing disturbance frequency. The letter refers to the region size (A largest to D smallest), and the number to disturbance rate (0no disturbance to 3, 3 times disturbed). 33  Figure 2.5.  G a m m a diversity across all region size treatments.  G a m m a diversity is  equivalent to the number of species at the landscape scale, or the pooled number  vii  of species across all 4 disturbance treatments, within a region. Error bars are ± 1 SE.  Figure 3.1.  Predictions for A) the impact of regional species richness on local richness;  53-54  B) the pattern of local species richness with assembly time, with expected relationships between local and regional richnesses at three points along the community assembly trajectory (adapted from Mouquet et al. 2003).  Figure 3.2.  Locations in Southwestern British Columbia of collection sites for moss  patches used in the experiment.  55  Also shown is a NMDS plot showing the  differences between community structure among the collected moss patches. 56  Figure 3.3.  Experimental design and relationship between number of patches in region  ' treatments and species present: White areas = differing regions collected. Dark areas = defaunated central core. Letters refer to the differing regions, from one region represented (All A's), to eight regions represented (regions A thru H; the locations of which are randomly determined).  There are 3 replicates of each  treatment. The combinations shown are for illustrative purposes only.  Figure 3.4.  Local richness versus time. Filled black circles show all data; open circles  57  are means +/- SE.  Figure 3.5.  Plots of local and regional species richness. A. 16 June 2003. B. 28 July  5 8  •2003. C. 25 August 2003. D. 22 September 2003. E. 12 January 2004. F. 23 August 2004.  Figure 4.1.  Simplified bromeliad tank food web showing common species found at  78  Estacion Biologica Pitilla, Costa Rica.  viii  Figure 4.2.  Predictions for number of emerging adults per treatment.  See Table 4-1 b  for Treatment legend.  Figure  4.3.  Average dry mass ± SE of sediment for the treatments  Chironomids alone  (C), Chironomid-Scirtids (CS), Chironomid-tipulids (CT), No insects (NI), Scirtids alone (S), and Tipulids alone (T). Figure 4.4.  Treatment effects on the mean number of emerging detritivore chironomids  (± SE). See Table 4-1 b for Treatment legend.  ACKNOWLEDGEMENTS Much of the research in this thesis was conducted with funding from the Natural Sciences and Engineering Research Council of Canada, the University of British Columbia, the Santa Fe Institute Complex Systems Summer School, and others. I am grateful for the support. My committee members (Michael Doebeli, Judy Myers, Dolph Schluter a n d Tony Sinclair) gave me a lot to think about before, during, and after committee meetings, comps, copies of paper drafts, and when they cornered me in the halls (incidentally, when cornered, I w a s almost never able to use the 'flight' of the 'fight or flight' response, and I am the better for sticking with them). I would like to also thank Jamie Smith for sharing some hard and timely advice with me over a few glasses of beer. The twinkle in his eye and his pointed questions will be missed. The graduate students and post-docs gathered about Hut B8 and the late Hut B6 were a source of inspiration throughout this project. For some reason, my office kept being located next to that of the Biodiversity post-doc, despite the moves as buildings were torn down and others erected. I can't begin to thank whoever was responsible for this, because having Jessica Hellmann, Ross T h o m p s o n , and then Luke H a r m o n next door made popping in for a pep talk or s o m e erudite science advice ridiculously easy. Every grad student should be so lucky. The Srivastava lab w a s a collection of remarkable people, the kind of group that one looks back on and wonders how so many smart and good folks ended up in one place. N S E R C provided a number of great students to work with, as did the UBC work-study program- Marius Aurelian, Derek Tan, Mike Janssen, and Raenelle Parker were smart and fun. Thanks to my office mate Katsky Venter for being so good natured and so intelligent, and to Ben Gilbert and Jackie 'Big Guy' Ngai for lots of laughs and good science talk. There's really no way for me to truly thank Diane for putting up with me for as long as she did. If there is a more intelligent, kind, and fun person to work with out there, I would be absolutely floored. Diane is remarkable, with an eye for surrounding herself with good people (present company excluded). The following grad students were especially important in shaping the w a y I think (and in many cases, the way I think with drink): Laura Prugh, Paul Smith, Risa Sargent, Aleeza Gerstein, Allyson Longmuir, Jessica Hill, Alistair Blatchford, Leanna W a r m a n , Brad Davis, and especially Tim Vines, Arianne Albert, Kerry Marchinko, Bob Lessard, Nathan Taylor and Mike Melnychuk. A wonderful bunch. Alain Joseph, Greg Sharam, Ryan Norris, Jon Shurin, Ross T h o m p s o n , Colin Bates- could there be a better group of people to have a relaxing sit with? The better the people you surround yourself with, the more fun you have as a grad student. I had a lot of fun. My family w a s incredibly supportive through all of this, and special thanks go to my parents, Larry and Margaret, w h o encouraged my very odd interests as a kid, and then unconditionally supported everything I did once I got to grad school, no matter how inexplicably bizarre. Finally, Meghan Mulcahy and Aimee Pelletier somehow made this several notches more incredible, and I thank them most of all.  x  CO-AUTHORSHIP STATEMENT CHAPTER 2- Landscape geometry determines community response to disturbance in a moss-microarthropod ecosystem: I acknowledge the contributions of Diane Srivastava to this manuscript. I had the original idea for the experiment, which I constructed a n d analysed. Diane assisted in initial identifications of mite samples, as well as with designing the analyses. I wrote the first draft of the manuscript. A version of this chapter has been submitted to the journal 0//cos.  CHAPTER 3- Does regional species richness determine local species richness? An experimental test of saturation theory: I acknowledge the contributions of Diane Srivastava a n d Raenelle Parker to this manuscript. Diane had the original idea for the experiment, which I modified, constructed, and analyzed. Raenelle worked on the original mite library and the collection of the original moss samples. I wrote the first draft of the manuscript. A version of this chapter is to be submitted to the journal  Ecology.  CHAPTER 4- Predation overwhelms a processing chain in a bromeliad-insect food web: I acknowledge the contributions of Diane Srivastava and Daryl Suen to this manuscript. the food w e b experiment, and analysed all of the data for both experiments.  I designed  Daryl contributed data  from an experiment conducted for a directed studies project, which I incorporated into the manuscript after reanalyzing his results. I wrote the first draft of the manuscript. A version of this chapter has been submitted to the journal  Oecologia.  xi  Chapter 1: General Introduction  1.1 INTRODUCTION OF CONCEPTS Species diversity is one of the most endearing characteristics of life on earth. W e are surrounded by "endless forms most beautiful and wonderful" (Darwin 1859), and for centuries scientists have sought to explain what controls the existence of different numbers of species in different locations. This thesis provides a small contribution to the accumulation of knowledge on w h a t factors in the environment determine the species that are present in local communities. Here I take an experimental approach to look at the relationship between local and regional species richness, the effects of disturbance and regional habitat loss on local species richness, and the interacting effects of positive and negative interactions between species on the relative abundances of those species. In this chapter, I briefly review some concepts necessary for understanding the rationale behind the studies, then describe some necessary theory and past research to set the stage for the experiments themselves, the results of which are shown in Chapters 2, 3 and 4. A final concluding chapter reviews the main findings of the thesis, describes how these results fit into the current body of ecological knowledge, and proposes areas of future study.  1.1.1 Species Diversity Species diversity is a term which combines both species richness and abundance to describe the variety of life present in a defined area. It is divided into three components: a, (3, and y diversity (Whittaker 19,70). a diversity refers to the number of species at a local scale (often arbitrarily defined), (3 diversity is the between-community diversity, and y diversity is the regional diversity (Whittaker 1970, Loreau 2000). In this thesis, I use a moss-microarthropod system with a high (3 diversity to create regions of different species diversities (y diversity), to look at the impact on local species diversity (a diversity) of both the changes in y diversity a n d local disturbance. These results are s h o w n in Chapters 2 a n d 3.  1.1.2 Local versus regional processes What is the role of the regional species pool in determining local species diversity? Do larger and more diverse regions lead to increased species diversity at the local scale? Do local communities saturate with species at a relatively low regional diversity, or is there no hard upper limit on the n u m b e r o f species in the local region (other than that determined by space)? Much of the work in this thesis seeks to  2  understand the relative roles of local and regional controls, in conjunction with disturbance frequency, on local species diversity. The pattern between local and regional richness can be of either a saturated or unsaturated form (Ricklefs 1987, Cornell and Lawton 1992). Though saturation of communities w a s first proposed by Elton in 1950 (as quoted in Terborgh and Faaborg 1980), the concept has been in vogue of late. Saturation refers to the total number of species that a community can contain before no more can be accommodated, due to factors such as space (Paine 1984), competition (MacArthur and Levins 1967), disturbance (Caswell and Cohen 1993), and dispersal (MacArthur and Wilson 1967). All of these factors interact and overlap in a variety of ways. Saturated patterns are demonstrated w h e r e local interactions dominate over dispersal in limiting local diversity. Unsaturated patterns may be found where local richness is not constrained by local interactions, but instead depends on the supply of colonists from the region. In the latter case, local and regional richness should be correlated over the entire range (Srivastava 1999). In general, large regions have more species than small regions (Brown 1995, Rosenzweig 1995). This is because large regions also likely have more, and a more diverse, collection of habitat and resources, and thus can support more individuals. If there are more individuals, probability tells us there will likely be more species (Srivastava and Lawton 1998). There are many examples, however, where this relationship does not hold. It is a prediction of island biogeography theory (MacArthur and Wilson 1967) that two small islands, one close and one far from a mainland, will not support the same number of species. The island closest to the mainland will have a higher equilibrium species number, due to the increased immigrant rain. It follows that the size of the region does not control the species present, though processes associated with the size of the region (such as immigration, or probability of disturbance [Wardle 1997]) may. Thus, it may be these covariates (e.g., disturbance in the form of seasonality) that are more important than differences in regional richness in structuring local richness. Chapter 3 will examine the relative contributions of these factors to local species richness.  1.1.3 Disturbance The role of disturbance in determining the structure of specific communities has'been studied for some time (e.g., Watt 1947, Connell 1978, Sousa 1979, Paine and Levin 1981). Disturbances are classified by  3  their frequency, duration, size, and their intensity or severity (Pickett and White 1985). Ecologists in the early part of the Twentieth century considered disturbance to be an unnatural phenomenon not particularly worthy of study, which simply kept communities from reaching their natural climax state (Clements 1936, O d u m 1969). The work of the last thirty years has made this view less and less acceptable. Instead, disturbance has increasingly been used to explain community structure in combination with competition/niche theory (MacArthur and Levins 1964). Disturbances are typically divided into two groups: pulse disturbance and press disturbance (Pickett and White 1985). Pulse disturbances tend to be naturally occurring in a given environment, and the organisms inhabiting an area are typically adapted to such infrequent disturbance (Pickett and White 1985). Examples of pulse disturbances include fire and large-scale insect defoliation (e.g., Holling 1986, Payette 1992), and even annual changes of the seasons. Press disturbances tend to be of human origin, including such things as acid rain, climate change, and human use of a habitat patch. Organisms subjected to these press disturbances are seldom adapted to them (Pickett and White 1985). Disturbances can structure the community in novel ways- by creating new habitat for some species, or through increasing spatial heterogeneity in the landscape (Pickett and White 1985). Spatially heterogeneous landscapes generally provide habitat for more species than spatially homogeneous landscapes (MacArthur 1958, MacArthur and MacArthur 1 9 6 1 , Rosenzweig 1995). After disturbance, species may recolonise formerly occupied patches (Pickett and White 1985). If the patch is within a small, undisturbed region, species may not be present and may not be able to recolonise. Similarly, if the disturbed patch is unconnected to habitat in which the species is found, recolonisation may be impossible (Gilbert et al. 1998). Disturbance thus plays an important role in maintaining species diversity in the landscape, and it is important to understand the mechanisms by which this happens, including the interaction between disturbance and region size. Chapter 2 examines the interactions between press disturbance and habitat loss patterns.  1.1.4 Habitat loss Habitat loss is an ongoing problem worldwide, and has been identified as the primary factor leading to species extirpations and extinctions (Dirzo and Raven 2003). A variety of processes lead to loss of habitat, such as conversion to agricultural or urban landscapes, and much habitat loss follows a  4  predictable pattern of encroachment, increasing fragmentation, loss of connectivity, and finally, complete disappearance (Dirzo and Raven 2003). Little is known, however, about the effects of local patch disturbance that may take place within regions as habitat around that patch is lost. Do the effects of loss of the undisturbed region surrounding a disturbed local patch accrue linearly, or is there a threshold point beyond which the processes that help maintain or rescue populations are lost? Does this threshold occur as a result of fragmentation, or because of loss of connectivity? These are the questions addressed in Chapter 2.  1.1.5 Positive versus negative interactions and processing chains Most studies have sought to explain species coexistence and determination of relative abundances through competition for one or more resources (MacArthur and Levins 1967), apparent competition through shared predators (Holt 1977), trophic interactions like strong predation (Paine 1966, Paine 1969, Shurin and Allen 2001), or spatial and temporal niche partitioning (Hastings 1977, Hastings 1980). All of these processes rely on the existence of negative interactions between species to explain how two or more species influence the abundances of one another (MacArthur 1972, Strong et al. 1984). Positive interactions between species, however, may be more important than is often realized (Bruno et al. 2003, Tirado and Pugnaire 2005). Processing chains, a type of facilitative effect, consist of at least two species, one of which conditions, or prepares, a resource for another species. The conditioning species has been termed the 'upstream' species (to denote its role 'up the stream' of resource processing), and this species often feeds on and prepares a resource for a 'downstream' species (to denote the species' position 'downstream' of the flow of resources- Heard 1994a). Heard (1994b) lists more than 50 examples of putative processing chains, from those involving microorganisms (yeasts, phytoplankton), to invertebrates (insects, mollusks), and vertebrates (gazelles, gulls, skuas). Processing chains have been.demonstrated in a variety of phytotelmata (standing water found in plants), including treeholes (Paradise and Dunson 1997, Paradise 1999), Heliconia  bracts (Seifert and Seifert 1976, 1979), and pitcher plants (Bradshaw  1983, Beaver 1984, Heard1994b). Typically these experiments are conducted with only two species (an upstream and a downstream species). Rarely are processing chains demonstrated in a food web context. In Chapter 4, I contrast the effect of positive horizontal (within trophic-level) and negative vertical  5  (between trophic-level) effects on species abundance in the food w e b of a Costa Rican bromeliad phytotelmata.  1.2 S t u d y s y s t e m s 1.2.1 British Columbia moss and associated microarthropods In various species of moss (including Polytrichum  and Racomitrium  species used in this thesis),  microarthropods, predominantly mites (Acari) and springtails (collembola), are found at high abundance and species richness (Davis 1 9 8 1 , Gilbert et al. 1998, Gonzalez and Chaneton 2002, Hoyle 2004). Patches of moss and associated microarthropods make a useful experimental system in two ways: 1) through being easily shaped into different combinations of region size and connectivity (Chapter 2), and 2) through changing the number of species found in experimental regions (Chapter 3). Treatments may be easily applied by collecting moss from various, widely spaced, locations and then using the differences in community composition between those collected patches to create different treatments in a controlled setting (as shown in Chapter 3). This compresses the natural (3 diversity into a small area, which can be used to create regions of different (y) species richness. Further, the microarthropods inhabiting the moss are easily extracted using a simple Tullgren funnel technique (Knudson 1972, Kethley 1991). The small size of mites and collembola relative to treatment landscapes makes this system a small-scale analogue of macro-landscapes with much larger organisms. A final point is that in coastal British Columbia (where the experiments were conducted), drought subjects the moss system to regular disturbances in its natural setting, and thus my manipulations are typical of the environmental conditions of the area. In fact, because of this wet and dry cycle, it appears that two different communities have developed in the moss-microarthropod system to take advantage of this seasonal cycle (Gonzalez 2000, pers. obs).  1.2.2 Costa Rican bromeliads and associated larval insects I conducted experiments at the Estacion Biologica Pitilla within the A r e a de C o n s e r v a t i o n Guanacaste, Guanacaste province, Costa Rica (10°59' N, 85°26'W). Here I contrasted the effects of predation and facilitation in the form of a processing chain, using insect larvae found in the water-filled tanks of bromeliads (described in Chapter 4). Bromeliads of the genera Vriesea and Guzmania  trap  water between the bases of their leaves, and likely contain the majority of the standing water in the wet  tropical forests of northwestern Costa Rica (~ 7000 l/ha, Srivastava, pers. comm.). This provides an important resource for a large number of species, including aquatic microorganisms like phyto- and zooplankton, insects that obligately oviposit in standing water, and a variety of Arachnids, reptiles, and amphibians (Frank 1983, Richardson 1999). Detritus forms the basal resource for the insect and microbial community (Kitching 1999, Richardson 1999). T h e bromeliads m a y accrue a benefit from many of these organisms, as bromeliads get much of their nutrient supply through specialized cells (trichomes) on the surface of the leaves in the water filled tanks that absorb nutrients produced in the breakdown of detritus by various invertebrates and microorganisms (Benzing 2000). T h e discrete nature of tankforming bromeliads and the aquatic insect larvae communities found within (Laessle 1 9 6 1 , Richardson 1999, Armbruster et al. 2002) makes them an ideal model system for examining the roles of facilitative processes in structuring a natural community. Past research has suggested that facilitation in the form of a processing chain may be occurring in the bromeliads  Vriesea  and  Guzmania  (Suen unpubl. obs.,  Srivastava 2006), and I contrast the effects of the processing chain (positive interactions between species) with the effects of predation (negative interactions between species). I used a subset of the bromeliad food w e b for my experiments, focusing on the following larval insect species: a tipulid which shreds leaf detritus, a scirtid beetle which scrapes detritus, two chironomid species which feed on detritus particles, a tanypodine chironomid predator, and the top predator, a damselfly.  