Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mycorrhizal facilitation of kin recognition in interior Douglas-fir (Pseudotsuga menziesii var. glauca) Asay, Amanda Karlene 2013

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

Item Metadata

Download

Media
24-ubc_2013_fall_asay_amanda.pdf [ 1013.75kB ]
Metadata
JSON: 24-1.0103374.json
JSON-LD: 24-1.0103374-ld.json
RDF/XML (Pretty): 24-1.0103374-rdf.xml
RDF/JSON: 24-1.0103374-rdf.json
Turtle: 24-1.0103374-turtle.txt
N-Triples: 24-1.0103374-rdf-ntriples.txt
Original Record: 24-1.0103374-source.json
Full Text
24-1.0103374-fulltext.txt
Citation
24-1.0103374.ris

Full Text

Mycorrhizal facilitation of kin recognition in interior Douglas-fir(Pseudotsuga menziesii var. glauca)byAMANDA KARLENE ASAYB.Sc., Brown University, 2010A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Forestry)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)October 2013? Amanda Karlene Asay 2013iiAbstractInsight into influences on successful seedling establishment could be essential to futureregeneration of British Columbia?s interior Douglas-fir (Pseudotsuga menziesii var. glauca)forests, particularly as climate changes. Areas of harsh climatic conditions have low regenerativecapacity and require management decisions leading to enhanced seedling establishment.Variable retention harvesting and natural regeneration from residual trees, for example, maybecome increasingly important for their locally adaptive traits as climate changes. Kinrecognition, mycorrhizal networks, or the combination of the two may be important mechanismsfor enhanced seedling establishment in these regions. We examined the effects of relationship(kin vs. non-kin) and mycorrhizal networks on regeneration from seed in greenhouse and fieldsettings. In the greenhouse, kin recognition was evident in differing foliar microelement (Fe, Mo,Al and Cu) and growth variables (total leaf area, volume and stem length) according torelationships between seedlings. Kin recognition was also weakly evident in the field, where itwas expressed as differential survivorship among kin versus non-kin seedlings. Kin selectionwas evident in the greenhouse, where microelement content of kin was greater than non-kin.Greater mycorrhizal colonization of kin compared to non-kin as well as greater donor total leafarea, volume and stem length also suggest kin selection, although not consistently in allexperiments. In the field, survivorship was greater among non-kin; however, detection of kinrecognition may have been masked by the large effects of site and seed origin on germinationand survival. Mycorrhizal networks and carbon transfer occurred within all greenhouse seedlingpairs, and enhanced mycorrhization of kin suggests network colonization was involved in kinselection, but our data does not strongly support our hypothesis that kin recognition wasfacilitated by mycorrhizal networks. While the mechanism of kin recognition is still not welliiiunderstood, we provided evidence of kin recognition in interior Douglas-fir seedlings,particularly those that originate from harsh climates, and observed subtle indicators of kinselection or reduction of competition due to a close genetic relationship.  Accounting for thesephenomena in forest management could be helpful to successful regeneration of interiorDouglas-fir forests as stresses associated with climate change increase.ivPrefaceContributions by Amanda Asay:This thesis is an original, unpublished product of the author, Amanda Asay. Theidentification of the research topic was done by supervisor Dr. Suzanne Simard and modified byAmanda Asay and Dr. Suzanne Simard in conjunction. The design of the research program wasconducted by Amanda Asay and Dr. Suzanne Simard with input from committee members Dr.Sally Aitken, Dr. Daniel Durall and Dr. Susan Dudley. Dr. Brian Pickles assisted with the designof the 13CO2 labelling design and morphotying procedures. Dr. William Mohn and RolandWilhelm assisted in the 13CO2 labelling design only. The performance of all of parts of theresearch, both field and greenhouse, were led by Amanda Asay with exception to the DNAanalysis, which was sent to the Irving K. Barber School at UBC-Okanagan, and the microwavedigestion/ICP (Inductively Coupled Plasma-Optical Emission Spectrometer) and % C, N and Sanalyses, which were sent to the British Columbia Ministry of Environment AnalyticalChemistry Laboratory, and the EA-IRMS (elemental analysis-isotope ratio mass spectrometry),which was sent to the Stable Isotope Facility at UBC-Vancouver. The morphotyping process wasshared by Dr. Brian Pickles (Chapter 2) and Amanda Asay (Chapter 2 and 3). The analysis of theresearch data was done by Amanda Asay with suggestions from Dr. Suzanne Simard andstatistical advice from Dr. Valerie LeMay.vTable of contentsAbstract ?????????????????????????.?????????iiPreface ?????...???????????????????..?????????.ivTable of contents ??????????????????????????????...vList of tables ???????????????????????????????...viiiList of figures ??????????????????????????.????...?xAcknowledgements ?????????????????????????????.xiiChapter 1: Introduction ???????????????????????????...1Study species ?????????????????????????????..1Kin recognition and kin selection ?????????????????????...3Mycorrhizal networks ??????????????????????????.5Plant interactions ????????????????????????????8Overview of the thesis ?????????????????????????..11Chapter 2: Control cross pollinated full-sibling seedling pairs may recognize kin throughmycorrhizal networks in a greenhouse setting ??????????????????.13Introduction ?????????????????????????????...13Methods ???????????????????????????????.17Experimental design and treatments ?????????????????.17Experimental setup ????????????????????????1813CO2 labelling ?????????????????????????..21Measurements ?????????????????????????...22Data analysis ??????????????????????????.24Results ???????????????????????????????...2413C transfer ???????????????????????????24Growth ????????????????????????????..25Nutrients ????????????????????????????27Soils ?????????????????????????????..28Discussion ??????????????????????????????.28Kin effects ???????????????????????????.28viMycorrhizal network effects ????????????????????.30Relationship x network effects ?????..??????????????31Family effects??????????????????????????.33Conclusions ...?????????????????????????????33Chapter 3: Field collected, greenhouse grown sibling pairs may recognize kin throughmycorrhizal networks in a greenhouse setting ..????????.?????????.46Introduction ?????????????????????????????...46Methods ???????????????????????????????.51Experimental design and treatments ?????????????????.51Experimental setup ????????????????????????52Measurements ?????????????????????????...55Data analysis ??????????????????????????.56Results ???????????????????????????????...57Mycorrhizal colonization ..?????????????????????57Growth ????????????????????????????..57Nutrients ????????????????????????????58Germination rates ????????????????????????..59Soils ?????????????????????????????..60Discussion ??????????????????????????????.60Kin effects ???????????????????????????.60Seed origin effects ????????????????????????.62Level of relatedness ???????????????????????..63Mycorrhizal network effects ????????????????????.64Conclusions ...?????????????????????????????66Chapter 4 Mature trees may preferentially support kin seedlings in germination andsurvival through mycorrhizal networks ?.??????????????.?????.82Introduction ?????????????????????????????...82Methods ???????????????????????????????.83Study sites ???????????????????????????.83Experimental design and treatments ?????????????????.84Experimental setup ????????????????????????85viiObservations and data collection ????...?????????????...86Data analysis ??????????????????????????.87Results ???????????????????????????????...87Germination and survival ..?????????????????????87Site and seed origin ????????.???????????????..89Climate, soil moisture and relationship ??..???.???.???????89Climate, soil moisture and networks ?...???????????????..90Discussion ??????????????????????????????.91Germination and survival rates ???????????????????.91Kin effects ???????????????????????????.92Mycorrhizal network effects ????????????????????.93Site and seed origin effects ????????????????????...93Climate effects ?????????????????????????..94Conclusions ...?????????????????????????????95Chapter 5: Summary and conclusions ?????????????????????109Review of objectives ?????????????????????????...109Summary of main findings ???????????????????????..110Major objectives ???????????????????????.?.110Minor objectives ????????????????????????.112Contributions to the field of study ????????????????????..115Limitations of studies????????????.?????????????.116Future directions ???????????????????????????.116References ????????????????????????????????..118viiiList of tablesTable 2.1. Family pairings to create kin and non-kin treatments in full sibling greenhouseexperiment ?????????????????????????????????.35Table 2.2. Analysis of variance for network, relationship and interaction (network x relationship)effects on foliar nutrient content and 13C transfer in recipient seedlings ??...??????...36Table 2.3. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in recipient seedlings ?????????????????....37Table 2.4. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in donor seedlings ???.??????????????..?..38Table 2.5. Analysis of variance for network, relationship and interaction (network x relationship)effects on foliar nutrient content in donor seedlings ?????????????????39Table 2.6. Analysis of variance, means ratio, and difference between means for the effect ofautoclaving on soil used in mesh bags accessed by recipients (autoclaved) and pots accessed bydonors (non-autoclaved) on soil nutrients, CEC (cation excahange capacity) and pH ????40Table 3.1. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in donor seedlings ????.???????????????68Table 3.2. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in recipient seedlings ??????????????????69Table 3.3. Analysis of variance for network, relationship and interaction (network x relationship)effects on foliar nutrient content and 13C transfer in recipient seedlings ?????????.70Table 3.4. Analysis of covariance for network, relationship and interaction (network xrelationship) effects, with seed origin included as the covariate and its effects on growth andgermination rates recipient seedlings ...??????????????????????.71Table 4.1. Geographic location and estimates of climatic variables for each of the study sites(Alex Fraser Research Forest, Farwell Canyon and Paska Lake) obtained from ClimateBC(Wang et al. 2012). ???????????????????...??????????.97Table 4.2. Climate data for each study site measured during the experimental period (May-August 2012) including ambient air temperature, soil temperature and moisture ??????98Table 4.3. Climate variables for each study site derived from data measured during theexperimental period (May-August 2012) including soil temperature:moisture index, degree daysbelow 4?C as well as the date range in which those degree days occurred ????..????99ixTable 4.4. Germination and Survival by total number (#) and percentage (%) for the main effectsof network and relationship (kin and non-kin) measured in # or % of treatment units (360 total)and # or % of total seeds sown (1800 total) or germinated (for survival) ..????????100Table 4.5. Logistic regression testing for the seed germination probability in response torelationship, mesh treatment (network), the combination of relationship and mesh as well as site?????????????????????????????????????..101Table 4.6. Logistic regression testing for the probability that a surviving seedling will be presentin response to relationship, mesh treatment (network), the combination of relationship and meshas well as site ???????????????????????????...??...?102Table 4.7. Germination and Survival by total number (#) and percentage (%) for each site and byseed origin location measured in # or % of treatment units (360 total) and # or % of total seedssown (1800 total) or germinated (for survival) ??????????????????...103Table 4.8. Total number of seeds that germinated in each site (columns) and which site thegerminated seed originated from (rows) ?????????????????????.104Table 4.9. Total number of seedlings that survived in each site (columns) and site of seed originof the surviving seedling (rows) ????????????????????????.105xList of figuresFigure 2.1. Photos of (a) a top view of an experimental unit (pot) just after the donor seedlingwas sown, (b) experimental unit at the onset of the 13CO2 labelling period, recipient age 4 months,with donor age 10 months and (c) fitting a donor seedling with a 13CO2 labelling bag ??...?41Figure 2.2. 13C transfer to recipient seedlings measured in excess 12C under the four network andrelationship combinations as well as all networked compared to all no network recipientseedlings ??????????????????????????????????42Figure 2.3. Least square means of recipient seedling (a) average weight per colonized root tip,(b) above ground biomass, (c) total biomass, (d) root weight, (e) stem length and (f) stem weightunder the four relationship and network combinations (interaction effects) ???.????...43Figure 2.4. . Least square means of height growth rate in donor seedlings under the fourrelationship and network combinations (interaction effects)??.????????????44Figure 2.5. Least square means of recipient seedling (a) iron (Fe), (b) molybdenum (Mo) , (c)magnesium (Mg), and (d) sulphur (S) content in grams under the four relationship and networkcombinations (interaction effects) ?????????????...??????????.45Figure 3.1. Google map (? 2013 Google) of south-western British Columbia including threefield research sites (Alex Fraser Research Forest, Farwell Canyon and Paska Lake sites), the seedsource for the full sibling experiment (Kalamalka seed orchard), the soil source for thegreenhouse experiments (Soil source: greenhouse experiments) and the location of thegreenhouse (UBC greenhouse) ?????????????????????????.72Figure 3.2. Least square means of recipient seedling (a) average weight per colonized root tipand (b) percent of total root weight colonized under the relationship effects without considerationof the network treatment used ??????????????????.??????..?73Figure 3.3. Least square means of donor seedling (a) volume (total leaf area x stem length) and(b) total leaf area under the relationship effects without consideration of the network treatmentused ????????????????????.???????????????..74Figure 3.4. Growth rate, total biomass, and root weight least square means of recipient seedlingsdifferentiated by network effect ?...???????????????????????75Figure 3.5. Least square means of root weight in donor seedlings under the four relationship andnetwork combinations (interaction effects) ?...??????????????????..76Figure 3.6. The variation of (a) growth rate and (b) germination rate (least square means)between seeds originating from Farwell Canyon, Alex Fraser Research Forest and Paska Lake??????????????????????????????????????77xiFigure 3.7. Aluminum (Al) and copper (Cu) content least square means of recipient seedlingsdifferentiated by relationship effect ???????????????????????..78Figure 3.8. same site. Least square means of (a) aluminum (Al) and (b) copper (Cu)differentiated by relationship effect between kin pairs (seeds from same parent trees), non-kin(same site) (seeds from different parent trees within the same site) and non-kin (seeds fromdifferent parent trees from different sites) ...????????????????????79Figure 3.9. Significant least square mean differences in recipient seedlings under the fourrelationship and network combinations (interaction effects) found in copper (Cu), iron (Fe) andzinc (Zn) ...?????????????????????????????????80Figure 3.10. Least square means of germination rate in recipient seedlings under the fourrelationship and network combinations (interaction effects) and seed origin effects??.??...79Figure 4.1. Odds ratio values, presented on a logarithmic scale, for relation, network, site andrelationship x network effects in the logistic regression model predicting (a) germination (b)survival ????????????????????????????????.?..106Figure 4.2. The number of (a) kin and non-kin germinants, (b) kin and non-kin survivors, (c)network and no netwrok germinants and (d) network and no network survivors across the rangeof soil temperature:moisture index values found in the study sites ??????...????107Figure 4.3. The number of (a) kin and non-kin germinants, (b) kin and non-kin survivors, (c)network and no netwrok germinants and (d) network and no network survivors across the rangeof soil moisture content found in the study sites ??...???????????????.108xiiAcknowledgementsTo start with I would like to thank my supervisor, Dr. Suzanne Simard, for the countlesshours of guidance and support in the completion of this thesis. From the onset of this project shehas gone above and beyond to inspire, collaborate, and provide assistance in all aspects of thisstudy. I could not have completed this thesis without her help. I would like to thank all membersof my committee for their assistance at various stages of this study. To Dr. Sally Aitken, for herassistance and expertise in the genetics and population distribution of interior Douglas-fir as wellas a source of guidance, whenever needed, throughout the study. To Dr. Dan Durall, for hisassistance and expertise in mycorrhizal networks and below ground communities and foraccepting me into group discussions furthering my knowledge on the subject. To Dr. SusanDudley, for her assistance and expertise in kin recognition and kin selection in plant species.Everyone on my committee provided valuable information on the design, process and analysesinvolved in this thesis. I would also like to thank the Belowground Ecosystems Group forsuggestions at various points along the study, in particular the assistance of the Simard lab group,Cory Wallace, Matt Zustovic, Julia Amerongen Maddison, Deon Louw, Colin Mahony, AaronZwiebel and Melissa Dergousoff. A special thanks to Dr. Brian Pickles for all his advicethroughout the study as well as essential collaboration on the labelling portion of the study. Iwould also like to thank Roland Wilhelm for his assistance and the access to equipment duringthe labelling study and the analysis of its results. Funding for my Masters was provided by aUBC Faculty of Forests Graduate Fellowship, NSERC Discovery Grant of S Simard, andNSERC Strategic Grant of W Mohn, S Simard and S Hallam.11 IntroductionThere are two main questions explored in this thesis: (1) whether kin recognition occursamong interior Douglas-fir (Pseudotsuga menziesii var. glauca-(Beissn.) Franco), and (2)whether kin recognition, if present, is facilitated by mycorrhizal networks.Study speciesThe study was conducted on a common and economically valuable species in the interiorof British Columbia, Canada: interior Douglas-fir. Interior Douglas-fir forests are widelydistributed across western North America, ranging from north-central British Columbia (55oN,up to 760 m elevation) to northern Mexico (19oN, up to 3260 m elevation).  Broad variationoccurs in climate (precipitation range 410?3400 mm per year; mean July temperature 7?30oC;mean January temperature -9 to 8oC), disturbance regimes (e.g., stand maintaining to standreplacing fires) and site quality (very dry and poor to very moist and rich) (Hermann andLavender, 1990).Both pollen and seed from interior Douglas-fir are wind dispersed, which allows for large,continuous populations within its range that can span relatively large, disconnected geographicareas (Hamrick et al. 1992).  Douglas-fir is highly genetically diverse compared to other coniferswith a considerable amount of genetic variability between varieties correlated with theenvironmental conditions of the seed?s origin (Campbell and Sorensen, 1978, Rehfeldt, 1978,Campbell, 1986, Krutovsky et al. 2009). Population boundaries of Douglas-fir tend to bedelineated by areas of low elevation, such as rivers and valleys and high elevation, such asmountain ranges (Rehfeldt, 1989). Elevational topographic features are often only a barrier toseed dispersal but not pollen dispersal (causing the difference between markers of mtDNA,2associated with seed dispersal, and cpDNA, associated with pollen dispersal) and therefore geneflow between populations can persist (Gugger et al. 2010).Most of the genetic variation of Douglas-fir is due to high regional diversity. There islittle variation among populations within the same region (< 1% using mtDNA markers and4.1% using cpDNA markers) or among individuals within a population (7.8 % using mtDNAmarkers and 27.5 % using cpDNA markers) when compared with the variation among groups(91.5% and 68.3% respectively) (Gugger et al. 2010). This is consistent with Rehfeldt (1989)who found that most genetic variation occurred over significant differences in geographicdistance or elevation, which translated into adaptive differentiation depending on the number offrost free days in the region.In this study, mature, seed bearing trees from Paska Lake, Farwell Canyon and the AlexFraser Research Forest, all within British Columbia, were used as well as control cross pollinatedseeds from the Kalamalka Research Station. The Alex Fraser Research Forest, Knife Creek block(121.88?W, 52.05?N) and Farwell Canyon sites (122.63?W, 51.79?N) are located outside ofWilliams Lake in the Cariboo Chilcotin Coast Region. The Paska Lake site (120.67?W, 50.50?N)is located approximately 30 km southwest of Kamloops in the Thompson Okanagan Region. Dueto weak differentiation among populations in studies of coastal Douglas-fir (Pseudotsugamenziesii var. menziesii (Mirb.) Franco) in BC, we expected the Alex Fraser Research Forest andFarwell Canyon locations to represent similar ?Cariboo? populations (Krutovsky et al. 2009). ThePaska Lake site falls in a different geographic region known as the Thompson-Okanagan, and itlikely represents a significantly different population. The two sites in the Cariboo are dry withmild summers and cold winters whereas the site in the Thompson-Okanagan is dry, but has hotsummers and mild winters (Spittlehouse 2006) (addition climate data can be found in Tables 4.1-34.3). Comparisons between these sites were conducted to examine regional and withinpopulation or subpopulation effects.Kin recognition and kin selectionKin selection describes cooperation between genetically related individuals that canenhance their combined fitness despite the potential individual fitness cost of the cooperativebehaviour. The tendency for an individual to participate in these cooperative behaviours,potentially leading to kin altruism, is described in Hamilton?s rule [Hamilton 1964]: rB > Cwhere C is the cost to the individual participating in the cooperative behaviour, B is the benefit tothe relative or group of relatives and r is the degree of relatedness. A common example of thisphenomenon is in social insects such as bees. Due to the high degree of relatedness in the hivefrom all individuals sharing maternal genes from the queen, each individual can suffer highindividual cost to defend or otherwise benefit the hive and still have an overall positive effect onits genes? fitness (Platt and Bever 2009).Kin selection is not nearly as well understood in plants due to the problem of kinrecognition. Kin recognition is the ability to distinguish between kin and non-kin individuals.Animals and many insects have the advantage of context in order to recognize which individualsare their kin, such as bees in a hive, eggs in a nest or, even more clearly, a mother giving birth.Depending on seed dispersal tendencies a plant may just as easily be growing next to a kinindividual, non-kin member of the same species or even a different species. Due to this challengeof recognition, before we can establish that kin selection is occurring, we first must establish thatthere can be recognition of an individual that is closely related. Biedrzycki et al (2010) hasprovided strong evidence that kin recognition occurs in Arabidopsis thaliana, a weedy4herbaceous annual. They also found evidence that kin recognition is conveyed through the activerelease of soluble root exudates. They determined this by growing seedlings in a medium withthe exudates of a kin, a stranger or their own exudates and examined the number of lateral rootsdeveloped. The seedlings that were grown in the presence of stranger exudates developedsignificantly more lateral roots than those grown in either the kin or self-root exudates. Theyconfirmed it was the exudates that were conveying the neighbour identity by adding a rootsecretion inhibitor that eliminated the differences. This distinct difference suggests that theseplants can differentiate between a kin and a stranger by recognizing their root exudates.These results can be interpreted as kin selection as well. Lateral root development isconsidered a competitive trait in plants. When a plant produces fewer lateral roots due to itsrecognition of a kin neighbour, it can be said that it is sacrificing some of its below groundcompetitive ability to allow its kin to also succeed in close proximity (kin selection). In a reviewof kin recognition in several plant species by File et al. (2011), nine studies showed that kingroups had outperformed strangers (suggesting kin selection) and eleven studies showed strangergroups outperforming kin groups, with twenty one studies showing no differences betweengroups. This has sparked a debate over what the dominant process is among plants growing inclose proximity. When strangers are more successful, it is suggested that there is a high level ofkin competition occurring. Because related individuals are more likely to be phenotypicallysimilar in traits such as rooting depth, kin will be competing in the same niche. Strangers,however, can take advantage of niche partitioning, thus, exploiting slightly different niches. Bothindividuals are thus provided with enough resources to be successful. There has been evidencethat both of these processes are occurring in plant systems. Platt and Bever (2009) suggest thatthe nature of the competitive behavior that occurs depends heavily on population density. When5space is very limited, kin competition is very high, which could favour niche partitioning, butwhen there is open space to be utilized, the evolution toward cooperation within a related groupmay be favoured.The study of kin recognition and kin selection is still in its infancy. There is still muchresearch to be done before predicting interaction outcomes between plants is possible, if it is everpossible with any certainty. Understanding kin recognition is also complicated by resultssuggesting that it may be mediated by belowground mechanisms, making it even morechallenging for researchers. This study adds to this body of knowledge through examination ofkin recognition in a coniferous tree species, interior Douglas-fir, and by the examination of apossible mechanism of facilitation, mycorrhizal networks.Mycorrhizal networksFossil records suggest that fungus-plant symbioses have been occurring ever since plantsbegan to colonize land. It was this symbiosis, where each of the partners used the others?specialized traits, which likely allowed for much of the colonization to occur (Fortin et al. 2009,Smith and Read 1997). Due to its pivotal role in the evolution of land plants, eventually leadingto the evolution of land animals, this symbiosis has been considered a more powerful driver forevolution than competition, parasitism or predation (Fortin et al. 2009, Margulis 1981).The term mycorrhiza comes from the Greek myco, meaning fungus and riza, meaning root.Mycorrhiza symbiosis literally is a root and fungus living in physical contact. It has beenestimated that over 90% of plant species form some type of mycorrhiza (Smith and Read 2007,Fortin et al. 2009). Mycorrhizae are thought to have many functions, including providing mineralnutrition and water acquisition, protection against pathogens, resistance to environmental stresses,6soil aggregation, and hormonal activity for the plant as well as providing an essential carbonsource for the fungus. There is a wide diversity of specialized mycorrhizal fungi includingarbuscular, ecto, ericoid and orchid mycorrhizae. The focus on this thesis is on theectomycorrhizal fungi. These fungi form associations with the roots of higher plants and theseassociations are typically characterized by apoplastic growth of the fungus, and formation of ahartig net and mantle. Ectomycorrhizal fungi have been classified into exploration types usingcharacteristics and quantity of emanating hyphae or rhizomorphs (dense groupings of hyphae).They range from contact-exploration types to long-distance exploration types (Agerer 2001). Awell-known long-distance exploratory ectomycorrhizal complex is Rhizopogonvinicolor/vesiculosis, which associates specifically with Douglas-fir.  