Welcome.  I'm  going  talk to  so  member  of  the  Pal. I'm  delighted  today  to introduce  my  colleague and  great  friend  Christina. I'm  going  to  start of  the  important  information will  be  the  parties  which  the  grafts. One,  the  address  is a  CC  and  anyone's  off  brings. Brit  skater  and  I  have known  each  other  for a  very  long  time  and  I  think a  real  signature  of  Brit's  work is  that  she  probes very  fundamental  questions  and evolution  ecology  by meeting  host  microsystem  systems. I  think  what  Britain  loves mostly  the  designing  experiments, and  they're  always  deceptively simple  but  beautiful  experiments. So  things  like  looking at  time  shift  experiments  of bacteriage  over  the  course of  the  season  and  a  leaf, which  in  retrospect  so  obvious, but  no  one  had  done  it  before  Brit did  until  these  fantastic  patterns. When  you  think  a  little bit  about  microbial  communities, you  realize  that  the  equivalent  of the  varying  straight  language, where  two  enormous  and  different  communities are  bumping  into  each  other, happens  all  the  time. Any  time  of  leaf  drops  on  the  ground, two  communities  are  crashing  into  each  other. But  you  can  interrogate  that using  experimental  systems  and  microbes, plants  or  other  things, this  is  the  token  of  our  incredible  and esteem  the  mystery  mug and  that  print.  Thanks. Thank  you.  It's  a  pleasure  to  be  here. And  really  had  a  wonderful  visit  so  far. I'm  looking  forward  to  meeting many  more  of  you  this  afternoon. And  with  that,  I'll  jump  right in  because  the  focus of  my  talk  today  is  going  to be  thinking  about  microbiomes, which  unless  you've  been  living  in  a  rock, you  know  the  term  by  now. And  some  people  think  of  it  as  a  bandwagon. And  I  can  see  why  it's gotten  that  reputation. But  fundamentally  why  I  love microbiomes  is  that as  evolutionary  biologist, as  an  ecologist,  I  think  it's  given  us a  chance  to  re,  question  everything. Often  the  answer  is  no,  they  don't  matter. But  it's  really  fun  to  think  about  the  fact that  all  plants  and animals  evolved  in  a  microbial  soup. And  everything  that  we  have initially  thought  of  as  a  host  trait, or  genetic  trait,  was likely  impacted  in  some  way, shape,  or  form,  by  the  vast  majority  of commensal  or  even  mutualistic  microbes with  which  a  host  interacts. So  I  have  a  lot  of  fun  teaching an  undergraduate  class  called Reimagining  Biology in  light  of  the  microbiome. I  often  think  about  that  when  I  read  papers. But  of  course,  the  reality  is  that microbes  matter  sometimes  and  other  times, everything  we  thought  was true  is  indeed  true. What  I  want  to  talk  about  today is  factors  that  shape  the  microbiome. And  what  do  I  mean  by  that? Well,  just  like  any  community, it's  going  to  be  affected  by an  abiotic  environment  for  the  microbiome. For  example,  of  our  skin,  we  might think  about  things  like  the  sunscreen that  we  just  applied  or indeed  the  amount  of  UB  exposure  we  get, et  cetera  for  a  plant. We  think  about  things like  water  availability, nutrients  in  the  soil  and  so  forth. But  of  course,  the  other  important selection  factors  that  shape microbial  community  composition is  the  biotic  factors. And  that  includes  importantly  the  host, if  you're  talking  about a  host  associated  microbiome. And  I'll  spend  a  lot  of  time today  thinking  about  how hosts  shape  the  microbome with  which  they  interact. But  other  biotic  interactions  as  well. Including  all  of  the  competition, bacteria, bacteria  competition  within  a  microbiome. And  what  you  maybe  have  heard  less  about, which  is  the  role  that  bacteria, phage,  viruses  play  in  shaping  the  micro  ome. I'm  going  to  go  somewhat quickly  through  the  first  half  of my  talk  with  the  hopes  that  I'm  protecting enough  time  to  do  phages  justice  at  the  end. Fundamentally,  one  of  the  reasons  that I'm  excited  about  understanding these  processes  is  that  it  means  that  we  can start  changing  microbiomes  in a  way  that  might  be  useful. Or  at  least  thinking  about  how  we unintentionally,  unintentionally shape  microbiomes. In  particular, in  agriculture  or  clinical  settings. Okay? Of  course,  in  a  room  full  of people  who  think  about evolution  all  the  time, selection  is  only  as  good as  the  variation  upon  which  it  acts. A  lot  of  what  we  spend time  thinking  about  is  how microbes  get  into  onto  a  host  to  begin  with. Because  until  you  understand  that, you  will  never  be  able  to  predict a  response  to  selection. When  you  look  at  any  microbiome, whether  it's  a  human  microbiome, as  I  show  here, or  a  plant  microbiome, the  general  signature  is the  change  is  the  norm. It's  very  rare  that  you  would  go back  and  sample  a  microbiome  a  month  later, or  even  a  day  later  and  see exactly  the  same  species in  the  same  composition. But  importantly,  what's  shaping these  changes  through  time can  be  both  ecological  species  sort,  right? Just  wholesale  turnover  of  species  based  on what  you  just  ate  or the  antibiotic  you  just  took. But  of  course,  it  can  also  be  evolutionary. Through  true  adaptation. And  we're  just  coming  to  terms  with the  relative  importance  of those  two  factors  in  shaping  the  microbiome. And  for  the  human  microbiome, it's  actually  still  very controversial  how  important  within host  evolution  is  versus just  species  sorting  and  acquisition. But  we  have  any  Muller  in the  audience  who's  been  one  of  the  leaders of  thinking  about  this  question and  what  we  can  learn  about adaptation  by  taking broad  phylogenetic  perspectives. This  is  an  example  from  a  different  system, but  this  is  the  idea  that you  can  look  at  the  phylogeny  of hosts  and  find  that  there  are broad  phylogenetic signatures  of  the  microbiome, which  is  