Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A study on the recovery of Tobago's coral reefs following the 2010 mass bleaching event Buglass, Salome 2014

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

Item Metadata


24-ubc_2015_february_buglass_salome.pdf [ 1.72MB ]
JSON: 24-1.0167267.json
JSON-LD: 24-1.0167267-ld.json
RDF/XML (Pretty): 24-1.0167267-rdf.xml
RDF/JSON: 24-1.0167267-rdf.json
Turtle: 24-1.0167267-turtle.txt
N-Triples: 24-1.0167267-rdf-ntriples.txt
Original Record: 24-1.0167267-source.json
Full Text

Full Text

A	  study	  on	  the	  recovery	  of	  Tobago's	  coral	  reefs	  following	  the	  2010	  mass	  bleaching	  event	  by	  Salome	  Buglass	  	  BSc.	  Geography,	  University	  College	  London,	  United	  Kingdom,	  2009	  	   A	  THESIS	  SUBMITTED	  IN	  PARTIAL	  FULFILLMENT	  OF	  THE	  REQUIREMENTS	  FOR	  THE	  DEGREE	  OF	  	  Master	  of	  Science	  in	  THE	  FACULTY	  OF	  GRADUATE	  AND	  POSTDOCTORAL	  STUDIES	  	  (Geography)	  	  The	  University	  of	  British	  Columbia	  (Vancouver)	  	  	   December	  2014	  	  	  ©	  Salome	  Buglass,	  2014	  ii	  	  Abstract	  The	  rise	  of	  ocean	  temperatures	  globally	  has	  become	  a	  grave	  threat	  to	  coral	  reefs,	  as	  it	  is	  increasing	  the	  severity	  and	  frequency	  of	  mass	  coral	  bleaching	  events	  and	  post-­‐bleaching	  coral	  mortality.	   The	   continued	   existence	   of	   productive	   coral	   reefs	   will	   rely	   on	   corals’	  ability	  to	  undergo	  recovery.	  In	  2010,	  Tobago’s	  coral	  reefs	  were	  exposed	  to	  severe	  heat	  stress	   leading	   to	  mass	   bleaching	   of	   up	   to	   29-­‐60%	   of	   colonies	   at	   observed	   sites.	   This	  study	  evaluated	  the	  impact	  of	  coral	  bleaching	  and	  recovery	  of	  coral	  communities	  across	  three	  major	   reef	   systems	   in	   Tobago	   that	   differ	   in	   their	   exposure	   to	   terrestrial	   runoff.	  Assessments	   were	   done	   on	   the	   1)	   density	   and	   composition	   of	   coral	   juveniles	   to	  characterise	   the	   levels	   of	   recruitment,	   2)	   sedimentation	   rates	   and	   composition	   to	  understand	  its	  potential	  impact	  on	  recovery,	  and	  3)	  species’	  size	  frequency	  distributions	  in	  2010,	  2011	  and	  2013	  to	  examine	  temporal	  changes	  among	  population	  size	  structure.	  	  In	  2013,	  low	  juvenile	  densities	  were	  observed	  (5.41	  ±	  6.31	  m-­‐2)	  at	  most	  reef	  sites,	  which	  were	   dominated	   by	   brooding	   genera	  while	   broadcasting	   genera	  were	   rare.	   Sediment	  material,	   measured	   in	   May	   and	   June	   (end	   of	   Tobago’s	   dry	   season)	   was	   mostly	  terrigenous	  and	  deposited	  at	  rates	  below	  coral	  stress	  threshold	  levels	  at	  most	  sites.	  Out	  of	   27	   species	   populations	   assessed	   between	   all	   sites,	   4	   populations	  mean	   colony	   size	  had	  significantly	  changed	  by	  the	  bleaching	  event,	  and	  only	  changed	  5	  populations	  over	  the	  two	  following	  years.	  The	  few	  populations	  that	  were	  significantly	  altered	  (mainly	  S.	  siderea	  and	  M.	   faveolata)	  after	   the	  bleaching	  saw	  a	   rise	   in	  small	   sized	  colonies,	  mostl	  likely	  as	  a	  result	  of	  colony	  fragmentation.	  	  This	   study	  highlights	   that	   recovery	   via	   sexually	  produced	   recruits	   among	  broadcasting	  species	  was	  limited.	  While	  sedimentation	  rates	  were	  low,	  it	  is	  likely	  they	  are	  significantly	  higher	   throughout	   the	   rainy	   season,	   thus	   a	   long-­‐term	   sedimentation	   study	   is	   highly	  recommended.	  Most	  coral	  populations	  resisted	  significant	  alteration	  from	  heat	  stress	  in	  2010.	  However,	  given	  that	  future	  thermal	  stress	   is	  projected	  to	  become	  more	  intense,	  this	   study	   shows	   that	   mass	   bleaching	   disturbance	   could	   lead	   to	   decline	   coral	  population’s	  mean	  colony	  size,	  which	  could	  affect	  coral	  recovery	  as	  smaller	  colonies	  are	  less	  fecund.	  	  	  iii	  	  Preface	  This	  thesis	  is	  based	  on	  field	  data	  collected	  in	  2013	  in	  collaboration	  with	  the	  Institute	  of	  Marine	  Affairs	  (IMA)	  under	  the	  guidance	  and	  supervision	  of	  Dr.	  Simon	  Donner.	  All	  data	  from	   the	   field	   were	   collected	   by	   myself	   with	   the	   assistance	   of	   volunteers.	   Benthic	  percent	   cover	   and	   coral	   colony	   size	   frequency	   data	   for	   2010,	   2011	   and	   2012	   were	  previously	   collected	   by	   the	   IMA	   as	   part	   of	   their	   Biodiversity	   and	   Ecology	   Research	  Programme.	   Laboratory	   sample	   processing	   was	   performed	   partly	   in	   the	   IMA’s	  Biodiversity	   and	   Ecology	   Laboratory	   and	   on	   UBC	   Vancouver	   campus.	   I	   undertook	   all	  laboratory	  sample	  processing,	  analysis	  of	  data,	  and	  writing	  of	  the	  thesis	  manuscript.	  iv	  	  Table	  of	  Contents	  	  Abstract	  ...................................................................................................................................	  ii	  Preface	  ...................................................................................................................................	  iii	  Table	  of	  Contents	  ...................................................................................................................	  iv	  List	  of	  Tables	  ..........................................................................................................................	  vi	  List	  of	  Figures	  ........................................................................................................................	  vii	  Acknowledgements	  ..............................................................................................................	  viii	  Chapter	  1.	  Introduction	  ..........................................................................................................	  1	  Chapter	   2.	   An	   assessment	   of	   coral	   juvenile	   community	   and	   sediment	   deposition	  among	  three	  major	  reef	  systems	  in	  Tobago	  ............................................................................	  4	  2.1	   Introduction	  ........................................................................................................................	  4	  2.2	   Methods	  .............................................................................................................................	  6	  2.2.1	   Study	  area	  ...................................................................................................................	  6	  2.2.2	   Sedimentation	  assessment	  .........................................................................................	  8	  2.2.3	   Juvenile	  community	  survey	  ......................................................................................	  10	  2.2.4	   Statistical	  analysis	  .....................................................................................................	  11	  2.3	   Results	  ..............................................................................................................................	  11	  2.3.1	   Juvenile	  density	  and	  composition	  ............................................................................	  11	  2.3.2	   Characterization	  of	  sedimentation	  ...........................................................................	  14	  2.4	   Discussion	  .........................................................................................................................	  16	  Chapter	  3.	  Using	  coral	  size	  distribution	  to	  assess	  the	  recovery	  from	  mass	  bleaching	  in	  the	  southern	  Caribbean	  .........................................................................................................	  21	  3.1	   Introduction	  ......................................................................................................................	  