Long-distance explorershave differentiated rhizomorphs allowing for efficient water and nutrient transport (Agerer 2001).Mycorrhizae have also been shown to connect roots from different plant individuals ofthe same or different species (Molina et al. 1992). Resource sharing between individuals throughmycorrhizal networks, such as from hub trees to seedlings, has been shown to occur in Douglas-fir forests (Teste et al. 2009, Querejita et al. 2003, Brooks et al. 2006).  Resources in some casescan move between plants along source-sink gradients governed by differences in plantphysiology, such as photosynthetic rates or nutrient contents in plants, and by fungal factors suchas exploration strategy or network density (van der Heijden and Horton 2009, Simard et al. 2012).Regeneration facilitation through mycorrhizal networks appears to increase the regenerativecapacity and aid in the self-organization and stability of forests (Simard et al. 2013). A source-sink gradient or size difference has also been shown work in the opposing direction (Merrild et al.2013). When a size discrepancy was established between two tomato plants connected by a7mycorrhizal network, it resulted in preferential P uptake by the larger plant and P deficiency inthe smaller plant.Mycorrhizal networks can facilitate regeneration either by increasing fungal colonizationof new seedlings for greater resource uptake capacity, or by directly transferring resources (wateror nutrients) or other compounds from large trees to regenerating seedlings.  Most studies showthat some resource transfer is occurring through mycorrhizal networks, although some argue thatthe results are inconclusive (Whitfield 2007). The bigger questions are how much transfer isoccurring, which resources or compounds are being transferred, which individual is benefitting,and what are the ecological or fitness consequences? Of the resources examined (mainly carbon,nitrogen, phosphorus or water), carbon transfer has been the central focus in ectomycorrhizalnetworks. Most studies show that carbon flows through mycorrhizal networks from the source(tree), into to the connecting fungus, and through to the sink (seedling) without significant cost tothe source tree.  One study showed that bi-directional transfer occurs between source and sinkplants, but that there was a net transfer from source to sink plants (Simard et al. 1997). Whensource-sink relationships between plants shift, such as over the growing season or on an annualbasis, the direction of net transfer changes with it (Philip 2006, Deslippe and Simard 2011).When two equivalent plants compete for transferred carbon, cost/benefit considerations ofmycorrhizal networks are more complex. Nitrogen transfer also appears to be quite complex. It isstill unclear whether nitrogen flows to nitrogen fixing plants or to non-nitrogen fixing plants, andit appears to depend on each of their nitrogen requirements, and therefore the cost-benefit ratio isstill unknown (Selosse et al. 2006, Van der Heijden and Horton 2009, Whitfield 2007).Aside from general function and capability, some scientists have discussed whether plantinteractions mediated by mycorrhizal networks are predominantly mutualistic or competitive,8socialist or capitalist (Van der Heijden and Horton 2009). Evidence for a ?socialist? perspectivecomes from studies showing that resources can be more evenly distributed among plantsinvolved in the network (Perry et al. 1989). On the other hand, evidence for a ?capitalist?perspective show that larger, more resource-demanding or more mycorrhizal-dependent plantsbenefit more from mycorrhizal networks through larger total biomass gains than smaller plants(Van der Heijden and Horton 2009). The mycorrhizal fungus can also benefit when it connectsdifferent plants by increasing the number of healthy hosts from which it acquires carbon (Selosseet al. 2006). These ideas are largely theoretical and more research is necessary.It is generally accepted that mycorrhizal systems are important to plants and ecosystemsbut much more research is needed to improve our understanding of processes and patterns. Allareas of research involving mycorrhizal networks have been reported as ?poorly understood?,and even where interaction outcomes could be measured, the underlying process has been?currently unknown? or ?not yet clear?. Furthermore, network processes that have been studied,such as carbon transfer, have been viewed with a very critical eye by some who are notconvinced this is the work of mycorrhizal networks at all. We are still a long way from widelyaccepted theories and predictions. Due to the infancy of the study of mycorrhizal networks, anyand all research in this field will be harshly critiqued but also necessary to increase knowledge ofthe importance of plant-fungal relationships in ecosystem function.Plant interactionsMycorrhizal symbiosis has been found to be a mutualistic relationship in most cases,however, in many systems ?cheaters? evolve to exploit the mutualism without contributing to therelationship (a parasite essentially). Kiers et al. (2011) sought to determine whether9communication occurred between a plant and a fungus to stabilize their mutualistic relationshipand to prevent ?cheaters? from exploiting the symbiosis. They found that both the plant and thefungus were able to detect how much benefit they were receiving from their respectivefungus/plant partner and preferentially transfer resources in demand to a more cooperativepartner. Host plants receiving higher amounts of phosphorus from a particular fungal individualtransferred more carbon to that individual. Likewise, a fungus receiving more carbon from aparticular host root would provide it with more phosphorus. This is much like a market economy,with higher quality services being rewarded bidirectionally. This study made two novelcontributions: first, that there is measurable recognition and reaction in the mycorrhizalsymbiosis and, second, that mutualisms are maintained because both partners are important inmaintaining the relationship, contrasting with previous theory that one dominant partner held thefate of multiple hosts and essentially forced them into participation (Kiers et al. 2011).Not only does signaling and communication occur between plant roots and fungi within asingle mycorrhiza; evidence is mounting that plants communicate with each other throughmycorrhizal networks.  The movement of nutrients or water between plants through networksalong source-sink gradients can be considered a form of communication. Song et al. (2010)found evidence for biochemical signalling between plants connected by an arbuscularmycorrhizal network. Tomato plants growing near a blight-infected tomato plant were able to?eavesdrop? through mycorrhizal networks that their neighbour had up-regulated defenseenzymes to fight off this disease, and as a result were able to up-regulate their own defenseenzymes and prime themselves for attack. This reaction decreased disease incidence and severityof the healthy neighbours compared to non-mycorrhizal neighbours, mycorrhizal neighbours thatwere not connected via a mycorrhizal network, or neighbours connected to another healthy plant.10These treatment comparisons showed that the communication or signal transmission likelyoccurred through mycorrhizal networks, not through airborne volatiles. Healthy plants connectedto the infected individual via mycorrhizal networks were able to increase defense enzyme levelsand increase defensive gene regulation previous to any attack on themselves.  Although thechemical identity of the signals remains unknown, this study provides needed insight to plant-plant communication and opens the door to further research in identification of signalcompounds that may also operate in kin recognition (Song et al. 2010).Inter-plant communication through mycorrhizal networks does not always benefit allindividuals involved. Barto et al. (2011) showed that allelochemicals can also be transmittedthrough mycorrhizal networks. Allelopathy is the production of compounds used to inhibit thegrowth of a neighbouring plant. Transfer of allelochemicals from one plant to another throughmycorrhizal networks rather than the soil matrix has the advantage of more efficient and fasterchemical transmission over greater distances due to the protection of allelochemicals from soilmicrobes and due to cytoplasmic streaming through hyphal cells. Over the course of twoexperiments they found significantly decreased biomass and higher allelochemicalconcentrations in leaves and surrounding soils of receiver plants connected via a mycorrhizalnetwork. The study was conducted using heterospecific plants. This strongly suggests thatallelochemicals are being transported through these networks to inhibit neighbouring receiverplants and give the supplier of the allelochemical a competitive advantage (Barto et al. 2011).These developments provide interesting insight and generate new hypotheses aboutmechanisms underlying plant ecosystem dynamics. It is conceivable, for example, that aparticular plant species could recognize and select for its kin through intraspecific mycorrhizalnetworks, while simultaneously increasing competitive effects on non-kin neighbours by11releasing allelochemicals or signals or acquiring more nutrients from the network for greatercompetitive growth.Overview of the thesisThe main objectives of this thesis, addressed in each of the research chapters (2-4), wereto: 1) determine whether kin recognition is detectable in interior Douglas-fir seedlings; 2)determine whether kin recognition, if present, would present in a way that supports the kinselection theory; 3) to determine whether mycorrhizal networks mediated kin recognitionbetween seedling pairs (chapters 2 and 3) or between seedlings and parent trees (chapter 4). Theminor objectives that were addressed in specific chapters were to: 1) determine if kin recognitionability varied among distinct genotypes or ?families? of seedlings (chapter 2); 2) determine if theregion of seed origin affects kin recognition among seedlings grown in a common greenhouseenvironment (chapter 3); 3) determine if kin recognition occurs along a gradient of relatedness(chapter 3); 4) determine if the region of seed origin affects kin recognition among seedlingsgrown in the field with a variety of growing conditions (different sites) (chapter 4). Twoadditional minor objectives were examined across chapters and discussed in the concludingchapter (5); 5) to determine whether full sibling kin pairs from control cross pollination exhibiteddiffering kin recognition effects than did kin pairs collected from open-pollinated parent trees inthe field and; 6) to determine whether effects seen in the controlled environment of thegreenhouse would be detectable under natural climatic conditions in the field. Chapter 2 wasdesigned to evaluate kin recognition among control cross pollinated sibling pairs in a greenhouseenvironment. Seedling pairs were sown six months apart in order to encourage a source-sinkgradient between the older, ?donor? seedling and the younger ?recipient? seedling. Mycorrhizalnetwork access was controlled using mesh bags of two pore sizes: one allowing for fungal12hyphae to penetrate the pores and connect the rooting systems of the two seedlings, butpreventing root on root contact; and the other pore size preventing fungal hyphae as well as rootsfrom crossing the mesh barrier. Chapter 3 was designed to evaluate kin recognition among openpollinated seeds collected from six parent trees from three field sites (two trees per site) ininterior British Columbia in a greenhouse setting. Chapter 4 was designed to evaluate kinrecognition in a field setting using naturally pollinated seeds collected from the same parent treesas were used as host centers for kin and non-kin seedlings in the study.132 Control cross pollinated full-sibling seedling pairs may recognize kinthrough mycorrhizal networks in a greenhouse settingIntroductionThe forest industry in western North America is benefitted by extensive and productiveinterior Douglas-fir (Pseudotsuga menziesii var. glauca) forests for saw and pulp logs (BCMinistry of Forests, Land and Natural Resource Operations, 2012 annual report). Forestregeneration of interior Douglas-fir has long been problematic due to the harsh growingenvironment, particularly climatic aridity (Newsome et al. 1991, Huggard et al. 2005, Vyse et al.2006). This has spawned research on forest regeneration and health for over half a century(Lavender et al. 1990), but these problems are expected to amplify with climate change (Hamannet al. 2011, Wang et al. 2012). Two concepts that may aid the regeneration and stability of theseforests are: (1) kin selection, stemming from the ability of plants to recognize others that areclosely genetically related to them and support their growth and health (Dudley and File, 2007),and (2) mycorrhizal networks connecting a community of trees and seedlings and facilitatingwater and nutrient transfer over a source-sink gradient (Perry et al. 1989, Horton and Bruns 2001,Querejeta et al. 2003). We explore both concepts using uneven-aged pairs of interior Douglas-firseedlings in a greenhouse setting.Kin recognition in plants is defined as the ability to discriminate kin in competitiveinteractions (Dudley and File, 2007). This is known to present in at least two ways. Kin selectiondescribes an altruistic behaviour between kin pairs or groups where an individual could suffer aloss to its individual fitness to increase the fitness level of the pair or group (Hamilton, 1964).This theory suggests kin groups will perform better than non-kin groups (File et al. 2011). Nichepartitioning theory (or the elbow-room model) describes pairs or groups of individuals that aredifferent genetically and occupy slightly different ecological niches (space and resources), thus14reducing direct competition (Young, 1981). This theory suggests that kin groups will beoutperformed by non-kin groups (File et al. 2011). A 2011 review of kin studies in plants foundevidence of kin selection in nine studies, evidence of niche partitioning in 11 studies and noevidence of kin recognition in 21 studies (File et al. 2011). The scenarios under which kinrecognition could take place are still uncertain. Stress gradients and population densities maycontribute to whether individuals put resources toward the recognition of neighbouring plantsand responses to them (Platt and Bever, 2009). The mechanisms of recognition are also stilllargely unknown, although there is evidence that it is a below ground process involving signalsfound in root exudates (Biedrzycki et al. 2012).We propose that mycorrhizal networks facilitate the recognition of kin within a connectedcommunity. Mycorrhizal networks form from a fungal-root symbiosis (literally ?myco?-?rhiza?),where extra-matrical hyphae of a single fungus connect two or more plants of the same ordifferent species (Francis and Read 1984, Perry et al. 1989, van der Heijden and Horton 2009).Mycorrhizal networks comprised of the well-known long-distance exploring ectomycorrhizalfungi, species in the Rhizopogon vinicolor/vesiculosis complex, are known to intra-specificallyconnect most trees in Douglas-fir forests (Kretzer et al. 2003, Teste et al. 2009, Beiler et al.2010). Rhizopogon networks are known to form differentiated rhizomorphs (dense groupings ofhyphae) that allow for efficient transport of water and nutrient, such as carbon and nitrogen,particularly along source-sink gradients (Perry et al. 1989, Molina et al. 1999, Agerer 2001,Teste et al. 2010, Bingham and Simard 2011). Resource transfer has been loosely associatedwith increased regenerative capacity of interior Douglas-fir forests (Teste et al. 2009, Binghamand Simard 2011). Other types of chemical transfer have also been observed in mycorrhizalnetworks. Song et al. (2010) showed that tomato plants use mycorrhizal networks to detect up-15regulated defense enzymes in blight-infected neighbouring plants, and up-regulate their owngenes to constitutively produce defense enzymes to prepare for a potential attack. These otherforms of transfer have led us to examine whether kin recognition results from chemicalsignalling through networks as well.Chemical signalling is a type of plant interaction that acts as a form of communication(Bruin and Sabelis, 2001). This has been shown to take place both above and below ground butthis study will only focus on below ground interactions (Shulaev et al. 1997, Kessler et al., 2006Song et al. 2010). Some scientists say that chemical signals are used as a means ofcommunication by plants to prime mycorrhizal inoculum to germinate or grow and form asymbiotic relationship (Keirs et al. 2011), to warn neighbours of an imposing threat (Song et al.2010), to detect the relatedness of a close neighbour (Dudley and File, 2007) or even inhibit aneighbour?s growth using allelochemicals (Barto et al. 2011). Biedrzycki et al (2010) showedthat the signalling involved in kin recognition in Arabidopsis thaliana happens through the rootexudates. We wanted to investigate if these root exudates travel through mycorrhizal networksbetween interior Doulas-fir seedlings. Resources such as water, carbon and nitrogen(Schoonmaker et al. 2007, Teste et al. 2010, Bingham and Simard 2011) as well as defensesignals (Song and Simard, unpublished data) have been shown to transfer between interiorDouglas-fir trees through mycorrhizal networks, leading us to test whether kin recognitionsignals also occur along this pathway. If kin recognition signalling and resource transfer isoccurring through these pathways, the extent of the colonization of network-forming fungi wouldlikely have an effect on signalling and transfer. File et al. (2012) showed that in Ambrosiaartemisiifolia L. (common ragweed), the greater the colonization of roots, the greater the growthof each sibling connected by the network.16Research into these areas could inform forest regeneration practices. Managingecosystems in a way that leaves ecological legacies has been proposed as a means of supportinga more natural regeneration process (Keeton and Franklin, 2005). With this research,identification of ideal ecological legacies to assist forest recovery may be improved. Large hubtrees that are highly connected to younger trees through mycorrhizal networks could be used tomaintain the below ground fungal community (Beiler et al. 2010). An array of high seedproducing, healthy, mature hub trees could also support the germination, survival and growth oftheir kin, resulting in a naturally regenerated forest community comprised of well adapted,supported and healthy seedlings.  If these seedlings can be supported by surrounding maturetrees, they may also be better able to deal with changing climate and more extreme conditionssuch as drought (Bingham and Simard, 2011).The objective of this study was to determine whether kin recognition occurs betweeninterior Douglas-fir seedlings, whether it is mediated by mycorrhizal networks, and whether itvaries among families. We examined kin recognition by comparing performance of full siblingand unrelated pairs using control-pollinated interior Douglas-fir seeds. Our first hypothesis wasthat the high level of relatedness between sibling pairs would result in shifts in plant behaviourcompared to non-sibling pairs. We expected enhanced performance of donor and receiver plantsin kin pairs compared to non-kin pairs. Our second hypothesis was that kin recognition would bemediated through mycorrhizal networks. We expected greater kin recognition between donor andreceiver plants connected in a mycorrhizal compared to those that were isolated from each other.We also expected the extent of the mycorrhizal network (amount of colonization) to enhance thekin recognition and network effects. Our third hypothesis was that kin recognition would varygenetically, i.e., among families. We expected kin recognition to vary according to the17competitiveness (e.g., shoot or root growth rates) of plant genotypes, where families with fastergrowth rates are better able to recognize kin. Finally, we expected kin recognition to be mediatedthrough resource transfer between seedlings in a pair, with preferential transfer from parent treesto kin seedlings compared to non-kin seedlings. We expected increased transfer between kinseedlings to be associated with enhanced kin performance.MethodsExperimental design and treatmentsThis experiment took place in the University of British Columbia greenhouse inVancouver, British Columbia, Canada over an 11 month period (March 2012 to February 2013).A total of 100 pots, approximately 2 L, height = 18 cm, diameter = 16.5 cm, were distributeduniformly over half of one greenhouse bench. The pots were re-randomized every two weeks. Nosupplementary light was provided.A 2 x 2 factorial design was used where mesh size (two levels) and relationship (twolevels) were applied in a completely randomized design. The mesh size factor (two levels)references the pore size of the specialized mesh bags (approximately 20cm x 8 cm) made byPlastok? (Meshes and Filtration) Ltd. (Birkenhead, UK) that separated the root systems of theolder, ?donor? seedling and the younger, ?recipient? seedling. A total of 50 mesh bags had a poresize of 35 ?m, which has previously been shown to allow fungal hyphae to penetrate the bag butprevent roots from growing through the barrier (Teste and Simard 2008). The other 50 mesh bagshad a pore size of 0.5 ?m, which has been shown to prevent fungal hyphae from penetrating andallows only soil water to flow in and out of the bag (Teste et al. 2006). The relationship factor(two levels) references the genetic closeness of the seedling pair grown in a single pot. A total of1840 pots were planted with a kin pairing, where both seedlings originated from the same family(designated 1/1, 2/2, etc.) (Table 2.1). The remaining 60 pots were planted with a non-kin pairing,where the seedlings came from two different families (1/2, 4/3, etc.) (Table 2.1). The seeds wereproduced by the interior spruce breeding program of the Ministry of Forests, Lands and NaturalResources Operations at the Kalamalka Research Station near Vernon, British Columbia. 