suggestive  that  there  has  been a  co  evolutionary  history or  that  the  microbiome  has  cost, be  seated  with  hosts which  shouldn't  be  a  surprise if  they  are  in  fact  important. One  of  the  ways  they  are  important  is  that they  can  reshape  host  phenotype. And  again,  this  is a  beautiful  paper  by  someone else  in  this  room  roles  who  had  this beautiful  schematic showing  how  a  host  genotype could  result  in  different  phenotypes depending  on  the  microbiome with  which  it  associates. And  of  course,  that  could  be  no  change. In  fact,  for  many  host  traits  we  care  about, that's  going  to  be  the  rule. But  in  other  ways,  you  might have  a  shift  in  a  phenotype based  on  the  different  microbial  composition of  a  microbiome, or  a  widening  of the  niche  space  and  so  forth. This  means  that  in  a  dooms  day, global  change  world, especially  Jess  and  I  spend  a  lot of  time  thinking  about  why  trees  exist. How  long  lived  organisms  could  ever keep  up  with the  rapid  change  that's  happening? One  glimmer  of  hope is  that  maybe  they  don't  have  to, at  least  in  the  short  term, if  the  microbiomes  associated with  them  can  make  up  or compensate  for  some  of  the  adaptation  that's required  in  a  highly  fluctuating  environment. That's  generally  why  I care  about  microbiomes. I  work  not  on  human  microbiomes, although  I  will  come  back  to  them  at  the  end. On  the  phage  side  of  things, I  work  on  the  plant  microbiome. I  don't  mean  this,  but  the majority  of  the  time  you  hear about  the  plant  microbiome, you're  hearing  about below  ground  associations. And  that's  because  below  ground  rises, sphere  microbiome  associations  are critically  important  to  plant  health. I  actually  like  to  think  about the  root  microbiome  as the  equivalent  of  the  human  gut  microbiome. This  is  where  digestion  happens and  so  forth,  nutrient  acquisition. But  in  fact,  in  my  group  work above  ground  on  the  Phils  sphere, microbiome  which  is  relatively understudied  but  equally  important. Aside  from  just  the  impact  on  plant  health, it's  a  huge  amount  of  biomass  on  the  planet, even  though  we  might  not  think  about  it as  that  important  of a  habitat  for  bacteria  in  terms  of  there's not  that  many  nutrients on  a  leaf,  it  turns  out. But  the  sheer  biomass of  leaves  on  this  planet  means  that  it's actually  a  pretty  great  place  to  be a  microbe  that  specialized  in  this  area. As  an  ecologist  though, I  really  love  the  phylosphere. If  anyone's  looking  for  a new  model  as  they're going  into  their  postdoc  or first  acutene  position. If  you're  interested  in  plant  microons, leaves  are  wear  at  that, I'll  tell  you  why  they're spatially  explicit  and  someone  who likes  to  know  ask  questions about  local  adaptation or  population  dynamics. It's  just  a  wonderful  hack. I  don't  need  to  convince  reviewers  that this  is  one  leaf and  this  is  a  different  leaf,  right? Can  we  can  all  agree  on  that, but  I  don't  know  what  that looks  like  below  ground, it's  a  mess  down  there.  All  right? So  it's  also  easy  to  manipulate. And  I'll  show  you  examples of  how  we  can  spray  a  leaf, and  the  microbiome  indeed  colonizes that  leaf  and  changes  through  time. And  we  can  recapitulate a  natural  microbiome  in  the  lab. There's  variation  in  leaves. Of  course,  you  can  look  outside  and see  that  also  because a  leaf  is  exposed  to oxygen  and  a  similar environments  to  a  petri  dish, the  majority  of  microbes  we find  the  leaves  are  indeed  cultural. That's  a  nice  little  experimental  hack. There's  also  a  critically  important  function, and  I  won't  actually spend  much  time  on  that  today. But  I'll  put  two  very  short  vignettes in  just  in  case  you  were  not convinced  of  that  to  start, I  like  to  think  about,  as  I  said, how  microboms  get  there  as  a  first  pass before  we  can  start  to  understand how  they're  responding  to  selection. This  is  a  photo  I  shamelessly  lifted  from the  Internet  of  a wild  tomato  growing  in  Peru. The  reason  I  do  this  is  because  we  work  on tomato  and  mainly  think about  it  in  agricultural  settings. But  this  is  the  context in  which  the  long  evolutionary  history of  the  plant  in  its  microbes  happens. This  is  where  I  like  to  think  about  how a  microbiome  is  born,  if  you  will. A  plant  comes  about  in  a  community. The  Philo  sphere,  believe  it  or  not, is  pretty  empty  of microbes  during  early  plant  development. The  majority  of  colonization, as  I'll  show  you, happens  through  mechanisms  after  germination. But  even  at  the  germination  stage, there's  the  possibility  that  you  have vertical  transmission  from  parental  plants, from  maternal  plants  through  seeds. Then,  of  course,  these  are  open  to  wind, rain,  insects  and  so  forth. Wind  is  very  great. Rain  is  a  great  way  of moving  microbes  around, but  they  still  have  to be  coming  from  somewhere. And  the  somewhere  is  most  likely  neighbors. So  this  was  our  first  best  guess  at all  the  different  places  that microbial  colonization might  come  from  and  matter. I'll  just  go  through  very quickly  three  vignettes to  convince  you  that  these  are  all important  sources. The  first  one  is  seed. This  is  work  done  by  Shirley  John, who's  now  a  Cornell  as a  graduate  student  but was  an  undergrad  at  the  time. She  ran  this  with  Norma  Morella, who's  now  at  Pivot  Bio. What  Norma  and  Shirley  did  is  they  went into  an  organic  farm,  collected  tomato, sterilized  the  outside  of the  fruit,  collected  the  seeds, and  simply  asked  what's on  the  outside  of  the  seed? The  answer  is  this  is  16  S  sequencing. Don't  worry  too  much  about  it. The  answer  is  actually  not  that  much. If  you  really  squint, you  might  notice  that  the  vast  majority of  these  are  a  pentoa  species, a  