21	  3.2	   Methods	  ...........................................................................................................................	  23	  3.2.1	   Benthic	  cover	  survey	  ................................................................................................	  24	  v	  	  3.2.2	   Colony	  size	  frequency	  survey	  ...................................................................................	  25	  3.2.3	   Statistical	  analysis	  .....................................................................................................	  25	  3.3	   Results	  ..............................................................................................................................	  26	  3.3.1	   Changes	  in	  percent	  coral	  cover	  ................................................................................	  26	  3.3.2	   Changes	  in	  coral	  population	  structure	  and	  community	  composition	  ......................	  27	  3.4	   Discussion	  .........................................................................................................................	  32	  Chapter	  4.	  Conclusion	  ............................................................................................................	  36	  Bibliography	  ..........................................................................................................................	  39	  Appendices	  ............................................................................................................................	  46	  	  vi	  	  List	  of	  Tables	  	  Table	   1.	   Information	   on	   the	   six	   reef	   sites	  where	   juvenile	   assessment	  was	   undertaken	  and	  sediment	  traps	  were	  installed	  .....................................................................................	  8	  Table	  2.	  Sieved	  size	  groups	  and	  sediment	  classes	  according	  to	  the	  Wentworth	  size	  class	  system.	  .............................................................................................................................	  10	  Table	  3.	  Coral	  juvenile	  data	  per	  site.	  ................................................................................	  11	  Table	  4.	  Number	  of	  juvenile	  taxa	  found	  at	  each	  site.	  .......................................................	  12	  Table	  5.	  Mean	  and	  standard	  deviations	  of	  sediment	  measurements	  ..............................	  15	  Table	  6.	  Mean	  percent	  cover	  of	  live	  coral	  and	  (±)	  standard	  deviation	  estimated	  at	  each	  site	  and	  year	  .....................................................................................................................	  27	  Table	  7.	  Total	  number	  of	  coral	  colony	  and	  species	  recorded	  per	  site	  ..............................	  27	  Table	  8.	  Summary	  of	  the	  colony	  abundance	  and	  size	  data	  collected	  for	  each	  dominant	  species	  present	  at	  each	  reef	  sites.	  ...................................................................................	  46	  Table	  9.	  Significant	  comparison	  of	  size	  frequency	  distributions	  and	  colony	  size	  between	  years	   (2010,	   2011	   and	   2013)	   determined	   using	   Kolmogorov-­‐Smirnov	   test	   (KS)	   and	  Kruskal-­‐Walis	  (KW)	  test	  respectively.	  ...............................................................................	  47	  Table	   10.	   Significant	   comparison	   size	   frequency	   distributions	   between	   reef	   sites	  determined	  using	  Kolmogorov-­‐Smirnov	  test	  (KS)	  test	  respectively.	  ................................	  48	  	  vii	  	  List	  of	  Figures	  	  Figure	  1.	  Map	  of	  Tobago	  and	  location	  of	  studied	  reef	  systems	  and	  site	  ........................................	  7	  Figure	   2	   Proportion	   of	   counted	   juveniles	   that	   were	   produced	   from	   broadcasting	   or	  brooding	  reproductive	  strategies	  ...................................................................................	  12	  Figure	  3.	  Relative	  abundance	  of	  major	  taxa	  (genus)	  groups	  at	  each	  site	  for	  (A)	   juvenile	  population	  based	   from	  count	  data	  and	   (B)	  adult	  population	  based	  on	  percent	  cover	  assessed	  in	  2013	  (see	  Chapter	  3	  for	  data	  collection	  methods)	  ............................	  13	  Figure	  4.	  Correspondence	  analysis	  (CA)	  biplot	  .............................................................................	  14	  Figure	  5.	  Boxplot	  of	  sediment	  accumulation	  rates	  per	  site	  ..........................................................	  15	  Figure	  6.	  Stacked	  barplots	  of	  average	  percent	  of	  sediment	  (left)	  composition	  and	  (right)	  particle	  size	  distribution	  from	  sediment	  data	  collected	  May-­‐June	  2013	  .......................	  15	  Figure	  7.	  Map	  of	  Tobago	  and	  location	  of	  studied	  reef	  systems	  and	  sites	  ....................................	  24	  Figure	  8.	   Size	   frequency	  distributions	  of	   coral	   taxa	  at	   sites	  with	   significant	  differences	  between	  years,	  as	  determined	  by	  the	  Kolmogorov-­‐Smirnov	  test	  ..................................	  29	  Figure	   9.	   Boxplot	   and	   mean	   size	   (white	   filled	   dots)	   per	   species	   per	   site	   indicating	  changing	  trends	  in	  colony	  size	  between	  each	  year..	  ......................................................	  30	  Figure	  10.	  Non-­‐metric	  multidimensional	  (NMDS)	  scaling	  using	  Bray-­‐Curtis	  dissimilarities	  plot	   of	   the	   qualitative	   changes	   among	   the	   coral	   communities	   at	   each	   of	   the	  sites	  per	  year	  (named	  and	  colour	  coded)	  across	  Tobago..	  .............................................	  31	  	  	  viii	  	  Acknowledgements	  	  I	   am	   really	   thankful	   to	   the	   many	   people	   who	   supported	   and	   inspired	   me	   while	  undertaking	   my	   master’s	   degree	   and	   writing	   up	   my	   thesis.	   I	   especially	   thank	   my	  supervisor	   Dr.	   Simon	   Donner	   for	   taking	   me	   on	   as	   his	   student	   and	   for	   his	   sound	  encouragement	  and	  guidance	  throughout	  the	  entire	  process.	  I	  am	  also	  grateful	  for	  the	  time	  and	  support	  I	  received	  my	  committee	  member	  Jennifer	  Williams.	  Additionally,	  I	  am	  very	   thankful	   to	   Jahson	   Alemu	   I	   for	   assisting	   me	   in	   field	   and	   providing	   me	   with	   key	  information	  about	  Tobago’s	  coral	  reefs.	  	  My	   gratitude	   is	   also	   extended	   to	  my	  many	   fellow	  geography	   students	   and	   friends	   for	  their	  help	  and	  support,	  especially	  Leonora	  King,	  Christopher	  Quick,	  Lawrence	  Bird	  and	  David	   West.	   Thanks	   also	   go	   to	   my	   father	   David	   Buglass	   and	   Giordano	   Mitchell	   for	  proofreading.	   Further	   thanks	   go	   to	   the	   staff,	   students	   and	   faculty	   of	   the	   Geography	  Department,	   particularly	   for	   the	   support	   by	   Suzanne	   Lawrence,	   Sandy	   Lapsky	   and	  Stefanie	   Ickert.	   Particular	   thanks	   go	   to	   my	   family	   and	   friends	   who	   have	   been	   so	  encouraging	  and	  supportive	  throughout	  my	  years	  as	  a	  graduate	  student.	  	  Finally	   this	   work	  would	   not	   have	   been	   possible	   without	   the	   funding	   from	   TerreWEB,	  BRITE,	   and	   from	   all	   those	   individuals	   who	   generously	   donated	   to	   the	   crowdfunding	  campaign	  to	  help	  finance	  my	  field	  work.	  	  	  	  	  	  	   	  	  	   1	  Chapter	  1.	   Introduction	  Caribbean	   coral	   reefs	   represent	   less	   than	   10%	   of	   all	   the	  world’s	   tropical	   coral	   reef	   systems,	  nonetheless	  they	  sustain	  critical	  habitats	  for	  the	  region’s	  marine	  biodiversity	  and	  provide	  vital	  goods	  and	  services	   for	  over	  43	  million	  people	   (Wilkinson	  &	  Souter	  2008).	  