1.3 Thesis objectives and overview This thesis looks at controls of local species diversity- that is, what factors s u m to produce the local species richnesses and abundances we see in nature. These factors can take many forms. Historically, local explanations were dominant in local control. For instance, competition, the niche, and the state of local habitat w e r e used to explain why groups of species were found in local regions (MacArthur 1958, MacArthur and Levins 1964, MacArthur 1972). More recently, the spatial scale of inference has been broadened. Dispersal, regional richness, and even ecological drift (a random march to extinction, sensu Hubbell 2001) have been proposed as explanations for local richness (e.g., Ricklefs and Schluter 1993). These two scales require different sorts of exploration to determine their relative roles in structuring local communities, a fact that has not always been appreciated. Robert MacArthur himself w a s guilty of explaining local richness as a result of local factors (e.g., habitat partitioning in MacArthur's warblers, c  7  MacArthur 1958), and as a result of regional factors (in the case of the Theory of Island Biogeography, using dispersal- MacArthur and Wilson 1967). This has been called 'MacArthur's Paradox' (Schoener 1983). I do not intend to resolve MacArthur's Paradox in this work; rather, my goal is to explore some of the effects of varying different regional and local factors to decrease or increase local species diversity. In any case, the paradox itself does not apply to natural communities. It is clear that a multiplicity of factors contribute to a full explanation of local species richness: MacArthur w a s simply exploring the strengths of various predictors for local species diversity in a variety of situations, as I do here. In this thesis I explore some regional-scale, and s o m e local-scale explanations for local species diversity. To study these various processes, I have conducted experiments in moss-microarthropod systems in British Columbia, and in water-filled tanks of bromeliads in Costa Rica. Specifically, I examine the roles of regional species diversity in determining local species diversity, the interacting effects of regional habitat loss and disturbance for local species diversity, and the role of facilitation in local habitats for determining local species abundances.  8  1.4 REFERENCES Armbruster P., R.A. Hutchinson and P. Cotgreave 2002. Factors influencing community structure in a South American tank bromeliad fauna. Oikos 96: 225-234. Beaver, R.A. 1983. The communities living in Nepenthes  pitcher plants: Fauna and food webs. In: J. H.  Frank and L.H. Lounibos (eds.) Phytotelmata: Terrestrial plants as hosts for aquatic insect communities, pp. 1 2 9 - 1 5 9 . Plexus Publishing, New Jersey. Benzing, D. 2000. Bromeliaceae: Profile of an Adaptive Radiation. Cambridge University Press. Bradshaw, W.E. 1983. Interaction between the mosquito Wyeomyia knabi, and their carnivorous host Sarracenia  purpurea.  smithii, the midge  Metriocnemus  In: J. H. Frank and L.H. Lounibos (eds.)  Phytotelmata: Terrestrial plants as hosts for aquatic insect communities, pp.161-189. Plexus Publishing, New Jersey. Brown, J.H. 1995. Macroecology. University of Chicago Press, Chicago Bruno, J.F., J.J. Stachowicz, and M.D. Bertness. 2003. Incorporating facilitation into ecological theory. Trends in ecology and evolution. 18:119-125 Caswell, H. and J.E. Cohen 1993. Local and regional regulation of species-area relations: a patch occupancy model. In: Ricklefs, R.E. and D. Schluter (eds.). Species diversity in ecological communities: historical a n d geographical perspectives. University of Chicago Press, Chicago. Clements, F.E. 1936. Nature and structure of the climax. Journal of Ecology 24: 252-284. Connell, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302-1310. 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 6 1 : 1-12. Darwin, C. 1859. The origin of species by means of natural selection or the preservation of favoured races in the struggle for life. Reprinted 1963, Washington Square Press, N e w York. Davis, R. C. 1 9 8 1 . Structure and function of two antarctic terrestrial moss communities. Ecological Monographs 51:125-143. Dirzo, R. and P.H. Raven 2003. Global state of biodiversity and loss. Annual review of environment and resources. 2 8 : 137-167. Elton, C.S. 1950. The ecology of animals. Methuen: London.  9  Frank J.H. 1983. Bromeliad phytotelmata and their biota, especially mosquitoes. In: J. H. Frank and L H . Lounibos (eds.) Phytotelmata: Terrestrial plants as hosts for aquatic insect communities, pp 101-128. Plexus Publishing, New Jersey. Gilbert, F.S., A. Gonzalez, and I. Evans-Freke. 1998. Corridors maintain species richness in the fragmented landscapes of a microecosystem. Proceedings of the Royal Society of London Series b 265: 577-582. Gonzalez, A. 2000. Community relaxation in fragmented landscapes: the relation between species, area and age. Ecology Letters 3: 441-448. Gonzalez, A. and E.J. Chaneton. 2002. Heterotroph species extinction, abundance and biomass dynamics in an experimentally fragmented microecosystem. Journal of Animal Ecology 7 1 : 594-602. Hastings, A. 1977. Spatial heterogeneity and the stability of predator-prey systems. Theoretical population biology 12: 37-48. Hastings, A. 1980. Disturbance, coexistence, history, and competition for space. Theoretical population biology 18: 363-373. Heard, S.B. 1994a. Pitcher-plant midges and mosquitoes: a processing chain commensalism. Ecology 75:1647-1660 Heard, S.B. 1994b. Processing chain ecology: resource condition and interspecific interactions. Journal of Animal Ecology 63:451-464 Holling, C.S. 1986. The resilience of terrestrial ecosystems: local surprise and global change. In: Clark, W . C . and M u n n , R.E. (eds.) Sustainable development of the biosphere. Cambridge University Press, New York. Pp. 292-317. Holt, R.D. 1977. Predation, apparent competition, and the structure of prey communities. Theoretical population biology 12:197-229. Hoyle, M. 2004. Causes of the species-area relationship by trophic level in a field-based microecosystem. Proceedings of the Royal Society of London Series B 2 7 1 : 1159-1164. Hubbell, S.P. 2 0 0 1 . The unified neutral theory of biodiversity and biogeography. Princeton University Press, New Jersey.  10  Kethley, J. 1 9 9 1 . A procedure for extraction of microarthropods from bulk soil samples with emphasis on inactive stages. Agriculture, ecosystems, environment 34:193-200. Kitching, R.L. 2001 Food w e b s in phytotelmata: "bottom-up" and "top-down" explanations for community structure. Annual review of entomology 46: 729-760 Knudsen, J.W. 1972. Collecting and preserving plants and animals. Harper and Row, New York. Laessle, A . M . 1961 A micro-limnological study of Jamaican bromeliads. Ecology 42:499-517 Loreau, M. 2000. Are communities saturated? On the relationship between a, (3, and y diversity. Ecology Letters 3: 73-76. MacArthur, R.H. 1958. Population ecology of some warblers of northeastern coniferous forests. Ecology 39: 599-619. MacArthur, R.H. 1972. Geographical ecology: patterns in the distribution of species. Harper and Row, New York. MacArthur, R.H. and J. MacArthur 1 9 6 1 . On bird species diversity. Ecology 42: 594-598. MacArthur, R.H. and R. Levins 1964. Competition, habitat selection, and character displacement in a patchy environment. Proceedings of the national academy of sciences of the United States of America, 516:1207-1210. MacArthur, R.H. and R. Levins. 1967. The limiting similarity, convergence, and divergence of coexisting species. The American Naturalist 1 0 1 : 377-385. MacArthur, R.H. and E.O. Wilson 1967. The theory of island biogeography.  Princeton  University Press, New Jersey. O d u m , E.P. 1969. The strategy of ecosystem development. Science. 164: 262-270. Paine, R.T. 1966. Food web complexity and species diversity. American Naturalist 100: 65-75. Paine, R.T. 1969. A note on trophic complexity and species diversity. A m e r i c a n Naturalist 103: 91-93. Paine, R.T. 1984. Ecological determinism in the competition for space. Ecology 65: 1339-1348. Paine, R. and S. Levin. 1 9 8 1 . Intertidal landscapes: disturbance and the dynamics of pattern. Ecological monographs 5 1 : 145-178. Paradise, C.J. 1999. Interactive effects of resources and a processing chain interaction in treehole habitats. Oikos 85:529-535  11  J  Paradise, C.J. and W.A. Dunson 1997. Insect species interactions and resource effects in treeholes: are helodid beetles bottom-up facilitators of midge populations? Oecologia 109:303-312 Payette, S. 1992. Fire as a controlling process in the North American boreal forest. In: Shugart, H.H., R. Leemans, and G.B. Bonan (eds.) A systems analysis of the global boreal forest. Cambridge University press, New York. Pp. 144-169. Pickett, S.T.A. a n d P.S. White 1985. T h e ecology of natural disturbance and patch dynamics. Academic Press, Orlando, Fl. Richardson, B.A. 1999 T h e bromeliad microcosm and the assessment of faunal diversity in a Neotropical forest. Biotropica 31:321-336 Ricklefs, R. E. 1987. Community diversity: relative roles of local a n d regional processes. Science 235: 167-171. Ricklefs, R. E. a n d D. Schluter (eds.) 1993. Species diversity in ecological communities. University of Chicago Press, Chicago. Rosenzweig, M.L. 1995. Species diversity in space and time. Cambridge University Press, Cambridge. Seifert, R.P. a n d F.H. Seifert 1976. A community matrix analysis of Heliconia insect communities. T h e American Naturalist 110: 461-483. Seifert, R.P. a n d F.H. Seifert 1979. A Heliconia insect community in a Venezuelan cloud forest. Ecology 60: 462-467. Shurin, J.B and E.G. Allen. 2 0 0 1 . Effects of competition, predation, and dispersal on local and regional species richness. The American Naturalist 158: 624-637. Sousa, W.P. 1979. Disturbance in marine intertidal boulder fields: the nonequilibrium maintenance of species diversity. Ecology 60: 1225-1239. Srivastava, D.S. 1999. Using local-regional richness plots to test for species saturation: pitfalls and potentials. Journal of Animal Ecology 68: 1-17. Srivastava, D.S. 2006. Habitat structure, trophic structure and ecosystem function: interactive effects in a bromeliad-insect community. Oecologia. Srivastava, D.S. and J.H. Lawton 1998. W h y more productive sites have more species: an experimental test of theory using tree-hole communities. The American Naturalist 152:510-529.  12  Strong, D.R. 1983. Natural variability and the manifold mechanisms of ecological communities. The American Naturalist 122: 636-660. Strong, D. R., D. Simberloff, L G. Abele, and A. B. Thistle (eds). 1984. Ecological communities: conceptual issues and the evidence. Princeton University Press, Princeton, N J . Terborgh, J.W. and J. Faaborg 1980. Saturation of Bird Communities in the West Indies. The American Naturalist 116:178-195. Tirado, R. and F.I Pugnaire 2005. Community structure and positive interactions in constraining environments. Oikos 1 1 1 : 437-444. Wardle, D.A. et al. 1997. The influence of island area on ecosystem properties. Science 277: 12961299. Watt, A . S . 1947. Pattern and process in the plant community. Journal of Ecology 35: 1-22. Whittaker, R.H. 1970. Communities and ecosystems. Macmillan, London.  13  C H A P T E R 2- L a n d s c a p e g e o m e t r y d e t e r m i n e s c o m m u n i t y r e s p o n s e t o disturbance in a m o s s - m i c r o a r t h r o p o d e c o s y s t e m  CHAPTER 2- Landscape geometry determines community response to disturbance in a moss-microarthropod ecosystem: I acknowledge the contributions of Diane Srivastava to this manuscript. I had the original idea for the experiment, which I constructed and analysed. Diane assisted in initial identifications of mite samples, as well as with designing the analyses. I wrote the first draft of the manuscript. A version of this chapter has been submitted to the journal Oikos.  14  2.1 INTRODUCTION A major goal of community ecology is to understand h o w species diversity is maintained (MacArthur 1972, Ricklefs and Schluter 1993, Simberloff 2004). T w o mechanisms thought to be important to the maintenance of species diversity are abiotic disturbance (Connell 1978, Pickett and White 1985, Mackey and Currie 2001), and connection to a larger pool of potential colonists (i.e. the metacommunity; Ricklefs and Schluter 1993, Hubbell 2 0 0 1 , Leibold et al. 2004). Theoretical studies have s h o w n that species diversity within metacommunities can be positively or negatively affected by disturbance. For example, the coexistence of competitors may be enhanced through disturbance via competition-colonization trade-offs (Hastings 1980, Nee and May 1992, Caswell and Cohen 1993, Dytham 1994, Tilman 1994, Kareiva and Wennergen 1995). Moilanen and Hanski (1995) showed that increased landscape connectivity can enhance the ability of a superior competitor to decrease diversity through its strong competitive effect. In an experimental study, Forbes a n d Chase (2002) demonstrated that increased connectivity between local patches decreased regional species richness, and increased similarity a m o n g local communities. At least one microcosm study has explored the interaction between pairwise combinations of metacommunity size, connectivity, and disturbance, though not all three together. Warren (1996) found that protist diversity in beakers w a s reduced by disturbance at low but not high dispersal levels. Landscape geometry, or the combination of scale and connectivity of the landscape, thus contributes to patterns of local species richness. This result is especially important given continuing human modification of the landscape, with much habitat currently being degraded and lost. Habitat loss has been pinpointed as the primary cause of species extirpations and extinctions worldwide (Dirzo and Raven 2003). M u c h habitat loss follows a predictable pattern of encroachment, increasing fragmentation, loss of connectivity, and finally, complete disappearance (Dirzo and Raven 2003). Here w e use an empirical approach to test the effects of declining region size (undisturbed habitat loss), connectivity, and disturbance rate on local species diversity. Different combinations of landscape geometry and disturbance lead to different predictions for local (a) and regional (y) species richness, and the turnover in species between areas (3 diversity) (Rosenzweig 1995). Increased undisturbed habitat surrounding a disturbed local patch (the location of  15  much of the metacommunity) is predicted to lead to faster recovery of local patch (a) diversity after disturbance due to the larger pool of potential species to recolonize the patch (Caswell and Cohen 1993). Under the same scenario, effects on p diversity may be opposing. Faster recovery from disturbance in larger regions may minimize spatial turnover of populations. Alternatively, a larger regional species pool may cause increased differences between local communities due to priority effects (Chase 2003, Fukami 2004). G a m m a diversity is most likely to increase with region size, due to species-area effects and increased habitat heterogeneity (Rosenzweig 1995). Connectivity is also predicted to affect community response to disturbance. Loss of connectivity can potentially decrease recolonization rate after disturbance, producing lower a richness. Loss of connectivity is predicted to increase p diversity, as dispersal between patches is effective in homogenizing composition. However, a richness may also be affected by which species fail to recolonize: loss of a keystone predator may further decrease a diversity (Shurin 2001), whereas loss of a dominant competitor may allow for compensatory increase in a diversity (MacArthur 1972). G a m m a diversity may be lower in unconnected than connected regions, as species go extinct first locally and then regionally due to diminished rescue effects between patches. Alternatively, y diversity may be higher in unconnected than connected regions if founder effects in the multiple isolated patches produce very different communities. Region size and connectivity often covary: decreased region size leads to increased fragmentation, a c o m m o n pattern in nature (Fahrig 2003). A threshold effect may be present as the effects of decreased region size grade into the effects of increased fragmentation, and ultimately loss of connectivity (theoretical studies show loss of connectivity to occur on landscapes w h e n greater than 4 0 % of the original region has been destroyed- With et al. 1997). This may lead to a stronger inverse relationship between disturbance rate and local species diversity than would be expected through the effects of decreased region size alone. In addition to effects on the entire community, individual species with different traits are predicted to respond differently to disturbance. These traits include trophic position, features associated with certain taxonomic groups (such as fragile exoskeletons that increase susceptibility to desiccation), rarity, and body size. A variety of predictions may be drawn from the literature concerning the effects of these  16  traits on susceptibility to extinction (for reviews, see Lawton 1994, Gaston and Blackburn 1995, Rosenzweig 1995). The predictions vary widely, and w e simply explore the utility of these traits in explaining a species' response to disturbance in our system. Because of the differing impacts predicted for different species, and given that w e expect reduced region size and loss of connectivity to decrease recolonization from the surrounding region following disturbance, w e predict that such changes in landscape geometry will result in decreased resilience (sensu Holling 1973) for the community. Resilience refers to the amount of impact a system can absorb before that system is pushed to an alternate domain or state (Holling 1973, Peterson et al. 1998). Here w e define a highly resilient community as one which returns to pre-disturbance levels of species richness, abundance, and composition, rather than being pushed to another state where dispersal cannot repair the effects of disturbance. Based on this background, w e make the following predictions:. 1. Alpha diversity should be highest in local patches within large, connected regions subject to low levels of disturbance. 2. Beta diversity should be highest in unconnected regions. 3. G a m m a diversity should be highest in large regions. 4. A threshold effect resulting in faster decline of species richness with disturbance should occur w h e n region size decreases so much that patches become disconnected. W e report the findings of an experiment conducted using the microarthropod community of a moss-based ecosystem. In this system, microarthropods, predominantly mites (Acari) and collembola, are found at high abundance and species richness (Davis 1 9 8 1 , Gilbert et al. 1998, Gonzalez and Chaneton 2002, Hoyle 2004). Patches of moss and associated microarthropods make a useful experimental landscape, as treatments are easily applied by removing moss from underlying rock, and microarthropods are easily extracted from moss. Further, drought subjects the moss system to regular disturbances in its natural setting, and thus our manipulations are typical of the environmental conditions of the area. The small size of mites and collembola relative to treatment landscapes makes this system a small-scale analogue of macro-landscapes with m u c h larger organisms. Based on the moss-microarthropod system and these observations, our question is:  17  W h a t are the effects of losing undisturbed habitat on species diversity in disturbed local patches?  2.2 MATERIALS AND METHODS 2.2.1 Study site and system The experiment w a s conducted using the community of microarthropods inhabiting a moss (Polytrichum and Bryum spp.) covered granitic outcrop in the University of British Columbia's Malcolm Knapp Research Forest in Haney, British Columbia, Canada (49.216 N, 122.515 W ) . The Research Forest is situated at an elevation of 600 m, with forest cover of Western Hemlock (Tsuqa heterophvlla (Raf.) Sarg.), Douglas-fir (Pseudotsuga menziesii (Mirbel) Franco), and Western Redcedar (Thuja plicata Donn ex D. Don). The microarthropods (predominantly mites [Acari] and collembola) inhabiting the moss mats are found in high abundances, and 200 or more morphospecies (hereafter referred to as species) can be found in small areas (less than 20 m , pers. obs.). This experimental model system (Ims and Stenseth 2  1989, Srivastava et al. 2004) has proven to be useful in studying community response to various processes including the effects of corridors on species richness (Gilbert et al. 1998), a n d the trajectory of community relaxation through time (Gonzalez 2000, Hoyle 2004). In an area measuring 4 m by 6 m, experimental moss patches of various shapes and sizes were created by scraping moss and soil from the rock surface until only bare rock w a s showing between t h e m . In similar moss systems such areas of bare rock constitute effective barriers to movement by soil-dwelling microarthropods (Gilbert et al. 1998).  