THE MICRO-ECOLOGY OF STREAM BIOFILM DYNAMICS: ENVIRONMENTAL
DRIVERS, SUCCESSIONAL PROCESSES, AND FORENSIC APPLICATIONS
Dissertation
Submitted to
The College of Arts and Sciences of the
UNIVERSITY OF DAYTON
In Partial Fulfillment of the Requirements for
The Degree of
Doctor of Philosophy in Biology
By
Jennifer M. Lang
Dayton, Ohio
August, 2015
THE MICRO-ECOLOGY OF STREAM BIOFILM DYNAMICS: ENVIRONMENTAL
DRIVERS, SUCCESSIONAL PROCESSES, AND FORENSIC APPLICATIONS
Name: Lang, Jennifer M.
APPROVED BY:
______Ryan W. McEwan, Ph.D Faculty Advisor
______M. Eric Benbow, Ph.D Faculty Advisor
______Robert J. Kearns, Ph.D Committee Member
______Thomas M. Williams, Ph.D Committee Member
______Heather R. Jordan, Ph.D Committee Member
ii
ABSTRACT
THE MICRO-ECOLOGY OF STREAM BIOFILM DYNAMICS: ENVIRONMENTAL
DRIVERS, SUCCESSIONAL PROCESSES, AND FORENSIC APPLICATIONS
Name: Lang, Jennifer M. University of Dayton
Advisor: Dr. Ryan W. McEwan
Microbial activity has an essential role in ecosystem processes, and in stream ecosystems, biofilms are the base of the food web that is fueled by photosynthesis and they are integral to nutrient processing. Stream biofilms are microbial communities of algae, bacteria, fungi, and protozoa encased in an extracellular polymeric substance
(EPS) (molecules secreted by the microbes) that are attached to a substrate (e.g. rocks, leaves) in an aqueous environment. The substrate categorizes the biofilm, and organic matter like leaves and carrion such as salmon carcasses are important substrates for nutrient dynamics. In special instances, human remains may be deposited into streams and colonized by biofilms; therefore, assessing these biofilms can have direct application to the forensic sciences. Stream ecologists have extensively investigated how environmental factors influence algal community composition, while environmental microbiologists have focused on the role of bacterial communities in nutrient dynamics.
My dissertation marries these two approaches by considering biofilm communities as a
iii
functioning ecosystem and uses ecological theory as a framework to understand the dynamics of this micro-ecosystem. This framework uses aspects from landscape ecology within a larger context of community ecology to explain how the development of biofilm communities is altered by environmental factors. In addition, this framework was used to investigate biofilm development on carrion (dead animal) in a forensic science context.
iv
ACKNOWLEDGEMENTS
I would like to thank my advisors Drs. M. Eric Benbow and Ryan W. McEwan for providing me the opportunity and support that has allowed me to develop into an independent scientist. First, I thank Dr. Benbow for inviting me into his lab and for the depth and breadth of experiences that ensued. Secondly, I am thankful to Dr. McEwan for graciously taking over and managing the final steps in finishing my degree. The unwavering support was truly appreciated. I would also like to thank my committee members Drs. Robert J. Kearns, Carl Friese, Thomas M. Williams, and Heather R. Jordan for their helpful input throughout this process.
Ample waves of gratitude are designated for the extra help provided by the undergraduates, graduates, and associated individuals. Rachel Erb, Tiffany Blair, Mary
Timko, Joe Rockner, Jon White, Allison Gansel, Nikki Henger, Jamie Alferi, Will
Kmetz, Lauren Shewhart, Alex Calteaux, Patrick Vrablik, and Ali Wright were all undergraduates that contributed to data sampling and processing. Andy Lewis, Rachel
McNeish, Kathy Gorbach, and Allissa Blystone were graduate students that provided helpful discussions and constant support. The same is true for Dr. Phil Nickell and Dr.
Jen Pechal. I would also like to thank the Baudendistel family for providing a site to conduct the aquatic decomposition study in Farmersville, OH, and Dr. John R. Wallace and Rachel Erb for conducting the decomposition study in Millersville, PA. Lastly, I am
v
extremely thankful to my friends and family for their constant support and understanding when my dedication to this endeavor meant I could not spend as much time with them as
I would have enjoyed.
Additional thanks go to the University of Dayton Biology Department for financial support and utilization of equipment. This work has also been supported in part by the University of Dayton Office of Graduate Academic Affairs through the Graduate
Student Summer Fellowship and the Dissertation Year Fellowship. In addition, the Penn
State University Swine Research facility is acknowledged for providing stillborn fetal piglets.
vi
TABLE OF CONTENTS
ABSTRACT ...... iii!
ACKNOWLEDGEMENTS ...... v!
LIST OF TABLES ...... x!
LIST OF FIGURES ...... xiii!
CHAPTER 1. INTRODUCTION ...... 1!
Overview ...... 1!
Stream Biofilms ...... 4!
Biofilm Succession ...... 5!
Biofilm Architecture ...... 7!
Biofilm Function ...... 9!
Community Ecology ...... 11!
Competition ...... 12!
Predation and Herbivory ...... 15!
Symbiosis: Mutualism, Commensalism, and Parasitism ...... 16!
Summary ...... 18!
Forensic Sciences ...... 19!
Forensic Entomology ...... 20!
Aquatic Forensic Entomology ...... 21!
vii
Alternative Methods for PMSI Estimates ...... 23!
Literature Cited ...... 27!
CHAPTER 2. ABIOTIC AUTUMNAL ORGANIC MATTER DEPOSITION AND
GRAZING DISTURBANCE EFFECTS ON EPILITHIC BIOFILM SUCCESSION .... 34!
Abstract ...... 34!
Introduction ...... 35!
Methods ...... 39!
Results ...... 48!
Discussion ...... 51!
Literature Cited ...... 57!
Tables ...... 66!
Figures ...... 71!
CHAPTER 3. FREQUENT DISTURBANCE FACILITATES THE EFFECT OF
ENVIRONMENTAL FACTORS DURING AUTUMN ON EPILITHIC BIOFILM
COMMUNITY ASSEMBLY ...... 76!
Abstract ...... 76!
Introduction ...... 77!
Methods ...... 81!
Results ...... 87!
Discussion ...... 90!
Literature Cited ...... 98!
Tables ...... 105!
Figures ...... 108!
viii
CHAPTER 4. BOTTOM-UP AND TOP-DOWN INTERACTIONS BETWEEN
INVERTEBRATE GRAZERS AND EPILITHIC BIOFILMS DICTATE BIOFILM
COMMUNITY STRUCTURE THROUGHOUT SUCCESSION ...... 114!
Abstract ...... 114!
Introduction ...... 116!
Methods ...... 120!
Results ...... 130!
Discussion ...... 134!
Literature Cited ...... 140!
Tables ...... 147!
Figures ...... 155!
CHAPTER 5. INFLUENCE OF ENVIRONMENTAL FACTORS ON EPINECROTIC
BIOFILMS ...... 161!
Abstract ...... 161!
Introduction ...... 162!
Methods ...... 165!
Results ...... 171!
Discussion ...... 173!
Literature Cited ...... 177!
Tables ...... 182!
Figures ...... 185!
CHAPTER 6. CONCLUSION ...... 189!
ix
LIST OF TABLES
CHAPTER 2
1. Table 1. Water characteristics of temperature (ºC), specific conductivity (SpCond
µS/cm), total dissolved solids (TDS mg/L), pH, turbidity (NTU), dissolved
oxygen (mg/L), depth (cm), and flow velocity (cm/s2) were measured (N = 3)
throughout the study. Asterisks denote significance at P < 0.001 (ANOVA)...... 66!
2. Table 2. The influences of time, treatment, and their interaction on ARISA and
pyrosequenced communities were assessed with PERMANOVA using Bray-
Curtis distance measure and 9999 permutations...... 67!
3. Table 3. ARISA communities (N = 3) were assessed with MRPP using Bray-
Curtis distance measure to determine pairwise comparisons within time and
treatment...... 68!
4. Table 4. Alpha and beta diversity measures of the family taxonomic level of
bacterial 454-pyrosequenced communities are separated into treatment and time
categories...... 69!
5. Table 5. Biofilm characteristics of algal biomass (chlorophyll a x 67), total
biomass (AFDM), and the ratio of algal/total biomass were analyzed (N = 3) by a
two-way repeated measures ANOVA...... 70
x
CHAPTER 3
6. Table 1. Water quality environmental factors are represented as the mean (± SE)
(N = 3) for each sampling date and deployment and for the entire study. Results
of the correlation of the factors to NMDS ordination are also reported...... 105
CHAPTER 4
7. Table 1. Water characteristics were measured (N = 3) on each deployment and
harvest date. Bolded values denote significant differences at P < 0.05 (ANOVA)
throughout the study...... 147!
8. Table 2. The influences of grazing treatment, the amount of time (days) biofilms
grew in the stream, the biofilm treatment, and the factor interactions on ARISA
communities were assessed with PERMANOVA using Bray-Curtis distance
measure and 9999 permutations. Significant P-values at < 0.05 are bolded...... 148!
9. Table 3. Percent relative abundance of genera that represented abundant taxa (>
1%) of the entire dataset was averaged (± SD) for each grazing treatment...... 149!
10. Table 4. Significant indicators (P < 0.05) of the grazing treatment were
determined using indicator analysis on 454-pyrosequeced bacterial communities
at the genus level...... 151!
11. Table 5. Significant indicators (P < 0.05) of the amount of time (days) biofilms
grew in the stream and the biofilm treatment were determined using indicator
analysis on 454-pyrosequeced bacterial communities at the genus level...... 153
xi
CHAPTER 5
12. Table 1. Water quality parameters in Farmersville, OH were measured when
substrates were deployed and at every sampling date 15 m above and below the
uppermost and lowermost carcasses, respectively...... 182!
13. Table 2. Water quality parameters in Millersville, PA of dissolved oxygen (mg/L),
pH, specific conductivity (µS/cm), water temperature (oC), total dissolved solids
(g/L), oxidation reduction potential (mV), and salinity (ppt) were measured at a
single location 30 m upstream of the uppermost carcass and 30 m downstream of
the lowermost carcass on each sampling day using a Horiba® (Kyoto, Japan)
Multi Water Quality Checker (U-50 Series)...... 184!
xii
LIST OF FIGURES
CHAPTER 2
1. Figure 1. The study occurred in a 35 m stream reach (A) during autumnal leaf
deposition...... 71!
2. Figure 2. Biofilm characteristics of algal biomass (chlorophyll a x 67), total
biomass (ash free dry mass), and the ratio of algal/total biomass were measured at
14, 24, and 38 days...... 72!
3. Figure 3. Bacterial (3-D, R2 = 0.880, stress = 0.08) and eukaryotic (2-D, R2 =
0.912, stress = 0.07) community profiles generated by ARISA were analyzed by
NMDS using Bray-Curtis distance (N = 3)...... 73!
4. Figure 4. Composite bacterial communities were identified by 454
pyrosequencing and are shown as relative abundance (%) at phylum and family
taxonomic levels (N = 1)...... 74!
5. Figure 5. Bacterial community similarity at phylum and family taxonomic levels
were analyzed by group average cluster analysis with Bray-Curtis distance after a
variance stabilizing transformation available in the DESeq2 R package...... 75
CHAPTER 3
6. Figure 1. Water quality parameters were correlated with time since the start of the
study...... 108!
xiii
7. Figure 2. Biofilm characteristics of ash free dry mass (AFDM) (top) and
chlorophyll a (bottom) are overlaid on stream mean daily discharge...... 109!
8. Figure 3. Biofilm development of the different categories were assessed using A)
time and B) undisturbed development time using accumulated degree days (ADD)
as the unit of measure to account for the affect of decreasing temperature
throughout the season...... 111!
9. Figure 4. Bacteria biofilm community structure visualized using nonmetric
multidimensional scaling (3-D, R2 = 0.98, stress = 0.14) and overlaid with biofilm
category and sampling date...... 112
CHAPTER 4
10. Figure 1. Modified flow plus dark treatments (A) were built with metal screws to
create heterogeneous flow and an overhead tarp to reduce direct light by 97%. 155!
11. Figure 2. The methods of the study are represented by a schematic depicting how
the unique biofilms were created...... 156!
12. Figure 3. Biofilm characteristics of algal biomass (chlorophyll a x 67) and total
biomass (AFDM) of the no grazing control treatment (N = 2) were analyzed by a
two-way ANOVA...... 157!
13. Figure 4. The amount of biomass loss due to grazing was determined by
subtracting the biomass levels of the grazing treatments (N = 3) from the mean of
the no grazing control treatment...... 158!
xiv
14. Figure 5. Bacterial community profiles generated by ARISA were analyzed by
NMDS (3-D, R2 = 0.880, stress = 0.13) using Bray-Curtis distance (N = 3). .... 159!
15. Figure 6. Eukaryotic community profiles generated by ARISA were analyzed by
NMDS (3-D, R2 = 0.744, stress = 0.15) using Bray-Curtis distance (N = 3). .... 160
CHAPTER 5
16. Figure 1. Bacteria community structure was visualized using nonmetric
multidimensional scaling (3-D, R2 = 0.97, stress = 0.18) and overlaid with
significant factors (PERMANOVA) of biofilm type and location of study...... 185!
17. Figure 2. Bacteria communities were separated into epinecrotic (3-D, R2 = 0.97,
stress = 0.17) and epilithic (3-D, R2 = 0.98, stress = 0.14) biofilms and community
structure was visualized using nonmetric multidimensional scaling...... 186!
18. Figure 3. Succession of Dayton (3-D, R2 = 0.98, stress = 0.14) and Millersville (3-
D, R2 = 0.99, stress = 0.11) epinecrotic biofilms were depicted with NMDS
ordination using the days of decomposition as an overly, which was significant
for both communities determined by PERMANOVA...... 187!
19. Figure 4. Bacterial community structure of epilithic biofilms in Dayton is depicted
using NMDS ordination (3-D, R2 = 0.98, stress = 0.13) and overlaid with days,
tile placement, and canopy...... 188!
xv
CHAPTER 1
INTRODUCTION
Overview
Microbial organisms are the behind-the-scenes drivers of nearly all biological processes, and it is becoming increasingly clear that their incredible impact results from cooperation within communities rather than individual species (Zarraonaindia et al.
2013). The interactions and processes are so dynamic that these microbial communities resemble ecosystems. In fact, stream biofilms have been described as “miniature complex forests” (Patrick and Roberts 1979). The idea of organisms functioning in concert is an old idea in the field of ecology, which has been grappling with understanding the complex interactions of organisms within “communities” for over a century. Indeed, plant ecologist Fredrick Clements described plant communities as “super organisms” in the early 1900s, and ecologists have since developed an increasingly sophisticated suite of theories and techniques to understand the development and function of complex assemblages of organisms. Microbial ecologists are now facing these same issues, and given the ubiquity and importance of microbial communities, understanding how these communities work is an increasingly pressing scientific need. It is the aim of my dissertation research to blend well-established ecological theory and state-of-the-art
1
molecular techniques to provide novel insights into the microbial systems of stream biofilms.
The key challenge in studying microbial systems is that the organisms composing the communities are microscopic and only 1% of environmental bacteria can be cultured in the lab (Staley and Konopka 1985). Therefore, understanding which organisms are present and what roles they play has been a nearly impossible challenge for biologists.
Recent advances in DNA based techniques have made analyzing microbial biofilms a tractable biological problem because organisms can be identified by their unique genetic sequences. Now that community composition can be determined, another challenge has been developing a conceptual outline for understanding these complex communities.
Ecology houses a body of theory that provides a usable framework to study microbial communities, especially biofilm development. Patterns of community development have been studied in forest ecology for over 100 years, and provide key concepts for understanding how biofilms interact and respond to the environment during development. Biofilm development is typically explained as a successional process.
Succession is the natural sequence of species change and replacement over time
(Clements 1916, 1936) and was first articulated by Thoreau in 1860 (Thoreau and Flint
1956). For example, forests develop in stages dominated by herbs/weeds, shrubs, shade intolerant trees, and lastly shade tolerant trees.
Borrowing ideas from community ecology has become necessary to understand and appreciate the complexity and dynamic nature of microbial communities.
Community ecology is the study of the distribution, abundance, and maintenance of species in a space and over time. Basically it asks the questions; where are species, why
2
are they there, and for how long? Yet this is extremely difficult because the sampled communities are “snapshots” within a time and space continuum. Community ecologists have described a variety of rules and mechanisms to explain the patterns of community assembly, but the result is a lattice of theories unique to the study, making it difficult to put the results into a larger context. One solution is to take a macroecological approach by broadening from the local scale to regions or ecosystems. This would “get above the mind-boggling details of local community assembly to find a bigger picture, whereby a kind of statistical order emerges from the scrum” (Lawton 1999). This view is well suited for microbial ecology because entire communities are typically sampled and the focus is on how the community functions or responds as a unit.
Stream biofilms have been studied extensively, but only recently has the entire community been considered, and rarely has it included an ecosystem approach.
Traditional research within stream ecology focused on the algal community but lacked theoretical concepts explaining community patterns (Larned 2010) and did not account for the role of bacteria. By treating biofilms as ecosystems, ecological theory can be used as a framework to answer how and why environmental factors dictate biofilm community patterns, filling a knowledge gap within stream ecology. This approach is essential because biofilms are not a simple, static layer of slime that is uniform throughout the stream, but rather a dynamic heterogeneous micro-ecosystem that interacts with and influences the larger stream ecosystem (Battin et al. 2003a).
In addition, by demonstrating that biofilms develop in a successional manner on a carrion resource, this process can potentially be used to estimate post mortem submersion interval (PMSI) in the forensic sciences. Currently, there is no standard protocol for
3
determining the PMSI of human remains found in aquatic habitats and methods with potential to address this problem need to be investigated.
Stream Biofilms
Stream biofilms are diverse microbial communities that function as micro- ecosystems. The composition and community dynamics of biofilms are influenced by both biotic (e.g. competition, predation) and abiotic (e.g. attachment substrate, stream chemistry, nutrient availability, disturbances) factors. For instance, biofilms on wood or leaves (i.e. epixylic) generally have higher heterotrophic biomass represented by bacteria and fungi compared to those on inorganic substrates like rocks (i.e. epilithic) (Giller and
Malmqvist 1998). Algae typically dominate epilithic communities and are the main source of primary (photosynthetic) production in streams (Larned 2010). Bacteria and fungi are important in decomposition and nutrient processing, while many protozoa are predators that feed on the attached microbes. This diversity within the biofilm is promoted by the ‘microbial loop’ or interactions between autotrophic (food-producing) and heterotrophic (food-consuming) organisms where inorganic nutrients like nitrogen are cycled by bacteria (Lock et al. 1984) and algae produce organic nutrients (Rier et al.
2014). Also, secreted enzymes can accrue in the matrix and break down larger organic particles, increasing nutrient availability and supporting both biofilm biomass and nutrient processing within the stream ecosystem (Rier et al. 2014).
Biofilms are found in a variety of aquatic environments, but streams are particularly interesting because they are characterized by high spatial and temporal variability, which influences biofilm ecology (Lear et al. 2008). Natural disturbances,
4
such as invertebrate grazing and flow velocity changes, are considered the most important sources of variability because they operate at multiple spatio-temporal scales
(Townsend 1989). Due to this variability, stream community structure and function are often explained within the context of patch dynamics (Pringle et al. 1988, Townsend
1989). In fact, the patchy distribution of flow velocity and the associated hydraulic parameters such as shear forces and lift promotes biofilms that differ in community composition (Besemer et al. 2007). The heterogeneous (i.e, patchy) nature of streams results in variable niche diversity, thus allowing coexistence and increased biodiversity with cascading effects on ecosystem processes (Cardinale et al. 2002). At a larger scale, increases in stream discharge that define spat and flood disturbances, as well as prolonged droughts and lower discharge, can shift the patch mosaic of biofilm communities at the stream reach scale (Lake 2000, 2003).
Biofilm Succession
Even though biofilm development is rich in complex interactions that will influence organism composition, some general trends characterize stream biofilm succession. The first colonizers, predominately bacteria, arrive after organic molecules cover the substrate surface. The bacteria actively attach quickly and dominate the pioneer community (Pohlon et al. 2010), remaining as the base layer for subsequent immigration of larger microorganisms by providing attachment sites (Hodoki 2005, Augspurger et al.
2010). As the algal canopy forms, the community becomes dominated by algal diatoms
(unicellular algae with a silica cell wall) (Besemer et al. 2007). Then, larger and filamentous algae are incorporated until filamentous algae dominate at the mature stage.
5
It is important to note that bacteria and other algae do not disappear throughout succession; they are just not always the dominant forms of biomass. This succession ranges from weeks to months depending on both abiotic and biotic factors, and might never reach the mature stage if disturbance (i.e., grazing or flow scouring) events occur too frequently. This process is similar to forest succession where organisms progressively increase from low (i.e., bacteria or weeds) to high (i.e., filamentous algae or trees) physical stature (Hoagland et al. 1982).
The process and outcome of biofilm successional development is influenced by both abiotic and biotic factors. Light, flow, and nutrients are the most important abiotic factors. Light determines algal presence, and the intensity can vary algal composition and biomass, which influences how invertebrate grazers (e.g. insect, snail) respond to the biofilm (Wellnitz et al. 1996). Flow is the primary physical force in streams and is categorized by both velocity and direction. Biofilms in slow moving water are typically thicker than those in fast moving water (Battin et al. 2003b) and lack long thin structures called streamers. Turbulent flow (water moving in various directions) compared to laminar flow (water moving in one direction) creates biofilms with increased surface heterogeneity (Besemer et al. 2007). Nutrients are responsible for the types of organisms that can be supported within biofilms, and studies have demonstrated that stream biofilms are limited by both nitrogen and phosphorus (Tank and Dodds 2003). When nutrients are added naturally through leaf decomposition, bacterial and eukaryotic microbial communities respond in sync rather than competing for nutrients (Danger et al. 2013), furthering the evidence that biofilms are micro-ecosystems.
6
Invertebrate grazing is the most influential biotic factor. Grazers reduce biomass and alter community composition, but the effect depends on the specific organism’s physiology and energy needs. In addition, grazers demonstrate optimal foraging of the biofilm resource by selecting patches (Kohler 1984) that are of a high food resource quality. Because community composition and physical structure are one in the same, all of these factors contribute to the diversity and overall architecture of biofilms.
Biofilm Architecture
Biofilms as microbial communities have a unique property in that their physical architecture is a reflection of community composition and the surrounding flow environment. Community composition influences structure formation of biofilm architecture (Besemer et al. 2009), but flow is just as important because it is the primary physical force in streams (Besemer et al. 2007). Together these factors dictate the architectural development of biofilms (Battin et al. 2003b, Besemer et al. 2009). Biofilms start as small, simple forms that develop into larger, relatively dynamic forms (Hoagland et al. 1982). As the pioneer community, bacterial biofilms will form cone and mushroom shaped cell clusters separated by channels and voids (Stoodley et al. 1994). This formation aids nutrient and oxygen uptake and minimizes diffusion through the matrix
(Costerton et al. 1995). In turbulent flow conditions, these structures condense into ridge- or ripple-like forms paralleling the direction of flow. This type of ordered formation is not typical in other flow conditions (Stoodley et al. 2002). As the biofilm matures, filamentous streamers and protrusions develop at the surface while ridges are joined by quasihexagonal structures (Battin et al. 2003b). Joint ridges may provide structure to the
7
biofilm that resists scouring by flow. Pennate diatoms are the main component of the quasihexagonal structures, while larger algae create the protrusions (Battin et al. 2003b).
Bacteria, including cyanobacteria, provide the scaffolding of streamers, which diatoms and other algae readily colonize (Besemer et al. 2007). Streamers can form up to several centimeters in length and allow the biofilm to increase in mass without sloughing as a consequence of increased hydraulic stress and friction (Besemer et al. 2009). In addition, new biofilms can form from bits of streamers that break off allowing for dispersal to new substrates.
The physical structure of biofilms also influences particle deposition from the water column. Protrusions create rough microtopography that changes the hydrodynamics around the surface of the biofilm by creating turbulence that influences settling and filtration of particles (Augspurger et al. 2010). ‘Landing sites’ of microbeads and stained bacterial cells were located between protruding diatoms that presumably created micro- eddies. Biofilms in both laminar and turbulent flow conditions showed similar surface roughness and landing distribution of both particles and bacteria, indicating that biofilm architecture changes hydrodynamics in a way that promotes nutrient uptake and colonization regardless of flow regime.
The physical three-dimensional structure changes throughout succession but small-scale structures may have an impact on successional trajectories (Hoagland 1982;
Besemer et al. 2007). Bacterial communities under different flow regimes had very similar pioneer communities that then diverged based on the type of flow condition
(Besemer et al. 2007). Yet, these communities converged into similar communities while increasing in diversity at the same time filamentous algae became dominant. The
8
conclusion was that algae were “ecosystem engineers” that were driving bacterial community succession through biophysical coupled controls (Besemer et al. 2007).
Presumably, the filamentous algae modified the flow around the biofilms creating similar flow microenvironments regardless of the surrounding flow regime, consequently affecting the bacterial community composition. In addition, the architecture created by algae may have simultaneously eliminated the physical stress created by flow and increased niche diversity through complex structures, resulting in a more diverse bacterial community. This suggests that biofilm architecture can affect biofilm resistance and resilience to flow variability and associated disturbances.
Biofilm Function
The complex relationships between abiotic factors, biofilm architecture, and community composition dictates the overall function of the biofilm as a micro-ecosystem, which then plays a role in the larger stream ecosystem. Within the biofilm, trophic diversity is maintained through interactions between autotrophic and heterotrophic organisms (Giller and Malmqvist 1998). For example, bacteria can cycle nitrogen through coupled nitrification/denitrification processes that require aerobic (i.e. surface of biofilm) and anaerobic (i.e inner part of biofilm) environments (Lock et al. 1984). This aids algal growth, and then algae provide the bacteria with inorganic and organic nutrients during metabolism and upon death, in turn creating a ‘microbial loop’ (Giller and Malmqvist 1998). Biofilm communities are also efficient at absorbing nutrients from the water column, which subsequently has bottom up effects on stream ecosystem nutrient processes (Battin et al. 2003a). The nature of the matrix traps and stores nutrients
9
and as biofilm mass increases, so does transient storage (Battin et al. 2003a). In addition, the matrix can accrue enzymes necessary to break down larger organic particles, therefore increasing the ability to attain nutrients and influencing organic matter processing in the stream (Lock et al. 1984, Sutherland 2001). This captures nutrients in their downstream path, allowing them to enter the food web through grazing, scavenging, filter-feeding and associated predation of the macroinvertebrate communities (Merritt et al. 2008).
The allochthonous organic matter inputs from the riparian and catchment runoff are important sources of nutrients and substrates for biofilms. The organic matter (OM) can be dissolved (DOM; <~0.45 um), fine particulates (FPOM; ~0.45um–1 mm), or coarse particulates (CPOM; >1 mm). DOM is the major nutrient source for biofilms, but
FPOM can be retained and degraded within the complex biofilm matrix. Epixylic biofilms use CPOM like leaves as a substrate and degrade it, subsequently releasing and absorbing nutrients and even creating FPOM (Giller and Malmqvist 1998). This makes biofilms involved in the majority of organic matter processing in streams. Their role goes even further because biofilms promote shredding invertebrates to ingest leaf material.
Invertebrates will not consume leaves until a biofilm ‘conditions’ it, presumably because physical and enzymatic degradation makes the material more suitable for digestion or uptake (Cummins 1974). Epilithic biofilms on rocks are also a food resource for grazing invertebrates. These biofilms are typically dominated by algae and are the main source of primary production within streams. These combined roles make biofilms the base of nutrient and energy flow within a stream ecosystem, which contributes to overall ecosystem function.
10
Community Ecology
Biofilms are complex communities that include all of the species interactions typical of macro-communities. Ecological communities are defined as an assemblage of populations of at least two different species that interact directly and indirectly within a defined geographic area (Agrawal et al. 2007, Ricklefs 2008, Brooker et al. 2009).
Species interactions form the basis of many ecosystem properties and processes such as nutrient cycling and food webs. The nature of these interactions can vary depending on the evolutionary context and environmental conditions in which they occur. As a result, ecological interactions between individual organisms and entire species are often difficult to define and measure and are frequently dependent on the scale and context of the interactions (Harrison and Cornell 2008, Ricklefs 2008, Brooker et al. 2009).
Nonetheless, there are several classes of interactions among organisms that are found throughout many habitats and ecosystems. Using these classes of interactions as a framework when studying an ecological community allows scientists to describe naturally occurring processes and aids in predicting how human alterations to the natural world may impact ecosystem properties and processes.
At the coarsest level, ecological interactions can be defined as either intra-specific or inter-specific. Intra-specific interactions are those that occur between individuals of the same species, while interactions that occur between two or more species are called inter- specific interactions. Because most species occur within ecological communities, these interactions can be affected by, and indirectly influence, other species and their interactions. The interactions that will be highlighted in this section are competition, predation, herbivory, and symbiosis. These are not the only types of species interactions,
11
and they are all pieces within a larger network of interactions that make up the complex relationships occurring in nature.
Competition
Competition is most typically considered the interaction of individuals that vie for a common resource that is in limited supply. More generally, it can be defined as the direct or indirect interaction of organisms that share the same resource and leads to a change in fitness. The outcome usually has negative effects on the weaker competitors.
There are three major forms of competition. Two forms, interference competition and exploitation competition, are categorized as real competition, while the third form, apparent competition, is not. Interference competition occurs directly between individuals while exploitation competition and apparent competition occur indirectly between individuals (Holomuzki et al. 2010).
When an individual directly alters the resource-attaining behavior of other individuals, the interaction is considered interference competition. For example, when a male gorilla prohibits other males from accessing a mate by using physical aggression or displays of aggression, the dominant male is directly altering the mating behavior of other males. This is also an example of an intra-specific interaction. Exploitation competition occurs when individuals interact indirectly as they compete for common resources, such as territory, prey, or food. Simply put, the use of the resource by one individual will decrease the amount available for other individuals. Whether by interference or exploitation, over time a superior competitor can eliminate an inferior one from the area, resulting in competitive exclusion (Hardin 1960). The outcomes of competition between
12
two species can be predicted using equations and one of the most well known is the
Lotka-Volterra model (Volterra 1926, Lotka 1932). This model relates the population density and carrying capacity of two species to each other and includes their overall effect on each other. The four outcomes of this model are the following: 1) species A competitively excludes species B; 2) species B competitively excludes species A; 3) either species wins based on population densities; or 4) coexistence occurs. Species can survive together if intra-specific is stronger than inter-specific competition. Meaning, each species will inhibit their own population growth before they inhibit that of the competitor, leading to coexistence.
Another mechanism that prevents competitive exclusion is alternative life history and dispersal strategies, which are usually reinforced through natural selection. This mechanism reduces competitive interactions and increases opportunities for new colonization and nutrient acquisition. The success of this is often dependent upon events
(such as tide, flood or fire disturbances) that create opportunities for dispersal and nutrient acquisition. Consider that Plant Species A is more efficient than Plant Species B at nutrient uptake, but Plant B is a better disperser. In this example, the resource under competition is nutrients. If a disturbance creates new space for colonization, Plant B is expected to arrive first and maintain its presence in the community until Plant A arrives and begins competing with Plant B. Eventually, Plant A will outcompete Plant B, perhaps by growing faster because Plant A is more efficient at nutrient acquisition. With an increasing Plant A population, the Plant B population will decline, and given enough time, can be excluded from that area. The exclusion of Plant B can be avoided if a local disturbance (e.g., prairie fires) consistently opens new opportunities (space) for
13
colonization. This often happens in nature, and thus disturbance can balance competitive interactions and prevent competitive exclusion by creating patches that will be readily colonized by species with better dispersal strategies (Roxburgh et al. 2004). The success of the dispersal versus nutrient acquisition tradeoff depends, however, on the frequency and spatial proximity of disturbance events relative to the dispersal rates of individuals of the competing species. Coexistence can be achieved when disturbances occur at a frequency or distance that allows the weaker, but often better dispersing, competitor to be maintained in a habitat. If the disturbance is too frequent the inferior competitor but better disperser remains dominant, but if the disturbance is rare then the superior competitor slowly outcompetes the inferior competitor, resulting in competitive exclusion. This phenomena, where certain levels of disturbance promotes diversity, is known as the intermediate disturbance hypothesis (Horn et al. 1975, Connell 1978). The overall theory predicts that the greatest amount of diversity within a community will be at intermediary levels of disturbance.
Apparent competition occurs when two individuals that do not directly compete for resources affect each other indirectly by being prey for the same predator. Consider a hawk that preys on both squirrels and mice. In this relationship, if the squirrel population increases, then the mouse population may be positively affected because more squirrels will be available as prey for the hawks. However, an increased squirrel population may eventually lead to a higher population of hawks requiring more prey, thus, negatively affecting the mice through increased predation pressure as the squirrel populations decline. The opposite effect could also occur through a decrease in food resources for the predator. If the squirrel population decreases, it can indirectly lead to a reduction in the
14
mouse population because they will be the more abundant food source for the hawks.
Apparent competition can be difficult to identify in nature, often because of the complexity of indirect interactions involving multiple species and changing environmental conditions.
Predation and Herbivory
Predation requires one individual, the predator, to kill and eat another individual, the prey. In most examples of this relationship, the predator and prey are both animals; however, protozoans are known to prey on bacteria and other protozoans and some plants are known to trap and digest insects (e.g., pitcher plant). Typically, this interaction occurs between species (inter-specific); but when it occurs within a species (intra-specific) it is cannibalism. Cannibalism is actually quite common in both aquatic and terrestrial food webs (Greenwood et al. 2010, Huss et al. 2010). It often occurs when food resources are scarce, forcing organisms of the same species to feed on each other. Surprisingly, this can actually benefit the species as a whole by sustaining the population through times of limited resources while simultaneously allowing the scarce resources to rebound through reduced feeding pressure (Huss et al. 2010). The predator-prey relationship can influence organismal body composition through sophisticated adaptations by both predators and prey, in what has been called an “evolutionary arms races.” Familiar predatory adaptations are sharp teeth and claws, stingers or poison, quick and agile bodies, camouflage coloration, and excellent olfactory, visual, or aural acuity. Prey species have evolved a variety of defenses including behavioral, morphological, physiological, mechanical, life history synchrony, and chemical defenses to avoid being preyed upon.
15
Another interaction that is much like predation is herbivory, which is when an individual feeds on all or part of a photosynthetic organism (plant or algae), possibly killing it (Gurevitch et al. 2006). An important difference between herbivory and predation is that herbivory does not always lead to death of the individual. Herbivory is often the foundation of food webs since it involves the consumption of primary producers
(organisms that convert light energy to chemical energy through photosynthesis). Plants, like prey, have also evolved adaptations to herbivory. Tolerance is the ability to minimize negative effects resulting from herbivory, while resistance means that plants use defenses to avoid being consumed. Physical (e.g., thorns, tough material, sticky substances) and chemical adaptations (e.g., irritating toxins on piercing structures, and bad tasting chemicals in leaves) are two common types of plant defenses (Gurevitch et al. 2006).
Symbiosis: Mutualism, Commensalism, and Parasitism
Symbiosis is an interaction characterized by two or more species living purposefully in direct contact with each other. The term symbiosis includes a broad range of species interactions but typically refers to three major types: mutualism, commensalism, and parasitism. Mutualism is a symbiotic interaction where both or all individuals benefit from the relationship and it can be obligate or facultative. Species involved in obligate mutualism cannot survive without the relationship, while facultative mutualistic species can survive individually when separated but often not as well (Ellison et al. 1996). For example, leafcutter ants and certain fungi have an obligate mutualistic relationship. The ant larvae eat only one kind of fungi, and the fungi cannot survive without the constant care of the ants. As a result, the colonies’ activities revolve around
16
cultivating the fungi. They provide it digested leaf material, can sense if a leaf species is harmful to it, and keep it free from pests. A good example of a facultative mutualistic relationship is found between mycorrhizal fungi and plant roots. It has been suggested that 80% of vascular plants form relationships with mycorrhizal fungi (Deacon 2006)
(Deacon 2006). Yet the relationship can turn parasitic when the environment of the fungi is nutrient rich because the plant takes nutrients from the fungi without providing a benefit (Johnson et al. 1997). Thus, the nature of the interactions between two species is often relative to the abiotic conditions and not always easily identified in nature.
Commensalism is an interaction in which one individual benefits while the other is neither helped nor harmed. For example, orchids found in tropical rainforests are epiphytes that grow on the branches of trees in order to access light, but the presence of the orchids does not affect the trees. Commensalism can be difficult to identify because the individual that benefits may have indirect effects on the other individual that are not readily noticeable or detectable. If the orchid from the previous example grew too large and inhibited tree growth, then the relationship would become parasitic.