Many	  coral	   reefs	   in	  this	   region,	   however,	   are	   at	   high	   risk	   of	   being	  degraded	   (Burke	  et	   al.	   2011).	   Since	   the	  1970s	  many	  Caribbean	   reefs	  have	  experienced	  unprecedented	   levels	  of	  decline	   in	   coral	   cover,	   from	  about	  50%	  to	  10%,	  and	  are	  instead	  turning	  into	  algal	  dominated	  environments	  (Gardner	  et	  al.	  2003;	   Roff	   et	   al.	   2011).	   This	   ecological	   deterioration	   has	   been	   attributed	   to	   the	   reduction	   of	  herbivory,	  due	  to	  overfishing	  and	  the	  regional	  die-­‐off	  of	  grazing	  urchins	  in	  the	  1980s,	  increased	  terrestrial	   runoff,	  marine	  pollution	  and	  disease	  outbreaks	   (Hughes	  1994;	   Jackson	  et	  al.	  2014).	  Furthermore,	   increased	   warming	   of	   ocean	   temperatures,	   driven	   by	   anthropogenic	   climate	  change	   in	   the	   last	   two	  decades,	  has	   increased	   the	   frequency	  of	  mass	   coral	  bleaching	  events,	  which	  are	  often	  followed	  by	  significant	  coral	  mortality.	  Consequently,	  these	  heat	  stress	  events	  have	   exacerbated	   the	   decline	   of	   coral	   communities	   and	   pose	   a	   grave	   threat	   to	   the	   already	  fragile	  coral	  reef	  ecosystems	  of	  this	  region.	  	  Considering	   that	   the	   frequency	   and	   intensity	   of	   mass	   coral	   bleaching	   events	   are	   likely	   to	  increase	  in	  the	  near	  future	  (Donner	  et	  al.	  2005;	  Hoegh-­‐Guldberg	  et	  al.	  2007),	  the	  post-­‐bleaching	  recovery	  of	  schleractinian	  coral	  population	  is	  critical	  to	  the	  survival	  of	  productive	  coral	  reefs	  in	  the	  Caribbean.	  Following	  post-­‐bleaching	  mortality,	  ideally,	  corals	  progressively	  recover	  to	  their	  pre-­‐disturbance	   state	   (Gilmour	   et	   al.	   2013).	   Alternatively,	   however,	   the	   coral	   community	  composition	   changes,	   due	   to	   differential	   bleaching	   impact	   and	   reproductive	   success	   among	  coral	   species	   (Shenkar	   et	   al.	   2005;	   Obura	   2005),	   or	   reefs	   become	   colonized	   by	   algae	   and	  sponges	  due	  to	  corals	  inability	  to	  undergo	  recovery	  (Norström	  et	  al.	  2009;	  McClanahan	  2000).	  Post-­‐disturbance	   recovery	   relies	  on	   coral	   communities	   re-­‐growing	  and	   colonizing	   the	   reef	   via	  sexual	   recruitment	   (Pearson	   1981).	   In	   turn	   this	   process	   is	   determined	   by	   the	   diversity,	  abundance	  and	  size	  of	  surviving	  coral	  colonies,	  their	  ability	  to	  grow	  and	  reproduce,	  their	  ability	  of	   larva	   to	   settle	   and	   survive,	   as	  well	   as	   the	  post-­‐settlement	   survival	  of	   recruits	   (Tamelander	  2002;	  Baker	  et	  al.	  2008;	  Crabbe	  2009).	  Additionally,	  recovery	  can	  also	  be	  very	  site	  specific	  due	  to	   secondary	   disturbances	   affecting	   coral	   community	   dynamics,	   especially	   in	   the	   case	   of	  disturbances	   that	   undermine	   coral	   reproductive	   processes	   such	   as	   terrestrial	   runoff	   and	  overfishing	   (Burt	  et	  al.	  2008).	  Consequently,	   the	  extent	  and	  direction	  of	   coral	   recovery	   is	  not	  easily	   predictable.	   Thus,	   to	   improve	   our	   understanding	   of	   post-­‐bleaching	   recovery	   it	   is	  important	   to	   assess	   the	   bleaching	   impact	   on	   coral	   assemblages	   and	   their	   ability	   to	   sexually	  	   2	  reproduce	  within	  their	  given	  environment	  (Irizarry-­‐soto	  &	  Weil	  2009;	  Birrell	  et	  al.	  2005;	  Smith	  et	  al.	  2005).	  	  In	   this	   thesis,	   I	   focus	   on	   understanding	   post-­‐bleaching	   recovery	   among	   the	   most	   southern	  Caribbean	   coral	   reef	   systems	   that	   fringe	   the	   island	   of	   Tobago.	   Tobago’s	   coral	   reefs	   have	  undergone	   the	   same	   degradation	   trajectory	   as	   the	  majority	   of	   their	   Caribbean	   counterparts.	  Nor	   were	   they	   spared	   from	   undergoing	   mass	   bleaching	   during	   the	   Caribbean	   wide	   ocean	  thermal	   stress	   events	   in	   1998,	   2005	   and	  2010	   (M.	   Eakin	   et	   al.	   2010).	   In	   2010	   Tobago’s	   reefs	  were	   reported	   to	   undergo	   severe	   bleaching	   of	   up	   to	   29-­‐60%	   of	   colonies	   at	   observed	   sites	  (Alemu	  I	  &	  Clement	  2014).	  Though	  bleaching	  induced	  mortality	  was	  estimated	  at	  only	  2-­‐8%	  of	  corals	   (Alemu	   I	  &	   Clement	   2014),	   a	   disease	   outbreak	   recorded	   following	   the	   bleaching	   likely	  lead	   to	   further	  mortality	   (Alemu	   I	   2011).	   Currently,	   little	   is	   known	   about	   the	   post-­‐bleaching	  impact	  or	   recovery	  process	  among	  Tobago’s	  coral	  communities.	  Most	  data	  on	  the	  health	  and	  disturbance	  history	  of	  Tobago’s	  coral	  reefs	  is	  related	  to	  changes	  in	  percent	  benthic	  cover,	  with	  the	  exception	  of	  one	  study	  done	  on	  recruitment	  (on	  artificial	  substrate)	  and	  growth	  modelling	  (Mallela	   &	   Crabbe	   2009).	   Furthermore,	   due	   to	   continuous	   coastal	   development	   taking	   place	  along	   the	   island’s	   south-­‐western	   coast,	   concerns	   have	   arisen	   about	   the	   impact	   of	   increased	  terrestrial	   runoff	  on	   some	  of	   the	   island’s	   coastal	  ecosystems	   (Parkinson	  2010;	   Lapointe	  et	  al.	  2010;	   Mallela	   et	   al.	   2010).	   Undoubtedly,	   these	   terrestrial	   runoff	   flows	   may	   also	   shape	   the	  recovery	  trajectory	  of	  some	  of	  the	  island’s	  coral	  reefs.	  This	  thesis	  explores	  the	  impacts	  on	  and	  recovery	  of	  coral	  reef	  communities	  across	  three	  distinct	  reef	  systems	  adjacent	  to	  land	  characterized	  by	  different	  levels	  of	  settlement	  and	  infrastructure	  development.	   The	   first	   objective	   (Chapter	   2)	   is	   to	   quantify	   the	   juvenile	   communities	   across	  these	   different	   reef	   systems,	   as	   a	   high	   abundance	   of	   coral	   juveniles	   indicates	   that	   coral	  populations	   are	   able	   to	   reproduce	   sexually	   and	   recruit,	   and	   thus	   able	   to	   support	   recovery	  processes	   (Arnold	   2011;	  Huitric	  &	  Mcfield	   2000;	   Ritson-­‐williams	   et	   al.	   2009).	  Given	   that	   high	  levels	   of	   sedimentation	   impede	   coral	   growth,	   reproduction	   and	   recruitment	   (Fabricius	   2005;	  Miller	  et	  al.	  2000;	  Wittenberg	  &	  Hunte	  1992)	  a	  second	  objective	  is	  to	  investigate	  the	  rate	  and	  composition	   of	   sedimentation	   across	   the	   three	   distinct	   reef	   systems	   (Chapter	   2).	   The	   final	  objective	  of	  this	  study	  is	  to	  assess	  the	  impact	  and	  recovery	  of	  the	  bleaching	  event	  on	  species’	  population	  demographics	  by	  examining	  the	  changes	   in	  the	  size	  and	  frequency	  distributions	  of	  species’	  populations	  before	  and	  after	   the	  bleaching	  event	  and	   in	   the	   two	  years	   following	   the	  bleaching.	  	  	   3	  Tobago’s	  coral	  reefs	  represent	  some	  of	  the	  most	  understudied	  reefs	  among	  Caribbean	  (Alemu	  I	  &	  Clement	  2014).	  This	  study	   is	  one	  of	  the	  few	   in	  the	  region	  that	   is	  able	  to	  examine	  temporal	  changes	  following	  a	  mass-­‐bleaching	  event.	  It	  provides	  valuable	  baseline	  datasets	  and	  analysis	  of	  juvenile	  communities,	  sediment	  composition	  and	  deposition	  rates,	  and	  corals’	  population	  size	  structure	   among	   some	   of	   Tobago’s	   key	   reef	   systems.	   Additionally,	   this	   research	   supports	   a	  growing	  body	  of	  research	  using	  population	  size	  structure	  to	  research	  effects	  of	  disturbance	  on	  coral	  population	  dynamics.	  	  	   4	  Chapter	  2.	   	