1240control cross-pollinated interior Douglas-fir seeds were used from four pairs of known parents(hereafter referred to as ?families? 1 through 4). Family 1 is a control cross pollinatedcombination of Fdi SA (Salmon Arm, British Columbia) 8069 x SA 8033, family 2 - Fdi SA8041 x 8016, family 3 - Fdi SA 8047 x SA8037 and family 4 - Fdi SA 8211 x SA 8006 (whereFdi=interior Douglas-fir and SA=Salmon Arm). All of the seedlings that emerged from seeds inthe same family were full siblings. The pairings and mesh sizes were combined in a way thatresulted in 20 kin pairs with 35 ?m mesh, 20 kin pairs with 0.5 ?m mesh, 30 non-kin pairs with35 ?m mesh and 30 non-kin pairs with 0.5 ?m mesh. The families were equally distributedamong all combinations.Experimental setupEach pot contained a mesh bag that was placed against the edge and gently packed with a3:1 mixture of greenhouse standard potting mix: field collected soil (Figure 2.1a). This mixturewas thoroughly blended in a mechanical soil mixer to achieve high homogeneity, and autoclavedfor one to 1.5 hours at 250?C to kill any fungal inoculum in the field collected soil. Autoclavingwas applied to the soil mixture inside the mesh bags to ensure that mycorrhization of recipientseedlings occurred primarily through contact with mycorrhizal networks of donor seedlings. Theremainder of the pot was filled with a 1:1 combination of non-autoclaved field collected soil andgreenhouse standard potting mix, also thoroughly blended. The soil was collected from an19interior Douglas-fir forest near Princeton, British Columbia (approximately 120.58?W,49.43?N)(Figure 3.1). The forest occurred in the Dry Cool Interior Douglas-fir subzone (IDFdk),was comprised of pure interior Douglas-fir, and was underlain by a Dystric Brunisol soil with asandy loam texture and a moder humus form (Canadian System of Soil Classification 1998). Thetop litter layer was scraped off and the underlying 10 to 15 cm (including the fermentation layer,humus layer and mineral soil) was collected and transported immediately to the UBC greenhousein Vancouver, BC. This soil was expected to contain a sufficient amount of interior Doulas-fircompatible fungal inoculum to encourage mycorrhizal colonization within the pot.Each pot was designed to contain a pair of seedlings (Figure 2.1b). One older seedlingwas established 8 months in advance, outside of the bag, to act as a donor to the younger,recipient seedling. The pairs were subject to the four treatment groups, described above,depending on the relationship of the pair of seeds sown (full siblings or from separate families)and the pore size of the mesh bag installed (0.5 ?m or 35 ?m).At the onset of the experiment, all pots were soaked to field capacity with water. A totalof 125 seeds from each of the four families described above were sown in 25 pots (100 pots intotal), with five seeds per pot. Before they were sown, seeds were sterilized with 10% H2O2 for10 minutes and then allowed to dry. An additional two seeds per pot were stratified and sown inthe mixed field soil after six weeks to make up for unsatisfactory germination rates. Thisinvolved a 24 hour soak in distilled water followed by drying and refrigeration (4?C) for threeweeks. Some seedlings were transplanted from pots with multiple germinants to those yet togerminate with the goal of having at least one seedling per pot. They were all sown in the mixedfield soil containing the inoculum. A thin layer of fine gravel known as ?forest sand? was spreadover all sown seeds to discourage the growth of damping off fungi.20Pots were lightly watered each day until each pot had at least one healthy seedling. Therewas at least one seedling per pot 10 weeks after the first seeds were sown, at which point the potswere thinned to one healthy seedling as close to the center of the pot as possible. No fertilizerwas provided at any point. The watering regime was shifted from a light watering daily toweekly watering to field capacity. The limited watering and absence of fertilization weredesigned to encourage mycorrhizal colonization. The seedlings were allowed to grow for anothertwo months with periodic thinning when necessary.When the ?donor? seedlings were 4.5 months old, they were transported into a ConvironPGV36 Plant Growth chamber at UBC to undergo a blackout regime in preparation for anartificial winter. The blackout regime consisted of a 10 day period with 10 hour days at 20?C and14 hour nights at 15?C. After the blackout period, the seedlings were returned to the greenhousefor three weeks to prepare for winter, and then returned to the growth chamber for the artificialwinter. The seedlings experienced 16 hour uninterrupted nights, 8 hour days at low light levelsand a temperature of 4?C for six weeks during the artificial winter provided by the growthchamber. The seedlings were then returned to the greenhouse. The soil and seedlings wereallowed 24 hours to adjust to the temperature difference, and the second set of ?recipient?seedlings were sown the next day. A total of five seeds from the corresponding kin or non-kinpairing were sown within the inside edges of the mesh bag. The seeds were from the same stockas the seedlings currently established and were stratified in the same manner.The seeds were sown in accordance with the family they were from and the family of thealready established seedling. After the second set of seeds was sown, they were allowed togerminate, grow and were thinned when necessary for the next four months. Heights wererecorded for both seedlings every three weeks.2113CO2 labellingThe 13CO2 labelling was conducted 8.5 months after the experiment was established. Atotal of 90 of the original 100 pots remained with a pair of healthy seedlings at the onset of the13CO2 labelling period (18 kin/network (35 ?m mesh), 16 kin/no network (0.5 ?m mesh), 29 non-kin/network and 28 non-kin/no network). Those 90 pots were split into three labelling treatments(1-day chase, 6-day chase and control). We used 37 pots (9 kin/network, 8 kin/no network and10 each of the non-kin pots) for a 10-hour labelling period followed by a 6-day chase periodbefore the seedlings were harvested. We used an additional 17 pots (9 non-kin/network and 8non-kin/no network) for an identical 10 hour labelling period followed by a 1-day chase periodbefore the seedlings were harvested. For cost purposes only the 6-day chase labelled seedlingswere used for further analyses in this study and the 1-day chase pots were harvested for RNAanalysis for another study being run in parallel. The remaining 36 pots were used as controls.During the labelling period, the older, donor seedling was sealed inside an inflated plasticFoodsaver? bag (approximately 11?? x 16??) (Figure 2.1c). The bag was transparent on one sideand translucent on the other side, so the pots were positioned in way that maximal light wouldenter the transparent side. The bags were pre-sealed on both sides and open on the ends. Wesealed one end using a Seal-A-Meal? food saver, melting the two layers together. After the bagwas placed over the donor seedling, the remainder of the bag was sealed along the end andaround the base of the seedling with Tuck? Contractors Sheathing Tape. Both the labeled andcontrol seedlings were sealed in the bags and inflated with ambient air. In addition, the labeledseedlings? bags received three injections, at equal time intervals, of 13C labeled CO2 (99% 13C) (<1% 18O) totaling 50 ml (Cambridge Isotope Laboratories Inc. Andver, MA). At the end of the 1022hour period, the bags were removed and the labeled and control pots were separated for a 6-daychase period.MeasurementsThe growth variables examined in both the donor and recipient seedlings were needle,stem and root weight (g), total and above ground biomass (g), average weight per colonized roottip (g per root tip), percent of total root weight colonized (%), stem length (cm), germination rate(germinants per seeds sown), average growth rate (cm per week), total leaf area (cm2) andvolume, calculated as total leaf area x stem length (cm3). Average leaf area per needle (cm2) wasexamined only in recipient seedlings. The seedlings were removed from the pots after theyounger seedlings had grown for four months. The below and above ground portions wereseparated. The soil was carefully removed from around the roots, which were then washed withtap water. The needles were removed and measured using a LICOR-3100 leaf area meter. Thenumber of needles was also recorded for the younger, recipient seedlings only. The needles weredried and biomass was recorded. The stems were measured for length, number of branches anddry biomass.The roots were examined for mycorrhizal colonization and the root tips that appeared tobe colonized were weighed and visually morphotyped. Sanger sequencing of fungal DNA wasperformed on a subset of tips at the Irving K. Barber School at UBC-Okanagan. A total of 65 tipswere sequenced with at least five tips from each visually morphotyped species and five root tipsidentified as non-mycorrhizal. Fungal DNA was extracted using the following protocol: 25 ?lE7526 SIGMA Extraction Solution (Sigma-Aldrich, ON, Canada) added per individual ECMroot tip, incubated at 95 ?C for 10 minutes then cooled to 4 ?C, 25 ?l D5688 Dilution Solution(Sigma-Aldrich) added prior to freezing at -20 ?C.  The internal transcribed spacer region (ITS)23was amplified for each DNA sample using ITS1 (White et al., 1990) and ITS4/ITS4B primers(Gardes & Bruns, 1993). Amplifications were performed on a Veriti 96-Well Thermal Cycler(Applied Biosystems, ON, Canada) in 12.5 ?l volumes containing: Nuclease-free H2O, 1.25 ?lof 25 mM MgCl2, 2.5 ?l GoTaq? Reaction Mix (Promega, WI, USA), 0.25 ?l of 10mM dNTPs,0.25 ?l of 10 mM each primer, 0.65 units GoTaq? DNA Polymerase (Promega), and 100 ngDNA.  Thermocycling conditions: 3 minutes denaturation (94 ?C), 35 cycles of denaturing,annealing and extension (94 ?C for 35 s, 51 ?C for 35 s, and 72 ?C for 50 s respectively), 10minutes final extension (72 oC) before cooling to 4 ?C. Purification was performed using theUSB ExoSAP-IT? PCR Product Clean-Up kit (Affymetrix, CA, USA) then sequenced usingABI BigDye v3.1 Terminator chemistry and an ABI 3130xl Genetic Analyser (AppliedBiosystems). Raw sequence data were analysed using the SEQUENCHER software packageVersion 4.7 (Gene Codes Corp., MI, USA) before comparison with NCBI and UNITE(Abarenkov et al., 2010) databases using the BLAST algorithm.  Names were assigned tomorphotypes based on the combination of morphological characteristics and minimum 97%sequence matches corresponding to the indicated species.  These sequence data are in the processof being submitted to the GenBank database. We identified Rhizopogon by comparing thesequences to NCBI and UNITE databases, only using the Kretzer entries to determine speciesnames. The best matching sequence was a 99% match to Rhizopogon vinicolor (accessionnumber: AF263933) [Kretzer ID].A suite of foliar nutrients were examined using microwave digestion/ICP (InductivelyCoupled Plasma-Optical Emission Spectrometer) and % C, N and S analyses at the BritishColumbia Ministry of Environment Analytical Chemistry Laboratory, Victoria, BC. Themacronutrients examined were C, N, K, CA, Mg, P and S. The micronutrients examined were Al,24B, Cu, Fe, Mn, Mo, Na and Zn. The extrapolated gram content per seedling and average sampleconcentrations were used in our analysis. C:N and N:P ratios were calculated and examined.The amount of 13C transferred from the donor seedlings to the recipient seedlings? rootsand stems was determined by EA-IRMS (elemental analysis-isotope ratio mass spectrometry) byUBC?s (Vancouver) stable isotope facility and examined in terms of excess 12C equivalent. Thiswas calculated according to the modified Button (1991) procedure described by Teste et al.(2009). Three subsamples each of mixed field soil and autoclaved soil mixture were analyzed fortotal C, N and S, pH, available P and mineral N. Exchangeable cations (Al, Ca, Fe, K, Mg, Mnand Na) and effective cation exchange capacity were also measured (0.1 M barium chloride).Data analysisAll statistical analyses were run using SAS version 9.3 (Cary, North Carolina). Two-wayanalysis of variance (ANOVA) using PROC GLM was run separately for each variable.  Thefactors included were relationship, network and relationship x network. An analysis ofcovariance (ANCOVA) was also run using PROC GLM to examine the effect of family as acovariate. Planned contrasts were carried out to determine the strength of the extreme treatmentsalong the relationship-network gradient: (1) kin/network versus all other treatments, and (2) no-kin/no network versus all other treatments.Results13C transferThere was a difference in the amount of 13C transferred to the recipient seedling stembased on the pore size of the mesh used to separate the seedling root systems. The seedlingsgrown in mesh bags of the larger pore size (35 ?m) allowing fungal hyphae to connect the root25systems received a significantly greater amount of 13C transfer than seedlings grown in meshbags with smaller pore size (0.5 ?m) preventing hyphal connections (p=0.0290, Table 2.2,Figure2.2). The genetic relationship between the seedlings in a pair had no effect on in theamount of 13C transferred to the recipient seedling stem. The interaction effect of relationship xnetwork was not significant for the amount of 13C transferred.  However, pairwise comparisonsshowed that differences in 13C transfer between mesh sizes was greater and of higher statisticalsignificance for non-kin than kin pairs (Figure 2.2).GrowthRelationship and network main effectsThere were no significant main effects or strong trends for any of the growth variablesdue to either the relationship factor or the network factor (p>0.05).Relationship x network effectsThere were significant or strong trends in relationship x network interactions for allgrowth variables measured in the recipient seedlings (Table 2.3). Needle, stem and root weight,total and above ground biomass, total leaf area, stem length, average growth rate and volume(total leaf area x stem length) all showed significant relationship x network interaction effects(p<0.05) (Figures 2.3, 2.4). There was also a strong interaction trend for average weight per roottip and average leaf area per needle in recipient seedlings (p<0.1) (Table 2.3). The kin/networktreatment (kin relationship between the seedling pair where the recipient was grown in a meshbag allowing for a network formation, 35 ?m) and the non-kin/no network (non-kin with 0.5 ?mmesh) unexpectedly had consistently lower least square (LS) mean value than the other two26treatments (kin/no network and non-kin/network)(Figure2.3b-f). The only exception was averageweight per root tip, where the recipient seedling in the kin/network treatment had a higher LSmean value (2.963 mg) compared to the other three treatments combined (2.078 mg)(p=0.0252)(Figure2.3a )For donor seedlings, there was a significant relationship x network effect on heightgrowth rate (p = 0.0393) (Table 2.4) with the same pattern among treatments as for the recipientseedlings (Figure 2.4). The planned contrasts showed there was a difference in donor stem lengthwhen each extreme treatment ((1) kin/network or (2) non-kin/no network) was compared to theother three treatments combined. Planned contrast 1 showed that kin/network donor seedlingshad shorter stem lengths (24.22 cm) than the other three treatments combined (27.07 cm) (p =0.0406).Alternatively, donor seedling in the non-kin/no network treatment (planned contrast 2)tended to have longer stems. This effect was coupled with shorter stems of recipient seedlings inthe same treatment (non-kin/no network) compared to the three other treatments combined. Inplanned contrast (2), the donor seedlings also had significantly greater root weight in the non-kin/no network treatment (p = 0.0450). Recipient seedlings, by contrast, tended to have lowerroot weight, stem weight, and total biomass in the non-kin/no network treatment (p< 0.10).Average growth rate did not follow this trend in the donor seedlings; however, the trendremained when considering just the recipient seedlings.Family effectsRelationship and network effects were tested for all variables with ANCOVA using thefamily of the recipient seedling as a covariate. These analyses revealed that family had no27significant effect on any of the variables tested. However,  the small sample size (less than 25units per family) this study may have been provided insufficient power to detect effects.NutrientsRelationship effectsThere was greater Al (p<0.0001) and Fe (p=0.0850) content in foliage of donor seedlingsof pairs that had a kin relationship compared to those that had a non-kin relationship (Table 2.5).Fe and Mo were significantly higher in recipient seedlings of kin than non-kin pairs (Table 2.2).Relationship had no other effect on the remaining nutrient variables.Network effectsThere was greater foliar Al content in donor seedlings of pairs that had networkformation capabilities (35 ?m mesh) compared to those that did not (0.05 ?m mesh)(Table 2.5).Networking capability had no other effect on the remaining nutrient variables.Relationship x network effectsThere were significant relationship x network interaction effects on C, K, N, S and Zncontent of recipient seedlings (p < 0.05). There were also strong tendencies for interaction effectson B, Ca, Mg and P (p < 0.1) (Table 2.2) (Figure 2.5). In the donor seedlings, there was arelationship x network effect on foliar Al (p=0.0375), Fe (p=0.0680) and C:N ratio (p=0.0317)(Table 2.5).Donor and recipient foliar nutrient concentrations28Donor seedlings had higher concentrations of S, P, Mg, Ca than recipient seedlings butrecipients had higher concentrations of Al, B, Cu, Mo, N and Zn (p<0.05). Donors and recipientshad the same concentrations of Fe, K and Mn. N content, as well as content of all other macro-and micronutrients, were considerably higher in the larger donor seedlings than recipientseedlings (p<0.0001).SoilsThe autoclaved soil had higher percent total C and S and available P but lower pH thannon-autoclaved mixed field soil. There was no difference in percent total N or mineral N.Autoclaved soil had greater exchangeable cations (CMol+/Kg), except Mn, which was higher inthe mixed field soil and Fe, which did not differ between autoclaved and non-autoclaved soils.The effective cation exchange capacity was also higher in the autoclaved soil (Table 2.6).DiscussionKin effectsThere was evidence of kin recognition in some of the micronutrient variables. Fe(recipient and donor), Mo (recipient) and Al (donor) were significantly different in the kinseedlings compared to the non-kin seedlings, suggesting that some recognition caused thediscrepancy. Interestingly, in all cases where there were significant differences, kin seedlings hadgreater micronutrient content than non-kin seedlings. Our hypothesis that kin recognition wouldlead to kin selection is thus supported by the nutrient results. We expected to see that same resultin the growth variables, suggesting that kin recognition leads to enhanced kin growthperformance. Our results, however, do not support that hypothesis. There were no significant29main kin effects on any of the growth variables. It is difficult to link beneficial kin effects todirect fitness gains in the recipient but instead these benefits have been thought to result frommore indirect characteristics such as altered morphology, allocation tendencies or the reductionof resource uptake by donors (File et al., 2011). The favourable growth conditions of thegreenhouse used in this experiment may have led to kin recognition that was only detectable insubtle ways, as we observed with Fe, Mo and Al. Fe is important for photosynthesis and as a co-factor in many enzymes, Mo is important in the nitrate reduction reaction, and Al is the mostabundant metallic element in the soil and is thought to be important in biochemical pathwaysinvolved in signal transduction. Al is known to have toxic effects on plants at certainconcentrations but those concentrations were not reached in this study (Guerinot and Li, 1994,Mulder et al., 1959, Soon, 1995). The benefits of Al on plant physiology have not been widelystudied and therefore, they are largely unknown.The differences in ?donor and recipient foliar nutrient concentrations? may have resultedfrom the reduction in pH following autoclaving the soil. The autoclaved soil, present only in themesh bags surrounding the recipient seedlings, had a significantly lower pH than the surroundingsoil at the onset of the experiment. B, Cu and Zn (higher concentrations in recipient seedlings) aswell as Fe and Mn (no difference) are more available to plants in mineral soils as pH decreases.More importantly, the content of all nutrients was greater in donor seedlings due to the largediscrepancy in size between the donors and recipients. This discrepancy represents a large carbonand nutritional source-sink gradient between donors and receivers that may drive interplanttransfers through mycorrhizal networks (Simard et al. 2002).30Mycorrhizal network effectsThe DNA sequencing performed on the colonized root tips showed that Rhizopogonvinicolor was the most common fungal association colonizing 95.6% of donor roots and 83.3%of recipient roots. This result was as expected; Rhizopogon vinicolor is a strong networkingfungal species and known to associate closely with Douglas-fir seedlings during forestdevelopment (Massicotte et al. 1994, Molina et al. 1999, Tweig et al. 2007, Beiler et al. 2012).Other taxa included Pyronemataceae (sp.) and ascomycete endophytes.  At least oneectomycorrhizal fungal species was shared between the donor and recipient in all networkedtreatments where both seedlings were colonized (only 2.2% were not colonized). This providessufficient evidence that the seedlings in a pair had the potential to form mycorrhizal networks.Transfer of carbon provided more definitive evidence for the presence of functionalmycorrhizal networks in the network treatments. There was significantly greater excess 12Cequivalent in recipient seedlings that were grown in mesh bags with the 35 ?m pore size, wherefungal hyphae could penetrate through the pores. These results agree with Teste et al. (2006),who tested the effect of mesh pore size on mycorrhizal network formation. The ability to detect13C in the recipient seedling also suggests the potential for transfer of other compounds throughnetworks. Nevertheless, there were no differences in any of the foliar macronutrients, in eitherdonor or recipient seedling, due to a main network effect. It is possible that the carbon and othernutrients are transferred from the donor seedling to networked recipient seedlings but that thenutrients are stored in the seedling stem rather than the roots (tested in excess 12C equivalentanalysis) and needles (tested in the foliar nutrient analysis). Teste et al. (2010) also found thatexcess 12C equivalent measured in the shoots of Douglas-fir seedlings was higher than that31detected in the roots. We also expect that detection of nutrient transfer through comparison offoliar nutrients would be obscured by larger variation and quantities than 13C.Our hypothesis that formation of networks between donor and recipient seedlings wouldlead to enhanced growth of recipients was not supported by our data. There were no differencesin any growth variables due to a network connection. Several studies have shown an increase insurvival, total biomass, root biomass, above ground biomass or height in autotrophic,ectomycorrhizal plants with an opportunity to form networks with other plants (Onguene andKuyper 2002, Dickie et al 2005, Booth and Hoeksema 2010, Nara 2006, McGuire 2007, Teste etal. 2009, Bingham and Simard 2011). It is possible that some of these same effects may havebecome apparent in the recipient seedlings had they been allowed to grow for a longer period oftime. Some studies have shown network facilitation to occur most strongly or only whereseedlings were growing under environmental stress.  For example, Bingham and Simard (2011)found improved growth of networked seedlings only under drought stress, McGuire (2007) andOnguene and Kuyper (2002) found increased performance under deep canopy shade, and Narafound increased survival in nutrient poor volcanic soils. Neither donor nor recipient seedlings inour study were water, light or nutrient stressed and therefore may have been less likely to benefitfrom carbon or nutrient transfer or enhanced nutrient or water uptake through mycorrhizalnetworks.Relationship x network effectsDespite few main effects, there was evidence for a significant interaction betweennetwork and relationship in colonization, growth and nutrient variables.  