Parasitism occurs when one individual, the parasite, benefits from another individual, the host, while harming the host in the process. Parasites that feed on host tissue or fluids and can be found within (endoparasites) or outside (ectoparasites) of the host body (Holomuzki et al. 2010). For example, different species of ticks are common ectoparasites on animals and humans. Parasites typically do not kill their hosts, but can significantly weaken them; indirectly causing the host to die via illness, effects on metabolism, lower overall health and increased predation potential (Holomuzki et al.
2010). For instance, a trematode that parasitizes certain aquatic snails will cause infected
17
snails to lose some of their characteristic behavior. The snails will remain on the tops of rocks where food is inadequate and even during peaks of waterfowl activity, making them easy prey for birds (Levri 1999). Further, parasitism of prey species can indirectly alter the interactions of associated predators, other prey of the predators, and their own prey. When a parasite influences the competitive interaction between two species, it is termed parasite-mediated competition, and the parasite can infect one or both of the involved species (Hatcher et al. 2006). For example, the malarial parasite Plasmodium azurophilum differentially infects two lizard species found in the Caribbean, Anolis gingivinius and Anolis wattsi. A. gingivinius is a better competitor than A. wattsi but is susceptible to P. azurophilum, while A. wattsi rarely contracts the parasite. These lizards are found coexisting only when the parasite is present, indicating that the parasite lowers the competitive ability of A. gingivinius’ (Schall 1992). In this case, the parasite prevents competitive exclusion, therefore maintaining species diversity in this community.
Summary
The species interactions discussed are the most commonly studied interactions that occur in nature and identifying their effects can get complicated because they can directly or indirectly influence other intra-specific and inter-specific interactions.
Additionally, the role of abiotic factors adds complexity to species interactions and how we understand them. That is to say, species interactions are part of the framework that forms the complexity of ecological communities and help shape community dynamics. It was originally thought that competition was the driving force of community structure, but it is now understood that all of the interactions discussed, along with their indirect effects
18
and the variation of responses within and between species, define communities and ecosystems (Agrawal et al. 2007).
Forensic Sciences
Estimating the amount of time human remains have been in an aquatic setting has proven difficult for forensic scientists, partly because the field is built upon a foundation of using insects as indicators. Forensic entomology is frequently utilized to estimate the postmortem interval of terrestrial carrion, but entomology has been less useful for decomposition within aquatic habitats. The main reason for the dearth of information is that terrestrial habitats are exposed to invertebrates that have evolved to feed solely on these carrion resources, while aquatic invertebrates have not. There is a clear successional pattern of invertebrates that utilize and breakdown terrestrial carrion. Yet, no distinct pattern has been described for aquatic carrion, creating the necessity for alternative techniques to estimate a post mortem submersion interval (PMSI). One of the most promising paths is through the use of epinecrotic (Pechal et al. 2014) biofilms that develop on the carrion resource. Although multiple authors have noted biofilm development on salmon (Claeson et al. 2006), waterfowl (Parmenter and Lamarra 1991), rats (Tomberlin and Adler 1998), and swine (Haefner et al. 2004), only a few studies have attempted to utilize them for estimating a PMSI. Even then, these studies have a narrow focus that has not addressed the more specific enivironmental factors that influence the establishment and succession of this unique biofilm community.
19
Forensic Entomology
The field of forensic entomology uses arthropods as a tool to provide insight and gain information about situations involving the judicial system. The branch most commonly associated with forensic entomology is termed medicocriminal entomology because it uses arthropods as evidence to solve crimes, which are typically violent crimes
(Hall 1990). There has been considerable research done in this field to characterize the process of decomposition with substantial amounts of literature published in the last decade, however, less than 20% of all forensic literature relates to aquatic habitats
(Merritt and Wallace 2010). Even so, most information related to aquatics pertains to case studies or terrestrial insects colonizing a corpse and are not actual studies of aquatic decomposition.
While the general decomposition process is similar in the two habitats, the ecology and components are very different. Decomposition in both habitats occurs in roughly five stages. In terrestrial habitats, the stages are fresh, bloat, active decay, advanced decay, and dry/remains (Payne 1965, Anderson and VanLaerhoven 1996), while the stages in aquatic habitats are submerged fresh, early floating, early floating decay, advance floating decay, and sunken remains (Haefner et al. 2004). Abiotic factors have very important influences on carrion decomposition within both habitats, especially temperature (Merritt and Wallace 2010), but the key difference is that terrestrial decomposition is driven by sarcophagous (flesh-eating) invertebrate colonizers that follow an ecological succession pattern (Payne 1965). A pattern of colonizing invertebrates has been well established in terrestrial carrion decomposition, and it was originally thought that similar patterns of invertebrate succession would be seen during
20
aquatic carrion decomposition (Haskell et al. 1989), yet, that has proven false. Right now, the most successful use of invertebrates in aquatic related crimes is on a case-by-case basis that involves sending samples to an aquatic entomologist. For example, black fly cocoons found on a submerged car proved a car was in the river when a suspect supposedly heard from the victim, which later contributed to his conviction (Merritt and
Wallace 2010). Another case involved caddisfly larvae of Pychnopsyche sp., and these organisms build cases as shelter made from materials like stones or twigs. A PMSI was estimated based on the life history characteristics of this caddisfly by using the time of year that they build their cases (Wallace et al. 2008). While these examples illustrate that aquatic invertebrates can be used in the forensic cases, the cases highlight the lack of understanding of the process of aquatic decomposition and its associated organisms.
Aquatic Forensic Entomology
Identifying a successional pattern of invertebrates associated with aquatic decomposition that parallels the patterns seen in terrestrial systems has been unsuccessful. There are some general trends, but nothing has been specific enough to warrant a useful pattern. Chironomidae midge larvae seem to be a commonality because they become dominate by day four on pig carcasses in a lake (Vance et al. 1995) and comprised > 95% of collected invertebrates on decomposing rat carcasses in streams
(Keiper et al. 1997). There was a pattern in diversity on the rats, but the invertebrate assemblages were different in the riffle (fast, turbulent flow) verses pool (stable, minimal flow) stream habitats (Keiper et al. 1997). Alternative invertebrate patterns were observed on pig carcasses that were clothed (Hobischak and Anderson 2002), and the
21
study was performed from December to April in British Columbia while the other studies were performed in spring and summer. Successional patterns have also been attempted using invertebrate assemblages categorized by functional feeding group (FFG), which organizes invertebrates based on their food resource and how it is obtained (Merritt and
Cummins 2006). Organic matter can be course particulates (greater than 1mm in size; e.g. leaves, twigs, plants), fine particulates (less than 1mm in size), microbes in biofilms, and other living invertebrates. For example, a collector would be an organism that uses fine particulate organic matter obtained by filtering the water or gathering the material. A study conducted in pool and riffle stream habitats in Indiana by Schultenover and
Wallace demonstrated that collectors dominated throughout decomposition, but proportions of the FFGs were different between habitats and predators were found only in the riffle habitat during late decomposition (Merritt and Wallace 2010). Also, the greatest diversity was observed at the beginning of decomposition (Merritt and Wallace 2010). In the Columbian Andes, there were similarly high frequency of collectors in a stream habitat, yet predators were just as abundant in a lake habitat (Barrios and Wolff 2011).
Once again, patterns were altered by habitat. The variation between these studies demonstrates that invertebrate assemblages on carrion do not follow any distinct pattern in aquatic habitats.
It appears that abiotic factors may prevent any replicable aquatic invertebrate successional pattern associated with carrion decomposition. Season (Tomberlin and Adler
1998), flow (Casamatta and Verb 2000, Haefner et al. 2004), habitat (Keiper et al. 1997,
Hobischak and Anderson 2002, Barrios and Wolff 2011), and temperature (Heaton et al.
2010) all influence the invertebrate taxa that are present in an aquatic system, and
22
therefore, what will colonize the carrion. This results in patterns that cannot be used as a
PMSI estimate and reiterates why a microbial focus may be more practical. Abiotic factors have more weight in dictating aquatic invertebrate assemblages on carrion because there are no purely sarcophagus aquatic invertebrates like there are in terrestrial habitats (Haskell et al. 1989, Chaloner et al. 2002). Environmental factors were the drivers of evolution for aquatic invertebrates, while the carrion resource itself was the driver of evolution for necrophagous terrestrial invertebrates. Insects like blow flies
(Calliphoridae) and carrion beetles (Silphidae) have evolved as decomposers that develop solely on carrion resources. These necrophagous invertebrates utilize the resource in a predictable successional pattern that directly influences decomposition, while aquatic invertebrates typically use carrion as a habitat substrate (Wallace et al. 2008, Barrios and
Wolff 2011). Aquatic invertebrates do not appear to be attracted to the carrion itself and may randomly encounter the carrion through natural drift or be attracted to the epinecrotic biofilm as a food resource. Multiple studies have noted the biofilm growth, and only a few have attempted to study it (Keiper et al. 1997, Hobischak and Anderson
2002, Barrios and Wolff 2011).
Alternative Methods for PMSI Estimates
The lack of a clear invertebrate pattern has lead researches to explore alternative methods to determine PMSI. Two areas of study pursued were using accumulated degree days (ADD) and microbes. Using ADD is essential in determining PMI in terrestrial habitats because the timing of insect colonization, larval growth rate, and survivorship all depend on accumulated heat (Greenberg 1991), and because insects drive terrestrial
23
decomposition, temperature is directly related to the rate and stage of decomposition. It was suggested that temperature is just as important to aquatic decomposition (Merritt and
Wallace 2010). A linear regression model correlating ADD to PMSI was developed by using information from 187 cases in the United Kingdom (Heaton et al. 2010). They illustrated that ADD significantly affected the decay process, and there was no significant difference of decomposition between the waterways (multiple canals and rivers).
The use of microbes as a forensic tool is not a novel concept. They have been used to 1) identify a crime scene 2) determine drowning as cause of death and 3) indicate
PMSI (Keiper and Casamatta 2001). For example, specific diatoms on a suspect’s shoes placed him at the crime scene and resulted in his confession (Siver et al. 1994).
Furthermore, detecting drowning is difficult and typically uses the presence of diatoms as confirmation (Piette and De Letter 2006). Because of this, marine bacterial presence was proposed to detect drowning because they are smaller than diatoms and are not naturally found in humans (Kakizaki et al. 2008). Continued research was able to distinguish between freshwater and marine bacterioplankton in the blood of drowning victims and established that these organisms do not easily invade the blood postmortem, indicating that their presence could aid drowning confirmation (Kakizaki et al. 2011). Lastly, microbes within epinecrotic biofilms that form on the surface of remains have been investigated as an indicator in determining PMSI (Merritt and Wallace 2010).
It is known that biofilm development follows a successional pattern akin to forest development (Hoagland et al. 1982), thus, a few studies have tried to utilize this to tease out a successional pattern that can be applied to estimating a PMSI. Most research has focused on using diatoms (unicellular photosynthetic organisms with cell walls made of
24
silica) because they are the most common kind of algae, do not degrade easily, and have been well studied (Lowe and LaLiberte 2006). A study on algae found, not surprisingly, that diatoms had the greatest abundance and diversity throughout decomposition
(Casamatta and Verb 2000). There was a lag phase of algal colonization with a drastic increase in taxa diversity at day 12 and cyanobacteria chlorophytes were not observed until day 17. This follows the general pattern of biofilm development where algae become established later during succession and indicates that bacteria and fungi are dominate in beginning stages.
The difference in algae development on carrion and an artificial substrate (i.e. ceramic tiles) has also been investigated. Algal growth was significantly greater on pig carcasses compared to ceramic tiles especially later in decomposition (Haefner et al.
2004) and algal diatom richness decreased over time (Zimmerman and Wallace 2008). In addition, diatom diversity was different on pigs compared to ceramic tiles (Zimmerman and Wallace 2008). Algal growth was also significantly increased on nutrient diffusing clay pots compared to control pots (Fairchild et al. 1985), which suggests that the algae may take up nutrients being released as the carrion decomposes. The differences in biomass and community composition of epinecrotic biofilms compared to natural ones suggest that these communities may be key in estimating a PMSI, and the role of bacteria and fungi in breaking down the carrion could be important in determining PMSI (Merritt and Wallace 2010).
Recent advances in genetic technologies have made investigating successional patterns of bacteria more feasible. The first study to investigate this was conducted in a marine habitat on swine heads and employed cloning technologies (Dickson et al. 2011).
25
Successional patterns were different based on season, and complex communities that formed by the first sampling date (day 3) were not comprised of natural skin biota.
Communities also varied according to tissue type; cheek represented superficial skin, snout represented secretious/membranous skin, and neck represented a wound site. These differences may have forensic applications when selecting a swab site on a corpse. The first study conducted in a stream setting used 454-pyrosequencing to provide taxonomic resolution of the entire bacterial community (Benbow et al. in press). Communities differed based on sampling date indicating successional development, and these patterns were observed in both winter and summer. Genera richness increased throughout decomposition during both seasons, but the overall influence of season was more important than successional changes (Benbow et al. in press). Successional patterns of bacteria appear to be consistent in epinecrotic biofilms but the community composition will be influenced by environmental factors. Further research is needed to investigate how to utilize this pattern in determining a PMSI.
26
Literature Cited
Agrawal, A. A., D. D. Ackerly, F. Adler, A. E. Arnold, C. Cáceres, D. F. Doak, E. Post, P. J. Hudson, J. Maron, K. A. Mooney, M. Power, D. Schemske, J. Stachowicz, S. Sharon, M. G. Turner, and E. Werner. 2007. Filling key gaps in population and community ecology. Frontiers in Ecology and the Environment 5:145–152.
Anderson, G. S., and S. L. VanLaerhoven. 1996. Initial studies on insect succession on carrion in southwestern British Columbia. J Forensic Sci 41:617–625.
Augspurger, C., C. Karwautz, M. MuBmann, H. Daims, and T. J. Battin. 2010. Drivers of bacterial colonization patterns in stream biofilms. FEMS Microbiology Ecology 72:47–57.
Barrios, M., and M. Wolff. 2011. Initial study of arthropods succession and pig carrion decomposition in two freshwater ecosystems in the Colombian Andes. Forensic Science International 212:164–172.
Battin, T. J., L. A. Kaplan, J. D. Newbold, X. Cheng, and C. M. E. Hansen. 2003a. Effects of current velocity on the nascent architecture of stream microbial biofilms. Appl Environ Microbiol 69:5443–5452.
Battin, T. J., L. A. Kaplan, J. D. Newbold, and C. M. E. Hansen. 2003b. Contributions of microbial biofilms to ecosystem processes in stream mesocosms. Nature 426:439– 42.
Benbow, M., J. Pechal, J. Lang, R. Erb, and J. Wallace. (n.d.). The potential of high- throughput metagenomic sequencing of aquatic bacterial communities to estimate the postmortem submersion interval. J. Forensic Sci.
Besemer, K., I. Hödl, G. Singer, and T. J. Battin. 2009. Architectural differentiation reflects bacterial community structure in stream biofilms. ISME J 3:1318–1324.
Besemer, K., G. Singer, R. Limberger, A.-K. K. Chlup, G. Hochedlinger, I. Hödl, C. Baranyi, and T. J. Battin. 2007. Biophysical controls on community succession in stream biofilms. Applied and Environmental Microbiology 73:4966–4974.
Brooker, R. W., R. M. Callaway, L. A. Cavieres, Z. Kikvidze, C. J. Lortie, R. Michalet, F. I. Pugnaire, A. Valiente-Banuet, and T. G. Whitham. 2009. Don’t diss integration: A comment on Ricklefs's disintegrating communities. American Naturalist 174:919– 927.
Cardinale, B. J., M. A. Palmer, C. M. Swan, S. Brooks, N. L. Poff, and F. Collins. 2002. The influence of substrate heterogeneity on biofilm metabolism in a stream ecosystem. Ecology 83:pp. 412–422.
27
Casamatta, D. A., and R. G. Verb. 2000. Algal colonization of submerged carcasses in a mid-order woodland stream. J Forensic Sci 45:1280–1285.
Chaloner, D. T., M. S. Wipfli, and J. P. Caouette. 2002. Mass loss and macroinvertebrate colonisation of Pacific salmon carcasses in south-eastern Alaskan streams. Freshwater Biology 47:263–273.
Clements, F. E. 1916. Plant succession: Analysis of the development of vegetation. Carnegie Institute of Washington, Washington D.C.
Clements, F. E. 1936. Nature and structure of the climax. Journal of Ecology 24:252– 284.
Connell, J. . 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302– 1310.
Costerton, J. W., Z. Lewandowski, D. E. Caldwell, D. R. Korber, and H. Lappin-Scott. 1995. Microbial Biofilms. Annual Review of Microbiology 49:711–745.
Cummins, K. W. 1974. Structure and function of stream ecosystems. BioScience 24:631– 641.
Danger, M., J. Cornut, E. Chauvet, P. Chavez, A. Elger, and A. Lecerf. 2013. Benthic algae stimulate leaf litter decomposition in detritus-based headwater streams: a case of aquatic priming effect? Ecology 94:1604–1613.
Deacon, J. 2006. Fungal Symbiosis. Page 256 Fungal Biology. Blackwell Publishing.
Dickson, G. C., R. T. M. Poulter, E. W. Maas, P. K. Probert, and J. A. Kieser. 2011. Marine bacterial succession as a potential indicator of postmortem submersion interval. Forensic Science International 209:1–10.
Ellison, A. M., E. J. Farnsworth, R. R. Twilley, S. Ecology, N. Dec, and M. E. Aaron. 1996. Facultative mutualism between red mangroves and root-fouling sponges in Belizean mangal. Ecology 77:2431.
Fairchild, G. W., R. L. Lowe, and W. B. Richardson. 1985. Algal periphyton growth on nutrient-diffusing substrates: an in situ bioassay. Ecology 66:465–472.
Giller, P. S., and B. Malmqvist. 1998. The biology of streams and rivers. (M. J. Crawley, C. Little, T. R. E. Southwood, and S. Ulfstrand, Eds.) Biology of Habitats. Oxford University Press Inc., New York.
Greenberg, B. 1991. Flies as forensic indicators. Journal of Medical Entomology:565– 577.
28
Greenwood, M. J., A. R. McIntosh, and J. S. Harding. 2010. Disturbance across an ecosystem boundary drives cannibalism propensity in a riparian consumer. Behavioral Ecology 21:1227–1235.
Gurevitch, J., S. M. Scheiner, and G. A. Fox. 2006. Herbivory and plant-pathogen interactions. Page 257 The Ecology of Plants. Sineaur Associates, Inc., Sunderland, Massachusetts.
Haefner, J. N., J. R. Wallace, and R. W. Merritt. 2004. Pig decomposition in lotic aquatic systems: the potential use of algal growth in establishing a postmortem submersion interval (PMSI). J. Forensic Sci. 49:1–7.
Hardin, G. 1960. The competitive exclusion principle. Science 131:1292–1297.
Harrison, S., and H. Cornell. 2008. Toward a better understanding of the regional causes of local community richness. Ecology Letters 11:969–979.
Haskell, N. H., D. G. McShaffrey, D. A. Hawley, R. E. Williams, and J. E. PLess. 1989. Use of aquatic insects in determining submersion interval. J Forensic Sci 34:622– 632.
Hatcher, M. J., J. T. A. Dick, and A. M. Dunn. 2006. How parasites affect interactions between competitors and predators. Ecology Letters 9:1253–1271.
Heaton, V., A. Lagden, C. Moffatt, and T. Simmons. 2010. Predicting the postmortem submersion interval for human remains recovered from U.K. waterways. J Forensic Sci 55:302–307.
Hoagland, K. D., S. C. Roemer, J. R. Rosowski, and R. Rosowski. 1982. Colonization and community structure of two periphyton assemblages, with emphasis on the diatoms (Bacillariophyceae). American Journal of Botany 69:pp. 188–213.
Hobischak, N. R., and G. S. Anderson. 2002. Time of submergence using aquatic invertebrate succession and decompositional changes. J Forensic Sci 47:142–151.
Hodoki, Y. 2005. Bacteria biofilm encourages algal immigration onto substrata in lotic systems. Hydrobiolgia 539:27–34.
Holomuzki, J. R., J. W. Feminella, and M. E. Power. 2010. Biotic interactions in freshwater benthic habitats. Journal of the North American Benthological Society 29:220–244.
Horn, H. S., M. L. Cody, and J. M. Diamond. 1975. Markovian properties of forest succession. (M. L. and D. Cody J.M., Ed.) Ecology and evolution of communites. Bellnap Press, Cambridge, MA.
29
Huss, M., T. Van Kooten, and L. Persson. 2010. Intra-cohort cannibalism and size bimodality: a balance between hatching synchrony and resource feedbacks. OIKOS 119:2000–2011.
Johnson, N. C., J. H. Graham, and F. A. Smith. 1997. Functioning of mycorrhizal associations along the mutualism-parasitism continuum. New Phytologist 135:pp. 575–586.
Kakizaki, E., S. Kozawa, N. Imamura, T. Uchiyama, S. Nishida, M. Sakai, and N. Yukawa. 2011. Detection of marine and freshwater bacterioplankton in immersed victims: Post-mortem bacterial invasion does not readily occur. Forensic Science International 211:9–18.
Kakizaki, E., K. Takahama, Y. Seo, S. Kozawa, M. Sakai, and N. Yukawa. 2008. Marine bacteria comprise a possible indicator of drowning in seawater. Forensic Science International 176:236–247.
Keiper, J. B., and D. A. Casamatta. 2001. Benthic organisms as forensic indicators. Journal of the North American Benthological Society 20:311–324.
Keiper, J. B., E. G. Chapman, and B. A. Foote. 1997. Midge larvae (Diptera: Chironomidae) as indicators of postmortem submersion interval of carcasses in a woodland stream: a preliminary report. J Forensic Sci 42:1074–1079.
Kohler, S. 1984. Search mechanism of a stream grazer in patchy environments: the role of food abundance. Oecologia 62:209–218.
Lake, P. S. 2000. Disturbance, patchiness, and diversity in streams. Journal of the North American Benthological Society 19:573–592.
Lake, P. S. 2003. Ecological effects of perturbation by drought in flowing waters. Freshwater Biology 48:1161–1172.
Larned, S. T. 2010. A prospectus for periphyton: recent and future ecological research. Journal of the North American Benthological Society 29:182–206.
Lawton, J. H. 1999. Are there general laws in ecology? Oikos 84:177–192.
Lear, G., M. J. Anderson, J. P. Smith, K. Boxen, and G. D. Lewis. 2008. Spatial and temporal heterogeneity of the bacterial communities in stream epilithic biofilms. FEMS Microbiology Ecology 65:463–473.
Levri, E. P. 1999. Parasite-induced change in host behavior of a freshwater snail: parasitic manipulation or byproduct of infection? Behavioral Ecology 10:234–241.
30
Lock, M. A., R. R. Wallace, J. W. Costerton, R. M. Ventullo, and S. E. Charlton. 1984. River epilithon: toward a structural-functional model. OIKOS 42:pp. 10–22.
Lotka, A. J. 1932. The growth of mixed populations: two species competing for a common food supply. Journal of the Washington Academy of Sciences 22:461–469.
Lowe, R. L., and G. D. LaLiberte. 2006. Benthic stream algae: distribution and structure. Pages 327–356 in R. F. Hauer and G. A. Lamerti, editors. Methods in Stream Ecology. Academic Press, Boston.
Merritt, R. W., and K. W. Cummins. 2006. Trophic relationships of macroinvertebrates. Pages 585–609 in F. R. Hauer and G. A. Lamberti, editors. Methods in stream ecology, 2nd edition. Academic Press, London.
Merritt, R. W., K. W. Cummins, and M. Berg. 2008. An introduction to the aquatic insects of North America. Kendall/Hunt Publ. Co., Dubuque, IA.
Merritt, R. W., and J. R. Wallace. 2010. The role of aquatic insects in forensic investigations. Pages 177–222 in J. H. Byrd and J. L. Castner, editors. Forensic Entomology: the utility of arthropods in legal investigations, 2nd edition. CRC Press, Washington D.C.
Patrick, R., and N. A. Roberts. 1979. Diatom communities in the Middle Atlantic States, U.S.A. Nova Hedwigia:265–283.
Payne, J. A. 1965. A summer carrion study of the baby pig Sus scrofa linnaeus. Ecology 46:592–602.
Pechal, J., T. Crippen, and M. Benbow. 2014. The potential use of bacterial community succession in forensics as described by high throughput metagenomic sequencing. Int J Legal Med 128:193–205.
Piette, M. H. A., and E. A. De Letter. 2006. Drowning: still a difficult autopsy diagnosis. Forensic Science International 163:1–9.
Pohlon, E., J. Marxsen, and K. Küsel. 2010. Pioneering bacterial and algal communities and potential extracellular enzyme activities of stream biofilms. FEMS Microbiology Ecology 71:364–373.
Pringle, C. M., R. J. Naiman, G. Bretschko, J. R. Karr, W. Oswood, J. R. Webster, R. L. Welcomme, and M. J. Winterbourn. 1988. Patch dynamics in lotic systems!: the stream as a mosaic. J. N. Am. Benthol. Soc. 7:503–524.
Ricklefs, R. E. 2008. Disintegration of the ecological community. American Naturalist 172:741–750.
31
Rier, S. T., J. M. Shirvinski, and K. C. Kinek. 2014. In situ light and phosphorus manipulations reveal potential role of biofilm algae in enhancing enzyme-mediated decomposition of organic matter in streams. Freshwater Biology 59:1039–1051.
Roxburgh, S. H., K. Shea, and J. B. Wilson. 2004. The intermediate disturbance hypothesis: patch dynamics and mechanisms of species coexistence. Ecology 85:359–371.
Schall, J. J. (n.d.). Parasite-mediated competition in Anolis lizards. 1992 92:64.
Siver, P. A., W. D. Lord, and D. J. McCarthy. 1994. Forensic limnology: the use of freshwater algal community ecology to link suspects to an aquatic crime scene in southern New England. Journal of forensic sciences 39:847.
Staley, J. T., and A. Konopka. 1985. Microoganisms in aquatic and terrestiral habitats. Ann. Rev. Microbiol.:321–46.
Stoodley, P., D. Debeer, and Z. Lewandowski. 1994. Liquid flow in biofilm systems. Applied and Environmental Microbiology 60:2711–2716.
Sutherland, I. W. 2001. Biofilm exopolysaccharides: a strong and sticky framework. Microbiology 147:3–9.
Tank, J. L., and W. K. Dodds. 2003. Nutrient limitation of epilithic and epixylic biofilms in ten North American streams. Freshwater Biology 48:1031–1049.
Thoreau, H. D., and C. L. Flint. 1956. The succession of forest trees. TM, courtesy of Thoreau’s publishers, Houghton, Mifflin Company.
Tomberlin, J. K., and P. H. Adler. 1998. Seasonal colonization and decomposition of rat carrion in water and on land in an open field in South Carolina. Journal of Medical Entomology 35:704–709.
Townsend, C. R. 1989. The patch dynamics concept of stream community ecology. J. N. Am. Benthol. Soc. 8:36.
Vance, G. M., J. K. VanDyk, and W. A. Rowley. 1995. A device for sampling aquatic insects associated with carrion in water. J Forensic Sci 40:479–482.
Volterra, V. 1926. Variations and fluctuations of the numbers of individuals in animal species living together. Animal Ecology.
Wallace, J. R., R. W. Merritt, R. Kimbirauskas, M. E. Benbow, and M. McIntosh. 2008. Caddisflies assist with homicide case: determining a postmortem submersion interval using aquatic insects. J Forensic Sci 53:219–221.
32
Wellnitz, T. A., R. B. Rader, and J. V Ward. 1996. Light and a grazing mayfly shape periphyton in a Rocky Mountain stream. Journal of the North American Benthological Society 15:pp. 496–507.
Zarraonaindia, I., D. Smith, and J. Gilbert. 2013. Beyond the genome: community-level analysis of the microbial world. Biology and Philosophy 28:261–282.
Zimmerman, K. A., and J. R. Wallace. 2008. The potential to determine a postmortem submersion interval based on algal/diatom diversity on decomposing mammalian carcasses in brackish ponds in Delaware. J Forensic Sci 53:935–941.
33
CHAPTER 2
ABIOTIC AUTUMNAL ORGANIC MATTER DEPOSITION AND GRAZING
DISTURBANCE EFFECTS ON EPILITHIC BIOFILM SUCCESSION
Abstract
Stream epilithic biofilm community assembly is influenced in part by environmental factors. Autumn leaf deposition is an annual resource subsidy to streams, but the physical effects of leaves settling on epilithic biofilms has not been investigated.
We hypothesized that bacterial and micro-eukaryotic community assembly would follow a successional sequence that was mediated by abiotic effects that were simulating leaf deposition (reduced light and flow) and by biotic (snail grazing) disturbance. This hypothesis was tested using an in situ experimental manipulation. Ambient biofilms had greater algal biomass and distinct ARISA community profiles compared to biofilms developed under manipulated conditions. There were no significant differences in biofilm characteristics associated with grazing, suggesting that results were driven by reduced light/flow rather than invertebrate disturbance; however, grazing appeared to increase bacterial taxon richness. Interestingly at day 38, all treatments grouped together in ordination space and had similar algal/total biomass ratios. We suggest algal priming promoted a shift in ambient biofilms but that this effect is dependent upon successional
34
timing of algal establishment. These data demonstrate that abiotic effects were more influential than local grazing disturbance and imply that leaf litter deposition may have bottom-up effects on the stream ecosystem through altered epilithic biofilms.
Introduction
Stream epilithic biofilms form on inorganic substrates and are matrix-enclosed microbial communities comprised of bacterial, algal, fungal, and protozoan organisms.
The abundance and diversity of these organisms is dictated by environmental factors including light (Steinman et al. 1991, Roeselers et al. 2007), flow (Arnon et al. 2007,
Besemer et al. 2007, 2009b), nutrients (Olapade and Leff 2005, Ardón and Pringle 2007,
Passy 2008), and invertebrate grazing (Tuchman and Stevenson 1991, Lawrence et al.
2002), which can all interact to drive biofilm community assembly (Rosemond et al.
1993, Wellnitz and Rader 2003, Lange et al. 2011). Understanding how these factors influence biofilms has been facilitated by treating biofilms as micro-ecosystems that follow ecological principles (Battin et al. 2007, Fierer et al. 2010).
Biofilm development is a result of dynamic community interactions and is commonly described using successional patterns. Separately, algae (DeNicola and
McIntire 1990, Wellnitz and Rader 2003, Sekar et al. 2004) and bacteria (Jackson et al.
2001, Besemer et al. 2007) exhibit independent successional sequences, but the algal- bacterial relationship can influence these patterns of community assembly. Indeed, algae have been proposed as “ecosystem engineers” that exert a biophysical control on bacteria because bacterial community composition under different flow regimes became more similar once filamentous algae was established (Besemer et al. 2007). Algae may also
35
influence community composition through a priming effect, where labile algae byproducts facilitate heterotrophic organisms to use complex organic matter; however, this has only been studied in epixylic biofilms (Rier et al. 2014, Kuehn et al. 2014) and has yet to be applied to interactions within epilithic biofilms. In general during epilithic biofilm succession, bacteria dominate the pioneer community (Stock and Ward 1989,
Pohlon et al. 2010) and are the base layer that promotes algal attachment (Hodoki 2005,
Roeselers et al. 2007). Protozoans are present to feed on biofilm organisms, and their densities correlate with diatom density (Kanavillil and Kurissery 2013) and biofilm biomass (Romaní et al. 2014). These micro-eukaryotic organisms include ciliates, flagellates, and amoeba and can have different feeding mechanisms like raptorial or direct interception that affects biofilm morphology and spatial arrangement (Böhme et al.
2009, Dopheide et al. 2011). Protozoan grazing can also influence biofilm community composition (Corno and Jürgens 2008, Wey et al. 2008), but environmental and seasonal differences have been shown to be more influential on composition than micro- eukaryotic grazing (Wey et al. 2012).
Physical disturbances and invertebrate grazing can alter stream biofilm successional trajectories. Specific grazing effects are dependent upon the type and density of the grazer species, but in general, they reduce biofilm biomass and alter community composition and spatial heterogeneity (Feminella and Hawkins 1995, Hillebrand 2008).
If grazing occurs at intermediate levels, biofilm diversity is expected to increase because the biofilm would contain patches of various successional stages as described by the successional mosaic hypothesis (Chesson and Huntly 1997) and intermediate disturbance hypothesis (Connel 1978, Resh et al. 1988). The prediction is that intermittent grazing
36
would allow organisms at early, late, and intermediary stages to be present at the same time; therefore, the diversity would be greater when compared to biofilms without grazing disturbance where the entire “landscape” is at the same successional stage.
Autumn leaf deposition is a naturally occurring seasonal pulse event that transfers nutrients into the aquatic habitat and also physically changes environmental conditions because leaves will reduce light and flow when they settle on the benthos. Light availability (Steinman et al. 1990, Hill and Dimick 2002, Lange et al. 2011) and flow velocity (Biggs and Thomsen 1995, Battin et al. 2003b, Arnon et al. 2007) are important in shaping biofilm community structure and function; algae use light during photosynthesis while flow directly influences physical architecture (Battin et al. 2003b,
Hödl et al. 2014) and therefore community composition (Besemer et al., 2009). Limiting flow also decreases the frequency and duration of flow scouring events at the biofilm scale (Biggs and Thomsen 1995, Hart et al. 2013) and indirectly affects nutrient availability by altering nutrient diffusion and uptake rates (Horner et al. 1990, Larned et al. 2004). This suggests that leaf deposition has the potential to affect stream biofilms in a way that is unrelated to the nutrients provided by the leaves.
Leaf deposition is typically studied within the context of decomposition, organic matter budgets, and ecosystem metabolism (Tank et al. 2010), where the leaves are considered energy inputs and substrates for epixylic biofilms and invertebrates (Tank et al. 2010). Nevertheless, an unaccounted for effect of leaf deposition on stream ecosystem processes could result from reduced light and flow altering the structure and function of epilithic biofilms. These biofilms within stream ecosystems are the predominant source of primary production, integral in nutrient processing, part of the food web, and drive
37
ecosystem processes (Battin et al. 2003a). Therefore, leaf deposition has the potential for bottom-up effects on the stream ecosystem by altering epilithic biofilms, but this idea has not been well studied.
We used an in situ manipulation experiment to assess the physical, abiotic effects of leaf deposition (reduced light and flow) and biotic disturbance (invertebrate grazing) governing succession and biomass development of epilithic biofilms developing during autumn leaf deposition. Biofilms were grown on unglazed ceramic tiles under three conditions: ambient (control – no mesh), inclusion (reduced light/flow from mesh with grazer snails included), and exclusion (reduced light/flow from mesh without snail grazing). Bacterial and eukaryotic community profiles were described using automatic ribosomal intergenic spacer analysis (ARISA) to elucidate patterns of community composition. Next generation 454-pyrosequencing was used to further describe the taxonomic diversity of the bacterial communities. We hypothesized that (H1) treatment effects would be more significant than time in driving community assembly because communities are expected to change over time while treatments would mediate the species pool regardless of time. We also expected (H2) community differences between treatments to be more pronounced later during succession when algal biomass peaks and the effect of reduced light conditions is maximized. In addition, the (H3) effect of abiotic factors was predicted to be more influential than invertebrate grazing disturbances because grazing is a small-scale “local” disturbance compared to the constant and permanent influence of reduced light and flow. Grazing was hypothesized (H4) to drive an increase in bacterial taxon richness along with disrupting biofilm community succession in a way that would result in unchanged community composition over time.
38
This disrupted succession is a result of applying the successional mosaic hypothesis over time where the overall bacterial communities would appear unchanged because grazers would be continuously creating new micro-patches. These patches would generate a mosaic of various stages of succession in the biofilm landscape that were continuously being created, hence, the entire biofilm would not proceed through succession as a unit and the net community composition would remain unchanged.
Methods
Site description
The study was conducted in a lower third order river section of the upper region of the Little Miami River in a small deciduous forest corridor of the Little Miami State
Forest Preserve in Xenia, Ohio, USA (39°76.552 83°90.062). The surrounding landcover of the catchment area was predominately agriculture, and the riparian forest was dominated by maple (Acer sp.) and elm (Ulmus sp.), but included hackberry (Celtis sp.), sycamore (Plantanus sp.), and walnut (Juglans sp.). The stream has been categorized as an Exceptional Warmwater Habitat by the Ohio Environmental Protection Agency because it supported high diversity of aquatic organisms (OEPA 2002). The study was conducted in a 35 x 20 m run habitat where the substrate was predominately gravel and cobble with several intermittent boulders.
Water quality and flow
The following water quality variables were recorded at the start of the study and on every sampling date: temperature (ºC), specific conductivity (SpCond µS/cm), total
39
dissolved solids (TDS mg/L), pH, turbidity (NTU), and dissolved oxygen (DO mg/L).