2.2.2 Experimental design To simulate different region sizes, 20 circular moss regions were created by scraping continuous moss from rock (Fig. 2.1). Four different types of region (n = 5 replicates per treatment) were created: a largeregion treatment (treatment A: 40 cm diameter), a medium-region treatment (B: 28 cm diameter), a smallregion treatment (C: with the four local regions, each 10 cm in diameter, with a perimeter of 1 cm-thick moss, and a 2 cm wide moss corridor between local regions), and a 'no' region treatment (treatment D), with no connections between local patches. From the original continuous m o s s carpet, replicated treatments were positioned randomly on the rock surface, with at least 10 cm of bare rock separating  18  each region replicate. A plastic template of each region size w a s used to ensure the same patch size was created for each replicate. Each region contained within it four 10 cm diameter circular 'plugs' of moss (the local patches for sampling) that were subjected to variable frequencies of disturbance over the course of 4 months. These plugs constitute a habitat patch of moss e m b e d d e d within the larger region of moss, and are thus an open community. The total undisturbed area for each of the regions w a s as follows: A- 942.5 c m , B- 301.6 c m , C- 154.2 c m , D- 0 c m . The response variables w e r e the number of 2  2  2  2  individuals, species richness, and community composition of microarthropods in the moss plugs. All replicates were left undisturbed for one month prior to the beginning of the disturbance treatments, to allow the communities within the moss to respond to the creation of the region-size treatments. To simulate different disturbance regimes, each of the 4 local patches (moss plugs) within the region was subjected to different rates of disturbance over 4 months, from D e c e m b e r 2002 to April 2003. Three disturbance levels were chosen, along with one control (termed disturbance level 0). The moss plug under high disturbance (disturbance level 3) had all microarthropods removed by Tullgren funnel extraction 3 times over 3 months (January-March). In the medium disturbance level (disturbance level 2), the moss plug was disturbed once a month for the final two months (February-March), and the low disturbance (disturbance level 1) treatment was disturbed once (in March). For the undisturbed treatment (disturbance level 0), the plug w a s extracted only at the end of the experiment (April). Every month each 10 cm moss plug w a s removed from the field, placed in a resealable plastic bag, and transported 60 km to a lab at the University of British Columbia. Each moss plug w a s then placed in a Tullgren funnel for 36 hours. Tullgren funnels use a humidity gradient caused by a combination of light and heat (Knudsen 1972) to cause the microarthropods found in the moss to drop into a vial of 7 0 % ethanol, 2 0 % glycerol, and 1 0 % water. If a disturbance w a s being conducted on an individual plug, the 4 0 W light bulbs of the Tullgren funnels were turned on for 36 hours to remove all microarthropods. For the undisturbed treatments, each plug w a s placed in the funnel, but the light w a s not turned on. At the end of four months (corresponding to a generation or more for many of the microarthropod taxa- Walter and Proctor 1999), all moss plugs were removed from the field and brought to the lab where microarthropods were removed by Tullgren funnel extraction. While the extraction efficiency of Tullgren funnels is very high, it is likely this efficiency w a s less than 100%, but consistent for all plugs. All plugs were extracted at the end of the  19  experiment, and microarthropods collected. All microarthropods w e r e stored in vials a t 4 ° C until manually sorted under a 60X dissecting scope. Microarthropods with distinct morphological characteristics were described as individual morphospecies, using various keys (Krantz 1978; unpublished keys provided by the Ohio State University Acarology Summer Program), and identification by experts (D. Walter, H. Proctor, H. Klompen, V. Behan-Pelletier, J. Addison). Descriptions and a key for all the morphospecies found in this study are available on the World Wide W e b at http://www.zooloqv.ubc.ca/~srivast/mites, and a list of the morphospecies is found in Appendix S1 in the Supplementary Material. All microarthropods were measured using a L e i c a ™ MZ16 A stereomicroscope and AutoMontage 3D reconstruction software (Syncroscopy Corporation).  2.2.3 Analyses A priori planned comparisons of richness and abundance in regions A, B, and C (together) vs D (to look for threshold effects that occur w h e n decreased region size causes fragmentation and loss of connectivity), A and B vs C (to look for effects of increased fragmentation), and A vs B (to look for the effect of region size) were performed using t-tests. A caveat must be mentioned here about covariance between region size and connectivity. W h e n the surrounding region b e c o m e s small enough to cease to exist, local patches necessarily become disconnected. C o m p a r e d to region C, region D has both decreased region size and decreased connectivity. Similar concerns may be raised about the change in edge:area ratio, or area of disturbed moss in this system. W e make the point that all these factors covary with loss of regional area in nature; w e therefore conservatively interpret our response as passing through a threshold caused by both loss of connectivity and region area, rather than being caused solely by one or the other. To look for trait predictors for differences in response to disturbance, microarthropods were placed in taxonomic groups (collembola, adult and juvenile oribatids, prostigmatids, and mesostigmatids). Morphospecies were assigned to a trophic level based on taxonomy (detritivorous/fungivorous oribatids and collembola were assigned to trophic level 1, small prostigmatid nematode feeders to trophic level 2, arthropod-feeding prostigmatids and mesostigmatids to trophic level 3, and pseudoscorpions, staphylinid beetles, and spiders were assigned to trophic level 4- Krantz 1978, Walter and Proctor 1999). For each  20  region treatment, effect sizes for each species' response to disturbance were calculated by taking the natural log of the quotient of the number of individuals present in the 3 times-disturbed plugs (pooled over all replicates) divided by those present in the undisturbed plugs. Analysis of the effect of body size on response to disturbance used a body size index defined as a species' body length multiplied by body width in millimetres (Davies et al. 2000), divided by 1000. We cannot assess the effects of abundance on disturbance response in the same way as the other traits because less abundant species (where less abundant species are defined as those with fewer than 5 individuals) are less likely than common species to be sampled in a post-disturbance plug, simply due to their lower numbers. Such sampling effects, however, should not affect the proportional decline in abundance as disturbance increases. We therefore compared the slopes of the regressions of mean proportion of individuals remaining between the undisturbed and disturbance levels 1, 2, and 3 separately for regions A through D. Statistical analyses were conducted using NMDS and ANOSIM routines in Primer 5 for Windows (Plymouth Marine Laboratories), t-tests, ANOVA, ANCOVA and linear regression in R version 2.0.1 (R Development Core Team 2004) and PopTools within Microsoft Excel (CSIRO and Microsoft corporation). For NMDS plots, square-root transformed abundance data were first used to construct a Bray-Curtis similarity matrix of region and disturbance levels then graphed using Hierarchical Agglomerative Clustering.  2.3  RESULTS  A total of 26 274 individuals and 163 species were counted in this study; an appendix containing species data can be accessed at the online site. We calculated the mean a richness in the highest disturbance treatment (disturbance level 3), for each region. Planned comparisons of all connected (mean richness across all disturbance level 3 plugs for regions A, B, and C) and unconnected (region D) regions showed that connected regions had a higher number of individuals (t-test with unequal variances, p = 0.043, DF = 77 ) and species (t-test with unequal variances, p = 0.024, DF = 77 ). There was no significant effect of reduced region size when corridors were present (mean richness across all disturbance level 3 plugs for regions A and B vs C) and region  21  size (A vs B) on species richness and number of individuals (t-tests, all p-values »  0.05, DF = 57).  Increased disturbance caused a statistically significant decrease in the number of species (Fig. 2.2a), though not in the number of individuals (Fig. 2.2b), across all region size treatments. There w a s a significant decrease in both richness and abundance in region D. This is particularly evident in the decreased slopes in treatment D (Figs. 3a, b). Species' abundances increased with fragmentation of the m o s s habitat. In the undisturbed plug of the disconnected treatment, D, total abundance (mean = 620, SD = 194, range = 427 to 863) increased above that seen in the undisturbed local patches of the larger and connected region sizes (A, B, C; mean = 372, SD = 140, range 184 to 639). With disturbance, however, abundances within the disconnected treatment were reduced below the levels of the others (Fig. 2.2b). Species richness and abundance in the smallest region samples (D- the unconnected region) decreased at a m u c h greater rate than those in the connected regions; for species richness the decrease was 2.5 times faster, and for number of individuals, 4 times faster (Figs. 3a, b). To avoid the potential of spatial covariance in the design (disturbance treatments are nested within the region treatments), the slopes of all species richness versus disturbance regressions were compared in an A N O V A , using one slope parameter per regional replicate (n = 5). The richness-disturbance slope of region D w a s significantly steeper than that of the other regions (region type: F  3 1 5  = 3 . 7 1 , p = 0.03, only slope for region D significantly different from the  regions A, B, and C: Fisher's LSD). Mean community composition (species richness and abundance) changed in response to disturbance, especially in the smallest and most fragmented regions (Fig. 2.4). Undisturbed patches had similar composition regardless of region type. Increasing disturbance frequency lead to greater change in community composition (2-way A N O S I M of Disturbance Frequency a n d Region Size: global R = 0.493, p = 0.002). This can be seen in the paths followed through N M D S space. N M D S plots show the relative community similarity between plots using Bray-Curtis similarities; the plot itself is dimensionless. For example, treatment AO vs A 3 has a relative similarity of 64.33%, and DO vs D3, 53.39%. Region A communities changed the least in response to disturbance frequency (AO vs A 3 ) , and region D communities changed dramatically, following much longer paths to their end states (DO vs D3; Fig. 2.4).  22  W e used a y diversity measure similar to that used in Type III species-area curves (i.e., we sampled constant areas within regions- the 4 local patches- and pooled the total of all species in all disturbance treatments within a region; Rosenzweig 1995). between region types (ANOVA: F  3 1 6  There w a s no difference in y diversity  = 0.585, p = 0.63; Fig. 2.4).  Various metrics of species identity showed few differences in response to fragmentation and disturbance (Table 2.1). Oribatids in the connected regions- A, B, and C ( A N O V A and LSD test) showed a w e a k e r decline following disturbance compared to other taxa. In region D, there w a s no difference in numerical response to disturbance amongst all taxa. In fact, for all taxa (save the Oribatid differences just mentioned), none of abundance, body size, or trophic position affected species response to disturbance (Table 2.1).  2.4 DISCUSSION In any landscape, a suite of factors, both biotic and abiotic, work together to structure the community of species found in a local patch (MacArthur 1972, Ricklefs and Schluter 1993, Fahrig 2003). W e consider the roles of region size (undisturbed habitat), connectivity, and disturbance in determining the local community composition of moss microarthropods, embedded within a larger region of similar, undisturbed habitat. Our study presents three main conclusions. First, disturbance caused local (a) species richness to significantly decline within all regions, but especially in the unconnected (region D) treatment. Only region D had a significant decline in number of individuals. A threshold effect is present here, as the decreased region size and increased fragmentation caused a much faster decline in species richness and abundance with disturbance. Surprisingly, fragmenting m o s s habitat initially caused a strong increase in species richness and total abundance, likely due to edge effects. G a m m a (y) diversity (the species richness across all disturbance levels within a region) showed no difference between the four region size treatments. Second, community composition changed with increased levels of disturbance, but especially in patches within small or unconnected regions. Third, for only one group (oribatids) did body size, taxonomic group, or trophic level of species predict a species' response (in abundance) to decreased region size and fragmentation. W e examine each of these conclusions in turn.  23  We observed a threshold effect between the large, connected treatments (A,B, and C), and the small, unconnected treatment D. Fragmentation interacted with disturbance to determine local community composition. Fragmented local regions initially increased in species richness. This is likely due to a combination of edaphic edge effects (Redding et al., 2003, Ries et al. 2004, Harper et al. 2005) and the inability of species to,disperse across barriers (Andren 1994, Gilbert et al. 1998, Gonzalez et al. 1998). The edges of the fragmented and unconnected local patches showed signs of drying, and this may have activated aestivating microarthropod stages, or caused the hatching of eggs of species whose reproductive season is triggered by a period of dry heat. A fence effect (Krebs et al. 1969, MacArthur 1972) may have been caused by the bare-rock dispersal barrier, confining to the patch species that would normally have dispersed. The greater decrease in species richness (with disturbance) in unconnected regions can likely be explained by the barrier to dispersal between isolated patches (even 1cm of bare rock serves as a dispersal barrier to many species- Gilbert et al. 1998), resulting in a decrease in the rescue effect (Brown and Kodric-Brown 1977), as well as decreased immigration of new species. We see this in the unconnected region D, where species richness declines at a much faster rate after disturbance, compared to the connected regions. In unconnected regions, many species do not recolonize local patches following disturbance as they do in the connected regions. In connected regions, there is still a negative effect of disturbance frequency on species richness, but not abundance, though this effect is not as strong as that in the unconnected regions. Some studies have shown that corridors between fragmented (though otherwise undisturbed) patches could facilitate movement for birds (Schmiegelow et al. 1997), maintain ecosystem processes such as seed dispersal by Eastern Bluebirds (Levey et al. 2005), and enhance species richness in small patches (e.g., as shown in a moss-microarthropod study by Gonzalez et al. 1998). Another study using the moss-microarthropod system, however, found no evidence for a difference in species richness between different connectivity treatments, and suggested that corridors connecting patches may be more useful during extreme conditions (Hoyle and Gilbert 2004). Our results confirm the conjecture of Hoyle and Gilbert (2004): fragmentation has the greatest effect in extreme drought conditions. It appears that on the short-term, at least, fragmented populations can persist if connected to source pools of potential colonizers. This has important implications for conservation: while corridors may  24  not seem a good conservation bargain under ideal conditions, they may become very important to maintaining species on a landscape in the face of extreme or novel conditions such as climate change. Size and connectivity of regions are often thought to be important in controlling species richness (MacArthur 1972, Caswell and Cohen 1993). Larger regions, and those regions that have connectivity, are predicted to have more individuals available to recolonize a region after a disturbance. The three largest and connected regions in our study showed no difference in a diversity at the same disturbance levels. The unconnected treatment, however, showed a difference in a richness in disturbed and undisturbed patches within a region. This indicates that we have passed a threshold from where region size and accompanying connectivity between local patches maintains a rescue effect, to where loss of both the surrounding region and connectivity decreases or eliminates rescue effect post-disturbance. This result suggests that connectivity between patches may be more important than region size per se in maintaining local species diversity. Further work is necessary to determine exactly where this threshold from region size to loss of connectivity exists: is it the loss of region size or connectivity that most strongly determines local species richness? We view the role of fragmentation, and especially loss of connectivity, in a landscape as a loss of resilience sensu Holling (1973), who described systems as having low resilience if they had low capacity to rebound from a stressor. In fact, the slope of the lines in Figs. 2 a and b represents a measure of the resilience of this system, with increased negative slope (Figs. 3a, b) equating with low resilience. In this study, therefore, fragmented regions have low resilience. The initial increased species richness and abundance when fragmented, due to edge effects, is quickly erased by disturbance (Fig. 2.2). The continued loss of the rescue effect decreases the ability of species to recolonize in the low-resilience fragmented regions, where recolonisation cannot compensate for disturbance rate. In addition to the decrease in species richness and species abundances, community composition changes with disturbance. Of the 4 region size treatments, regions B (medium) through D (unconnected) change substantially from their tight cluster of undisturbed states, with the community composition of these three treatments diverging with increased disturbance level. With smaller region sizes to supply recolonists post-disturbance, and decreased resilience in the fragmented regions, dispersal cannot compensate for the frequency of disturbance.  25  While species composition of individual local regions changed with disturbance, total (y) diversity within region treatments did not. Unconnected local patches had decreased a, and increased 3 diversity, resulting in unchanged y diversity. This large 3 diversity in unconnected region D suggests that founder effects play a large role in community structure, as a result of the limited dispersal due to the loss of connectivity. Following Davies et al. (2000), we equated larger declines in abundance with increased extinction risk. The finding that neither body size nor trophic position predicted species response to fragmentation is in contrast to that of Davies et al. (2000) and Didham et al. (1998). All taxonomic groups save oribatids decreased in abundance similarly; oribatids increased, or decreased at a lower rate relative to other taxa. This is likely due to the ability of hard-exoskeleton oribatids to withstand some of the disturbances in situ. Species lost a similar proportion of their individuals between the undisturbed and greatest disturbance treatments, regardless of their abundance. We might expect that less abundant species would be extirpated first due to stochastic effects associated with their low numbers, but our contrary result may indicate a competition/colonization trade-off: poor competitors in situ (the less abundant species), may in fact be able to recolonize disturbed areas at a faster rate than common species. The short return time of disturbance may not have allowed the community to reach a competitive equilibrium, and thus the low abundance species persisted or recolonized post-disturbance relatively better than the common species. Didham et al. (1998) and Gonzalez and Chaneton (2002) found that extinction was biased toward rare species following fragmentation, though their work did not explicitly disturb local patches postfragmentation. It appears the species are responding to region size and disturbance frequency treatments in our study in a broadly similar manner.  2.4.1 Conclusions We experimentally examined the interactive effects of region size, fragmentation, and disturbance on local species richness. Disturbance decreased the number of individuals across all region treatments, and the number of species in fragmented, and especially disconnected, regions. We view this effect of decreased region size and fragmentation as a loss of resilience in the system. While our results do not agree with all studies on the effects of connectivity (Hoyle 2005), they do agree with those studies that  26  include extreme conditions (Gonzalez et al. 1998): connectivity provides increased resilience in the face of disturbances. W e consider the maintenance of resilience by means of local habitat connectivity to the regional pool of species to be an important conservation investment in managed landscapes.  27  Table 2.1. Trait effects for body size, species identity, trophic level, and abundance, with associated pvalues from regressions (continuous variables: Body Size), ANOVAs (discrete variables: Trophic Position, Taxonomic Group) and an ANCOVA (Abundance). P-values >0.05 indicate there is no difference in mean abundance for the various traits between disturbance level 0 and 3.  