The interaction patternswere also evident in planned contrasts.  Recipient seedling colonization (measured as average32weight per colonized root tip) was greater among networked kin than any other treatment,supporting our hypothesis that mycorrhizal networks would facilitate mycorrhizal colonization ofkin. Enhanced colonization by mycorrhizal fungi in kin pairings has also been observed inAmbrosia artemisiifolia L. (common ragweed) seedlings (File et al. 2012). However, ourhypothesis that mycorrhizal networks would result in enhanced growth performance of recipientkin was not supported. Indeed, we found the opposite. Both the kin/network and non-kin/nonetwork recipient seedlings had the lowest growth rates. This is an interesting result particularlywhen the donor growth traits are examined in parallel. The kin/network donor seedlings tendedto have lower and slower growth (average growth rate and stem length) as observed in therecipients, however, the non-kin/no network donors tended toward to have higher growth values(root weight). This suggests that some recognition of kin is occurring between the kin seedlingpairs. However, this appears not to directly enhance recipient performance, but rather to reducethe competitive environment of both kin seedlings (kin cooperation). It is possible that the non-kin/no network donor seedlings, having recognized a stranger seedling in its neighbourhood,only detects a competitor. The donor seedling, having been established earlier, continues tooutcompete its unrelated competitor. Dudley and File (2007) identified this same aspect of kinrecognition and selection in the root allocation of Cakile edentula var. lacustris (Brassicaceae), aself-fertilizing annual plant. They found that allocation to fine root mass between stranger (non-kin) pairs did not differ from root allocation of an individual grown alone, but allocation to fineroot mass in kin pairs was lower providing a less competitive environment amongst kin.33Family effectsWe expected to see genetic differences in both growth and nutrient variables dependingon the family in which the seedling came from. That differentiation was also expected to affectthe seedling pairs? ability to recognized and cooperate with kin neighbours (Donohue, 2003).Our analysis did not support this hypothesis; there were no differences in the growth or nutrientvariables according to the family the seed came from. All seeds were obtained from a seedorchard and received relatively identical treatment before the onset of the experiment, therefore,these seeds may have already been too similar in competitive ability to detect differences amongfamilies. Our sample size for families, however, may have been too small to detect family effectsin this experiment.ConclusionsWe found subtle evidence of kin recognition expressed in mycorrhizal colonization, foliarmicronutrient and growth traits of donor and recipient seedlings.  However, our hypothesis thatkin recognition would result in greater growth of recipient kin seedlings (i.e., kin selection) wasnot supported.  Instead, kin recognition was evident in reduced growth of both donors andrecipient seedlings.  It appears that seedlings may have reduced their competitive environmentwhen in the neighbourhood of kin (i.e., increasing elbow room), but donors were increasing theircompetitive ability when in the neighbourhood of non-kin recipients. We found that mycorrhizalnetworks formed between donor and recipient seedlings, as shown in shared ectomycorrhizalfungal species and 13C transfer.  We also found that networks facilitated mycorrhizalcolonization of kin but not non-kin seedlings. However, we did not find any other evidence thatkin selection was facilitated by mycorrhizal networks. A competitive advantage was not gained34by any group of kin (family) nor did the high level of genetic relatedness (full sibling pairs)enhance kin recognition. Stronger relationships may have been apparent if we had grown therecipient seedlings for longer or under environmental stress.While no significant differences translated directly to growth, the enhancement ofmirconutrients in kin seedlings may have an effect on other processes involved in seedling health,although that was not a conclusion of our study. It is possible that forest regeneration of interiorDouglas-fir forests could improve with management practices that encourage reproduction of kinseedlings near their parents. This could include the retention of healthy, cone-bearing legacytrees to supply seed for natural regeneration.  Harvesting practices could be timed to coincidewith mast seed years and sites could be prepared in the neighbourhood of legacy trees to providesuitable mixed mineral seedbeds and minimal competition from native grasses (Simard et al.2003).  These practices would also help conserve mycorrhizal networks, which have been shownin other studies to benefit natural and planted regeneration (Horton and Bruns 2001, Teste et al.2009, Bingham and Simard 2011). Retention of large legacy trees would create strong source-sink gradients with neighbouring seedlings, which could result in greater carbon transfer. Thiscould also create the potential for greater kin selection than we observed between our months-oldseedlings. These practices may be particularly effective on sites with low productivity either dueto micronutrient deficiencies or drought.35Table 2.1. Family pairings to create kin and non-kin treatments in full sibling greenhouseexperiment. All families are control cross pollinated seeds from known interior Douglas-firparents, both different for each family. Family 1 - Fdi SA 8069 x SA 8033 (1), family 2 - Fdi SA8041 x 8016 (2), family 3 - Fdi SA 8047 x SA8037 (3) and family 4 - Fdi SA 8211 x SA 8006(4). Each family is paired with itself (kin), seen in the top row, and each of the other threefamilies both as donors (first number in pair) and recipients (second number in pair) (non-kin),seen in the bottom three rows.1/1 2/2 3/3 4/41/2 2/3 3/4 4/31/3 2/4 3/2 4/21/4 2/1 3/1 4/136Table 2.2. Analysis of variance for network, relationship and interaction (network x relationship)effects on foliar nutrient content and 13C transfer in recipient seedlings. * p < 0.05.Network Relationship Network x RelationshipF P F P F PAl (g) 2.04 0.1571 1.77 0.184 0.03 0.8659B (g) 0.23 0.6312 0.83 0.3645 3.47 0.0661C (g) 0 0.9988 0.5 0.8186 5.15 0.0259*Ca (g) 0.11 0.7418 0.01 0.9294 3.45 0.0669Cu (g) 0.36 0.5477 0.05 0.8288 0 0.977Fe (g) 1.86 0.1767 4.68 0.0333* 2.37 0.1271K (g) 0.11 0.742 0.02 0.8864 5.28 0.0241*Mg (g) 0.3 0.5846 0.05 0.8244 3.44 0.0617Mn (g) 0.16 0.6878 0.58 0.4492 1.13 0.2915Mo (g) 0.02 0.9012 10.36 0.0019* 0.66 0.4202N (g) 0.03 0.872 0.12 0.7353 5.47 0.0216*Na (g) 0.54 0.4647 0.73 0.3963 0.53 0.4705P (g) 0.04 0.8464 0.11 0.7461 2.87 0.0939S (g) 0.06 0.8071 0.02 0.885 5.3 0.0238*Zn (g) 0.11 0.741 0.08 0.7734 5.39 0.0227*C:N 0.03 0.8553 0.14 0.713 0.04 0.5271N:P 1.99 0.1623 0.74 0.3917 1.51 0.222413C transferred(excess 12Cequivalent instem) (g) 5.76 .0290** 0.03 0.8718 0.82 0.374237Table 2.3. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in recipient seedlings. * p < 0.05.Network RelationshipNetwork xRelationshipF P F P F PNeedle weight (g) 0.6 0.4405 0.02 0.8805 4.84 0.0304*Stem weight (g) 1.02 0.3165 0.38 0.5407 7.19 0.0088*Root weight (g) 0.05 0.8201 0.27 0.6074 6.06 0.0158*Total biomass (g) 0.5 0.481 0.01 0.9102 6.47 0.0127*Above ground biomass (g) 0.72 0.3998 0.07 0.7914 5.57 0.0205*Total leaf area (cm2) 0.48 0.492 0.01 0.9189 5.33 0.0234*Stem length (cm) 0.45 0.5044 0.34 0.5624 5.83 0.0178*Average weight per colonized root tip (g) 1.36 0.2476 1.01 0.3169 2.98 0.0881Average growth rate (cm/week) 0.04 0.8335 0.29 0.5943 4.72 0.0326*Volume (cm3) -(total leaf area x stem length) 0.59 0.4443 0.19 0.6658 5.04 0.0273*Average leaf area per needle (cm2) 0.02 0.8776 0.06 0.8145 2.81 0.097338Table 2.4. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in donor seedlings. * p < 0.05.Network RelationshipNetwork xRelationshipF P F P F PNeedle weight (g) 0.2 0.6534 0.19 0.6623 1.2 0.2759Stem weight (g) 0.12 0.7283 1.32 0.254 0.17 0.6843Root weight (g) 1.52 0.2204 0.79 0.3779 0.94 0.334Total biomass (g) 0.69 0.4072 0.31 0.5813 0.92 0.3414Above ground biomass (g) 0.19 0.6662 0.05 0.8153 0.74 0.3925Total leaf area (cm2) 0.75 0.3884 0.02 0.8753 0.38 0.5367Stem length (cm) 2.5 0.1176 1.93 0.1686 0.65 0.4219Average weight per root tip (g) 1.09 0.3001 0.23 0.6319 0.44 0.5066Average growth rate (cm/week) 1.03 0.3136 0.66 0.4194 4.38 0.0393*Volume (cm3) ?(total leaf area x stem length) 1.32 0.2533 0.6554 0.2 0.01 0.919339Table 2.5. Analysis of variance for network, relationship and interaction (network x relationship)effects on foliar nutrient content in donor seedlings. * p < 0.05.Network Relationship Network x RelationshipF P F P F PAl (g) 4.62 0.0345** 38.16 <.0001* 4.47 0.0374*B (g) 0.03 0.8538 1.59 0.2102 0.2 0.6558C (g) 0.23 0.6339 0.26 0.6133 1.09 0.2997Ca (g) 0.44 0.5109 0 0.9662 0 0.9829Cu (g) 0.75 0.39 0.15 0.695 2.53 0.1155Fe (g) 0.19 0.6658 3.04 0.0850 3.42 0.068*K (g) 0.08 0.7724 0.22 0.6378 1.63 0.2047Mg (g) 0.16 0.6945 0.04 0.8467 0.01 0.9356Mn (g) 0.08 0.7762 0.33 0.5654 0.14 0.7109Mo (g) 0.47 0.4961 2.26 0.1367 0.04 0.8372N (g) 0 0.9653 0 0.9452 0.05 0.8212Na (g) 0.04 0.837 1.39 0.2409 1.15 0.2856P (g) 0.07 0.7961 0.16 0.6893 0 0.9642S (g) 0.06 0.8049 0.02 0.8853 0.82 0.3684Zn (g) 0.01 0.9049 0.33 0.5654 0.1 0.7511C:N 0.38 0.541 0.69 0.4072 4.77 0.0317*N:P 0 0.9983 0.01 0.9356 0.07 0.798740Table 2.6. Analysis of variance, means ratio, and difference between means for the effect ofautoclaving on soil used in mesh bags accessed by recipients (autoclaved) and pots accessed bydonors (non-autoclaved) on soil nutrients, CEC (cation excahange capacity) and pH. Positivedifference in means values indicate a higher value in autoclaved soil compared to non-autoclavedsoil. * p < 0.05.ANOVA ? Autoclaving soilMeans ratio Difference inmeans F PAl (CMOL+/Kg) 1.21 0.003 4.77 0.0943Ca (CMOL+/Kg) 1.05 2.240 7.11 0.0560Fe (CMOL+/Kg) 1.40 0.001 1.73 0.2587K (CMOL+/Kg) 1.04 0.150 8.01 0.0473*Mg (CMOL+/Kg) 1.38 3.742 236.87 0.0001*Mn (CMOL+/Kg) 0.91 -0.124 21.27 0.0099*Na (CMOL+/Kg) 1.76 0.361 223.6 0.0001*CEC (CMOL+/Kg) 1.10 6.373 32.14 0.0048*C (%) 1.12 2.848 11.66 0.0269*N (%) 0.96 -0.025 3.82 0.1222S (%) 1.52 0.048 9.91 0.0346*Available P - PO4/P (mg/Kg) 1.26 130.446 64.08 0.0013*Mineral N (mg/Kg) 0.56 -31.469 3.58 0.1315pH (1:1 H2O) 0.93 -0.413 549.14 <0.0001*pH (1:2 CaCl2) 0.94 -0.350 126.72 0.0004*41a                                  b                                     cFigure 2.1. Photos of (a) a top view of an experimental unit (pot) just after the donor seedlingwas sown, (b) experimental unit at the onset of the 13CO2 labelling period, recipient age 4 months,with donor age 10 months and (c) fitting a donor seedling with a 13CO2 labelling bag.42Figure 2.2. 13C transfer to recipient seedlings measured in excess 12C under the four network andrelationship combinations as well as all networked compared to all no network recipientseedlings. Different letters indicate bars that are significantly different at p > 0.05. Error bars areone standard error above and below.-0.000200.00020.00040.00060.00080.0010.00120.0014Network No networkExcess12C equivalent KinNon-kinKin and non-kincombinedAAB BABAB43a bc                                                                      de                                                                      fFigure 2.3. Least square means of recipient seedling (a) average weight per colonized root tip,(b) above ground biomass, (c) total biomass, (d) root weight, (e) stem length and (f) stem weightunder the four relationship and network combinations (interaction effects). Different lettersindicate bars that are significantly different at p > 0.05. Error bars are one standard error aboveand below.00.00050.0010.00150.0020.00250.0030.00350.004Average weight per roottip (g/tip) ABB ABA00.020.040.060.080.10.120.14Above ground biomass (g)ABAABB00.050.10.150.20.25Total Biomass(g)AB BAAB00.010.020.030.040.050.060.070.080.090.1Root weight (g) ABABAB012345678910Stem length(cm)BAABA00.0050.010.0150.020.0250.030.035Stem weight (g)ABABA44Figure 2.4. . Least square means of height growth rate in donor seedlings under the fourrelationship and network combinations (interaction effects). Different letters indicate bars thatare significantly different at p > 0.05. Error bars are one standard error above and below.00.10.20.30.40.50.6kin/network kin/no network non-kin/network non-kin/no networkGrowth rate (cm/week)AB ABAB45a bc                                                                            dFigure 2.5. Least square means of recipient seedling (a) iron (Fe), (b) molybdenum (Mo) , (c)magnesium (Mg), and (d) sulphur (S) content in grams under the four relationship and networkcombinations (interaction effects). Different letters indicate bars that are significantly different atp > 0.05. Error bars are one standard error above and below.0.00E+001.00E-062.00E-063.00E-064.00E-065.00E-066.00E-067.00E-068.00E-069.00E-061.00E-05Fe (g)AB BAB0.00E+002.00E-084.00E-086.00E-088.00E-081.00E-071.20E-07Mo (g)ABABCC00.000020.000040.000060.000080.00010.00012Mg (g)ABABAB00.000020.000040.000060.000080.00010.000120.000140.00016S (g)AB ABAB463 Field collected, greenhouse grown sibling pairs may recognize kin throughmycorrhizal networks in a greenhouse settingIntroductionThe interior Douglas-fir (Pseudostuga menziesii var. glauca (Beissn.) Franco) forests arewidely distributed across western North America, ranging from north-central British Columbia(55oN, up to 760 m elevation) to northern Mexico (19oN, up to 3260 m elevation).  They varybroadly in climate (precipitation range 410?3400 mm per year; mean July temperature 7?30oC;mean January temperature -9 to 8oC), disturbance regimes (e.g., stand maintaining to standreplacing fires) and site quality (very dry and poor to very moist and rich) (Hermann andLavender, 1990). The forests are highly sensitive to these climatic, disturbance and sitevariations, and this is reflected in the composition (relatively pure to mixed stands), structure(multi-storied and uneven-aged to single-storied and even-aged) and species ecophysiology (e.g.,shade tolerant to shade intolerant). Within British Columbia, the Interior Douglas-fir (IDF)biogeoclimatic zone includes very dry climatic regions, known as the ?interior drybelt?, whereinterior Douglas-fir is shade tolerant and forms relatively pure, uneven-aged forests, as well aswetter, more productive climatic regions, known as the ?interior wetbelt?, where interiorDouglas-fir is shade intolerant and grows as an early or mid-seral species in mixture with up toeleven other conifer and broadleaf species. Establishment and growth of interior Douglas-fir islimited in the drybelt primarily by drought and cold temperatures, and in the wetbelt byinterspecific competition for light. This variation has led to distinct management considerations(Klenner et al. 2009).  For example, selective harvesting and natural regeneration of interiorDouglas-fir dominates silviculture systems in the drybelt, whereas clearcutting and plantinginterior Douglas-fir are favoured in the wetbelt.  Natural regeneration is highly variable due tohigh interannual variation in seed production and climate, and planting success varies from very47low (<50% 5-year survival) in the drybelt to very high (>80%) in the wetbelt (Newsome et al.1991, Vyse et al. 2006). In addition to climatic, site quality and interspecific variation, there areseveral other ecological factors that may influence regeneration success, including kinrecognition and the presence of mycorrhizal networks, the main concepts explored in this study.Kin recognition among interior Douglas-fir could be a factor affecting managementdecisions in IDF forests. Kin recognition is the ability of individuals to discriminate kin, orclosely related neighbours (i.e., siblings), in competitive interactions (Dudley and File, 2007).This is a well-known concept among social insects and other animals and is usually translatedinto kin selection. Kin selection is an altruistic behaviour defined by Hamilton?s rule; if the costof an action to an individual is less than the benefit to another individual multiplied by the levelof relatedness, the individual will proceed with the costly behaviour (Hamilton, 1964). This isthought to have evolved from the increase in indirect fitness of the genes shared by the relatedindividuals. Kin recognition and selection are more difficult and nuanced processes to observe inplants than in animals. Nonetheless recent research has shown evidence for kin recognition insome plant species (van der Heijden and Horton 2009, File et al. 2011). Kin recognition haspresented in both increased competition (Wilson et al. 1987, Tonsor 1989, Donohue 2003) andincreased cooperation between kin individuals (Cheplick and Kane 2004, Boyden et al. 2008,Milla et al. 2009). The ability of interior Douglas-fir to recognize kin and the response to kinwithin a stand could inform future management decisions.Interior Douglas-fir has large, continuous populations spanning disconnected geographicareas due to wind dispersal of pollen and seed (Hamrick et al. 1992). Gene flow due to thewidespread wind dispersal of pollen, and to lesser degree of seed, has led to relatively littlegenetic variation for selectively neutral genetic markers within or among populations in the same48region (Gugger et al. 2010). The majority of the among population genetic variation for adaptivetraits is due to high regional variation stemming from significant adaptive phenotypic variationlinked to the local environmental conditions. We used interior Douglas-fir as a test species in thisstudy because these characteristics may influence the ability or necessity of seedlings torecognize kin; we asked (1) level of relatedness - how genetically close do the seedlings need tobe for kin recognition to take place? and (2) seed origin ? will seeds coming from areas ofdiffering climatic regions, or levels of environmental stress differ in kin recognition ability? Asopposed to simpler self/non-self recognition (i.e. prevention of self-pollination) with a binaryoutcome, kin recognition involves several possible states along a gradient of relatedness (Chen etal. 2012). Regeneration success may be improved with management practices that consider, forexample, the effect of degree of relatedness among individuals on kin recognition. Kinrecognition and selection may decrease along a relationship gradient from strong recognition andselection between siblings/progeny (siblings, from the same parent trees for this study) with thehighest degree of relatedness, to within-population (different parent trees within the same site),to among-populations (different parent trees from different sites, thus with the lowest degree ofrelatedness).The high regional genetic variability and plasticity of interior Douglas-fir allows thespecies to persist in environmental conditions that vary greatly throughout its range. Due to thisvariability, populations vary widely in germination, growth and biomass allocation patterns(Kremer 1994, St. Claire et al. 2005, Gugger et al. 2010). The level of environmental stressamplifies the variation. The stress-gradient hypothesis (Greenlee and Callaway, 1996) refers to aconceptual model allowing for greater facilitation among individuals in a community withgreater environmental stress (Bertness and Callaway, 1994). Essentially, the harsher the49conditions, the more likely neighbours will benefit from cooperation. Kin selection is a form ofcooperation between related individuals in a community, and therefore could be influenced bythe amount of stress experienced by the community. Greater environmental stress could lead togreater kin recognition or selection.Mycorrhizal networks with mature trees have been shown to enhance survival and growthof nearby interior Douglas-fir seedlings more so in the drybelt than the wetbelt of the IDF zone,in keeping with the stress gradient hypothesis (Bingham and Simard 2011). Mycorrhizalnetworks are underground pathways formed by the hyphae of fungi that symbiotically associatewith plant roots (from the Greek ?myco? for fungus and ?rhiza? for root) and connect two ormore plants of the same or different species (Francis and Read 1984, Perry et al. 1989, Hortonand van der Heijden 2009). Many studies have documented mycorrhizal networks as mediatorsfor interplant transfer of water, nutrients and other chemicals, such as defense signals orallelochemicals (Agerer 2001, Querejeta et al. 2003, Teste et al. 2010, Bingham and Simard2011, Barto et al. 2012). Here we investigate whether kin recognition can also be mediatedthrough these networks. Given that access to a mycorrhizal network has been seen to improveinterior Douglas-fir regeneration under drought stress (Bingham and Simard, 2011), weexamined whether mycorrhizal networks facilitated kin recognition differentially along a stressgradient, where seeds originating from harsh climatic conditions would produce seedlings moreapt to form beneficial mycorrhizal associations and result in greater kin recognition than thosefrom more favourable regional climates. In the face of predicted climate change, understandingresponses to these environmental stresses could prove important for successful regeneration offuture forests.50The objective of this study was to determine whether kin recognition or kin selectionwould be detectable between greenhouse-grown interior Douglas-fir seedling pairs from fieldcollected seed, whether it is mediated by mycorrhizal networks, and whether seed origin or levelof relatedness affects these processes. We used seeds from six mature interior Douglas-fir treescollected from three locations (2 trees per location) that differed in regional climate in interiorBritish Columbia, to yield sibling (seeds originating from the same parent tree (kin)) and non-sibling (seeds originating from different parent trees (non-kin)) pairs. To examine kinrecognition along the relationship gradient, we examined the non-kin pairs in two groups whereseedlings originated from different parent trees either from (a) the same site or (b) a different site.Our first hypothesis was that kin pairs would have greater growth and foliar nutrient content,indicating kin selection, in comparison to non-kin pairs. The enhancement in growth andnutrition was expected to increase along a relationship gradient, with the greatest valuesoccurring among kin pairs, followed by non-kin originating from the same site (withinpopulation), then non-kin originating from different sites (among populations). Our secondhypothesis was that kin recognition, and all factors that affect it, would be mediated bymycorrhizal networks. We predicted that kin recognition would be greater in seedling pairs thatwere connected in a mycorrhizal network compared to those where the seedlings were isolatedfrom each other. We expected that recognition would also be affected by the extent of networkformation (amount of colonization). Our third hypothesis was that kin recognition wouldincrease with the competitiveness (e.g., growth rates) of plant genotypes. We expected greaterkin recognition among genotypes originating from sites that yielded larger seedlings.  Weexpected seedling root allocation to decrease with increasing regional precipitation of genotypeorigin, providing seedlings the ability acquire greater soil resources under drought-stressed51conditions through mycorrhizal networks and kin recognition, in agreement with the stress-gradient hypothesis.MethodsExperimental design and treatmentsThis experiment took place in the University of British Columbia greenhouse inVancouver, British Columbia, Canada over an 11 month period (March 2012 to February 2013).Half of one bench was used to equally distribute 180 pots, approximately 2 L with a height of 18cm and diameter of16.5 cm. The pots were rotated randomly every two weeks. Nosupplementary light was provided. The seeds were acquired in the fall of 2011 from matureinterior Douglas-fir trees in three locations in the interior of British Columbia. The Alex FraserResearch Forest, Knife Creek block (121.88?W, 52.05?N) and Farwell Canyon locations(122.63?W, 51.79?N) are outside of Williams Lake in the Cariboo Chilcotin Coast region. ThePaska Lake location (120.67?W, 50.50?N) is approximately 30 km southwest of Kamloops in theThompson Okanagan Region (Figure 3.1). ClimateBC provided the mean annual temperature(MAT) and precipitation (MAP) for each location in the 1981-2009 climate reference period(Wang et al. 2012). MAT of the Alex Fraser site is 4.8?C and MAP = 470 mm. MAT of theFarwell Canyon site is 3.5?C and MAP = 402 mm. MAT of the Paska Lake site is 3.5?C andMAP = 411 mm. These locations will be discussed in greater detail in chapter 4 (Figure 4.1, 4.2).The top two seed producing trees from each location (six total) were used as the seed source forthe donor seedlings for this experiment.A 2 x 3 factorial design was used, where mesh size (two levels) and relationship (threelevels) were applied in a completely randomized design.  