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  An	  assessment	  of	  coral	   juvenile	  community	  and	  sediment	  deposition	  among	  three	  major	  reef	  systems	  in	  Tobago	  2.1 	  Introduction	  Tropical	   coral	   reefs	   are	   highly	   productive	   ecosystems	   that	   act	   as	   important	   natural	   assets,	  which	  provide	  vital	  goods	  and	  services	  to	  coastal	  communities.	  This	  is	  especially	  true	  for	  small	  island	   states	   in	   the	   Caribbean,	   such	   as	   Tobago,	  which	   rely	   heavily	   on	   healthy	   coral	   reefs	   for	  protecting	  coastal	  property	  from	  storms,	  supporting	  near-­‐shore	  fisheries	  and	  attracting	  tourism	  (Burke	   et	   al.	   2008).	   Despite	   their	   value,	   in	   the	   last	   three	   decades	   Caribbean	   reefs	   have	  experienced	   a	   dramatic	   decline	   in	   coral	   cover,	   from	   approximately	   50%	   to	   10%,	   and	  consequently	  have	  gone	  from	  being	  coral-­‐dominated	  to	  algae-­‐dominated	  reefs	  (Côté	  &	  Darling	  2010).	  This	  shift	  has	  been	  attributed	  to	  historical	  overfishing	   leading	  to	  the	  reduction	  of	   large	  sized	  grazing	  herbivore	  fish	  populations,	  die-­‐off	  of	  grazing	  Diadema	  urchins	  in	  the	  early	  1980s,	  increased	  sedimentation	  and	  nutrient	  enrichment,	  and	  the	  spread	  of	  coral	  disease	  (Hughes	  &	  Connell	   1999;	   Norström	   et	   al.	   2009;	   Jackson	   et	   al.	   2014).	   In	   the	   last	   two	   decades,	   mass	  bleaching	  events	  are	  often	  followed	  by	  substantial	  coral	  mortality,	  have	  become	  an	  additional	  threat,	   further	   testing	   the	   resilience	   of	   the	   remaining	   Caribbean	   coral	   communities	   (Hoegh-­‐Guldberg	  1999;	  Eakin	  et	  al.	  2010).	  Bleaching	  events	  are	  irrefutably	  linked	  to	  ocean	  warming	  as	  a	  result	  of	  global	  climate	  change,	  and	  there	  is	  strong	  evidence	  that	  the	  frequency	  and	  intensity	  of	  mass	  coral	  bleaching	  events	  are	  likely	  to	  increase	  in	  the	  near	  future	  (Donner	  et	  al.	  2005;	  Hoegh-­‐Guldberg	  et	  al.	  2007).	  Coral	   communities’	   ability	   to	   recover	   following	   disturbances	   like	   bleaching	   events	   will	  determine	   the	   long-­‐term	   survival	   of	   these	   ecosystems.	   Maintenance	   of	   coral	   populations	  depends	  critically	  on	  corals'	  sexual	  production	  of	  larvae,	  recruitment	  and	  capability	  of	  recruits	  to	  survive	  and	  grow	  into	  adult	  colonies	  (Hughes	  &	  Tanner	  2000;	  Arnold	  2011).	  A	  variety	  of	  biotic	  and	   abiotic	   factors	   across	   time	   and	   space	   can	   impact	   the	   success	   of	   recruitment	   and	   post-­‐settlement	  survival,	  including	  the	  size	  and	  health	  of	  parent	  colonies,	  availability	  and	  complexity	  of	  substrate,	  competition,	  predation	  and	   light	  availability	  (Sammarco	  1985;	  Tamelander	  2002;	  Babcock	  &	  Smith	  2000).	  Coral	  larvae	  and	  recruits	  thus	  tend	  to	  be	  very	  susceptible	  to	  mortality	  (Trapon	   et	   al.	   2013).	   In	   comparison	   to	   Indo-­‐Pacific	   reefs,	   post-­‐disturbance	   recovery	   of	  Caribbean	  coral	  reefs	  has	  generally	  been	  low	  (Baker	  et	  al.	  2008).	  According	  to	  the	  literature	  this	  disparity	  is	  driven	  by	  high	  algal	  cover	  and	  land-­‐based	  marine	  pollution,	  making	  Caribbean	  reefs	  	   5	  hostile	  environments	  for	  recruitment	  and	  juvenile	  survival	  (Arnold	  &	  Steneck	  2011;	  Jackson	  et	  al.	  2014).	  	  Increased	  terrestrial	  runoff	  along	  Caribbean	  coastlines,	  has	  been	  a	  growing	  problem	  in	  a	  region	  where	  continuous	  coastal	  development	  and	  agriculture	  intensification	  is	  taking	  place	  (Burke	  et	  al.	  2011;	  Begin	  2012;	  Hernandez	  et	  al.	  2009).	  Regardless	  of	  whether	  sediments	  settle	  or	  remain	  in	  suspension,	  they	  can	  affect	  all	  growing	  stages	  of	  a	  coral’s	  lifecycle.	  Turbidity	  decreases	  light	  penetration,	  reducing	  coral	  growth	  (Cortes	  &	  Risk	  1985;	  Fabricius	  et	  al.	  2003),	  whilst	  particles	  deposited	  on	  coral	  colonies	  weakens	  their	  health	  and	  in	  excess	  can	  smother	  corals	  (Hernandez	  et	  al.	  2009;	  Erftemeijer	  et	  al.	  2012).	  Corals’	  reproductive	  and	  recruitment	  stages	  are	  believed	  to	  be	  especially	  sensitive	  to	  sedimentation	  levels,	  as	  reefs	  with	  high	  sedimentation	  levels	  have	  low	  recruitment	  and	  juvenile	  densities	  (Fabricius	  2005).	  It	  is	  likely	  that	  larvae	  and	  recruits	  are	  very	  susceptible	  to	  being	  smothered	  or	  damaged	  by	  sediment,	  and	  particles	  covering	  hard	  surfaces	  inhibit	  the	  settling	  of	  larvae	  (Babcock	  &	  Smith	  2000;	  Torres	  &	  Morelock	  2002;	  Fabricius	  2005).	  Additionally,	  recent	  studies	  indicate	  that	  sedimentation	  suppresses	  herbivory	  and	  promotes	  the	  growth	  of	   algae,	   thereby	   reducing	   settling	   space	   for	   recruitment	   (Goatley	  &	  Bellwood	  2013).	  Sedimentation-­‐based	   stress	   on	   coral	   depends	  on	  multiple	   physical	   factors	   such	   as	   the	   timing	  and	   amount	   of	   sediment	   deposition,	   reef	   depth	   and	   distance	   to	   shore,	   ocean	   currents	   and	  waves,	  and	  sediment	  grain	  size	  and	  composition	  (Hernandez	  et	  al.	  2009;	  Abdullah	  et	  al.	  2011;	  Waheed	  et	   al.	   1998).	   The	   size	  of	   sediment	  particles	   is	   a	   particularly	   important	   property	   as	   it	  determines	  the	  transport	  mode	  and	  potential	  impact	  on	  coral	  communities.	  For	  instance,	  corals	  can	  remove	  sand-­‐sized	  grains	  with	  more	  ease	  than	  very	  fine	  sediment	  like	  silt/clay	  (Weber	  et	  al.	  2006).	  Fine	  sediment	  also	  tends	  to	  carry	  greater	  concentrations	  of	  toxic	  contaminants	  that	  can	  be	  lethal	  to	  corals	  (Fabricius	  et	  al.	  2003;	  Rogers	  1990).	  Most	  of	  Tobago’s	  reefs	  share	  the	  same	  history	  of	  degradation	  as	  the	  rest	  of	  Caribbean	  and	  thus	  are	   characterized	  by	   low	  coral	   cover	   that	   ranges	   from	  10-­‐30%	   (Mallela	   et	   al.	   2010).	  Recently	  there	   have	   been	   rising	   concerns	   that	   increased	   terrestrial	   runoff,	   due	   to	   continuous	   urban	  developments,	  may	  play	  an	  important	  role	  in	  shaping	  the	  trajectory	  of	  some	  the	  island’s	  coral	  reefs	  (Parkinson	  2010;	  Lapointe	  et	  al.	  2010;	  Mallela	  et	  al.	  2010).	  Data	  about	  sedimentation	  on	  Tobago’s	  reefs	  is	  limited	  to	  one	  study,	  however	  ,	  which	  found	  sediment	  rates	  along	  11	  separate	  reef	  systems	  to	  be	  all	  below	  levels	   	  that	  tend	  to	  stress	  corals	  (Mallela	  et	  al.	  2010).	  No	  further	  investigation	  was	  conducted	  into	  the	  sediments’	  particle	  size	  distribution	  or	  composition,	  which	  can	  provide	  key	  information	  such	  as	  the	  origin	  of	  sediments	  (Begin	  2012)	  	   6	  Tobago’s	   reefs	  were	   affected	  by	   the	   regional	  mass	  bleaching	   events	   in	   1998,	   2005	   and	  most	  recently	   in	  2010.	   	  The	  heat	  stress	   in	  2010	  lasted	  almost	  5	  months,	  during	  which	  29	  to	  60%	  of	  coral	   communities	   among	   Tobago’s	   reefs	   experienced	   severe	   bleaching.	   While	   bleaching	  induced	  mortality	  was	  about	  2-­‐8%	  (Alemu	  I	  &	  Clement	  2014),	  the	  bleaching	  event	  was	  followed	  by	  a	  disease	  outbreak	  possibly	  causing	  further	  mortality	  (Alemu	  I	  2011).	  Knowledge	  about	  coral	  recruitment	   processes	   on	   Tobago’s	   coral	   reefs	   comes	   exclusively	   from	   studies	   assessing	  recruitment	  rates	  on	  artificial	  substrate,	  which	  quantify	  the	  early	  settling	  stages	  of	  coral	  recruits	  (Babcock	  &	  Smith	  2000).	  