couple  of  strains  of  pentoa  that was  pretty  consistent  across these  three  different  cultivars  we  looked  at. But  nonetheless,  bacteria  were  there, which  allows  us  to ask  questions  about  whether that  initial  colonization  is important  to  future  assembly or  other  plant  functions. Let's  start  with  the  impact  on  assembly. It's  true  across  systems,  including  humans, that  we  indeed  acquire  a  microbome  at  birth. This  is  generally  true in  the  plant  and  animal  kingdom, but  the  majority  of those  vertical  transmission  events  are  a  few, certainly  never  a  whole  microbome. One  question  you  might  ask  is, does  that  small  seeding  of just  a  few  individuals  early  in  life  matter? And  one  of  the  ways  it  matters is  through  things  like  priority  effects. This  is  Rena  de  Bray  who  was a  graduate  student  with  me  and is  now  at  the  Max  plank. If  you're  interested  in priority  effects  in  microboes, she  put  this  really  beautiful  review  paper together  and  it's  thought  about  all  of the  different  ways  that  the  arrival  of one  species  would  of  course affect  subsequent  successional  dynamics. Everything  that  we  have  learned  from plant  and  animal  community  assembly  seems  to translate  over  quite  well to  microbial  community  assembly. Well,  actually  maybe  the  advantage is  that  in  microbiomes, we  can  study  this  at  a  scale  that's  pretty impressive  compared to  some  of  the  other  plants  and animal  datasets  which  are  harder  to come  by,  what  do  I  mean  by  that? Reno  went  in  and  found  a  public  dataset on  human  infant  gut  microbiomes. And  what  she  did  is  for each  taxa  in  the  infant  gut, she  asked  whether  these  are samples  through  time  from  the  same  infants. And  she  asked  whether  within  a  window, once  you  first  see  that  taxa  appear, how  often  do  you  continually  see  it? All  right,  The  top  is  an  example  of a  bacterial  species  that had  very  high  persistent. After  you  see  it,  you  always  find  it  there in  that  particular  infant  in  the  bottom. This  is  an  opposite  example  where  you see  something  pop  up  but  it  does  not  persist, it's  out  competed,  or  otherwise  disappears. When  she  did  this  for  all  of  the  taxa, what  she  then  did  was look  for  signatures  of  priority  effects. So  I'm  just  going  to  focus  on this  one  species  as  an  example. But  what  she  could  then  do  is quantify  persistence as  I've  already  shown  you. And  then  ask  whether  the  persistence  could  be predicted  based  on  the  microbiome at  the  time  it  first  arrived. This  is  just  a  principal component  plot  showing, it's  a  nice  way  of visualizing  microbiome  dissimilarity. Really  all  you  need  to  know  about these  plots  is  if two  points  are  close  together, those  microbiomes  are  very  similar, they're  far  apart,  they're  very  dissimilar. What  you're  showing  here  is  that  persist, if  this  particular  tax  persisted, that's  because  it  arrived in  a  microbiome  that  had a  very  different  community  composition than  if  it  did  not  persist. And  she  did  this  with  a  number  of other  datasets  from  different  organisms, and  we  could  find  very  strong and  clear  signatures  of  priority  effects. This  is  just  to  say  that  if you're  working  on  a  system  or  a  microbiome where  vertical  transmission  doesn't seem  very  important  because it's  only  one  of  tax  or  very  low  density, it  could  end  up  being  critically important  to  shaping the  entire  assembly  of the  microbiome  during  development. The  other  way  that these  initial  priority  effects in  vertically  transmitted  organisms might  be  important  is  of  course, actually  shaping  the  health  outcome  of, in  this  case,  seedlings. The  other  side  of  this  experiment, which  I  just  introduced  is  that  after Shirley  took  off the  seed  epiphytic  community  of  these  tomato, she  then  surfaced  sterilized  seeds. Which  by  the  way  we do  in  agricultural  settings almost  all  the  time. You  then  take your  sterilized  seeds  and  then  she remoculated  half  of  them with  their  endogenous  microbiome. She  just  put  the  microbiome  back  on, allowed  the  seedlings  to  germinate, and  then  challenge  them with  Pseudomon  syringe, which  is  a  nice  model  pathogen  in  tomato. And  asked  how  disease progressed  in  tomatoes  that  came from  sterile  seed  or  those that  had  been  given  their  microbiome  back. The  punch  line  is  if  you  look  at the  disease  curves  over  time, These  are  four  different  cultivars, each  with  their  own  unique microbiome  given  back. The  general  rule  was  that  the  seedlings  that were  only  given  buffer  were more  susceptible  to  the  pathogen. They  had  more  disease  over  time than  those  had  been given  their  microbiome  back. Even  though  again,  it's just  a  few  pentoa  species  that  was  enough to  protect  these  early  seedlings against  a  very  common pathogen  in  the  environment. To  the  first  question,  is vertical  transmission  important  to microbiome  assembly?  The  answer  is  yes. In  both  humans  and  plants, and  the  majority  of systems  where  this  has  been  looked  at, it  matters  in  two  ways. Both  functionally  for  phenotype but  also  in  terms  of  shaping  future  assembly. In  terms  of  the  second  questions  of horizontal  transmission coming  from  the  environment, this  is  a  more  challenging  question to  ask  because, well,  it's  just  messy,  right? As  soon  as  you  put  something  outside. The  first  way  we  ask  this  is by  thinking  about  systems  in  which this  wind  rain  suspersal avenue  has  been  disrupted. This  is  tomato  growing in  a  starter  greenhouse. It  turns  out  I  didn't know  this  till  a  very  simply, long  time  into  my  career  working  on  tomato. That  all  tomato  plants  that are  grown  in  commercial  settings  start in  a  greenhouse  and  they're planted  once  they're  about  high. So  this  early  assembly  is completely  