The mesh size factor (two levels)52references the pore size of the specialized mesh bags (approximately 20cm x 8 cm) made byPlastok? (Meshes and Filtration) Ltd. (Birkenhead, UK) that separated the root systems of theolder, ?donor? seedling and the younger, ?recipient? seedling. A total of 90 mesh bags had a poresize of 35 ?m, which has been shown to allow fungal hyphae to penetrate the bag but preventroots from growing through the barrier (Teste and Simard 2008). The other 90 mesh bags had apore size of 0.5 ?m, which has been shown to prevent fungal hyphae from penetrating but allowsoil water to flow in and out of the bag (Teste et al. 2006).  The relationship factor (three levels)references the genetic closeness of the seedling pair grown in a single pot. A total of 90 of thepots received a kin pairing, with both seedlings grown from seeds originating from the same tree.These trees were not control pollinated, therefore, the seedling pairs of kin relationship could befull or half siblings. We also examined the relationship of seedling pairs that did not originatefrom the same tree but came from the same location (i.e., non-kin but from the same site). Thesewere expected to be genetically closer than seeds from different sites but are not guaranteed to behalf siblings (because of the lack of controlled pollination, they could be half siblings). Theremaining 90 pots received a non-kin pairing (i.e., non-kin from different locations). Therelationship pairings and mesh sizes were combined in a way that resulted in 48 kin pairs, 36non-kin intra-site pairs and 96 non-kin inter-site pairs that were evenly split between the network(35 ?m pore size mesh) and no network (0.5 ?m pore size mesh) treatments and evenlydistributed among the six donor seedling origins.Experimental setupEach pot contained a mesh bag that was placed against the pot edge and gently packedwith a 3:1 mixture of the greenhouse standard potting mix: field collected soil (Figure 2.1a). This53mixture was blended thoroughly in a mechanical soil mixer to achieve high homogeneity. Themixture was autoclaved for one to 1.5 hours at 250?C to kill any fungal inoculum provided bythe field collected soil. The remainder of the pot was filled with a 1:1 combination of non-autoclaved field collected soil and greenhouse standard potting mix that was thoroughly mixed.This design ensured that mycorrhizal colonization of receiver seedlings planted inside the meshbags occurred through mycorrhizal hyphal growth from the field soil-inoculated donor seedlings.The soil was collected from an interior Douglas-fir stand just outside of Princeton, BritishColumbia (approximately 120.58?W, 49.43?N) (Figure 3.1). The forest occurred in the Dry CoolInterior Douglas-fir subzone (IDFdk), was comprised of pure interior Douglas-fir, and wasunderlain by a Dystric Brunisol soil with sandy loam texture and a moder humus form (CanadianSystem of Soil Classification 1998). The top litter layer was scraped off and the underlying 10 to15 cm (including the fermentation layer, humus layer and mineral soil) was collected andtransported immediately to the UBC greenhouse in Vancouver, BC. This soil was expected tocontain a sufficient amount of interior Doulas-fir compatible fungal inoculum to encouragemycorrhizal colonization of donor seedlings within the pot.Each pot was designed to contain a pair of seedlings (Figure 2.1b). One older seedlingwas established 8 months in advance to act as a donor to the younger, recipient seedling. Thepairs were subject to one of four treatment groups, described above, depending on therelationship of the pair of seeds sown (full siblings or from separate families) and the pore size ofthe mesh bag installed in the pot (0.5 ?m or 35 ?m).At the onset of the experiment, all pots were soaked thoroughly with water. For the firstsowing, a total of 150 seeds from each of the six donor seed origins described above were sownin 30 pots (180 pots in total), with five seeds per pot. Before sown, the seeds were sterilized with5410% H2O2 for 10 minutes then dried. An additional three seeds per pot were stratified for asecond sowing, which entailed a 24 hour soak in distilled water followed by drying andrefrigeration (4?C) for three weeks. These were sown after six weeks due to unsatisfactorygermination rates in the first sowing. We also transplanted some seedlings from pots withmultiple germinants to those yet to germinate with the goal of having at least one seedling perpot. These were all sown in the mixed field soil containing the inoculum. A thin layer of finegravel known as ?forest sand? was spread over the sown seeds to discourage the growth ofdamping off fungi.The pots were lightly watered each day until each pot had at least one healthy seedling.The watering regime shifted from a light watering daily to weekly watering to field capacity. Theabsence of fertilizer and the limited watering regime were meant to encourage mycorrhizalfungal colonization. The seedlings were allowed to grow for another two months with periodicthinning when necessary.When the seedlings were 4.5 months old they were transported into a Conviron PGV36Plant Growth chamber at UBC to undergo a blackout regime to induce budset and the onset ofdormancy in preparation for an artificial winter. The blackout regime consisted of a 10 dayperiod with 10 hour days at 20?C and 14 hour nights at 15?C. After the blackout period, theseedlings were returned to the greenhouse for three weeks to allow them to become dormant forwinter, and then were returned to the growth chamber for an artificial winter. During theartificial winter, the seedlings experienced 16 hour uninterrupted nights, 8 hour days at low lightlevels and a temperature of 4?C for six weeks to meet their chilling requirement to breakdormancy and enter quiescence. The seedlings were then returned again to the greenhouse. Thesoil and these ?donor? seedlings were allowed 24 hours to adjust to the temperature difference,55and the second set of seeds intended as ?recipient? seedlings were sown the next day. Fivestratified seeds from the corresponding kin, non-kin same site or non-kin different site pairingwere sown within the edges of the mesh bag. The seeds were from the same stock as theseedlings currently established.The seeds were sown in accordance with the tree from which the donor seedlingoriginated and the tree from which the new (recipient) seedling originated. After the second setof seeds was sown, they were allowed to germinate, grow and were thinned when necessary overa four month period. Germination rate for all recipient seeds was recorded. Height was alsorecorded for both seedlings every three weeks. A total of 89 seedling pairs remained healthy andwere harvested at the end of the experiment, which represented 46.4% establishment success ratefor the pots. There were 21 kin/network, 20 kin/no network, 19 non-kin/network (5 of whichwere non-kin from the same site) and 29 non-kin/no network (8 of which were non-kin from thesame site).MeasurementsThe growth variables examined in both the donor and recipient were needle, stem androot weight (g), total and above ground biomass (g), average weight per colonized root tip (g perroot tip), percent of total root weight colonized (%), stem length (cm), germination rate(germinants per seeds sown), average growth rate (cm per week), total leaf area (cm2) andvolume, calculated as total leaf area x stem length (cm3) in order to estimate the overall effect ofthe height and conical volume of the seedling. Average leaf area per needle (cm2) was examinedonly in recipient seedlings. The seedlings were removed from the pots after the youngerseedlings were grown for four months. The below and above ground biomasses were separated.56The soil was carefully removed from around the roots, which were then washed with tap water.The needles were removed and measured using a LICOR-3100 leaf area meter. The number ofneedles was also recorded for the younger, recipient, seedlings only. The needles were dried andbiomass was recorded. The stems were measured for length, number of branches and drybiomass.The root tips were examined for mycorrhizal colonization and those that appeared to becolonized were weighed and morphotyped. The DNA of a subset of tips was sequenced at theIrving K. Barber School at UBC-Okanagan. For further details refer to Chapter 2 methods.A suite of foliar nutrients were examined using microwave digestion/ICP (InductivelyCoupled Plasma-Optical Emission Spectrometer) and % C, N and S analyses at the BritishColumbia Ministry of Environment Analytical Chemistry Laboratory. The macronutrientsexamined were C, N, K, Ca, Mg, P and S. The micronutrients examined were Al, B, Cu, Fe, Mn,Mo, Na and Zn. The extrapolated gram content per seedling and average sample concentrationswere used in our analysis. C:N and N:P ratios were calculated. Foliar nutrients were measuredonly for recipient seedlings.Data analysisAll statistical analyses were conducted using SAS version 9.3 (Cary, North Carolina).Two-way analysis of variance (ANOVA) was run separately for each variable.  The factorsincluded were relationship, network and relationship x network. An analysis of covariance(ANCOVA) was also run to examine the effect of seed origin (location) as a covariate.Significance was determined at p ? 0.05 and trends were noted where p ? 0.10.57ResultsMycorrhizal colonizationTwo variables were used to evaluate colonization: percent of total root weight colonizedand average weight per colonized root tip. In the donor seedlings, there was a main effect ofmesh size (network) (Table 3.1). A higher percent of the donor root weight comprised ofcolonized root tips and greater average weight per colonized root tip occurred in the 0.5 ?m meshbag where mycorrhizal network formation was prevented (p<0.05). In the recipient seedlings,both variables were higher in kin than non-kin seedlings (p<0.05) (Table 3.2, Figure3.2). The kineffects were not significant in the donor seedlings and the network effect was not significant inthe recipient seedlings. The interaction between network formation and seedling relationship wasnot significant for either variable in the donors or recipients.GrowthRelationship effectsIn donor seedlings, there was a significant effect of relationship on total leaf area andvolume (p<0.05) as well as a trend in stem length (p<0.10) (Table 3.1). Donor seedlings with akin relationship with recipients (i.e., seeds of each seedling coming from the same tree) hadgreater leaf area, volume and stem length than those with a non-kin relationship (Figure 3.3).There were no significant kin effects for any recipient seedling variables (Table 3.2). We alsotested the difference between non-kin from same site (both with and without networks) againstall other treatments, including non-kin from different sites, for all growth variables. There wereno differences between the same site non-kin seedlings and any other treatment. For all otheranalyses, the non-kin seedlings from the same and different sites were simply considered non-kin.58Network effectsRoot weight, total biomass and growth rate were lower in recipient seedlings of pairs thatwere separated by 35 ?m mesh than 0.05 ?m mesh (p<0.05)(Table 3.2, Figure 3.4). Aboveground biomass of networked recipient seedlings also tended to be lower (p<0.10) (Table 3.2).Mycorrhizal networks had no significant effect on any of the donor seedling variables (Table3.1).Relationship x network effectsThe only significant relationship x network interaction effect occurred in root weight ofdonor seedlings (Table 3.1). Donor seedlings with a kin relationship and no network (0.5 ?mmesh) had more root biomass than those with a non-kin relationship and no network (Figure 3.5).Networked non-kin donors also tended to have more root biomass than isolated non-kin donors.ANCOVA showed no origin effects on any growth variables.Seed origin effectsSeed origin had a significant effect on the growth rate of recipient seedlings (p = 0.0440)(Table 3.4). The seedlings from Paska Lake seeds grew faster than seedlings from Alex FraserResearch Forest seeds. The growth rate of seedlings from Farwell Canyon seeds did not differentfrom either seed origin location (Figure 3.6a)NutrientsOnly recipient seedlings were measured for foliar nutrient content. There was greateraluminum and copper content in recipient seedlings with kin than non-kin relationships (Figure593.7). The same site, non-kin seedlings were first tested separately from the different site, non-kinseedlings. The kin seedlings, regardless of their ability to form networks, had significantlygreater aluminum than non-kin, regardless of whether they came from the same or different sites.The networked kin seedlings had significantly greater foliar copper content than all othertreatments (Figure 3.8). There no difference between the same site, non-kin and the different site,non-kin for any variables (p>0.10), so all non-kin were combined for all other analyses.There was lower boron and manganese content in recipient seedlings that were connectedby mycorrhizal networks (35 ?m mesh) than those isolated (0.5 ?m mesh), paralleling networkeffects on growth responses. The opposite effect occurred with copper, where recipient seedlingshad higher copper where they could form a network with the donor.Copper, zinc (p<0.05) and iron content (p<0.10) were affected by interactions betweenrelationship and network (Figure 3.9). The origin of the seed was tested as a covariate, but hadno effect on any of the nutrient variables.Germination ratesGermination rate differed among seed origins (locations) (p<0.0001) (Table 3.4). TheFarwell Canyon site had a significantly higher germination rate than the Alex Fraser ResearchForest or Paska Lake sites (Figure 3.6b). There were also significant interaction effects; however,these are not likely ecologically relevant since relationship and network effects are unlikely toaffect germination (Figure 3.10).60SoilsThe autoclaved soil had higher percent total carbon, sulphur and available phosphorus butlower pH than non-autoclaved mixed field soil. There was no difference in percent total nitrogenor mineral nitrogen. Autoclaved soil had greater exchangeable cations (CMol+/Kg), exceptmanganese which was higher in the mixed field soil, and iron which was unaffected byautoclaving. The effective cation exchange capacity was also higher in the autoclaved soil. Thesoil in this experiment was identical to that of the full sibling experiment described in chapter 2(Table 2.6).DiscussionKin effectsEvidence of kin recognition was found in both foliar nutrients and seedling growth. Wedefined kin recognition as a significant difference (greater or lesser) in a variable depending onthe seedling?s relationship to its pair. We expected the difference to present as kin selection,where kin seedlings, recipients in particular, have greater nutrient content or growth ratescompared to non-kin seedlings. We found kin selection expressed in greater foliar content of themicronutrients, aluminum and copper, in kin than non-kin recipient seedlings. Copper is knownto be essential for foliar enzyme activity, chlorophyll formation, and is important in seedformation and production (Mengel and Kirkby 2001, Marschner 2012). Aluminum has apotential role in protecting roots against disease (Marschner 2012). Both copper and aluminumare known to be involved in biochemical signaling, particularly ethylene pathways, involved inphotosynthesis and plant growth and development (Hirayama et al. 1990). It is possible thatincreased foliar copper and aluminum among receiver kin played a role via signal transduction or61root physiology in enhancing mycorrhizal colonization and health of receiver roots.Alternatively, increased mycorrhization has been shown to improve copper nutrition in plants (Liet al. 1991). There was no evidence for expression of kin recognition in foliar macronutrients,similar to previous studies showing kin effects on other plant traits such as plant physiology,morphology or growth strategies (File et al 2011, Chen et al. 2012).We also observed kin selection expressed in greater mycorrhizal colonization of kin thannon-kin recipient seedlings (i.e., greater average weight per colonized root tip and greater percentof total root weight colonized). Enhanced mycorrhizal colonization suggests that the kinrelationship is providing the recipient seedlings with some advantage, such as improved rootphysiology, that is recognized and sought by the mycorrhizal fungi. File et al. (2012) also foundmore extensive mycorrhizal network formation in kin pairs of ragweed seedlings.Kin recognition was also evident in some donor seedling growth traits. Total leaf area,volume and stem length of donor seedlings were greater where they were paired with kin thannon-kin seedlings. In contrast to our expectation, the enhanced growth traits of kin donors didnot affect the growth of kin recipients. However, greater growth rates of donor kin may haveplayed a role in the increased mycorrhization and micronutrient content of recipient kin.  Donorfacilitation of recipient kin, or vise versa, could have occurred through interplant carbon transferas shown in Chapter 2, or through a priming effect between donor and recipient roots.  With norecognition detectable in the recipient seedling growth traits, it is difficult to speculate as towhether the enhanced growth in the donor seedlings was due to kin selection or reducedcompetition (i.e., niche partitioning) (File et al. 2011).62Seed origin effectsGermination rates were greatest among recipient seeds originating from the driestlocation, Farwell Canyon (Figure 3.6b). Moreover, kin seedlings originating in Farwell Canyonhad significantly higher germination rates than the non-kin seeds (Figure 3.9.).  These kin effectswere not evident among seed originating from the other two wetter sites.  The effect ongermination rates suggests that seeds from the harsher climate were better adapted to massgermination in the greenhouse, as many germinants would die in normal field conditions. Theyare also likely to support greater mycorrhizal colonization for enhanced uptake of limited waterand nutrients. These results support our hypothesis that kin selection would be greatest amonggenotypes that are more competitive in limited resource acquisition. The growth rates amongdifferent seed origins corroborated this hypothesis. Seedlings originating from seed at FarwellCanyon grew statistically at the same rate as those from the wetter, Paska Lake origin.Considering the harsh conditions of the Farwell Canyon site, we could expect slower growthrates, but we did not observe this. Rehfeldt (1989) showed that in common garden experimentsseedlings from regions with fewer frost free days had reduced growth compared to those fromregions with a greater number of frost free days. As the Farwell Canyon site had a shorter frostfree period (84 days) than did Paska Lake (98 days), this was an unexpected result (Wang et al.2012).This result supports the stress gradient hypothesis, which suggests that environmentalstress is a key driver in inter-plant facilitation (Greenlee and Callaway 1996, Callaway et al 2002,Liancourt et al. 2005, Cavieres et al. 2006). It is possible that seeds from Farewell Canyon weregenetically primed for kin recognition to facilitate regeneration under climatically stressful63conditions. The transfer of water from donor to recipient plants through mycorrhizal networkshas also been shown to increase with environmental stress (Bingham and Simard 2011).Level of relatednessWe reject our hypothesis that the strength of kin effects would be detectable along agenetic gradient of relatedness, with decreasing strength from kin, to non-kin from the same site,to non-kin from different sites. There was a difference between the kin seedlings and all the non-kin seedlings combined, but no differentiation between non-kin seedlings originating from thesame versus different sites. The noted differences were only observed in variables that had kineffects regardless of the separation of non-kin due to site. There could be a number ofexplanations for this result. Kin recognition could only be sensitive at the sibling/progeny leveland therefore relatedness would be undetectable at the population level. If non-kin seedling pairsoriginating from the same site did not share paternal nor maternal genes, they would not bedifferentiated from the non-kin pairs originating from different sites at that level of sensitivity.Alternatively, if the recognition process heavily relies on maternal effects and is only sensitive tothe sibling/progeny level, the non-kin originating from same versus different sites would beindistinguishable (Donohue, 2003). Due to the wind dispersal process of both the pollen and theseeds of interior Douglas-fir, the populations can be large and continuous, covering significantgeographic distances (Hammerick et al. 1992). If these sites are experiencing gene flow amongstthem, which is likely the case between the Farwell Canyon and Alex Fraser Research Forest sitesas they are relatively close in proximity (approximately 55 km apart), they could be consideredone population and therefore, the non-kin originating from the same or different sites would besimilar in genetic makeup. Little variation among populations of Douglas-fir has been found64within the same region or among individuals within a population (Gugger et al. 2010) inselectively neutral markers. This is consistent with Rehfeldt (1989), who found that most geneticvariation in growth traits such as height and growth strategy occurred over significant differencesin geographic distance or elevation, which translated into adaptive differentiation depending onthe number of frost free days in the region. However, our results must be considered with cautiongiven the small sample size and the possibility that the condition of the parent tree and thequality or maturation of the seed collected may have influenced the results.Mycorrhizal network effectsThe DNA sequencing performed on the colonized root tips showed that Rhizopogonvinicolor was the most common fungal association, colonizing 92.2% of donor roots and 95.6%of recipient roots. This result was expected; Rhizopogon vinicolor is a strong networking fungalspecies that dominates interior Douglas-fir throughout forest development (Tweig et al. 2007,Beiler et al. 2010). The fungal community also included Pyronemataceae (sp.) and ascomyceteendophytes. In networked treatments, 39 of 40 recipients were colonized by at least one fungalspecies and had at least one ectomycorrhizal species shared in common between the donor andrecipient. This provides sufficient evidence that the seedlings in a pair had the potential to formmycorrhizal networks.Our hypothesis that the kin effects were facilitated by mycorrhizal networks, where thekin effect would only be detectable where seedlings were connected by a network, was partiallysupported by our data. Recipient seedlings had higher copper, and to a lesser degree iron contentin the kin/network treatment than the other treatments. It is possible that copper and iron werepreferentially transferred to kin seedlings through mycorrhizal networks in enzymes or hormones65acting as chemical signals or nutritional supplements for receiver plants. However, the ability toform mycorrhizal networks had no effect on any other nutrient variable where kin recognitionwas detectable.Contrary to our expectations, we found that the presence of mycorrhizal networksreduced the growth rate, total biomass and root weight, as well as some foliar nutrients (boronand manganese), of recipient seedlings.  We expected that mycorrhizal networks would facilitaterecipient seedling growth based on previous research showing enhanced germination, survivaland growth of seedlings where they were networked with older trees (Teste et al. 2009, Binghamand Simard 2011). In these studies, increased seedling performance was associated with resourcetransfer along carbon and nitrogen source-sink gradients. The opposite growth effect that wefound agrees with a study by Merrild et al. (2013), who showed a size discrepancy between twotomato plants connected by a mycorrhizal network was associated with preferential P uptake bythe larger and P deficiency in the smaller plant. We did not measure foliar P content in bothdonor and recipient seedlings in this experiment; however, foliar P concentration and contentwere greater in the larger donors than the smaller recipient seedlings in the full siblingexperiment (Chapter 2). The recognition of another nearby seedling via signaling throughmycorrhizal networks could prompt the donor seedling to increase its competitive effect onrecipient seedlings, thus reducing recipient seedling growth via the niche partitioning or ?elbowroom? hypothesis (Young 1981, File et al. 2011). In our study, this resulted in reduced rootweight, total and above ground biomass, and growth rate of the recipient seedlings withmycorrhizal networks.  This did not coincide with network effects on donor seedling growthrates. However, donors that formed networks with recipient seedlings had lower mycorrhizal66colonization rates than those that were isolated, potentially resulting from network-enhancedinterplant competition.ConclusionKin recognition was evident through enhanced mycorrhizal colonization and foliarmicronutrients of recipient seedlings and greater growth rates of donor seedlings.  