These	  studies	  have	   indicated	  that	  the	  number	  of	  recruits	  settling	  on	  some	  of	  Tobago’s	  reefs	  has	  decreased	  over	  last	  two	  decades	  (Mallela	  &	  Crabbe	  2009).	  However	  no	  research	  has	  yet	  focused	  on	  the	  early	  life-­‐stages	  of	  corals	  on	  natural	  substrata,	  which	  can	  be	  done	  by	  assessing	  the	  juvenile	  populations.	  As	  juveniles	  are	  4-­‐7	  years	  old,	  they	  serve	  as	  a	  proxy	  measurement	  of	  the	  integrated	  outcome	  of	  corals	  sexual	  reproduction,	   larval	  settlement,	  and	  recruits	  post-­‐settlement	  survivorship	  within	  a	  multiple	  year	  time	  period	  (Vermeij	  et	  al.	  2011)	  Given	  that	  both	  bleaching	  and	  disease	  have	  negative	  impacts	  on	  the	  fecundity	  of	  coral	  colonies	  (Weil	  &	  Vargas	  2009)	  and	  that	  recruitment	  and	  post	  settlement	  survival	  often	  can	  be	  adversely	  impacted	   by	   sediment	   deposition,	   the	   objectives	   of	   this	   study	  were	   to:	   (1)	   quantify	   juvenile	  densities	   and	   taxa	   composition,	   and	   (2)	   estimate	   the	   rate	   and	   composition	   of	   sedimentation	  across	  three	  distinct	  reef	  system	  adjacent	  to	  land	  with	  different	  levels	  of	  urbanization.	  2.2 Methods	  2.2.1 Study	  area	  Tobago	  is	  a	  300	  km2	   large	  hilly	   island	  of	  volcanic	  origin	  that	   is	  surrounded	  by	  fringing	  shallow	  reefs.	  These	   reefs	  evolved	  under	   the	   influence	  of	  nutrient	  and	  sediment	   rich	   inputs	   from	  the	  Orinoco	  and	  Amazon	  Rivers,	  and	  consequently	  Tobago’s	  coral	  communities	  have	  lower	  species	  diversity	   in	   comparison	   to	  other	  Caribbean	   reefs	   (Moses	  &	  Swart	  2006;	   Lapointe	  et	   al.	   2003;	  Potts	   et	   al.	   2004).	  Whilst	   the	  majority	   of	   Tobago	   is	   covered	   in	   forest	   and	   shrubs	   lands,	   the	  south-­‐western	   part	   of	   the	   island	   has	   undergone	   significant	   urbanization	   and	   agricultural	  development.	  The	  study	  was	  conducted	  on	  three	  major	  reef	  systems,	  chosen	  for	  local	  perceived	  importance	  and	  their	  potentially	  different	  exposure	  to	  sedimentation	  due	  to	  coastal	  land	  uses.	  These	   included	   Caribbean	   Sea	   facing	   Buccoo	   and	   Culloden	   Reef,	   and	   Atlantic	   Sea	   facing	  Speyside	  Reef	  (Figure	  1).	  	  	  	   7	  	  Figure	  1.	  Map	  of	  Tobago	  and	  location	  of	  studied	  reef	  systems	  and	  sites	  Buccoo	  Reef	  is	  comprised	  of	  five	  large,	  sloping	  reef	  flats	  covering	  about	  7	  km2	  and	  is	  Tobago’s	  only	   official	   marine	   protected	   park	   (since	   1973).	   Despite	   being	   a	   major	   economic	   asset,	  attracting	  over	  10,000	  visitors	  annually,	  this	  park	  has	  received	  little	  protection	  or	  safeguarding	  against	  land-­‐based	  pollution	  (Lapointe	  et	  al.	  2010).	  In	  the	  last	  three	  decades,	  Land	  adjacent	  to	  this	   reef	   has	   experienced	   rapid	   urbanization	   and	   untreated	   sewage	   and	   uncontrolled	   storm	  waters	   drain	   into	   Buccoo	   Bay	   (Potts	   et	   al	   2004,	   Lapointe	   et	   al.	   2010,	   Parkinson	   2010).	   The	  horseshoe	  shaped	  reef	  of	  Culloden	  covers	  ~5.8	  ha	  and	   is	   located	   in	  a	  remote	  bay	  surrounded	  mostly	  by	  forested	  hills	  (Laydoo	  1991).	  The	  bay	  is	  accessible	  via	  a	  dirt	  road	  and	  human	  activities	  are	   limited	   to	   occasional	   recreational	   divers,	   artisanal	   fishing	   and	   boat	   anchorage.	   Speyside	  features	  a	  large	  network	  of	  fringing	  reefs	  along	  small	  islands	  and	  rocky	  outcrops	  on	  the	  north-­‐eastern	  side	  of	  the	  island.	  Like	  Culloden,	  Speyside	  coastal	  lands	  remain	  relatively	  undeveloped	  comprising	  of	  a	  hilly	  forested	  landscape	  apart	  from	  Speyside	  village	  (a	  fishing	  community)	  and	  2	  medium-­‐sized	  hotels.	  	  	   8	  Table	  1.	  Information	  on	  the	  six	  reef	  sites	  where	  juvenile	  assessment	  was	  undertaken	  and	  sediment	  traps	  were	  installed	  between	  8-­‐12m	  depths	  Reef	  system/site	  	   	   Coordinates	   %Coral	  cover	  /	  Dominant	  coral	  taxa	   Distance	   to	   shore	   and	   potential	   sources	   of	  sediment	  (Parkinson	  2010)	  Buccoo	  Reef	   	   	   	  Outer	  Buccoo	  (OB)	  	   11°11.371’	  N	  60°49.412’	  W	   20.24±5.83	  M.faveolata,	  S.	  siderea,	  C.Natans,	  Agaricia	  spp.	  ~1.5km	   from	   mainland.	   Two	   streams	   and	  urban	   wastewaters	   discharge	   in	   Buccoo	   bay	  (Parkinson	  2010).	  	  Western	  Buccoo	  (WB)	  	   11°11.043’	  N	  60°50.782’	  W	  	  13.64±8.38	  M.faveolata,	  D.strigosa,	  C.natans,	  S.	  siderea	  	  Culloden	  Reef	   	   	   	  Culloden	  East	  (CE)	  	   11°14.833’	  N	  60°45.086’	  W	   12.19±5.23	  M.faveolata,	  M.cavernosa,	  A.	  palmate,	  D.	  strigosa	  ~280m	  from	  mainland.	  One	  stream	  terminates	  in	  this	  bay.	  Dirt	  road	  access	  to	  the	  bay.	  Culloden	  West	  (CW)	  	   11°14.982’	  N	  60°44.904’	  W	   16.02±7.57	  M.	  faveolata,	  D.	  strigosa,	  C.natans,	  M.	  cavernosa	  	  Speyside	  Reef	   	   	   	  Black	  Jack	  Hole	  (BJH)	  	   11°18.072’	  N	  60°31.223’	  W	   13.01±7.12	  M.faveolata,	  S.siderea,	  P.astreoides,	  M.alcicornis	  ~1km	  from	  mainland.	  Paved	  road	  all	  along	  the	  coast	   line.	   Doctors	   River	   and	   several	   small	  streams	  drain	  into	  Speyside	  bay.	  Patches	  forest	  have	  been	  Angel	  Reef	  (AR)	  	   11°17.688’	  N	  60°30.001’	  W	   46.42±14.67	  M.mirabilis,	  M.	  faveolata,	  M.cavernosa,	  C.natans	  increasing	  cleared	  for	  growing	  crops	  	  Two	  study	  sites	  were	  established	  at	  each	  reef	  system	  (Figure	  1,	  Table	  1).	  At	   the	  six	   reef	  sites	  juvenile	  population	  and	  sedimentation	  assessments	  were	  carried	  out	  from	  May	  to	  June	  in	  2013	  using	  SCUBA.	  All	  data	  were	  collected	  at	  the	  six	  sites	  between	  the	  depths	  of	  8-­‐12m	  during	  the	  months	  of	  May	  and	  June	  2013,	  the	  end	  of	  the	  dry	  season	  in	  Tobago.	  	  2.2.2 Sedimentation	  assessment	  	  Sediment	  traps	  were	  deployed	  for	  roughly	  one	  month,	  following	  published	  methods	  (English	  et	  al.,	   1977,	  Hill	   and	  Wilkinson	  2004,	   and	   Storlazzi	   et	   al	   2009).	   Each	   sediment	   trap	   consisted	  of	  three	  cylindrical	  5-­‐cm-­‐wide	  and	  20-­‐cm-­‐long	  PVC	  pipes	  with	  wire	  mesh	  at	   the	  top	  to	  allow	  for	  sediment	   deposition	   but	   deter	   large	  marine	   organisms.	   Zip	   ties	  were	   used	   to	   fix	   three	   pipes	  onto	   a	   1	   m	   tall	   metal	   rod	   to	   form	   a	   trap	   set.	   At	   each	   site	   three	   sediment	   trap	   sets	   were	  hammered	   into	   non-­‐living	   reef	   substrate,	   leaving	   the	   pipe	   traps	   about	   0.75	   m	   above	   the	  substrate.	  Traps	  were	  spaced	  ~	  30	  m	  from	  each	  other,	  maintaining	  adequate	  coverage	  of	   the	  reef	  site.	  	  	   9	  From	  the	  14th	  to	  26th	  of	  May	  2013,	  three	  trap	  sets	  (9	  tubes)	  were	  set	  up	  at	  each	  site	  –	  a	  total	  of	  54	  individual	  tubes.	  After	  30-­‐37	  days	  (20-­‐27th	  of	  June	  2013)	  the	  sediment	  traps	  were	  recovered	  by	  capping	  each	  pipe	  underwater	  before	  bringing	  them	  up	  to	  the	  surface.	  A	  total	  of	  45	  tubes	  were	   recovered.	   Only	   6	   and	   5	   tubes	   were	   recovered	   safely	   from	   Culloden	   East	   and	   West,	  respectfully,	  as	  the	  rest	  were	  either	  dislodged	  or	  disappeared.	  After	  collection	  the	  content	  from	  each	   tube	   washed	   out	   with	   distilled	   water	   and	   filtered	   through	   a	   funnel	   and	   filter	   paper.	  