Region A  B  C  D  Trophic position  0.21  0.75  0.21  0.06  Body Size  0.71  0.63  0.31  0.48  Taxonomic Group  0.042  Abundance  0.001  TRAIT  1 2  2  1  0.005  0.001  1  2  0.004  0.15  1  0.72  0.28  Oribatids are less likely to decrease in abundance (LSD test) Oribatids have higher extinction risk when found at low abundances (<5 individuals) (p = 0.001)  28  C  D 10 cm  F i g u r e 2.1. Experimental design. A = large region, B = Medium region, C = Small region, D region. 0 = Not disturbed, 1 = Disturbed once/month for 1 month (last 1 month), 2 = Disturbed once/month for 2 months (last 2 months), 3 = Disturbed once/month for 3 months.  0  -I  ,  ,  0  1  , 2  ,  —,  3  Disturbance Level  b. Figure 2.2.  The relationship between disturbance and species richness (a), and disturbance and  abundance (b), as a function of region size. Disturbance measured from 0 (no disturbance) to 3 (highest disturbance), and region size from A (largest region) to D (no region). For species vs. disturbance: region A - y = -2.88x + 35.5, R = 0.89; B-y =-2.18x + 28.79, R = 0.90; C-y =-2.84x + 33.04, R = 0.90; D-y - 2  2  2  6.5x = 41.85, R = 0.94. For individuals vs. disturbance: region A- y = -32.32x + 428.50, R = 0.50; B- y = 2  2  -34.78x + 333.26, R = 0.70; C- y = -36.66x + 446.83, R = 0.27; D- y = -139.66x = 727.77, R = 0.93. 2  2  2  Error bars of slope values are shown in Fig. 2.3.  30  Region Size  Region Size  CJ  >  CJ -I  oac> nre  -40  3 ^ +J  -SO  (fl h  a. SS  re 3  •g  -120  '> T3 C  -140  O  -160  a O  -180  b. Figure  2.3.  Mean (± SE) of the slopes of disturbance frequency and species richness per region (a),  and disturbance frequency and abundance per region (b). For region sizes, A (largest region) to D (no region).  31  Stress: 0.15  Figure 2.4.  Ordination (NMDS ) plot of all treatment communities, showing mean species richness per  treatment. The undisturbed patches are enclosed with a stippled line, and arrows show increasing disturbance frequency. The letter refers to the region size (A largest to D smallest), and the number to disturbance rate (0- no disturbance to 3, 3 times disturbed).  32  oj 70 re o w 0) 60 Q.  re o  %  re  re .2  50 40  30 Q. (A o 20 i— a)  X! E 10 3 B  C Region  F i g u r e 2.5. G a m m a diversity across all region size treatments. G a m m a diversity is equivalent to the number of species at the landscape scale, or the pooled number of species across all 4 disturbance treatments, within a region. Error bars are ± 1 SE.  •33  2.5 REFERENCES Andren, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: a review. Oikos 71: 355-366. Brown, J.H. and Kodric-Brown, A. 1977. Turnover rates in insular biogeography: effect of immigration on extinction. Ecology 58: 445-449. Caswell, H. and Cohen, J.E. 1993. Local and regional regulation of species-area relations: a patch occupancy model. In: Ricklefs, R.E. and Schluter, D. (eds.), Species diversity in ecological communities: historical and geographical perspectives. University of Chicago Press, Chicago, pp.99-107. Chase, J.M. 2003. Community assembly: When does history matter? - Oecologia 136: 489-498. Connell, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199: 1302-1310. Davies, K.F., Margules, C.R., and Lawrence, J.F. 2000. Which traits of species predict population declines in experimental forest fragments? Ecology 81: 1450-1461. Davis, R.C. 1981. Structure and function of two Antarctic terrestrial moss communities. Ecological Monographs 51: 125-143. Didham, R.K. et al. 1998. Beetle species responses to tropical forest fragmentation. Ecological Monographs 68: 295-323. Dytham, C. 1994. Habitat destruction and competitive coexistence: A cellular model. Journal of Animal Ecology 63: 490-491. Fahrig, L. 2003. Effects of habitat fragmentation on biodiversity. Annual Reviews of Ecology and Systematics 34: 487-515. Forbes, A.E. and Chase, J.M. 2002. The role of habitat connectivity and landscape geometry in experimental zooplankton metacommunities. Oikos 96: 433-440. Fukami, T. 2004. Community assembly along a species pool gradient: implications for multiple-scale patterns of species diversity. Population Ecology 46: 137-147. Gaston, K.J., and Blackburn, T.M.. 1995. Birds, body size and the threat of extinction. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 347: 205-212.  34  Gilbert, F.S., Gonzalez, A., and Evans-Freke, I. 1998. Corridors maintain species richness in the fragmented landscapes of a microecosystem. Proceedings of the Royal Society of London Series B 265: 577-582. Gonzalez, A. 2000. Community relaxation in fragmented landscapes: the relation between species, area and age. Ecology Letters 3: 441-448. Gonzalez, A. and Chaneton, E.J. 2002. Heterotroph species extinction, abundance and biomass dynamics in an experimentally fragmented microecosystem. Journal of Animal Ecology 71: 594-602. Harper, K.A. et al. 2005. Edge Influence on Forest Structure and Composition in Fragmented Landscapes. Conservation Biology 19: 1-15. Hastings, A. 1980. Disturbance, coexistence, history, and competition for space. Theoretical Population Biology 18: 363-373. Holling, C.S. 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4: 1-23. Hoyle, M. 2004. Causes of the species-area relationship by trophic level in a field-based microecosystem. Proceedings of the Royal Society of London Series B 271: 1159-1164. Hoyle, M. 2005. Experimentally fragmented communities are more aggregated. Journal of Animal Ecology 74: 430-442. Hoyle, M. and Gilbert, F. 2004: Species richness of moss landscapes unaffected by short-term fragmentation. Oikos 105: 359-367. Ims, R.A. and Stenseth, N.C. 1989. Divided the fruitflies fall. Nature 342: 21-22. Kareiva, P. and Wennergren, U. 1995. Connecting landscape patterns to ecosystem and population processes. Nature 373: 299-302. Knudsen, J.W. 1972. Collecting and preserving plants and animals. Harper and Row. Krantz, G.W. 1978. A manual of acarology. Oregon State University Bookstores. Krebs, C.J., Keller, B. and Tamarin, R. 1969. Microtus population biology. Ecology: 50, 587-607. Lawton, J.H. 1994. Population dynamic principles. Philosophical Transactions of the Royal Society of London Series B 344: 61-68.  35  Leibold, M.A. et al. 2004. The metacommunity concept: a framework for multi-scale community ecology. Ecology Letters 7: 601-613. Levey, D.J., et al. 2005. Effects of Landscape Corridors on Seed Dispersal by Birds. Science 309:146148. MacArthur, R.H. 1972. Geographical ecology: Patterns in the distribution of species. Harper and Row. Mackey, R.L and Currie, D.J. 2001: The diversity-disturbance relationship: is it generally strong and peaked? Ecology 82: 3479-3492. Moilanen, A. and Hanski, I. 1995. Habitat destruction and competitive coexistence in a spatially realistic metapopulation model. Journal of Animal Ecology 64:141-144. Nee, S. and May, R.M. 1992. Dynamics of metapopulations: habitat destruction and competitive coexistence. Journal of Animal Ecology 61: 37-40. Peterson, G., Allen, C R . and Holling, C.S. 1998. Ecological resilience, biodiversity and scale. Ecosystems 1: 6-18. Pickett, S.T.A. and White, P.S. 1985. The ecology of natural disturbance and patch dynamics. Academic Press, Orlando, Fl. Redding, T.E., et al. 2003. Spatial patterns of soil temperature and moisture across subalpine forestclearcut edges in the southern interior of British Columbia. Canadian Journal of Soil Science 83: 1 2 1 130. Ries, L., et al. 2004. Ecological responses to habitat edges: mechanisms, models and variability explained. Annual Review of Ecology, Evolution and Systematics 35: 491-522. Ricklefs, R.E. and Schluter, D. (eds.) 1993. Species diversity in ecological communities. Univ. of Chicago Press. Rosenzweig, M.L. 1995. Species diversity in space and time. Cambridge Univ. Press. Schmiegelow, F.K.A., Machtans, C.S. and Hannon, S.J. 2001. Are boreal birds resilient to forest fragmentation? An experimental study of short-term community responses. Ecology 78: 1914-1932 Shurin, J.B. 2001. Interactive effects of predation and dispersal on zooplankton communities. Ecology 82: 3404-3416.  36  Simberloff, D. 2004. Community ecology: Is it time to move on? American Naturalist 163:787-799. Srivastava, D.S., et al. 2004. Are natural microcosms useful model systems for ecology? Trends in Ecology and Evolution 19: 379-384. Tilman, D. 1994. Competition and biodiversity in spatially structured habitats. Ecology 75: 2-16. Walter, D.E. and Proctor, H.C. 1999. Mites: Ecology, Evolution and Behaviour. Univ. of New South Wales Press and CAB International. Warren, P. H. 1996. Dispersal and destruction in a multiple habitat system: an experimental approach using protist communities. Oikos 77: 317-325. With, K. A., Gardner, R. H. and Turner, M. G. 1997. Landscape connectivity and population distributions in heterogeneous environments. Oikos 78: 151-169.  37  CHAPTER 3- Does regional species richness determine local species richness? An experimental test of saturation theory.  CHAPTER 3- Does regional species richness determine local species richness? An experimental test of saturation theory: I acknowledge the contributions of Diane Srivastava and Raenelle Parker to this manuscript. Diane had the original idea for the experiment, which I modified, constructed, and analyzed. Raenelle worked on the original mite library and the collection of the original moss samples. I wrote the first draft of the manuscript. A version of this chapter is to be submitted to the journal Ecology.  38  3.1 INTRODUCTION Much research in community ecology aims to understand the factors structuring species diversity in any local area (MacArthur 1972, Ricklefs 1987, Holyoak et al. 2005). Classically, local species richness has been explained in one of two ways: by referring to local processes or regional processes. Local processes have been most strongly invoked through niche theory (Pianka 1966, MacArthur and Levins 1967, Schoener 1974), where strong competition for niche space limits the number of species that can coexist at the local scale. Regional processes, on the other hand, can result from evolution through time (Jablonski and Sepkoski 1996, Ricklefs 2004), or at ecological time scales, through dispersal to the local community from a larger region (the theory of island biogeography- MacArthur and Wilson 1967). Interestingly, R.H. MacArthur played a key role in the development of theory at both scales, and this dual vision is now referred to as 'MacArthur's Paradox' (Schoener 1983). A number of experimental studies show the importance of biotic interactions in structuring local communities, and Carpenter et al. (1995) and Brown et al. (2001) review in detail the findings of their large-scale and long-term studies. Fewer studies examine how differences in regional richness structure local species richness through processes like dispersal. Nevertheless, there exist a number of empirical studies where the relationship between local and regional richness is demonstrated using regressions of local richness on regional richness (Terborgh and Faaborg 1980, Cornell 1985, Ricklefs 1987, Shurin and Srivastava 2005). Generally, linear relationships between local and regional richness have been interpreted to indicate a dominant role for regional processes (dispersal, biogeographical processes). A decelerating curvilinear line, where maximum local species richness is reached at a low level of regional species richness, indicates that local processes like competition among species are dominant, and is typically interpreted as evidence for saturation (Fig 3.1a). Tests for species saturation have been largely observational (e.g., Terborgh and Faaborg 1980, Cornell 1985, Ricklefs 1987, Cornell 1993, Kiflawi et at 2003, Caley and Schluter 1997, Cornell and Karlson 2000). The observational approach has been criticized on many fronts (Srivastava 1999, Hillebrand and Blenckner2002, Shurin and Srivastava 2005). Many of these critiques concern confounding covariates of regional richness. For example, some regions may contain more species than others simply because they are larger (Rosenzweig 1995). Similarly, locations may differ in richness because of area of, or time since, disturbance. Several models have shown that variation in either local  39  or regional area (Srivastava 1999, Hillebrand and Blenckner 1999, He et al. 2005, Shurin and Srivastava 2005) or time since disturbance (Caswell and Cohen 1993) can lead to misleading results when using the graphical method. The problems with observational approaches to testing saturation theory may be solved by experimentally manipulating regional species richness. However, there are few manipulations of experimental richness, save for seed addition experiments (Turnbull et al. 2000) and coral reef manipulations (Abele 1984). Even fewer change regional species composition. We expand on these studies by using a novel experimental approach to construct a gradient of regional richness and composition without changing regional area. We then account for time since disturbance by following the community assembly of local communities overtime. Several theoretical and observational studies have looked at the interaction between assembly time and regional species richness on local species richness (Mouquet et al. 2003, Fukami 2004, Munguia 2004). In the species competition model of Mouquet et al. (2003) the size of the regional pool is varied, and the relationship between local and regional richness changes depending on the time when sampled (Fig. 3.1b). Local versus regional richness patterns appear saturated early in the assembly process because only a subset of the regional pool has colonized any local patch. At intermediate time periods communities appear unsaturated because competitive exclusion has not gone to completion. Finally, at equilibrium, communities are again saturated due to competitive exclusion. Local richness has a predictable relationship with regional richness through time, moving from dominance by regional (dispersal) processes early in community assembly, to dominance by local (competitive interactions) processes late in the assembly process. Fukami (2004) shows that due to the effects of assembly history, local communities increase in their dissimilarity as regional pool size increases. In an observational study of the fouling community on pen shells in Florida, Munguia (2004) demonstrated different relationships between local and regional richness depending on the dispersal ability of species- motile species showed a saturating relationship, while sessile species remained unsaturated. Thus, when the assembly history of the community is short, unsaturated relationships between local and regional richness are expected; as time along the assembly trajectory increases, the relationship becomes saturated. One potential problem with these studies, however, is that they do not change community composition between regions, only  40  richness. The composition of the community may also impact the assembly of communities (Drake 1991), and may have important implications for the relationship between local and regional richness and community structure (Fukami 2004). Here we report the findings of an experiment conducted using the microarthropod community of a moss-based ecosystem. In this system, microarthropods, predominantly mites (Acari) and collembola, are found at high abundance and species richness (Davis 1981, Gilbert et al. 1998, Gonzalez and Chaneton 2002, Hoyle 2004). Patches of moss and associated microarthropods make a useful experimental system, as treatments are easily applied by collecting moss from various locations in the field spaced far apart, and then using these differences in community composition to create different treatments in a more controlled setting. This compresses the natural (3 diversity into a small area, which can be used to create regions of different species richness, as well as different community composition between similar regional richnesses. Further, the microarthropods inhabiting the moss are easily extracted using a simple Tullgren funnel technique (Kethley 1991, Knudson 1972). The small size of mites and collembola relative to treatment landscapes makes this system a small-scale analogue of macro-landscapes with much larger organisms. We address the following questions: Does local diversity reflect the impact of the regional pool (both in terms of richness and community composition), or environmental conditions (like seasonality) mediated through local interactions? How do differences in regional species richness and composition change local species richness? How does the trajectory of species richness in community assembly change due to differences in regional species richness/composition? If local determinism is dominant, then we expect the following 2 predictions to hold: 1) local richness in all regional richness treatments should be similar, due to similarity in local habitat and environmental conditions; 2) local diversity should increase through time to an asymptote (i.e., saturate). If regional species richness and composition structures local richness, then the following predictions should hold: 1) local richness will be higher in those treatments with greater regional species richness; 2) local richness will change through time as a function of differences in regional richness, due to dispersal effects (Fig. 3.1). Further, differences in regional community composition may be reflected locally, as similar regional richnesses may produce very different local communities.  41  3.2 METHODS 3.2.1 Study system The experiment was conducted using the microarthropod community of Roadside rock moss,  Racomitrium  canescens  (Hedw.) Brid., a common moss species growing on cliff faces and exposed  bedrock in coastal British Columbia (Fig. 3.2). Roadside rock moss hosts over 200 species of mites and other microarthropods (Starzomski and Srivastava, unpublished data), dominated by the Oribatid, Mesostigmatid, and Prostigmatid mites, and Arthropleonid and Symphypleonid Collembola, as well as a variety of spiders (Arachnida), and small insects (especially larval Coleoptera). The ease of manipulation of the moss into various shapes and sizes, as well as the high community diversity makes this system an excellent one for doing community ecology experiments. For this experiment we collected Roadside rock moss from several field locations and brought it to the University of British Columbia to conduct our experiment. 3.2.2 Experimental design We manipulated regional richness independently of region area by creating large patches of moss (35.5 cm 'regions') of standard area by combining moss collected from up to eight different sites (labeled A through H). Regional richness treatments were created from moss originating from 1, 2, 4 or 8 sites (Fig. 3.3). These treatments followed the random assembly method of biodiversity-ecosystem function studies (Schmid et al. 2002), since we were unable to examine all combinations of 1, 2, 4 and 8 sites. We randomly drew sites from 8 possible locations for each site number treatment. Thus, the 1 site treatment was represented by 3 site origins (C, F and H), the 2 site treatment was represented by 3 site combinations (AB, AH and BD), the 4 site treatment was represented by site combinations CDEG, CDEH, and DFGH, and the 8 site treatment had 3 replicates of the full eight-site combination (ABCDEFGH). To estimate the regional species richnesses that resulted from the creation of the regional richness treatments, samples of each moss collection site were sorted and enumerated (Table 3.1). Each of these 12 site combinations (9 different site combinations plus 3 ABCDEFGH) was reproduced 7 times to allow destructive harvesting through time. Community assembly in the treatments was followed over time by sampling a small central area of moss in each patch that we had initially defaunated.  42  3.2.3 Moss collection We collected Roadside rock moss along a latitudinal gradient (Fig. 3.2) along Highway 99 between West Vancouver and Whistler in southwestern British Columbia (between 49.30 N, 123.13 W and 50.07N, 123.08W). It is well known that communities of species in similar habitats, sampled at increasing distance from one another, tend to have progressively fewer species in common (Whittaker 1970, Rosenzweig 1995). This turnover in species richness between sites, known as f3 diversity (Whittaker 1970), allows us to sample several different sites to collect a number of different communities in similar habitat, and later add these communities together to create treatments of different regional richness. To ensure that communities of microarthropods were different between sites, the moss patches were collected from eight separate locations along Highway 99, with distances between individual sites ranging from 4.5 km to 71.5 km. A ninth location was sampled to provide moss for the local patch in the experiment. The Roadside rock moss was collected by driving to 9 sites of flat, sunny exposures containing moss overlaying a well drained granitic outcrop. A knife was used to cut 70cm by 120cm rectangles of moss from the underlying rock. This moss was stored on a plastic sheet, covered with a plastic bag, and then transported to a laboratory at the University of British Columbia.  3.2.4 Patch construction Over the course of 3 days in late May 2003, the rectangles of moss were cut into triangular 'pie-slices' of moss; these became 1 of 8 pieces of moss used to construct treatments varying in the number of ' collection regions represented. Eight 'pie-slices' of moss were placed in a 35.5 cm diameter plastic planting tray. The centre of the moss circle consisted of a 10 cm diameter plug of moss collected from a ninth location along Highway 99 (Fig. 3.2). This plug of moss was defaunated in a Tullgren funnel (for 72 hours) prior to being placed in the centre of the tray (Knudsen 1972, Kethley 1991). Eighty-four of the plastic planting trays, each with a 5 cm lip, were placed on flat particle board, spaced a minimum of 20 cm apart. A gap of even 1 cm between moss habitat patches constitutes an effective barrier to dispersal (Gonzalez et al. 1998); the combination of the 5cm lip and the 20 cm of bare particle board prevented most terrestrial dispersal between replicates. The location of the experiment was on the grounds of the south campus of the University of British Columbia, Vancouver, British Columbia (49.27N 123.26W).  