Measurements were recorded every 30 sec for 10 min at upper, middle, and lower points along the reach using a YSI 6600 v2 Sonde (YSI Inc, Yellow Springs, OH, USA). In addition, water depth (cm) and flow velocity (cm/s2) (FlowTracker® Handheld-ADV®;
SonTek/Xylem, San Diego, CA, USA) were measured at each experimental board (see below).
Experimental design
Epilithic biofilms were allowed to naturally develop on two unglazed ceramic tiles (4.8 x 4.8 x 0.5 cm) attached to brick pavers (19.2 x 9 x 1.3 cm) with 100% silicone under three conditions: ambient (control – no mesh), inclusion (reduced light/flow from mesh netting with grazer snails included), exclusion (reduced light/flow from mesh netting without snail grazing). The ambient condition was subjected to natural light and flow conditions and served as the control. These biofilms represent natural biofilms found on rocks that are prominent in the water column where leaf packs would not form.
Abiotic effects simulating leaf deposition were created with 500 µm nylon mesh netting
(BioQuip Products, CA, USA) that was folded into an enclosure (20 x 10 x 10 cm) and secured by hot glue. The mesh enclosure was attached to the paver with 100% silicone and the resulting effect was reduced light and flow conditions and excluded grazers.
Goniobasis (Elimia) sp. snails (1.5 cm; N = 2) at natural densities were added to the inclusion treatment enclosures through a slit that was sewn closed with fishing line.
Natural snail density was determined on 1 October 2010 by quantifying the number of snails within nine, 50 cm2 areas of benthic substrate throughout the study reach. Using
40
VELCRO® tape, the pavers of each treatment were randomly distributed on wood boards
(1 x 0.3 x 0.025 m) that were anchored to the streambed approximately 3.6 m apart along the thalweg (deepest section of the stream) to ensure as similar flow conditions as possible. Pavers were randomly sampled (N = 3 per treatment) at 14, 24, and 38 days of growth. Tiles were removed, transported in the dark on ice to the lab, and stored at 4°C until processing within 24 h.
The study was conducted from 15 October 2010 to 22 November 2010 and autumnal leaf deposition occurred during the study (Figure 1). Leaf packs that formed against the boards and biofilm material that developed on enclosure structures were gently removed every four days (Figure 1) to prevent the organic matter buildup from interfering with results by creating stagnant and anoxic conditions inside the enclosures.
This did not remove the visually obvious biofilm material inside the enclosures so we feel the setup was still able to reflect physical effects of leaves while still ensuring grazer manipulation. Also, leaves were rarely found on ambient biofilms because there was not a protrusion for leaves to settle against and the slight elevation of the boards in the water column subjected them to flow that was not conducive for leaf deposition.
Biofilm biomass
Biofilm biomass from one tile per paver was removed using a sterile razor blade and toothbrush and suspended in ultrapure water (NANOpure II; Barnstead, Boston, MA,
USA). Two sub samples for total biomass as ash free dry mass (AFDM) and algal- associated biomass as chlorophyll a and were collected on GB-140 glass membrane filters (diameter, 25 mm; pore size, 0.4 µm; Sterlitech, Kent, WA, USA) following
41
techniques used by Steinman et al., (2007). Chlorophyll a filters were stored at -20°C until extraction within three months and were used to determine algal biomass by multiplying the chlorophyll a value by the biomass factor of 67 (APHA 1999). Biofilm from the second tile was removed and stored in 90% ethanol at -20°C until DNA extraction in autumn 2013 to determine community profiles.
DNA extraction
DNA was extracted from 0.1 - 0.3 g of ethanol evaporated, dried biofilm using a combination of methods from Miller et al., (1999) and Zhou et al., (1996) as suggested by Lear et al., (2010). Samples were lysed by bead beating (0.5 g each of 0.1 mm and 0.5 mm glass lysis beads; RPI, Mount Prospect, IL, USA) for 15 min on a horizontal vortex adaptor (MO BIO Laboratories, Carlsbad, CA, USA) at full speed in 1.2 mL of extraction buffer (100 mM Tris-HCl [pH 8.0], 100 mM EDTA disodium salt [pH 8.0], 100mM sodium phosphate [pH 8.0], 1.5 M sodium chloride, 1% CTAB), 12 µL proteinase K (20 mg/mL), and 30 µL SDS (20%). The mixture was incubated at 65°C for 1 hr with gentle end-over-end inversions by hand every 15 m. Then, 4 µL of RNase (100 mg/µL) was added. DNA was isolated from organic debris with chloroform/isoamyl alcohol extraction and was precipitated overnight at -20°C with isopropanol. The mixture was warmed to 37°C to dissolve salt precipitates, and DNA was pelleted at 15,000 g for 30 min. The DNA pellet was washed twice with ice cold 70% EtOH and dissolved in 25-50
µL ultrapure water (NANOpure II; Barnstead, Boston, MA, USA). Samples were purified using PowerClean® DNA Clean-Up Kit (MO BIO Laboratories, Carlsbad, CA, USA)
42
with a modified protocol for low amounts of DNA to remove PCR inhibitors typical of biofilm samples.
ARISA
Bacterial and eukaryotic communities were assessed using profiles created by automated ribosomal intergenic spacer analysis (ARISA) (Fisher and Triplett 1999), which generates a unique ‘fingerprint’ of microbial communities using the 16S-23S intergenic space in bacteria and the ITS1-5.8S-ITS2 region in eukaryotes (Ranjard et al.
2001). While ARISA does not taxonomically identify organisms like next generation sequencing methods, this method generates community profiles and produces similar patterns in results (Bienhold et al. 2012, van Dorst et al. 2014). Approximately 15-20 ng of DNA quantified by spectrophotometer (NanoPhotometerTM Pearl; Denville Scientific
Inc., South Plainfield, NJ, USA) was amplified by PCR using 25 µL GoTaq® Colorless
Master Mix (Promega, Madison, WI, USA) with 0.5 µM of forward and reverse primers.
Bacteria ribosomal intergenic space regions were amplified with primers ITSF (5’-
GTCGTAACAAGGTAGCCGTA-3’) labeled with FAM at the 5’ end (IDT, Coralville,
IA, USA) and ITSReub (5’-GCCAAGGCATCCACC-3’) (Cardinale et al. 2004).
Eukaryote ribosomal intergenic space regions were amplified with primers 2234C (5’-
GTTTCCGTAGGTGAACCTGC-3’) labeled with ATTOTM 550 (IDT, Coralville, IA,
USA) and 3126T (5’- ATATGCTTAAGTTCAGCGGGT-3’) (Ranjard et al. 2001).
While the eukaryotic primer set has typically been used for soil fungal communities
(Ranjard et al. 2001), sequenced clones of fragments from freshwater biofilms revealed it targets various algae and ciliates, and therefore can be used to describe the general
43
eukaryotic community (Fechner et al. 2010). Fragments were created with the following
PCR conditions: (i) 94°C for 3 min, (ii) 35 cycles of 94°C for 1 min, 56°C (57.5°C for eukaryotes) for 1 min, 72°C for 2 min, and finally (iii) 72°C for 10 min (Fechner et al.
2010). Equal volumes of bacterial and eukaryotic PCR products from a sample were combined and sent to DNA Analysis, LLC (Cincinnati, OH, USA) for multiplexed fragment analysis on an ABI 3100 (Life Technologies, Carlsbad, CA, USA). Fragments were interpreted using Genescan v 3.7 using the Local Southern Size Calling Method with a peak height threshold of 100 fluorescence units to remove background fluorescence and formulated using GeneMapper v 2.5 (Life Technologies, Carlsbad, CA,
USA).
Fragment peak length and area was converted to column format using the treeflap
Excel Macro (http://www.wsc.monash.edu.au/~cwalsh/treeflap.xls) and processed with the automatic_binner script to determine binning window size and the interactive_binner script to determine the best starting window position (Ramette 2009) in R v 3.1.0 (R Core
Team 2014). This method was used to account for inherent imprecision of analyzer machines. Peak area was converted to relative abundance of each fragment as part of the entire sample, fragments < 0.09% relative abundance were removed (Ramette 2009), and window size was calculated to be 1.5 base pairs. Within the eukaryotic samples, a 14 day exclusion sample had an excessive number of fragments (5x the average) and so was removed from analysis because it was considered an unsuccessful reaction due to technical artifacts, and a 24 day ambient had a failed result. All bacterial reactions were successful.
44
Pyrosequencing
In order to provide taxonomic descriptions of bacteria in the biofilms in a way to supplement the ARISA community profiles, we employed 454-pyrosequencing by using methods of Pechal et al., (2014) and Benbow et al., (in press) and described briefly here.
Bacterial diversity was determined using amplicon pyrosequencing (Dowd et al. 2008a) at an offsite lab (MR DNA, Shallowater, TX, USA) as previously described (Dowd et al.
2008b, Swanson et al. 2011, Handl et al. 2011). Samples were pooled using equal amounts of DNA from replicates that were quantified using Quant-iT TM PicoGreen® dsDNA assay kit (Invitrogen, Carlsbad, CA, USA). A unidirectional barcoding strategy using the forward primer was utilized to multiplex samples. Universal Eubacterial primers Gray28F (5’TTTGATCNTGGCTCAG) and Gray519r (5’
GTNTTACNGCGGCKGCTG) amplified the bacterial V1-3 regions of the 16S rRNA using HotStarTaq Plus Master Mix Kit (Qiagen, Valencia, CA, USA) with the following single-step PCR protocol: (i) 94°C for 3 min, (ii) 28 cycles of 94°C for 30 sec, 53°C for
40 sec, 72°C for 1 min, and finally (iii) 72°C for 5 min. Amplicon products were purified using Agencourt Ampure beads (Agencourt Bioscience Corporation, MA, USA) and sequenced with Roche 454 FLX titanium instruments and reagents following manufacturer’s guidelines. Sequence data was processed with a proprietary analysis pipeline (www.mrdnalab.com, MR DNA, Shallowater, TX, USA). All barcodes, primers, short sequences (< 200bps), sequences with ambiguous base calls, and sequences with homopolymer runs exceeding 6 bp were removed. Sequences were then denoised and chimeras removed. Operational taxonomic units clustered at 97% similarity were classified using BLASTn against a curated GreenGenes database (DeSantis et al. 2006).
45
The sequences were submitted to the European Nucleotide Archive with the acquisition number PRJEB8214.
Statistical analysis
Biofilm biomass characteristics were analyzed using a two-way repeated measures analysis of variance (ANOVA) with Bonferroni multiple comparisons to test factors of treatment, time, and their interaction and performed within GraphPad Prism v
5.0 (GraphPad Software Inc, San Diego, CA, USA).
ARISA microbial community patterns were visualized using nonmetric multidimensional scaling (NMDS) with Bray-Curtis (Sorensen) distance. NMDS was used because it is a nonparametric multivariate approach useful in evaluating nonlinear relationships of data with high numbers of zeros by using ranked distances (McCune and
Grace 2002). The final configuration was determined following Peck (2010) where
NMDS was run with three axes for bacteria and two axes for eukaryotes, 250 runs with real data, unstability criterion of 0.00001, and 250 Monte Carlo runs of random data. For bacteria, the two axes with the best orthogonality or lowest stress were used for representation. Outliers were identified as samples outside two standard deviations of the mean of average distances of all samples, but were not removed until NMDS confirmation that the sample had a substantial influence on the results (McCune and
Grace 2002). A bacterial 14 day exclusion sample and an eukaryotic 14 day ambient sample were removed as outliers. The effects of time, treatment, and their interaction were tested with nonparametric multivariate analysis of variance (PERMANOVA)
(Anderson 2001) using the adonis function in the vegan 2.2-0 library in R v 3.1.0 (R Core
46
Team 2014). This test was chosen because it takes into account interactions between the time and treatment factors. Multi-response permutation procedure (MRPP) (Zimmerman et al. 1985) was used for pairwise comparisons to determine which groups within a factor were significantly different. This test is sensitive to differences in group location in ordination space and group dispersion, but the betadisper function in the vegan library revealed group dispersions were not significantly different (R Core Team 2014).
Therefore, this analysis was useful in identifying which groups within time and treatment were separated and driving significant PERMANOVA results. Analyses, unless designated otherwise, were conducted using PC-ORD 6 (MjM Software, Gleneden
Beach, OR, USA).
Bacterial communities described by 454-pyrosequencing were analyzed at both the phylum and family level as outlined by Wang et al. (2007) using non-rarefied libraries. Rarefying is argued as being “statistically inadmissible because it requires the omission of available valid data” (McMurdie and Holmes 2014) and so was avoided.
Rarefying is used in analysis to account for samples with different library sizes and the usually associated heteroscedasticity issues. These pitfalls were addressed by applying a
Variance Stabilizing Transformation available in the DESeq2 R package (Love et al.
2014). This approach is advantageous because it retains all sequencing reads and equalizes the variances without increasing the uncertainty, which occurs when randomly eliminating read counts (McMurdie and Holmes 2014). We were especially interested in retaining all data because samples were pooled replicates. Clustering of samples was investigated using group average method cluster analysis with Bray-Curtis distance to determine how samples grouped according to the factors tested. Community composition
47
of abundant taxa was represented with stacked bar graphs using relative abundance (%) created in GraphPad Prism v 5.0 (Graphpad Software Inc., San Diego, CA, USA).
Groups with > 1% of the total abundance were considered abundant taxa (Pascault et al.
2014) and the remaining taxa were grouped into a Rare Taxa category. While rare taxa has been classified as < 0.01% (Galand et al. 2009, Pascault et al. 2014), we feel that it is appropriate for our purposes to describe a category of taxa representing ≤ 1% relative abundance as rare taxa. Alpha diversity measured by richness was further investigated at the family level and analyzed with a two-way ANOVA where samples were grouped by common treatment or time factor. These groups consisted of the three libraries with the designated factor type and were not comprised of true replicates because the nine libraries were created by pooling the DNA sample replicates. Mean and total richness was calculated using the same groups. Total richness was determined by adding the taxa present in any of the three designated communities. Variation within the groups was assessed using Whittaker’s beta diversity (βW) where the smaller the value the more similar the samples are (Whittaker 1972, Anderson et al. 2011).
Results
Study site and water parameters
The study occurred during leaf deposition (Figure 1) where the streambed contained few leaves at the beginning of the experiment, was intermittently covered by
14 days, and was well covered by 24 days (Figure 1, personal observation JML). A brief storm producing very strong winds on day 10 (25 October 2010) facilitated leaf deposition. Throughout the study, temperature significantly decreased, while specific
48
conductivity, total dissolved solids, pH, and dissolved oxygen significantly increased
(ANOVA, P <0.001; Table 1); however, these changes stabilized at 24 days. Depth and flow remained unchanged. Additionally, particulate matter (turbidity) decreased throughout the study while dissolved solids (TDS and conductivity) increased.
Biofilm biomass
Treatment, time, and their interaction had statistically significant effects on algal biomass, total biofilm biomass, and the ratio of algal/total biomass (repeated measures two-way ANOVA, P < 0.05; Table 5), except the interaction effect on algal biomass was only nearly significant (P = 0.068). Algal biomass was reduced in the inclusion and exclusion treatments (both with mesh netting enclosures) over the study period (Figure
2). At 24 days of colonization, the total biofilm biomass was the same across all treatments, but algal biomass peaked under ambient conditions. Interestingly, ambient total biomass at 38 days was significantly greater than enclosure treatments (Bonferroni pairwise comparisons, P < 0.05), and this was coupled with a slight decrease in algal biomass, which resulted in an algal/total biomass ratio that was similar to the enclosure treatments. Exclusion and inclusion treatments were not significantly different, but inclusion levels were lower on days 24 and 38. Grazer effects may not have been detected because of adjusted statistical power associated with low replication and Bonferroni corrections (Figure 2).
49
Biofilm community structure
There were significant treatment and time effects on both bacterial and eukaryotic
ARISA community profiles (PERMANOVA, P < 0.01; Table 2), and bacteria generated
3 axes while eukaryotes generated 2 axes. This could reflect the amount of information available for analyses because more fragments were produced for bacteria than eukaryotes, where bacteria totaled 255 operational taxonomic units (OTUs) with a mean of 44.4 ± 19.8 OTUs per sample and eukaryotes had 79 OTUs with a mean of 11.4 ± 3.9
OTUs per sample. This may have resulted in the bacterial data being more sensitive to patterns and could explain why NMDS ordination resulted in 3 axes for bacteria and only
2 axes were generated for eukaryotes.
Unique microbial communities were associated with both treatment and time as indicated by NMDS ordination (Figure 3) and MRPP (Table 3). Bacterial communities were significantly different between the ambient and enclosure treatments, which were not different from each other. Only ambient and inclusion treatments were significantly different for eukaryotes. Both bacterial and eukaryotic late succession communities at 38 days were significantly (MRPP, P < 0.05; Table 3) different from days 14 and 24 (Figure
3), which were statistically indistinguishable from each other (MRPP, P > 0.05; Table 3).
Bacterial community composition and diversity
Taxon richness determined by 454 pyrosequencing at both phylum and family levels collectively decreased from day 14 to 38, and the inclusion (grazing) treatment had the greatest diversity (Figure 4; Table 4). Grazing was also associated with less variation as the inclusion treatment had the lowest beta diversity (Table 4). Family richness was
50
significantly different (two-way ANOVA, P = 0.0416) among treatments but not through time (two-way ANOVA, P = 0.1510). In relation to phyla, Proteobacteria dominated all biofilm communities, but Verrucomicrobia and the collective group of Rare Taxa were more abundant in the enclosure treatments. Oceanospirillaceae was the dominant family across all treatments and times, but with slightly less apparent dominance on day 38 in both exclusion and inclusion treatments (Figure 4). Ambient biofilms has less family
Rare Taxa than the enclosure treatments, which could reflect levels of alpha diversity because ambient biofilms had the least average and total richness (Table 4). The bacteria communities of ambient treatments were more closely related to each other than to the other treatments, as demonstrated by cluster analysis, suggesting that treatment effects were stronger than time effects (Figure 5). Nonetheless, time was more significant than treatment (PERMANOVA, P < 0.05, Table 2) at the phylum level, and there was not an interaction affect at either phylum or family levels. These data give a glimpse into patterns of bacterial taxonomic structure within epilithic biofilms. Yet the variability cannot be addressed because libraries were pooled replicates, and so this information provides a foundation that can be built upon in future studies.
Discussion
Simulated abiotic effects from leaf deposition on epilithic biofilm succession
Abiotic effects from inclusion and exclusion treatments (reduced light/flow) were predicted to have strong effects on communities that were magnified through time resulting in diverging successional trajectories from ambient treatments. But the last date of bacterial and eukaryotic ARISA communities grouped together in multidimensional
51
space (Figure 3) rather than being distinctly separated on different trajectories as predicted (H2). Time was more significant based on PERMANOVA and MRPP results of both ARISA and pyrosequenced communities, but treatment was still a significant factor in influencing communities and created distinguishable clusters in cluster analysis of pyrosequenced bacterial communities. Both factors were important in driving community structure, and treatment was not more influential as predicted (H1). It is possible the changing environmental parameters masked some of the treatment effects by creating different species pools throughout the study, but conditions stabilized by 24 days and were not different at day 38. This grouping at day 38 was a robust pattern observed in both bacteria and eukaryotic profiles, and the synchrony suggests that the entire micro- ecosystem was changing.
The clustering of community profiles coupled with the decrease in the ambient algal/biomass ratio indicates that the ambient biofilms became more similar to the enclosure biofilms at day 38, even though algal levels were still elevated. Algal establishment was expected to fuel bacterial nutrient requirements (Sobczak 1996) and select communities that reflected this relationship (Kalscheur et al. 2012), but the shifts at
38 days suggest otherwise. It is possible nutrient enrichment from an influx of various forms of organic matter during autumn through leaf breakdown (Gessner et al. 1999) released bacteria from depending on algae (Scott et al. 2008, Lang et al. 2012), but this would mean algae presence was unimportant and does not explain why the shift occurred after algae become established. It appears algae was the catalyst for the heterotrophic shift, and it has been reported that algal presence increases enzyme activity within biofilms (Romaní and Sabater 2000, Ylla et al. 2009, Rier et al. 2014) and decomposition
52
rates of leaf litter (Danger et al. 2013, Kuehn et al. 2014). Algae could promote biochemical interactions with heterotrophic organisms by creating a structure that facilitates the incorporation of dissolved and particular organic matter through micro- eddies (Augspurger et al. 2010), oscillating streamers (Besemer et al., 2009), and a sticky, anionic matrix (Freeman et al. 1995, Sutherland 2001). Algae also influence the extracellular enzymes in the matrix (Rier et al. 2007), which could impact nutrient cycling (Mulholland et al. 1991, Ziegler and Lyon 2010) and coexistence (Carr et al.
2005) and promote bacteria to rely less on algal derived carbon (Ziegler and Lyon 2010).
A combination of these factors may have shifted ambient biofilms toward dominance by heterotrophic organisms late in succession when mature biofilms contained an established algal-bacterial relationship.
Algal-bacterial relationship during epilithic biofilm succession
Relationships among biofilm organisms are complex, and the results of this study suggest community assembly was influenced by a mechanism that promoted a heterotrophic shift later in succession. We suggest that the increased total biomass coupled with a similar algal/total biomass ratio at day 38 in the ambient treatment is a result of a priming effect of algae (Danger et al. 2013, Kuehn et al. 2014). Algae is known to support bacterial colonization through increased substrate area (Rier and
Stevenson 2001) and niche space (Besemer et al. 2007), but a priming effect results when labile organic matter facilitates the use of recalcitrant organic matter (Guenet et al. 2010).
This means that the algal exudates utilized by bacteria (Haack and McFeters 1982) would facilitate the breakdown of complex organic matter, like nutrients derived from leaf
53
material. But this cannot occur until algae is established, which suggests a successional timing component that includes an algal biomass threshold (Findlay et al. 1993, Sobczak
1996). Studies have demonstrated that the correlation between bacterial and algal biomass was not observed when algal levels were < 5 µg chlorophyll a/cm2 (Rier and
Stevenson 2001). In our study, this threshold was calculated to be at 0.33 mg/cm2 of algal biomass by multiplying 5 µg chlorophyll a/cm2 by the biomass factor of 67 (APHA
1999), and only the ambient treatment at 24 days was clearly above this level (Figure 2).
We suggest algal priming as a mechanism that propels epilithic biofilm succession in a heterotrophic direction when sufficient recalcitrant nutrients are available (Lyon and
Ziegler 2009, Ziegler and Lyon 2010) and once algae have been established.
Physical abiotic vs. biotic disturbance factors on epilithic biofilm community succession
We hypothesized that the abiotic effects of mesh netting enclosures would be more influential than the “local” grazing disturbance on epilithic biofilm succession (H3) because it was a constant and permanent treatment effect. In general, enclosure communities were separated from ambient treatments in ordination space and had less biomass than ambient treatments. There were no statistical differences between inclusion and exclusion biomass levels; though, this may be due to low statistical power from low replication and Bonferroni correction. The Bonferroni correction used for post hoc tests is very conservative (Perneger 1998), but using a Holm correction (Holm 1979) did not give different results either. Overall, this suggests that the reduced light and flow effects masked the influence of grazing.
54
Grazing resulted in increased richness and less variation of 454 pyrosequenced bacterial communities (Figure 4) as was predicted (H4) by the successional mosaic hypothesis; however, this was not statistically tested because samples were pooled.
Grazing likely created patches of various successional stages that prevented the entire community from progressing along the same successional sequence because the inclusion treatment samples appear visually similar (Figure 4) and the variation assessed through beta diversity (βW) was the lowest (Table 4). Results indicate that grazing had the predicted effect, and few studies have investigated the relationship between bacteria of stream biofilms and invertebrate grazers (Carman and Guckert 1994, Alfaro et al. 2007).
With the continuous improvement of next generation sequencing methods, there is an opportunity to further explore this relationship.
Conclusion and future directions
Our results demonstrate that the abiotic factors associated with the physical effects of autumnal leaf deposition are more influential than local invertebrate grazing disturbances on epilithic biofilm community composition and succession in a temperate stream. This implies that leaf deposition includes significant effects that are unaccounted for when leaf deposition is treated strictly as a nutrient addition. The impact would be bottom-up effects on ecosystem processes as a result of reduced light and flow altering the structure and function of epilithic biofilms. There is also the interaction between epilithic biofilms and the epixylic biofilms on the leaves that needs to be considered. This interaction could further alter epilithic biofilm community composition and function with a possible role of fungi in these biofilms (Rier et al. 2007, Artigas et al. 2009, Miura and
55
Urabe 2014). Additionally, we hypothesize that algal priming within light-exposed epilithic biofilms promoted the heterotrophic community to utilize organic matter. This idea is speculative and requires further testing. Future work that focuses on elucidating these patterns are needed to determine if a priming effect could be a general mechanism of succession or if it relies upon a large amount of available organic matter.
Funding
This work was supported in part by the University of Dayton Office of Graduate
Professional and Continuing Education through the Graduate Student Summer
Fellowship Program awarded to JML and through discretionary funds of the Department of Biology at the University of Dayton.
Acknowledgements
We are grateful to A. Lewis for support in the field and J. Pechal for assistance in extractions, analyses, and manuscript preparation. We are also thankful for the helpful and thoughtful comments provided by three anonymous reviewers.
Conflict of Interest
The authors declare no conflict of interest.
56
Literature Cited
Alfaro, A. C., S. E. Dewas, and F. Thomas. 2007. Food and habitat partitioning in grazing snails (Turbo smaragdus), Northern New Zealand. Estuaries and Coasts 30:431–440.
Anderson, M. J. 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecology 26:32–46.
Anderson, M. J., T. O. Crist, J. M. Chase, M. Vellend, B. D. Inouye, A. L. Freestone, N. J. Sanders, H. V Cornell, L. S. Comita, K. F. Davies, S. P. Harrison, N. J. B. Kraft, J. C. Stegen, and N. G. Swenson. 2011. Navigating the multiple meanings of β diversity: a roadmap for the practicing ecologist. Ecology letters 14:19–28.
APHA. 1999. Standard methods for the evaluation of water and wastewater, 20th edition. American Public Health Association, Washington, DC.
Ardón, M., and C. M. Pringle. 2007. The quality of organic matter mediates the response of heterotrophic biofilms to phosphorus enrichment of the water column and substratum. Freshwater Biology 52:1762–1772.
Arnon, S., A. I. Packman, C. G. Peterson, and K. A. Gray. 2007. Effects of overlying velocity on periphyton structure and denitrification. Journal of Geophysical Research 112:G01002.
Artigas, J., A. M. Romaní, A. Gaudes, I. Muñoz, and S. Sabater. 2009. Organic matter availability structures microbial biomass and activity in a Mediterranean stream. Freshwater Biology 54:2025–2036.
Augspurger, C., C. Karwautz, M. MuBmann, H. Daims, and T. J. Battin. 2010. Drivers of bacterial colonization patterns in stream biofilms. FEMS Microbiology Ecology 72:47–57.
Battin, T. J., L. A. Kaplan, J. D. Newbold, X. Cheng, and C. M. E. Hansen. 2003a. Effects of current velocity on the nascent architecture of stream microbial biofilms. Appl Environ Microbiol 69:5443–5452.
Battin, T. J., L. A. Kaplan, J. D. Newbold, and C. M. E. Hansen. 2003b. Contributions of microbial biofilms to ecosystem processes in stream mesocosms. Nature 426:439– 42.
Battin, T. J., W. T. Sloan, S. Kjelleberg, H. Daims, I. M. Head, T. P. Cutis, and L. Eberl. 2007. Microbial landscapes: new paths to biofilm research. Nature Reviews: Microbiology 5:76–81.
57
Benbow, M., J. Pechal, J. Lang, R. Erb, and J. Wallace. (n.d.). The potential of high- throughput metagenomic sequencing of aquatic bacterial communities to estimate the postmortem submersion interval. J. Forensic Sci.
Besemer, K., I. Hödl, G. Singer, and T. J. Battin. 2009a. Architectural differentiation reflects bacterial community structure in stream biofilms. ISME J 3:1318–1324.
Besemer, K., G. Singer, I. Hödl, and T. J. Battin. 2009b. Bacterial community composition of stream biofilms in spatially variable-flow environments. Appl Environ Microbiol 75:7189–7195.
Besemer, K., G. Singer, R. Limberger, A.-K. K. Chlup, G. Hochedlinger, I. Hödl, C. Baranyi, and T. J. Battin. 2007. Biophysical controls on community succession in stream biofilms. Applied and Environmental Microbiology 73:4966–4974.
Bienhold, C., A. Boetius, and A. Ramette. 2012. The energy-diversity relationship of complex bacterial communities in Arctic deep-sea sediments. ISME J 6:724–732.
Biggs, B. J. F., and H. A. Thomsen. 1995. Disturbance of stream periphyton by perturbations in shear stress time to structural failure and differences in community resistance. Journal of Phycology 31:233–241.
Böhme, A., U. Risse-Buhl, and K. Kusel. 2009. Protists with different feeding modes change biofilm morphology. FEMS Microbiology Ecology 69:158–169.
Cardinale, M., L. Brusetti, P. Quatrini, S. Borin, A. M. Puglia, A. Rizzi, E. Zanardini, C. Sorlini, C. Corselli, and D. Daffonchio. 2004. Comparison of different primer sets for use in automated ribosomal intergenic spacer analysis of complex bacterial communities. Applied and Environmental Microbiology 70:6147–6156.
Carman, K. R., and J. B. Guckert. 1994. Radiotracer determination of ingestion and assimilation of periphytic algae, bacteria, and adsorbed amino acids by snails. Journal of the North American Benthological Society 13:80–88.
Carr, G. M., A. Morin, and P. A. Chambers. 2005. Bacteria and algae in stream periphyton along a nutrient gradient. Freshwater Biology 50:1337–1350.
Chesson, P., and N. Huntly. 1997. The roles of harsh and fluctuating conditions in the dynamics of ecological communities. The American Naturalist 150:519–553.
Connel, J. H. 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302– 1310.
Corno, G., and K. Jürgens. 2008. Structural and functional patterns of bacterial communities in response to protist predation along an experimental productivity gradient. Environmental Microbiology 10:2857–2871.
58
Danger, M., J. Cornut, E. Chauvet, P. Chavez, A. Elger, and A. Lecerf. 2013. Benthic algae stimulate leaf litter decomposition in detritus-based headwater streams: a case of aquatic priming effect? Ecology 94:1604–1613.
DeNicola, D. M., and C. D. McIntire. 1990. Effects of substrate relief on the distribution of periphyton in laboratory streams. II. Interactions with irradiance. Journal of Phycology 26:624–633.
DeSantis, T. Z., P. Hugenholtz, N. Larsen, M. Rojas, E. L. Brodie, K. Keller, T. Huber, D. Dalevi, P. Hu, and G. L. Andersen. 2006. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Applied and environmental microbiology 72:5069–72.
Dopheide, A., G. Lear, R. Stott, and G. Lewis. 2011. Preferential feeding by the ciliates Chilodonella and Tetrahymena spp. and effects of these protozoa on bacterial biofilm structure and composition. Applied and Environmental Microbiology 77:4564–72.
Van Dorst, J., A. Bissett, A. S. Palmer, M. Brown, I. Snape, J. S. Stark, B. Raymond, J. McKinlay, M. Ji, T. Winsley, and B. C. Ferrari. 2014. Community fingerprinting in a sequencing world. FEMS Mcrobiology Ecology 89:316–330.
Dowd, S. E., Y. Sun, and P. R. Secor. 2008a. Survey of bacterial diversity in chronic wounds using Pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiology 8:43.
Dowd, S. E., Y. Sun, P. R. Secor, R. D. Wolcott, A. Domingo, and J. A. Carroll. 2008b. Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: bacterial diversity in the ileum of newly weaned Salmonella infected pigs. Foodborne Pathogens and Disease 5:1–14.
Fechner, L. C., F. Vincent-Hubert, P. Gaubert, T. Bouchez, C. Gourlay-Francé, and M.- H. Tusseau-Vuillemin. 2010. Combined eukaryotic and bacterial community fingerprinting of natural freshwater biofilms using automated ribosomal intergenic spacer analysis. FEMS microbiology ecology 74:542–53.
Feminella, J. W., and C. P. Hawkins. 1995. Interactions between stream herbivores and periphyton: a quantitative analysis of past experiments. J. N. Am. Benthol. Soc. 14:465–509.
Fierer, N., D. Nemergut, R. Knight, and J. M. Craine. 2010. Changes through time: integrating microorganisms into the study of succession. Research in Microbiology 161:635–642.
Findlay, S., K. Howe, and D. Fontvielle. 1993. Bacterial-algal relationships in streams of the Hubbard Brook experimental forest. Ecology 74:2326–2336.
59
Fisher, M. M., and E. W. Triplett. 1999. Automated approach for ribosomal intergenic spacer analysis of microbial diversity and its application to freshwater bacterial communities. Applied and Environmental Microbiology 65:4630–4636.
Freeman, C., P. J. Chapman, K. Gilman, M. A. Lock, B. Reynolds, and H. S. Wheater. 1995. Ion exchange mechanisms and the entrapment of nutrients by river biofilms. Hydrobiologia:61–65.
Galand, P. E., E. O. Casamayor, D. L. Kirchman, and C. Lovejoy. 2009. Ecology of the rare microbial biosphere of the Arctic Ocean. Proceedings of the National Academy of Sciences of the United States of America 106:22427–32.
Gessner, M. O., E. Chauvet, M. Dobson, S. Oikos, and F. May. 1999. A perspective on leaf litter breakdown in streams. OIKOS 85:377–384.
Guenet, B., M. Danger, L. Abbadie, and G. Lacroix. 2010. Priming effect: bridging the gap between terrestrial and aquatic ecology. Ecology 91:2850–2861.
Haack, T. K., and G. A. McFeters. 1982. Nutritional relationships among microorganisms in an epilithic biofilm community. Microbial Ecology 8:115–26.
Handl, S., S. E. Dowd, J. F. Garcia-Mazcorro, J. M. Steiner, and J. S. Suchodolski. 2011. Massive parallel 16S rRNA gene pyrosequencing reveals highly diverse fecal bacterial and fungal communities in healthy dogs and cats. FEMS Mcrobiology Ecology 76:301–10.
Hart, D. D., B. J. F. Biggs, V. I. Nikora, and C. A. Flinders. 2013. Flow effects on periphyton patches and their ecological consequences in a New Zealand river. Freshwater Biology 58:1588–1602.
Hill, W., and S. Dimick. 2002. Effects of riparian leaf dynamics on periphyton photosynthesis and light utilisation efficiency. Freshwater Biology:1245–1256.
Hillebrand, H. 2008. Grazing regulates the spatial variability of periphyton biomass. Ecology 89:165–173.
Hödl, I., L. Mari, E. Bertuzzo, S. Suweis, K. Besemer, A. Rinaldo, and T. J. Battin. 2014. Biophysical controls on cluster dynamics and architectural differentiation of microbial biofilms in contrasting flow environments. Environmental Microbiology 16:802–12.
Hodoki, Y. 2005. Bacteria biofilm encourages algal immigration onto substrata in lotic systems. Hydrobiolgia 539:27–34.
Holm, S. 1979. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics 6:65–70.
60
Horner, R. R., E. B. Welch, M. R. Seeley, and J. M. Jacoby. 1990. Responses of periphyton to changes in current velocity, suspended sediment and phosphorus concentration. Freshwater Biology 24:215–232.
Jackson, C. R., P. F. Churchill, and E. E. Roden. 2001. Successional changes in bacterial assemblage structure during epilithic biofilm development. Ecology 82:555–566.
Kalscheur, K. N., M. Rojas, C. G. Peterson, J. J. Kelly, and K. A. Gray. 2012. Algal exudates and stream organic matter influence the structure and function of denitrifying bacterial communities. Microbial Ecology 64:881–892.
Kanavillil, N., and S. Kurissery. 2013. Dynamics of grazing protozoa follow that of microalgae in natural biofilm communities. Hydrobiologia 718:93–107.
Kuehn, K. A., S. N. Francoeur, R. H. Findlay, and R. K. Neely. 2014. Priming in the microbial landscape: periphytic algal stimulation of litter-associated microbial decomposers. Ecology 95:749–762.
Lang, D. A., R. S. King, and J. T. Scott. 2012. Divergent responses of biomass and enzyme activities suggest differential nutrient limitation in stream periphyton. Freshwater Science 31:1096–1104.
Lange, K., A. Liess, J. J. Piggott, C. R. Townsend, and C. D. Matthaei. 2011. Light, nutrients and grazing interact to determine stream diatom community composition and functional group structure. Freshwater Biology 56:264–278.