However ourhypothesis that kin recognition would be expressed as kin selection through enhanced growthrates of recipient seedlings was not supported. The effect of seedling relatedness was evidentwhen comparing sibling and non-sibling relationships, but did not appear to act along a gradientof genetic relatedness.  Kin recognition appeared to be present in germination rates of seedsoriginating from the driest location but not the two wetter locations, suggesting that competitiveability of genotype affects kin selection. This finding also supports the stress gradient hypothesis.The variation in germination may be attributed to maternal effects such as the health of the treeand the size of the seeds it produced. Mycorrhizal networks played an unexpected role in kinrecognition by reducing growth rate of recipient seedlings and thus increasing asymmetricalcompetition between donor and recipient seedlings.  They also were associated with improvedrecipient copper and iron nutrition, which perhaps played a role in chemical signaling betweenplants.The ability of interior Douglas-fir to recognize kin may have forest managementimplications. The positive relationship that kin recognition had on aluminum and copper contentas well as germination rate and mycorrhizal colonization suggests that, in some cases,particularly where germination is low, or aluminum or copper is scarce, preferentially promotingkin regeneration may be beneficial. Retaining large, healthy legacy trees for natural regeneration67(especially in dry climates), or pre-planning to grow and plant kin seedlings originating fromthose legacy hub trees (especially in wetter climates), may result in higher germination andsurvival rates but further research is necessary to substantiate these claims. Variable retentionharvesting may be particularly effective in dry sites, such as the Farwell Canyon region, whereenvironmental stress may drive cooperation within the biotic community in accordance with thestress gradient hypothesis (Greenlee and Callaway, 1996). This management approach wouldalso conserve mycorrhizal networks, which have been shown to benefit seedling establishment,particularly in dry sites (Bingham and Simard, 2012). The results from this study also suggestthat mycorrhizal networks and kin recognition would enhance growth variation amongneighbouring seedlings, perhaps leading to more diverse stand structures.68Table 3.1. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in donor seedlings. *p <0.05Network RelationshipNetwork xRelationshipF P F P F PNeedle weight (g) 0.03 0.8723 0.27 0.6075 0.85 0.3604Stem weight (g) 0.01 0.9318 0.19 0.6671 1.04 0.3115Root weight (g) 0.10 0.7534 0.58 0.4477 4.11 0.0458*Total biomass (g) 0.03 0.8560 0.41 0.5226 2.08 0.1531Above ground biomass (g) 0.01 0.9253 0.26 0.6098 0.98 0.3259Total leaf area (cm2) 0.52 0.4747 5.31 0.0237* 0.55 0.4596Stem length (cm) 0.00 0.9882 2.99 0.0875 0.32 0.5745Average weight per colonized root tip (g) 7.56 0.0073* 2.33 0.1309 1.20 0.2769Percent of root weight colonized 5.51 0.0212* 0.01 0.9297 0.24 0.6228Average growth rate (cm/week) 0.87 0.3539 0.43 0.5130 0.00 0.9945Volume (cm3) - (total leaf area x stem length) 0.17 0.6824 4.89 0.0297* 0.26 0.611369Table 3.2. Analysis of variance for network, relationship and interaction (network x relationship)effects on growth variables in recipient seedlings. *p <0.05TTNetwork RelationshipNetwork xRelationshipF P F P F PNeedle weight (g) 2.22 0.1403 1.18 0.2797 1.21 0.2738Stem weight (g) 1.72 0.1929 0.75 0.3899 0.61 0.4378Root weight (g) 4.21 0.0432* 0.03 0.8676 2.70 0.1043Total biomass (g) 5.88 0.0174* 0.51 0.4777 0.05 0.8294Above ground biomass (g) 3.01 00865 1.53 0.2193 1.50 0.2247Total leaf area (cm2) 2.04 0.1568 1.14 0.2886 0.61 0.4356Stem length (cm) 1.24 0.2679 0.37 0.5441 2.13 0.1485Average weight per colonized root tip (g) 0.40 0.5313 7.06 0.0094* 0.25 0.6215Percent of root weight colonized 0.04 0.8341 12.59 0.0006* 0.50 0.4834Average growth rate (cm/week) 4.19 0.0437* 1.33 0.2518 1.18 0.2796Volume (cm3) - (total leaf area x stem length) 0.43 0.5149 1.41 0.2386 0.06 0.8135Average leaf area per needle (cm2) 0.67 0.4158 0.02 0.8952 1.39 0.241570Table 3.3. Analysis of variance for network, relationship and interaction (network x relationship)effects on foliar nutrient content and 13C transfer in recipient seedlings. *p <0.05.Network Relationship Network x RelationshipF P F P F PAl (g) 0.27 0.6019 26.07 <0.0001* 1.59 0.2115B (g) 3.97 0.0496* 2.10 0.151 2.05 0.1555C (g) 2.11 0.1503 1.03 0.3132 1.20 0.2771Ca (g) 1.51 0.2222 1.49 0.2263 1.27 0.2637Cu (g) 4.81 0.0311* 5.48 0.0216* 8.40 0.0048*Fe (g) 0.13 0.7185 1.09 0.2987 3.35 0.0707K (g) 1.62 0.2064 1.49 0.2251 1.13 0.2901Mg (g) 1.21 0.2744 0.65 0.4235 1.32 0.2532Mn (g) 6.04 0.0160* 0.00 0.9890 0.79 0.3753Mo (g) 0.18 0.6711 0.27 0.6065 0.96 0.3312N (g) 1.92 0.1697 1.13 0.2905 0.87 0.3533Na (g) 0.77 0.3836 0.02 0.8853 0.30 0.5856P (g) 1.96 0.1647 1.42 0.2369 1.28 0.2606S (g) 0.67 0.4168 1.43 0.2352 2.00 0.1605Zn (g) 0.40 0.5293 0.48 0.4911 6.21 0.0147*C:N 0.00 0.9880 0.23 0.6361 0.56 0.4547N:P 1.30 0.2579 0.44 0.5073 0.91 0.343871Table 3.4. Analysis of covariance for network, relationship and interaction (network xrelationship) effects, with seed origin included as the covariate and its effects on growth andgermination rates recipient seedlings.*p <0.05.Growth rate Germination ratedf F P df F PNetwork 1 4.87 0.0303* 1 1.82 0.1801Relationship 1 0.00 0.9442 1 1.95 0.1648Network x relationship 1 3.92 0.0513 1 0.15 0.7018Seed origin 2 3.25 0.0440* 2 23.99 <0.0001*Network x seed origin 2 0.51 0.5996 2 0.26 0.7713Relationship x seed origin 2 4.19 0.0187* 2 21.79 <0.0001*Network x relationship x seed origin 2 4.51 0.0140* 2 0.14 0.873472Figure 3.1. Google map (? 2013 Google) of south-western British Columbia including threefield research sites (Alex Fraser Research Forest, Farwell Canyon and Paska Lake sites), the seedsource for the full sibling experiment (Kalamalka seed orchard), the soil source for thegreenhouse experiments (Soil source: greenhouse experiments) and the location of thegreenhouse (UBC greenhouse).Alex FraserResearch ForestsiteFarwell CanyonsitePaska LakesiteKalamalkaseed orchardUBC greenhouseMap data ?2013 GoogleSoil source:greenhouseexperiments73a                                                                        bFigure 3.2. Least square means of recipient seedling (a) average weight per colonized root tipand (b) percent of total root weight colonized under the main relationship effects withoutconsideration of the network treatment used. Different letters indicate bars that are significantlydifferent at p > 0.05. Error bars are one standard error above and below.00.000250.00050.000750.0010.001250.00150.001750.002Kin Non-kinRecipientAverage weight per colonizedroot tipAB051015202530Kin Non-kinRecipientPercent of total root weightcolonizedAB74a                                                                            bFigure 3.3. Least square means of donor seedling (a) volume (total leaf area x stem length) and(b) total leaf area under the main relationship effects without consideration of the networktreatment used. Different letters indicate bars that are significantly different at p > 0.05. Errorbars are one standard error above and below.020040060080010001200140016001800Kin Non-kinDonor volume (cm3 )AB01020304050607080Kin Non-kinDonorTotal leaf area (cm2)AB75Figure 3.4. The growth rate, total biomass, and root weight least square means of recipientseedlings differentiated by network effect. Different letters indicate bars that are significantlydifferent at p > 0.05. Each of the variables were compared by network effect, but were notcompared to each other. Error bars are one standard error above and below.00.050.10.150.20.250.30.350.40.450.5Growth rate  Total biomass  Root weight(g) NetworkNo networkABABAB76Figure 3.5. Least square means of root weight in donor seedlings under the four relationship andnetwork combinations (interaction effects). Different upper case letters indicate bars that aresignificantly different at p > 0.05. Different lower case letters indicate bars that are significantlydifferent at p > 0.10. Error bars are one standard error above and below.00.20.40.60.811.21.4Donor root weight (g)A aBbABKin Non-kinNetworkNo network77a                                                                               bFigure 3.6. The variation of (a) growth rate and (b) germination rate (least square means)between seeds originating from Farwell Canyon, Alex Fraser Research Forest and Paska Lake.Different letters indicate bars that are significantly different at p > 0.05. Error bars are onestandard error above and below.00.050.10.150.20.250.30.350.40.450.50.55FarwellCanyonAlex Fraser Paska LakeGrowth rate (cm/week)ABAB00.050.10.150.20.250.30.350.4FarwellCanyonAlex Fraser Paska LakeGermination rate(germinants/seeds sown)AB B78Figure 3.7. Aluminum (Al) and copper (Cu) content least square means of recipient seedlingsdifferentiated by relationship effect. Different letters indicate bars that are significantly differentat p > 0.05. Each of the variables were compared by relationship effect, but were not comparedto each other. Error bars are one standard error above and below.0.00E+005.00E-071.00E-061.50E-062.00E-062.50E-063.00E-063.50E-06Al Cu(g) KinNon-kinABAB79abFigure 3.8. same site. Least square means of (a) aluminum (Al) and (b) copper (Cu)differentiated by relationship effect between kin pairs (seeds from same parent trees), non-kin(same site) (seeds from different parent trees within the same site) and non-kin (seeds fromdifferent parent trees from different sites). Different upper case letters indicate bars that aresignificantly different at p > 0.05. Different lower case letters indicate bars that are significantlydifferent at p > 0.10. Error bars are one standard error above and below.0.00E+005.00E-071.00E-061.50E-062.00E-062.50E-063.00E-063.50E-064.00E-06Kin Kin Non-kin(same site)Non-kin(same site)Non-kin Non-kinAluminum (g)aBBbBBA0.00E+005.00E-071.00E-061.50E-062.00E-062.50E-063.00E-063.50E-064.00E-064.50E-06Kin Kin Non-kin(same site)Non-kin(same site)Non-kin Non-kinCopper (g)B B B BBA80Figure 3.9. Significant least square mean differences in recipient seedlings under the fourrelationship and network combinations (interaction effects) found in copper (Cu), iron (Fe) andzinc (Zn). Different upper case letters indicate bars that are significantly different at p > 0.05.Different lower case letters indicate bars that are significantly different at p > 0.10. Each of thevariables were compared by relationship effect, but were not compared to each other. Error barsare one standard error above and below.0.00E+002.00E-064.00E-066.00E-068.00E-061.00E-051.20E-051.40E-051.60E-05Cu  Fe  Zn(g)Kin/networkKin/no networkNon-kin/networkNon-kin/no networkAB B BaabbabABABA81Figure 3.10. Least square means of germination rate in recipient seedlings under the fourrelationship and network combinations (interaction effects) and seed origin effects. Differentletters indicate bars that are significantly different at p > 0.05. Error bars are one standard errorabove and below.00.050.10.150.20.250.30.350.40.450.50.55kin/no network kin/network non-kin/nonetworknon-kin/networkGermination rate(germinants/seeds sown)Farwell CanyonAlex FraserPaska LakeAABC BCBBCCCBC BC BCBC824 Mature trees may preferentially support kin seedlings in germination andsurvival through mycorrhizal networksIntroductionInterior Douglas-fir (Pseudostuga menziesii var. glauca (Beissn.) Franco) grows in awide variety of climatic conditions and therefore exhibits high genetic variability (Gugger et al.2010). The variation in climatic conditions has also created distinctive regeneration patterns. Indrier sites, seedlings tend to regenerate in clusters around larger, mature trees. In wetter sites,regenerating seedlings are more evenly distributed through disturbed openings (Simard 2009).Since interior Douglas-fir pollen and seed are wind dispersed (Hamrick et al. 1992) and it isunlikely that dispersal would be lower in dry than wet sites, factors other than seed dispersal arelikely influencing regeneration patterns. Under identical conditions, clustering would increasecompetition among regenerating seedlings, suggesting some advantage is gained by regeneratingnear mature trees. These areas may be reservoirs for water, nutrients and shade, perhapsexplaining why a tree grew to maturity there, but other facilitative processes may be present aswell. We are interested in the facilitative effect that kin recognition, mycorrhizal networks or thecombination of the two may have on regeneration of disturbed sites in the interior Douglas-firzone (IDF) of British Columbia.While greenhouse studies can isolate effects and control for growing conditions andgenetic variability, it is important to test whether these patterns occur in natural conditions in thefield. In this study, we examined kin recognition and mycorrhizal network influence on seedlingsgrown in the field. Our main hypotheses were identical to those of the greenhouse studies(Chapters 2 and 3). We predicted there would be a difference in germination and survival due tothe relationship of the seedling to the mature tree it was sown near (parent tree). We predictedthat the kin seedlings would have higher germination and survival than non-kin seedlings (kin83selection). The effects of relationship were predicted to be mediated by mycorrhizal networks. Inaddition we expected the relationship effect to be stronger where the growing conditions wereharsher in accordance with the stress gradient hypothesis.MethodsStudy sitesThree study sites were located in interior British Columbia, Canada. All sites hadretention of individual tress due to either partial harvesting (Paska Lake and Alex FraserResearch Forest) or natural disturbances (Farwell Canyon). None of the sites had been siteprepared or planted. Understory plant community composition, primarily native grass species,and abundance were similar among all sites. The Paska Lake site (120.67?W, 50.50?N) isapproximately 30 km southwest of Kamloops. The Alex Fraser Research Forest site in the KnifeCreek block (121.88?W, 52.05?N) is approximately 20 km southeast of Williams Lake in theCariboo Chilcotin Coast region of British Columbia. The Farwell Canyon site (122.63?W,51.79?N) is approximately 50 km southwest of Williams Lake also in the Cariboo ChilcotinCoast region of British Columbia. The Alex Fraser Research forest site is the warmest and thewettest of the three followed by the drier and cooler Paska Lake site and the driest site, FarwellCanyon (with same mean annual temperature as Paska Lake).All temperature and precipitationmeasures were retrieved from ClimateBC, over the most recent 30-year climate period, 1981-2009 (Wang et al. 2012) (Figure 3.1 & Table 4.1). The Alex Fraser Research forest and FarwellCanyon sites are located in the interior Douglas-fir (IDF) biogeoclimatic zone, and Paska Lake isin the Montane Spruce (MS) zone, however all sites are located in interior Douglas-fir forests.84They are found in three separate subzones; IDFdk (Dry Cool IDF), IDFxm (Very Dry Mild IDF)and MSxk (Very Dry Cool MS), respectively.Experimental design and treatmentsIn the fall of 2011, cones were collected from 10 to 16 trees in each of the study sites andseeds extracted. Five of these trees were selected at each of the three sites as experimental?parent trees? based on filled seed production, health and absence of neighbours. In May of 2012,all experimental units were installed at all three sites.A 2 x 2 factorial design was used where mesh size (two levels) and relationship (twolevels) were applied in a randomized complete block design, where the three sites served asblocks. The mesh size factor (two levels) references the pore size of the specialized mesh bags(approximately 20cm x 8 cm) made by Plastok? (Meshes and Filtration) Ltd. (Birkenhead, UK)that separated the root system of the seedling from those of the surrounding below groundcommunity. These bags would provide some protection from below-ground granivores andherbivores as they would not be able to penetrate the mesh from below ground. The relationshipfactor (two levels) references the genetic closeness of the seedling to the parent tree it was sownnear. A kin relationship refers to a seedling grown from seed originating from the same tree inwhich it was sown near, whereas a non-kin relationship refers to one grown from seedoriginating from a different mature tree at a different site.  A total of 24 nylon mesh bags(Plastok? (Meshes and Filtration) Ltd.), 12 at a pore size of 0.5 ?m and 12 at 35 ?m, werespaced evenly around each of the 15 trees at a radius of three meters centered on the tree bole. Aradius of three meters was selected to accommodate placement of all mesh bags and theirseedlings just outside the drip line of the crown. Previous research suggests this distance85represents a zone of optimal trade-off between minimal competition and maximal mycorrhizalcolonization (Teste and Simard 2009, Bingham and Simard 2011).Experimental setupThe mesh bags were placed in holes that were excavated to approximately the size of thebags (approximately 20 cm deep by 8 cm diameter). The soil removed was isolated, mixed andused to fill the mesh bag. The mesh bag was placed into the hole and any remaining soil wasused to secure to bag, fill the hole and obscure the view of the bag from wildlife. There was atotal of 360 mesh bags installed (24 per tree, five trees per site and three sites). The relationshippairing and mesh size combinations resulted in 180 kin and 180 non-kin relationships evenlysplit between the network (35 ?m pore size mesh) and no network (0.5 ?m pore size mesh)treatments.Five seeds were sown per mesh bag around each of the 15 trees over a three-day period inmid May 2012. For each of the mature trees, 12 of the bags were sown with stratified seeds thatwere collected from the proximal parent tree (i.e., kin seeds). The remaining 12 bags were sownwith seeds collected from the cones of three different mature trees from each of the other twosites (i.e., non-kin non-site seeds). The three trees from each site that provide the seed for thenon-kin seedlings were chosen purely by number of seeds obtained in the initial seed collection.The top three seed producers out of the five trees per site were used to provide the non-kin seed.The treatments were split evenly among the following four treatment combinations: (1) kin seedsin 35 ?m mesh pore size (kin/network), (2) non-kin seeds in 35 ?m mesh (non-kin/network), (3)kin seeds in 0.5 ?m mesh (kin/no network) and (4) non-kin seeds in 0.5 ?m mesh (non-kin/no86network). We also secured flexible, plastic mosquito netting around each bag in a dome-likeshape to reduce seed predation.Observations and data collectionOver four months in the summer of 2012 (May ?August), the sites were checked everythree to four weeks for germination, survival, soil moisture content and any disturbance to thetreatments. When the first germinant emerged from a bag, the mosquito netting was removed toensure it would not hinder growth. Soil moisture was measured at three consistent locationsaround each mesh bag with a HH2 moisture meter with an ML2x Theta probe (Delta-T DevicesLtd.) and averaged for the site (Table 4.2). The most common disturbances were cattle tramplingand bag tearing or removal by wildlife (mostly bears). In those cases, if the bag was not damaged,it was returned to its position and, if no seeds were visible, the appropriate seeds for thetreatment assigned to that bag were sown again. If the bag was damaged, the appropriate meshsized bag and seeds were replaced in its original position. At the Paska Lake site, there was someevidence of browsing, but most seedlings that germinated but did not survive appeared to havebeen under severe drought stress. Mid-way through June 2012, one data logger was installed ineach site at representative locations to record ambient air temperature at the surface of the soiland one to record soil temperature at a depth of 20 cm (approximately the maximum bag depth).Ambient air temperature was measured by a HOBO Pendant? Temperature/Alarm Data Logger(UA-001-64) and soil temperature was measured by a HOBO U23 Pro v2 External TemperatureData Logger (U23-004) (Table 4.2). Degree days below 4?C as well and soiltemperature:moisture index were derived from the climate data above. Degree days below 4?Care the number of days in which the ambient air temperature dipped below 4?C during the87measurement period (mid-June through August 2012). A more well-known climate variable isdegree days below 5?C, however we used 4?C because it better represented climatic differencesbetween sites. Soil temperature:moisture index (T:M) was calculated for each site by dividing theaverage soil temperature by the average soil moisture (Table 4.3).Data analysisTotal numbers of germinants and survivors as well as the number of experimental unitsthat contained a germinant and survivor were tallied. A germinant was defined as an emergingseedling with a defined stem. The existence of a germinant(s) for that experiment unit was notedat first observation. If that germinant persisted until the last observation it was considered asurvivor for the experimental period. This data was analyzed to detect effects of relationship,network, and the combination of relationship and network using simple statistics. The data wasanalyzed for site and seed origin effects. Germination and survival rates were calculated bypercent. Logistic regression (PROC LOGISTIC) was used to test seed germination probabilityand probability of presence of a survivor, in response to relationship, mesh treatment (network),the combination of relationship and mesh, as well as site. Odds ratios were calculated using theentire data set. Significance was determined at p < 0.05 and trends were noted at p < 0.10. Alldata analysis was conducted using SAS version 9.3 (Cary, North Carolina).ResultsGermination and survivalOf the 1800 stratified seeds sown, 157 germinated (8.72% germination rate) in 96experimental units (26.67% of experimental units with a germinant). Of the 157 germinants, 4088seedlings survived, yielding a survival rate of 25.5%. Of the 360 experimental units, 26 had atleast one surviving seedling, yielding 26.8% of germinating units with a survivor, but only 7.2%of all experimental units with a survivor. 9.8% of seeds with access to a network (35 ?m mesh)germinated, and of those, 19.3% survived. 7.7% of seeds without network access (0.5 ?m mesh)germinated, and of those, 33.3% survived. 8.9% of seeds with a kin relationship to the parent treegerminated, and of those, 17.5% survived. 8.6% of seeds with a non-kin relationship to theparent tree germinated, and of those, 33.7% survived (Table 4.4).In the logistic regression, site was a significant factor in both the germination of seedsand the presence of surviving seedlings (p = 0.0004 and 0.0001, respectively)(Table 4.5, 4.6).Network tended to have an effect on germination (p = 0.0699) (Table 4.5) and relationshiptended to have an effect survival (p = 0.0960) (Table 4.6). The odds ratios showed that seedswere 2.58 times more likely to germinate in the Paska Lake site than either of the other two sites.The seeds sown at Farwell Canyon were more likely to germinate than those sown at the AlexFraser Research Forest (Table 4.5, Figure 4.1a). Paska Lake was 7.80 times more likely to havesurvivors present in the experimental units than either of the other two sites. The Alex FraserResearch Forest was more likely to have surviving seedlings than Farwell Canyon (Table 4.6,Figure 4.1b). Seeds that were grown in a 35 ?m mesh bag (network) were 1.54 times more likelyto germinate than those in a 0.5 ?m mesh bag (no network)(Table 4.5, Figure 4.1a). Seedlingsthat had a non-kin relationship to the parent tree were 1.98 times more likely to be a survivorthan non-kin (Table 4.6, Figure 4.1b). There were no significant interaction (relationship xnetwork) effects.A separate logistic regression was conducted to determine if seed origin affectedgermination or survival. There was a significant effect of seed origin on presence of germinants89(p <0.0001).  The odds ratio indicated that, at present, a germinant was 3.73 time more likely tobe from seed originating from Farwell Canyon.  