disrupted  in  a  greenhouse  setting. There's  no  rain,  there's  no  wind. This  allowed  us  to  do  something  interesting, which  is  to  say,  well, if  we  grow  tomato  in  a  greenhouse, does  a  phylosphere  microbiome  establish? If  not,  can  we  make  an  amendment? This  is  Eli  Melferberg who  ran  this  project  for  his  Phd. He's  now  with  Sam  Brown  at  Georgia  Tech. What  Eli  did  is  he  created a  synthetic  community, which  is  just  a  collection  of isolates  that  we  understand  and  culture. We  grew  tomato  in a  protected  greenhouse  environment  and did  a  very  simple  experiment where  half  of  them you  just  spray  early  in development  with  buffer. And  the  other  half  we  put  on the  16  bacterial  taxa  that  we know  are  widespread  in a  natural  tomato  microbiome. I  want  to  spend  much  time  on  this. The  main  punch  line  I  want  you  to take  away  is  that  we've  done  this  when, if  you  do  16  S  sequencing, they  both  have  a  microbiome. But  when  we  use  digital  droplet  PCR  to quantify  the  absolute  abundance of  bacteria  on  those  leaves, we  found  that  when  we  don't inoculate  them  with  anything  early  in  life, they  have  essentially  no  bacteria on  their  leaves.  Very  little. But  if  we  inoculate  at  fairly  low  dose  and just  in  the  first  few  weeks  of life  are  a  little  synthetic  community, Then  those  bacteria months  later  establish  and thrive  and  are  actually quite  high  density  important. And  this  is  the  only  agriculturally relevant  slide  I'll  show  today. But  importantly,  and  really a  massive  surprise  to  us. These  plants  that  we inoculated  with the  synthetic  community  early  in life  ended  up  having more  fruit  yield  is  much  higher. What's  really  cool  is  there  were more  fruit  of  the  same  size  and  weight. This  additional  microbial  philosphere is  essentially  acting  as a  resource  for  the  plant. Or  otherwise  changing  far  to  fruit. Again,  important  function. But  for  the  point  of  today's  talk, the  main  thing  I  want  to  take  away from  this  is  that  all  of those  environmental dispersal  mechanisms  matter. And  if  you  protect  a  plant from  wind  and  rain  and  insects, then  indeed  they  have a  depoperate  microbiome. The  final  part  of  that  is not  just  the  wind  and  rain, but  as  I  already  said  from where  to  answer  that, Kyle  Meyer,  who  was  a  poc, in  the  lab,  ran a  beautiful  experiment  where we  basically  had  the  fun  of taking  an  entire  city block  and  building  neighborhoods, little  plant  neighborhoods. And  this  is  what our  plant  neighborhood  looks  like. We  created  these  rings  in a  huge  replicated  block  design across  the  city  block. Where  we  had  a  tomato  plant  surrounded by  a  bean  plant surrounded  by  tomato  or  pepper. And  all  combinations  including tomatoes  surrounded  by  tomato  and  so  forth. But  the  other  thing  that  Kyle did  is  that  we  grew  the  focal  plant. We  put  everything  into the  field  at  one  month  old. We  allowed  everything  to  grow  for  a  month, and  then  we  harvested  the  focal  plant that  allowed  the  rest  of  the neighborhood  to  continue  growing. And  we  put  back  into  the  center, planted  the  same  species and  same  original  age  one. And  then  measured  the  microbiome  of  that. We  had  three  times  where  we transplanted  the  focal  plant. I  think  of  it  as a  Dorian  Gray  experiment,  right? The  neighborhood  grows  up,  but the  focal  plant  stays  the  same  age. Asked  how  important  the  neighborhood is  in  shaping the  microbiome  of  that  focal  plant. Now  I've  glossed  over  something  important, which  is  just  like  human  hosts, plants  are  not  just  a  petri  dish,  right? They  have  very  specific  mechanisms  by  which they  select  for  the  microbes that  land  in  and  on  them, including  leaf  exidatestroot  exidates as  well  as  defensive  chemicals. But  the  question  is  whether  that  selection  is the  most  important  factor  or  whether the  environment  can  overwhelm  that  selection. I'm  just  going  to  show  you  one  piece of  data  from  this  experiment. But  these  are  our  three  harvests where  the  focal  plant was  always  one  month  old, but  the  neighborhood  was growing  up  around  it. What  we  looked  at  is  how  much  of the  variation  in microbiome  community  composition. We  could  explain  by  the  genotype, whether  it  was  pepper,  tomato,  or  bean. The  neighborhood  who  it  was  surrounded by  or  where  it  was  in  the  city  block. Initially,  at  that  first  harvest, where  everyone's  essentially  the  same  age and  the  same  biomass, we  found  that  plants  were  pretty  darn  good at  selecting  a  microbiome. Tomato  were  able  to  select a  tomato  microbiome  being a  bean  microbiome,  and  so  forth. But  as  that  neighborhood became  larger  and  larger, and  you  had  these  mass  effects  of  dispersal, the  plant  lost  the  ability  essentially, to  select  for  something  that resembled  a  tomato  microbiome. We  might  start  to  care  about  this  a  lot  when we  think  about  rare  plants  in  a  community. As  plants  become  more  and  more  rare, this  ability  to  recruit might  become  more  and  more difficult  if  the  neighbors in  your  neighborhood  is  important. So  all  of  this  matters. I  am  going  to  No,  I'll  do  it. Okay. All  of  this  matters.  And  I've already  alluded  to  this  a  little  bit. But  what's  arriving  there is  just  the  start  and then  comes  the  selection. I've  mentioned  how  plants  themselves  select. And  I'll  just  give  you  one  little  vignette  of how  that  happens.  