Larned, S. T., V. I. Nikora, and B. J. F. Biggs. 2004. Mass-transfer-limited nitrogen and phosphorus uptake by stream periphyton: A conceptual model and experimental evidence. Limnol. Oceanogr. 49:1992–2000.
Lawrence, J. R., B. Scharf, G. Packroff, T. R. Neu, and J. R. Lawerence. 2002. Microscale evaluation of the effects of grazing by invertebrates with contrasting feeding modes on river biofilm architecture and composition. Microbial Ecology 44:pp. 199–207.
Lear, G., Y. Dong, and G. Lewis. 2010. Comparison of methods for the extraction of DNA from stream epilithic biofilms. Antonie van Leeuwenhoek 98:567–571.
Love, M. I., W. Huber, and S. Anders. 2014. Moderated estimation of fold change and dispersion for RNA-Seq data with DESeq2. bioRxiv.
Lyon, D. R., and S. E. Ziegler. 2009. Carbon cycling within epilithic biofilm communities across a nutrient gradient of headwater streams. Limnology and Oceanography 54:4.
61
McCune, B., and J. B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR.
McMurdie, P. J., and S. Holmes. 2014. Waste not, want not: Why rarefying microbiome data is inadmissible. PLoS Comput Biol 10:e1003531.
Miller, D. N., J. E. Bryant, E. L. Madsen, and W. C. Ghiorse. 1999. Evaluation and optimization of DNA extraction and purification procedures for soil and sediment samples. Applied and Environmental Microbiology 65:4715–4724.
Miura, A., and J. Urabe. 2014. Spatial and seasonal changes in species diversity of epilithic fungi along environmental gradients of a river. Freshwater Biology.
Mulholland, P. J., A. D. Steinman, A. V Palumbo, J. W. Elwood, and D. B. Kirschtel. 1991. Role of nutrient cycling and herbivory in regulating periphyton communities in laboratory streams. Ecology 72:966–982.
OEPA. 2002. Total maximum daily loads for the upper Little Miami River. Columbus.
Olapade, O. A., and L. G. Leff. 2005. Seasonal response of stream biofilm communities to dissolved organic matter and nutrient enrichments. Applied and Environmental Microbiology 71:2278–2287.
Pascault, N., S. Roux, J. Artigas, S. Pesce, J. Leloup, R. D. Tadonleke, D. Debroas, A. Bouchez, and J.-F. Humbert. 2014. A high-throughput sequencing ecotoxicology study of freshwater bacterial communities and their responses to tebuconazole. FEMS microbiology ecology 90:563–574.
Passy, S. I. 2008. Continental diatom biodiversity in stream benthos declines as more nutrients become limiting. Proceedings of the National Academy of Sciences of the United States of America 105:9663–7.
Pechal, J., T. Crippen, and M. Benbow. 2014. The potential use of bacterial community succession in forensics as described by high throughput metagenomic sequencing. Int J Legal Med 128:193–205.
Peck, J. E. 2010. Multivariate analysis for community ecologists: Step-by-step using PC- ORD. MjM Software Design, Gleneden Beach, OR.
Perneger, T. V. 1998. What’s wrong with Bonferroni adjustments. BMJ 316:1236–1238.
Pohlon, E., J. Marxsen, and K. Küsel. 2010. Pioneering bacterial and algal communities and potential extracellular enzyme activities of stream biofilms. FEMS Microbiology Ecology 71:364–373.
62
Ramette, A. 2009. Quantitative community fingerprinting methods for estimating the abundance of operational taxonomic units in natural microbial communities. Applied and Environmental Microbiology 75:2495–505.
Ranjard, L., F. Poly, J. C. Lata, C. Mougel, J. Thioulouse, and S. Nazaret. 2001. Characterization of bacterial and fungal soil communities by automated ribosomal intergenic spacer analysis fingerprints: biological and methodological variability. Applied and Environmental Microbiology 67:4479–4487.
Resh, V., A. Brown, A. Covich, and M. Gurtz. 1988. The role of disturbance in stream ecology. J. N. Am. Benthol. Soc. 7:433–455.
Rier, S. T., K. A. Kuehn, and S. N. Francoeur. 2007. Algal regulation of extracellular enzyme activity in stream microbial communities associated with inert substrata and detritus. J N Am Benthol Soc 26:439–449.
Rier, S. T., J. M. Shirvinski, and K. C. Kinek. 2014. In situ light and phosphorus manipulations reveal potential role of biofilm algae in enhancing enzyme-mediated decomposition of organic matter in streams. Freshwater Biology 59:1039–1051.
Rier, S. T., and R. J. Stevenson. 2001. Relation of environmental factors to density of epilithic lotic bacteria in 2 ecoregions. Journal of the North American Benthological Society 20:520–532.
Roeselers, G., M. C. M. van Loosdrecht, and G. Muyzer. 2007. Heterotrophic pioneers facilitate phototrophic biofilm development. Microbial Ecology 54:578–585.
Romaní, A. M., C. M. Borrego, V. Díaz-Villanueva, A. Freixa, F. Gich, and I. Ylla. 2014. Shifts in microbial community structure and function in light and dark grown biofilms driven by warming. Environmental Microbiology 16:2550–2567.
Romaní, A. M., and S. Sabater. 2000. Influence of algal biomass on extracellular enzyme activity in river biofilms. Microbial Ecology 40:16–24.
Rosemond, A. D., P. J. Mulholland, and J. W. Elwood. 1993. Top-down and bottom-up control of stream periphyton: effects of nutrients and herbivores. Ecology 74:1264– 1280.
Scott, J. T., J. A. Back, J. M. Taylor, and R. S. King. 2008. Does nutrient enrichment decouple algal–bacterial production in periphyton? Journal of the North American Benthological Society 27:332–344.
Sekar, R., V. Venugopalan, K. Nandakumar, K. Nair, and V. Rao. 2004. Early stages of biofilm succession in a lentic freshwater environment. Hydrobiologia 512:97–108.
63
Sobczak, W. V. 1996. Epilithic bacterial responses to variations in algal biomass and labile dissolved organic carbon during biofilm colonization. J. N. Am. Benthol. Soc. 15:143–154.
Steinman, A. D., G. A. Lamberti, and L. P.R. 2007. Biomass and pigments of benthic algae. Pages 357–380 in F. R. Hauer and G. A. Lamberti, editors. Methods in Stream Ecology, 2nd edition. Academic Press, New York.
Steinman, A., P. J. Mulholland, A. V Palumbo, T. F. Flum, and D. DeAngelis. 1991. Resilience of lotic ecosystems to a light-elimination disturbance. Ecology 72:1299– 1313.
Steinman, A., P. J. Mulholland, A. V Palumbo, T. F. Flum, J. W. Elwood, and D. L. DeAngelis. 1990. Resistance of lotic ecosystems to a light elimination disturbance: a laboratory stream study. OIKOS 58:80–90.
Stock, M. S., and A. K. Ward. 1989. Establishment of a bedrock epilithic community in a small stream: microbial (algal and bacterial) metabolism and physical structure. Canadian Journal of Fisheries and Aquatic Sciences 46:1874–1883.
Sutherland, I. W. 2001. The biofilm matrix – an immobilized but dynamic microbial environment. Trends in Microbiology 9:222–227.
Swanson, K. S., S. E. Dowd, J. S. Suchodolski, I. S. Middelbos, B. M. Vester, K. A. Barry, K. E. Nelson, M. Torralba, B. Henrissat, P. M. Coutinho, I. K. O. Cann, B. A. White, and G. C. Fahey. 2011. Phylogenetic and gene-centric metagenomics of the canine intestinal microbiome reveals similarities with humans and mice. The ISME Journal 5:639–49.
Tank, J. L., E. J. Rosi-Marshall, N. A. Griffiths, S. A. Entrekin, and M. L. Stephen. 2010. A review of allochthonous organic matter dynamics and metabolism in streams. Journal of the North American Benthological Society 29:118–146.
Tuchman, N. C., and R. J. Stevenson. 1991. Effects of selective grazing by snails on benthic algal succession. Journal of the North American Benthological Society 10:430–443.
Wang, Q., G. M. Garrity, J. M. Tiedje, and J. R. Cole. 2007. Naïve Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and Environmental Microbiology 73:5261–5267.
Wellnitz, T., and R. B. Rader. 2003. Mechanisms influencing community composition and succession in mountain stream periphyton: interactions between scouring history, grazing, and irradiance. Journal of the North American Benthological Society 22:528–541.
64
Wey, J. K., K. Jürgens, and M. Weitere. 2012. Seasonal and successional influences on bacterial community composition exceed that of protozoan grazing in river biofilms. Applied and Environmental Microbiology 78:2013–24.
Wey, J. K., A. Scherwass, H. Norf, H. Arndt, and M. Weitere. 2008. Effects of protozoan grazing within river biofilms under semi-natural conditions. Aquatic Microbial Ecology 52:283–296.
Whittaker, R. H. 1972. Evolution and measurement of species diversity. Taxon 21:213– 251.
Ylla, I., C. Borrego, A. M. Romaní, and S. Sabater. 2009. Availability of glucose and light modulates the structure and function of a microbial biofilm. FEMS Microbiology Ecology 69:27–42.
Zhou, J., M. A. Bruns, and J. M. Tiedje. 1996. DNA recovery from soils of diverse composition. Applied and Environmental Microbiology 62:316–322.
Ziegler, S. E., and D. R. Lyon. 2010. Factors regulating epilithic biofilm carbon cycling and release with nutrient enrichment in headwater streams. Hydrobiologia 657:71– 88.
Zimmerman, G. M., H. Goetz, and P. W. Mielke. 1985. Use of an improved statistical method for group comparisons to study effects of prairie fire. Ecology 66:606–611.
65
Tables
Table 1. Water characteristics of temperature (ºC), specific conductivity (SpCond µS/cm), total dissolved solids (TDS mg/L), pH, turbidity (NTU), dissolved oxygen (mg/L), depth (cm), and flow velocity (cm/s2) were measured (N = 3) throughout the study.
Asterisks denote significance at P < 0.001 (ANOVA).
Time Temp * SpCond * TDS * pH * Turb DO * Depth Flow (Days) (ºC) (µS/cm) (mg/L) (NTU) (mg/L) (cm) (cm/s2)
0 15.38 ± 0.02 666.8 ± 0.3 433.0 ± 0.0 8.42 ± 0.02 1.45 ± 0.22 11.22 ± 0.03 45.8 ± 5.8 12.5 ± 1.4
14 10.31 ± 0.02 670.0 ± 0.4 435.6 ± 0.2 8.49 ± 0.02 0.80 ± 0.35 11.64 ± 0.07 46.5 ± 5.6 9.6 ± 2.8
24 5.76 ± 0.05 679.2 ± 0.5 441.5 ± 0.2 8.80 ± 0.03 0.74 ± 0.54 15.25 ± 0.16 46.3 ± 5.6 9.0 ± 1.7
38 5.76 ± 0.05 679.1 ± 0.5 441.5 ± 0.2 8.80 ± 0.03 0.75 ± 0.52 15.27 ± 0.07 45.5 ± 5.7 12.9 ± 5.2
Average 9.3 ± 4.58 673.8 ± 6.3 438.0 ± 4.2 8.63 ± 0.21 0.94 ± 0.34 13.35 ± 2.22 46.0 ± 0.5 10.5 ± 1.4
66
Table 2. The influences of time, treatment, and their interaction on ARISA and pyrosequenced communities were assessed with
PERMANOVA using Bray-Curtis distance measure and 9999 permutations. ARISA data are comprised of true replicates (N= 3) but pyrosequencing data are not because libraries were created by pooling the replicates. Groups for pyrosequencing data included all samples that had a common time or treatment level (N = 3). Significant P-values at < 0.05 are bolded.
!! ARISA Pyrosequencing Bacteria Eukaryotes Phylum Family d.f. pseudo-F P psuedo-F P pseudo-F P psuedo-F P
Time 1 9.646 0.0001 22.106 0.0001 7.53 0.0004 1.884 0.0493
Treatment 2 4.964 0.0009 3.018 0.0174 2.529 0.0509 1.733 0.0433
Interaction 2 2.124 0.0432 1.584 0.1450 0.8461 0.6009 2.553 0.8141
67
Table 3. ARISA communities (N = 3) were assessed with MRPP using Bray-Curtis distance measure to determine pairwise comparisons within time and treatment. MRPP reports a test statistic (T) where the more negative the value the more separation between groups; a chance-corrected within-group agreement (A) where if within group heterogeneity equals that by chance, A is 0, and if all samples are identical, A is 1; and a P-value. Significant values are bolded where P < 0.05.
Bacteria Eukaryotes T A P T A P Time Overall -6.80 0.113 0.0000 -7.75 0.211 0.0000
14 vs. 24 -0.69 0.012 0.2068 -0.39 0.009 0.2949
14 vs. 38 -7.93 0.136 0.0000 -8.12 0.241 0.0001
24 vs. 38 -5.92 0.107 0.0004 -7.36 0.215 0.0001
Treatment Overall -3.68 0.061 0.0040 -1.21 0.033 0.1162
A vs. I -3.72 0.062 0.0052 -2.39 0.036 0.0326
A vs. E -4.28 0.079 0.0023 -1.30 0.037 0.1009
I vs. E 0.28 -0.006 0.5092 0.94 -0.033 1.0000
68
Table 4. Alpha and beta diversity measures of the family taxonomic level of bacterial 454-pyrosequenced communities are separated into treatment and time categories. The libraries are composites of replicate samples, and therefore there are no true replicates in the analyses. Measures were calculated with libraries that had the common treatment or time category (N = 3). Alpha diversity is represented by richness, and total richness was determined by additively counting the taxa present within the three samples. Variation within the groups was assessed by Wittaker’s beta diversity (βW).
Treatment Time (Days) Ambient Inclusion Exclusion 14 24 28
Average Richness 93.3 ± 24.8 131.7 ± 11.9 121.3 ± 5.1 129.7 ± 9.3 111.3 ± 28.9 105.3 ± 23.7
Total Richness 132 171 158 167 158 144
Beta Diversity (βW) 2.04 1.44 1.57 1.47 1.71 1.8
69
Table 5. Biofilm characteristics of algal biomass (chlorophyll a x 67), total biomass (AFDM), and the ratio of algal/total biomass were analyzed (N = 3) by a two-way repeated measures ANOVA. Significant P-values at < 0.05 are bolded.
Algal Biomass Total Biomass Algal/Total Biomass Effect df F P F P F P Time 2 7.96 0.0063 10.96 0.002 6.06 0.0151 Treatment 2 29.09 0.0008 6.38 0.0328 29.69 0.0008 Interaction 4 2.90' 0.068 3.75 0.0334 4.24 0.0229
70
Figures
Figure 1. The study occurred in a 35 m stream reach (A) during autumnal leaf deposition.
Leaves were still green fourteen days before the study (A; 1 October 2010) but had finished deposition 24 days into the study (B, C, D; 8 November 2010). Leaf packs occasionally formed on boards and enclosure treatments (D) but rarely on ambient treatments. Biofilms became thick and filamentous by 24 days and covered all surfaces
(C and D). Leaf packs and biofilm buildup on enclosures (D) were removed every four days to prevent buildup effects from interfering with results. Arrows denote direction of water flow.
71
Figure 2. Biofilm characteristics of algal biomass (chlorophyll a x 67), total biomass (ash free dry mass), and the ratio of algal/total biomass were measured at 14, 24, and 38 days.
Data were analyzed by a repeated measures two-way ANOVA (N = 3), and letters denote significant differences after Bonferroni corrected multiple comparisons at P < 0.05.
72
Figure 3. Bacterial (3-D, R2 = 0.880, stress = 0.08) and eukaryotic (2-D, R2 = 0.912, stress = 0.07) community profiles generated by ARISA were analyzed by NMDS using
Bray-Curtis distance (N = 3). Community patterns varied with time (top) and treatment
(bottom). Numbers in parentheses are the R2 values for the respective axis.
73
Figure 4. Composite bacterial communities were identified by 454 pyrosequencing and are shown as relative abundance (%) at phylum and family taxonomic levels (N = 1). The reported taxa are represented by > 1% relative abundance. Samples are organized by treatment and nested numbers are time (days). Numbers at the top of each column represent the phyla and family richness for each respective graph. The communities are composites of replicates and therefore represent the average community.
74
Figure 5. Bacterial community similarity at phylum and family taxonomic levels were analyzed by group average cluster analysis with
Bray-Curtis distance after a variance stabilizing transformation available in the DESeq2 R package. Ambient (black circle), inclusion
(gray square), and exclusion (open triangle) treatments are represented by symbols. The sample labels include the first letter of the treatment and the time (days).
75
CHAPTER 3
FREQUENT DISTURBANCE FACILITATES THE EFFECT OF ENVIRONMENTAL
FACTORS DURING AUTUMN ON EPILITHIC BIOFILM COMMUNITY
ASSEMBLY
Abstract
Disturbance and environmental factors dictate the pattern of successional development in ecological communities. We investigated epilithic biofilm succession throughout the autumn season, and hypothesized that over an identical timeframe that biofilms grown in late autumn would have increased biomass and different successional trajectories than biofilms in early autumn. Biofilms developed on tiles that were deployed twice during the study: 7 September 2011 for early autumn biofilms and 25 October 2011 for late autumn biofilms. Biofilms established in early autumn were expected to converge to reflect late autumn biofilms by the end of the season due to changes in environmental factors; however in response to frequent spate events, we hypothesized that frequent disturbance would disrupt succession and mask the effect of changing seasonal factors.
As predicted, epilithic biofilms grown in late autumn initially had increased biomass compared to early autumn biofilms, but the first spate event eliminated the difference. In addition, the rate of development was significantly greater during late autumn compared
76
to early autumn. The bacterial community profiles of established late autumn biofilms were located between early autumn and late autumn groups indicating it was a blend of these two communities. This also demonstrates that disturbances did not mask effects of environmental factors. Biofilm category, date, and their interaction significantly influenced community composition (PERMANOVA, p < 0.001). These results suggest that frequent disturbance facilitated the effect of environmental factors by creating physical or niche space that then supported alternative communities.
Introduction
Stream biofilms are complex ecosystems that respond to variation in environmental conditions and disturbance events. Community assembly in biofilms is of considerable practical and theoretical importance (Fierer et al. 2010) and recent scientific interest has focused on understanding development in these micro-ecosystems
(Augspurger et al. 2010; Besemer et al. 2012; Lee et al. 2013). Biofilms exhibit clear successional patterns (McCormick and Stevenson 1991; Jackson et al. 2001; Besemer et al. 2007), which is the replacement of species over time that results in changing community composition throughout various stages (Clements 1916, 1936). In general for epilithic biofilms, bacteria dominate the “pioneer” community (Stock and Ward 1989;
Pohlon et al. 2010) and are the base layer that subsequently promotes algal attachment of diatoms (Hodoki 2005; Roeselers et al. 2007). Then at the “mature” stage, filamentous and green algae are dominant (Hoagland et al. 1982; Stock and Ward 1989). All of these organisms are present throughout succession, but their relative abundances and community composition changes, which is described as “initial floristic composition” in
77
vegetation ecology (Egler 1954). While succession is considered a relatively directional process, environmental perturbations can interrupt the process and/or alter the successional trajectory.
Disturbances are discrete events in time that alter community structure and change resources or the physical environment (Pickett and White 1985) and disrupt successional processes. There are different scales of disturbances, but the focus of this study is on high flow spates that would affect the entire biofilm community in a stream system and be categorized as a pulse event (Lake 2000). Spates are brief increases in stream discharge resulting from rain events and are responsible for removing biofilm material due to the scouring forces of increased flow (Biggs and Thomsen 1995). The resulting community after a disturbance event typically reflects a previous stage of succession, and the process is subsequently reinitiated but not necessarily on the same trajectory (Power and Stewart 1987; Biggs et al. 1999). Organisms are affected differently by disturbance processes, and shear stress during a spate event can select low stature non filamentous diatoms that are resistant to these forces (Power and Stewart
1987; Biggs and Close 1989; Peterson and Stevenson 1990). In addition, alternative communities can form following a disturbance event if the species pool or other environmental factors changed.
There are a multitude of environmental factors within stream ecosystems that influence epilithic biofilms, but the main abiotic factors are light, flow, and nutrients, which can all interact to dictate community assembly (Rosemond et al. 1993; Wellnitz and Rader 2003; Lange et al. 2011). Light availability (Steinman et al. 1990; Hill and
Dimick 2002; Lange et al. 2011) and flow velocity (Biggs and Thomsen 1995; Battin et
78
al. 2003; Arnon et al. 2007) are important in shaping biofilm community structure and function; algae use light during photosynthesis while flow directly influences physical architecture (Battin et al. 2003; Hödl et al. 2014) and therefore community composition
(Biggs et al. 1998; Besemer et al. 2009). Nutrient availability is important for the entire community but autotrophs and heterotrophs are influenced differently. For instance, algae are limited by phosphate in freshwater systems (Van der Grinten et al. 2004) and bacteria respond mostly to dissolved organic matter (Romaní et al. 2004; Docherty et al. 2006;
Olapade and Leff 2006). The direct response of biofilm organisms to nutrients may be augmented by indirect synergistic or antagonistic interactions within the bacterial-algal relationship.
Autumn is a time of changing environmental factors and, most notably, temperature decreases and leaf deposition occurs. Leaf deposition will increase light availability as canopy cover decreases (Hill and Dimick 2002) but physical effects can decrease light availability and flow velocity to the benthos subsequently altering epilithic biofilm community composition (Lang et al. In press). Nutrients are also released as leaves decompose (Gessner et al. 1999) and bacteria can utilize these nutrients, which can produce different community compositions when bacterial responses differ (Benner et al.
1986; McNamara and Leff 2004; Wu et al. 2009). Temperature affects enzyme function and cellular processes and as temperature decreases the availability of dissolved oxygen increases. Decreased temperature has also been associated with decreased respiration
(Rosa et al. 2013), denitrification (Boulêtreau et al. 2012), and utilization of recalcitrant organic matter (Ylla et al. 2012) in epilithic biofilms. Thus the environmental changes
79
associated with autumn have strong potential to impact stream biofilms; however, little is known about successional processes in these biofilms in relation to disturbance dynamics.
The main objective of this study was to monitor biofilm development and community composition throughout the changing autumn season and compare early vs. late autumn biofilms. A third biofilm category was labeled “established late autumn biofilms”. This category was used to reflect natural conditions of established biofilms at a mature stage that were subjected to the changing environmental conditions in late autumn. Barring a large-scale disturbance event, there should be a substantial amount of biofilm at a mature stage at any given time. This established community may respond differently or take longer to respond to changes in environmental conditions, and we were interested in investigating that. Biofilms developed on unglazed tiles that were deployed twice during the study and subjected to ambient stream conditions: 7 September 2011 for early autumn biofilms and 25 October 2011 for late autumn biofilms. Additionally, frequent spate events presented the opportunity to investigate the relationship of disturbance dynamics and biofilms. Bacterial community profiles were described using automatic ribosomal intergenic spacer analysis (ARISA) to elucidate patterns of community composition. We hypothesized that (H1) biofilms grown in late autumn would have increased biomass and different communities than biofilms in early autumn due to changing environmental factors. We also hypothesized that (H2) biofilms established in early autumn would slowly change to resemble those established in late autumn by the end of the season. Additionally, (H3) frequent disturbances were expected to disrupt succession and mask differences in environmental factors because the communities would be constantly at a pioneer stage of succession where they would be dominated by
80
efficient colonizers as predicted by the intermediate disturbance hypothesis (Connell
1978).
Methods
Site description
The study was conducted in a lower third order river section of the upper region of the Little Miami River in a small deciduous forest corridor of the Little Miami State
Forest Preserve in Xenia, Ohio, USA (39°76.552 83°90.062). The surrounding landcover of the catchment area was predominately agriculture, and the riparian forest was dominated by maple (Acer sp.) and elm (Ulmus sp.), but included hackberry (Celtis sp.), sycamore (Plantanus sp.), and walnut (Juglans sp.). The stream has been categorized as an Exceptional Warmwater Habitat by the Ohio Environmental Protection Agency because it supported high diversity of aquatic organisms (OEPA 2002). The study was conducted in a run habitat where the substrate was predominately gravel and cobble with several intermittent boulders from 7 September 2011 to 13 December 2011.
Environmental parameters
Steam mean daily discharge data was collected from the USGS using the
03240000 Little Miami River near Oldtown, OH gage (39°44.55, 83°55.50). It is located about 2 miles upstream of the study site. Environmental water quality parameters were measured at the start of the study and on every sampling date at upper, middle, and lower sections along the reach. Specific conductivity (SpCond µS/cm), total dissolved solids
(TDS mg/L), turbidity (NTU), pH, and temperature (°C) were recorded using a YSI 6600
81
v2 Sonde (YSI Inc, Yellow Springs, OH) every 30 sec for 10 min, and water was
3- 3- collected to measure phosphorus (mg/L PO4 ), chlorine (mg/L), nitrate (mg/L NO -N),
- nitrite (mg/L NO2 -N), sulfate (mg/L), ammonia (mg/L NH3-N), alkalinity (mg/L
CaCO3), and total suspended solids (TSS mg/L) in the lab (Hach® Company, Loveland,
CO). Also at each experimental board (see below), light availability (lux) and canopy cover (densitometer) before leaf deposition was determined. Water temperature was recorded (iButton Temperature Logger; Fondriest Environmental Inc., Fairborn, OH) continuously throughout the study at 30 min intervals at the first, middle, and last experimental boards.
Biofilm development and processing
Epilithic biofilms developed naturally on hexagonal unglazed porcelain tiles (N =
6) attached to brick pavers (19.2 x 9 x 1.3 cm) with 100% silicone. Each paver was attached to a wood board (1.2 x 0.3 x 0.02 m) with VELCRO® tape under natural ambient conditions. Tiles were deployed twice during the study: 7 September 2011 for early autumn biofilms and 25 October 2011 for late autumn biofilms. The date of October 25th was chosen to designate the beginning of late autumn because riparian leaf deposition was mostly complete. Tiles (N = 4) were randomly collected weekly, but timing was contingent upon water levels allowing accessibility (Figure 1). Biofilms were transported to the lab and placed at -20°C until processing. Biofilm biomass was removed using a sterile razor blade and toothbrush and suspended in ultrapure water (NANOpure II;
Barnstead, Boston, MA, USA). Two sub samples were collected on GB-140 glass membrane filters (diameter, 25mm; pore size, 0.4 µm; Sterlitech, Kent, WA, USA) to
82
determine total biomass as ash free dry mass (AFDM) and algal-associated biomass as chlorophyll a following techniques used by (Steinman et al. 2007). A third subsample was collected for DNA extractions, and these filters were stored in 90% ethanol at -80°C until extraction. Although for DNA extractions, we highly suggest using a mixed nitrocellulose filter and placing a few drops of ethanol on the sample that evaporates before storage at -80°C.
DNA extraction
DNA was extracted from ethanol evaporated, dried filters using a combination of methods from Miller et al., (1999) and Zhou et al., (1996) as suggested by Lear et al.,
(2010). Samples were lysed by bead beating (0.5 g each of 0.1 mm and 0.5 mm glass lysis beads; RPI, Mount Prospect, IL, USA) for 15 min on a horizontal vortex adaptor
(MO BIO Laboratories, Carlsbad, CA, USA) at full speed in 1.2 mL of extraction buffer
(100 mM Tris-HCl [pH 8.0], 100 mM EDTA disodium salt [pH 8.0], 100mM sodium phosphate [pH 8.0], 1.5 M sodium chloride, 1% CTAB), 12 µL proteinase K (20 mg/mL), and 30 µL SDS (20%). The mixture was incubated at 65°C for 1 hr with gentle end-over- end inversions by hand every 15 min. Then, 4 µL of RNase (100 mg/µL) was added.
DNA was isolated from organic debris with chloroform/isoamyl alcohol extraction and was precipitated overnight at -20°C with isopropanol. The mixture was warmed to 37°C to dissolve salt precipitates, and DNA was pelleted at 15,000 g for 30 min. The DNA pellet was washed twice with ice cold 70% EtOH and dissolved in 25-50 µL ultrapure water (NANOpure II; Barnstead, Boston, MA, USA). Samples were purified using
PowerClean® DNA Clean-Up Kit (MO BIO Laboratories, Carlsbad, CA, USA) with a
83
modified protocol for low amounts of DNA to remove PCR inhibitors typical of biofilm samples.
ARISA
Bacterial communities were assessed using profiles created by automated ribosomal intergenic spacer analysis (ARISA) (Fisher and Triplett 1999), which generates a unique ‘fingerprint’ of microbial communities using the 16S-23S intergenic space in bacteria. While ARISA does not taxonomically identify organisms like next generation sequencing methods, this method generates community profiles that produce similar patterns in results (Bienhold et al. 2012; van Dorst et al. 2014). Approximately
15-20 ng of DNA quantified by spectrophotometer (NanoPhotometerTM Pearl; Denville
Scientific Inc., South Plainfield, NJ, USA) was amplified by PCR using 25 µL GoTaq®
Colorless Master Mix (Promega, Madison, WI, USA) with 0.5 µM of forward and reverse primers. Bacteria ribosomal intergenic space regions were amplified with primers
ITSF (5’-GTCGTAACAAGGTAGCCGTA-3’) labeled with FAM at the 5’ end (IDT,
Coralville, IA, USA) and ITSReub (5’-GCCAAGGCATCCACC-3’) (Cardinale et al.
2004). Fragments were created with the following PCR conditions: (i) 94°C for 3 min,
(ii) 35 cycles of 94°C for 1 min, 56°C (57.5°C for eukaryotes) for 1 min, 72°C for 2 min, and finally (iii) 72°C for 10 min (Fechner et al. 2010). PCR products were sent to DNA
Analysis, LLC (Cincinnati, OH, USA) for fragment analysis on an ABI 3100 (Life
Technologies, Carlsbad, CA, USA). Fragments were interpreted using Genescan v 3.7 using the Local Southern Size Calling Method with a peak height threshold of 100
84
fluorescence units to remove background fluorescence and formulated using
GeneMapper v 2.5 (Life Technologies, Carlsbad, CA, USA).
Fragment peak length and area was converted to column format using the treeflap
Excel Macro (http://www.wsc.monash.edu.au/~cwalsh/treeflap.xls) and processed with the automatic_binner script to determine binning window size and the interactive_binner script to determine the best starting window position (Ramette 2009) in R v 3.1.0 (R Core
Team 2014). This method was used to account for inherent imprecision of analyzer machines. Peak area was converted to relative abundance of each fragment as part of the entire sample, fragments < 0.09% relative abundance were removed (Ramette 2009), and window size was calculated to be 1.5 base pairs.
Data analysis
Water quality parameters and time since study initiation were analyzed with a
Pearson correlation to identify trends during the season. Chlorine was the only parameter without a normal distribution and so a Spearman’s correlation was used. All analyses were conducted using GraphPad Prism v 5.0 (GraphPad Software Inc, San Diego, CA,
USA).
Biofilms were categorized into three groups using a combination of when growth was initiated and the time of the season. Early autumn biofilms were deployed on 7
September 2011 and late autumn biofilms were deployed on 25 October 2011 (Figure 2).
Established late autumn biofilms were deployed on 7 September 2011, and so were established in early autumn but sampled during late autumn; hence, they were categorized as established late autumn biofilms. The effect of pre-growth of the
85
established late autumn biofilms was determined using a student’s t-test for both AFDM and chlorophyll a. Biofilm development between the biofilm categories was compared to determine the effect of seasonal differences and spate disturbances. Development was represented by AFDM and compared using a scale of accumulated degree days (ADD) to account for the effect of changing temperature throughout the study. Early autumn and late autumn biofilms were compared to assess the influence of seasonal differences. The rate of biofilm development for each category was determined by a linear regression of
AFDM with undisturbed development time measured in ADD and slopes were compared for significant differences to determine if biofilms groups were growing at different rates.
Undisturbed development time was defined as the amount of time since deployment or the last disturbance event. Only early autumn biofilms were not normal when tested using the D’Agostino normality test. Analyses were performed in R v 3.1.2 (R Core Team
2014) with vegan v 2.2-1 (Oksanen et al. 2015) , simba v 0.3-5 (Jurasinski and Retzer
2012), and fBasics v 3011.87 (Rmetrics Core Team et al. 2014) packages. Figures were created within GraphPad Prism v 5.0 (GraphPad Software Inc, San Diego, CA, USA).
Microbial community patterns were visualized using nonmetric multidimensional scaling (NMDS) with Bray-Curtis (Sørensen) distance because it is a nonparametric approach useful in evaluating nonlinear relationships of data with high numbers of zeros
(McCune and Grace 2002). Differences in biofilm category were tested with nonparametric multivariate analysis of variance (PERMANOVA) (Anderson 2001) and ellipses representing standard error 95% confidence intervals of groups are displayed on the NMDS ordination. Sampling date was also used as an overlay on the ordination; however, date was not tested with PERMANOVA because group dispersions were
86
significantly different. All environmental factors and description categories were correlated to the ordination and significant factors (P < 0.05) were fitted as vectors.
Analyses were conducted in R v 3.1.2 (R Core Team 2014) with the vegan v 2.2-1
(Oksanen et al. 2015) package.
Results
Autumnal changes in environmental conditions
In temperate regions, autumn is generally characterized by cooling temperatures and tree senescence accompanied by leaf fall. Distinguishable relationships between water quality parameters and time (correlation; Figure 1) in our study site included temperature (R2 = 0.88, P < 0.0001), phosphorus (R2 = 0.86, P < 0.0001), total suspended solids (R2 = 0.48, P = 0.0122), specific conductivity (R2 = 0.48, P = 0.0122), and turbidity (R2 = 0.51, P = 0.0132) all of which decreased throughout the study period. Our data indicate that pH (R2 = 0.44, P = 0.0195) and chlorine (R2 = 0.84, P = 0.0009) increased throughout autumn.
There were also frequent disturbances as a result of spate events during this study, and they were stronger (greater discharge) and more frequent (~ every 7 days) towards the end of the study (indicated by discharge spikes in Figure 2). These disturbances removed the deposited leaf material, and there were few leaves in the benthos by the end of the study, in contrast to the streambed being predominately covered at the beginning of late autumn (JML, personal observation).
87
Biofilm biomass dynamics
Biofilm growth patterns indicated by AFDM and chlorophyll a were similar and reflect seasonal differences and spate disturbances. In general, biomass increased until a disturbance event removed material. Established late autumn biofilms had significantly greater AFDM than late autumn biofilms (t-test, P < 0.05) because they had been previously growing, but this effect was soon eliminated by frequent disturbances (Figure
2). Chlorophyll a was also increased in established late autumn biofilms but only significantly greater on the first sampling date. These levels were relatively low throughout the study, but this reflects the fact that algae become established later during successional development and that spates removed material by scouring the biofilm.
Biofilm development assessed using AFDM was different between the biofilm categories. Initial development of late autumn biofilms was greater than early autumn biofilms, but spate disturbances reduced biomass by the third sampling date and so we do not know if this trend would have continued (Figure 3A). Biofilm AFDM levels were significantly predicted (linear regression) by undisturbed development time (ADD) for established late autumn (R2 = 0.59, P < 0.0001) and late autumn (R2 = 0.77, P < 0.0001) biofilms but not early autumn biofilms (R2 = 0.09, P = 0.1967) (Figure 3B). Also, the slopes of established late autumn and late autumn biofilms were significantly different from early autumn biofilms (P = 0.001, P = 0.027; respectively) but not from each other
(P = 0.119), therefore indicating that the rate of development was dependent upon seasonal timing.
88
Community assembly patterns
Bacterial biofilm communities displayed patterns that were related to season and environmental factors. Biofilm category was significant in driving community composition (PERMANOVA, F = 3.5886, P < 0.0001), and the distinctly separate ellipses of standard error (95% CI) indicate differences among the three biofilm communities (Figure 4). The three biofilm communities were arrayed in a distinguishable fashion along NMDS axis one with the early autumn biofilm communities located on the right side of the ordination diagram and the late autumn biofilms found on the left side.
The established late autumn biofilms are located between the other categories, which indicates that they are a blend of early autumn and late autumn communities. In addition, each sampling date of early autumn biofilms clustered together while the two late autumn categories were overlapping (Figure 4, middle panel). This suggests that succession is occurring but that the biofilm category was the most influential.