None of the other factors tested were significantfor germination or survival.Site and seed originThe Paska Lake site had the highest germination rate (13.2%) followed by FarwellCanyon (8.8%) and Alex Fraser (3.8%). Paska Lake also had the highest survival rate (40.5%)followed by Alex Fraser (21.7%) and Farwell Canyon (5.6%). The seeds that originated fromFarwell Canyon had the highest germination rate (16.7%) when considering all sites, followed bythose from Paska Lake (4.8%) and Alex Fraser (4.6%). The seeds originating from Paska Lakehad the highest survival rate (41.4%) followed by those from Alex Fraser (32.1%) and FarwellCanyon (19.0%) (Table 4.7). The seeds originating from Farwell Canyon germinated best in theFarwell Canyon and Paska Lake sites, with much less success in the Alex Fraser site. The seedsoriginating from the Alex Fraser Research Forest site had the most germinants in the Alex Frasersite and Paska Lake sites, with less than half in the Farwell Canyon site. The seeds originatingfrom Paska Lake germinated much better in the Paska Lake site, but with few germinants ineither of the other two sites (Table 4.8). Seedlings originating from any site survived best in thePaska Lake site; however, Farwell Canyon seeds had the most survivors in the Paska Lake sitefollowed by Paska Lake seeds and lastly Alex Fraser seeds (Table 4.9).Climate, soil moisture and relationshipSoil temperature:moisture (T:M) index had an interesting effect on the number of kingerminants versus non-kin germinants (Figure 4.2a). Kin seeds increased in number of90germinants as soil T:M increased, whereas non-kin seeds peaked at the intermediate value (PaskaLake site) and was lower at both the extremes. For both kin and non-kin seedlings, survivalpeaked at the intermediate soil T:M and had under five survivors at either of the extremes(Figure 4.2b). More non-kin survivors were present at the intermediate soil T:M value at PaskaLake due to the greater number of germinants, however, the survival rate of kin seedlings wasactually greater (47.8%) than that of the non-kin seedlings (37.5%) at that site.For soil moisture content measured over the experimental period, a similar patternemerged; as soil moisture decreased, kin seed germination increased (Figure 4.3a). At theintermediate soil moisture site, Paska Lake with 0.172 m3 per m3, there was greater germinationof non-kin seeds, but at the driest site (Farwell Canyon, 0.153 m3 per m3) the opposite occurred,with greater germination of kin than non-kin seeds. At the wettest site (Alex Fraser), there wasno effect of relationship on number of germinants. The greatest difference in survival betweenkin and non-kin seedlings (21 non-kin seedlings compared to 11 kin seedlings) occurred at thesite with intermediate soil moisture, Paska Lake. The number of survivors at the extremes did notvary by relationship (Figure 4.3b).Climate, soil moisture and networksSoil T:M index differentially affected seed germination depending on whether or not theyhad mycorrhizal network access to the parent tree (Figure 4.2c). Due to the low germination ratesand very low survivorship, no significant relationships were found; only trends were noted. Atthe two wetter sites (Alex Fraser and Paska Lake, respectively), seeds with access to a networkgerminated at a higher rate than isolated seeds. At the most drought stressed site (highest soilT:M), Farwell Canyon, there was no effect of network on number of germinants. As with91survival, both network and non-network seedling survival peaked at the intermediate soil T:M(Paska Lake) (Figure 4.2d). Isolated seedlings survived in greater numbers than networkedseedlings.Soil moisture content measured over the experimental period had the same effect onsurvival as soil T:M index. Germination at the driest site, Farwell Canyon, with an average soilmoisture content of 0.153 m3 per m3, was unaffected by the presence of networks (Figure 4.3c).As soil moisture increased (Paska Lake (0.172 m3per m3) then Alex Fraser (0.195 m3 per m3)),networks provided germinants an advantage over those that were not networked. The greatestadvantage networks provided for germination occurred at the intermediate site, Paska Lake. Thegreatest relationship effect also occurred at the intermediate site, where 19 non-kin seedlingssurvived compared to 13 kin seedlings. The number or survivors at the extremes did not vary byrelationship (Figure 4.3d).DiscussionGermination and survival ratesIn comparison to the greenhouse experiments discussed in the previous two chapters,germination and survival rates were very low. However, the germination and survival rates weobserved are typical for natural regeneration in dry interior Douglas-fir forests (Huggard et al.2005). In studies examining regeneration from seed in the IDF zone, Teste et al. (2009) foundseedling survival rates under 40% and Bingham and Simard (2012) found approximately 50%survival of nursery-grown container stock. Success of natural regeneration is highly variable andconsiderably lower than artificial regeneration (Simard 2009).  Moreover, regeneration success isdramatically reduced with climatic aridity and extreme climate events. As a result, any92advantaged gained by kin recognition or mycorrhizal networks could make a difference inoverall seedling establishment in the field. This could allow for regeneration improvements,particularly in regions of high drought stress.Kin effectsKin recognition, but not kin selection, was weakly evident as increased probability ofsurvival in kin compared to non-kin experimental units (Table 4.6). The odds ratio showed thatnon-kin had slightly greater predicted survival than kin (Figure 4.1b); however, this does notnecessarily provide evidence of either niche partitioning or non-kin facilitation (sensu plantdefense hypothesis). By consistently using only the seed from the original 15 parent trees, eitheras kin or non-kin depending on where it the seed was sown, we were able to compare thesurvival and germination across sites. Site and seed origin had a greater influence on survivalthan did relationship. Most survival, regardless of relationship, occurred at the Paska Lake site,and because more non-kin than kin surviving seedlings were present at that site, the results wereskewed. Survival rate of kin compared to non-kin, as opposed to the total number of survivors, inour data also contradicts this result. If we take into account higher non-kin germination at thePaska Lake site, kin survival rate (not total number) is actually greater than non-kin (Table 4.8,4.9). We acknowledge that using the top seed producers may have impacted unmeasured factors,such as seed size, particularly if those trees were producing stress crops, so these results must beinterpreted with caution. However, due to the constraints of the experiment, it was necessary touse the top seed producers to acquire enough seed for even the small sample size that we had.93Mycorrhizal network effectsNetwork effects were weakly evident as increased probability of a germinant beingobserved in networked versus non-networked experimental units (Table 4.5). The odds ratio alsoshowed that seeds with access to a mycorrhizal network had greater predicted germination(Figure 4.1a). How mycorrhizal networks may affect germination is unclear because colonizationis known to occur in these forests only after several months (Barker et al. 2013).  Furtherresearch is needed to determine whether mycorrhizal networks are involved in biochemicalsignaling to germinating seeds. We expected to see a difference in survival given that theseedlings could access a mycorrhizal network and therefore nutrients that may otherwise beunavailable with only uncolonized roots. Greater survival of seedlings accessing mycorrhizalnetworks has been observed in other studies in IDF forests (Teste et al. 2009, Bingham andSimard 2011). Our hypothesis that kin recognition is facilitated by mycorrhizal networks,however, was not supported. There were no significant kin x network interaction effects in thelogistic regression. In addition, no network effects were observed when relationship effects wereobserved (Table 4.6) and no relationship effects were observed when network effects wereobserved (Table 4.5).Site and seed origin effectsSeed origin and site had the greatest effect on both germination and survival. Thesouthern-most site, Paska Lake, had the greatest predicted presence of germinants and survivors.Interestingly, the seeds that originated from Farwell Canyon, regardless of sowing site, hadsignificantly higher germination rate than any other seed origin region. Seedlings originatingfrom Farwell Canyon also had the highest number of survivors, but this did not affect survival94rate because of the high germination and high mortality rates (kin and no-kin). As noted inchapter 3, seeds originating from Farwell Canyon and the seedlings they produce may be betteradapted to deal with harsh climatic conditions (extreme temperatures and drought stress, Tables4.1-3) than seeds and seedlings from the other sites. Essentially, Farwell Canyon seeds cangerminate in the other sites, but the other seeds cannot germinate at Farwell Canyon. Thissupports our hypothesis that with greater environmental stress, kin recognition and kin selectionbecomes more important. It also supports, possibly to a greater extent, that benefit (or detriment)of relation to a parent tree is dwarfed by more significant processes such as regional geneticadaptations.Climate effectsClimate had interesting effects on both germination and survival. As expected, survivalwas low in the hottest, driest location. However, it was also low at the coolest, wettest location.This pattern was consistent in kin, non-kin, networked and isolated seedlings. The intermediatesite, Paska Lake, was also at the highest elevation. It was the southernmost location and furtheraway from the range limit for interior Douglas-fir; therefore, there may be other sitecharacteristics, such as soil type or availability of soil nutrients, not considered here, thatpromote greater germination and survival rate. Relation and networking appeared to have agreater influence on germination than survival across the range of climate represented by oursites. As drought and heat increased (higher soil T:M index), germination of kin seeds increased.Again, this may have more to do with regional genetic adaption to the site than actual relation tothe parent tree, and an additional study including same site non-kin (different parent tree at thesame site) is needed to test this hypothesis. Seeds with access to a network tended to germinate at95a higher rate in the two wetter sites (lower soil T:M index), but these did not differ significantlythan the hottest, driest site (Farwell Canyon). This is contrary to the stress gradient hypothesis;however, this may be irrelevant to germination because mycorrhizal fungi require four month-old seedling roots for colonization (Barker et al. 2013).ConclusionRelationship and network effects were weakly evident in survivorship and germination,respectively. These effects were slight in comparison to the effect of site on these variables. Siteand seed origin were the major determinants of germination and survival, and whether or not thegerminant or survivor was kin in relation to the parent tree. Climate interacted with relationshipand network factors to affect isolated germinants and survivors in interesting patterns that did notfollow a strict stress gradient.The site and seed origin results have interesting management implications. Seedlingsgrown in nurseries for outplanting originate from widely diverse seed sources within a seed zone,resulting in considerable variability.  Our results suggest this is a good strategy for productivesites, as seedlings from both favourable and harsh growing conditions are able to persist in aproductive site. However, sites with harsh growing conditions may benefit from partialharvesting and natural regeneration to ensure a local seed source. Leaving mature trees to act asseed sources, for kin seedlings as well as seedlings with the appropriate genetic adaptations tothe region, may facilitate natural regeneration, particularly under harsh conditions. Leave treeswould provide shelter from extreme microclimatic conditions and act as reservoirs formycorrhizal fungi. Due to the low overall germination and survival rates in the IDF zone,96providing seedlings any advantage toward establishment could make a large difference inregeneration success.97Table 4.1. Geographic location and estimates of climatic variables for each of the study sites(Alex Fraser Research Forest, Farwell Canyon and Paska Lake) obtained from ClimateBC(Wang et al. 2012).Site Mean Annual SiteCoordinatesElevation(m)Temperature(?C)Precipitaion(mm)Temperaturerange (?C)Alex FraserResearch Forest121.88?W,52.05?N 862 4.8 470 -10.0 to 22.6Farwell Canyon122.63?W,51.79?N 1184 3.5 402 -12.3 to 21.1Paska Lake120.67?W,50.50?N 1435 3.5 411 -9.4 to 19.398Table 4.2. Climate data for each study site measured during the experimental period (May-August 2012) including maximum, minimum and average ambient air temperature, soiltemperature and soil moisture. Ambient air temperature was measures at ground level, soiltemperature was measured at a depth of approximately 20 cm and soil moisture was measuredwith 10 cm of the soil surface.Site Ambient air temperature (?C) Soil Temperature (?C) Soil moisture (m3.m3)Max Min Ave Max Min Ave Max Min AveAlex FraserResearch Forest 46.5 0.23 17.9 21.8 11.8 15.9 0.599 0.001 0.195Farwell Canyon 59.1 -1.00 22.3 25.7 9.80 16.2 0.474 0.001 0.153Paska Lake 44.5 0.01 16.3 25.6 8.10 16.2 0.489 0.001 0.17299Table 4.3. Climate variables for each study site derived from data measured during theexperimental period (May-August 2012) including soil temperature:moisture index, degree days4?C as well as the date range in which those degree days occurred. Soil temperature:moistureindex was calculated as soil temperature (?C) divided by soil moisture (m3per m3).Site Degree days below Date range of degree days below Soil temperature:moisture4?C 4?C indexAlex Fraser ResearchForest 10 06/26-08/23 81.5Farwell Canyon 10 06/30-08/23 105.9Paska Lake 8 06/18-07/04 94.2100Table 4.4. Germination and survival by total number (#) and percentage (%) for the main effectsof network and relationship (kin and non-kin) measured in # or % of treatment units (360 total)and # or % of total seeds sown (1800 total) or germinated (for survival).Germination Survival# % # %Network units 55 30.6% 13 23.6%No network units 41 22.8% 13 31.7%Kin units 45 25.0% 9 20.0%Non-kin units 51 28.3% 17 33.3%Total units 96 26.7% 26 27.1%Network seedlings 88 9.7% 17 19.3%No network seedlings 69 7.7% 23 33.3%Kin seedlings 80 8.9% 14 17.5%Non-kin seedlings 77 8.6% 26 33.7%Total seedlings 157 8.7% 40 25.5%101Table 4.5. Logistic regression testing for the seed germination probability in response torelationship, mesh treatment (network), the combination of relationship and mesh as well as site.Odds ratios are given for each category. P values are given for the effect as a whole to test thenull hypothesis that the odds ratio is equal to one. * p < 0.05.Logistic regression: c = 0.642 Likelihood ratio P = 0.0007Effect Odds ratios df Wald ?2 P > ?2Kin 0.87 1 0.3774 0.5390Non-kin 1.15______________________0.5 ?m mesh 0.65 1 3.3457 0.067435 ?m mesh 1.54______________________Paska Lake site 2.58 2 15.4899 0.0004*Farwell Canyon site 0.73Alex Fraser Research Forest site 0.48102Table 4.6. Logistic regression testing for the probability that a surviving seedling will be presentin response to relationship, mesh treatment (network), the combination of relationship and meshas well as site. Odds ratios are given for each category. P values are given for the effect as awhole to test the null hypothesis that the odds ratio is equal to one. * p < 0.05.Logistic regression: c = 0.776 Likelihood ratio P < 0.0001Effect Odds ratios df Wald ?2 P > ?2Kin 0.50 1 2.7709 0.0960Non-kin 1.980.5 ?m mesh 1.00 1 0.0138 0.906435 ?m mesh 1.00Kin x 0.5 ?m mesh 0.70 1 0.1935 0.6600Kin x 35 ?m mesh 0.52Non-kin x 0.5 ?m mesh 1.37Non-kin x 35 ?m mesh 1.65Paska Lake site 7.80 2 18.0632 0.0001*Farwell Canyon site 0.15Alex Fraser Research Forest site 0.34103Table 4.7. Germination and Survival by total number (#) and percentage (%) for each site and byseed origin location measured in # or % of treatment units (360 total) and # or % of total seedssown (1800 total) or germinated (for survival).Germination Survival# % # %Farwell Canyon site 53 8.8% 3 5.7%Alex Fraser Research Forest site 23 3.8% 5 21.7%Paska Lake site 79 13.2% 32 40.5%Farwell Canyon origin 100 16.7% 19 19.0%Alex Fraser Research Forest origin 28 4.7% 9 32.1%Paska Lake origin 29 4.8% 12 41.4%104Table 4.8. Total number of seeds that germinated in each site (columns) and which site the germinated seedoriginated from (rows). Kin germinants (the seed originated and germinanted in the same site, around the sameparent tree it came from) are shaded and located along the diagonal.  Non-kin germinants (seed origin and site ofgermination differed) are located off-diagonal. There were equal numbers of kin and non-kin sown in each site,however, the non-kin seed origin was split between the two other sites.Farwell Canyon site Alex Fraser Research Forest site Paska Lake site TotalFarwell Canyon origin 46 10 44 100Alex Fraser Research Forest origin 5 11 12 28Paska Lake origin 2 4 23 29Total 53 25 79 157105Table 4.9. Total number of seedlings that survived in each site (columns) and site of seed origin of the survivingseedling (rows). Kin survivors (the seed originated and the seedling survived in the same site, around the sameparent tree it came from) are shaded and located along the diagonal.  Non-kin survivors (seed origin and site ofsurvival differed) are located off-diagonal.Farwell Canyon site Alex Fraser Research Forest site Paska Lake site TotalFarwell Canyon origin 2 3 14 19Alex Fraser Research Forest origin 1 1 7 9Paska Lake origin 0 1 11 12Total 3 5 32 40Figure 4.1. Odds ratio values, presented on a logarithmic scale, for relationship, network, site andrelationship x network effects in the logistic regression model predicting (a) germination (b)survival.0.1110b0.1110Relation  Network  Site  Relation x NetworkaKin                               No network Paska Lake                    Kin/no networkNon-kin                        Network                  Farwell Canyon            Kin/networkAlex Fraser Non-kin/no networkNon-kin/network107a                                                                           bc                                                                           dFigure 4.2. The number of (a) kin and non-kin germinants, (b) kin and non-kin survivors, (c)network and no network germinants and (d) network and no network survivors across the rangeof soil temperature:moisture index values found in the study sites.010203040506075 85 95 105Number of germinants051015202575 85 95 105Number of survivors0102030405075 85 95 105Number of germinantsSoil temperature:moisture index0510152075 85 95 105Number of survivorsSoil temperature:moisture indexKinNon-kinNetworkNo network108Figure 4.3. The number of (a) kin and non-kin germinants, (b) kin and non-kin survivors, (c)network and no netwrok germinants and (d) network and no network survivors across the rangeof averaged soil moisture content found in the study sites.01020304050600.14 0.16 0.18 0.2Number of germinants05101520250.14 0.16 0.18 0.2Number or survivors010203040500.14 0.16 0.18 0.2Number of germinantsSoil moisture content (m3/m3)051015200.14 0.16 0.18 0.2Number of survivorsSoil moisture content (m3/m3)KinNon-kinNetworkNo Network1095 Summary and conclusionsThe regenerative capacity of interior Douglas-fir in harsh climates has been a concern offorest ecology researchers and managers alike. Insight into influences on successful seedlingestablishment could be essential to future management decisions as climate changes. Interplantcommunication has recently generated considerable interest and research results, includingevidence of kin recognition and, in some cases, kin selection (File et al 2011). Whether kinrecognition occurs and has influence on seedling success in interior Douglas-fir is a new andexciting area of research. While the mechanism of kin recognition is still not well understood, wehave provided evidence of kin recognition in interior Douglas-fir seedlings, particularly thosethat originate from regions of harsh climate, and have observed indirect indicators of kinselection or reduction of competition due to a close genetic relationship.Review of objectivesThere were three main objectives that were addressed in each research chapter. The firstwas to determine whether kin recognition is detectable in interior Douglas-fir seedlings. Thesecond was to determine whether kin recognition, if present, would present in a way supportingkin selection theory. The third was to determine whether mycorrhizal networks mediated kinrecognition between seedling pairs (chapters 2 and 3) or between seedlings and parent trees(chapter 4).The Chapter 2 minor objective was to determine if kin recognition ability varied amongdistinct genotypes or ?families? of seedlings (chapter 2).110The Chapter 3 minor objectives were to determine if the region of seed origin affects kinrecognition among seedlings grown in a common greenhouse environment and to determine ifkin recognition occurs along a gradient of relatedness.The Chapter 4 minor objective was to determine if the region of seed origin affects kinrecognition among seedlings grown in the field with a variety of growing conditions (differentsites) (chapter 4).The minor objective comparing Chapter 2 and 3 was to determine whether full sibling kinpairs from control cross pollination exhibited differing kin recognition effects than did kin pairscollected from parent trees in the field with natural pollination.The minor objective comparing Chapters 2 and 3 to Chapter 4 was to determine whethereffects seen in the controlled environment of the greenhouse would be detectable under naturalclimatic conditions in the field.Summary of main findingsMajor objectivesObjective 1 ? Kin recognitionChapter 2 provided evidence of kin recognition by significant differences found in somefoliar micronutrients (Fe, Mo, and Al) according to the relationship, either kin or non-kin,between the seedling pairs. Evidence of kin recognition was found in foliar micronutrients,seedling growth and mycorrhizal colonization in Chapter 3. Kin recognition was weakly evidentin Chapter 4 by difference in probability that a survivor would be present in a kin experimentalunit compared to a non-kin one.111Objective 2 ? Kin selectionOverall the evidence of kin selection was very weak but we were able to detect somesignificant effects that could be interpreted as kin selection. The significant differences inmicronutrients between kin and non-kin seedlings observed in Chapter 2 and 3 suggest kinselection occurred, as all nutrients that differed according to relationship were greater in kincompared to non-kin seedlings. Greater mycorrhizal colonization of kin compared to non-kinseedlings as well as greater donor total leaf area, volume and stem length was observed inChapter 3. No evidence of kin selection was evident in the growth of recipient seedlings. InChapter 4, greater predicted presence of a survivor occurred in non-kin experimental units;however, these results must be considered cautiously because of the significantly greaterinfluence of site and seed origin on germination and survival in the field.Objective 3 ? Mediation by mycorrhizal networksBoth donor and recipient seedlings were colonized predominantly by a singlemycorrhizal fungus, Rhizopogon vincolor, providing sufficient evidence that mycorrhizalnetworks formed between seedling pairs (Chapter 2 and 3). More definitive evidence for thepresence of functional networks was greater transfer of labelled 13C between networkedseedlings than isolated pairs (Chapter 2), although this did not result in enhanced performance bynetworked seedlings. There were significant differences among the four treatment combinationsinvolving relationship and networks.  