43  3.2.5  Sampling  The centre plug of the moss circles was destructively sampled from each treatment at 0.5 months, 2 month, 3 months, 4 months, 8 months, 11 months and 16 months after the beginning of the experiment. Dates of sampling were divided into wet and dry season. Wet season months included September through January, with average monthly precipitation of the defined wet season of 103.6 mm. Dry season months included May through August, with average monthly precipitation of 48.1 mm. To sample a plug of moss, the centre 10cm plug of a treatment was cut out of the plastic planting tray, placed in a labeled plastic Zip-Loc  Tm  bag, and taken to our lab at the University of British Columbia. Each plug was placed in  a Tullgren funnel for 41 hours (17 hours at low-intensity, and 24 at mid-intensity) under a 40W light bulb, and the microarthropods collected and fixed in a solution of 70% Ethanol, 20% Glycerol, and 10% water. All microarthropods were stored in 30 ml vials at 4°C until manually sorted under a 60X dissecting scope. Microarthropods with distinct morphological characteristics were described as individual morphospecies, using various keys (Krantz 1978; unpublished keys provided by the Ohio State University Acarology Summer Program), and identification by experts (D. Walter, H. Proctor, H. Klompen, V. Behan-Pelletier, J. Addison). Descriptions and a key for all the morphospecies found in this study are available on the World Wide Web at http://www.zooloqy.ubc.ca/~srivastymites, and a list of the morphospecies is found in Appendix S1 in Supplementary Material. 3.2.6  Analyses  Analyses were conducted using non-metric multidimensional scaling (NMDS), Analysis of similarity (ANOSIM), and the Similarity Percentages routine in PRIMER 5 for Windows (Plymouth Marine labs) to determine differences in community structure in both the original and treatment samples. In all cases, abundance data were square-root transformed before constructing Bray-Curtis similarity matrices. ANOSIM produces a statistic called R, which is the difference in the mean ranks of similarities between groups to those within groups. Analysis of Variance (ANOVA) was used on log-transformed species richness data (various measures of species diversity were also examined, with no qualitative difference in the results, and are not considered further) to test for differences in species richness between treatments. Linear regression was used to look at the relationship between number of regions represented and  44  regional species richness. ANOVAs and linear regression analyses were conducted using R version 2.0.1 (r-project.org).  3.3 Results Two hundred thirty four species were present in the regional pool, of which 205 were subsequently found in one of the experimental treatments. Roadside rock moss microarthropod communities were significantly different amongst sites spanning Vancouver and Whistler, British Columbia (ANOSIM: R = 0.797, p = 0.001, Fig. 3.2). The number of species found in community samples from each of the 8 collection sites ranged from 26 to 83 per 314 cm of moss, with a mean of 50. Consequently, when moss 2  from differing sites was assembled into 'regions', regional species richness varied between 34 and 234 species (Fig. 3.3). In general, the regional species richness increases in a linear fashion with increased numbers of sites represented by moss pie-slices (Fig. 3.3, Table 3.1a). Therefore, in the remainder of the results we use number of sites per region as a proxy for regional richness. There were differences in local species richness overtime  (F ,7 6  6  = 20.74, p < 0.0001; Fig. 3.4).  Species richness increased quickly from June to August, 2003, after which it rapidly decreased to a low in September 2003, and then increased through to the end of the experiment in August 2004. Differences in regional richness had no overall effect on local richness over the course of the experiment (ANOVAF  3i7g  = 1.54, p = 0.22; Fig. 3.4), nor on any particular date (Table 3.2). The community composition of the  original surrounding region, as represented by each of the 10 different combinations of original sites in the region (e.g., C, AB, or DFGH) also did not significantly impact local richness ( F  9]73  = 1.32, p = 0.24).  Importantly, neither regional richness (number of sites represented in region) nor regional composition (identity of sites represented in region) determined changes in local richness overtime (2-way ANOVA; F ,56 3  = 0.85, p = 0.47 for Number of sites X Time; F  9|70  = 0.82, p = 0.60 for Identity of site X Time).  At 2 weeks after the beginning of the experiment, local richness showed a negative relationship with the initial regional richness determined from our presampling of the moss 'pie-slices' (Fig. 3.5). At 4 weeks, this changed to a positive relationship between local and regional richness, while at 8 weeks, the relationship again is negative. At 6 months, there is a hump shaped relationship, while at 16 months, the relationship is a flat line. Overall, the relationship between local and regional richness, where local  45  richness is regressed on regional richness, is a flat line, indicating no effect of regional richness on local richness. Temporal differences in species richness appeared to follow seasonal patterns. We therefore divided the dates into the wet season (September through April) and the dry season (May through August). There was an effect on local species richness of the composition of the original regional community in the wet season (ANOVA, F g = 2.89, p = 0.04), though not in the dry season (F g = 0.55, 1 4  9]4  p = 0.83). There was no effect on local richness of the regional richness treatments in either the wet (F3,2o  = 1.10, p = 0.31), or dry ( F  3 5 5  = 0.75, p = 0.39) season. There was a greater number of species in  the dry season of 2004 compared to 2003 (single-tailed t-test with unequal variance, p < 0.0001). To assess the impact of species arriving with wind currents (as part of the so-called 'aerial plankton'- Hardy and Milne 1945), we calculated, for each sample, the proportion of species not originally recorded in the surrounding region ('new species'). We would always expect some new species, as even within the same region treatment there will be some variation in the exact composition of the regional pool between replicates. However, the probability of finding a new species should be constant over time. By contrast, if aerial plankton is an important contributor of new species, we would expect the proportion of new species to increase with time. The proportion of new species, however, was constant through time (regression R = 0.34, p = 0.30), suggesting that aerial plankton are important over much longer time 2  scales than the length of the experiment. Finally, we examined how local community composition was affected by regional richness, regional composition, and time. Overall, local community composition was significantly different between dates (ANOSIM, R = 0.303, P = 0.001), and regional composition (R = 0.166, P = 0.002). In the wet season, differences in regional composition did not cause differences in local composition (R = 0.031, P = 0.406). In the dry season, however, regional composition did cause a significant difference in community composition (R = 0.13, P = 0.013).  3.4  Discussion  An outstanding question in community ecology (MacArthur 1972, Ricklefs and Schluter 1993, Holyoak et al. 2005) is 'what are the local community consequences of changes in regional species richness and composition?' We took advantage of the ease of manipulating species-rich patches of moss from various  46  sites, shrinking the distance between different communities to combine many different spatially-distinct communities into regional treatments. By decreasing the spatial scale of this p diversity, we created significant differences in regional richnesses between our treatments. In our study, regional effects were found to be less important than the effects of seasonality for local richness. Despite initial differences in the regional richness and composition of experimental moss systems, our study showed no direct effect of these differences on the local species richnesses at any time along a successional gradient of 0.5 to 16 months. Differences were seen, however, in the number of species present locally between dates along the successional gradient, indicating, as we explain shortly, a complicated interaction between competition, regional community composition, and seasonality. Unlike local richness, local composition was affected by the regional species pool. Importantly in this system, the effects of biogeography (the initial composition of the regional pool) on local community composition were evident only in certain seasons. Theory predicts that the effects of regional richness will be strongest midway through succession (Mouquet et al. 2003), though there was no effect of regional richness on local richness at any our sampling dates. A potential explanation is that community assembly happens so quickly that local areas were fully saturated with species in just two weeks. Species richness continued to increase, however, more than a year after the beginning of the experiment (compare June-August 2003 with June-August 2004; Fig. 3.4). A more plausible explanation is that this system does not have a large enough window for the transitory dynamics predicted by Mouquet et al. (2003) to result in linear local-regional plots. Rather there are two states, a rapid colonization by common, fast-dispersing species, and a slow colonization by the remaining species. Several authors (Lawton and Strong 1981, Cornell 1985, Mouquet et al. 2003) have predicted that when only a subset of the species pool can quickly colonize an area, the initial pattern will be saturation. This is similar to the pattern found by Munguia (2004), who showed that in a pen shell community in Florida, motile species initially showed a saturating relationship, while slower, more sessile species had an unsaturated relationship. Contrary to this explanation, it might be argued that more species-rich regions are likely to have any particular fast-dispersing species, and thus supply more fast-dispersers to the local area (a type of sampling effect). This is true for randomly assembled species pools. Species, however, are not randomly assembled in nature or our experiment. Fast-  47  dispersing species are more likely to be common between spatially-distinct sites; it is the slower, less common species that are likely to make up the majority of the p diversity between sites (Rosenzweig 1995). As a result, while our regional treatments differed substantially in overall species richness, they may have contained similar numbers of fast-dispersing species (i.e., all of the (3 diversity is in the slowmoving species). While local richness was independent of regional richness within sampling periods, local richness was significantly different between dates. There was a general pattern of a fast rise in species richness early in the experiment (June 2003), followed by a rapid decrease in September 2003, to a further rise in species richness through to the end of the experiment (August 2004). This pattern was due to a combination of competition and the environmental changes of the wet-dry cycle this moss system experiences in nature. It is contrary to Mouquet et al.'s (2003) prediction of convergence of local patch richness after transitory assembly dynamics. In line with the predictions of Mouquet et al. (2003), competition likely took place between the early arriving fast dispersers, decreasing their total species richness in local patches. This was perhaps also enhanced by the effects of the change in season. In the moss microecosystem, there are two partially distinct communities of microarthropods found in the wet and dry seasons (Gonzalez 2000, Starzomski, pers. obs.). During the wet season, greater amounts of fungal hyphae can be observed in the soil supporting the moss, and fungal feeding microarthropods may come to dominate at this time. During the dry season, fewer fungi are observed, and detritivorous microarthropods may play a larger role. Seasonal changes may thus have a larger impact on local species richness than differing regional species pools. Assembly appears to be ongoing, 2 to 20 microarthropod generations after the beginning of the experiment. As a result, no local equilibrium was reached, unlike in Mouquet et al. (2003). This is not due to the influx of the so-called 'aerial plankton'. Many microarthropods are known to disperse in air currents as part of this 'aerial plankton' (Hardy and Milne 1945, Russell and Wilson 1996), but we could find no evidence of an increase in species through this mechanism. Instead, this pattern may be explained by the arrival of the poor dispersers from the region into the local patch. Mite species are known to move between 1 and 4 cm day" (Berthet 1964, Katsky Venter, pers. comm.), and it may simply 1  take many of these species weeks or months to move from the regional area to the local patch. Other  48  experiments with this system also suggest that the main source of mite species is contiguous moss habitat (Gonzalez 2000, Starzomski and Srivastava, in review). Further, the arrival of these slow-moving species occurs during and after competitive effects between the common, fast-dispersing species, and thus these slow moving species bring the species richness back to the original maximum, and beyond. Although regional richness does not impact local communities, regional composition does. This is a seasonally-dependent process, shown strongly in the dry season (April to August). This may result from locally- and regionally-distinct aestivating and egg stages of mites found in the soil, which our initial sampling did not pick up. Alternatively, though we took care to collect moss from similar conditions along a latitudinal gradient, it may have been that those moss samples had experienced different disturbance regimes, and thus had different compositions that were adapted to conditions along a disturbance gradient. These differing community compositions may have enabled the regional pools that contained these species to differentially contribute species after changes in the season. For instance, those sites that had not experienced a disturbance for a long time would be expected to have a higher proportion of slow dispersing species. Though many of the fast-dispersing species may have been lost due to competitive effects, the slow-dispersers could bring the species richness in each site to a similar level (e.g., similar to the equal species richness in our experiment at August 2003 and August 2004; Fig. 3.4). By having more of these slow-dispersers, dynamics of community assembly would be different seasonally, due to the different adaptations of the species present in the regional pool. Munguia (2004) found a similar pattern: different regions contained different local communities even when local richness was independent of regional richness. Thus, even though local richness is similar between regional richnesses, this does not mean that biogeography is unimportant. Changes in relative abundance and identity of species in the regional pool may be more important than the number of species rjer se in determining the structure of local communities. As a result, saturation figures of local species richness regressed on regional species richness will under-represent the role of history. Under the expectations of Mouquet et al.'s (2003) model, all local patches should eventually converge to a global equilibrium of local species richness, independent of regional richness, due to strong competitive effects between species. Our local patches did not settle on a local equilibrium. Instead, a secondary increase was seen where the final local patch species richness was higher than at any other  49  point during succession. Differences in species dispersal ability and the existence of two distinct seasonal communities (one adapted to wet season conditions, the other to dry) would prevent the mossmicroarthropod system from ever settling to a local equilibrium, as seasonal changes would favor different species and communities. This illustrates an important point in the interpretation of local on regional regressions- environmental changes may obscure the overall impacts of the regional community on the local, due to the transition between community states. This is largely born out by our figures showing local vs. regional species richness: we never see a classic straight or decelerating curvilinear line. Rather, the relationship between local and regional richness is varied, reflecting the combined effects of seasonality and community assembly in filtering species from the region into any local patch. We contend that various processes, not simply the size or composition of the regional pool, contribute to structuring local species richness in local patches embedded within regions of differing species richness and composition (MacArthur 1972, Drake 1991, Ricklefs and Schluter 1993, Shurin et al. 2000, Fukami 2004, Holyoak et al. 2005, Shurin and Srivastava 2005). Our results show the effects of regional composition in structuring local species richness, but further demonstrate that these effects of biogeography are seasonally dependent. In our moss-microarthropod system, there are at least two distinct seasonal communities: one adapted to wet conditions found between September and April, and another adapted to dry conditions found between May and August. We demonstrate that the effects of time and space (MacArthur 1972, Ricklefs and Schluter 1993, Peterson et al. 1998), in the forms of community assembly and seasonality, are responsible for dispersing and resident individuals, and can dramatically impact the interpretations of processes responsible for the formation of local species richness.  50  Table 3.1. A. Initial estimated regional diversities from sorted microarthropods. / Each of the eight sample sites is referred to by its letter code, A through H. B. Locations of the moss collection sites in southwestern British Columbia.  A. Treatment 1 1 1 2 2 2 4 4 4 8 8 8  Patch Composition  REGION REGION REGION REGIONS REGIONS REGIONS REGIONS REGIONS REGIONS REGIONS REGIONS REGIONS  Total species 34 115 97 90 128 96 206 151 132 234 234 234  c F H AB AH BD DFGH CDEH CDEG ABCDEFGH ABCDEFGH ABCDEFGH  B. Site# A B C D E F G H Centre  Location Horseshoe Bay Tantalus look-off Cheakamus Lake Shannon Falls Cheakamus Bridge Cypress #16 Squamish Sign Upper-levels  N 49 49 49 49 49 49 49 49 49  50.766 54.339 59.422 39.968 38.17 30.818 22.688 20.954 20.873  W 123 123 123 123 123 123 123 123 123  08.900 09.753 08.442 10.125 12.748 15.590 16.096 13.393 13.467  51  Table 3.2.  Summary of ANOVA results for local species richness at each date.  Groups  F-value  P-value  16 June 2003 28 July 2003 25 August 2003 22 September 2003 12 January 2004 3 May 2004 23 August 2004  1.10 0.11 0.03 0.44 0.91 2.33 0.0004  0.32 0.75 0.86 0.52 0.37 0.16 0.98  DF 3,8 3^8 3,8 3's 3^8 3^8 3,8  Regional species richness A.  Figure 3.1. Predictions for A) the impact of regional species richness on local richness; B) the pattern of local species richness with assembly time, with expected relationships between local and regional richnesses at two points along the community assembly trajectory (adapted from Mouquet et al. 2003).  53  Regional species  richness  Regional species  richness  Community assembly time B.  Figure 3.1. Predictions for A) the impact of regional species richness on local richness; B) the pattern of local species richness with assembly time, with expected relationships between local and regional richnesses at two points along the community assembly trajectory (adapted from Mouquet et al. 2003).  54  Streu: D ?  G  H  H  Figure 3.2.  Locations in Southwestern British Columbia of collection sites for moss patches used  the experiment. Also shown is a NMDS plot showing the differences between community structure among the collected moss patches.  300  0  1  2  4  8  Number of sites per region  Figure 3.3.  Experimental design and relationship between number of sites represented in region  treatments and species present. White areas = differing regions collected. Dark areas = defaunated central core. Letters refer to the differing regions, from one region represented (All A's), to eight regions represented (regions A thru H; the locations of which are randomly determined). There are 3 replicates of each treatment. The combinations shown are for illustrative purposes only.  56  ^  o /  Af  ^ nrJ <vf v^ <$? ^  *J<  e\r  nr  rJ?  Date  Figure 3.4.  Local richness versus time. Filled black circles show all data; open circles are means +/-  SE.  57  o  22 Sept 2003  CN  O  (/>  c o a: 0  • •••  O 0) Q. CO  J  i—i—rn—i—i—i  r  1 2  8  4  D O  CO  o  CO  o  CN  0 o  o  "ro o o  12 Jan 2004 o o  o o  J8  i—i—i—r  i r 1  2  6 CO  23 Aug 2004 o o  °  8  LO CN  o 0  8 LO  o  i—i—i—i—i—i—i—r 1 2  4  8  Number of sites per region Figure 3.5. Plots of local and regional species richness. A. 16 June 2003. B. 28 July 2003. C. 25 August2003. D. 22 September 2003. E.12 January 2004. F. 23 August 2004.  58  3.5 REFERENCES Berthet, P.L. 1964. Field study of the mobility of Oribatei (Acari), using radioactive tagging. Journal of Animal Ecology. 