Many environmental factors were correlated to the NMDS ordination with vectors reflecting the seasonal gradient (Figure 4; Table 1). Most vectors were arrayed diagonally along NMDS Axis one, indicating the separation was linked with seasonal differences of biofilms. Turbidity and sulfate do not follow this pattern but elevated levels most likely drove this departure. Turbidity is faced directly towards 14 September, which has the highest turbidity value for the entire study. Similarly for sulfate, the greatest value was on
14 September and the second highest value was on 9 November. There are also parameters that displayed very similar results. Variables related to particulate matter
(total suspended solids, total dissolved solids, and specific conductivity) were clustered together, along with the categorical variables of biofilm category (R2 = 0.670, P = 0.001)
89
and seasonal time (R2 = 0.673, P = 0.001) (Fig. 4). Biofilm category is dependent upon seasonal timing so this is expected, but the date vector (R2 = 0.09, P = 0.1967) is shifted away from these two categories. Other insignificant, categorical variables tested were undisturbed development time (R2 = 0.048, P = 0.538) and AFDM (R2 = 0.080, P =
0.356).
Discussion
Biofilm development and autumnal changes in environmental factors
Throughout the autumn season, increased light availability from reduced canopy cover could promote algal growth (Hill and Dimick 2002) and increased nutrients from decomposing leaves could promote bacterial growth (Lock and Hynes 1976; Benner et al.
1986; McNamara and Leff 2004; Wu et al. 2009), which would drive alterations in the biofilm community. We hypothesized (H1) that biofilms in late autumn would have both increased biomass and a unique community compared to early autumn biofilms as a result of these changing environmental factors. Temperature decreased throughout the study, but this was not expected to have an overall impact on biomass in comparison to the other factors. It could change community dynamics within biofilms because temperature affects respiration (Rosa et al. 2013), denitrification (Boulêtreau et al. 2012), and utilization of recalcitrant organic matter (Ylla et al. 2012). We did not measure any environmental change over autumn that would indicate increasing resources, but it is still probable that increased nutrients from leaching had an effect. Although spate events removed the deposited leaf material and very little remained by the end of the season, which may explain the decrease in particulate matter measures (total suspended solids
90
and turbidity) and possibly unmeasured nutrients. Indeed, just one storm event removed about 64% of the coarse particulate matter (primarily leaves) in a Tennessee stream
(Mulholland et al. 1985) so it is not surprising that there was minimal leaf material after multiple spate events.
The storm events may have also been responsible for the trend of decreasing phosphorus throughout the season. Phosphorus export from soil is driven by extreme discharges from storm flow where other nutrients like nitrate are released during non- storm flow periods (Pionke et al. 1996; Royer et al. 2006). For example, about 70-80% of soil phosphorus was exported during storms events (Pionke et al. 1996; Royer et al. 2006) and it is suggested that 50-80% of all incidental loss of phosphorus from soil is a result of storm runoff with the greatest amount occurring during the first event following a fertilizer application (Withers et al. 2003). Because the catchment area is predominately agriculture, stream phosphorus levels could have been influenced by fertilizer runoff. The stabilized phosphorus levels by the end of the season could indicate that the inputs of agriculture phosphorus through runoff were previously exhausted.
We suspect that algae drove the elevated AFDM levels in the late autumn biofilms. Algae typically have greater biomass than bacteria (Romaní and Sabater 2000) and also provide substrate for bacterial colonization (Rier and Stevenson 2001).
Therefore, when algae become dominate later in succession, biomass levels will increase.
The chlorophyll a values at and near zero at the beginning of development reflect these successional patterns. Although, the increase in chlorophyll a by the second date of late autumn (9 Nov) indicates that the algal community became established much earlier during the successional sequence in late autumn biofilms versus early autumn biofilms.
91
This earlier establishment is an indication that changed conditions promoted algal growth, and we postulate that this is a result of a light effect because photosynthetic efficiency can increase after leaf fall (Division and Ridge 2002). Leaf fall should increase light availability at the stream surface as canopy cover decreases (Hill and Dimick 2002); however, we found no measurable differences in light availability through autumn. Two caveats to this finding are that the time of day of measurements was inconsistent and there was changing cloud cover. Thus, while light availability at the stream surface probably increased over time, we were unable to quantify this change.
Established late autumn biofilms had elevated levels of biomass in comparison to late autumn biofilms but similar rates of development. Also, both of these rates were different than the early autumn biofilm rate, which indicates that the rate of development was driven by seasonal timing and differences in environmental conditions. This demonstrates that developing communities, regardless if there was previously established biomass or not, developed at similar rates during the same time period. But the question is whether similar rates would have occurred in established late autumn biofilms that were not disturbed. Would seasonal differences support increased biomass or would the physical stress of flow prevent increased growth? The prediction of increased biomass was based on previous field observations (JML) where biofilms became very thick with long filamentous streams during autumn. Future studies can investigate why this occurs and can determine if biofilms function differently during early or late autumn. Results would have implications for the role of epilithic biofilms in ecosystem processes during the autumn season.
92
Community Composition
Community composition represented by bacterial ARISA profiles was driven by seasonal timing and reflects changing environmental factors. The established late autumn biofilms are located between the other categories, indicating that they are a blend of early autumn and late autumn communities. This supports the prediction that early autumn biofilms would be distinctly different from late autumn biofilms (H1) and that biofilm composition of established biofilms would shift to resemble those of late autumn (H2).
Individual environmental factors can influence biofilm community assembly, but seasonal changes reflect the cumulative effect of multiple factors. Seasonal differences in community composition have been documented for diatoms (Hoagland et al. 1982), algae
(Ledger and Hildrew 1998), and bacteria (Olapade and Leff 2004) indicating that the changes affect the entire biofilm community. There were also differences in nutrient utilization of dissolved organic matter (DOM) by bacteria (Olapade and Leff 2006) that included the use of leaf leachate (Koetsier et al. 1997), which indicates that biofilm function may change as well. Studies spanning multiple seasons show cumulative effects of changing factors can influence biofilms, but our study shows that changes are also happening within a season.
Environmental factors can influence community composition by acting as a filter that selects particular organisms. For example, pH was the most important driving factor when comparing epilithic communities from 17 streams that spanned a gradient of acid mine drainage impact (Lear et al. 2009). Both bacteria and algae were affected as indicated by DNA clone sequences. Increases in the acid tolerant bacteria and algae genera were reflective of very acidic streams (Lear et al. 2009). Different communities
93
have also developed based on catchment land use (Lear and Lewis 2009) and stream location (Lear et al. 2008). The differences in community composition based on stream location were greater than spatial and temporal variation during10 weeks within the streams, and the driving factor was determined to be temperature (Lear et al. 2008).
Nutrients can also influence community structure. Phosphate regime altered diatom composition, and certain species at high phosphate levels overgrew the biofilm as a result of their advantageous filamentous structure (Van der Grinten et al. 2004). Dissolved organic matter composition selected similar bacterial communities even though inoculums with varying bacterial composition were tested (Docherty et al. 2006). In addition, bacteria responded more to forms of DOM rather than inorganic nutrients
(Olapade and Leff 2006). These studies demonstrate the importance of environmental factors in dictating biofilm community composition and that each factor has the potential to be important.
Alternative communities and the role of disturbance
Disturbances facilitated the effect of environmental factors rather than masking them, which is contrary to the hypothesis (H3) that frequent disturbance would disrupt succession and mask differences in environmental factors. The reasoning is that disturbances would constantly be creating early stage pioneer communities, therefore selecting early arrival organisms regardless of environmental factors. These communities were predicted to be the same because biofilm communities develop in a non stochastic manner from the source communities in the water column (Besemer et al. 2012).
Bacterial biofilm communities from various stream types were more similar to each other
94
than their water column source communities (Besemer et al. 2012), which suggests that species sorting is an assembly mechanism in stream biofilms (Besemer et al. 2012; Lee et al. 2013). Yet, results demonstrate that even the pioneer communities were different throughout this study. Whether the species pool itself had changed is unknown because water column communities were not sampled, but even if it remained the same, environmental factors selected which organisms flourished. This suggests that the changing environmental factors promoted different communities by providing the opportunity to proliferate in the physical or niche space created by the disturbances.
The formation of alternative communities may also impact the function of these communities. This is an important consideration because biofilms are an essential component in stream ecosystem processes (Battin et al. 2003) and autumn includes an influx of nutrients from leaf deposition. These nutrients, especially in the form of dissolved organic matter, can be utilized by biofilms (Lock and Hynes 1976) and influence the community composition (McNamara and Leff 2004). In addition, enzyme activity changes in response to the available organic matter and the presence of algae
(Jones and Lock 1993; Romaní et al. 2004). Light-grown biofilms are a net DOM consumer (Romaní et al. 2004) and have increased enzyme activity in comparison to dark-grown biofilms (Espeland et al. 2001). This suggests that algae not only increase biofilm biomass in late autumn, but also can change biofilm function by promoting organic matter utilization through increased enzymatic activity.
95
Conclusion and future direction
Biofilm community composition and development was different throughout the autumn season and this reflected differences in environmental variables. The majority of the measured environmental factors were correlated to the ordination, which suggests that differences in early and late autumn community composition was a cumulative result of these changing factors. It appears these variables exerted a selective pressure, subsequently leading to alternative community composition. Further, disturbances that disrupted community composition facilitated the influence of environmental factors by providing an opportunity for alternative communities to form. Autumn is characterized by increases in light availability, decreased temperature, and leaf deposition, which could all lead to changes in biofilm function, and because biofilm function is integral in ecosystem processes, this can have bottom-up effects on higher-level stream function and processes. The relationship between structure and function and possible differences in biofilm function between early and late autumn biofilms is an area that should be further explored.
Funding
This work was supported in part by the University of Dayton Office of Graduate
Professional and Continuing Education through the Graduate Student Summer
Fellowship Program awarded to JML and through discretionary funds of the Department of Biology at the University of Dayton.
96
Acknowledgements
We are grateful to Will Kmetz, Jon White, and Tiffany Blair for support in the field and lab and Jen Pechal for assistance in extractions and analyses. We would also like to thank
Allison Gansel, Lauren Shewhart, Alex Calteaux, and Patrick Vrablik for help with sample processing.
Conflict of interest
The authors declare no conflict of interest.
97
Literature Cited
Anderson, M. J. 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecology 26:32–46.
Arnon, S., A. I. Packman, C. G. Peterson, and K. A. Gray. 2007. Effects of overlying velocity on periphyton structure and denitrification. Journal of Geophysical Research 112:G01002.
Augspurger, C., C. Karwautz, M. MuBmann, H. Daims, and T. J. Battin. 2010. Drivers of bacterial colonization patterns in stream biofilms. FEMS Microbiology Ecology 72:47–57.
Battin, T. J., L. A. Kaplan, J. D. Newbold, X. Cheng, and C. M. E. Hansen. 2003a. Effects of current velocity on the nascent architecture of stream microbial biofilms. Appl Environ Microbiol 69:5443–5452.
Battin, T. J., L. A. Kaplan, J. D. Newbold, and C. M. E. Hansen. 2003b. Contributions of microbial biofilms to ecosystem processes in stream mesocosms. Nature 426:439– 42.
Benner, R., E. R. Peele, and R. E. Hodson. 1986. Microbial utilization of dissolved organic matter from leaves of the red mangrove, Rhizophora mangle, in the Fresh Creek estuary, Bahamas. Estuarine, Coastal and Shelf Science 23:607–619.
Besemer, K., I. Hödl, G. Singer, and T. J. Battin. 2009. Architectural differentiation reflects bacterial community structure in stream biofilms. ISME J 3:1318–1324.
Besemer, K., H. Peter, J. B. Logue, S. Langenheder, E. S. Lindström, L. J. Tranvik, T. J. Battin, and E. S. Lindstrom. 2012. Unraveling assembly of stream biofilm communities. ISME J 6:1459–68.
Besemer, K., G. Singer, R. Limberger, A.-K. K. Chlup, G. Hochedlinger, I. Hödl, C. Baranyi, and T. J. Battin. 2007. Biophysical controls on community succession in stream biofilms. Applied and Environmental Microbiology 73:4966–4974.
Bienhold, C., A. Boetius, and A. Ramette. 2012. The energy-diversity relationship of complex bacterial communities in Arctic deep-sea sediments. ISME J 6:724–732.
Biggs, B. J. F., and M. E. Close. 1989. Periphyton biomass dynamics in gravel bed rivers: the relative effects of flows and nutrients. Freshwater Biology 22:209–231.
Biggs, B. J. F., D. G. Goring, and V. I. Nikora. 1998. Subsidy and stress responses of stream periphyton to gradients in water velocity as a function of community growth form. J. Phycol. 34:598–607.
98
Biggs, B. J. F., and H. A. Thomsen. 1995. Disturbance of stream periphyton by perturbations in shear stress time to structural failure and differences in community resistance. Journal of Phycology 31:233–241.
Biggs, B. J. F., N. C. Tuchman, R. L. Lowe, and R. J. Stevenson. 1999. Resource stress alters hydrological disturbance effects in a stream periphyton community. Oikos 85:95–108.
Boulêtreau, S., E. Salvo, E. Lyautey, S. Mastrorillo, and F. Garabetian. 2012. Temperature dependence of denitrification in phototrophic river biofilms. Science of the Total Environment 416:323–8.
Cardinale, M., L. Brusetti, P. Quatrini, S. Borin, A. M. Puglia, A. Rizzi, E. Zanardini, C. Sorlini, C. Corselli, and D. Daffonchio. 2004. Comparison of different primer sets for use in automated ribosomal intergenic spacer analysis of complex bacterial communities. Applied and Environmental Microbiology 70:6147–6156.
Clements, F. E. 1916. Plant succession: Analysis of the development of vegetation. Carnegie Institute of Washington, Washington D.C.
Clements, F. E. 1936. Nature and structure of the climax. Ecology 24:252–284.
Connell, J. . 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302– 1310.
Docherty, K. M., K. C. Young, P. A. Maurice, and S. D. Bridgham. 2006. Dissolved organic matter concentration and quality influences upon structure and function of freshwater microbial communities. Microbial Ecology 52:378–88.
Van Dorst, J., A. Bissett, A. S. Palmer, M. Brown, I. Snape, J. S. Stark, B. Raymond, J. McKinlay, M. Ji, T. Winsley, and B. C. Ferrari. 2014. Community fingerprinting in a sequencing world. FEMS Microbiology Ecology 89:316–330.
Egler, F. E. 1954. Vegetation science concepts I. Initial floristic composition, a factor in old-field vegetation development with 2 figs. Vegetatio 4:412–417.
Espeland, E. M., S. N. Francoeur, and R. G. Wetzel. 2001. Influence of algal photosynthesis on biofilm bacterial production and associated glucosidase and xylosidase activities. Microbial Ecology 42:524–530.
Fechner, L. C., F. Vincent-Hubert, P. Gaubert, T. Bouchez, C. Gourlay-Francé, and M.- H. Tusseau-Vuillemin. 2010. Combined eukaryotic and bacterial community fingerprinting of natural freshwater biofilms using automated ribosomal intergenic spacer analysis. FEMS Microbiology Ecology 74:542–53.
99
Fierer, N., D. Nemergut, R. Knight, and J. M. Craine. 2010. Changes through time: integrating microorganisms into the study of succession. Research in Microbiology 161:635–642.
Fisher, M. M., and E. W. Triplett. 1999. Automated approach for ribosomal intergenic spacer analysis of microbial diversity and its application to freshwater bacterial communities. Applied and Environmental Microbiology 65:4630–4636.
Gessner, M. O., E. Chauvet, M. Dobson, S. Oikos, and F. May. 1999. A perspective on leaf litter breakdown in streams. OIKOS 85:377–384.
Van der Grinten, E., M. Janssen, S. G. H. Simis, C. Barranguet, and W. Admiraal. 2004. Phosphate regime structures species composition in cultured phototrophic biofilms. Freshwater Biology 49:369–381.
Hill, W., S. Dimick, E. S. Division, and O. Ridge. 2002. Effects of riparian leaf dynamics on periphyton photosynthesis and light utilisation efficiency. Freshwater Biology:1245–1256.
Hoagland, K. D., S. C. Roemer, J. R. Rosowski, and R. Rosowski. 1982. Colonization and community structure of two periphyton assemblages, with emphasis on the diatoms (Bacillariophyceae). American Journal of Botany 69:pp. 188–213.
Hödl, I., L. Mari, E. Bertuzzo, S. Suweis, K. Besemer, A. Rinaldo, and T. J. Battin. 2014. Biophysical controls on cluster dynamics and architectural differentiation of microbial biofilms in contrasting flow environments. Environmental Microbiology 16:802–12.
Hodoki, Y. 2005. Bacteria biofilm encourages algal immigration onto substrata in lotic systems. Hydrobiolgia 539:27–34.
Jackson, C. R., P. F. Churchill, and E. E. Roden. 2001. Successional changes in bacterial assemblage structure during epilithic biofilm development. Ecology 82:555–566.
Jones, S. E., and M. A. Lock. 1993. Seasonal determinations of extracellular hydrolytic activities in heterotrophic and mixed heterotrophic/autotrophic biofilms from two contrasting rivers. Hydrobiologia 257:1–16.
Jurasinski, G., and V. Retzer. 2012. simba: A collection of functions for similarity analysis of vegetation data.
Koetsier, P. I. I. I., J. V McArthur, and L. G. Leff. 1997. Spatial and temporal response of stream bacteria to sources of dissolved organic carbon in a blackwater stream system. Freshwater Biology 37:79–89.
100
Lake, P. S. 2000. Disturbance, patchiness, and diversity in streams. Journal of the North American Benthological Society 19:573–592.
Lang, J., R. McEwan, and M. Benbow. (n.d.). Abiotic autumnal organic matter deposition and grazing disturbance effects on epilithic biofilm succession. FEMS Microbiology Ecology.
Lange, K., A. Liess, J. J. Piggott, C. R. Townsend, and C. D. Matthaei. 2011. Light, nutrients and grazing interact to determine stream diatom community composition and functional group structure. Freshwater Biology 56:264–278.
Lear, G., M. J. Anderson, J. P. Smith, K. Boxen, and G. D. Lewis. 2008. Spatial and temporal heterogeneity of the bacterial communities in stream epilithic biofilms. FEMS Microbiology Ecology 65:463–473.
Lear, G., Y. Dong, and G. Lewis. 2010. Comparison of methods for the extraction of DNA from stream epilithic biofilms. Antonie van Leeuwenhoek 98:567–571.
Lear, G., and G. D. Lewis. 2009. Impact of catchment land use on bacterial communities within stream biofilms. Ecological Indicators 9:848–855.
Lear, G., D. Niyogi, J. Harding, Y. Dong, and G. Lewis. 2009. Biofilm bacterial community structure in streams affected by acid mine drainage. Applied and Environmental Microbiology 75:3455–3460.
Ledger, M. E., and A. G. Hildrew. 1998. Temporal and spatial variation in the epilithic biofilm of an acid stream. Freshwater Biology 40:655–670.
Lee, J. E., H. L. Buckley, R. S. Etienne, and G. Lear. 2013. Both species sorting and neutral processes drive assembly of bacterial communities in aquatic microcosms. FEMS Microbiology Ecology 86:288–302.
Lock, M., and H. Hynes. 1976. The fate of “dissolved” organic carbon derived from autumn-shed maple leaves (Acer saccharurn) in a temperate hard-water stream1. Limnology and Oceanography 21:436–443.
McCormick, P. V, and R. J. Stevenson. 1991. Mechanisms of benthic algal succession in lotic environments. Ecology 72:1835–1848.
McCune, B., and J. B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR.
McNamara, C. J., and L. G. Leff. 2004. Response of biofilm bacteria to dissolved organic matter from decomposing maple leaves. Microbial Ecology 48:324–330.
101
Miller, D. N., J. E. Bryant, E. L. Madsen, and W. C. Ghiorse. 1999. Evaluation and optimization of DNA extraction and purification procedures for soil and sediment samples. Applied and Environmental Microbiology 65:4715–4724.
Mulholland, P. J., J. D. Newbold, J. W. Elwood, L. A. Ferren, R. W. Jackson, and J. R. Webster. 1985. Phosphorus spiralling in a woodland stream: seasonal variations. Ecology 66:pp. 1012–1023.
OEPA. 2002. Total maximum daily loads for the upper Little Miami River. Columbus.
Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O’Hara, G. L. Simpson, P. Solymos, M. H. Stevens, and H. Wagner. 2015. vegan: Community Ecology Package.
Olapade, O. A., and L. G. Leff. 2004. Seasonal dynamics of bacterial assemblages in epilithic biofilms in a northeastern Ohio stream. Journal of the North American Benthological Society 23:686–700.
Olapade, O. A., and L. G. Leff. 2006. Influence of dissolved organic matter and inorganic nutrients on the biofilm bacterial community on artificial substrates in a northeastern Ohio, USA, stream. Canadian Journal of Microbiology 52:540–9.
Peterson, C. G., and R. J. Stevenson. 1990. Post-spate development of epilithic algal communities in different current environments. Canadian Journal of Botany 68:2092–102.
Pickett, S. T. A., and P. S. White. 1985. The ecology of natural disturbance and patch dynamics. Academic Press, Orlando, Florida.
Pionke, H. B., W. J. Gburek, A. N. Sharpley, and R. R. Schnabel. 1996. Flow and nutrient export patterns for an agricultural hill-land watershed. Water Resources Research 32:1795–1804.
Pohlon, E., J. Marxsen, and K. Küsel. 2010. Pioneering bacterial and algal communities and potential extracellular enzyme activities of stream biofilms. FEMS Microbiology Ecology 71:364–373.
Power, M. E., and A. J. Stewart. 1987. Disturbance and recovery of an algal assemblage following flooding in an Oklahoma stream. American Midland Naturalist 117:333– 345.
Ramette, A. 2009. Quantitative community fingerprinting methods for estimating the abundance of operational taxonomic units in natural microbial communities. Applied and Environmental Microbiology 75:2495–505.
102
Rier, S. T., and R. J. Stevenson. 2001. Relation of environmental factors to density of epilithic lotic bacteria in 2 ecoregions. Journal of the North American Benthological Society 20:520–532.
Rmetrics Core Team, D. Wuertz, T. Setz, and Y. Chalabi. 2014. fBasics: Rmetrics - Markets and Basic Statistics.
Roeselers, G., M. C. M. van Loosdrecht, and G. Muyzer. 2007. Heterotrophic pioneers facilitate phototrophic biofilm development. Microbial Ecology 54:578–585.
Romaní, A. M., H. Guasch, I. Muñoz, J. Ruana, E. Vilalta, T. Schwartz, F. Emtiazi, and S. Sabater. 2004. Biofilm structure and function and possible implications for riverine DOC dynamics. Microbial Ecology 47:316–28.
Romaní, A. M., and S. Sabater. 2000. Influence of algal biomass on extracellular enzyme activity in river biofilms. Microbial Ecology 40:16–24.
Rosa, J., V. Ferreira, C. Canhoto, and M. A. S. Graça. 2013. Combined effects of water temperature and nutrients concentration on periphyton respiration - implications of global change. International Review of Hydrobiology 98:14–23.
Rosemond, A. D., P. J. Mulholland, and J. W. Elwood. 1993. Top-down and bottom-up control of stream periphyton: effects of nutrients and herbivores. Ecology 74:1264– 1280.
Royer, T. V, M. B. David, and L. E. Gentry. 2006. Timing of riverine export of nitrate and phosphorus from agricultural watersheds in Illinois: Implications for reducing nutrient loading to the Mississippi River. Environmental Science and Technology 40:4126–4131.
Steinman, A. D., G. A. Lamberti, and L. P.R. 2007. Biomass and pigments of benthic algae. Pages 357–380 in F. R. Hauer and G. A. Lamberti, editors. Methods in Stream Ecology, 2nd edition. Academic Press, New York.
Steinman, A., P. J. Mulholland, A. V Palumbo, T. F. Flum, J. W. Elwood, and D. L. DeAngelis. 1990. Resistance of lotic ecosystems to a light elimination disturbance: a laboratory stream study. OIKOS 58:80–90.
Stock, M. S., and A. K. Ward. 1989. Establishment of a bedrock epilithic community in a small stream: microbial (algal and bacterial) metabolism and physical structure. Canadian Journal of Fisheries and Aquatic Sciences 46:1874–1883.
Wellnitz, T., and R. B. Rader. 2003. Mechanisms influencing community composition and succession in mountain stream periphyton: interactions between scouring history, grazing, and irradiance. Journal of the North American Benthological Society 22:528–541.
103
Withers, P. J. A., B. Ulén, C. Stamm, and M. Bechmann. 2003. Incidental phosphorus losses–are they significant and can they be predicted? Journal of Plant Nutrition and Soil Science 166:459–468.
Wu, L., C. B. Blackwood, and L. G. Leff. 2009. Effect of single-species and mixed- species leaf leachate on bacterial communities in biofilms. Hydrobiologia 636:65– 76.
Ylla, I., A. M. Romaní, and S. Sabater. 2012. Labile and recalcitrant organic matter utilization by river biofilm under increasing water temperature. Microbial Ecology 64:593–604.
Zhou, J., M. A. Bruns, and J. M. Tiedje. 1996. DNA recovery from soils of diverse composition. Applied and Environmental Microbiology 62:316–322.
104
Tables
Table 1. Water quality environmental factors are represented as the mean (± SE) (N = 3) for each sampling date and deployment and for the entire study. Results of the correlation of the factors to NMDS ordination are also reported.
Date Temperature SpCond TDS pH Turbidity Phosphorus Chlorine 3- (ºC) (µS/cm) (mg/L) (NTU) (mg/L PO4 ) (mg/L) 7 Sept 16.81 ± 0.01 592.8 ± 0.9 385.4 ± 0.7 8.31 ± 0.01 8.5 ± 0.4 0.29 ± 0.05 0.02 ± 0.02 14 Sept 19.19 ± 0.04 669.8 ± 0.4 435.5 ± 0.5 8.34 ± 0.01 10.9 ± 5.3 0.31 ± 0.03 0.02 ± 0.01 22 Sept 17.32 ± 0.04 664.0 ± 0.2 431.4 ± 0.5 8.05 ± 0.05 4.5 ± 1.8 0.23 ± 0.01 0.02 ± 0.02 29 Sept 13.82 ± 0.02 617.0 ± 0.2 401.0 ± 0.0 8.37 ± 0.02 7.0 ± 0.6 0.29 ± 0.06 0.03 ± 0.01 6 Oct 12.89 ± 0.01 673.9 ± 0.3 438.0 ± 0.0 8.35 ± 0.01 3.4 ± 1.0 0.20 ± 0.02 0.03 ± 0.01 12 Oct 15.57 ± 0.03 666.0 ± 0.2 433.0 ± 0.0 8.25 ± 0.01 3.0 ± 1.4 0.14 ± 0.02 0.04 ± 0.02 25 Oct 12.10 ± 0.01 655.5 ± 0.5 426.0 ± 0.0 8.34 ± 0.05 4.4 ± 1.0 0.15 ± 0.01 0.03 ± 0.01 2 Nov 9.79 ± 0.10 680.5 ± 1.1 442.3 ± 0.7 8.36 ± 0.07 - 0.17 ± 0.05 0.02 ± 0.01 9 Nov 11.60 ± 0.04 655.2 ± 1.0 425.9 ± 0.8 8.41 ± 0.04 3.6 ± 2.7 0.02 ± 0.01 0.13 ± 0.04 21 Nov 10.43 ± 0.01 645.6 ± 0.5 419.8 ± 0.4 8.52 ± 0.03 3.1 ± 0.4 0.04 ± 0.01 0.30 ± 0.17 3 Dec 7.67 ± 0.02 603.1 ± 0.3 392.0 ± 0.0 8.46 ± 0.02 3.4 ± 0.1 0.04 ± 0.01 0.55 ± 0.35 13 Dec 6.03 ± 0.03 628.6 ± 0.5 408.6 ± 0.4 8.45 ± 0.02 2.0 ± 0.4 0.02 ± 0.01 0.17 ± 0.04
105
Mean 12.77 ± 3.99 646.0 ± 29.0 408.6 ± 0.5 8.35 ± 0.12 4.9 ± 2.7 0.16 ± 0.11 0.11 ± 0.16 R2 0.593 0.261 0.263 0.492 0.543 0.574 0.281 P 0.001 0.027 0.027 0.002 0.001 0.001 0.023
106
Date Nitrate Sulfate Nitrite Ammonia Alkalinity TSS 3- - (mg/L NO -N) (mg/L) (mg/L NO2 -N) (mg/L NH3-N) (mg/L CaCO3) (mg/L) 7 Sept 2.3 ± 0.6 43 ± 1 0.010 ± 0.002 0.10 ± 0.03 227 ± 10 23 ± 3 14 Sept 1.3 ± 0.1 46 ± 1 0.006 ± 0.001 0.25 ± 0.06 231 ± 7 18 ± 1 22 Sept 0.1 ± 0.1 42 ± 1 0.004 ± 0.003 0.08 ± 0.05 222 ± 7 18 ± 0 29 Sept 0.8 ± 0.6 39 ± 2 0.005 ± 0.001 0.11 ± 0.06 215 ± 3 12 ± 1 6 Oct 0.8 ± 0.5 43 ± 2 0.006 ± 0.001 0.27 ± 0.13 254 ± 32 18 ± 4 12 Oct 1.2 ± 0.1 42 ± 1 0.008 ± 0.001 0.09 ± 0.08 248 ± 2 7 ± 1 25 Oct 1.2 ± 0.5 - 0.010 ± 0.002 0.04 ± 0.01 249 ± 6 2 ± 1 2 Nov 1.5 ± 0.7 44 ± 0 0.006 ± 0.001 0.04 ± 0.03 256 ± 3 4 ± 1 9 Nov 0.8 ± 0.3 45 ± 0 0.009 ± 0.007 0.11 ± 0.05 247 ± 10 2 ± 1 21 Nov 0.8 ± 0.6 43 ± 9 0.007 ± 0.001 0.15 ± 0.01 262 ± 3 11 ± 1 3 Dec 1.5 ± 0.7 36 ± 1 0.007 ± 0.001 0.19 ± 0.05 245 ± 16 9 ± 2 13 Dec 2.5 ± 0.4 40 ± 1 0.007 ± 0.001 0.07 ± 0.04 238 ± 14 7 ± 2 Mean 1.2 ± 0.7 42 ± 3 0.007 ± 0.001 0.13 ± 0.08 241 ± 15 11 ± 7 R2 0.147 0.268 0.233 0.115 0.268 0.283 P 0.148 0.024 0.04 0.212 0.022 0.016
107
Figures
Figure 1. Water quality parameters were correlated with time since the start of the study.
Dashed lines when present denote significant correlations (P < 0.05). Specific conductivity was omitted in the figure because it is directly linked to total dissolved solids because it measures dissolved particles that carry an electrochemical charge and the results were nearly identical.
108
Figure 2. Biofilm characteristics of ash free dry mass (AFDM) (top) and chlorophyll a
(bottom) are overlaid on stream mean daily discharge. Red asterisks symbolize tile deployment, and tiles were deployed on 7 September 2011 for early autumn biofilms and on 25 October 2011 for late autumn biofilms. Established late autumn biofilms were deployed on 7 September 2011, and so were established in early autumn but sampled
109
during late autumn; hence, they were categorized as established late autumn biofilms.
Established late autumn biofilms were compared to late autumn biofilms using a student’s t-test and significant differences (P < 0.05) are denoted by black asterisks.
Circles designate dates with community profile data for only early autumn/established late autumn biofilms (open circles) or for both established late autumn biofilms and late autumn biofilms (closed circles).
110
Figure 3. Biofilm development of the different categories were assessed using A) time and B) undisturbed development time using accumulated degree days (ADD) as the unit of measure to account for the affect of decreasing temperature throughout the season.
Undisturbed development time was defined as the amount of time since deployment or the last disturbance event. Early autumn biofilms and late autumn biofilms were compared to each other A) to reflect differences in development based on time during the season. A linear regression of AFDM with undisturbed development time B) was used to assess differences in the rate of development between the biofilm categories. Both late autumn (R2 = 0.77, P < 0.0001) and established late autumn (R2 = 0.59, P < 0.0001) had significant linear models, but early autumn did not (R2 = 0.09, P = 0.1967). The slopes were compared to each other and stars denote a significant difference.
111
Figure 4. Bacteria biofilm community structure visualized using nonmetric multidimensional scaling (3-D, R2 = 0.98, stress = 0.14) and overlaid with biofilm
112
category and sampling date. Biofilms were categorized into three groups using a combination of when growth was initiated and the time of the season and significantly influenced patterns (PERMANOVA, pseudo-F = 3.5886, P < 0.0001). Ellipses represent standard error 95% confidence intervals. Environmental factors were correlated to the ordination, and significant vectors (P < 0.0001) are displayed. Vector direction indicates direction of the gradient and vector length is proportional to the strength of the correlation.
113
CHAPTER 4
BOTTOM-UP AND TOP-DOWN INTERACTIONS BETWEEN INVERTEBRATE
GRAZERS AND EPILITHIC BIOFILMS DICTATE BIOFILM COMMUNITY
STRUCTURE THROUGHOUT SUCCESSION
Abstract
Bottom-up and top-down forces interact to dictate ecosystem dynamics. Biofilms are micro-ecosystems that are subjected to bottom-up controls of abiotic factors during succession and top-down controls of invertebrate grazers. Our objective was to determine if the top-down influence of grazers would be mediated by the bottom-up influences of abiotic factors on biofilm succession demonstrated by preferential feeding. Epilithic biofilms were grown on unglazed porcelain tiles under ambient, modified flow (increased turbulence), dark, and modified flow plus dark treatment conditions for 7, 14, 21, and 28 days. Laboratory microcosms containing all communities were subjected to no grazing, mayfly (Maccafertium sp.), snail (Goniobasis sp.), and multi-species, snails plus mayflies, grazing for 6 days. After grazing, the grazing effect on biomass was calculated as the difference between the no grazing control and the grazing treatment and community structure of bacteria and eukaryotes was characterized using genetic
114
techniques of automated ribosomal intergenic spacer analysis (ARISA) and 454 pyrosequencing. In the absence of grazing, lighted treatments had elevated biomass levels compared to treatments under dark conditions, and modified flow had increased algal and total biomass at 21 and 28 days. Mayfly grazing significantly reduced biofilm communities that developed under modified flow conditions at 21 days (two-way
ANOVA, P < 0.05), while snails reduced all communities at 21 and 28 days the most
(two-way ANOVA, P < 0.05). When the two grazers were combined, a distinct tiered pattern emerged where modified flow communities at 21 days were significantly grazed the most, communities at 7 days were grazed the least, and the remaining communities were grazed at an intermediate level. Grazing was the most influential on both bacterial
(PERMANOVA, pseudo-F = 8.05, P = 0.0001) and eukaryotic (PERMANOVA, pseudo-
F = 10.23, P = 0.0001) communities but only bacterial communities were affected by time (PERMANOVA, pseudo-F = 1.82, P = 0.0055). Indicator analysis at the genera taxonomic level produced indicators for all levels of the grazing treatments, time of growth in streams, and the biofilm treatments. The most striking result was that there were 38 indicator taxa for the control biofilms in comparison to seven indicators for snails, three for mayflies, and eight for combined species grazing pressure. Grazing may have masked the influence of abiotic treatment effects and successional timing differences that were still apparent in the no grazing control treatment. In total, these results demonstrate that grazers had a preference for late stage biofilms that was influenced by abiotic factors and shows that grazers alter both bacterial and eukaryotic biofilm community composition.
115
Introduction
Ecological communities are governed by a multitude of processes but the interplay between bottom-up and top-down effects frequently drives food web dynamics and community structure. Bottom-up controls exert an influence through resource availability while consumption from predators or herbivores are the source of top-down control (Hairston et al. 1960, Power 1992). These processes can have cascading effects that extend to multiple trophic levels (Hunter and Price 1992). An example of a system driven by top-down control is when large mouth bass indirectly determined the stream biofilm biomass by feeding on and controlling minnow populations that grazed the biofilm (Power et al. 1985). On the other hand, bottom-up control through nutrient limitation can also influence stream biofilms (Rosemond et al. 1993). While each of these forces can be the driver of natural communities, they commonly interact and the general view is that top-down forces influence a bottom-up template because primary producers need to be present to have an ecosystem (Hunter and Price 1992).
Stream biofilms are complex communities of microorganisms that function as micro-ecosystems following the principles and theories of ecology (Battin et al. 2007,
Fierer et al. 2010). These communities are great representatives for larger ecological systems and can be investigated in a similar manner. The interplay of bottom-up and top- down controls are important components dictating biofilm community assembly and biomass (Rosemond et al. 1993, McIntyre et al. 2006, Winkelmann et al. 2013); however, succession is also a prominent pattern that governs biofilm development (McCormick and Stevenson 1991, Jackson et al. 2001, Besemer et al. 2007). Ecosystem development resulting from succession gives rise to species being replaced over time along with
116
altered physiognomy and community composition (Clements 1916, 1936). A general successional pattern for epilithic biofilms is that bacteria dominate the pioneer community (Stock and Ward 1989, Pohlon et al. 2010) and are the base layer that subsequently promotes algal attachment of diatoms (Hodoki 2005, Roeselers et al. 2007).