Sediment	   samples	  were	   dried	   in	   an	   oven	   at	   60°C	   for	   three	   hours,	   allowed	   to	   cool	   and	  were	  sealed	  in	  zip-­‐lock	  bags	  for	  transport	  to	  Trinidad.	  At	  the	  laboratory	  the	  sediments	  samples	  were	  rinsed	  out	  twice	  with	  distilled	  water	  to	  remove	  salts	  and	  oven-­‐dried	  at	  105°C	  overnight.	  The	  dry	  sediment	   was	   transferred	   from	   filter	   paper	   into	   a	   petri	   dish	   with	   a	   fine	   brush	   and	   kept	   in	  desiccator	   overnight.	   The	   total	   weight	   of	   the	   sediment	   collected	   from	   each	   pipe	   trap	   was	  determined	  using	  an	  analytical	  balance.	  Sedimentation	  rates	  (mg	  cm-­‐2d-­‐1)	  were	  determined	  by	  dividing	  the	  dry	  weight	  (in	  mg)	  by	  the	  area	  of	  the	  sediment	  trap	  aperture	  width	  (in	  cm2)	  over	  the	  duration	  (in	  days)	  the	  sediment	  trap	  was	  installed	  at	  each	  site	  (Abdullah	  et	  al.	  2011).	  	  Sediment	   composition	   was	   analyzed	   employing	   loss	   on	   ignition	   (LOI)	   methods	   to	   determine	  what	   fraction	   of	   the	   sediment	   was	   composed	   of	   organic	   and	   carbonate	   matter,	   leaving	   the	  remaining	  non-­‐carbonate	  material	  as	  the	  terrigenous	  fraction	  of	  the	  sample.	  It	  is	  important	  to	  note	  that	  this	  method	  is	  predominantly	  used	  by	  paleolimnologists	  on	  lake	  core	  samples	  and	  is	  known	  to	  provide	  rough	  estimates	  of	  sediment	  composition	  (Santisteban	  et	  al.	  2004).	  As	  many	  individual	  pipe	  traps	  contained	  <2	  grams	  of	  sediment,	  samples	  from	  each	  trap	  set	  were	  pooled	  to	  form	  a	  composite	  sample.	  About	  3	  grams	  of	  each	  composite	  sediment	  sample	  were	  placed	  in	  a	  pre-­‐weighed	  ceramic	  crucible.	  To	  determine	  dry	  sediment	  weight	  all	  samples	  were	  dried	  in	  an	  oven	  at	  105°C	  for	  three	  hours	  (to	  remove	  moisture)	  and	  cooled	   in	  a	  desiccator	  before	  weight	  was	  recorded.	  Afterwards	  organic	  matter	  was	  combusted	  from	  the	  samples	  by	  placing	  them	  in	  a	  furnace	  at	  550°C	  for	  six	  hours,	  cooling	  in	  a	  desiccator	  and	  then	  weighed	  (Brooks	  et	  al.	  2007).	  This	   procedure	   was	   followed	   by	   ashing	   samples	   at	   925°C	   to	   determine	   carbonate	   content	  (Luczak	  and	  Kupka	  1997).	  The	  percent	  of	  organic	  and	  carbonate	  content	  was	  calculated	  using	  the	  following	  formula	  (Heiri	  et	  al.	  2001):	  	  Organic  content  (%) = ((DW105  –DW550)/DW105)×100	  Carbonate  content  (%) = ((DW550–DW950)/DW105)  ×   𝑐𝑜? 𝑐𝑜?×100  	  Where	  DW105	   is	   the	  dry	  weight	  of	   the	   sample	  before	   combustion,	  DW550	   is	   the	  dry	  weight	  after	  combustion	  at	  550°C,	  DW950	  is	  the	  dry	  weight	  after	  combustion	  at	  925°C	  and	  CO3/CO2	  is	  the	   ratio	   between	   the	  molecular	  weights	   of	   CO3	   (60u)	   and	   CO2	   (44u)	   being	   1.36	   (as	   cited	   in	  	   10	  Veres	   2002).	   A	   total	   of	   16	   composite	   sub-­‐samples	   were	   ashed.	   Particle	   size	   analysis	   was	  conducted	  using	  the	  wet	  sieving	  method	  (Syvitski	  2007),	  using	  ~0.5	  grams	  from	  each	  composite	  sediment	   sample,	  which	  were	   separated	   into	   five	   fractions	   (Table	  2).	  Before	   sieving,	   samples	  were	  dried	  for	  3	  hours	  to	  remove	  all	  moisture	  and	  weighed.	  Table	  2.	  Sieved	  size	  groups	  and	  sediment	  classes	  according	  to	  the	  Wentworth	  size	  class	  system.	  	  Grain	  size	  groups	  sieved	   Sediment	  type	  >500	  μm	   coarse	  sand	  500-­‐250μm	   medium	  sand	  250-­‐125μm	   fine	  sand	  125-­‐63μm	   very	  fine	  sand	  <63μm	   silt-­‐clay	  	   	  Dry	   samples	  were	   then	  emptied	   into	   a	   glass	  beaker	   filled	  with	  distilled	  water	   and	   left	   for	   an	  hour	   for	   the	   sediment	   particles	   to	   disaggregate.	   The	   content	   was	   then	   poured	   through	   the	  stacked	  sieve	  set,	  in	  the	  order	  of	  mesh	  sizes	  500,	  250,	  125	  and	  63	  microns.	  Care	  was	  taken	  to	  ensure	  that	  no	  particles	  remained	  in	  the	  beaker	  by	  washing	  with	  a	  fine	  tip	  squeeze	  bottle.	  	  Each	  sieve	  was	  carefully	  separated,	  placed	  in	  a	  ceramic	  beaker	  to	  retain	  the	  washings,	  and	  set	  to	  dry	  in	  an	  oven	  for	  3	  hours	  at	  100°C	  until	  all	  water	  was	  evaporated.	  Once	  dry,	  the	  sediment	  particles	  were	  moved	  from	  the	  sieves	   into	  the	  ceramic	  dish	  using	  a	  fine	  brush	  to	  ensure	  every	  particle	  was	   retained.	   The	   weight	   of	   each	   sediment	   sample	   fraction	   was	   determined	   using	   a	   high	  precision	  analytical	  balance.	  	  2.2.3 Juvenile	  community	  survey	  Juvenile	  coral	  colonies,	  hereafter	  referred	  to	  as	   juveniles,	  were	  enumerated	  and	   identified	  by	  carefully	   scrutinizing	   the	   reef	   benthos	   in	   sixty	   randomly-­‐placed	   0.25m²	   quadrats	   per	   site	  (Carpenter	  &	  Edmunds	  2006).	  Quadrats	  were	  dropped	  from	  a	  height	  of	  3m	  above	  the	  substrate	  while	  ensuring	  that	  no	  quadrats	  overlapped,	  as	  per	  McClanahan	  (2000).	  A	  pilot	  survey	  of	   this	  method	  was	  undertaken	   in	  advance	   to	  determine	   sample	   sizes	  by	  plotting	   the	   running	  mean	  against	  the	  number	  of	  quadrats	  surveyed	  (Edmunds	  et	  al.	  1998).	  	  Juveniles	  were	  defined	  as	  any	  colony	  of	  large-­‐sized	  coral	  taxa	  visible	  to	  the	  naked	  eye	  with	  a	  maximum	  diameter	  of	  5	  cm	  (e.g.,	  Montastraea	   spp.,	   Diploria	   spp.,	   and	   Siderastrea	   spp.,	   etc.)	   or	   of	   2	   cm	   in	   diameter	   for	   small	  sized	  coral	  taxa	  (Porites	  spp.,	  Favia	  fragum,	  and	  Agaricia	  spp.,	  etc.).	  This	  distinction	  was	  made	  as	   small-­‐sized	   corals	   tend	   to	   be	   sexually	  mature	   adults	   once	   larger	   than	   2	   cm	   (Chiappone	  &	  Sullivan	  1996;	  Miller	  et	  al.	  2000;	  Irizarry-­‐soto	  &	  Weil	  2009).	  Small	  living	  coral	  fragments	  that	  are	  not	   a	   product	   of	   sexual	   reproduction	   were	   omitted	   from	   the	   count	   (Trapon	   et	   al.	   2013).	  	  Juveniles	  were	  identified	  to	  the	  species	  level	  when	  possible	  and	  to	  the	  genus	  when	  they	  were	  	   11	  taxonomically	   difficult	   to	   distinguish.	   Care	   was	   taken	   to	   inspect	   cryptic	   habitats	   such	   as	  crevasses	  and	  underneath	  shelves.	  If	  more	  than	  three	  quarters	  of	  the	  benthic	  cover	  within	  the	  quadrats	   was	   covered	   by	   non-­‐settling	   substrate,	   e.g.	   sand	   or	   living	   coral,	   the	   quadrat	   was	  moved	  to	  the	  side	  or	  the	  aforementioned	  process	  was	  repeated	  (Edmunds	  et	  al.	  1998).	  	  2.2.4 Statistical	  analysis	  	  All	   summary	   statistics	   and	  analyses	  were	  done	   in	  R	   version	  2.15.1.	  All	   juvenile	   and	   sediment	  data	  were	  tested	  for	  normality	  using	  the	  Shapiro-­‐Wilk	  test	  and	  homogeneity	  of	  variance	  using	  graphical	  methods.	  Sedimentation	   rate	  data	  were	   found	   to	  be	  normally	  distributed,	  although	  juvenile	  data	  did	  not	  follow	  a	  normal	  destruction.	   	  Differences	  in	  sedimentation	  rates	  per	  site	  were	   tested	   employing	   a	   one-­‐way	  ANOVA,	   followed	   by	   a	   Tukey	  HSD	   post-­‐hoc	   test	   to	   detect	  pairwise	   differences.	   Differences	   in	   juvenile	   densities	   per	   site	   were	   tested	   using	   the	   non-­‐parametric	  methods.	  A	  Kruskal-­‐Wallis	  test	  was	  employed	  followed	  by	  Dunn’s	  post-­‐hoc	  pairwise	  test	   (using	  packages	  multcomp 1.2-­‐17	   and	  coin 1.0-­‐23)	   (Miller	  et	  al.	  2000).	  