This  is  Banca. Banca  was  a  post  graduate  student  in  the  lab. What  they  did  is  went  out  to the  beautiful  UC  Botanic  Gardens. If  you  come  visit  us  in  Berkeley, we'll  take  you  for  a  walk  there. It's  gorgeous. They  chose  24  plant  species and  sampled  leaves  did  16  S  sequencing. As  the  first  pass,  just  ask, how  important  is  the  plant  species  identity in  this  common  garden  in shaping  the  microbes  that  live  on  the  leaves. The  answer  is,  don't  worry too  much  about  the  bar  plot  here, it's  just  showing  you  the  variation in  terms  of  bacterial  tax  up. What  I  want  you  to  know  is  that  21%  of the  microbiome  variation  on  leaves  is explained  by  what  species  of  plant  it  is. The  plants  are  definitely  not  petri  dishes, they  are  very  selective  media. Okay.  All  right. That  was  interesting,  but we  still  don't  really  understand how  they're  selecting  ca, being  very  clever  and  thinking  about the  ways the  leaf  morphology  might  shape  this. Had  this  idea  to  separately  sample the  top  of  a  leaf  and the  underside  of  a  leaf. You  might  expect,  if  you  think  about  it, that  the  top  of  the  leaf  is more  prone  to  dispersal, and  just  whatever  arrives  there, the  underside  of  the  leaf  is where  all  the  stomata  are, where  all  the  leaf  exhibits  are likely  be  coming  out  of  the  plant. We  had  an  idea  that  the  top  of  the  leaf should  be  more  neutral  in  terms  of  dispersal, and  the  bottom  might  be  where  you  see these  more  deterministic  processes. Indeed,  that's  what  we  found. On  the  left,  you're  looking at  just  diversity. Overall,  16  S how  many  bacterial  species  were  there. Indeed,  the  top  of  leaves  across  all  24  of these  species  generally had  higher  alpha  diversity. Maybe  you  could  think  of them  as  being  more  open  to dispersal  and  so  forth than  the  underside  of  the  leaf. But  the  other  neat  thing  is  we used  a  series  of  occupancy  abundancy  curves, which  I  won't  show  you  today, but  a  way  of  asking  about the  bacteria  that  are highly  species  specific, those  that  showed  high  endemism across  plant  species. Here  we  found  the  likelihood  of  finding these  endemic  species  to be  much  higher  on the  underside  of  the  plant  leaf. We're  still  trying  to  wrap  our head  around  exactly  how plants  really  culture  or select  for  their  microviome, but  this  is  at  least  suggesting that  where  they're  doing  that  is primarily  on  the  underside  of  the  leaf  where the  most  likely  chemical  interactions between  a  microb  and  a  plant  occur. So  anyone  in  the  room  who works  on  root  microvimes, this  should  be  reminiscent  to the  idea  of  root  exidates, really  shaping  who  can colonize  on  root  surfaces. If  a  plants  can  select  a  microbiome, then  we  can  too. I  like  doing  experimental  evolution a  couple  of  years  before  the  paper  I'm going  to  show  you  today,  Norma  Morella, did  a  beautiful  experiment  where she  took  a  tomato  microbiome and  passaged  it  over  time  in the  greenhouse  on  different  tomato  genotypes. The  punch  line  of  that  is  that  it  worked, we  could  select  for  a  microbiome that  was  much  less  diverse than  the  starting  microbiome  from  the  field, but  that  was  highly  adapted  to  the  tomato. Okay,  and  I'm  not  going  to show  you  that  data  today, but  you'll  have  to  take  my  word  for  it. You  can  read Norma's  paper  if  you're  interested. But  Meyer  came  into  the  lab just  after  we  had  published  that  paper. And  we  were  convinced  we  could select  for  a  well  adapted  microbiome. But  now  we  were  ready  for  a  rich  hypothesis. And  one  of  the  questions in  the  microbiome  field, going  back  to  this  question  of phylosymbiosis  and  specialization, is,  can  and  does a  microbiome  adapt  to  a  host  plant? If  the  answer  is  yes, then  you  should  be  able  to, just  like  we  select  for  general  parasites or  specific  parasites, we  should  actually  be  able  to  select  for a  more  specific  microbiome or  a  more  general  microbiome. It's  a  little  high  risk, I  admit,  but  we  went  for  it. What  Kyle  did  is  he started  with  a  tomato  microbiome, a  rich,  diverse  microbiome  from  the  field. And  apply  that  microbiome  to either  tomato  or  the  wild  tomato, which  I  showed  you  the  picture  of poll  folia,  pepper  or  bean. Then  at  each  passage at  the  end  of  eight  weeks, he  would  cull  the  whole  plant to  wash  off  all  the  microbes, pellet  those  microbes,  and put  that  back  onto  just  one  plant, all  biologically  independent  lines, six  lines  per  treatment. He  did  that  for  six  passages, this  is  non,  triviulus  is almost  two  years  in  the  greenhouse. We  did  that  on  con,  specific lines  where  the  microbiome always  saw  the  same plant  generation  after  generation. Or  we  did  a  host  swapping experiment  where the  microbiome  started  on  being, for  example,  it  went  to  tomato, then  back  to  being  tomato  bean  and  so  forth. We  dropped  the  digital PCR  to  measure  density  and then  we  did  16  S  sequencing to  quantify  composition, although  we  now  have  the  metagenomes, but  we're  not  quite  ready  to  show  it  today. And  then  we  got  to the  end  of  the  experiment  and realized  we  had  a  really  big  problem, which  is  how  do  you  measure  a  microbiome? Specialism  versus  generalism. What  even  is  that? How  do  we  measure  it? We  scratched  our  head  of a  lot  and  came  up with  a  few  different  ways  of  asking  this. The  first  way  was similar  to  the  neighborhood  experiment, where  we  simply  asked  how  much  of the  community  composition  variation could  be  explained  by  host  identity. With  the  idea  being a  generalist  microbiome  would have  less  of  an  impact, a  host  signature, than  a  specialist  microbiome. This  is  the  same  thing. This  is  the  percent  of  variation  of community  composition  that  can be  explained  by, in  this  case,  the  host  gena  type, whether  it  was  tomato,  pepper  or  bean. This  is  the  vertical  lines  that  are the  conspecific  lines  that  stayed  on tomato  or  stayed  on  bean  and  so  forth. In  those  cases,  when we  compared  the  microomes, there  was  a  strong  host  signature. Another  way  of  saying  that  is  we  could easily  differentiate  a  tomato  microbiome from  a  