33: 443-49. Brown, J.H., T.G. Whitham, S.K.M. Ernest and C.A.Gehring. 2001. Complex species interactions and the dynamics of ecological systems: long-term experiments. Science 293:643-650. Carpenter, S.R. 1996. Microcosm experiments have limited relevance for community and ecosystem ecology. Ecology 77:667-680. Carpenter, S.R., S.W. Chisholm, C.J. Krebs, D.W. Schindler, and R.F, Wright. 1995. Ecosystem experiments. Science 269:324-327. Caswell, H. and J:E. Cohen 1993. Local and regional regulation of species-area relations: a patch occupancy model. In: Ricklefs, R.E. and D. Schluter. Species diversity in ecological communities: historical and geographical perspectives. University of Chicago Press, Chicago. Cornell, H.V. 1985. Local and regional richness of cynipine gall wasps on California oaks. Ecology 66: 1247-1260. Cornell, HV 1999. Unsaturation and regional influences on species richness in ecological communities: a review of the evidence. Ecoscience 6: 303-315. Davis, R. C. 1981. Structure and function of two Antarctic terrestrial moss communities. Ecological Monographs 51:125-143. Drake, J.A. 1991. Community assembly mechanics and the structure of an experimental species ensemble. The American Naturalist 137: 1-26. Fukami, T. 2004 Community assembly along a species pool gradient: implications for multiple-scale patterns of species diversity. Population Ecology 46: 137-147. Gilbert, F.S., A. Gonzalez, and I. Evans-Freke. 1998. Corridors maintain species richness in the fragmented landscapes of a microecosystem. Proceedings of the Royal Society of London Series b 265: 577-582. Gonzalez, A., J.H. Lawton, F.S. Gilbert, T.M. Blackburn, and I. Evans-Freke. 1998. Metapopulation dynamics, abundance and distribution in a microecosystem. Science: 281, 2045-2047.  59  Gonzalez, A. 2000. Community relaxation in fragmented landscapes: the relation between species, area and age. Ecology Letters 3: 441-448. Gonzalez, A. and E.J. Chaneton. 2002. Heterotroph species extinction, abundance and biomass dynamics in an experimentally fragmented microecosystem. Journal of Animal Ecology 71: 594-602. Hardy, A. C. and P. S. Milne 1938. Studies in the distribution of insects by serial currents. Experiments in serial tow-netting from kites. Journal of Animal Ecology 7: 199-229. He, F., K.J. Gaston, E.F. Connor, D.S. Srivastava 2005. The local-regional relationship: immigration, extinction, and scale. Ecology 86: 360-365. Hillebrand H. and T. Blenckner. 2002. Regional impact on local diversity - from pattern to process. Oecologia 132: 479-491 Hoyle, M. 2004. Causes of the species-area relationship by trophic level in a field-based microecosystem. - Proceedings of the Royal Society of London Series B 271: 1159-1164. Holyoak, M., M.A. Leibold and R.D. Holt. 2005 Metacommunities: Spatial dynamics and ecological communities. University of Chicago Press, Chicago. Hubbell, S.P. 2001. The unified neutral theory of biodiversity and biogeography. Princeton University Press, Princeton, New Jersey. 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. Jablonski, D. and J.J. Sepkoski, Jr. 1996. Paleobiology, community ecology, and scales of ecological pattern. Ecology 77: 1367-1378. Kethley, J. 1991. A procedure for extraction of microarthropods from bulk soil samples with emphasis on inactive stages. Agriculture, Ecosystems, Environment 34:193-200. Kiflawi, M., A. Eitam, and L. Blaustein 2003. The relative impact of local and regional processes on macro-invertebrate species richness in temporary pools. Journal of Animal Ecology 72: 447-452. Knudsen, J.W. 1972. Collecting and preserving plants and animals. Harper and Row, New York. Krantz, G.W. 1978. A manual of acarology. Oregon State University Bookstores, Corvallis, Oregon. Lawton, J.H. and D.R. Strong 1981. Community patterns and competition in folivorous insects. The American Naturalist 118: 317-338.  60  Loreau, M. 2000. Are communities saturated? On the relationship between a, p, and y diversity. Ecology Letters 3: 73-76. MacArthur, R.H. 1972. Geographical ecology: Patterns in the distribution of species. Harper and Row, New York. 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, New Jersey. Miller, T.E., J.M. Kneitel, and J.H. Burns. 2002. Effect of community structure on invasion success and rate. Ecology 83:898-905. Morton, R.D. and R. Law. 1997. Regional species pools and the assembly of local ecological communities. Journal of Theoretical Biology 187: 321-333. Mouquet N., P. Munguia, J.M. Kneitel and T.E. Miller 2003. Community assembly time and the relationship between local and regional species richness. Oikos 103: 618-626. Munguia, P. 2004. Successional patterns of pen shell communities at local and regional scales. Journal of Animal Ecology 73:64-74. Peterson, G., C. R. Allen, and C.S. Holling 1998. Ecological resilience, biodiversity and scale. Ecosystems 1:6-18 Pianka, E.R. 1966. Latitudinal gradients in species diversity: a review of concepts. American Naturalist. 100: 33-46. Ricklefs, R. E. 1987. Community diversity: Relative roles of local and regional processes. Science 235: 167-171. Ricklefs, R.E. 2004. A comprehensive framework for global patterns in biodiversity. Ecology Letters 7: 115 Ricklefs, R. E. and D. Schluter eds. 1993. Species diversity in ecological communities. University of Chicago Press, Chicago. Rosenzweig, M.L. 1995. Species diversity in space and time. Cambridge University Press, Cambridge.  61  Roughgarden, J. 1989 The structure and assembly of communities. In Roughgarden, R. M. May, and S. A. Levin (Eds.). Perspectives in Ecological Theory, Princeton University Press, Princeton, pp. 203-226. Russell, R.W. and X Wilson. 1996. Aerial plankton detected by radar. Nature 381: 200-201. Schluter D. and R.E. Ricklefs 1993. Convergence and the regional component of species diversity. In: Ricklefs RE & D Schluter (Eds.). Species diversity in ecological communities. University of Chicago Press, Chicago, pp. 230-240. Schmid, B., Hector, A:, Huston, M A , Inchausti, P., Nijs, I., Leadley, P.W. & Tilman, D. 2002. The design and analysis of biodiversity experiments. In: M. Loreau, S. Naeem & P. Inchausti (Eds.) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford. Pp. 61-75. Schoener, T.W. 1974. Resource partitioning in ecological communities. Science 185: 27-39. Schoener, T.W. 1983. Field experiments on interspecific competition. American Naturalist 122: 240-285. Shurin, J. B. 2000. Dispersal limitation, invasion resistance, and the structure of pond zooplankton communities. Ecology 81:3074-3086. Shurin, J. B. 2001. Interactive effects of predation and dispersal on zooplankton communities. Ecology 82: 3404-3416. Shurin, J.B and E.G. Allen. 2001. Effects of competition, predation, dispersal on local and regional species richness. American Naturalist 158: 624-637. 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. Shurin, J and D. S. Srivastava. 2005 New Perspectives on Local and Regional Diversity: Beyond Saturation. In: Metacommunities In Holyoak, M. M. Leibold and R. Holt (Eds.). University of Chicago Press, Chicago, pp. XXX-XXX. Srivastava, D.S. 1999. Using local-regional richness plots to test for species saturation: Pitfalls and potentials. Journal of Animal Ecology 68: 1-17. Srivastava, D.S., et al. 2004. Are natural microcosms useful model systems for ecology? - Trends in Ecology and Evolution 19: 379-384. Starzomski, B M . and D.S. Srivastava. Landscape geometry determines community response to disturbance.  62  Terborgh, J.W. and J. Faaborg 1980. Saturation of bird communities in the West Indies. The American Naturalist 116:178-195. Turnbull, L A , M.J. Crawley, and M. Rees. 2000. Are plant populations seed-limited? A review of seed sowing experiments. Oikos 88: 225-238. Whittaker, R.H. 1970. Communities and ecosystems. Macmillan, London.  63  CHAPTER 4- Predation overwhelms a processing chain in a bromeliad-insect food web  CHAPTER 4- Predation overwhelms a processing chain in a bromeliad-insect food web: I acknowledge the contributions of Diane Srivastava and Daryl Suen to this manuscript. I designed the food web experiment, and analysed all of the data for both experiments. Daryl contributed data from an experiment conducted for a directed studies project, which I incorporated into the manuscript after reanalyzing his results. I wrote the first draft of the manuscript. A version of this chapter has been submitted to the journal Oecologia.  64  4.1 INTRODUCTION The processes structuring ecological communities remain the fundamental research topic for community ecologists (MacArthur 1972, Strong et al. 1984, Ricklefs and Schluter 1993, Hubbell 2001, Holyoak et al. 2005). Most experimental and theoretical studies have sought to explain how these processes work through competition for one or more resources (MacArthur and Levins 1967), apparent competition through shared predators (Holt 1977), trophic interactions like strong predation (Paine 1966, Paine 1969, Shurin and Allen 2001), or spatial and temporal niche partitioning (Hastings 1977, Hastings 1980). All of these processes rely on the existence of negative interactions between species to explain how two or more species influence the abundances of one another (MacArthur 1972, Strong et al. 1984). Few studies look at the role of positive or facilitative interactions between species, and their role in structuring communities (Bertness and Callaway 1994, Heard 1994a). Even rarer are contrasts of positive horizontal (within trophic-level) and negative vertical (between trophic-level) effects on species abundance (but see Stachowicz and Hay 1999). Indeed, facilitative effects may play a greater role in community structure than is often realized (Bruno et al. 2003, Tirado and Pugnaire 2005). Processing chains, a type of facilitative effect, consist of at least two species, one of which conditions, or prepares, a resource for another species. The conditioning species has been termed the 'upstream' species (to denote its role 'up the stream' of resource processing), and this species often feeds on, and prepares a resource for, a 'downstream' species (to denote the species' position 'downstream' of the flow of resources- Heard 1994a). Heard (1994b) lists more than 50 examples of putative processing chains, from those involving microorganisms (yeasts, phytoplankton), to invertebrates (insects, mollusks), and vertebrates (gazelles, gulls, skuas). Processing chains have been demonstrated in a variety of phytotelmata (standing water found in plants), including treeholes (Paradise and Dunson 1997, Paradise 1999), Heliconia  bracts (Seifert and Seifert  1976, 1979), and pitcher plants (Bradshaw 1983, Beaver 1984, Heard1994b). Processing chains may even occur across trophic levels: a recent study has shown that bromeliad-dwelling spiders positively affect the growth of their plant hosts, by providing up to 18% of the nitrogen used by the plant, and causing up to 15% longer leaves in bromeliads with spiders (Romero et al. 2006).  65  Demonstrating the existence of facilitative effects, often in the form of processing chains, is only one part of showing their importance to community structure. Most 'evidence' of processing chains comes from observational accounts (Heard 1994a), and awaits experimental confirmation. The small amount of experimental evidence that does exist has typically involved very simplified systems, with only two species (the upstream and downstream component of the processing chain), rather than within a food web in which other factors, such as predation, can be measured. Experiments that involve trophic complexity can provide detailed answers to specific questions about the roles of various processes in communities, though they can be difficult to undertake. Experiments that test more realistic and complete communities are necessary (Carpenter 1997), and more are being performed (e.g., Kneitel and Miller 2003, Shurin 2001, Chase and Knight 2003). We conducted experiments over two years (2003-2004) to contrast the effects of predation and facilitation in the form of a processing chain, using insect larvae found in the water-filled tanks of Costa Rican bromeliads. The discrete nature of tank-forming bromeliads and the aquatic insect larvae communities found within them (Laessle 1961, Richardson 1999, Armbruster et al. 2002) makes them an ideal model system for examining the roles of facilitative processes in structuring a natural community. Several trophic levels of insects are represented, including detritivores and primary and secondary predators. Detritus forms the basal resource for the insect and microbial community (Kitching 1999, Richardson 1999), and is the main source of nitrogen for the bromeliad (Reich et al. 2003), One species found in bromeliad tanks, a Polypedilum spp. chironomid, uses the detritus particles to create protective anti-predator cases. Further, tank-forming bromeliads of the genera Vriesea and Guzmania likely contain the majority of the standing water in the wet tropical forests of northwestern Costa Rica, and provide an important resource for a large number of organisms, from aquatic microorganisms, to insects (e.g., zygopteran odonates, chironomids and scirtid beetles), and a variety of Arachnids, reptiles, and amphibians (Frank 1983, Richardson 1999). The bromeliads may accrue a benefit from many of these organisms, as bromeliads get much of their nutrient supply through specialized cells (trichomes) on the • surface of the leaves in the water filled tanks. These trichomes absorb nutrients produced in the breakdown of leaves by various invertebrates and microorganisms (Benzing 2000).  66  Past research has suggested that facilitation in the form of a processing chain may be occurring in the bromeliads Vriesea and Guzmania (Suen unpubl. obs., Srivastava 2006). We used a novel experimental approach that involved manipulating a food web to test the roles of various species in the processing chain. The combination of testing the size of particles created by upstream components of the processing chain, and the growth and emergence of the downstream species allowed us to simulate a realistic community of species and their interactions, as well as isolate the mechanism driving the facilitation effect. Finally, manipulating a full food web allowed us to separate the effects of 3 levels of trophic interactions to determine the most important components structuring an aquatic insect community. We predicted that predation would decrease the number of emerging chironomids, but that the presence of the processing chain in the absence of predation would greatly increase the emergence of detritivore chironomids. Specifically, we hypothesized that 1) the upstream components of the processing chain (tipulids and scirtids) reduced the size of leaf detritus to that usable by collector-gatherer chironomids, 2) the presence of the processing chain with natural abundances of insect larvae positively affects the emergence rate of downstream chironomids, and 3) predation by the top predator Mecistogaster modesta Selys (Pseudostigmatidae: Odonata) (Melnychuk and Srivastava 2002) would overwhelm the positive effects of the processing chain.  4.2 METHODS 4.2.1 S t u d y s i t e a n d s y s t e m The experiments were performed at the Estacion Biologica Pitilla within the A r e a de Conservation Guanacaste, Guanacaste province, Costa Rica (10°59' N, 85°26'W). Bromeliads are common in primary and secondary forest surrounding the station (Ngai et al. in review, Srivastava et al. 2005). The food webs of phytotelm bromeliads (genera Vriesea and Guzmania) are based on a detritus resource, and the insect larvae and other invertebrates found within the water-filled tanks form a complex food web of more than 70 species. We use a subset of this food web for our experiment (basing our experiment on insect natural abundances found over 7 years by D.S. Srivastava; Fig. 4.1), focusing on the following species: a tipulid (Trentepholia sp., undescribed: Diptera), two chironomid species (especially Polypedilum sp.: Diptera, which both feeds on detritus particles as well as uses them to build protective cases, and an  67  undescribed species referred to as 'yellow chironomid'), one scirtid species (Coleoptera), a tanypodine species (Chironomidae: Diptera), and the top predator, M. modesta.  The insect and rotifer community  has been studied in the past (Srivastava 2006, Srivastava and Bell unpubl. data), and it has been shown that the plant receives a benefit through the nutrients released during breakdown of the detritus (Ngai and Srivastava, in review).  4.2.2 E x p e r i m e n t a l d e s i g n 4.2.3 Experiment 1 An experiment was done in September and October 2003 to examine the role of tipulid and scirtid insect larvae in processing detritus for chironomids. This experiment examined the role of the various insect larvae in creating detritus particles of two sizes: > 102 um and < 102 um. The following 6 treatments were used, in a 2 X 3 factorial design (Table 4.1 A):  With Chironomids: 1) Scirtid-Chironomidtreatment: five medium Polypedilum chironomids (average 7mm length) and five small Polypedilum chironomids (average 5mm length), four 5mm scirtids, two 4mm scirtids. 2) Tipulid-Chironomid treatment. Polypedilum chironomids same as ScirtidChironomidtreatment, one large Tipulid (>18mm length), one medium Tipulid (1316mm length), one small Tipulid (8-10mm length). 3) Chironomids alone treatment: Only Polypedilum Chironomid treatment.  Chironomids, as in Scirtid-  Without Chironomids: 1) Scirtids alone treatment: Only scirtids, same abundance as treatment.  Scirtid-Chironomid  2) Tipulids alone treatment: Only tipulids, same abundance as for Tipulid Chironomid treatment. 3) No insects treatment: no insect larvae included.  The experiment was done using artificial microcosms (n = 5 per treatment, except for the No insects treatment, which had 10 replicates) that modeled a single bromeliad tank. Each artificial tank consisted of a 100 ml plastic cup covered in light-blocking black plastic, with a single bromeliad leaf attached to the inside to provide a route for insect emergences. Each artificial tank included, as detrital input, three dead and dried leaves collected from the tree Conostegia  xalapensis  Bonpl.  68  (Melastomataceae) (mean dry weight of 0.675 g per leaf), and one C. xalapensis 0.600 g) collected from the tanks of a wild bromeliad. The latter C. xalapensis  leaf (mean dry weight  leaves were used to  introduce natural populations of micro-organisms (algae, bacteria, and fungi feeding on the leaves) from the bromeliad tank. Each microcosm was inoculated with 55ml of stream water and 7ml of bromeliad water (containing natural micro-organisms found in wild bromeliads, but with mosquito larvae removed). Mesh coverings were placed over each artificial bromeliad to prevent oviposition and the escape of larval insects and emerged adults. The experiment ran from 15 September to 15 October, with daily examination for emerged insects. Harvesting of both experiments was completed in a single day. Detrital fragments were rinsed from intact leaves, leaf fragments larger than 1cm X 1cm, and the artificial leaf-well. The sediment from each artificial leaf-well was separated into two size ranges: <102 pm and >102 pm but <1720 pm (hereafter >102 pm). This was done by filtering each sample through a 1720 pm sieve stacked over a 102 pm sieve and finally through a qualitative Whatman filter disc (<102 pm sediment) in a micropore glass Buchner funnel suctioned by an Emerson electric pneumatic pump. Debris caught in the 1720 pm sieve was discarded. Sediment trapped in the 102 pm Nytex sieve was transferred onto another qualitative Whatman filter disc. The filter discs were dried at ~70°C for 24 hours and then weighed using a Mettler Toledo AB104-S scale. The <102 pm particle size range was assumed to contain the particles that were collected and used by chironomids because in a gut analysis of Polypedilum  chironomids, particles were smaller than the mentum width (~ 97 pm).  4.2.4 A n a l y s i s Data were log-transformed to meet assumptions of homogeneity of variances and normality, and analyzed using a two-way MANOVA with <102 pm and >102 pm sediment sizes as the dependent variables, followed by univariate two-way ANOVAs and planned comparisons.  