Correspondence	  analysis	  (CA)	  was	  performed	  on	  eight	  genus	  groups	  found	  at	  each	  site,	  after	  removing	  all	  rare	  taxa,	   to	   explore	   the	   distribution	   of	   species	   composition	   across	   all	   sites	   (Irizarry-­‐soto	   &	  Weil	  2009)	  using	  ordination	  analysis	  tools	  in	  the	  Vegan	  Package	  in	  R	  software.	  	  2.3 Results	  2.3.1 Juvenile	  density	  and	  composition	  A	  total	  of	  428	  of	  juveniles	  between	  the	  sizes	  of	  ≥0.5	  to	  ≤5	  cm	  were	  counted	  over	  a	  total	  area	  of	  90	  m2	  (15	  m2	  at	  each	  site).	  Overall,	  mean	  juvenile	  density	  was	  5.4±	  6.3	  m-­‐2	  and	  ranged	  from	  0	  to	  32	  m-­‐2.	  The	  juvenile	  distribution	  was	  patchy	  with	  20-­‐49%	  of	  the	  0.25	  m2	  quadrats	  featuring	  no	  juvenile	  colonies	  across	  the	  assessed	  sites.	  The	  Buccoo	  Reef	  sites	  had	  the	  lowest	  abundance	  of	  juveniles,	  number	  of	  taxa	  and	  diversity	  (Table	  3).	  	  Table	  3.	  Coral	  juvenile	  data	  per	  site	  (15m2	  	  surveyed	  area).	  Site	   No.	  of	  juveniles	  	   Density(±SD)m-­‐2	   No.	  of	  taxa	  	   Shannon	  H'	  Outer	  Buccoo	   59	   3.9±	  4.8	   5	   1.8	  Western	  Buccoo	   50	   3.3±	  4.2	   6	   3.7	  Culloden	  East	   64	   4.3±	  4.2	   7	   4.2	  Culloden	  West	   99	   6.6±	  7.7	   8	   5.6	  Black	  Jack	  Hole	   69	   4.6±	  5.9	   8	   5.4	  Angel	  Reef	   146	   9.7±	  7.9	   10	   4.0	  	  	   12	  Whilst	  Angel	   Reef	   had	   the	  highest	   abundance	   and	   richness	   in	   taxa,	   Culloden	  West	   and	  Black	  Jack	  Hole	  had	  the	  most	  even	  diversity	  according	  to	  the	  Shannon	  Diversity	  Index.	  Mean	  juvenile	  abundances	  were	   similar	   between	  most	   sites	   except	   at	   Angel	   Reef,	  where	   abundances	  were	  significantly	   higher	  when	   compared	   to	   all	   other	   sites	   (Dunn’s	   test,	   p	   <	   0.05)	   except	   at	  West	  Culloden.	  Table	  4.	  Number	  of	  juvenile	  taxa	  found	  at	  each	  site.	   	   	  	  	  	  	  	  	  The	   majority	   of	   the	   juvenile	   colonies	   belonged	   to	   brooding	   taxa	   (72.9%),	   such	   as	   from	   the	  genera	  Agaricia,	  Porites,	  Madracis,	  Scolymia,	  and	  Favia.	  Broadcasting	  juvenile	  taxa	  represented	  the	  minority	  (27.1	  %)	  such	  as	  Siderastrea,	  Diploria,	  Montastrea	  and	  Colpophyllia.	  Brooding	  taxa	  dominated	   at	   Outer	   Buccoo,	   Culloden	   East	   and	   Angel	   Reef	   while	   West	   Culloden,	   Western	  Buccoo	  and	  Black	  Jack	  Hole	  sites	  had	  similar	  proportions	  of	  brooders	  and	  broadcasters,	  due	  to	  moderate	   abundances	   of	   Diploria	   and	   Siderastrea	   juveniles	   (Table	   4,	   Figure	   2).	   The	  predominant	  species	  were	  Agaricia	  spp.	  (45.4%)	  followed	  by	  Siderastrea	  spp.	  (12.7%),	  Diploria	  spp.	  (9.7%),	  and	  Porites	  spp.	  (8%).	  The	  remaining	  species	  altogether	  only	  represented	  24.2%	  of	  all	  juveniles.	  Agaricia	  spp.	  was	  most	  predominant	  at	  most	  sites	  with	  the	  exception	  of	  Black	  Jack	  Hole,	  where	  Porites	  spp.	  (mainly	  P.	  astreoides)	  dominated.	  	  Taxa/site	   OB	  	   WB	   CE	   CW	   BJH	   AR	  Agaricia	  sp.	   51	   24	   34	   26	   1	   85	  Colpophyllia	  sp.	   -­‐	   2	   -­‐	   -­‐	   -­‐	   -­‐	  Diploria	  sp.	   -­‐	   7	   8	   16	   15	   1	  Eusmilia	  fastigiata	  	   -­‐	   1	   -­‐	   -­‐	   -­‐	   -­‐	  Favia	  fragum	  	   -­‐	   -­‐	   6	   1	   8	   3	  Madracis	  decatis	  	   -­‐	   -­‐	   4	   3	   4	   5	  Madracis	  mirabillis	  	   -­‐	   -­‐	   -­‐	   -­‐	   -­‐	   27	  Montastraea	  cavernosa	  	   -­‐	   -­‐	   1	   5	   2	   -­‐	  Montastraea	  sp.	   2	   -­‐	   -­‐	   -­‐	   5	   5	  Mycetophyllia	  sp.	  	   1	   -­‐	   -­‐	   -­‐	   -­‐	   -­‐	  Porites	  astreoides	   1	   -­‐	   -­‐	   3	   27	   8	  Porites	  porites	  	   -­‐	   -­‐	   -­‐	   -­‐	   -­‐	   1	  Scolymia	  sp.	   -­‐	   2	   9	   19	   -­‐	   2	  Siderastrea	  sp.	   4	   14	   2	   26	   7	   9	  Total	   59	   50	   64	   99	   69	   146	  	   0	  20	  40	  60	  80	  100	  120	  140	  160	  OB	  	   WB	   CE	   CW	   BJH	   AR	  Total	  juveniles	  counted	  Reproduc?on	  mode	  Broadcast	  Brooder	  Figure	  2	  Proportion	  of	  counted	  juveniles	  that	  were	  produced	  from	  broadcasting	  or	  brooding	  reproductive	  strategies	  	   13	  	   	  Figure	  3.	  Relative	  abundance	  of	  major	  taxa	  (genus)	  groups	  at	  each	  site	   for	   (A)	   juvenile	  population	  based	  from	  count	  data	  and	  (B)	  adult	  population	  based	  on	  percent	  cover	  assessed	  in	  2013	  (see	  Chapter	  3	  for	  data	  collection	  methods)	  The	   assessed	   juvenile	   taxa	   composition	   differed	   greatly	   from	   the	   adult	   coral	   community	  composition	  (Figure	  3).	  Dominant	  adult	  coral	  species,	  like	  Montastrea	  spp.,	  were	  unrepresented	  in	   the	   juvenile	   sample	   counted.	   On	   the	   other	   hand	   adult	   coral	   from	   the	   Agaricia	   genus	  represent	   less	   than	   5%	   of	   adult	   coral	   cover	   at	   each	   site,	   despite	   being	   the	   most	   dominant	  juvenile	  group.	  The	  small-­‐sized	  brooding	  Scolymia	  spp.,	  had	  moderate	  abundances	  of	  juveniles,	  mainly	   at	   Culloden	   sites.	   The	   abundance	   of	   Siderastrea	   and	  Diploria	   juveniles	  were	   the	   only	  genera	  to	  somewhat	  reflect	  the	  adult	  percent	  cover.	  Correspondence	   analysis	   shows	   the	   relationship	   between	   the	   abundance	   of	   juvenile	   taxa	   (8	  genus	   groups)	   across	   the	   six	   sampled	   sites	   (Figure	   4).	   The	   ordination	   plot	   show	   close	  correspondence	  between	  Outer	  Buccoo	  and	  Angel	  Reef	  due	  to	  their	  similar	  high	  abundance	  of	  Agaricia.	  West	  Buccoo	  and	  Culloden	  differ	   from	  the	  aforementioned	  sites	  due	  to	   their	  higher	  abundance	  of	  Scolymia,	  Siderastrea	  and	  Diploria.	  The	  far	  right	  position	  of	  Black	  Jack	  Hole	  shows	  this	  site’s	  taxa	  composition	  differs	  from	  all	  other	  sites	  because	  of	  high	  Porites	  abundance	  and	  a	  lack	  of	  Agaricia	  juveniles.	  	  0 20 40 60 80 100 OB  WB CE CW BJH AR % Juvenile taxa A 	   14	  	   	  Figure	  4.	  Correspondence	  analysis	  (CA)	  biplot	  showing	  the	  ordination	  of	  in	  situ	  juvenile	  genera	  in	  2013	  along	  the	  first	   and	   second	   axis,	   which	   explains	   51%	   and	   31%	   of	   the	   variance,	   respectively.	   	   (the	   group	   “other”	  compromises	  of	  Favia	  fragum,	  Eusmilia	  fastigiata,	  Colpophyllia	  natans,	  and	  Mycetophyllia	  ssp.)	  2.3.2 Characterization	  of	  sedimentation	  	  The	  average	   sediment	  accumulation	   rate	  was	  5.6	  ±	  4.2	  mg	  cm-­‐2	  d-­‐1	   (Table	  5).	   Sediment	   rates	  were	  not	   significantly	   different	   at	   all	   sites,	  with	   the	   exception	  of	   Culloden	  West	   (Tukey	  HSD,	  p>0.05).	   Accumulation	   rates	   at	   Culloden	  West	  were	   four	   times	   higher	   than	   at	   all	   other	   sites	  (15.1	  ±	  2.9	  mg	  cm-­‐2	  d-­‐1).	  While	  sedimentation	  rates	  were	  higher	  in	  one	  trap-­‐set	  (17.6	  ±0.9	  mg	  cm-­‐2	  d-­‐1)	  than	  the	  other	  (12.5	  ±0.4	  mg	  cm-­‐2	  d-­‐1),	  the	  overall	  accumulation	  rates	  in	  all	  tube	  traps	  were	  much	   higher	   than	   those	   rates	  measured	   in	   the	   other	   sites.	  Whilst	   at	   Buccoo	   sediment	  rates	   did	   not	   vary	   greatly	   among	   trap-­‐sets,	   the	   opposite	   was	   true	   at	   Culloden	   and	   Speyside	  (Fig.5).	  	  	  The	  composition	  analysis	   (Table	  5,	  Figure	  6)	   indicated	  that	  terrigenous	  material	  dominated	  at	  all	  sites	  and	  represented	  the	  most	  dominant	  fraction	  of	  the	  collected	  material,	  ranging	  between	  51.6	   ±0.2-­‐72.98	   ±0.3	   %	   being	   highest	   at	   Culloden	   Reef	   which	   were	   the	   sites	   nearest	   to	  mainland.	  