bean  microbiome  from a  pepper  microbiome  explained  about  40%  of the  variation  in community  composition  every  other  generation. We  could  do  the  same  analysis for  the  host  swapping lines  because  every  other  generation they  were  indeed  on five  different  species,  right? So  we  can  do  the  same  analysis and  in  that  case  we found  a  decreasing  signature of  the  host  on  community  composition. It's  the  best  we  can  do  at  the  time to  really  ask  this  question  of  generalism. It's  a  fairly  nuanced  result, but  again  this  is  only  five  passages. This  is  a  short  period  of  time  to  be swapping  a  microbial  community  around  hosts. To  me,  it  was  pretty  compelling  and  exciting that  even  under  such  a short  selection  regime, we  managed  to  find  a  microbiome that  you  want  to. And  this  could  establish equally  well  on  tomato, pepper,  and  bean,  let's  say. Then  following  that  logic  through. If  that's  true,  then if  we  put  these  microbiomes onto  new  plant  species that  they  never  saw  in  the  experiment, we  should  see  the  same  thing. The  generalist  or  host swapped  microbiome  should  be able  to  establish  on an  entirely  new  host  plant. We  did  that  by  using  sum, corn  and  Canola  species that  were  not  involved  in  the  experiment. In  this  case,  whereas  our  con, specific  line  was  very differentiated  depending  on  which of  those  hosts  it  was  applied  to. For  our  generalist  host  swap  lines, we  saw  less  of  a  host  signature, again  indicating  that  we  were selecting  for  something  of  a  generalist. All  right,  the  last  part of  this  though  is  the  flip  side, the  specialism,  did  we  get a  well  adapted  microbiome? Just  did  a  nice  job  of  introducing  this, which  is  the  idea  of these  community  coalescence  experiments, or  we  call  them  community  crashing  or community  smashing  experiments. They're  really  good  fun. The  basic  idea  behind  it  is we  take  one  microbiome  and  another and  we  use  droplet  digital  live dead  PCR  to  quantify how  many  live  cells there  are  in  each  of  those. And  we  put  a  50,  50  mix  of  live  cells. We  inoculate  plants  in  this  case  with either  the  tomato  adapted  microbiome, a  bean  adapted  microbiome, or  a  50,  50  mix  of  the  two. We  apply  them  to  tomato and  we  let  them  duke  it  out, and  we  ask  who  wins? All  right. When  you  do  that,  what we  find  is  that  if  you  look  at  this  chair, you  read  the  cluster  over  here in  the  middle  of  that  is  the  tomato  con, specific  line  on  tomato  plants. That's  what  a  tomato  microbome looks  like  if  you  put  it  on  tomato. What  you'll  see  is  that  when  we take  that  tomato  line  and  we crash  it  with  either be  microbiome  or  pepper  microbiome. The  tomato  adapted  microbiome  winds out,  it  does  not  ****. That's  very  different. For  example,  if  we  spray  pepper alone  onto  tomato  or  being  alone  onto  tomato, those  cluster  out  very  far. If  you're  ever  interested  in adaptation  at a  whole  microbiome  community  level, please  reach  out  to  me  because these  community  coalescence  experiments  are really  powerful  and  exciting  ways to  convince  yourself that  there's  an  adaptation there,  there's  something  important. The  microbes  that  you're  studying  are indeed  well  adapted  to  that  space. In  this  experiment,  we  were  able  to  show  that over  only  five  passages, we  could  get  a  pepper  adapted  microbiome  or a  bean  adapted  microome  from something  that originally  started  from  tomato. Okay,  so  good,  I've  got  time to  do  due  diligence  on  bacteria  phages. So  the  microbes  get  there from  all  of the  different  mechanisms  I've  shown  you, the  host  selects  that  screen. The  microbiome  itself  is  selecting  by shaping  the  competition within  the  microbiome. But  something  that  is  missing  from the  vast  majority  of microbiome  studies  is  that there's  another  player  in every  microbiome  that  is  phages. If  you're  not  familiar,  phages are  viruses  that  infect  bacteria. The  phages  we  work  with  are  lytic  phages. They  actually  bind  to  infect and  kill  or  live  bacterial  cells, just  like  any  good  parasite  they  are. Or  they  have  the  potential  to  be an  important  selective  force on  their  host  population. Okay,  One  way  that  they  can  be a  force  is  through negative  frequency  dependent  selection. For  my  Phd,  I  was  working  on trematodes  as a  driver  of  sexual  recombination. And  I  couldn't  shake the  theory  of  red  queen  theory. When  I  started  working  on  pages, the  obvious  question,  are phages  parasites  like  any  other  parasite? Are  they  diversity  generating  mechanisms? The  theory  behind  it  is  quite  simple. If  you  have  parasites in  a  population  or  in  a  community, there's  going  to  be  selection  on those  parasites  to  infect  common  host  types. In  infecting  common  host  types, that  brings  the  density  of  those  common, or  the  frequency  of  those  host  down, makes  room  for  rare  hosts, and  therefore  maintains  diversity. We  often  think  about it  at  the  population  level, but  the  same  thing  holds true  potentially  at  the  community  level. This  is  the  older  study,  but  I'm  going  to update  it  with  some  unpublished  results. Next,  I've  already  introduced Norma  Morella  who's  a  Phd  student. The  question  we  had  is  if  pages  have the  potential  to  select against  common  bacteria and  increase  diversity, then  we  should  be  able  to  recreate that  at  the  microbiome  level. The  way  Norma  tested this  is  went  out  to  the  field, collected  a  tomato  microbiome as  I've  already  described, but  in  this  case,  before  inoculating our  nottobiotic  tomato  plants or  our  germ  free  tomato  plants, she  did  a  series  of  size  separations. She  captured the  bacterial  and  fungal  community on  a  filter  0.4  or  five  micron  filter. And  then  the  flow  through  which includes  all  the  virus  like  