Then at the mature stage, filamentous and green algae are dominant (Hoagland et al.
1982, Stock and Ward 1989). All of these organisms can be present throughout succession, but their relative abundances and community composition changes as described by initial floristic composition (Egler 1954). The successional process can be affected or altered by the bottom-up influences of light, flow, and nutrients and the top- down influences of grazers. Few studies have investigated the interplay of these controls within the context of succession and whether or not bottom-up effects through biofilms can affect the top-down influence of grazing activity.
Many abiotic factors within stream ecosystems influence epilithic biofilms, and these changes can have bottom-up effects on stream dynamics because biofilms are the base of the stream food web and main source of primary production. Light and flow are two of the most important abiotic factors that can affect biofilm structure and function
(Steinman et al. 1990, Biggs and Thomsen 1995, Hill and Dimick 2002, Battin et al.
2003, Arnon et al. 2007, Lange et al. 2011). Algae is directly affected by light because it is a resource used during photosynthesis, while flow directly influences physical architecture (Battin et al. 2003, Hödl et al. 2014) and therefore community composition
(Biggs et al. 1998, Besemer et al. 2009a) and indirectly affects nutrient availability by altering nutrient diffusion and uptake rates (Horner et al. 1990, Larned et al. 2004). These
117
factors are important dictators of biofilm community composition, but they are also important within the context of grazers because they influence the grazer’s food resource.
The interaction between the community of grazers and primary producers is a key area of ecological research. Biofilms are not uniform and are patchy by nature, and grazers have been shown to respond to these differences. Caddisfly larvae (Discosmoecus gilvipes) stayed within high food resource patches longer and spent less time in recently grazed areas (Hart 1981). There was also a strong correlation between the amount of time spent in a patch and time since that patch had been grazed (Hart 1981). Baetid mayflies
(Baetis tricaudatus) demonstrated search behavior that was selective for food resource patches because movement rate was high between patches but low within patches (Kohler
1984). In addition, search intensity was affected by the quality of the patch that was left
(Kohler 1984). These studies demonstrated that grazers were following optimal foraging principles. Additionally, mayflies chose high resource rocks (Álvarez and Peckarsky
2005) even in the presence of predators (Peckarsky 1996), and biofilm abundance rather than water velocity and substrate size determined invertebrate distribution patterns
(Richards and Minshall 1988). Grazers have also responded to biofilm composition. The presence of sediment reduced overall grazing, but snails (Potamopyrgus antipodarum) preferred biofilms dominated by filamentous green algae while mayflies (Deleatidium sp.) preferred ones dominated by diatoms (Suren 2005). Clearly, grazers have the ability to sense their food resource and respond behaviorally, and this suggests that they can respond to changes driven by abiotic factors and biofilm succession.
The top down effects of grazing activity is generally a reduction of biofilm biomass and altered community composition (Feminella and Hawkins 1995, Hillebrand
118
2008), but specific effects depend on mouthpart, motility, and foraging differences of the grazers (Hart 1985, Steinman et al. 1987, Lawrence et al. 2002). Grazers typically alter the algal community by reducing larger, filamentous forms and leaving small, low stature diatoms (Steinman et al. 1987, Underwood and Thomas 1990). In fact, the results of snail grazing on algal communities were dependent upon growth forms related to stage of succession (Tuchman and Stevenson 1991). The removal of filamentous cyanobacteria arrested succession of non-diatom algae, but diatom succession was enhanced when grazers removed early stage diatoms and left small, motile late diatoms (Tuchman and
Stevenson 1991). Grazing has the ability to affect succession, but there have been few studies focused on how succession can influence grazing (Díaz Villanueva and Modenutti
2004). If biofilm composition and biomass changes throughout succession and grazers respond to their food resource, then it begs the question; can top-down effects of grazers be mediated by bottom-up forces driven by biofilm successional trajectories influenced by abiotic factors?
Microbial biofilm communities offer an outstanding opportunity to assess the relative roles of bottom-up and top-down ecosystem effects. We used an experimental design that included two procedures to address these interactions. First, biofilms developed on unglazed tiles in the stream for 7, 14, 21, and 28 days under four treatment conditions of ambient (control), modified flow, dark, and modified flow plus dark, which resulted in 16 biofilm types. The tiles were deployed at multiple time periods so that they could be harvested on the same date and brought into the lab for the second procedure.
The biofilms where then subjected to grazing treatments of none (no grazing control), snails (Goniobasis sp.), mayflies (Maccaffertium sp.), and snails plus mayflies. Patterns
119
of community composition were described using bacterial and eukaryotic community profiles generated using automatic ribosomal intergenic spacer analysis (ARISA). Next generation 454-pyrosequencing was used to further describe the taxonomic diversity of the bacterial communities. We hypothesized that (H1) bottom-up effects of epilithic biofilms mediated by succession stage and abiotic factors would influence grazing activity. We also expected (H2) grazers to have top-down influences on biofilm community composition by altering the structure of grazed communities.
Methods
Site description
The study was conducted in a lower third order river section of the upper region of the Little Miami River in a small deciduous forest corridor of the Little Miami State
Forest Preserve in Xenia, Ohio, USA (39°76.552 83°90.062). The surrounding landcover of the catchment area was predominately agriculture, and the riparian forest was dominated by maple (Acer sp.) and elm (Ulmus sp.), but included hackberry (Celtis sp.), sycamore (Plantanus sp.), and walnut (Juglans sp.). The stream has been categorized as an Exceptional Warmwater Habitat by the Ohio Environmental Protection Agency because it supported high diversity of aquatic organisms (OEPA 2002). The study was conducted in a run habitat where the substrate was predominately gravel and cobble with several intermittent boulders.
120
Environmental parameters
Environmental water quality parameters were measured at every deployment and the harvest date (see below) at upper, middle, and lower sections along the reach.
Specific conductivity (SpCond µS/cm), total dissolved solids (TDS mg/L), turbidity
(NTU), pH, and temperature (°C) were recorded using a YSI 6600 v2 Sonde (YSI Inc,
Yellow Springs, OH) every 30 sec for 10 min, and water was collected to measure
3- 3- - phosphorus (mg/L PO4 ), chlorine (mg/L), nitrate (mg/L NO -N), nitrite (mg/L NO2 -N), sulfate (mg/L), ammonia (mg/L NH3-N), alkalinity (mg/L CaCO3), and total suspended solids (TSS mg/L) in the lab (Hach® Company, Loveland, CO). Also, depth was measured at each experimental board (see below).
Biofilm development
Epilithic biofilms developed naturally on hexagonal unglazed porcelain tiles (N =
6) attached to brick pavers (19.2 x 9 x 1.3 cm) with 100% silicone. Pavers were attached to wood boards (1.2 x 0.3 x 0.02 m) with VELCRO® tape under four treatments: ambient
(control), modified flow, dark, and modified flow plus dark (Figure 1). The ambient condition was subjected to natural light and flow conditions and served as the control.
The modified flow treatment was created by placing seven, 3.8 cm metal screws upstream of the paver and spaced between the tiles (Figure 1). A tarp placed above the paver created the dark treatment by reducing direct light by 97%, and the modified flow plus dark treatment contained both screws and tarp (Figure 1). Pavers for each treatment
(N=4) were randomly deployed every seven days beginning 20 July 2011 and collected at once on 17 August 2011. The result was 16 unique biofilms: the four abiotic treatments at
121
7, 14, 21, and 28 days of growth (Figure 2). Biofilms were transported to the lab and placed at 8°C overnight before being distributed to microcosms the next day (see below).
No samples were processed at this point and so there were no controls of biofilms before they were placed in the microcosms.
The ability to generate various biofilms required an experimental design of multiple tile deployments and one harvesting event. While the stream conditions were not identical throughout the growth period and there was a small storm that elevated stream level for a few days, this was not a primary concern because stream ecosystems are naturally heterogeneous. The importance was placed on creating different communities that represent multiple stages of succession under different abiotic conditions, rather than whether or not the successional trajectories were identical. The alternative option would have been to deploy all biofilms at once and then harvest them at multiple dates. But this was deemed unacceptable because biofilms would have needed to be stored in the freezer until the grazing treatment, and we wanted live biofilms.
Preferential biofilm grazing
Microcosms containing tiles (N = 2) of all 16 unique biofilms were subjected to four grazing treatments for 6 days: none (no grazing control), snails (Goniobasis sp.; N =
3), mayflies (Maccaffertium sp.; N = 6), and a combination treatment with both. These grazing invertebrates were chosen because they were the most abundant grazers during this time and numbers used reflect natural stream densities (KJ. Gorbach, unpublished data). All grazing treatments had N = 3 microcosms, but the control had N = 2.
Invertebrates were collected the day of biofilm harvesting and starved until introduction
122
to microcosms (~24 h). Microcosms were placed into environmentally controlled chambers set at 19°C to reflect stream temperature with 12 hr alternating light/dark cycle to allow equal opportunity of diurnal and nocturnal feeding behavior. Filter sterilized stream water was circulated by water pumps and refreshed daily to maintain consistent levels. After 6 days, tiles were removed and stored at -20°C until further processing.
Biofilm processing
Biofilms were processed at study completion after the grazing treatment (Figure
2) so there are no reference control communities that were processed immediately after being removed from the stream. Biofilm biomass was removed using a sterile razor blade and toothbrush and suspended in ultrapure water (NANOpure II; Barnstead, Boston, MA,
USA). Two sub samples were collected on GB-140 glass membrane filters (diameter,
25mm; pore size, 0.4 µm; Sterlitech, Kent, WA, USA) to determine total biomass as ash free dry mass (AFDM) and algal-associated biomass as chlorophyll a multiplied by the conversion factor of 67 (APHA 1999) following techniques used by (Steinman et al.
2007). A third subsample was collected for DNA extractions, and these filters were stored in 90% ethanol at -80°C until extraction. Although for the DNA extraction filter, we highly suggest using a mixed nitrocellulose filter and placing a few drops of ethanol on the sample that evaporates before storage at -80°C.
DNA extraction
DNA was extracted from dried filters following ethanol evaporation using a combination of methods from Miller et al., (1999) and Zhou et al., (1996) as suggested by
123
Lear et al., (2010). Samples were lysed by bead beating (0.5 g each of 0.1 mm and 0.5 mm glass lysis beads; RPI, Mount Prospect, IL, USA) for 15 min on a horizontal vortex adaptor (MO BIO Laboratories, Carlsbad, CA, USA) at full speed in 1.2 mL of extraction buffer (100 mM Tris-HCl [pH 8.0], 100 mM EDTA disodium salt [pH 8.0], 100mM sodium phosphate [pH 8.0], 1.5 M sodium chloride, 1% CTAB), 12 µL proteinase K (20 mg/mL), and 30 µL SDS (20%). The mixture was incubated at 65°C for 1 hr with gentle end-over-end inversions by hand every 15 min. Then, 4 µL of RNase (100 mg/µL) was added. DNA was isolated from organic debris with chloroform/isoamyl alcohol extraction and was precipitated overnight at -20°C with isopropanol. The mixture was warmed to 37°C to dissolve salt precipitates, and DNA was pelleted at 15,000 g for 30 m.
The DNA pellet was washed twice with ice cold 70% EtOH and dissolved in 25-50 µL ultrapure water (NANOpure II; Barnstead, Boston, MA, USA). Samples were purified using PowerClean® DNA Clean-Up Kit (MO BIO Laboratories, Carlsbad, CA, USA) with a modified protocol for low amounts of DNA to remove PCR inhibitors typical of biofilm samples.
Samples used in genetic analyses
Only ambient and modified flow biofilms were used in genetic analyses because dark and modified flow plus dark biofilms did not produce quality extractions due to low biomass. In addition, there was no measureable grazing activity of biofilms that developed in dark conditions and we wanted to focus on the influence of grazers on community dynamics. Unfortunately, day 7 biofilms also produced unsuccessful
124
reactions so the only samples used in ARISA and pyrosequencing were biofilms at 14, 21 and 28 days of development that were exposed to light.
ARISA
Bacterial and eukaryotic communities were assessed using profiles created by automated ribosomal intergenic spacer analysis (ARISA) (Fisher and Triplett 1999), which generates a unique ‘fingerprint’ of microbial communities using the 16S-23S intergenic space in bacteria and the ITS1-5.8S-ITS2 region in eukaryotes (Ranjard et al.
2001). While ARISA does not taxonomically identify organisms like next generation sequencing methods, this method generates community profiles and produces similar patterns in results (Bienhold et al. 2012, van Dorst et al. 2014). Approximately 15-20 ng of DNA quantified by spectrophotometer (NanoPhotometerTM Pearl; Denville Scientific
Inc., South Plainfield, NJ, USA) was amplified by PCR using 25 µL GoTaq® Colorless
Master Mix (Promega, Madison, WI, USA) with 0.5 µM of forward and reverse primers.
Bacteria ribosomal intergenic space regions were amplified with primers ITSF (5’-
GTCGTAACAAGGTAGCCGTA-3’) labeled with FAM at the 5’ end (IDT, Coralville,
IA, USA) and ITSReub (5’-GCCAAGGCATCCACC-3’) (Cardinale et al. 2004).
Eukaryote ribosomal intergenic space regions were amplified with primers 2234C (5’-
GTTTCCGTAGGTGAACCTGC-3’) labeled with ATTOTM 550 (IDT, Coralville, IA,
USA) and 3126T (5’- ATATGCTTAAGTTCAGCGGGT-3’) (Ranjard et al. 2001).
While the eukaryotic primer set has typically been used for soil fungal communities
(Ranjard et al. 2001), sequenced clones of fragments from freshwater biofilms revealed it targets various algae and ciliates, and therefore can be used to describe the general
125
eukaryotic community (Fechner et al. 2010). Fragments were created with the following
PCR conditions: (i) 94°C for 3 min, (ii) 35 cycles of 94°C for 1 min, 56°C (57.5°C for eukaryotes) for 1 min, 72°C for 2 min, and finally (iii) 72°C for 10 min (Fechner et al.
2010). Equal volumes of bacterial and eukaryotic PCR products from a sample were combined and sent to DNA Analysis, LLC (Cincinnati, OH, USA) for multiplexed fragment analysis on an ABI 3100 (Life Technologies, Carlsbad, CA, USA). Fragments were interpreted using Genescan v 3.7 using the Local Southern Size Calling Method with a peak height threshold of 100 fluorescence units to remove background fluorescence and formulated using GeneMapper v 2.5 (Life Technologies, Carlsbad, CA,
USA).
Fragment peak length and area was converted to column format using the treeflap
Excel Macro (http://www.wsc.monash.edu.au/~cwalsh/treeflap.xls) and processed with the automatic_binner script to determine binning window size and the interactive_binner script to determine the best starting window position (Ramette 2009) in R v 3.1.0 (R Core
Team 2014). This method was used to account for inherent imprecision of analyzer machines. Peak area was converted to relative abundance of each fragment as part of the entire sample, fragments with < 0.09% relative abundance were removed (Ramette 2009), and window size was calculated to be 1 base pair.
Pyrosequencing
In order to provide taxonomic descriptions of bacteria in a way to supplement the
ARISA community profiles, we employed 454-pyrosequencing by using methods of
Pechal et al. (2014) and Benbow et al. (in press) and described briefly here. Bacterial
126
diversity was determined using amplicon pyrosequencing (Dowd et al. 2008a) at an offsite lab (MR DNA, Shallowater, TX, USA) as previously described (Dowd et al.
2008b, Swanson et al. 2011, Handl et al. 2011). Samples were pooled using equal amounts of DNA from replicates that were quantified using Quant-iT TM PicoGreen® dsDNA assay kit (Invitrogen, Carlsbad, CA, USA). A unidirectional barcoding strategy using the forward primer was utilized to multiplex samples. Universal Eubacterial primers Gray28F (5’TTTGATCNTGGCTCAG) and Gray519r (5’
GTNTTACNGCGGCKGCTG) amplified the bacterial V1-3 regions of the 16S rRNA using HotStarTaq Plus Master Mix Kit (Qiagen, Valencia, CA, USA) with the following single-step PCR protocol: (i) 94°C for 3 min, (ii) 28 cycles of 94°C for 30 sec, 53°C for
40 sec, 72°C for 1 min, and finally (iii) 72°C for 5 min. Amplicon products were purified using Agencourt Ampure beads (Agencourt Bioscience Corporation, MA, USA) and sequenced with Roche 454 FLX titanium instruments and reagents following manufacturer’s guidelines. Sequence data was processed with a proprietary analysis pipeline (www.mrdnalab.com, MR DNA, Shallowater, TX, USA). All barcodes, primers, short sequences (<200bps), sequences with ambiguous base calls, and sequences with homopolymer runs exceeding 6 bp were removed. Sequences were then denoised and chimeras removed. Operational taxonomic units clustered at 97% similarity were classified using BLASTn against a curated GreenGenes database (DeSantis et al. 2006).
127
Statistical analyses
The influence of date on water quality characteristics measured during biofilm development was analyzed with an analysis of variance (ANOVA) in GraphPad Prism v
5.0 (GraphPad Software Inc, San Diego, CA, USA).
Biofilm biomass characteristics were analyzed using a two-way ANOVA with
Bonferroni multiple comparisons to test factors of treatment, time, and their interaction.
All characteristics were measured at study completion after the grazing treatment (Figure
2) so there are no reference control communities that were processed immediately after being removed from the stream. While it is likely the biofilms changed during this time period, this design was necessary to accurately determine the effect of grazing. The amount of biomass loss due to grazing was determined by subtracting the biomass values of the invertebrate grazing treatments from the average of the no grazing control. We are also assuming the decrease is due mainly to consumption. There is the possibility of non- consumptive loss, but the recirculating flow was gentle and no-consumptive losses due to grazing activity still reflects preference because the grazers had to be present. Net biomass loss was chosen as the response metric rather than percent decrease because it is more reflective of grazing effort. Biomass levels of the different communities in the no grazing control were not equal (Figure 3); therefore, a greater amount of material had to be removed from late stage biofilms than early stage biofilms to achieve 50% loss. We wanted a metric that reflected this difference in grazing effort. All analyses were performed within GraphPad Prism v 5.0 (GraphPad Software Inc, San Diego, CA, USA).
Microbial community patterns generated by ARISA were visualized using nonmetric multidimensional scaling (NMDS) with Bray-Curtis (Sørensen) distance on
128
data that was arcsine square root transformed. Rare fragments present in only one sample were removed in an effort to reduce the number of zeros in the dataset and minimize noise. NMDS was used because it is a nonparametric multivariate approach useful in evaluating nonlinear relationships of data with high numbers of zeros by using ranked distances (McCune and Grace 2002). The final configuration was determined following
Peck (2010) and where NMDS was run with three axes, 250 runs with real data, unstability criterion of 0.00001, and 250 Monte Carlo runs of random data. The two axes with the best orthogonality or lowest stress were used for representation. Outliers were identified as samples outside two standard deviations of the mean of average distances of all samples, but were not removed until NMDS confirmation that the sample had a substantial influence on the results (McCune and Grace 2002). There were four samples each with bacteria and eukaryotes that were removed as outliers. The effects of grazing treatment, time of growth in the stream, biofilm treatment, and their interaction were tested with nonparametric multivariate analysis of variance (PERMANOVA) (Anderson
2001) using the vegan v 2.2-1 package (Oksanen et al. 2015) in R v 3.1.0 (R Core Team
2014). This test was chosen because it takes into account interactions between the time and treatment factors. Analyses, unless designated otherwise, were conducted using PC-
ORD 6 (MjM Software, Gleneden Beach, OR, USA).
Bacterial communities described by 454-pyrosequencing were analyzed at both the phylum and genus level using non-rarefied libraries. We were interested in retaining all data because samples were pooled replicates, and rarefying omits available data
(McMurdie and Holmes 2014). Groups for analyses were created using samples with the designated factor type. Data was converted into relative abundance within sample and
129
analyzed by indicator analysis to determine which taxa were indicative of the different factors using labdsv v 1.6-1 package (Roberts 2013) in R v 3.1.0 (R Core Team 2014).
Results
Biomass patterns of the no grazing control
In-stream biofilm development was obvious over the course of the 28 day field experiment (Figure 2) and effects of flow and light treatments were clearly influential
(Figure 3). Biofilms are labeled by the amount of time they grew in the natural stream setting even though it includes the 6 day grazing period within the microcosoms. The no grazing control provides a reference for what biomass levels would have been without grazing and seems to reflect levels observed when biofilms were removed from the stream. Beginning at day 14, treatment effects became statistically distinguishable between light and dark conditions (Figure 3). Dark treatment biomass levels (both dark alone and modified flow plus dark) exhibited significantly lower biomass than the light treatments from day 14 to day 28 which seems to be a reflection of the missing algal component (Figure 3). The modified flow treatment resulted in apparently elevated total and algal biomass levels on days 21 and 28; however, it was only statistically significant for total biomass at 21 days. The overall pattern (Figure 3) of the no grazing control suggests that (a) light was more important than modified flow and that (b) modified flow increased biomass in an ambient light environment.
130
Biomass patterns of the grazing treatments
The bottom-up effects of biofilms influenced grazer activity as demonstrated by patterns of net biomass loss (Figure 4). We report data for only biofilms that developed under lighted conditions because biomass of dark and modified flow plus dark biofilms were very low and there was no measurable grazing activity. This indicates the importance of the algal community as a food resource. Later stage biofilms (day 21 and
28) were reduced the most by snails, but mayflies reduced day 21 biofilms the most
(Figure 4). Interestingly, a distinct pattern of biomass loss emerged when snails and mayflies grazed together where modified flow biofilms at day 21 were grazed the most and biofilms at day 7 were grazed the least. This was a robust pattern that was consistent for both algal and total biomass. Even when considering the percent decrease, modified flow day 21 biofilms had the greatest decrease (Figure 4). The results of percent decrease were not statistically significant (two-way ANOVA, P > 0.05), but this still indicates that modified flow biofilms at 21 days of growth were preferred.
Bacterial and eukaryotic community profile patterns
Bacterial and eukaryotic biofilm community composition (as estimated using
ARISA data) were related to time and grazing type (Table 2). Both bacteria (3-D, R2 =
0.880, stress = 0.13; Figure 5) and eukaryotes (3-D, R2 = 0.744, stress = 0.15; Figure 6) were most strongly influenced by the grazing treatment (PERMANOVA; Table 2).
Grazing was the only parameter to influence eukaryotes, while the bacterial community was also significantly influenced by time and the interaction between time x grazing and time x biofilm (Table 2). This suggests that either the bacterial community changed more
131
drastically throughout succession or that grazing impacts on the eukaryotic community overshadowed successional changes. Grazing appeared to have different effects on the bacterial and eukaryotic communities that are highlighted by grazing pressure (Figure 5;
Figure 6). Within bacteria, each grazing treatment is separated, however, the combination grazing treatment is the farthest from the no grazing control (Figure 5). In addition, grazing pressure shows a distinct gradient from zero grazers to two grazers. This indicates that snails and mayflies have different effects on the biofilm bacterial community but that their effects are additive. The pattern is slightly different for the eukaryotic community (Figure 6). The no grazing control is distinctly separated from all grazing treatment, and each grazing treatment forms a distinct cluster when viewed in three dimensions. There is a lack of grazer pressure gradient, but the profiles are still grouped together based on the number of grazers present. This suggests that snails and mayflies have different effects on the eukaryotic community but that the presence of any grazing is the most important. Grazers influenced the bacterial and eukaryotic communities differently, but they also had different effects themselves that could reflect grazing mechanisms or their own microbiome.
Bacterial community composition and diversity
Pyrosequencing was used to provide taxonomic information of the biofilm bacterial communities. There were a few trends when comparing the relative abundances of the genera between the no grazing control and any of the grazing treatments. The grazing treatments were averaged into an “all grazing” category for easier comparison
(Table 3). The two most abundant genera in the no grazing control (Rhodobacter,
132
Flavobacterium) were less in comparison to all grazing, and the two most abundant genera in the grazing treatments (Nannocystineae and Balneatrix) were greater than the no grazing control (Table 3). Rhodobacter decreased from 8.16% ± 3.37 without grazing to 4.11% ± 1.28 with grazing, and Flavobacterium showed a similar decrease from
8.05% ± 3.47 to 3.8% ± 1.07. Grazing increased the average relative abundance of
Nannocystineae from 2.28% ± 0.88 in the no grazing control to 12.43% ± 2.78 in all grazing treatments. A similar trend occurred with Balneatrix where it increased from
1.24% ± 0.34 to 9.36% ± 2.82 with grazing.
Indicator analysis produced groups indicative of the different treatment categories at both phylum and genus taxonomic levels. The Actinobacteria phylum (IV = 0.3175, P =
0.036) was an indicator of the no grazing control, and the mayfly treatment was indicated by SPAM (candidate division) (IV = 0.5000, P = 0.044). SPAM is an uncultured newly classified division (Lipson and Schmidt 2004) and has been found in alpine soil (Lipson and Schmidt 2004), Minnesota farm soil (Tringe et al. 2005), a Spanish coastal aquifer
(López-Archilla et al. 2007), and soil near a Himalayan glacier (Shivaji et al. 2011) to name a few. For the amount of time biofilms grew in the stream, day 21 was indicated by
Planctomycetes (IV = 0.4262, P = 0.038) and Bacteriodetes (IV = 0.3879, P = 0.048), and
Deinococcus (IV = 0.5436, P = 0.002) was indicative of day 28. There were no indicators for the biofilm treatment. The genus level generated indicators for all levels of the grazing treatments (Table 3), the time (days) of growth in stream (Table 4), and the biofilm treatments (Table 4), but all of the indicators could be considered non abundant taxa. The majority of the indicators were < 0.2% relative abundance within a sample and all were < 0.5%. The most striking result was that there were 38 indicators for the no
133
grazing control in comparison to 7 for snails, 3 for mayflies, and 8 for snails plus mayflies. Out of the 38 indicators for no grazing, 35 were present within at least one sample of all other grazing treatments and only one genus, Paenibacillus, was present only in the no grazing control. Chondromyces and Flavobacteriaceae were indicators of the snail treatment that were only present in the snail treatment, and Melitea and SPAM
(candidate division) were similarity indicators of mayfly treatment only present in that treatment.
Discussion
Bottom-up effects of biofilm communities on grazers
Bottom-up effects of primary producers on grazers and higher-level food web components are well-known from a variety of systems. Biofilms are a good model system for understanding bottom-up effects and we found that grazers exhibited a clear responses to bottom-up forces on the biofilm community. In particular, snails and mayflies when grazed together exhibited a preference for both succession stage (21 days) and environmental condition (modified flow). Why this particular community was selected requires further investigation, but there are a few aspects to be considered. First, optimal foraging theory suggests that grazers would choose communities that provide the most energy with the least cost to the grazer (Schoener 1971, Pyke et al. 1977). If biomass is a surrogate for energy, then we would predict that the most heavily grazed biofilm would be the one with the highest biomass. In fact, snails and caddisflies grew 5x and 7x, respectively, faster on high-biomass biofilms compared to low-biomass biofilms (Hill et al. 1992). The nutrient stoichiometry could also be important because mayfly growth rate
134
was affected by carbon to phosphorus ratio (Frost and Elser 2002), and it would not be surprising if these rates changed throughout succession.
Yet, modified flow biofilms at 28 days had the same level of biomass so this suggests that biomass alone was not the only factor influencing grazing activity and that biofilm community composition played a role. Community composition changes throughout succession, and grazing preference based on successional stage in our study is supported by previous findings (Díaz Villanueva and Modenutti 2004). It was suggested that the “herbivore-periphyton” interaction may depend on successional state because grazing effects on biomass and algal community differed based on successional timing
(Díaz Villanueva and Modenutti 2004). Algae become established later during succession and different forms of diatoms and filamentous algae are more easily utilized by different grazers (Lamberti et al. 1987), suggesting that certain stages of succession may be utilized easier by grazers. In addition, our study had an added component of different successional trajectories created by bottom-up effects. Modified flow treatment increased heterogeneity of the flow regime and this may have increased niche availability and therefore diversity, possibly increasing the quality of the biofilm as a resource for grazers
(Besemer et al. 2007, 2009b). Algae were not taxonomically identified in this study, but algal composition may prove to be an important factor that influences grazer response to biofilms. In summary, changes in light and flow were used to experimentally create varying biofilms, and grazers responded positively to modified flow at later stages of succession. This suggests that changes in epilithic biofilms driven by succession and abiotic factors can have bottom-up effects on the stream food-web as demonstrated by the response of invertebrate grazing activity.
135
Top-down effects of grazers on biofilm community composition
Grazing treatment drove the patterns in community composition ARISA profiles of bacteria and eukaryotes indicating top-down effects were more prominent than bottom-up effects in driving community composition. When considering the influence of snails and mayflies, each had a unique effect because the treatments were consistently separated. This may be a reflection of grazing mechanisms or preference because snails have radula mouthparts that have fine teeth and feed in a rasping motion (Barnese et al.
1990) while mayflies have mandibles that can produce a precise pinching motion. Studies have documented that snails and mayflies affect biofilms differently. Snails significantly removed large rosette-forming diatoms and a cyanophyte with loose filaments while mayflies reduced particular diatoms without changing abundance of filaments (Lamberti et al. 1987). This corroborates the finding that snails left adnate diatoms and short filaments of Stigeoclonium tenue (Steinman et al. 1987). The physical effects are also different because snails reduced biofilm thickness to 11 µm ± 5.2 while mayflies reduced thickness to 42 µm ± 9 (Lawrence et al. 2002). In addition, mayflies demonstrated selective grazing because diatom communities in feces were different compared to the source communities (Peterson et al. 1998). Our findings further support that grazers differ in their effect on biofilm community composition, but they also demonstrate that these effects occur in both the bacterial and eukaryotic populations.
Grazer effects on the bacterial and eukaryotic communities were unique as demonstrated by different configurations in the NMS ordinations. Grazers seemed to have a cumulative effect on the bacterial community, while any grazing was clearly separated from the no grazing control within the eukaryotic community. It is possible that
136
the eukaryotic community reflects grazing mechanisms or preference while the bacterial community reflects secondary grazer effects. A distinct bacterial layer remains after grazing (Lawrence et al. 2002). Both algae and bacteria will be assimilated, although the proportions would be slightly more algal (Morales and Ward 2000). Algae are considered to be the main resource that is grazed upon while bacteria are the bystanders that are inadvertently consumed. This is reflected in the fact that virtually all previous grazing studies have focused on the algal community (Feminella and Hawkins 1995, Hillebrand
2008); however, technologies to investigate the bacterial community were not readily available during these studies. Bacteria are an important component within epilithic biofilms that has been neglected when studying grazing effects. It is possible that communities were influenced secondarily through regurgitation or defecation; simply put, the grazers may be inoculating the biofilms with their own microbiome. This is a previously unstudied topic that can be investigated more readily with the advancement of sequencing technologies.
Implications for grazing in the natural environment
The bottom-up effects of epilithic biofilms mediated by abiotic factors and successional stage impacted invertebrate grazing. This suggests that there are nutritional differences associated with biofilm succession and that abiotic factors can impact the food web by altering the biofilm food resource. This also suggests that biomass removed by disturbances like spate events would create inferior quality biofilms because they would have low biomass (Hill et al. 1992). Another implication of grazers preferring late stage biofilms is that effects determined by enclosed manipulation studies may be
137
misrepresentative. These studies typically allowed a period of approximately 6-14 days of biofilm pre-growth before grazers were added and did not consider successional stage
(Steinman McIntire Lowry 1987; Steinman et al 1987). If grazers do not naturally feed on early stage biofilms, then overconsumption or altered foraging behavior may have resulted in response to a low quality food resource. Most of these studies allowed much longer grazing durations between 20-50 days so that may have allowed effects to balance out, but grazing based on successional timing in the natural setting is an important consideration. In fact, we observed anecdotal evidence of selective grazing of late stage biofilms. Many modified flow biofilms at 28 days had visible scrapings on them (Figure
1), and the patterns suggest crayfish are responsible. While the origin of these scrapings is not conclusive, this is still irrefutable evidence that preferential grazing of late stage biofilms occurred in the natural setting.
Conclusion and future directions
The relationship between grazers and biofilms may be more complex than originally thought, and our results suggest that biofilms and grazers influence each other.
First, preferentially grazing of select biofilm indicates that bottom-up effects through biofilm successional trajectory can modify grazing behavior. The most distinct pattern emerged when snails and mayflies grazed together. Why the particular biofilm of modified flow at 21 days was reduced the most still needs to be elucidated, but it may be a combination of biofilm biomass and a diverse community composition resulting from heterogeneous flow. Second, top-down effects from grazers themselves were different and the effects modified bacterial and eukaryotic communities differently. Eukaryotic
138
community patterns may have been driven by consumption or preference, while bacterial community patterns could include a combination of consumption effects and inoculation by grazer microflora. The idea that grazers can alter the composition of their food resource by adding additional microbes has not been well studied, but it can now be addressed with the advancements in sequencing technology.
The implications for these results suggest that there may be a timing component to bottom-up and top-down controls throughout biofilm succession. The bottom-up factors may be more important during the beginning of succession when grazers appear indifferent, while top-down effects from grazers may become more important later during succession. This may mean that grazer effects in the natural setting are slightly different than suggested by previous grazing manipulations because grazers were provided early stage biofilms. In addition, the light resource was required to elicit a grazing response and supports the idea that top-down effects influence a bottom-up template (Hunter and Price
1992).
139
Literature Cited
Álvarez, M., and B. L. Peckarsky. 2005. How do grazers affect periphyton heterogeneity in streams? Oecologia 142:pp. 576–587.
Anderson, M. J. 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecology 26:32–46.
APHA. 1999. Standard methods for the evaluation of water and wastewater, 20th edition. American Public Health Association, Washington, DC.
Arnon, S., A. I. Packman, C. G. Peterson, and K. A. Gray. 2007. Effects of overlying velocity on periphyton structure and denitrification. Journal of Geophysical Research 112:G01002.
Barnese, L. E., R. L. Lowe, and R. D. Hunter. 1990. Comparative grazing efficiency of pulmonate and prosobranch snails. Journal of the North American Benthological Society 9:35–44.
Battin, T. J., L. A. Kaplan, J. D. Newbold, X. Cheng, and C. M. E. Hansen. 2003. Effects of current velocity on the nascent architecture of stream microbial biofilms. Appl Environ Microbiol 69:5443–5452.
Battin, T. J., W. T. Sloan, S. Kjelleberg, H. Daims, I. M. Head, T. P. Cutis, and L. Eberl. 2007. Microbial landscapes: new paths to biofilm research. Nature Reviews: Microbiology 5:76–81.
Benbow, M., J. Pechal, J. Lang, R. Erb, and J. Wallace. (n.d.). The potential of high- throughput metagenomic sequencing of aquatic bacterial communities to estimate the postmortem submersion interval. J. Forensic Sci.
Besemer, K., I. Hödl, G. Singer, and T. J. Battin. 2009a. Architectural differentiation reflects bacterial community structure in stream biofilms. ISME J 3:1318–1324.
Besemer, K., G. Singer, I. Hödl, and T. J. Battin. 2009b. Bacterial community composition of stream biofilms in spatially variable-flow environments. Appl Environ Microbiol 75:7189–7195.
Besemer, K., G. Singer, R. Limberger, A.-K. K. Chlup, G. Hochedlinger, I. Hödl, C. Baranyi, and T. J. Battin. 2007. Biophysical controls on community succession in stream biofilms. Applied and Environmental Microbiology 73:4966–4974.
Bienhold, C., A. Boetius, and A. Ramette. 2012. The energy-diversity relationship of complex bacterial communities in Arctic deep-sea sediments. ISME J 6:724–732.
140
Biggs, B. J. F., D. G. Goring, and V. I. Nikora. 1998. Subsidy and stress responses of stream periphyton to gradients in water velocity as a function of community growth form. J. Phycol. 34:598–607.
Biggs, B. J. F., and H. A. Thomsen. 1995. Disturbance of stream periphyton by perturbations in shear stress time to structural failure and differences in community resistance. Journal of Phycology 31:233–241.
Cardinale, M., L. Brusetti, P. Quatrini, S. Borin, A. M. Puglia, A. Rizzi, E. Zanardini, C. Sorlini, C. Corselli, and D. Daffonchio. 2004. Comparison of different primer sets for use in automated ribosomal intergenic spacer analysis of complex bacterial communities. Applied and Environmental Microbiology 70:6147–6156.
Clements, F. E. 1916. Plant succession: Analysis of the development of vegetation. Carnegie Institute of Washington, Washington D.C.
Clements, F. E. 1936. Nature and structure of the climax. Ecology 24:252–284.
DeSantis, T. Z., P. Hugenholtz, N. Larsen, M. Rojas, E. L. Brodie, K. Keller, T. Huber, D. Dalevi, P. Hu, and G. L. Andersen. 2006. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Applied and Environmental Microbiology 72:5069–72.