When non-kin seedlings were grown in isolating meshbags, donor seedlings exhibited enhanced competition toward recipient seedlings. This did notoccur with kin seedlings grown with networks, providing evidence that kin recognition could befacilitated by mycorrhizal networks (Chapter 2). Further evidence of facilitation was found in112Chapter 3, where foliar copper and iron concentrations were significantly higher in kin seedlingsgrown with networks than without. In the field collected sibling seedlings (Chapter 3), theenhanced mycorrhizal colonization of recipient seedlings could only occur through linkage intomycorrhizal networks of donor seedlings. It is unclear if the greater mycorrhizal colonization isevidence for mediation of kin recognition through mycorrhizal networks or is a response due tokin recognition. In Chapter 4, network effects were weakly evident as greater probability that agerminant would be present in a networked experimental unit than an isolated one. In the fieldexperiment, however, kin recognition did not appear to be facilitated by mycorrhizal networks asno interaction effects occurred in the logistic regression. In addition, no network effects wereobserved when relationship effects occurred and no relationship effects were observed whennetwork effects occurred.Minor objectivesObjective 1 ? Effect of distinct genotypes (families) on kin recognitionChapter 2 was designed to examine the effect of distinct genotypes, or ?families? on kinrecognition. The four distinct families did not differ in growth traits or foliar nutrient content,however, which may explain why we were unable to detect a genotype effect on kin recognition.This may have simply been a result of the lack of statistical power. It is also possible these seedswere too similar in competitive ability to influence the growth traits we examined.Objective 2 ? Effect of the region of seed origin on kin recognitionChapter 3 was designed to examine the effect of seed origin on kin recognition whenseedlings were grown in common, favourable conditions (greenhouse setting). Seed origin had asignificant effect on germination rate among field collected, greenhouse grown recipient113seedlings. Seeds originating from Farwell Canyon had the greatest germination rate amongregions. Seed origin also had an effect on kin recognition in terms of germination rate, as kinoriginating from Farwell Canyon had significantly higher germination rates than non-kin fromFarwell Canyon or any other region tested. Mycorrhizal effects on germination are consideredminimal but more research is needed before appropriate conclusions can be drawn from thisresult. Farwell Canyon had the driest and hottest climate and kin recognition and selection (i.e.,cooperation among genetically similar individuals) may be more necessary in harsh climates inaccordance with the stress gradient hypothesis.Objective 3 ? Effect of the level of relatedness on kin recognitionChapter 3 was designed to examine the effect of the level of relatedness on kinrecognition. Kin recognition was not affected by the gradient of relatedness we tested. Kinrecognition was only detectable at the sibling/ non-sibling level; by contrast, whether the non-sibling originated from a different parent tree within the same site (within population) or adifferent site (among populations) did not affect any variables with relationship effects, but therewas little statistical power to test these effects.Objective 4 ? Effect of growing conditions (site) and the region of seed origin on kin recognitionChapter 4 was designed to examine the effect of site, or growing conditions due tovariable climate, and seed origin, on kin recognition in a field setting. Site and seed origin bothhad a significant effect on germination. The greatest number of seeds germinated in thesouthernmost site, Paska Lake. The greatest number of germinated seeds originated from the sitethat had the harshest growing conditions (most extreme air temperatures, highest soiltemperature:moisture index and lowest soil moisture content) during the experimental period,114Farwell Canyon. Site had a significant effect on presence of survivors, with the most survival atPaska Lake. Seed origin did not have a significant effect on presence of survivors. The largesignificant effects of site and seed origin we observed on the distribution of kin and non-kingermination and survival may have overshadowed potential relationship effects in the field.Objective 5 ? Kin recognition in greenhouse grown full sibling vs. field collected sibling pairsDespite the fact that control cross pollinated seeds yield full sibling pairs, more evidenceof kin recognition was detectable in the field collected seedling pairs. This suggests that theadaptations of seeds originating in different climatic regions are more important to kinrecognition than the closer relationship of full siblings originating in similar climatic conditions(all Chapter 2 seeds obtained from the Kalamalka Research station).Objective 6 ? Broad comparison of kin recognition and mycorrhizal networks in a greenhouse vs.a field settingMore evidence of kin recognition was detectable in the greenhouse experimentscompared to the field study. At this point, it is difficult to compare the effect of the setting asstated in the objective as low survival rates prevented us from harvesting the field seedlings andtherefore comparing all of the same variables. We did see a trend toward kin recognition in thepresence of germinants in the field; therefore, we suggest that kin recognition is not aphenomenon only detectable in highly controlled, greenhouse growing conditions. The strongestevidence that mycorrhizal networks are involved in kin recognition was through increasedmycorrhizal colonization of kin versus non-kin in the greenhouse (Chapter 3). It is still unclearfrom this evidence whether mycorrhizal networks have a facilitative role in kin recognition or theenhanced mycorrhizal colonization was a response to kin recognition.115Contributions to the field of studyTo our knowledge, this is the first study that examined kin recognition in any coniferoustree species. In a species such as interior Douglas-fir, where there is large variation in climaticconditions across its range and areas of very low regenerative capacity, small advantages toregenerating seedling could prove important to regeneration success. This study contributedevidence that kin recognition occurs among interior Douglas-fir seedlings, although it has veryminor effects on the seedlings compared to regional, climatic and other factors. There have beenmany studies on the effect of mycorrhizal networks on interior Douglas-fir. Our goal was to addto the base of knowledge of the scope of effects mycorrhizal networks have on interior Douglas-fir seedlings. There is evidence that these networks transport water, nutrients and defense signalsbetween conspecific individuals connected by mycorrhizal hyphae. We wanted to determinewhether mycorrhizal networks and resource transfer also play a role in facilitating kinrecognition. We found increased mycorrhizal colonization of kin compared with non-kinseedlings in the field collected, greenhouse grown sibling greenhouse experiment, which couldonly have occurred through mycorrhizal network formation, suggestive of kin recognition.Otherwise, there was little evidence that mycorrhizal networks facilitated kin selection in interiorDouglas-fir.The results of this study may have implications for management practices that encouragereproduction of kin seedlings near their parents, including the retention of healthy, cone-bearinglegacy trees during harvest to supply seed as well as act as refuges (shelter, nutrients andmycorrhizal inoculum) for natural regeneration. However, more research into kin selection in thefield is necessary before these practices could be implemented. These practices may be116particularly effective in areas where natural regeneration and productivity are low, either due tomicronutrient deficiencies or drought.Limitations of studiesThe minor objectives included in both greenhouse experiments (Chapter 2 and 3) sufferedfrom small sample size. This was a major limitation both in the greenhouse, but particularly inthe field study as we were constrained by the number of seeds produced by the parent trees at thetime of collection. More useful information could have also been gathered at the time ofcollection, such as seed size and weight, which may have proved useful in explaining the highvariation in field results. Weather and wildlife also proved to be factors that led to small samplesizes in the field.Time availability was also a major limitation to these studies. Had these studies beenconducted over a longer term, I believe both minor and major objectives could have been testedwith much more clarity. More growing time would have allowed for more growth and nutrientdifferentiation between treatments in the greenhouse.  Moreover, greater sample size in the fieldstudy would have increased our power to test our hypotheses.Future directionsMore information can be gained from the continuation of the field study to a point wheresurvival is sufficient for harvest and evaluation of all variables. Studies looking into equal siblingcompared to non-sibling pairs (seeds sown at the same time) could help to parse out the effectsof relationship, network and size discrepancies in a greenhouse study. In the field, studies117examining differential effects of kin recognition and regional adaptation to climate would helpdistinguish whether kin recognition truly provides an advantage to regeneration in harsh relativeto favourable climatic conditions.118ReferencesAgerer R. (2001) Exploration types of ectomycorrhizae: A proposal to classify ectomycorrhizalmycelial systems according to their patterns of differentiation and putative ecological importance.Mycorrhiza 11 pp.107?114Barker JS., Simard SW., Jones MD., Durall DM. (2013) Ectomycorrhizal fungal communityassembly on regenerating Douglas-fir after wildfire and clearcut harvesting. Oecologia. 172:4,pp 1179-1189Barto EK., Hilker M., Muller F., Mohney BK., Weidenhamer JD., Rillig MC. (2011) TheFungal Fast Lane: Common Mycorrhizal Networks Extend Bioactive Zones of Allelochemicalsin Soils. PLoS ONE 6:11 pp.880-882 doi:10.1371/journal.pone.0027195British Columbia Ministry of Forests (2012) Annual report. British Columbia Ministry ofForests, Lands and Natural Resource Operations, Victoria, BC.Beiler KJ., Durall DM., Simard SW., Maxwell SA., Kretzer AM. (2010) Mapping the wood-wide web: mycorrhizal networks link multiple Douglas-fir cohorts New Phytologist, 185 pp. 543-553.Beiler KJ., Simard SW., Lemay V., Durall DM. (2012) Vertical partitioning between sisterspecies of Rhizopogon fungi on mesic and xeric sites in an interior Douglas-fir forest, MolecularEcology doi: 10.1111/mec.12076Bertness M., Callaway RM. (1994) Positive interactions in communities. Trends in Ecologyand Evolution 9 pp. 191?193Biedrzycki ML., Jilany TA., Dudley SA., Bais HP. (2010). Root exudates mediate kinrecognition in plants. Communicative & Integrative Biology 3:1 pp.28-35Bingham MA., Simard SW. (2011) Do mycorrhizal network benefits to survival and growth ofinterior Douglas-fir seedlings increase with soil moisture stress? Ecology and Evolution 1 pp.306-316.Bingham MA., Simard SW. (2012) Ectomycorrhizal networks of old Pseudotsuga menziesii var.glauca trees facilitate establishment of conspecific seedlings under drought. Ecosystems15 pp.188-199 doi: 10.1007/s10021-011-9502-2Booth MG., Hoeksema JD., (2010) Ectomycorrhizal networks counteract competitive effects ofcanopy trees on seedling survival. Ecology 91 pp. 2294-2302.Boyden S., Binkley D., Stape JL. (2008) Competition among Eucalyptus trees depends ongenetic variation and resource supply. Ecology. 89 pp. 2850?2859. doi:10.1890/07-1733.1119Brooks R., Meinzer FC., Warren JM., Domec J-C., Coulombe R. (2006) Hydraulicredistribution in a Douglas-fir forest: lessons from system manipulations. Plant Cell Environ. 29pp. 138-150.Bruin J., Sabelis MW. (2001) Meta-analysis of laboratory experiments on plant?plantinformation transfer Biochemical Systematics and Ecology 29:10 pp.1089?1102Callaway RM., Brooker RW., Choler P., Kikvidze Z., Lortie CJ., Michalet R., Paolini L.,Pugnaire FI., Newingham B., Aschehoug ET., Armas C., Kikodze D., Cook BJ. (2002)Positive interactions among alpine plants increase with stress. Nature. 417 pp.844?848.Campbell RK. (1986) Mapped genetic variation of Douglas-fir to guide seed transfer insouthwest Oregon. Silvae Genet. 35 pp.85-96Campbell RK., Sorensen FC. (1978) Effects of test environment on expression of clines and ondelimitation of seed zones in Douglas-fir. Theor. Appl. Genet. 51 pp.233-246Cavieres LA., Badano EI., Sierra-Almeida A., Gomez-Gonzalez S., Molina-MontenegroMA. (2006) Positive interactions between alpine plant species and the nurse cushion plantLaretia acaulis do not increase with altitude in the Andes of central Chile. New Phytologist. 169pp. 59?69.Chen BJW., During HJ., Anten NPR. (2012) Detect thy neighbor: Identity recognition at theroot level in plants. Plant Science. 195 pp.157? 167Cheplick GP.,  Kane KH. (2004) Genetic relatedness and competition in Triplasis purpurea(Poaceae): resource partitioning or kin selection? Int. J. Plant Sci. 165 pp. 623?630.doi:10.1086/386556Deslippe JR., Simard SW. (2011) Below-ground carbon transfer among Betula nana mayincrease with warming in Arctic tundra. New Phytologist, 192:3 pp.689-698.Dickie IA., Schnitzer SA., Reich PB., Hobbie SE. (2005). Spatially disjunct effects of co-occurring competition and facilitation. Ecology Letters 8 pp.1191-1200.Donohue K. (2003) The influence of neighbor relatedness on multilevel selection in the GreatLakes sea rocket. American Naturalist 162 pp. 77?92.Dudley SA., File AL., (2007) Kin recognition in an annual plant. Biology Letters. 3, pp.435?438.(doi:10.1098/rsbl.2007.0232)File AL., Klironomos J., Maherali H., Dudley SA. (2012) Plant kin recognition enhancesabundance of symbiotic microbial partner. PLoS ONE 7:9 e45648 doi:10.1371/journal.pone.0045648120File AL., Murphy GP., Dudley SA. (2011) Fitness consequences of plants growing withsiblings: reconciling kin selection, niche partitioning and competitive ability. Proc R Soc B doi:10.1098/rspb.2011.1995 pp.1-10Fortin JA., Plenchette C., Piche Y. (2009) Mycorrhizas: the new green revolution. ?ditionsMultiMondesFrancis R., Read DJ. (1984) Direct transfer of carbon between plants connected by vesicular?arbuscular mycorrhizal mycelium. Nature 307 pp. 53 - 56 doi:10.1038/307053a0Giovannetti M, Mosse B. (1998) An evaluation of techniques for measuring vesicular-arbuscular mycorrhizal infection in roots. New Phytologist 84 pp.489-500.Greenlee JT., Callaway RM. (1996) Abiotic stress and the relative importance of interferenceand facilitation in montane bunchgrass communities in western Montana. American Naturalist.148 pp. 386?396.Guerinot ML., Yi Y. (1994) Iron: Nutritious, noxious, and not readily available. PlantPhysiology, 104, pp. 815?820Gugger PF., Sugita, S., Cavender-Bares, J. (2010) Phylogeography of Douglas-fir based onmitochondrial and chloroplast DNA sequences: testing hypotheses from the fossil record.Molecular Ecology 19, pp. 1877?1897 doi: 10.1111/j.1365-294X.2010.04622.xHamann A., Gylander T., Chen P. (2011) Developing seed zones and transfer guidelines withmultivariate regression trees. Tree Genetics & Genomes. 7 pp. 399?408 doi:10.1007/s11295-010-0341-7Hamilton WD. (1964) The genetical evolution of social behaviour. I. J. Theor. Biol. 7, 1-16Hamrick JL., Godt MJW., Sherman-Broyles SL. (1992) Factors influencing levels of geneticdiversity in woody plant species, New Forests 6 pp.95-124Hermann RK., Lavender DP. (1990). Pseudostuga menziesii (Mirb.) Franco. In: Burns, R.M.,Honkola, B.H. (Eds.), Silvics of North America, Volume 1, Conifers. USDA Forest Service,Agricultural Handbook 654, pp. 527?540.Hirayama T., Kieber JJ., Hirayama N., Kogan M., Guzman P., Nourizadeh S., Alonso JM.,Dailey WP., Dancis A., Ecker J R. (1999). RESPONSIVE-TO-ANTAGONIST1, aMenkes/Wilson Disease?Related Copper Transporter, Is Required for Ethylene Signaling in< i>Arabidopsis</i>. Cell. 97:3 pp. 383-393.Horton TR., Bruns TD. (2001) The molecular revolution in ectomycorrhizal ecology: peekinginto the black-box. Molecular Ecology 10 pp.1855-1871.121Huggard DJ., Arsenault A., Vyse A., Klenner W. (2005) The Opax Mountain SilviculturalSystems Project: Preliminary Results for Managing Complex, Dry Interior Douglas-fir Forests.B.C. Ministry of Forests and Range, Forest Sciences Program, Victoria, BC. Extension Note 72.Keeton, WS. and Franklin JF. (2005) Do remnant old-growth trees accelerate rates ofsuccession in mature Douglas-fir forests? Ecological Monographs 75:1 pp. 103-118 doi:10.1890/03-0626Kessler A., Halitschke R., Diezel C., Baldwin IT. (2006) Priming of plant defense responses innature by airborne signaling between Artemisia tridentata and Nicotiana attenuata. Oecologia148 pp. 280?292. doi: 10.1007/s00442-006-0365-8.Kiers ET., Franken O., Hart MM., Duhamel M., Verbruggen E., Bago A.,Vandenkoornhuyse P., Beesetty Y., Fellbaum CR., Palmer TM., Jansa J., Mensah JA.,Kowalchuk GA., West SA., B?cking H. (2011) Reciprocal Rewards Stabilize Cooperation inthe Mycorrhizal Symbiosis. Science 333:880 doi: 10.1126/science.1208473Klenner W., Walton R., Arsenault A., Kremsater L. (2008) Dry forests of the southerninterior of British Columbia: historic disturbances and implications for restoration andmanagement. Forest Ecology and Management 256, pp. 1711?1722.Kremer A. (1994) Genetic diversity and phenotypic variability of forest trees. GeneticsSelection Evolution. 26:1 pp. S105-S123 doi: 10.1051/gse:19940708Kretzer AM., Luoma DL., Molina R., Spatafora JW. (2003) Taxonomy of the Rhizopogonvinicolor species complex based on analysis of its sequences and microsatellite loci. Mycologia95 pp.480?487.Krutovsky, KV., St. Clair, JB., Saich, R., Hipkins, VD., Neale, DB. (2009) Estimation ofpopulation structure in coastal Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco var. menziesii]using allozyme and microsatellite markers, Tree Genetics & Genomes, 5:4 pp.641-658, doi:10.1007/s11295-009-0216-yLavender DP., Parish R., Johnson CM., Montgomery G., Vyse A., Willis RA., Winston D.(1990) Regenerating British Columbia's forests. University of British Columbia Press,Vancouver 1990Li X., Marschner H., George E. (1991) Acquisition of phosphorus and copper by VA-mycorrhizal hyphae and root-to-shoot transport in white clover. Plant and Soil, 136:1, pp 49-57Liancourt P., Callaway RM., Michalet R. (2005) Stress tolerance and competitive-responseability determine the outcome of biotic interactions. Ecology. 86 pp. 1611?1618.Marschner P., (2012) (1st edition by Marschner H., 1986) Marschner's Mineral Nutrition ofHigher Plants (Third Edition) Copyright ? 2012 Elsevier Ltd. All rights reserved122Margulis, L. (1981) Symbiosis in cell evolution: Life and its environment on the early earth.p.419.Massicotte HB., Molina R., Luoma DL., Smith JE. (1994) Biology of the ectomycorrhizalgenus, Rhizopogon. II Patterns of host-fungus specificity following spore inoculation of diversehosts grown in monoculture and dual culture. New Phytologist 126:4 pp. 677-690McGuire KL., (2007) Common ectomycorrhizal networks may maintain mono-dominance in atropical rain forest. Ecology. 88, pp. 567-574.Mengel K., Kirkby EA. (2001) Principles of Plant Nutrition. ? 2001 Kluwer AcademicPublishersMilla R., Forero DM., Escudero A., Iriondo JM. (2009) Growing with siblings: a commonground for cooperation or for fiercer competition among plants? Proc. R. Soc. B. 276 pp.2531-2540. doi:10.1098/rspb.2009.0369Molina R, Massicotte H, Trappe JM (1992) Specificity phenomenon in mycorrhizalsymbiosis: community-ecological consequences and practical implications. In Allen MF (ed)Mycorrhizal Functioning: An Integrative Plant-Fungal Process. Chapman Hall, New York. pp357-423Molina R., Trappe JM., Grubisha LC., Spatafora JW. (1999) Rhizopogon. In: Cairney, J.,Chambers, S.M. (Eds.), Ectomycorrhizal Fungi: Key Genera in Profile. Springer Verlag, Berlin.Mulder EG., Boxma R., Van Veen WL. (1959) The effect of molybdenum and nitrogendeficiencies on nitrate reduction in plant tissues. Plant and Soil 10:4 pp.335-355Nara K. (2006) Ectomycorrhizal networks and seedling establishment during early primarysuccession. New Phytologist 169 pp. 169-178.Newsome TA., Sutherland DC., Vyse A. (1991) Establishing Douglas-fir plantations in the drybelt of interior British Columbia. In: Baumgartner, D.M., Lotan, J.E. (Eds.), Interior Douglas-fir:The Species and its Management. Washington State University, Cooperative Extension Service,Seattle, WA, pp. 227?233.Onguene NA., Kuyper TW. (2002) Importance of the ectomycorrhizal network for seedlingsurvival and ectomycorrhiza formation in rain forests of south Cameroon. Mycorrhiza 12 pp.13-17.Perry DA., Margolis H., Choquette C., Molina R., Trappe LM. (1989) EctomycorrhizalMediation of Competition between Coniferous Tree Species. New Phytologist. 112:4 pp.501-511Philip LJ. (2006) The role of ectomycorrhizal fungi in carbon transfer within commonmycorrhizal networks. PhD Dissertation, The University of British Columbia 152 pp.123Platt TG., Bever JD. (2009) Kin competition and the evolution of cooperation. Trends inEcology and Evolution 24:7 pp.370-377Querejeta JI., Barea JM., Allen MF., Caravaca F., Rolda?n A. (2003) Differential responseof d13C and water use efficiency to arbuscular mycorrhizal infection in two aridland woodyplant species. Oecologia 135 pp.510?515Rehfeldt GE. (1978) Genetic differentiation of Douglas-fir populations from the NorthernRocky Mountains. Ecology 59 pp.1264-1270.Rehfeldt GE. (1989) Ecological Adaptations in Douglas-Fir (Pseudotsuga menziesii var.glauca): a Synthesis. Forest Ecology and Management 28 pp.203-215Roche L. (1969) A genecological study of the genus Picea in British Columbia. New Phytologist68 pp.505-554.Schoonmaker AL., Teste FP., Simard SW., Guy RD., (2007) Tree proximity, soil pathwaysand common mycorrhizal networks: their influence on utilization of redistributed water byunderstory seedlings. Oecologia 154 pp. 455-466.Selosse MA., Richard F., He X., Simard SW. (2006) Mycorrhizal network: des liaisonsdangereuses? Trends in Ecology and Evolution 21:21 pp.621-628Shulaev V., Silverman P., Raskin I. (1997) Airborne signalling by methyl salicylate in plantpathogen resistance. Nature 385 pp.718?721. doi: 10.1038/385718a0.Simard, SW. (2009) The foundational role of mycorrhizas in the self-organization of the interiorDouglas-fir forests. Forest Ecology & Management 258 pp.S95-S107Simard SW., Beiler KJ., Bingham MA., Deslippe JR., Philip LJ., Teste FP. (2012)Mycorrhizal networks: Mechanisms, ecology and modelling. Fungal Biology Reviews 26 pp. 39-60Simard SW, Jones MD, Durall DM. (2002) Carbon and nutrient fluxes within and betweenmycorrhizal plants. In van der Heijden M, Sanders I (eds) Mycorrhizal Ecology. Springer-Verlag,Heidelberg. Ecological Studies 157 pp.33-61Simard SW., Jones MD., Durall DM., Hope GD., Stathers RJ., Sorensen NS., Zimonick BJ.(2003) Chemical and mechanical site preparation: effects on Pinus contorta growth, physiology,and microsite quality on steep forest sites in British Columbia. Canadian Journal of ForestResearch 33 pp.1495-1515.Simard SW., Martin K., Vyse A., Larson B. (2013) Meta-networks of fungi, fauna and flora asagents of complex adaptive systems Chapter 7, pages 133-164. In: Managing World Forests as124Complex Adaptive Systems: Building Resilience to the Challenge of Global Change. Edited byPuettmann, K, Messier, C, and Coates, KD. Routledge, NY. ISBN 978-0-415-51977. 369 pp.Simard SW., Perry DA., Jones MD., Myrold DD., Durall DM., and Molina R. (1997) Nettransfer of carbon between tree species with shared ectomycorrhizal fungi. Nature, 388 pp.579-582Smith, S.E., Read, D.J. (1997) Mycorrhizal Symbiosis. Copyright ? 1997 Elsevier Ltd. Allrights reservedSong, Y.Y., Zeng, R.S., Xu, J.F., Li, J., Yihdego, W.G. (2010) Interplant Communication ofTomato Plants through Underground Common Mycorrhizal Networks. PloS ONE 5:10 e13324.doi:10.1371/journal .pone.0013324Soon YK. (1995) Forms of extractable aluminium in Canadian acid soils and their relations toplant growth. Developments in Plant and Soil Sciences 64 pp. 65-70Sorensen FC. (1979) Provenance variation in Pseudotsuga menziesii seedlings from the var.menziesii-var. glaucam transition zone in Oregon. Silvae Genetica 28 pp.37?119.Spittlehouse, D. (2006) ClimateBC: Your Access to Interpolated Climate Data for BC.Streamline Watershed Management Bulletin 9:2 pp.16-20.St. Clair JB., Mandel NL., Vance-Borland KW (2005) Genecology of Douglas Fir in WesternOregon and Washington. Annals of Botany. 96:7 pp.1199-1214 doi:10.1093/aob/mci278Teste FP., Karst J., Jones MD., Simard SW., Durall DM. (2006) Methods to controlectomycorrhizal colonization: effectiveness of chemical and physical barriers. Mycorrhiza. 17pp.51-65Teste FP, Simard SW., Durall DM., Guy RD., Jones MD. (2009) Access to mycorrhizalnetworks and roots of trees: importance for seedling survival and resource transfer. Ecology90:10 pp.2808-2822Teste FP, Simard SW., Durall DM., Guy RD., Berch SM. (2010) Net carbon transfer betweenPseudotsuga menziesii var. glauca seedlings in the field is influced be soil disturbance. Journalof Ecology 98 pp.429-439Teste FP, Simard SW. (2008) Mycorrhizal networks and distance from mature trees alterpatterns of competition and facilitation in dry Douglas-fir forests. Oecologia 158 pp.193-203.Tonsor SJ. (1989) Relatedness and intraspecific competition in Plantago lanceolata. TheAmerican Naturalist 134 pp.897?906 doi:10.1086/285020125Twieg BD., Durall DM., Simard SW. (2007) Ectomycorrhizal fungal succession in mixedtemperate forests. New Phytologist 176 pp.437-447.Van der Heijden, MGA., Horton, TR. (2009) Socialism in soil? The importance ofmycorrhizal fungal networks for facilitation in natural ecosystems. Journal of Ecology 97pp.1139-1150Vyse, A., Ferguson, C., Simard, S.W., Kano, T., and Puttonen, P. (2006) Growth of Douglas-fir, lodgepole pine, and ponderosa pine seedlings underplanted in a partially-cut, dry Douglas-firstand in south-central British Columbia. Forestry Chronicle 82 pp.723-732.Wang T., Campbell EM., O?Neill GA., Aitken SN., (2012) Projecting future distributions ofecosystem climate niches: Uncertainties and management applications. Forest Ecology andManagement 279 pp.128?140Wang T., Hamann, A., Spittlehouse, D. (2012) ClimateBC_Map, A Interactice Platform forVisualization and Data Access, Copyright(2012) Wang,T., Hamann, A., Spittlehouse, D. Allright reserved., http://www.genetics.forestry.ubc.ca/cfcg/ClimateBC40/Default.aspx, Annual andSeasonal VariablesWhitfield J. (2007) Underground Networking. Nature 449 pp.136-138Willson MF., Hoppes WG., Goldman DA., Thomas PA., Katusic-Malmborg PL., BothwellJL. (1987) Sibling competition in plants: an experimental study. The American Naturalist 129pp.304?311 doi:10.1086/284636Young, J. P. W. (1981) Sib competition can favour sex in two ways. J. Theor. Biol. 88 pp.755?756. doi:10.1016/0022-5193(81)90249-6

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0103374/manifest

Comment

Related Items