particles, all  the  phages  that  may  or may  not  have  existed  in  the  field. At  this  point  we  didn't  know  she concentrated  them  using  an  Amico  filter. And  then  we  inoculated  tomato with  either  a  bacterial  phage depleted  microbiome  or  we  put the  phages  back  together  with their  bacterial  community and  inoculated  them  together. We  then  after  one  day, used  dropoleate  digital  PCR  to ask  whether  the  phages  killed  anything. Were  there  even  phages  there? And  then  after  a  week,  we  asked how  the  presence  or  absence  of  phage shaped  microbiome  community assembly  to  our  relief. After  one  day,  when  we  looked  at the  number  of  bacterial  cells  on  leaves, we  found  that  when  we  recombined our  microbial  community  with the  resonant  phages  from  the  natural  setting, there  were  indeed  a  lot  of  phages there  capable  of  killing  bacteria. The  overall  density  of these  microbiomes  was  lower, which  is  a  release  because  it  means that  the  selection  potential  is  there. The  phages  are  there  in  nature. And  then  the  next  question is,  do  they  matter? Here  we  just  use  16  S  amplicon sequencing  to  again, just  ask  how  many  bacterial  taxa  there are  with  the  alpha  diversity of  these  communities  look  like, just  like  we  predict  with red  queen  theory  or negative  frequency  dependent  selection. When  the  phages  were  there, we  had  a  more  diverse  microbiome than  when  the  phages  were  depleted. We  were  really  excited about  this  for  many  reasons. It's  always  nice  when  theory  works,  right, But  also  it  means  that  it  opens the  opportunity  to  use bacteria  phages  to shape  and  reshape  microbomes. The  second  figure  here  is actually  asking a  slightly  different  question, which  is  if  we  look  at  two  microbomes  that received  the  bacteria  phage depleted  community, the  same  bacteria  but  without  their  phage. After  a  week  we  ask  how dissimilar  those  two  communities  are. We  find  that  they're  actually  quite dissimilar  when you  recombine  them  with  their  phage. They're  far  more  similar. The  phages  are  not  only  increasing  diversity, but  they're  shaping  where the  microbiom  goes  during  assembly, which  is  again,  pretty  exciting  from a  microbiome  engineering  perspective. More  recently,  we  combine this  phage  diversity  work with  another  finding  in  the  lab, which  I  won't  show  you,  but  is pretty  common  across plant  and  animal  systems. Which  is  that  we  know  the  microbiome can  be  disease  protective. I've  already  shown  you  earlier on  one  little  vignette. A  more  diverse  microbiome  on  average, is  less  susceptible  to invasion  by  a  pathogen. This  holds  true,  and  again, the  human  gut  microbiome  and  so  forth. Often  dysbiosis  or  a  disease associated  microbiome  state is  less  diverse  microbiome. There  are  some  neat  exceptions, but  that's  the  general  rule. If  that's  true  and  you combine  that  logic  with  this  logic, you  come  up  with  this  hypothesis, which  is  if  phages  impacts diversity  and  the  more  diverse  microbiome is  more  disease  protective, then  phages  should  reduce  disease. I've  already  introduced  Rena, Debra,  and  she  led  the  study. What  Rena  did  is  very similar  to  what  Norma  did  with  one  twist. She  went  out  to  the  field, she  collected  a  microbiome, she  size  separated  it, inoculated  plants  with either  just  Pseudomonas  syringe, so  this  is  just  above her  control  with  the  pathogen, the  pathogen  with  a  microbiome that  was  phage  depleted. A  microbiome  that  had  been  recombined  with its  phage  or  just  the  phage  alone. What  she  found  was exactly  as  we  would  predict. Under  the  theory,  the  phages created  a  more  disease  protective  microbiome. The  Y  axis  here  is the  density  of  the  pathogen pseudomonaseringe. You  see  the  least  Pseudomonaseringe when  the  microbiome  is recombined  with  its  phage. Importantly,  that  wasn't  because the  phages  are  directly  killing  the  pathogen, it  was  actually  an  indirect effect  through  modulating the  microbiome  for  various  reasons which  I  won't  get  into  right  now. We  sat  on  this  data  for almost  two  years  because  there's  other  things that  get  combined  when  you  do these  phage  depletion  experiments like  LPS  and  other  Nts  and  pants. But  we  figured  out  a  way  to rerun  it  to  convince ourselves  that  this  wasn't a  byproduct  of  our  method. We  did  it  two  years  later with  an  entirely  different  set  of microbiomes  from  a  different field  two  years  later, and  found  the  same  thing  a  second  time. Now  I'm  really  getting convinced  that  this  is  a  robust, very  real  result  where  again, we  saw  the  most  disease  protection when  the  microbiome  was present  with  its  phage. I  want  to  show  you  the  data  here, but  an  interesting  aside  was  that  we both  combined  the  microbiome with  its  own  phages  from  the  same  plant. We  sampled  it,  or an  allopatric  phage  population from  a  nearby  plant. We  actually  found  the  allopatric  or nearby  phage  more  disease  protective. We  think  there's  a  Goldilocks  effect where  phages  killed  but  not  too  much. All  right. More  or  less  already  said  this, but  just  to  say  occasionally, like  this  outlier  here, we  did  actually  find a  naturally  occurring  phage  that does  kill  and  infect  the  pathogen. This  is  a  pathogen  plate with  these  little  bacteria  phage  plaques. It's  not  to  say,  it's  not  always  direct. It  can  be  a  direct  protective effect  of  phage  as  well. But  for  the  majority  of  them  we  did  not find  any  phages  capable  of  killing  sing. It  really  was  to  this  modulation  of the  microbiome  fund  system to  be  able  to  test  this  idea. But  maybe  not  the most  impactful  because  indeed, tomato  microbiomes  outside  of agriculture  aren't  that  important. While  I  was  at  Senso  College, where  Jess  and  I  spent  a  lot of  time  chatting  about  science, I  