Díaz Villanueva, V., and B. Modenutti. 2004. Experimental analysis of grazing by the mayfly Meridialaris chiloeensis on different successional stages of stream periphyton. International Review of Hydrobiology 89:263–277.
Van Dorst, J., A. Bissett, A. S. Palmer, M. Brown, I. Snape, J. S. Stark, B. Raymond, J. McKinlay, M. Ji, T. Winsley, and B. C. Ferrari. 2014. Community fingerprinting in a sequencing world. FEMS Microbiology Ecology 89:316–330.
Dowd, S. E., Y. Sun, and P. R. Secor. 2008a. Survey of bacterial diversity in chronic wounds using Pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiology 8:43.
Dowd, S. E., Y. Sun, P. R. Secor, R. D. Wolcott, A. Domingo, and J. A. Carroll. 2008b. Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: bacterial diversity in the ileum of newly weaned Salmonella infected pigs. Foodborne Pathogens and Disease 5:1–14.
Egler, F. E. 1954. Vegetation science concepts I. Initial floristic composition, a factor in old-field vegetation development with 2 figs. Vegetatio 4:412–417.
Fechner, L. C., F. Vincent-Hubert, P. Gaubert, T. Bouchez, C. Gourlay-Francé, and M.- H. Tusseau-Vuillemin. 2010. Combined eukaryotic and bacterial community
141
fingerprinting of natural freshwater biofilms using automated ribosomal intergenic spacer analysis. FEMS Microbiology Ecology 74:542–53.
Feminella, J. W., and C. P. Hawkins. 1995. Interactions between stream herbivores and periphyton: a quantitative analysis of past experiments. Journal of the North American Benthological Society 14:465–509.
Fierer, N., D. Nemergut, R. Knight, and J. M. Craine. 2010. Changes through time: integrating microorganisms into the study of succession. Research in Microbiology 161:635–642.
Fisher, M. M., and E. W. Triplett. 1999. Automated approach for ribosomal intergenic spacer analysis of microbial diversity and its application to freshwater bacterial communities. Applied and Environmental Microbiology 65:4630–4636.
Frost, P. C., and J. J. Elser. 2002. Growth responses of littoral mayflies to the phosphorus content of their food. Ecology Letters 5:232.
Hairston, N. G., F. E. Smith, and L. B. Slobkin. 1960. Community structure, population control and competition. American Naturalist 94:421–425.
Handl, S., S. E. Dowd, J. F. Garcia-Mazcorro, J. M. Steiner, and J. S. Suchodolski. 2011. Massive parallel 16S rRNA gene pyrosequencing reveals highly diverse fecal bacterial and fungal communities in healthy dogs and cats. FEMS Mcrobiology Ecology 76:301–10.
Hart, D. D. 1981. Foraging and resource patchiness: field experiments with a grazing stream insect. Oikos:46–52.
Hart, D. D. 1985. Grazing insects mediate algal interactions in a stream benthic community. Oikos 44:40–46.
Hill, W., S. Dimick, E. S. Division, and O. Ridge. 2002. Effects of riparian leaf dynamics on periphyton photosynthesis and light utilisation efficiency. Freshwater Biology:1245–1256.
Hill, W. R., S. C. Weber, and A. J. Stewart. 1992. Food limitation of two lotic grazers: quantity, quality, and size-specificity. Journal of the North American Benthological Society 11:420–432.
Hillebrand, H. 2008. Grazing regulates the spatial variability of periphyton biomass. Ecology 89:165–173.
Hoagland, K. D., S. C. Roemer, J. R. Rosowski, and R. Rosowski. 1982. Colonization and community structure of two periphyton assemblages, with emphasis on the diatoms (Bacillariophyceae). American Journal of Botany 69:188–213.
142
Hödl, I., L. Mari, E. Bertuzzo, S. Suweis, K. Besemer, A. Rinaldo, and T. J. Battin. 2014. Biophysical controls on cluster dynamics and architectural differentiation of microbial biofilms in contrasting flow environments. Environmental Microbiology 16:802–12.
Hodoki, Y. 2005. Bacteria biofilm encourages algal immigration onto substrata in lotic systems. Hydrobiolgia 539:27–34.
Horner, R. R., E. B. Welch, M. R. Seeley, and J. M. Jacoby. 1990. Responses of periphyton to changes in current velocity, suspended sediment and phosphorus concentration. Freshwater Biology 24:215–232.
Hunter, M. D., and P. W. Price. 1992. Playing chutes and ladders: Heterogeneity and the relative roles of bottom-up and top-down forces in natural communities. Ecology 73:724–732.
Jackson, C. R., P. F. Churchill, and E. E. Roden. 2001. Successional changes in bacterial assemblage structure during epilithic biofilm development. Ecology 82:555–566.
Kohler, S. 1984. Search mechanism of a stream grazer in patchy environments: the role of food abundance. Oecologia 62:209–218.
Lamberti, G. A., L. R. Ashkenas, S. V Gregory, A. D. Steinman, S. Journal, N. American, B. Society, and N. Jun. 1987. Effects of three herbivores on periphyton communities in laboratory streams. Journal of the North American Benthological Society 6:92–104.
Lange, K., A. Liess, J. J. Piggott, C. R. Townsend, and C. D. Matthaei. 2011. Light, nutrients and grazing interact to determine stream diatom community composition and functional group structure. Freshwater Biology 56:264–278.
Larned, S. T., V. I. Nikora, and B. J. F. Biggs. 2004. Mass-transfer-limited nitrogen and phosphorus uptake by stream periphyton: A conceptual model and experimental evidence. Limnol. Oceanogr. 49:1992–2000.
Lawrence, J. R., B. Scharf, G. Packroff, T. R. Neu, and J. R. Lawerence. 2002. Microscale evaluation of the effects of grazing by invertebrates with contrasting feeding modes on river biofilm architecture and composition. Microbial Ecology 44:pp. 199–207.
Lear, G., Y. Dong, and G. Lewis. 2010. Comparison of methods for the extraction of DNA from stream epilithic biofilms. Antonie van Leeuwenhoek 98:567–571.
Lipson, D. A., and S. K. Schmidt. 2004. Seasonal changes in an alpine soil bacterial community in the Colorado Rocky Mountains. Applied and Environmental Microbiology 70:2867–2879.
143
López-Archilla, A. I., D. Moreira, S. Velasco, and P. López-García. 2007. Archaeal and bacterial community composition of a pristine coastal aquifer in Donana National Park, Spain. Aquatic Microbial Ecology 47:123–139.
McCormick, P. V, and R. J. Stevenson. 1991. Mechanisms of benthic algal succession in lotic environments. Ecology 72:1835–1848.
McCune, B., and J. B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR.
McIntyre, P. B., E. Michel, and M. Olsgard. 2006. Top-down and bottom-up controls on periphyton biomass and productivity in Lake Tanganyika. Limnology and Oceanography 51:27.
McMurdie, P. J., and S. Holmes. 2014. Waste not, want not: Why rarefying microbiome data is inadmissible. PLoS Comput Biol 10:e1003531.
Miller, D. N., J. E. Bryant, E. L. Madsen, and W. C. Ghiorse. 1999. Evaluation and optimization of DNA extraction and purification procedures for soil and sediment samples. Applied and Environmental Microbiology 65:4715–4724.
Morales, J. B. T., and A. K. Ward. 2000. Differential incorporation of algae and bacteria by Elimia clara (Prosobranchia:Pleuroceridae)—a study using dual-labeled epilithon. Journal of the North American Benthological Society 19:289–297.
OEPA. 2002. Total maximum daily loads for the upper Little Miami River. Columbus.
Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O’Hara, G. L. Simpson, P. Solymos, M. H. Stevens, and H. Wagner. 2015. vegan: Community Ecology Package.
Pechal, J., T. Crippen, and M. Benbow. 2014. The potential use of bacterial community succession in forensics as described by high throughput metagenomic sequencing. Int J Legal Med 128:193–205.
Peck, J. E. 2010. Multivariate analysis for community ecologists: Step-by-step using PC- ORD. MjM Software Design, Gleneden Beach, OR.
Peckarsky, B. L. 1996. Alternative predator avoidance syndromes of stream-dwelling mayfly larvae. Ecology:1888–1905.
Peterson, C. G., K. A. Vormittag, and H. M. Valett. 1998. Ingestion and digestion of epilithic algae by larval insects in a heavily grazed montane stream. Freshwater Biology 40:607–623.
144
Pohlon, E., J. Marxsen, and K. Küsel. 2010. Pioneering bacterial and algal communities and potential extracellular enzyme activities of stream biofilms. FEMS Microbiology Ecology 71:364–373.
Power, M. E. 1992. Top-down and bottom-up forces in food webs: do plants have primacy. Ecology 73:733–746.
Power, M. E., W. J. Matthews, and A. J. Stewart. 1985. Grazing minnows, piscivorous bass, and stream algae: dynamics of a strong interaction. Ecology 66:1448–1456.
Pyke, G. H., H. R. Pulliam, and E. L. Charnov. 1977. Optimal foraging: a selective review of theory and tests. The Quarterly Review of Biology 52:137–154.
Ramette, A. 2009. Quantitative community fingerprinting methods for estimating the abundance of operational taxonomic units in natural microbial communities. Applied and Environmental Microbiology 75:2495–505.
Ranjard, L., F. Poly, J. C. Lata, C. Mougel, J. Thioulouse, and S. Nazaret. 2001. Characterization of bacterial and fungal soil communities by automated ribosomal intergenic spacer analysis fingerprints: biological and methodological variability. Applied and Environmental Microbiology 67:4479–4487.
Richards, C., and G. W. Minshall. 1988. The influence of periphyton abundance on Baetis bicaudatus distribution and colonization in a small stream. Journal of the North American Benthological Society:77–86.
Roberts, D. W. 2013. labdsv: Ordination and multivariate analysis for ecology.
Roeselers, G., M. C. M. van Loosdrecht, and G. Muyzer. 2007. Heterotrophic pioneers facilitate phototrophic biofilm development. Microbial Ecology 54:578–585.
Rosemond, A. D., P. J. Mulholland, and J. W. Elwood. 1993. Top-down and bottom-up control of stream periphyton: effects of nutrients and herbivores. Ecology 74:1264– 1280.
Schoener, T. W. 1971. Theory of feeding strategies. Annual Review of Ecology and Systematics 2:369–404.
Shivaji, S., M. S. Pratibha, B. Sailaja, K. H. Kishore, A. K. Singh, Z. Begum, U. Anarasi, S. R. Prabagaran, G. S. N. Reddy, and T. N. R. Srinivas. 2011. Bacterial diversity of soil in the vicinity of Pindari glacier, Himalayan mountain ranges, India, using culturable bacteria and soil 16S rRNA gene clones. Extremophiles 15:1–22.
Steinman, A. ., C. D. McIntire, S. V. Gregory, G. . Lamberti, and L. R. Ashkenas. 1987. Effects of herbivore type and density on taxonomic structure and physiognomy of algal assemblages in laboratory streams. J. N. Am. Benthol. Soc. 6:175–188.
145
Steinman, A. D., G. A. Lamberti, and L. P.R. 2007. Biomass and pigments of benthic algae. Pages 357–380 in F. R. Hauer and G. A. Lamberti, editors. Methods in Stream Ecology, 2nd edition. Academic Press, New York.
Steinman, A., P. J. Mulholland, A. V Palumbo, T. F. Flum, J. W. Elwood, and D. L. DeAngelis. 1990. Resistance of lotic ecosystems to a light elimination disturbance: a laboratory stream study. OIKOS 58:80–90.
Stock, M. S., and A. K. Ward. 1989. Establishment of a bedrock epilithic community in a small stream: microbial (algal and bacterial) metabolism and physical structure. Canadian Journal of Fisheries and Aquatic Sciences 46:1874–1883.
Suren, A. M. 2005. Effects of deposited sediment on patch selection by two grazing stream invertebrates. Hydrobiologia 549:205–218.
Swanson, K. S., S. E. Dowd, J. S. Suchodolski, I. S. Middelbos, B. M. Vester, K. A. Barry, K. E. Nelson, M. Torralba, B. Henrissat, P. M. Coutinho, I. K. O. Cann, B. A. White, and G. C. Fahey. 2011. Phylogenetic and gene-centric metagenomics of the canine intestinal microbiome reveals similarities with humans and mice. The ISME Journal 5:639–49.
Tringe, S. G., C. Von Mering, A. Kobayashi, A. A. Salamov, K. Chen, H. W. Chang, M. Podar, J. M. Short, E. J. Mathur, and J. C. Detter. 2005. Comparative metagenomics of microbial communities. Science 308:554–557.
Tuchman, N. C., and R. J. Stevenson. 1991. Effects of selective grazing by snails on benthic algal succession. Journal of the North American Benthological Society 10:430–443.
Underwood, G. J. C., and J. D. Thomas. 1990. Grazing interactions between pulmonate snails and epiphytic algae and bacteria. Freshwater Biology 23:505–522.
Winkelmann, C., J. Schneider, D. Mewes, S. I. Schmidt, S. Worischka, C. Hellmann, and J. Benndorf. 2013. Top-down and bottom-up control of periphyton by benthivorous fish and light supply in two streams. Freshwater Biology:1–16.
Zhou, J., M. A. Bruns, and J. M. Tiedje. 1996. DNA recovery from soils of diverse composition. Applied and Environmental Microbiology 62:316–322.
146
Tables
Table 1. Water characteristics were measured (N = 3) on each deployment and harvest date. Bolded values denote significant differences at P < 0.05 (ANOVA) throughout the study.
Date Temperature SpCond TDS pH Turbidity Phosphorus Chlorine Nitrate (ºC) (µS/cm) (mg/L) (NTU) (mg/L PO43- ) (mg/L) (mg/L NO3--N) 20 July 24.18 ± 0.00 554.0 ± 0.0 360.0 ± 0.0 8.12 ± 0.00 - 0.15 ± 0.02 0.04 ± 0.01 1.0 ± 0.1 27 July 22.32 ± 0.01 600.0 ± 0.4 390.0 ± 0.0 8.11 ± 0.02 11.5 ± 0.4 0.42 ± 0.11 0.04 ± 0.01 2.4 ± 0.3 3 Aug 22.84 ± 0.05 590.0 ± 0.0 378.4 ± 0.0 8.11 ± 0.03 15.2 ± 0.1 0.29 ± 0.01 0.04 ± 0.01 1.4 ± 0.3 10 Aug 21.14 ± 0.02 643.9 ± 0.1 418.9 ± 0.0 8.12 ± 0.02 7.4 ± 0.0 0.30 ± 0.05 0.04 ± 0.01 1.3 ± 0.5 17 Aug 18.93 ± 0.01 650.5 ± 0.6 423.0 ± 0.0 8.08 ± 0.04 4.6 ± 0.3 0.18 ± 0.02 0.02 ± 0.01 0.8 ± 0.3 Mean 21.88 ± 1.98 607.7 ± 40.0 394.1 ± 0.0 8.10 ± 0.02 9.7 ± 4.6 0.27 ± 0.11 0.04 ± 0.01 1.4 ± 0.6
Date Sulfate Nitrite Ammonia Alkalinity TSS DO Depth (mg/L) (mg/L NO2--N) (mg/L NH3-N) (mg/L CaCO3) (mg/L) (mg/L) (cm) 20 July 39 ± 1 0.006 ± 0.001 0.23 ± 0.07 167 ± 6 10 ± 6 9.02 ± 0.02 34 ± 5 27 July 31 ± 1 0.009 ± 0.002 0.17 ± 0.03 233 ± 8 17 ± 6 8.51 ± 0.03 28 ± 5 3 Aug 33 ± 2 0.007 ± 0.001 0.15 ± 0.02 231 ± 12 10 ± 3 8.98 ± 0.09 30 ± 5 10 Aug 42 ± 3 0.010 ± 0.002 0.13 ± 0.06 248 ± 10 16 ± 3 9.97 ± 0.06 24 ± 5 17 Aug 46 ± 2 0.007 ± 0.002 0.16 ± 0.04 238 ± 5 15 ± 2 14.30 ± 0.43 20 ± 5 Mean 38 ± 6 0.008 ± 0.001 0.17 ± 0.04 223 ± 32 14 ± 4 10.0 ± 2.37 27 ± 5
147
Table 2. The influences of grazing treatment, the amount of time (days) biofilms grew in the stream, the biofilm treatment, and the factor interactions on ARISA communities were assessed with PERMANOVA using Bray-Curtis distance measure and 9999 permutations. Significant P-values at < 0.05 are bolded.
Bacteria Eukaryotes Df SS pseudo-F P SS pseudo-F P Grazing 3 5.3117 8.0565 0.0001 7.1708 10.2287 0.0001 Time 2 0.7992 1.8182 0.0055 0.6071 1.299 0.1451 Biofilm 1 0.2832 1.2888 0.1666 0.2323 0.994 0.4299 Grazing x Time 6 2.0296 1.5392 0.0016 1.3428 0.9577 0.564 Grazing x Biofilm 3 0.6512 0.9877 0.4814 0.4455 0.6355 0.9702 Time x Biofilm 2 0.7902 1.7979 0.0054 0.6429 1.3756 0.1074 Grazing x Time x Biofilm 6 1.4918 1.1313 0.1752 1.5735 1.1222 0.2305
148
Table 3. Percent relative abundance of genera that represented abundant taxa (> 1%) of the entire dataset was averaged (± SD) for each grazing treatment. An additional category of All Grazing combined grazing treatments for easier comparison between treatments with grazing and without grazing.
Genus None All Grazing Snail Mayfly Snail+Mayfly Nannocystineae 2.28 ± 0.88 12.43 ± 2.78 13.03 ± 3.85 12.12 ± 2.53 12.09 ± 1.87 Balneatrix 1.14 ± 0.34 9.36 ± 2.82 9.28 ± 3.67 10.03 ± 2.65 8.65 ± 2.20 Rhodobacter 8.16 ± 3.97 4.11 ± 1.28 4.45 ± 1.83 3.88 ± 0.63 3.99 ± 1.28 Comamonadaceae 5.44 ± 0.85 4.89 ± 0.57 5.29 ± 0.52 4.69 ± 0.66 4.64 ± 0.20 Flavobacterium 8.05 ± 3.48 3.83 ± 1.07 4.64 ± 0.83 3.18 ± 0.85 3.62 ± 1.07 Pleurocapsa 3.84 ± 1.58 4.27 ± 1.28 5.37 ± 1.43 3.54 ± 0.71 3.82 ± 0.69 Chamaesiphon 1.98 ± 0.47 2.85 ± 1.02 2.51 ± 0.77 2.69 ± 1.19 3.46 ± 0.96 Hydrogenophaga 4.39 ± 1.71 1.92 ± 0.62 1.84 ± 0.35 2.13 ± 0.93 1.78 ± 0.46 Acidovorax 3.70 ± 1.31 2.14 ± 0.68 2.24 ± 0.75 2.05 ± 0.85 2.14 ± 0.50 Mitochondria 0.57 ± 0.15 2.64 ± 1.47 1.99 ± 0.87 2.77 ± 2.07 3.26 ± 1.14 Rhodovulum 2.93 ± 1.18 1.42 ± 0.36 1.42 ± 0.49 1.47 ± 0.35 1.35 ± 0.20 Arenimonas 2.31 ± 0.72 1.54 ± 0.51 1.62 ± 0.49 1.76 ± 0.56 1.20 ± 0.33 Zymomonas 2.56 ± 0.76 1.21 ± 0.68 1.23 ± 0.55 1.21 ± 0.94 1.18 ± 0.60 Leptothrix 1.46 ± 0.51 1.36 ± 0.33 1.38 ± 0.44 1.40 ± 0.23 1.30 ± 0.33 Anabaena 0.98 ± 0.58 1.41 ± 0.61 1.19 ± 0.36 1.25 ± 0.56 1.88 ± 0.74 Chitinophagaceae 1.18 ± 0.45 1.38 ± 0.35 1.41 ± 0.33 1.36 ± 0.39 1.38 ± 0.42 Xanthomonadaceae 1.35 ± 0.74 1.26 ± 0.45 1.47 ± 0.55 1.28 ± 0.37 1.00 ± 0.36 Candidatus_amoebophilus 0.10 ± 0.12 1.64 ± 1.27 1.80 ± 1.87 1.90 ± 0.99 1.13 ± 0.60 Variovorax 1.55 ± 0.28 1.10 ± 0.29 1.24 ± 0.25 1.05 ± 0.40 1.01 ± 0.10 Flexibacter 0.54 ± 0.16 1.50 ± 0.43 1.52 ± 0.29 1.40 ± 0.52 1.61 ± 0.52
149
Alcaligenaceae 1.93 ± 1.07 0.61 ± 0.39 0.72 ± 0.55 0.51 ± 0.33 0.59 ± 0.26 Sphingobium 3.04 ± 2.38 0.56 ± 0.31 0.49 ± 0.19 0.55 ± 0.21 0.64 ± 0.52 Chroococcidiopsis 0.65 ± 0.49 1.14 ± 0.46 0.97 ± 0.14 1.05 ± 0.40 1.43 ± 0.69 Chlorochromatium 0.58 ± 0.31 1.27 ± 0.42 1.08 ± 0.19 1.58 ± 0.53 1.15 ± 0.32
150
Table 4. Significant indicators (P < 0.05) of the grazing treatment were determined using indicator analysis on 454-pyrosequeced bacterial communities at the genus level. The indicator value (IV) is indicative of the relative abundance and constancy within each group where 1 is perfect indication and 0 is no indication. All indicators were non-abundant taxa because they represented < 0.05% relative abundance of the sample.
Grazing Treatment Grazing Treatment Level IV P Genus Level IV P Genus None 0.636 0.004 Verrucomicrobiaceae Snail 0.6667 0.001 Flavobacteriaceae 0.6022 0.001 Erythrobacteraceae 0.5704 0.007 Aureispira 0.600 0.013 Paenibacillus 0.5000 0.047 Chondromyces 0.5409 0.001 Sphaerotilus 0.4885 0.027 Rhodanobacter 0.5305 0.012 Microvirga 0.4592 0.028 Soonwooa 0.5084 0.012 Albidiferax 0.4578 0.009 Acidocella 0.4654 0.002 Phyllobacteriaceae 0.3481 0.033 Kangiella 0.4638 0.003 Simplicispira 0.453 0.018 Rubellimicrobium Mayfly 0.6667 0.007 Melitea 0.453 0.001 Sphingobium 0.5408 0.013 Rhodoferax 0.4429 0.001 Pseudoxanthomonas 0.5000 0.045 Spam (candidate division) 0.4392 0.005 Ramlibacter 0.4325 0.013 Alcaligenaceae Snail+Mayfly 0.6000 0.018 Nitrosomonas 0.4271 0.011 Pilimelia 0.4899 0.001 Rivularia 0.4156 0.001 Meganema 0.4574 0.039 Pseudomonadaceae 0.3968 0.034 Yonghaparkia 0.445 0.046 Aminobacterium 0.3913 0.004 Beijerinckiaceae 0.386 0.041 Scherffelia
151
0.3851 0.034 Xanthobacteraceae 0.3853 0.001 Leptolyngbya 0.384 0.032 Rhizobacter 0.3506 0.008 Pseudomonas 0.3812 0.002 Hyphomicrobium 0.3353 0.017 Ferruginibacter 0.3746 0.003 Devosia 0.3734 0.009 Methylibium 0.3698 0.026 Alicycliphilus 0.3665 0.003 Brevundimonas 0.366 0.001 Zymomonas 0.3659 0.022 Rhodovulum 0.3646 0.005 Brachymonas 0.36 0.048 Rhodobacter 0.3552 0.002 Hydrogenophaga 0.3544 0.012 Nitrosomonadaceae 0.3525 0.038 Roseomonas 0.3508 0.011 Chloracidobacterium 0.35 0.008 Flavobacterium 0.3496 0.002 Methylotenera 0.322 0.011 Acidovorax 0.3118 0.013 Novosphingobium 0.3118 0.043 Thermomonas 0.3095 0.049 Arenimonas
152
Table 5. Significant indicators (P < 0.05) of the amount of time (days) biofilms grew in the stream and the biofilm treatment were determined using indicator analysis on 454-pyrosequeced bacterial communities at the genus level. The indicator value (IV) is indicative of the relative abundance and constancy within each group where 1 is perfect indication and 0 is no indication. All indicators were non-abundant taxa because they represented < 0.05% relative abundance of the sample.
Time of Growth In Stream Biofilm Treatment Level IV P Genus Level IV P Genus 14 Days 0.6358 0.008 Acidobacteriaceae Ambient 0.6542 0.004 Mitsuaria 0.5784 0.011 Alishewanella 0.6298 0.027 Cosmarium 0.5167 0.023 Clostridium 0.6097 0.026 Desulfobacteraceae 0.4574 0.037 Pedomicrobium 0.4204 0.037 Thioploca 0.4213 0.048 Saprospiraceae Modified Flow 0.6703 0.004 Sporichthya 21 Days 0.6555 0.003 Nostoc 0.668 0.025 Hymenobacter 0.5592 0.019 Rickettsia 0.6045 0.037 Methyloversatilis 0.5302 0.006 Holospora 0.6005 0.015 Acetobacteraceae 0.5096 0.02 Gemmata 0.5158 0.021 Anabaenopsis 0.4588 0.003 Lacibacter 0.4747 0.043 Porticoccus 0.4565 0.037 Aeromonas 0.4509 0.007 Achromobacter 0.4035 0.037 Thermomonas 0.3875 0.04 Chitinophagaceae
28 Days 0.6354 0.002 Haliangium 0.6121 0.004 Bacteriovoraceae
153
0.6111 0.003 Cytophaga 0.5436 0.003 Deinococcus 0.524 0.019 Pedobacter 0.5093 0.031 Sorangium 0.4286 0.038 Ancalomicrobium
154
Figures
Figure 1. Modified flow plus dark treatments (A) were built with metal screws to create heterogeneous flow and an overhead tarp to reduce direct light by 97%. Leaves and debris that were caught on the screws were removed weekly and sometimes invertebrate grazers were observed on the pavers (B). Black arrows denote snail (Goniobasis sp.) grazers. Modified flow biofilms at 21 days (C) were the preferred biofilm when both snail and mayflies were grazing together, but modified flow biofilms at 28 (D) days frequently displayed grazing patterns that occurred in the natural setting.
155
4 Abiotic Treatments X 4 Time Points Treatments • Ambient- Natural (control) Field • Modified Flow- Metal screws • Dark- Tarp (97% reduced) • Modified Flow + Dark
16 Biofilm Combinations Treatment A MF D MF+D
7
14
Time (Days) Time 21
28
16 Biofilm Combinations x 4 Grazing Treatments
Treatments • No Grazing- (control) Lab • Snails- Goniobasis sp. (3) • Mayflies- Maccaffertium sp. (6) • Snails (3) + Mayflies (6)
64 UNIQUE BIOFILMS
Figure 2. The methods of the study are represented by a schematic depicting how the unique biofilms were created. This study encompasses two experimental procedures where biofilms developed in the field under four abiotic conditions and then were brought into the lab to be subjected to four grazing treatments.
156
Figure 3. Biofilm characteristics of algal biomass (chlorophyll a x 67) and total biomass
(AFDM) of the no grazing control treatment (N = 2) were analyzed by a two-way
ANOVA. These values were determined after the 6-day grazing period, and letters denote significant differences after Bonferroni corrected multiple comparisons at P < 0.05.
157
Algal Biomass Loss Total Biomass Loss
0.50 1.8 Ambient Snail Snail Modified Flow b b b 1.2 b 0.25 ab ab a ab 0.6 a a
) 0.00 ) 2 2 0.0
0.50 Mayfly 1.8 Mayfly
b bc bc 1.2
0.25 ab ab ab a a a a 0.6 a a SE) Total Biomass Loss (mg/cm Loss SE) Biomass Total SE) Algal Biomass Loss (mg/cm Loss Biomass SE) Algal ± ± 0.00 0.0 Mean ( Mean Mean ( Mean c Snail+Mayfly 1.8 Snail+Mayfly 0.50 c
bc bc 1.2 b ab b ab b 0.25 0.6 a a 45 54 56 68 63 73 53 56 46 56 56 58 56 68 50 48 0.00 0.0 7 14 21 28 7 14 21 28 Time of Growth in Stream (Days) Time of Growth in Stream (Days)
Figure 4. The amount of biomass loss due to grazing was determined by subtracting the biomass levels of the grazing treatments (N = 3) from the mean of the no grazing control treatment. Only ambient and modified flow biofilms were included in analyses because grazers showed no response to biofilms that developed under dark conditions. Biofilm characteristics of algal biomass (chlorophyll a x 67) and total biomass (AFDM) were analyzed by a two-way ANOVA, and letters denote significant differences after
Bonferroni corrected multiple comparisons at P < 0.05. Numbers within bars of the snail plus mayfly treatment represent the percent decrease in biomass.
158
PBG.B.ALL1.R1OUTALL.FINAL Grazing None Snail Mayfly Snail+Mayfly Axis 2 (0.263)
PBG.B.ALL1.R1OUTALL.FINAL Axis 1 (0.350) Grazing Pressure 0 1 2 Axis 2 (0.263)
Axis 1 (0.350)
Figure 5. Bacterial community profiles generated by ARISA were analyzed by NMDS
(3-D, R2 = 0.880, stress = 0.13) using Bray-Curtis distance (N = 3). Data was arcsine square root transformed and rare fragments present in only one sample were removed in an attempt to reduce the effect of an abundance of zeros and minimize noise. Community patterns varied with grazing treatment (top) and number of grazers present (bottom).
Numbers in parentheses are the R2 values for the respective axis.
159
PBG.E.ALL.R1O.FINAL PBG.E.ALL.R1O.FINAL Grazing Grazing Pressure None 0 Snail 1 Mayfly 2 Snail+Mayfly Axis 3 (0.420) Axis 3 (0.490)
Axis 2 (0.212) Axis 2 (0.148)
Figure 6. Eukaryotic community profiles generated by ARISA were analyzed by NMDS
(3-D, R2 = 0.744, stress = 0.15) using Bray-Curtis distance (N = 3). Data was arcsine square root transformed and rare fragments present in only one sample were removed in an attempt to reduce the effect of an abundance of zeros and minimize noise. Community patterns varied with grazing treatment (left) and number of grazers present (right).
Numbers in parentheses are the R2 values for the respective axis.
160
CHAPTER 5
INFLUENCE OF ENVIRONMENTAL FACTORS ON EPINECROTIC BIOFILMS
Abstract
Biofilms are a formation of microbial communities that are found on all surfaces in aqueous environments. Human remains are sometimes found in aquatic settings and are not exempt as a biofilm substrate. It has proven difficult to determine a post mortem submersion interval (PMSI) because there is not a predictable succession of insects that use aquatic carrion as a resource; however, biofilms proceed through successional development, and this has been proposed as an alternative method to determine PMSI.
We sought to compare the response of epinecrotic and epilithic communities by assessing community composition through time and in relation to varying environmental conditions to investigate general principles of epinecrotic communities. Epinecrotic communities were distinctly different from epilithic communities (PERMANOVA, pseudo-F = 9.31, P
< 0.0001) even though location had an influence (PERMANOVA, pseudo-F = 17.31, P <
0.0001) indicating that selective forces of the substrate were greater than the influence of environmental variability. Yet, epinecrotic communities were influenced by environmental variation associated with location (pseudo-F = 11.46, P < 0.0001). These
161
communities did exhibit patterns of succession suggesting that succession is a robust process. The implications are that epinecrotic communities have the potential to be used for forensic applications by associating successional changes with time to determine a post mortem submersion interval; however, the influence of environmental factors indicates that taxa composition of the successional patterns are shaped by the environment. Future studies can investigate functional diversity rather than taxonomic diversity to determine if functional redundancy of different bacterial species produces a common functional successional pattern that can be applied universally in aquatic habitats to determine a PMSI.
Introduction
Understanding pattern and process in ecological community development is a long-established area of scientific inquiry that has new and important implications in the forensic sciences. The predictability of temporal community development (succession) is a topic with a long-history of scientific debate (Gleason 1926, Clements 1936), and linking community patterns to resources and disturbance across space and time has generated large bodies of theory and empirical results (e.g., Bormann and Likens 1979,
Weiher and Keddy 1999, among many others).
The type of evidence that is collected at a crime scene varies greatly, and biological evidence that is based on successional principles (most often of invertebrates) is an important tool for estimating a post mortem interval (PMI, the amount of time since death). Human remains are sometimes found submerged and forensic entomologists have been unsuccessful in identifying a predictable sequence of invertebrates that can be used
162
to estimate a postmortem submersion interval (Vance et al. 1995, Keiper et al. 1997,
Hobischak and Anderson 2002). This is mainly because aquatic invertebrates typically use the remains as a substrate and not an energy resource like the necrophagus terrestrial invertebrates (Wallace et al. 2008, Barrios and Wolff 2011). A commonality between the studies conducted in aquatic settings has been the development of a microbial layer on the carrion surface. These communities, recently classified as “epinecrotic” (Pechal et al.
2014), have been noted on salmon (Claeson et al. 2006), waterfowl (Parmenter and
Lamarra 1991), rats (Tomberlin and Adler 1998), and swine (Haefner et al. 2004), and only a few studies have focused on them. Documenting pattern and process of microbial community development has conceptual and practical scientific importance in the forensic sciences.
In aquatic ecosystems, microbes are found predominantly within biofilms rather than free floating forms (Costerton et al. 1995, Giller and Malmqvist 1998). Biofilms are matrix-enclosed microbial communities that are both trophically (heterotrophs and autotrophs) and phylogenetically (prokaryotes and eukaryotes) diverse, containing algae, bacteria, fungi, and protozoa (Hoagland et al. 1982, Giller and Malmqvist 1998). These microorganisms are encased in an extracellular polymeric substance (EPS) that is protective against changing conditions (Freeman and Lock 1995, Sutherland 2001), traps and stores nutrients (Freeman et al. 1995, Battin et al. 2003, Augspurger et al. 2010), and accrues enzymes that break down organic matter (Sinsabaugh et al. 1991b, Romaní et al.
2006). The type of surface classifies the biofilm and influences community composition and energy dynamics by dictating the dominant community. For example, epilithic biofilms are found on inorganic substrates like rocks and are more autotrophic with an
163
abundant algal community, while epixylic biofilms are found on decomposing plant matter and are more heterotrophic with a substantial fungal community (Sinsabaugh et al.
1991a, Tank and Dodds 2003, Hladyz et al. 2011). Epinecrotic biofilms are considered heterotrophic, but light availability can produce substantial algal communities (Haefner et al. 2004, Zimmerman and Wallace 2008). There is still much to learn about these epinecrotic biofilms.
Biofilms develop in a successional manner where community composition changes over time, and this sequence has been investigated as an alternative approach to measure PMSI. Studies of epinecrotic communities in terrestrial settings have identified successional patterns in bacterial communities (Hyde et al. 2013, Metcalf et al. 2013,
Pechal et al. 2014). Patterns of diversity and abundance in epinecrotic biofilms throughout aquatic decomposition have been demonstrated for diatoms (unicellular algae with a silica cell wall) in freshwater systems (Casamatta and Verb 2000, Zimmerman and
Wallace 2008) as well as bacteria in a marine setting (Dickson et al. 2011). The potential for using succession was further described for bacteria in freshwater streams where pyrosequencing revealed changing communities in both summer and winter seasons; however, communities were more similar to each other within a season suggesting that environmental factors influence composition more than the decomposition process
(Benbow et al. in press). Little is known about the microbial community development of epinecrotic biofilms, and, to our knowledge, the role of environmental factors in influencing community composition of epinecrotic biofilms is an unexplored area of scientific research.
164
The principle objective of this study was to assess microbial community development of epinecrotic biofilms under varying environmental conditions. Previous work has indicated clear evidence of successional pattern and environmental influence in epilithic biofilm communities, and in this study we sought to compare the response of epilithic and epinecrotic communities by assessing community composition through time and in relation to varying environmental conditions. We hypothesized (H1) that epinecrotic and epilithic biofilms would have substantially different biofilm community composition regardless of differences in environmental factors, i.e., resource substrate is a stronger ecological control than environmental conditions. We hypothesized that (H2) environmental variation would drive community composition within biofilm type. We further hypothesized (H3) that both the epilithic and epinecrotic biofilms would exhibit community differentiation based on development time (succession). This work has both practical implications in the forensic sciences and the potential to advance understanding the ecology of stream microbial communities.
Methods
Study site descriptions
The experiment was conducted in two locations (Dayton, OH and Millersville,
PA) to capture the effects of differing environmental factors. The Dayton study was conducted in Farmersville, OH from 29 June 2012 to 27 July 2012 in a first order headwater stream that joined Little Twin Creek (39°39'53.8"N 84°23'44.3"W). The surrounding landcover of the catchment area was predominately agriculture, and there was a 2-5 m riparian forest that was dominated by maple (Acer sp.) and elm (Ulmus sp.)