4.2.5 Experiment 2 Aquatic insect larvae comprising the most common members of the food web (Fig. 4.1) of water-filled bromeliad phytotelmata were manipulated to form 6 treatments (Table 4.1). Insect larvae were collected from water and detritus of the bromeliads Vriesea gladioliflora  H. Wendl, Guzmania  scherzeriana  Mez,  69  and Vriesea sanguinolenta  Cogn. and Marchal, collected from primary and secondary forest surrounding  Estacion Biologica Pitilla, between 1 and 13 October 2004. Each sampled bromeliad was carefully cut apart and washed to remove insect larvae clinging to leaves within the bromeliad tanks. Ten terrestrial bromeliads of the species Vriesea gladifolia, a common tank species in Costa Rica (Janzen 1983), were used as blocks. Bromeliads were chosen by size (all greater than approximately 1500 ml in volume), and by location (all bromeliads used in this study were located on the ground within secondary forest surrounding the research station). Six leaf wells per bromeliad were chosen for experimental treatments (n = 1 replicate/treatment/block). Leaf well enclosures maintained our food web manipulations, and were constructed using 50 ml plastic centrifuge tubes. Two holes 7mm in diameter were drilled in the side of the tubes and covered with 80 um Nytex™ mesh, to allow water to flow between the tube and the water of the leaf well in which it was placed. The mesh was small enough to prohibit experimental insect larvae from leaving the tube (though eggs and possibly early instars of individual species were able to enter, it was later discovered). The top of each tube was similarly covered with 80 um mesh, to prevent reinvasion of the tube by egg-laying adults or dispersing larvae. Additionally, by enclosing the tube in mesh, emerging individuals could be captured, to differentiate loss by predation from loss by emergence as adults. Four 2 x 2 cm and eight 1 x 1 cm samples of dead leaves of the tree C.  xalapensiswere  added as a detritus resource to each treatment. These leaves were collected from C. xalapensis  trees  surrounding the station, cut into the desired shapes, and then conditioned by placing them in a bath of stream water for two days to allow colonization by native bacteria, algae and fungi. Finally, a 10 cm section of popsicle stick was included in the tube to allow emerging insects a place to pupate out of the water. By sequentially removing the top predators, the strength of predation was examined (Treatments A and B; Fig. 4.2, Table 4.1b). To test the existence of the processing chain, the number of emerged detritivore chironomids (both Polypedilum  and yellow) was compared in treatments composing the full  processing chain (Treatment C; using naturally occurring abundances of insect larvae as determined by past surveys: Srivastava, unpublished data), with a treatment with half the naturally occurring abundances of tipulid and scirtid larvae (Treatment D), as well as without the processing chain (Treatment E; with detritivore chironomids only). We predicted that M. modesta and tanypodine  70  chironomid predation would have a strong negative effect on detritivore chironomid emergence, as would the absence of the processing chain (Fig. 4.2). We further predicted that detritivore chironomid emergence would be highest with natural abundances of the upstream component of the processing chain, and significantly reduced when these upstream abundances were halved (Fig. 4.2). Emergence of the detritivore chironomid species was chosen as a proxy for the strength of the processing chain, as well as the strength of predation, as previous work has shown that these species are at the end of the processing chain in this system. The experiment consisted of six treatments to test the roles of predation and a processing chain in the emergence of the detritivore chironomid species. In this community, detritivore chironomids are the presumed beneficiaries of the processing chain from the tipulid and scirtid species, due to their small mentum width (< 100 pm, Suen unpubL data). One species of chironomid (Polypedilum) also uses the small detritus particles to create protective cases as an antipredator defense. To test for the role of predation in structuring the community, we used the entire food web including M. modesta, a tanypodine chironomid, a tipulid, a scirtid, and the two detritivore chironomid species (Table 4.1). M. modesta individuals measured on average 12 mm from labrum to beginning of caudal gills. The average length of tanypodine chironomids was 4.5 mm. In those treatments containing 4 tipulids, one each of 7.5 mm, 9.5 mm, 11.5 mm and 13.5 mm was used; in the treatment containing half abundances, one each of 9.5 mm and 11.5 mm tipulids was used. In the treatments containing twelve scirtids, six 3 mm individuals, four 4 mm, and two 5 mm individuals were used; in the half abundances treatment, three 3 mm, two 4 mm, and one 5 mm individual were put into each centrifuge tube. Polypedilum chironomids totaled 7 in each treatment. Of these, three measured 2.5 mm, three 3.5 mm, and one 4.5 mm. Yellow chironomids totaled three individuals in all treatments, one each of 2 mm, 3 mm, and 4mm. 4.2.6 Sampling Emerged adult insects in the treatments were trapped by the fine mesh on top of the centrifuge tubes. Each experimental block was visited daily to check for new emergences, which were removed for identification and stored for later analysis. In the rare case of an emerged insect being unable to be collected, the insect was identified in the field. The majority were collected for later analysis. The  71  duration of the experiment was 25 days; after this time, other insect larvae became apparent, possibly by passing through the Nytex mesh as eggs or early instars. Dry mass of remaining detritus was determined at the end of the experiment by drying the detritus on a Whatman filter for 9 hours in an oven at ~ 70°C. All insect larvae were collected and lengths measured at the end of the experiment.  4.2.7 A n a l y s i s The chironomid emergence data were analyzed using a generalized linear model with a Poisson distribution and log-link function. Overdispersion was corrected using an empirically determined scale parameter. Orthogonal planned comparisons were performed to look for the effect of predation (M. modesta treatment vs. all other treatments, and tanypodine treatment vs. all predator-free treatments), the effect of the full processing chain (full processing chain treatment vs. half- and no-processing chain treatment), and the effect of reduced abundances of the upstream portion (the tipulids and scirtids) of the processing chain (half-processing chain treatment vs. no-processing chain). All analyses were conducted in R 2.0.1 (r-project.org), using an a of 0.05. Insect larvae body size data were log-transformed to conform to ANOVA assumptions of normality and homoscedasticity. Dry leaf detritus mass and larval body size were analysed using an ANOVA (a of 0.05).  4.3 RESULTS 4.3.1 Experiment 1: Analysis of sediment produced in artificial leaf-wells The average dry mass (grams) +/- SE of <102 um, >102 um and total sediment for each treatment in the scirtid and tipulid experiments is shown in Figure 3. Overall, there are significant differences in the production of <102 um and >102 um detritus particles (2-way MANOVA, F  5|34  = 4.11, P  = 0.005), with the type of shredder or scraper (scirtid, tipulid, neither) determining the differences (2-way MANOVA,  F ,34 2  = 8.53, P = 0.001). As predicted, the Chironomids Alone treatment created an equal  amount of <102 um sediment as the No Insect treatments (Chironomids Alone effect in MANOVA, F 1.00 , P = 0.32). There were no significant interactions in either the MANOVA  (F  2 | 3  4  =  1 3 4  = 0.73, p = 0.49) or  subsequent univariate ANOVAs (> 102 um: F , = 3.16, p = 0.06; < 102 um: F , = 0.77, p = 0.47). 2  39  2  39  72  We performed planned comparisons to look for the effects of presence of processing insects (tipulids and scirtids) and insect type on detrital processing. The presence of the upstream component had a significant impact on overall production of < 102 pm and > 102 pm detritus ( F  1|34  = 4.17, p = 0.049).  Tipulids created more > 102 pm detritus than Scirtids, but an equivalent amount of < 102 pm detritus (univariate ANOVAs; , 102 pm: F  1i19  = 1.60, p = 0.12; > 102 pm: F  1|19  = 4.90, p «  0.001).  4.3.2 Experiment 2: Food web manipulation and chironomid emergences Detritivore chironomid emergences were strongly affected by predation by M. modesta, as well as by the processing chain, though only with natural abundances of the upstream component (tipulid and scirtid species; Figure 4). There was a significant difference among treatments (generalized linear model ANOVA,  F445  = 3.117, p = 0.024). Predation by M. modesta significantly reduced the number of  emerging detritivore chironomids (planned comparison, M. modesta treatment (treatment A) versus all other treatments (treatments B, C, D, and E); F  1 4 5  = 6.339, p = 0.015), while predation by the tanypodine  chironomid had no effect (planned comparison, tanypodine treatment (B) versus no predator treatments (C, D, and E); F  1 4 5  = 0.014, p = 0.905). Natural abundances of tipulids (tipulid to detritivore chironomid  ratio = 0.4) and scirtids (scirtid to detritivore chironomid ratio = 1.2) significantly increased the number of emerging detritivore chironomids over the course of the experiment (planned comparison, full processing chain (C) versus half (D) and no processing chain (E) treatments; F  1|45  = 6.094, p = 0.017). When the  natural abundances of the tipulids and scirtids were halved (tipulid to detritivore chironomid ratio = 0.2, scirtid to detritivore chironomid ratio = 0.6), there was no facilitative effect by the upstream component (planned comparison, half processing chain (D) versus no processing chain (E) treatments; F  1i45  = 0.022,  p = 0.884). The dry mass of remaining leaves showed no significant difference among treatments ( F  4 4 9  =  2.096, p = 0.097), nor did the size of the remaining chironomid larvae at the end of the experiment differ among treatments ( F  4 9 9  = 1.597, p = 0.181).  4.4 Discussion In bromeliad phytotelmata, the positive effect of a processing chain was strong and increased the emergence rate of downstream species, though only with natural abundances of upstream species. The  73  mechanism by which this occurred was likely through leaf particle size creation by upstream species. This is similar to the effects seen in other systems such as streams (Richardson and Neill 1991), pitcher plants (Beaver 1983, Heard 1994a) and treeholes (Paradise and Dunson 1997, Paradise 1999). In addition to the processing chain, predation by the top predator, M. modesta, was very strong and overrode the effects of the processing chain in almost completely removing all other insect larvae present. Our experiment determined that both tipulids and scirtids create detritus particles < 102 um in size, the size below which particles are able to pass through the chironomid mentum. In addition, tipulids create a great deal of detritus particles > 102 um (but < 1720 um) in size, which likely facilitates the growth of increased amounts of bacteria, algae and fungi as a result of the increase in surface to volume ratio, relative to intact detritus. Decreasing the size of the detritus particles may thus have accrued benefits in 3 ways to the downstream chironomids: 1) through decreasing the size of detritus particles to a size that could pass through the mentum, and thus gut, of the chironomid, 2) through decreasing the size of the detritus particles for use in protective anti-predator cases, and 3) through increasing the surface:volume ratio of the detritus, providing increased habitat for bacteria, algae and fungi, also consumed by the chironomids, in addition to detritus (Merritt and Cummins 1996). In a study in Pennsylvania treeholes, Paradise and Dunson (1997) also found a positive role of detritus processing by scirtids on chironomid growth rates. In the food web manipulation, we showed that this process holds in the absence of the top predator M. modesta when tipulids and scirtids are present in natural abundances, but is overwhelmed by the negative effects of predation in M. modesta's presence. A reduction in the number of upstream insects below natural abundances resulted in the disappearance of the processing chain. It appears that the decrease in the abundance of the tipulid and scirtid component of the processing chain decreases the production of smaller particles of detritus to the extent that any positive effects on the growth rate of the chironomids is below detection limits. We expected that tanypodine predators would also play a large role in decreasing the number of emerged chironomids, but this was not the case. The tipulids, scirtids, and chironomids were all likely too large for this secondary predator. It appears that this predator only plays a significant predatory role on the chironomid species when the chironomids are in their early instars. Because we seeded our  74  experiment with larger instars (see Methods), we may have overridden the role of the tanypodine chironomid predator by introducing its prey in instars that were larger than it could predate (though tanypodines do predate insect larvae at least as large as themselves- Srivastava, unpubl. data). Nevertheless, the tanypodines that were introduced into the experimental treatments appeared to survive (at least in the absence of M. modesta;  M. modesta completely predated the tanypodines when present);  tanypodines were present in 8 of 10 replicates of the tanypodine treatment at the end of the experiment. It may be that larger instars of tipulids, Scirtids, and chironomids do not form a large part of the tanypodine diet. The strong effect of the top predator, M. modesta, overrode the effect of the processing chain by feeding on virtually all available prey in the treatments in which it was included. This has interesting implications for the ecology of the bromeliad. On the one hand, the processing chain is slowed, and as a result, the processing of leaves into nutrients available for uptake by the plant is likely also slowed. However, by feeding on those insect larvae that might also have emerged and left the plant (with their loads of nutrients such as nitrogen), M. modesta may benefit the plant by releasing some of the nutrients of the predated insect larvae into the phytotelm in its feces (Ngai and Srivastava, in review). In natural bromeliad systems, several water-filled tanks are found in one plant, and insect larvae can move from tank to tank to greater or lesser extents, depending on the species. In our experimental setup, all insect larvae were confined to 50ml centrifuge tubes, and thus opportunities for avoiding predation through hiding or spatial heterogeneity (moving to another tank where M. modesta is not present) were lost. Interestingly, other experiments have shown that the effects of predation and habitat complexity (the number of water-filled leaf wells) effectively cancel one another out over short periods of time (Srivastava 2006). Additionally, decisions possibly made by adults at the oviposition stage (to avoid ovipositing in > 100ml bromeliads that are likely to host M. modesta, for example) were also overridden by the design. Thus, the effects of M. modesta predation were likely amplified relative to completely natural conditions. Studies in phytotelm systems typically show facilitative effects between detritivores, and processing chains may be a general feature of these systems. Research has shown the existence of processing chains in a variety of phytotelmata, including Heliconia bracts (Seifert and Seifert 1976, 1979), pitcher plants (Bradshaw 1983, Heard1994a), and treeholes (Barrera 1996, Paradise and Dunson 1997,  75  Paradise 1999), We take a step beyond the methods of these studies to look for the presence of a processing chain in a food web context. We use multi-trophic food web manipulations to show under what conditions the processing chain can exist. When strong predation is present, or natural abundances of the upstream component of the processing chain are halved, the processing chain effect is overwhelmed. Further research is necessary to determine the effect of spatial and temporal structure on processing chains, to determine whether facilitative effects are detectable in the presence of predators if spatial refugia are present. In summary, we show that interactions between species can play positive roles for one or more of the interacting species. In the bromeliad system of Costa Rica, the presence of natural abundances of upstream tipulids and scirtids can facilitate an increased emergence rate of chironomids. This processing chain occurs through the tipulid and scirtid component preparing the resource (leaf detritus) for the chironomids by breaking the leaves into smaller particles. This results in two main benefits: 1) increased surface:volume ratio provides more surface area for decomposer bacteria and fungi, which are eaten by the chironomids, and 2) through decreasing the size of leaves to enable them to pass through the mentum of the chironomids, as well as be used for the protective cases of the Polypedilum  chironomids.  In the presence of an efficient top predator, however, this positive effect can be overridden by the strong effect of predation. When examining the processes responsible for co-occurrence and relative abundance of species in communities, both traditional explanations invoking negative interactions among species (like competition for niche space, predation, and spatial heterogeneity), coupled with positive interactions like facilitation through processing chains, should be examined.  76  A. Experiment 1  Treatment  Chironomids present  Chironomids absent  Tipulid Scirtid Neither  CT: 3 Tipulids, 10 Chironomids CS: 6 Scirtids, 10 Chironomids C: 10 Chironomids  T: 3 Tipulids, 0 Chironomids S: 6 Scirtids, 0 Chironomids NI: 0 total insects  NI = no insects treatment; C = Chironomids alone treatment; S = Scirtids alone; T = tipulids alone; CS = Chironomids and Scirtids; CT = Chironomids and Tipulids  B. Experiment 2  Treatment A  MM 1  Tan 1  Tip 4  Sc 12  PC 7  YC 3  Effect Predation & Intra-guild predation & processing chain  B  0  1  4  12  7  3  Predation & processing chain  C  0  D  0  E  0  0  4  1  0 0  2  2 0  7  6 0  3 7  7  Full processing chain 3  3  Half processing chain No processing chain  Table 4.1. Treatments used in Experiment 1 (A) and Experiment 2 (B): MM = Mecistogaster Tan = tanypodine, Tip = tipulids, Sc = scirtids, PC = Polypedilum  modesta,  chironomids, YC = Yellow chironomids.  77  Mecistogaster modesta  Tanypodine Chironomid (1 sp.)  Shredders: Tipulid (1 sp)  Scrapers: Helodids (1 sp.)  Mosquito Larvae (6 spp.)  Collector-gatherers Chironomids (2 s p p .  Ciliates Flagellates Rotifers  Bacteria Algae Fungi  Detritus  Figure 4.1.  60 total species feeding relationship processing chain  Simplified bromeliad tank food web showing common species found at Estacion Biologica  Pitilla, Costa Rica.  78  CA CD U C (D DI L_  CB  E CD  .a  E  <u  a:  Figure 4.2.  Predictions for number of emerging adults per treatment. See Table 4.1b for Treatment  legend.  79  0.05  3  «T CO  • <102 um LI >102 um  0.04  n E >. t-  •o -•-» c  0.03  £ "S  0.02  CO  cu 5  0.01  CS  CT  NI  Treatment  Figure 4.3.  Average dry mass ± SE of sediment for the treatments Chironomids  Chironomid-Scirtids  (CS), Chironomid-tipulids  alone (C),  (CT), No insects (NI), Sc/rf/cte a/one (S), and  T/pw/iofs a/one  (T).  80  3.5  2.5  CJ  cu CU  1.5  • 0.5  B  Treatment  Figure  4.4.  Treatment effects on the mean number of emerging detritivore chironomids (± SE). See  Table 4.1b for Treatment legend.  81  4.5  REFERENCES  Armbruster P., R.A. Hutchinson RA and P. Cotgreave 2002. Factors influencing community structure in a South American tank bromeliad fauna. Oikos 96: 225-234. Barrera R. 1996. Species coocurrence and the structure of a community of aquatic insects in tree holes. Journal of Vector Biology 21:66-80. Beaver, R.A. 1983. The communities living in Frank (Ed.).  Nepenthes  pitcher plants: Fauna and food webs.  In  J. H.  Phytotelmata: Terrestrial plants as hosts for aquatic insect communities,  pp. 1 2 9 - 1 5 9 .  Plexus Publishing, New Jersey. Benzing, D. 2000. Bromeliaceae: Profile of an Adaptive Radiation. Cambridge University Press. Bertness M. D. and R. Calloway 1994. Positive interactions in communities. Trends in ecology and evolution 9:191-193 Bradshaw, W . E . 1983. Interaction between the mosquito  knabi,  and their carnivorous host  Wyeomyia smithii,  Sarracenia purpurea.  the midge  Metriocnemus  Phytotelmata: Terrestrial Plants as Hosts for  Aquatic Insect Communities (eds J.H. Frank & L.P. Lounibos), pp. 161-189. Plexus, Medford, NJ. Bruno, J.F..J.J. Stachowicz, and M.D. Bertness. 2003. Incorporating facilitation into ecological theory. Trends in ecology and evolution. 18:119-125 Carpenter, S.R. 1996. Microcosm experiments have limited relevance for community and ecosystem ecology. Ecology 77: 667-680. Chase, J.M and T.M. Knight. 2003. Drought-induced mosquito outbreaks in wetlands. Ecology Letters 6: 1017-1024. Cummins K.W. and M.J. Klug 1979. Feeding ecology of stream invertebrates. A n n u a l Review of Ecology and Systematics 10:147-172 Daugherty M.P. and S.A. Juliano 2002. Testing for context-dependance in a processing chain interaction among detritus-feeding aquatic insects. Ecological Entomology 27:541-553 Fincke O.M. 1992. Consequences of larval ecology for territoriality and reproductive success of a neotropical damselfly. Ecology 73:449-462 Frank J.H. 1983. Bromeliad phytotelmata and their biota, especially mosquitoes. In: Frank J H , Lounibos LP (eds) Phytotelmata: Terrestrial plants as hosts of aquatic insect communities. Plexus, Medford USA, pp 101-128 82  Hastings, A. 1977. Spatial heterogeneity and the stability of predator-prey systems. Theoretical Population Biology 12: 37-48. Hastings, A. 1980. Disturbance, coexistence, history, and competition for space. Theoretical Population Biology 18: 363-373. Heard S.B. 1994a. Pitcher-plant midges and mosquitoes: a processing chain commensalism. Ecology 75:1647-1660 Heard S.B. 1994b. Processing chain ecology: resource condition and interspecific interactions. Journal of Animal Ecology 63:451-464 Holt, R.D. 1977. Predation, apparent competition, and the structure of prey communities. Theoretical Population Biology 12:197-229. Holyoak, M., M.A. Leibold and R.D. Holt. 2005 Metacommunities: Spatial dynamics and ecological communities. Chicago University Press. Hubbell, S.P. 2001. The unified neutral theory of biodiversity and biogeography. Princeton University Press, Princeton, New Jersey. Janzen, D.H. ed. 1983. Costa Rican Natural History, University of Chicago Press, Chicago. Kitching, R.L. 2001 Food webs In phytotelmata: "Bottom-up" and "top-down" explanations for community structure. Annual Review of Entomology 46: 729-760 Kneitel, J. M. and T. E. Miller. 2003. Dispersal rates affect species composition in metacommunities of Sarracenia  purpurea  inquilines. American Naturalist 162: 165-171.  