revisited  a  paper  that  I  had read  many  years  below  before, and  Dobson's  up  there, that  put  forward  a  very  simple  idea, but  a  beautifully  well  written  up one  that  if  you  have  a  healthy  ecosystem, a  diverse  and  thriving  ecosystem, that  should  actually  be  one  that's also  rich  in  parasites. This  is  quite  hard  to show  or  test  in  empiric, in  plant  and  animal  systems, but  very  easy  to  test  in microbial  systems where  the  phage  is  the  parasite. And  we  really  understand microbiome  health  and  pybiosisor  get  in there  in  collaboration  with Rachel  Wheatley  who  at Oxford,  Dominic  Coltapelzo's  postdoc. In  my  lab,  we  ran through  all  of the  existing  datasets  that  we  could get  our  hands  on  that  had looked  at  a  microbiome  in  health and  under  disturbance  had sequenced  both  the  bacteria  and  the  Vi, all  of  the  pages  associated with  that  microbiome. I  want  to  just  say  right  up  front  that  this was  one  of  these  meta  analyses where  the  only  thing  we're  going  to see  is  very  broad  signatures because  we  were  looking  across all  sorts  of  different  dispiosesdspiosis, especially  in  terms  of the  human  gut  literature. It  can  mean  all  sorts  of  different  things. It  can  mean  a  bacterial pathogen  that  came  in  and establishes  and  wipes  out  resident  microbes. It  can  mean  a  shift  of the  resident  microbiome  to a  different  state  that somehow  leads  to  inflammation or  other  disease  or  disorders. It  can  just  mean  loss  of particular  key  taxa  that  had a  function  that  no  longer  do their  function,  et  cetera. But  we  looked  across  all  of  them, just  asked,  are  there any  commonalities  in  terms of  what  happens  at  the  phage  level? Indeed,  are  phages  a  signature of  a  healthy  functioning  microbiome? The  first  result  made  us  a  little  sad, which  is  to  say  that  our  first  thought  was, surely  dyspiosis  should  mean less  rich  phage  diversity. For  some  dyspioss,  we  indeed see  a  strong  decrease in  the  alpha  diversity  of  the  virrome. By  the  way,  you  can  see how  many  different  dyspioses we  were  looking  across. Covid,  celiac  disease,  obesity, spinal  injury,  HIV, these  are  very  different  dyspioss, but  nonetheless,  in some  cases  we  actually  found more  viruses  in  a  dyspiotic  system or  a  disturbed  system  in  the  absence. As  a  first  pass,  we thought  no  real  overarching  signature  of phages  in a  healthy  versus  disturbed  microbiome. When  we  looked  at  beta  diversity, there's  no  nice  figure  to  show  here, so  I'll  just  give  you  the  actual  data. When  we  looked  at  beta  diversity, we  found  indeed  that  the  majority  of  systems, 63%  found  a  shift in  the  viral  beta  diversity. Which  viruses  were  there, which  phages  were  there? Under  disturbance  and  in health,  they  are  responding. But  that's  no  surprise.  If  you suddenly  change  which  hosts  are  there, you  should  expect  to  see a  change  in  the  viruses  that  are  there  too. Similarly,  the  papers  that  look  for enrichment  of  particular  viral  taxes, are  there  specific  viruses  that  you only  find  in  biasis  compared  to  health? Here,  the  real  majority, 83%  There  was  indeed one  virus  that  you  didn't find  in  this  page in  health  that  you  did  find  in  disturbance. But  again,  not  that satisfying  if  we're  trying  to sell  that  there's a  real  overarching  signature. The  only  thing  up  until  now was  just  that  some  viruses  change. The  final  analysis  we  did is  the  one  that  we  found  most  exciting. When  we  looked  across  all  of our  healthy  control  systems  cohorts, what  we  found  is  that  on  average, there  was  a  strong  positive  association between  bacterial  diversity and  phage  diversity. Makes  sense. If  you  have  a  written  posts, you  expect  to  have  more  phages  in a  natural  and  equilibrium  state. What  we  found  is  that  that  association breaks  down  under  dysbiosis. Where  is  that? What  you're  looking  at  is  the  correlation between  the  Rome  and  bacterium  diversity. How  tightly  correlated  the  two  are  in the  control  cohorts  and  in the  disturbance  cohorts For  each  of  those  studies. In  the  majority  of cases  we  saw  strong  decrease. Interestingly,  the  cases  where  we  didn't  see decreases  were  those  that were  non  pathogen  involved. Obesity,  metabolic  disease, and  high  fat  diets. Where  you  see  a  pathogen or  a  disease  associated  state, an  infectious  disease  associated  state. One  of  the  clearest  signatures  is this  breakdown  of the  relationship  between  the  two. Going  back  to  the  paper, it  suggests  that,  yes, indeed  there  is  a  broad  signature of  phage  diversity. But  maybe  it's  not  just that  more  is  better  or  less  is  better, it  rather  has  to  do  with  that  balance between  the  parasites  and the  hosts  in  the  system. All  right,  which  just brings  me  back  to  the  end, which  is  as  we learn  more  and  more  about  the  microbiomes, I  am  a  real  believer  that  we  can leverage  a  lot  of this  knowledge  for  solutions, societal  solutions, especially  with  such  rapid  global  change. But  also  thinking  about  it  in an  agricultural  setting  where we've  worked  pretty  hard to  break the  longstanding  evolutionary  association between  plants  and animals  and  their  microbiomes. Now  as  we  learn  more and  more  about  their  importance, we  can  start  to  redress  some  of those  problems  and  engineer the  microbiome  and  potentially  create more  adapted  hosts  to  change. And  with  that,  I  will  thank all  my  collaborators  that  especially  Jess, who's  sitting  here,  who  was  involved  in the  vast  majority  of  the  microphone studies  that  I  showed. I  haven't  really  enjoyed  collaborating with  and  talking  theory, which  I  didn't  get  to spend  much  time  on  today. And  everyone  who  contributed  to  the  work, most  of  whom  I  thanked  as I  went  through  that. I'll  take  questions  and  thank  you.