165
and invasive honeysuckle (Lonicera maackii). A drought eliminated flow to the upper portions of the stream and the study sites had to be relocated further down stream, which resulted in two different habitats that were intersected by a concrete divide under a bridge. The upstream, closed canopy site was shaded by trees, had clear low flowing water, and the substrate consisted of pebbles and cobbles. The downstream, open canopy site was a pool subjected to direct sunlight, had clouded almost stagnant water, and the substrate was cobbles and pebbles covered by sediments. The upstream site was more similar to the Millersville site than the downstream one.
The Millersville study is previously described in Benbow et al. (in press), and was conducted from 26 June 2012 to 17 July 2012 in a first order tributary to the west branch of Big Spring Run (39°59'29.1"N 76°15'49.0"W) within the Conestoga River watershed of Lancaster, Pennsylvania. The watershed is a mix of suburban/agricultural land, and the stream was bordered by 10-20 m of riparian forest and shrub vegetation. The riparian forest was dominated by silver maple (Acer saccharinum L.), box elder (Acer negundo
L.), and sycamore (Platanus occidentalis L.) with a ground cover of multiflora rose (Rosa multiflora Thunb.). The stream substrate consisted of a mixture of pebble and cobble.
Environmental water parameters
In Dayton, specific conductivity (SpCond µS/cm), total dissolved solids (TDS mg/L), pH, and temperature (°C) were recorded using a YSI 6600 v2 Sonde (YSI Inc,
Yellow Springs, OH) 15 m above and below the uppermost and lowermost carcasses, respectively. Water was also collected at these points to measure nitrate (mg/L NO3--N),
- nitrite (mg/L NO2 -N), sulfate (mg/L), ammonia (mg/L NH3-N), alkalinity (mg/L
166
CaCO3), and total suspended solids (TSS mg/L) in the lab using EPA approved protocols
(Hach® Company, Loveland, CO). In Millersville, water quality parameters of dissolved oxygen (mg/L), pH, specific conductivity (µS/cm), water temperature (oC), total dissolved solids (g/L), oxidation reduction potential (mV), and salinity (ppt) were measured at a single location 30 m upstream of the uppermost carcass and 30 m downstream of the lowermost carcass on each sampling day using a Horiba® (Kyoto,
Japan) Multi Water Quality Checker (U-50 Series).
Epinecrotic biofilm study design and sampling
Epinecrotic biofilms developed on stillborn Sus suscrofa carcases (N = 4) obtained from the Penn State University Swine Research Facility (State College,
Pennsylvania). Swine skin has been accepted as a surrogate for human skin (Dick and
Scott 1992, Sekkat et al. 2002, Jacobi et al. 2007) and carcasses are frequently used in place of human cadavers (Schoenly et al. 2007). Carcasses were placed on plastic drawer organizer trays (0.38 x 0.15 x 0.05 m) inside metal small game traps (Havahart®, Animals
B-Gone, Orrstown, PA) (0.61 x 0.18 x 0.18 m) to facilitate sampling as the carcass disarticulated and to prevent scavenger removal of the carcasses. Traps were anchored to the streambed on previously secured rebar, which allowed easy trap removal for sampling. Carcasses were placed ~15 m downstream from each other in run habitats of the stream to avoid being silted over in pools and the abrasive action of riffle habitats.
Epinecrotic biofilms were randomly sampled before being placed into streams
(Day 0) and bi-weekly alternating 3 or 4 days from the abdomen/rib cage using individually packaged sterile swabs. The area was swabbed with six strokes where the
167
swab was rotated 180° after the third stroke and one direction counted as a stroke. The swabs was placed individually into sterile microfuge tubes and transported on ice and kept at -20°C until DNA extraction. Carcasses were immediately submerged upon sample completion, and new gloves were worn for each sample collection.
Epilithic biofilm study design and processing
Epilithic biofilms developed naturally on hexagonal unglazed porcelain tiles (N =
6) attached to brick pavers (19.2 x 9 x 1.3 cm) with 100% silicone. Pavers were placed
0.3 m upstream of the cages and immediately downstream of the cages. Tiles (N = 4) were removed at each sampling date, transported to the lab on ice, and placed at -20°C until processing. Biofilm biomass was removed using a sterile razor blade and toothbrush and suspended in ultrapure water (NANOpure II; Barnstead, Boston, MA, USA). Two sub samples were collected on GB-140 glass membrane filters (diameter, 25mm; pore size, 0.4 µm; Sterlitech, Kent, WA, USA) to determine total biomass as ash free dry mass
(AFDM) and algal-associated biomass as chlorophyll a following techniques used by
(Steinman et al. 2007). A third subsample was collected for DNA extractions, and these filters were stored in 90% ethanol at -80°C until extraction. Although for DNA extractions, we highly suggest using a mixed nitrocellulose filter and placing a few drops of ethanol on the sample that evaporates before storage.
DNA extraction
DNA was extracted from ethanol evaporated, dried filters and swabs using a combination of methods from Miller et al., (1999) and Zhou et al., (1996) as suggested by
168
Lear et al., (2010). Samples were lysed in 1 ml extraction buffer made of 100 mM Tris-
HCl (pH 8.0), 100 mM EDTA disodium salt (pH 8.0), 100mM sodium phosphate (pH
8.0), 1.5 M sodium chloride and 1% CTAB; 20 µl of proteinase K (10mg/mL) and 25 µl of SDS (20%) were added to each sample in the extraction buffer prior to bead beating
(0.25 g each of 0.1 mm and 0.5 mm glass beads) for 15 min on a horizontal vortex adaptor (MO BIO Laboratories, Carlsbad, CA) at full speed. The samples were incubated at 60°C for 30 min with gentle end-over-end inversions by hand after 15 min; 250 µL of supernatant was collected in a new microcentrifuge tube after centrifugation at 6,000 x g for 10 min. For epinecrotic swab samples only, the lysis process was repeated without vortexing with an additional 500 µL extraction buffer, 10 µL proteinase K, and 12.5 µL
SDS to obtain a final supernatant volume of approximately 750 µL. DNA was separated from organic debris with a chloroform isoamyl alcohol (24:1 vol/vol) extraction and precipitated overnight at -20°C using isopropanol. Samples were removed from the -20°C and warmed to 37°C to dissolve salt precipitates, and the DNA was pelleted at 15,000 x g for 30 m. Finally, the DNA pellet was washed twice with ice cold 70% ethanol and dissolved in 50-100 µL ultrapure water (NANOpure II™, Thermo Scientific, Waltham,
MA, USA) water, depending on the DNA pellet size.
ARISA
Bacterial communities were assessed using profiles created by automated ribosomal intergenic spacer analysis (ARISA) (Fisher and Triplett 1999), which generates a unique ‘fingerprint’ of microbial communities using the 16S-23S intergenic space in bacteria. While ARISA does not taxonomically identify organisms like next
169
generation sequencing methods, this method generates community profiles that produce similar patterns in results (Bienhold et al. 2012, van Dorst et al. 2014). Approximately
15-20 ng of DNA quantified by spectrophotometer (NanoPhotometerTM Pearl; Denville
Scientific Inc., South Plainfield, NJ, USA) was amplified by PCR using 25 µL GoTaq®
Colorless Master Mix (Promega, Madison, WI, USA) with 0.5 µM of forward and reverse primers. Bacteria ribosomal intergenic space regions were amplified with primers
ITSF (5’-GTCGTAACAAGGTAGCCGTA-3’) labeled with FAM at the 5’ end (IDT,
Coralville, IA, USA) and ITSReub (5’-GCCAAGGCATCCACC-3’) (Cardinale et al.
2004). Fragments were created with the following PCR conditions: (i) 94°C for 3 min,
(ii) 35 cycles of 94°C for 1 min, 56°C (57.5°C for eukaryotes) for 1 min, 72°C for 2 min, and finally (iii) 72°C for 10 min (Fechner et al. 2010). PCR products were sent to DNA
Analysis, LLC (Cincinnati, OH, USA) for fragment analysis on an ABI 3100 (Life
Technologies, Carlsbad, CA, USA). Fragments were interpreted using Genescan v 3.7 using the Local Southern Size Calling Method with a peak height threshold of 100 fluorescence units to remove background fluorescence and formulated using
GeneMapper v 2.5 (Life Technologies, Carlsbad, CA, USA).
Fragment peak length and area was converted to column format using the treeflap
Excel Macro (http://www.wsc.monash.edu.au/~cwalsh/treeflap.xls) and processed with the automatic_binner script to determine binning window size and the interactive_binner script to determine the best starting window position (Ramette 2009) in R v 3.1.0 (R Core
Team 2014). This method was used to account for inherent imprecision of analyzer machines. Peak area was converted to relative abundance of each fragment as part of the
170
entire sample, fragments < 0.09% relative abundance were removed (Ramette 2009), and window size was calculated to be 1.5 base pairs.
Statistical Analyses
Environmental water parameters throughout the study within a site were averaged and compared using a students t-test. Values were also compared within the Dayton, OH site to compare the upstream/closed canopy habitat to the downstream/open habitat.
Microbial community patterns were visualized using nonmetric multidimensional scaling (NMDS) with Bray-Curtis (Sørensen) distance because it is a nonparametric approach useful in evaluating nonlinear relationships of data with high numbers of zeros
(McCune and Grace 2002). Differences in configuration based on categorical overlays were tested with nonparametric multivariate analysis of variance (PERMANOVA)
(Anderson 2001). Analyses were conducted in R v 3.1.2 (R Core Team 2014) with the vegan v 2.2-1 (Oksanen et al. 2015) package.
Results
Environmental water parameters
The study sites, including the upstream and downstream sites in Dayton, differed in water parameters (two-tailed t-test). Between Dayton (Table 1) and Millersville (Table
2), there were only a few common parameters measured due to access of different measurement devices. Temperature was the only non-significant (P = 0.8546) parameter while dissolved oxygen (P = 0.0347), pH (P = 0.0091), specific conductivity (P <
0.0001), and total dissolved solids (P = 0.0066) were significantly different. Within
171
Dayton, the upstream and downstream sites significantly differed in total suspended solids (P = 0.0178), specific conductivity (P = 0.0004), total dissolved solids (P =
0.0003), dissolved oxygen (P = 0.0008), pH (P < 0.0001), and temperature (P < 0.0001).
Community assembly patterns
Bacterial community composition determined by ARISA was influenced by time and location, but the substrate itself produced the most drastic grouping in NMDS ordination (Figure 1). Epinecrotic and epilithic biofilms were distinctly different
(PERMANOVA, pseudo-F = 9.31, P < 0.0001) but location was also influential
(PERMANOVA, pseudo-F = 17.31, P < 0.0001), and there was an interaction effect
(PERMANOVA, pseudo-F = 7.02, P < 0.0001) from these two factors (Figure 1). When samples were separated into epinecrotic and epilithic biofilms (Figure 2), both biofilm types were influenced by location (PERMANOVA, pseudo-F = 4.68, P < 0.0001; pseudo-F = 11.46, P < 0.0001; respectively) but only epilithic biofilms were influenced by canopy (PERMANOVA, pseudo-F = 3.66, P = 0.0002). This is not visible in the ordination because the groups align on the third axis (see below). This indicates that the biofilms on the same substrates are more similar to each other regardless of environmental factors associated with location or habitat, but within a biofilm type, environmental factors influence community composition.
Succession occurred in epinecrotic biofilms as indicated by each day grouping together in NMDS ordination and confirmed by PERMONVA (Figure 3). Epinecrotic biofilms in Dayton were influenced by days of decomposition (pseudo-F = 3.16, P <
0.0001; Figure 3) but not canopy (pseudo-F = 1.02, P = 0.3933; not shown). This
172
indicates that there is some robustness against environmental variability at the same location. Epinecrotic biofilms in Millersville were also influenced by day of decomposition (pseudo-F = 3.93, P < 0.0001; Figure 3). Succession is easily seen with the day overlays because groups are aligned in a chronological fashion. These data suggest that epinecrotic biofilms are influenced by location and possible changes in species pool but not by local environmental factors.
Epilithic biofilms in Dayton were influenced by days of decomposition (pseudo-F
= 12.40, P < 0.0001) and canopy (pseudo-F = 4.15, P < 0.0001), which had an interaction effect (pseudo-F = 2.31, P = 0.0093; Figure 4). But whether the tiles were placed upstream or downstream of the swine carcasses had no effect (pseudo-F = 1.61, P =
0.0695). This indicates that the carrion resource is not affecting the surrounding environment in a way that altered epilithic biofilm community composition. In addition, it is important to note that canopy aligned with NMDS Axis 3 (Figure 4), but the effect is not observed until 10 days. Millersville epilithic biofilms were not analyzed because there was a low sample size (N = 16), only three days were represented, and the group dispersions were significantly different (ANOVA, P = 0.0225).
Discussion
Epinecrotic communities are clearly discernible from epilithic communities
The development of ecological communities is known to be linked to resource substrates. One motivating objective in this study was to determine the discernibility of epinecrotic biofilms in relation to epilithic biofilms, which are virtually ubiquitous in streams. The results of our analysis suggest clear divergence among the two communities
173
based on substrate type (inorganic versus carrion) supporting H1. This is not entirely surprising because it has been established that epilithic and epixylic (on decaying plant material) biofilms differ in limiting nutrients, exoenzyme activity, and fungal biomass
(Sinsabaugh et al. 1991a, Tank and Dodds 2003). There are also production differences where net primary productivity was greater in epilithic biofilms while respiration was higher in epixylic biofilms reflecting differences in functional dynamics (Sabater et al.
1998). It was beyond the scope of this study to identify the constituent organisms driving the separation of the epinecrotic and epilithic microbial communities; however, one potentially important difference is that organisms dominating the epinecrotic community likely include a large portion of heterotrophs or detritovores whereas the epilithic community likely has a larger portion of autotrophs. Fungal organisms have been noted as an important component in structuring biofilms on plant derived organic substrates
(Golladay and Sinsabaugh 1991), and they may be important on carrion substrates as well. This biofilm component has yet to be studied in epinecrotic biofilms and may provide a foundation for future studies. These results suggest a potential utility of microbial profiling for forensic applications given that the epinecrotic biofilms were clearly discernible from the ubiquitous epilithic biofilms.
Epinecrotic communities exhibit clear variation in response to environmental conditions
Ecological communities are known to vary in accordance with environmental conditions; however, little is known about these effects in relation to epinecrotic biofilms.
We hypothesized (H2) that within each biofilm type environmental conditions could create discernible community patterns. The study sites (Dayton and Millersville) were
174
geographically distant from one another and we show evidence of clustering of the communities that reflect these disparate study sites. Dissolved oxygen, specific conductivity, and total dissolved solids were significantly different between the sites, and it is likely other unmeasured parameters like nutrients and chemistry were different as well (Van der Grinten et al. 2004). Dissolved nutrients are readily utilized by biofilms
(Lock and Hynes 1976) and the composition influences community structure (McNamara and Leff 2004). Bacterial communities were selected by dissolved organic matter (DOM) composition even though inoculums with varying bacterial composition were tested
(Docherty et al. 2006). In addition, bacteria responded more to forms of DOM rather than inorganic nutrients (Olapade and Leff 2006). Algae are also important because light- grown epilithic biofilms are a net DOM consumer (Romaní et al. 2004) and have increased enzyme activity in comparison to dark-grown epilithic biofilms (Espeland et al.
2001). The differences observed in community composition could be related to geological substrate, land use history, or variation in the species pool, but most likely is a cumulative result of the environmental variability.
Epinecrotic community succession
Succession is a foundational process in most ecosystems and has been clearly demonstrated in microbial communities (Fierer et al. 2010). We hypothesized that (H3) discernible variation over time would be found in the epinecrotic biofilm, which was clearly demonstrated in our analysis. Indeed, in both sites, community separation was clear across the timeframe of the study demonstrating that succession is a universal occurrence. This is further supported by succession of bacterial communities within
175
epinecrotic biofilms during winter in streams (Benbow et al. in press) and in a marine habitat (Dickson et al. 2011). Although temporal resolution was relatively coarse (days) and the communities were not always clearly separated (e.g., days 7, 10 and 14 were very similar in the Millersville site), these data do suggest the possibility of developing a microbe-based timeline for submerged remains.
Conclusions and potential for forensic application
Epinecrotic communities were distinctly different from epilithic communities regardless of the location indicating that selective forces of the substrate were greater than the influence of environmental variability. Yet, epinecrotic communities were influenced by environmental variation associated with location. These communities did exhibit clear patterns of succession suggesting that this is a robust process that will consistently occur. The implications are that epinecrotic communities have the potential to be used for forensic applications by associating successional changes with time to determine a post mortem submersion interval; however, the influence of environmental factors indicates that there is not one overarching successional pattern. Further investigation is required to determine how to utilize this pattern in a universal manner to determine the post-mortem submersion interval. Community structure is related to function and so biofilm functional abilities may also change throughout succession.
Future studies can investigate functional succession using metagenomics or metatranscriptomics to try and determine if functional redundancy of different bacterial species produces a common functional profile successional pattern.
176
Literature Cited
Anderson, M. J. 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecology 26:32–46.
Augspurger, C., C. Karwautz, M. MuBmann, H. Daims, and T. J. Battin. 2010. Drivers of bacterial colonization patterns in stream biofilms. FEMS Microbiology Ecology 72:47–57.
Barrios, M., and M. Wolff. 2011. Initial study of arthropods succession and pig carrion decomposition in two freshwater ecosystems in the Colombian Andes. Forensic Science International 212:164–172.
Battin, T. J., L. A. Kaplan, J. D. Newbold, and C. M. E. Hansen. 2003. Contributions of microbial biofilms to ecosystem processes in stream mesocosms. Nature 426:439– 42.
Benbow, M., J. Pechal, J. Lang, R. Erb, and J. Wallace. (n.d.). The potential of high- throughput metagenomic sequencing of aquatic bacterial communities to estimate the postmortem submersion interval. Journal of Forensic Sciences.
Bienhold, C., A. Boetius, and A. Ramette. 2012. The energy-diversity relationship of complex bacterial communities in Arctic deep-sea sediments. ISME J 6:724–732.
Bormann, F., and G. Likens. 1979. Patterns and process in a forested ecosystem. Springer-Verlag, New York, NY.
Cardinale, M., L. Brusetti, P. Quatrini, S. Borin, A. M. Puglia, A. Rizzi, E. Zanardini, C. Sorlini, C. Corselli, and D. Daffonchio. 2004. Comparison of different primer sets for use in automated ribosomal intergenic spacer analysis of complex bacterial communities. Applied and Environmental Microbiology 70:6147–6156.
Casamatta, D. A., and R. G. Verb. 2000. Algal colonization of submerged carcasses in a mid-order woodland stream. Journal of Forensic Sciences 45:1280–1285.
Claeson, S. M., J. L. Li, J. E. Compton, and P. A. Bisson. 2006. Response of nutrients, biofilm, and benthic insects to salmon carcass addition. Canadian Journal of Fisheries and Aquatic Sciences 63:1230–1241.
Clements, F. E. 1936. Nature and structure of the climax. Ecology 24:252–284.
Costerton, J. W., Z. Lewandowski, D. E. Caldwell, D. R. Korber, and H. Lappin-Scott. 1995. Microbial Biofilms. Annual Review of Microbiology 49:711–745.
177
Dick, I. P., and R. C. Scott. 1992. Pig ear skin: as an in‐vitro model for human skin permeability. Journal of Pharmacy and Pharmacology 44:640–645.
Dickson, G. C., R. T. M. Poulter, E. W. Maas, P. K. Probert, and J. A. Kieser. 2011. Marine bacterial succession as a potential indicator of postmortem submersion interval. Forensic Science International 209:1–10.
Docherty, K. M., K. C. Young, P. A. Maurice, and S. D. Bridgham. 2006. Dissolved organic matter concentration and quality influences upon structure and function of freshwater microbial communities. Microbial Ecology 52:378–88.
Van Dorst, J., A. Bissett, A. S. Palmer, M. Brown, I. Snape, J. S. Stark, B. Raymond, J. McKinlay, M. Ji, T. Winsley, and B. C. Ferrari. 2014. Community fingerprinting in a sequencing world. FEMS Microbiology Ecology 89:316–330.
Espeland, E. M., S. N. Francoeur, and R. G. Wetzel. 2001. Influence of algal photosynthesis on biofilm bacterial production and associated glucosidase and xylosidase activities. Microbial Ecology 42:524–530.
Fechner, L. C., F. Vincent-Hubert, P. Gaubert, T. Bouchez, C. Gourlay-Francé, and M.- H. Tusseau-Vuillemin. 2010. Combined eukaryotic and bacterial community fingerprinting of natural freshwater biofilms using automated ribosomal intergenic spacer analysis. FEMS Microbiology Ecology 74:542–53.
Fierer, N., D. Nemergut, R. Knight, and J. M. Craine. 2010. Changes through time: integrating microorganisms into the study of succession. Research in Microbiology 161:635–642.
Fisher, M. M., and E. W. Triplett. 1999. Automated approach for ribosomal intergenic spacer analysis of microbial diversity and its application to freshwater bacterial communities. Applied and Environmental Microbiology 65:4630–4636.
Freeman, C., P. J. Chapman, K. Gilman, M. A. Lock, B. Reynolds, and H. S. Wheater. 1995. Ion exchange mechanisms and the entrapment of nutrients by river biofilms. Hydrobiologia:61–65.
Freeman, C., and M. a. Lock. 1995. The biofilm polysaccharide matrix: A buffer against changing organic substrate supply? Limnology and Oceanography 40:273–278.
Giller, P. S., and B. Malmqvist. 1998. The biology of streams and rivers. (M. J. Crawley, C. Little, T. R. E. Southwood, and S. Ulfstrand, Eds.) Biology of Habitats. Oxford University Press Inc., New York.
Gleason, H. A. 1926. The individualistic concept of the plant association. Bulletin of the Torrey Botanical Club 53:7–26.
178
Golladay, S. W., and R. L. Sinsabaugh. 1991. Biofilm development on leaf and wood surfaces in a boreal river. Freshwater Biology 25:437–450.
Van der Grinten, E., M. Janssen, S. G. H. Simis, C. Barranguet, and W. Admiraal. 2004. Phosphate regime structures species composition in cultured phototrophic biofilms. Freshwater Biology 49:369–381.
Haefner, J. N., J. R. Wallace, and R. W. Merritt. 2004. Pig decomposition in lotic aquatic systems: the potential use of algal growth in establishing a postmortem submersion interval (PMSI). Journal of Forensic Sciences 49:1–7.
Hladyz, S., R. a. Cook, R. Petrie, and D. L. Nielsen. 2011. Influence of substratum on the variability of benthic biofilm stable isotope signatures: implications for energy flow to a primary consumer. Hydrobiologia 664:135–146.
Hoagland, K. D., S. C. Roemer, J. R. Rosowski, and R. Rosowski. 1982. Colonization and community structure of two periphyton assemblages, with emphasis on the diatoms (Bacillariophyceae). American Journal of Botany 69:188–213.
Hobischak, N. R., and G. S. Anderson. 2002. Time of submergence using aquatic invertebrate succession and decompositional changes. Journal of Forensic Sciences 47:142–151.
Hyde, E. R., D. P. Haarmann, A. M. Lynne, S. R. Bucheli, and J. F. Petrosino. 2013. The living dead: bacterial community structure of a cadaver at the onset and end of the bloat stage of decomposition. PLOS ONE 8:1–10.
Jacobi, U., M. Kaiser, R. Toll, S. Mangelsdorf, H. Audring, N. Otberg, W. Sterry, and J. Lademann. 2007. Porcine ear skin: an in vitro model for human skin. Skin Research and Technology 13:19–24.
Keiper, J. B., E. G. Chapman, and B. A. Foote. 1997. Midge larvae (Diptera: Chironomidae) as indicators of postmortem submersion interval of carcasses in a woodland stream: a preliminary report. Journal of Forensic Sciences 42:1074–1079.
Lear, G., Y. Dong, and G. Lewis. 2010. Comparison of methods for the extraction of DNA from stream epilithic biofilms. Antonie van Leeuwenhoek 98:567–571.
Lock, M., and H. Hynes. 1976. The fate of “dissolved” organic carbon derived from autumn-shed maple leaves (Acer saccharurn) in a temperate hard-water stream. Limnology and Oceanography 21:436–443.
McCune, B., and J. B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR.
179
McNamara, C. J., and L. G. Leff. 2004. Response of biofilm bacteria to dissolved organic matter from decomposing maple leaves. Microbial Ecology 48:324–330.
Metcalf, J. L., L. W. Parfrey, A. Gonzalez, C. L. Lauber, D. Knights, G. Ackermann, G. C. Humphrey, M. J. Gebert, W. Van Treuren, and D. Berg-Lyons. 2013. A microbial clock provides an accurate estimate of the postmortem interval in a mouse model system. Elife 2:e01104.
Miller, D. N., J. E. Bryant, E. L. Madsen, and W. C. Ghiorse. 1999. Evaluation and optimization of DNA extraction and purification procedures for soil and sediment samples. Applied and Environmental Microbiology 65:4715–4724.
Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O’Hara, G. L. Simpson, P. Solymos, M. H. Stevens, and H. Wagner. 2015. vegan: Community Ecology Package.
Olapade, O. A., and L. G. Leff. 2006. Influence of dissolved organic matter and inorganic nutrients on the biofilm bacterial community on artificial substrates in a northeastern Ohio, USA, stream. Canadian Journal of Microbiology 52:540–9.
Parmenter, R. R., and V. A. Lamarra. 1991. Nutrient cycling in a freshwater marsh: The decomposition of fish and waterfowl carrion. Limnol. Oceanogr. 36:976–987.
Pechal, J., T. Crippen, and M. Benbow. 2014. The potential use of bacterial community succession in forensics as described by high throughput metagenomic sequencing. Int J Legal Med 128:193–205.
Ramette, A. 2009. Quantitative community fingerprinting methods for estimating the abundance of operational taxonomic units in natural microbial communities. Applied and Environmental Microbiology 75:2495–505.
Romaní, A. M., H. Fischer, C. Mille-lindblom, L. J. Tranvik, J. Tranvik, and A. M. Romani. 2006. Interactions of bacteria and fungi on decomposing litter: differential extracellular enzyme activities. Ecology 87:2559–2569.
Romaní, A. M., H. Guasch, I. Muñoz, J. Ruana, E. Vilalta, T. Schwartz, F. Emtiazi, and S. Sabater. 2004. Biofilm structure and function and possible implications for riverine DOC dynamics. Microbial Ecology 47:316–28.
Sabater, S., S. V Gregory, and J. R. Sedell. 1998. Community dynamics and metabolism of benthic algae colonizing wood and rock substrata in a forest stream. Journal of Phycology 34:561–567.
Schoenly, K. G., N. H. Haskell, R. D. Hall, and J. R. Gbur. 2007. Comparative performance and complementarity of four sampling methods and arthropod
180
preference tests from human and porcine remains at the Forensic Anthropology Center in Knoxville, Tennessee. Journal of Medical Entomology 44:881–894.
Sekkat, N., Y. N. Kalia, and R. H. Guy. 2002. Biophysical study of porcine ear skin in vitro and its comparison to human skin in vivo. Journal of Pharmaceutical Sciences 91:2376–2381.
Sinsabaugh, R. L., S. W. Golladay, and A. E. Linkins. 1991a. Comparison of epilithic and epixylic biofilm development in a boreal river. Freshwater Biology 25:179–187.
Sinsabaugh, R. L., D. Repert, T. Weiland, S. W. Golladay, and A. E. Linkins. 1991b. Exoenzyme accumulation in epilithic biofilms. Hydrobiologia 222:29–37.
Steinman, A. D., G. A. Lamberti, and L. P.R. 2007. Biomass and pigments of benthic algae. Pages 357–380 in F. R. Hauer and G. A. Lamberti, editors. Methods in Stream Ecology, 2nd edition. Academic Press, New York.
Sutherland, I. W. 2001. The biofilm matrix – an immobilized but dynamic microbial environment. Trends in Microbiology 9:222–227.
Tank, J. L., and W. K. Dodds. 2003. Nutrient limitation of epilithic and epixylic biofilms in ten North American streams. Freshwater Biology 48:1031–1049.
Tomberlin, J. K., and P. H. Adler. 1998. Seasonal colonization and decomposition of rat carrion in water and on land in an open field in South Carolina. Journal of Medical Entomology 35:704–709.
Vance, G. M., J. K. VanDyk, and W. A. Rowley. 1995. A device for sampling aquatic insects associated with carrion in water. Journal of Forensic Sciences 40:479–482.
Wallace, J. R., R. W. Merritt, R. Kimbirauskas, M. E. Benbow, and M. McIntosh. 2008. Caddisflies assist with homicide case: determining a postmortem submersion interval using aquatic insects. Journal of Forensic Sciences 53:219–221.
Weiher, E., and P. Keddy. 1999. Ecological assembly rules; Perspectives, advances, retreats. Cambridge University Press, Cambridge, UK.
Zhou, J., M. A. Bruns, and J. M. Tiedje. 1996. DNA recovery from soils of diverse composition. Applied and Environmental Microbiology 62:316–322.
Zimmerman, K. A., and J. R. Wallace. 2008. The potential to determine a postmortem submersion interval based on algal/diatom diversity on decomposing mammalian carcasses in brackish ponds in Delaware. Journal of Forensic Sciences 53:935–941.
181
Tables
Table 1. Water quality parameters in Farmersville, OH were measured when substrates were deployed and at every sampling date 15
3- - m above and below the uppermost and lowermost carcasses, respectively. Nitrate (mg/L NO -N), nitrite (mg/L NO2 -N), sulfate
2- (SO4 mg/L), ammonia (mg/L NH3-N), alkalinity (mg/L CaCO3), and total suspended solids (TSS mg/L) were measured in the lab using EPA approved protocols (Hach® Company, Loveland, CO). Specific conductivity (SpCond µS/cm), total dissolved solids (TDS mg/L), dissolved oxygen (DO mg/L), pH, and temperature (°C) were recorded using a YSI 6600 v2 Sonde (YSI Inc, Yellow Springs,
OH). Bolded means denote significant differences (P < 0.05) between the upstream and downstream sites using a two tailed t-test.
3- 2- Date Time NO NO2 SO4 NH3 CaCO3 TSS SpCond TDS DO pH Temp 29 June 0 2.35 0.054 21 0.095 291 26 617 0.4 3.78 7.92 20.6 02 July 3 2.95 0.025 20.5 0.09 295 10 636 0.41 5.56 7.99 21.3 06 July 7 2.4 0.049 22 0.22 332 20 652 0.42 3.21 7.91 21.5 09 July 10 2.55 0.0155 23 0.125 294 10 652 0.42 5.28 8.02 21.8 13 July 14 3.4 0.0335 23 0.255 307 - 650 0.42 4.57 8.17 18.6 16 July 17 2.3 0.02 23.5 0.18 335 27 644 0.42 3.65 8.01 21.6 20 July 21 1.95 0.023 27 0.15 308 20 644 0.42 2.84 7.96 21.5 23 July 24 1.95 0.019 27 0.135 295 27.5 653 0.43 3.68 8.04 21.3
182
2.48 ± 0.030 ± 0.16 ± 307 ± 20 ± 643 ± 0.42 ± 4.07 ± 8.00 ± 21.0 ± Mean - 23 ± 2 0.49 0.014 0.06 17 7 12 0.01 0.97 0.08 1.1 Upstream 2.39 ± 0.034 ± 0.14 ± 313 ± 15 ± 666 ± 0.43 ± 2.87 ± 7.85 ± 19.7 ± - 24 ± 4 / Shaded 0.66 0.024 0.05 22 6 25 0.01 0.96 0.10 1.4 Downstre 2.58 ± 0.025 ± 0.17 ± 302 ± 34 ± 621 ± 0.40 ± 5.27 ± 8.16 ± 22.3 ± - 22 ± 1 am/Open 0.67 0.020 0.11 18 19 13 0.01 1.36 0.08 0.7
183
Table 2. Water quality parameters in Millersville, PA of dissolved oxygen (mg/L), pH, specific conductivity (µS/cm), water temperature (oC), total dissolved solids (g/L), oxidation reduction potential (mV), and salinity (ppt) were measured at a single location
30 m upstream of the uppermost carcass and 30 m downstream of the lowermost carcass on each sampling day using a Horiba®
(Kyoto, Japan) Multi Water Quality Checker (U-50 Series).
Date Time (Days) DO SpCond ORP pH Salinity TDS Temperature 26 June 0 6.94 966 93 8.2 0.48 0.483 18.3 29 June 3 4 928 64.5 8.1 0.46 0.464 20.5 03 July 7 6.8 1038 66.8 8.5 0.52 0.519 20.1 06 July 10 6.55 950 47.8 8.2 0.47 0.475 20.7 10 July 14 4.05 908 37.2 8.1 0.45 0.454 22.2 13 July 17 4.34 1130 37.9 8.1 0.57 0.565 20.1 17 July 21 4.78 963 36.7 8.1 0.48 0.459 23.5 20 July 24 5.59 779 29.6 8.1 0.38 0.389 21.8 Mean - 5.38 ± 2.25 957 ± 101 51.7 ± 21.5 8.2 ± 0.1 0.47 ± 0.05 0.476 ± 0.051 20.9 ± 1.6
184
Figures
Figure 1. Bacteria community structure was visualized using nonmetric multidimensional scaling (3-D, R2 = 0.97, stress = 0.18) and overlaid with significant factors
(PERMANOVA) of biofilm type and location of study.
185
Figure 2. Bacteria communities were separated into epinecrotic (3-D, R2 = 0.97, stress =
0.17) and epilithic (3-D, R2 = 0.98, stress = 0.14) biofilms and community structure was visualized using nonmetric multidimensional scaling. Location of study significantly influenced community structure of both biofilm types as determined by PERMANOVA, but only epilithic biofilms were significantly affected by canopy.
186
Figure 3. Succession of Dayton (3-D, R2 = 0.98, stress = 0.14) and Millersville (3-D, R2 =
0.99, stress = 0.11) epinecrotic biofilms were depicted with NMDS ordination using the days of decomposition as an overly, which was significant for both communities determined by PERMANOVA.
187
Figure 4. Bacterial community structure of epilithic biofilms in Dayton is depicted using
NMDS ordination (3-D, R2 = 0.98, stress = 0.13) and overlaid with days, tile placement, and canopy. Both days and canopy were significant determined by PERMANOVA. Note that canopy ordinates along the NMDS Axis 3.
188
CHAPTER 6
CONCLUSION
Stream biofilms were studied within the context of an ecological theory framework that was grounded primarily in community ecology. These complex communities function as micro-ecosystems, making the approach in studying how environmental factors affect biofilms feasible by using established ecological theories and principles. Successional development is a foundation for many biofilm studies and studying these patterns was incorporated into each chapter of this dissertation. Succession occurred in each instance and further supports that this is a universal pattern of developing biofilms. Nevertheless, successional trajectories were highly influenced by environmental factors. Light, flow, and grazing were of particular focus, and each factor affected biofilm communities in a way that altered community composition and was more influential than successional changes.
The impact of environmental factors on biofilms can play a larger role within the stream ecosystem by affecting ecosystem processes. The abiotic effects of reduced light and flow from leaf deposition can have bottom up effects on stream ecosystem processes through altered epilithic biofilms. This phenomenon is rarely addressed when investigating leaf deposition impacts on streams and warrants consideration. In addition,
189
the changing abiotic factors during autumn that were not associated with leaf deposition more important than frequent spate disturbances in influencing community composition.
Different pioneer communities were selected during early and late autumn, which suggests that environmental factors acted as a filter by either changing the species pool or affecting what organisms could survive and thrive.
Bottom-up effects of biofilms can also impact the food web. Light, flow, and successional state were all important factors in determining grazing activity. In addition, the top-down effects of grazers influenced community composition of both bacterial and eukaryotic communities. The connection between grazers and the bacterial community is an area that has not been well investigated and may provide important insights into the relationship between biofilms and grazers, especially within the context of the invertebrate microbiome. The results suggest that biofilm community assembly follows successional patterns but there may be a timing component of the different controls where bottom-up forces of abiotic factors are more important during the beginning stages and top-down forces of grazers are more important during later stages.
The influence of environmental factors on stream biofilms can even play a role in the forensic sciences. Epinecrotic biofilms have been proposed as a way to determine the
PMSI of aquatic human remains, and it was found that these biofilms are unique to the swine substrate regardless of location. Epinecrotic biofilms were clearly different than epilithic biofilms; however, location did influence community composition. This suggests that environmental factors dictate the species pool and that there is not one overarching successional pattern. It may be more prudent to investigate functional traits, rather than taxonomic, diversity throughout succession for the purposes of establishing a PMSI.
190