Laessle A.M. 1961 A micro-limnological study of Jamaican bromeliads. Ecology 42:499-517 MacArthur R.H. 1972. Geographical ecology. Harper and Row, New York MacArthur, R.H. and R. Levins. 1967. The limiting similarity, convergence, and divergence of coexisting species. American Naturalist 101: 377-385. Melnychuk M.C. and D.S. Srivastava 2002 Abundance and vertical distribution of a bromeliad-dwelling zygopteran larva, Mecistogaster  modesta, in a Costa Rican rainforest (Odonata: Pseudostigmatidae).  International Journal of Odbnatology 5:81-97 Merritt, R.W. and K.W. Cummins 1996. An Introduction to the Aquatic Insects of North America, Third Edition. Kendal/Hunt Publishing Company, Dubuque, Iowa.  83  Ngai, J.T., K. Kirby, B. Gilbert, B.M. Starzomski, A.J.D. Pelletier, and R. Connor. In review. The effects of land-use change on larval insect communities in Costa Rican bromeliads. Ngai, J.T. and D.S. Srivastava. In review. Predators accelerate nutrient cycling in an insect-bromeliad ecosystem Paine, R.T. 1966. Food web complexity and species diversity. American Naturalist 100: 65-75. Paine, R.T. 1969. A note on trophic complexity and species diversity. American Naturalist 103: 91-93. Paradise C.J. 1999. Interactive effects of resources and a processing chain interaction in treehole habitats. Oikos 85:529-535 Paradise C.J. and W.A. Dunson 1997. Insect species interactions and resource effects in treeholes: are helodid beetles bottom-up facilitators of midge populations? Oecologia 109:303-312 Reich A., J.J. Ewel, N.M. Nadkarni, T. Dawson, R.D. Evans 2003. Nitrogen isotope ratios shift with plant size in tropical bromeliads. Oecologia 137:587-590 Richardson B.A. 1999 The bromeliad microcosm and the assessment of faunal diversity in a Neotropical forest. Biotropica 31:321-336 Richardson, J.S. and W.E. Neill. 1991. Indirect effects of detritus manipulations in a montane stream. Canadian Journal of Fisheries and Aquatic Sciences 48:776-783. Romero, G.Q., Mazzafera, P., Vasconcellos-Neto, J., Trivelin, P.C.O.. 2006: Bromeliad-living spiders improve host plant nutrition and growth. Ecology: Vol. 87, No. 4, pp. 803-808. Ricklefs, R. E. and D. Schluter eds. 1993. Species diversity in ecological communities. University of Chicago Press, Chicago. Seifert, R.P. and F.H. Seifert 1976. A community matrix analysis of Heliconia  insect communities.  American Naturalist, 110, 461-483. Seifert, R.P. and F.H. Seifert 1979. A Heliconia insect community in a Venezuelan cloud forest. Ecology, 60, 462-467. Shurin, J. B. 2000. Dispersal limitation, invasion resistance, and the structure of pond zooplankton communities. Ecology 81:3074-3086. Shurin, J.B and E.G. Allen. 2001. Effects of competition, predation, and dispersal on local and regional species richness. American Naturalist 158: 624-637.  84  Srivastava, D.S. 2006. Habitat structure, trophic structure and ecosystem function: interactive effects in a bromeliad-insect community. Oecologia. Srivastava, D.S., J.Kolasa, J. Bengtsson, A. Gonzalez, S.P. Lawler, T.E. Miller, P. Munguia, T. Romanuk, D.C. Schneider, M.K. Trzcinski. 2004. Are natural microcosms useful model systems for ecology? Trends in Ecology and Evolution. 19:379-384. Srivastava D.S., Melnychuk M.C., Ngai J.T. 2005. Landscape variation in the larval density of a bromeliad-dwelling zygopteran, Mecistogaster modesta (Odonata: Pseudostigmatidae). International Journal of Odonatology Stachowicz, J.J. and M.E. Hay. 1999. Mutualism and coral persistence in algal-dominated habitats: the role of herbivore resistance to algal chemical defense. Ecology 80: 2085-2101. Strong, D. R., D. Simberloff, L. G. Abele, and A. B. Thistle (editors). 1984. Ecological communities: conceptual issues and the evidence. Princeton University Press, Princeton, NJ. Tirado, R. and F.I Pugnaire 2005. Community structure and positive interactions in constraining environments. Oikos 111: 437-444.  85  Chapter 5: General Conclusions  5.1 Overview of thesis In this thesis I used experimental manipulations of species-rich moss-microarthropod systems to elucidate the mechanisms structuring local diversity, using regional richness, disturbance, and habitat loss treatments. Further, in a Costa Rican bromeliad system, I examined the relative roles of positive and negative interactions between species in determining relative abundances of species in an aquatic insect food web. The use of two different systems to answer these questions may suggest that the thesis itself is disconnected; on the contrary, I argue that the use of different systems is absolutely essential to examining these factors. Furthermore, examining many factors that can structure species richness and abundance points at once to the complexity of natural systems, and secondarily to similar factors that may hold in different systems. These two points suggest there may be ways to reduce this complexity. We may be able to produce cross-system predictive models yet. The most general and well-studied of ecological models are those that apply to a variety of systems (Levins 1966). For example, the Intermediate Disturbance Hypothesis was suggested as a way to explain the effects of disturbance on sessile marine invertebrates (Connell 1978); it has since been applied to a variety of systems, with some success (Sousa 1979), and some failure (Mackey and Currie 2001). The species-area relationship (represented by log-normal species-abundance relationships and the formula S = cA (Preston 1962, Rosenzweig 1995) is considered to be one of the few 'laws' of z  ecology (Lawton 1999), and has been found to hold in many systems (Rosenzweig 1995). As a further example, few models have elicited as much attention (more than 1000 citations) across as many systems (e.g., snails: Lassen 1975; Amazon forest fragments: Zimmerman and Bierregaard 1986; mammals: Lomolino 1986) as the Theory of Island Biogeography (MacArthur and Wilson 1963, 1967). I think it is fair to say that a large subset of ecologists wish that simple rules held in complex ecosystems. That this is not true is demonstrable; however, there are structuring factors which do hold across many ecosystems (see the above examples, and the below for more specific cases), and it is worth pursuing these factors across as many systems as possible. This was my goal in this thesis- to look at those factors that determine local species richness and abundance in two separate and very different systems. Different systems have proven useful for answering different questions in ecology. For example, whole-lake manipulations (e.g., Schindler 1977, Carpenter and Kitchell 1993) have taught ecologists  87  about the important effects of nutrients (e.g., phosphorous- Schindler 1977) and acid rain in aquatic systems (Schindler 1989), as well as the importance of trophic regulation of aquatic food webs (Carpenter and Kitchell 1993). Manipulations of large terrestrial sites have found strong effects of both bottom-up and top-down controls of terrestrial biomass (Krebs et al. 2003). Observational studies of desert rodents have shown the importance of habitat size in maintaining viable populations (Brown et al. 2000), and experimental studies in the same system have demonstrated a number of complicated interactions between trophic structure and biomass (Brown 1984, Brown et al. 2001). This list can be extended, but an important conclusion is that similar factors may structure communities across different systems, including across the terrestrial-aquatic boundary. Thus, it may be most powerful to employ research programs that utilize a variety of different study systems to learn about general structuring factors in nature. Various factors affect local species richness and abundance. Generally, they can be divided into two groups- abiotic and biotic. Abiotic factors are those non-living factors, like fragmentation, weather and disturbance (whether fire/storm/geological, etc.). Biotic factors run a gamut of interactions between species, and include predation, competition for resources, dispersal rate, and facilitation. In this thesis, I have used disturbance and fragmentation (abiotic factors), as well as differences in the regional richness of an ecosystem (biotic factor), predation (biotic factor), and facilitative interactions among species (along a processing chain- a biotic factor), to examine how each determines local species richness and abundance. To analyze the effects of these various factors on natural communities, two systems were necessary: landscape effects were most easily measured in the moss-microarthropod system (where landscapes could be miniaturized), and interactions between individual species were best examined using the bromeliad system (where the abundances of species could be manipulated). The 3 major findings of this thesis are: Loss of habitat, combined with fragmentation of habitat, decreases the resilience of local communities to disturbance. There is a threshold between loss of undisturbed habitat and full fragmentation where processes that allow the maintenance of background species diversity are lost. Increasing the loss of undisturbed habitat decreases the number of species found to recolonise locally disturbed patches. With complete loss of undisturbed surrounding habitat, and thus loss of connectivity between disturbed and undisturbed  88  habitat, the recolonisation of locally disturbed patches is much lower in the moss-microarthropod system (Chapter 2). Seasonality is more important than differences in regional species richness in controlling local richness. In moss microecosystems, differing regional richness treatments do not lead to differing local species richnesses. Rather seasonality, in the form of moving between wet and dry seasons, exerts stronger control over local species richness, as does the point along a community assembly trajectory. Despite differences in biogeography (regional species richness), the effects of differences in regional species richness are overwhelmed by the impact of changes in season, and the effect of the differing regional community composition is only evident in the dry season (possibly as a result of the activation of aestivating mite stages, as well as the presence of eggs in the soil- Chapter 3). Predation overwhelms the effects of a processing chain in Costa Rican bromeliads. The strong impact of a top predator in water-filled bromeliad tanks greatly decreases the strength of a detritus-based processing chain. In the absence of the predator, however, the effect of the processing chain is strong, which increases the emergence of adult chironomid species at the end of the processing chain. The likely mechanism of the processing chain is the creation, by tipulids, of smaller pieces of detritus that are fed upon by chironomid species. These smaller pieces of detritus also increase the surface:volume ratio of the remaining detritus fragments in bromeliads, which may facilitate the increased growth of bacteria, algae, and fungi, also fed upon by chironomids (Chapter 4).  5.2 General implications Loss of habitat, combined with fragmentation of habitat decreases the resilience of local communities to disturbance. Loss and fragmentation of habitat is a global process, and is the main factor in explanations of loss of species at local and global scales (Dirzo and Raven 2003). Much remaining habitat is subject to multiple uses and disturbances (Dirzo and Raven 2003). The results of this thesis (Chapter 2) show that once an area of undisturbed habitat surrounding disturbed local patches has declined past a threshold, there is a negative effect for disturbed local patches, which cannot rebound to pre-disturbance levels of local richness. Perhaps the most important implication of this from the point of view of community ecology  89  theory is that removal of the links that connect a local community to the metacommunity causes dramatic impact. While a simple result, it is of profound importance in understanding h o w populations and communities interact within metacommunities (Holyoak et al. 2005), and suggests that model ecosystems like moss and associated microarthropods, as well as phytotelmata like bromeliad tank communities, may have a large role to play as models for spatial effects in metacommunity research. M u c h more research is likely to come out of these systems on these questions, as described in further detail below.  Seasonality is more important than differences in regional species richness in controlling local richness. Much research has been conducted on the role that regional diversity plays in structuring local diversity, resulting in over 100 published studies to date (reviewed in Srivastava 1999, Huston 1999, Cornell 1999). Many of these studies regress local species diversity on regional species diversity to look for the dominance of regional (dispersal) or local processes (competition). Several recent papers have pointed out flaws in this graphical method, resulting from differences in covariates of regional diversity (e.g., area, disturbance rate, differences in points along an assembly trajectory- reviewed in Srivastava 1999, Hillebrand and Blenckner 2002, He et al. 2005, Shurin and Srivastava 2005). In Chapter 3, I show that despite differences in regional richness of almost an order of magnitude, local patches within these different regional richness treatments were not significantly different. Instead, local patches differed amongst different dates, due to a combination of point along an assembly trajectory, as well as changes in season. These results suggest that even between regions of different community structure, local community structure may be largely the same if it is subjected to similar seasonal effects. As a result, interpretation of local versus regional regressions should take differences in seasonality into account. W e must look beyond the simple dichotomy of local (niche-based interactions) and regional (dispersal) explanations for the factors that structure local species richness.  Predation overwhelms the effects of a processing chain in Costa Rican bromeliads. The experiments of Chapter 4 show that positive interactions between species, in the form of a processing chain w h e r e one or more species prepares a resource for another species, can be important in determining the abundances of species in a community. I show that by having m e m b e r s of the  90  upstream component of a processing chain (detritivorous tipulids and scirtids) in a Costa Rican bromeliad phytotelmata, the emergence of the downstream component (detritivorous chironomids) is enhanced. However, in the presence of the top predator, a damselfly larva, the processing chain disappears, due to intense predation. Thus, both positive and negative interactions can be important in community structure. Much research is being conducted into these sorts of questions, with the intent to produce simple and usable models of the relative contributions of habitat structure and species interactions (e.g., Gotelli and Ellison 2006, Shurin 2006). These effects may be moderated by the inclusion of spatial structure in the system (Srivastava 2006), and more research into the effect of species interactions and habitat structure will be necessary to produce useful heuristic models of these complex food webs.  5.3 Future directions Future research is needed to: •  Determine the point in loss of regional habitat and increased fragmentation where the resilience of a system is lost. An experiment to examine this would add one treatment to the work done in Chapter 2: a set of 4 locally disturbed patches, surrounded by a perimeter of 1 cm-thick moss, though otherwise disconnected from one another.  •  Continue to pinpoint the factors controlling the relationship between local and regional species richness, and how plots of local versus regional richness may be used to look at other factors beyond the roles of dispersal versus niche structuring in local community composition. Further work could be conducted to look at the effects of regional heterogeneity in reducing the impacts of seasonality on regional species richness, as well as an experiment similar to that in Chapter 3, with addition of regional richness samples through time.  •  Determine the roles of time and space in mitigating the effects of strong predation on processing chain strength in Costa Rican bromeliads. Does spatial complexity in terms of the number of tanks/plant decrease the strong top-down effects of predation (Srivastava 2006)? Does temporal partitioning of the peak breeding times by members of the processing chain from the top predator,  Mecistogaster  modesta occur? Do ovipositing adults of the species present in the processing chain  make decisions on what size, location and type of bromeliad to oviposit in, based on the likelihood  91  of the presence of  M. modesta!  Experiments can be envisioned that manipulate all of these  factors.  Community ecology has embraced a theory- and experimental-approach to exploring the processes structuring the membership of natural communities. More complex models that incorporate two or more factors, yet still retain a useful degree of simplicity, are commonly produced, and experimental work tests this theory (as in this thesis), and extends the number of included factors to describe experimental systems. A s models b e c o m e more numerous and able to incorporate more factors, w e gain a new understanding of what controls local species richness and abundance. This thesis makes a small contribution to the growing literature on the structure of local communities. It is hoped that more questions like these will be answered in the coming years, untangling and further revealing the complexity and beauty of the processes that control species diversity.  92  5.4  REFERENCES  Brown, J.H. 1984. Desert rodents: a model system. Acta Zool. Fennica 172:45-49. Brown, J.H., B.J. Fox and D.A. Kelt. 2000. Assembly rules: desert rodent communities are structured at scales from local to continental. A m . Nat. 156: 314-321. Brown, J.H., T.G. Whitham, S.K.M. Ernest and C.A.Gehring. 2 0 0 1 . Complex species interactions and the dynamics of ecological systems: long-term experiments. Science 293:643-650. Carpenter, S R . and J.F. Kitchell (eds.). 1993. The Trophic Cascade in Lakes. Cambridge University Press, Cambridge, England Carpenter, S.R., J.F. Kitchell, and J.R. Hodgson. 1985. Cascading trophic interactions and lake productivity. Bioscience 35: 634-639. Cornell, HV 1999. Unsaturation and regional influences on species richness in ecological communities: a review of the evidence. Ecoscience 6: 3 0 3 - 3 1 5 . Dirzo, R. a n d P.H. Raven 2003. Global state of biodiversity and loss. A n n u a l review of environment and resources. 28: 137-167. Gotelli, N.J. and A . M . Ellison 2006. Food-web models predict species abundances in reposne to habitat change. P L O S Biology He, F., K.J. Gaston, E.F. Connor, D.S. Srivastava 2005. The local-regional relationship: immigration, extinction, and scale. Ecology 86: 360-365. Hillebrand H. and T. Blenckner 2002. Regional impact on local diversity - f r o m pattern to process. Oecologia 132: 479-491 Holyoak, M., M.A. Leibold and R.D. Holt. 2005 Metacommunities: Spatial dynamics and ecological communities. University of Chicago Press, Chicago. Huston, M. A. 1999. Local processes and regional patterns: appropriate scales for understanding variation in the diversity of plants and animals. Oikos 86: 3 9 3 - 4 0 1 . Krebs, C.J., S. Boutin, and R. Boonstra (eds). 2 0 0 1 . Ecosystem Dynamics of the Boreal Forest: The Kluane Project. University of Oxford Press, New York. Lassen, H.H. 1975. Diversity of freshwater snails in viewof equilibrium theory of island biogeography. Oecologia 19: 1-8.  93 I  Lawton, J.H. 1999. Are there general laws in ecology? Oikos 84: 177-192. Levins, R. 1966. .The strategy of model building in population biology. American Scientist 54: 4 2 1 - 4 3 1 . Lomolino, M.V. 1986. Mammalian community structure on islands: the importance of immigration, extinction and interactive effects. Biological Journal of the Linnean Society. 28: 1-21. MacArthur, R.H. and E.O. Wilson 1963. A n equilibrium theory of insular zoogeography. Evolution 17: 373-387. MacArthur, R.H. and E.O. Wilson 1967. The theory of island biogeography.  Princeton  University Press, New Jersey. Mackey, R.L and D.J. Currie 2 0 0 1 : The diversity-disturbance relationship: is it generally strong a n d peaked? Ecology: 82: 3 4 7 9 - 3 4 9 2 . Preston, F.W. 1962. The canonical distribution of commonness and rarity. Ecology 43: 185-215. Rosenzweig, M.L. 1995. Species diversity in space and time. Cambridge University Press, Cambridge. Schindler, D.W. 1977. Evolution of phosphorus limitation in lakes. Science 195: 260-262. Schindler, D.W. 1988. Effects of Acid Rain on Freshwater Ecosystems. Science 239: 149-157 Shurin, J and D. S. Srivastava. 2005 New Perspectives on Local and Regional Diversity: Beyond Saturation. In: Metacommunities M. Holyoak, M. Leibold and R. Holt (Eds.). University of Chicago Press, Chicago, pp. 399-417. Sousa, W.P. 1979. Disturbance in marine intertidal boulder fields: the nonequilibrium maintenance of species diversity. Ecology 60: 1225-1239. Srivastava, D.S. 1999. Using local-regional richness plots to test for species saturation: Pitfalls and potentials. Journal of Animal Ecology 68: 1-17. Srivastava, D.S. 2006. Habitat structure, trophic structure and e c o s y s t e m function: interactive effects in a bromeliad-insect community. Oecologia. Z i m m e r m a n , B.L. and R.O. Bierregaard. 1986. Relevance of the equilibrium-theory of island biogeography and species-are relations to conservation with a case from A m a z o n i a . Journal of Biogeography 13:133-143.  94  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092973/manifest

Comment

Related Items