The	  total	  percentage	  of	  carbonate	  materials	  ranged	  from	  22.03	  ±0.7	   -­‐	  40.1	  ±0.2	  %,	  and	   were	   highest	   at	   Buccoo	   sites.	   Organic	   matter	   represented	   <8%	   of	   the	   sediment	  composition	  of	  the	  samples	  at	  all	  the	  sites;	  this	  small	  percentage	  could	  be	  made	  up	  of	  the	  thin	  layers	   of	   turf	   that	   had	   grown	   inside	   the	   pipe	   trap.	   Apart	   from	   the	   greater	   proportion	   of	  terrigenous	   material	   being	   slightly	   higher	   at	   the	   Culloden	   sites,	   overall	   the	   composition	   of	  sediment	  collected	  did	  not	  vary	  greatly	  at	  each	  site.	  	  	   15	  Table	  5.	  Mean	  and	  standard	  deviations	  of	  sediment	  measurements	  	  	  	  No.	  of	  	  traps	  	  recovered	   Outer	  Buccoo	  9	   Western	  Buccoo	  9	   Culloden	  East	  5	   Culloden	  West	  6	   Black	  Jack	  Hole	  8	   Angel	  Reef	  8	  Mean	  weight(g)	  	   2.5	  ±	  0.73	   3.2	  ±	  0.5	   3.1	  ±	  1.0	   9.2	  ±	  1.8	   3.62	  ±	  2.2	   2.7	  ±	  1.3	  Rate	  (mg	  cm-­‐2	  d-­‐1)	  	   3.4	  ±	  0.9	   4.3	  ±	  0.7	   4.8	  ±1.6	   15.1	  ±2.9	   4.9	  ±3.1	   3.6	  ±1.8	  Sediment	  composition	  (%)	   	   	   	   	   	   	  Terrigenous	  	   56.5	  ±	  0.2	   51.6	  ±	  0.3	   70.98	  ±	  0.6	   72.98	  ±	  0.3	   65.32	  ±	  2.4	   58.76	  ±	  4.6	  CaCo3	   35.6	  ±	  0.5	   40.5	  ±	  0.2	   23.49	  ±	  0.3	   22.03	  ±	  0.7	   26.96	  ±	  3.9	   35.90	  ±	  5.7	  Organic	   7.9	  ±	  0.2	   7.9	  ±	  0.2	   5.53	  ±	  0.1	   4.99	  ±	  0.4	   7.72	  ±	  1.6	   5.34	  ±	  1.1	  Grain	  size	  distribution	  (%)	   	   	   	   	   	   	  Course	  sand	  	  	  	  	  	  	  	  >500	  µm	   1.4	  ±	  0.1	   5.5	  ±	  1.7	   3.3	  ±	  1.5	   3.5	  ±	  3.1	   15.9	  ±	  5.9	   8.8	  ±	  1.8	  Medium	  sand	  250-­‐500µm	   3.1	  ±	  1.3	   10.5	  ±	  1.6	   7.6	  ±	  1.1	   4.1	  ±	  0.0	   20.2	  ±	  2.2	   21.7	  ±	  4.1	  Fine	  sand	  	  	  	  	  	  	  	  125-­‐250µm	   22.6	  ±	  4.5	   18.2	  ±	  1.6	   26.0	  ±	  1.0	   29.8	  ±	  1.3	   35.0	  ±	  3.8	   42.9	  ±	  11.2	  Very	  fine	  sand	  	  63-­‐125µm	   21.8	  ±	  8.9	   29.0	  ±	  0.7	   44.4	  ±	  6.8	   45.5	  ±	  2.4	   20.5	  ±	  5.2	   15.3	  ±	  5.3	  Silt/clay	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  <63	  µm	   51.0	  ±	  10.7	   36.7	  ±	  4.3	   18.8	  ±	  3.3	   17.1	  ±	  0.7	   8.4	  ±	  4.9	   11.3	  ±	  7.6	  	   	   	   	   	   	   	  	  Figure	  5.	  Boxplot	  of	  sediment	  accumulation	  rates	  per	  site	  	  Figure	  6.	  Stacked	  barplots	  of	  average	  percent	  of	  sediment	  (left)	  composition	  and	  (right)	  particle	  size	  distribution	  from	  sediment	  data	  collected	  May-­‐June	  2013	  0	   20	   40	   60	   80	   100	  OB	  WB	  CE	  CW	  BJH	  AR	  %	  Distribu?on	  of	  par?cle	  sizes	  coarse	  sand	  	   mediam	  sand	   fine	  sand	  	   very	  fine	  sand	   silt/clay	  	  0	   20	   40	   60	   80	   100	  OB	  WB	  CE	  CW	  BJH	  AR	  %	  Sediment	  composi?on	  	  Terrigenous	   Calcareous	   Organic	  Sediment	  rate	  mg	  cm-­‐2 	  d-­‐1 	  	   16	  The	   sediment	   grain	   size	   distributions	   differed	   between	   the	   three	   reef	   systems.	   Culloden	   and	  Buccoo	   Reef,	   both	   on	   the	   western	   side	   of	   Tobago	   facing	   the	   Caribbean	   Sea,	   had	   a	   greater	  proportion	   of	   very	   fine	   sand	   (63-­‐125-­‐µm)	   and	   silt/clay	   (<63μm).	   At	   Buccoo	   Reef	   sites	   the	  proportion	  of	  silt/clay	  particles	  was	  the	  highest	  and	  represented	  the	  largest	  grain	  size	  fraction	  (36-­‐50%).	  At	   the	  Culloden	  sites	   the	  dominant	  grain	  size	   fraction	  was	  very	   fine	  sand	   (44-­‐45%).	  The	   Speyside	   sites,	   on	   the	   north-­‐eastern	   side	   of	   the	   island,	   were	   dominated	   by	   fine	   sand	  sediment,	  250-­‐500µm	  (35-­‐44%).	  Additionally,	  these	  sites	  had	  the	  largest	  proportion	  of	  coarser	  materials,	   >125μm,	   and	   had	   almost	   twice	   the	   amount	   of	   course	   and	   medium	   sand	   (>500-­‐250µm)	   in	   comparison	  with	   the	   other	   four	   sites.	   Although	   offshore	   from	   Tobago,	   Black	   Jack	  Hole	   and	   Angel	   Reef	   are	   50m	   or	   less	   from	   the	   shore	   of	   small	   islands,	   so	   there	   is	   a	   greater	  potential	   for	   coarse	   and	   medium	   sand	   particles	   reaching	   the	   fore	   reef	   zones	   with	   enough	  current	  and	  wave	  energy.	  	  2.4 Discussion	  While	  healthy	   coral	   reefs	   typically	   have	  high	  numbers	  of	   juvenile	   coral	   colonies,	   at	   degraded	  reefs	   their	   numbers	   tend	   to	   be	   limited	   (Jackson	   et	   al.	   2014).	   From	   reviewing	   the	   literature	  across	   the	   Caribbean,	   Ruiz-­‐Zarate	   and	   Gonzales	   (2004)	   found	   that	   juvenile	   densities	   ranged	  from	  13–274	  m-­‐2	  in	  what	  used	  to	  be	  healthier	  reefs	  up	  before	  the	  die-­‐off	  of	  the	  grazing	  urchin	  Diadema	  antillarum	  in	  the	  1980s.	  Most	  studies	  since	  then	  have	  reported	  much	  lower	  densities	  that	   range	   between	   0.8-­‐12	   juveniles	   m-­‐2.	   This	   decline	   has	   been	   attributed	   to	   the	   decline	   in	  parental	  stock,	  decrease	  in	  herbivory,	  increase	  in	  algal	  growth	  and	  terrestrial	  runoff	  (Jackson	  et	  al.	   2014).	   Our	   study	   found	   juvenile	   density	   across	   Tobago’s	   major	   reef	   system	   to	   be	   low,	  averaging	  around	  5.71±2.39	  m-­‐2,	  which	  was	  similar	  to	  the	  densities	  found	  in	  Florida	  (Moulding	  2005;	  Miller	  et	   al.	   2000;	  Chiappone	  &	  Sullivan	  1996),	  Bermuda	   (Smith	  1997),	  Cayman	   Islands	  	  (Manfrino	  et	  al.	  2013)	  and	  Belize	  (Irizarry-­‐soto	  &	  Weil	  2009).	  	  Densities	  did	  not	  vary	  among	  the	  different	  reef	  sites	  assessed	  with	  the	  exception	  of	  Angel	  reef,	  where	  juvenile	  abundances	  were	  significantly	  higher.	  This	  difference	  may	  be	  explained	  by	  Angel	  Reef’s	  higher	  level	  of	  coral	  cover,	  which,	  based	  on	  2013	  assessments	  (see	  Table	  1	  in	  Chapter	  2),	  represents	  about	  46%	  of	  the	  reef	  benthos,	  whereas	  at	  the	  other	  sites	  it	  was	  ≤	  21%.	  Additionally,	  algal	  coverage	  was	  <14%	  of	  Angel	  Reef’s	  benthos	  as	  opposed	  to	  28-­‐51%	  at	  the	  other	  reef	  sites.	  This	   supports	   the	   notion	   that	   higher	   coral	   cover	   and	   lower	   algal	   cover	   is	   a	  more	   hospitable	  environment	  for	  larvae	  and	  juveniles	  to	  develop	  (Jackson	  et	  al.	  2014).	  	  	   17	  Based	  on	  a	  growth	  rate	  of	  0.37	  to	  0.73	  mm/month	  (Birkeland	  1977)	  most	  juveniles	  recorded	  in	  this	  study	  were	  about	  5	  to	  10	  years	  old.	  Accordingly,	  the	  assessed	  juvenile	  communities	  were	  likely	   impacted	  by	  the	  2005	  bleaching	  event,	  which	  was	   followed	  by	  a	  coral	  disease	  outbreak	  (Bouchon	  et	  al.	  2008).	  Whilst	  coral	  juveniles	  tend	  not	  to	  suffer	  greatly	  from	  bleaching	  induced	  mortality	  (Mumby	  1999;	  Shenkar	  et	  al.	  2005),	  the	  bleaching	  of	  adult	  colonies	  however	  is	  known	  to	   reduce	   coral’s	   reproductive	   output	   in	   the	   years	   following	   the	   event	   (Ritson-­‐williams	   et	   al.	  2009;	   Ward	   et	   al.	   2000a;	   Mallela	   &	   Crabbe	   2009).	   Broadcasting	   taxa	   in	   the	   Caribbean	   are	  especially	   affected,	   as	   most	   bleaching	   events	   occur	   during	   their	   yearly	   spawning	   period	  between	   August	   and	   October	   (Szmant	   &	   Gassman	   1990).	   Thus	   it	   is	   likely	   that	   the	   juvenile	  densities	   reported	   in	   this	   study	  were	   even	   lower	   than	   usual,	   due	   to	   the	   impact	   of	   the	   2005	  bleaching	   event.	   This	   notion	   is	   supported	   by	   studies	  measuring	   recruitment	   rates	   on	   tiles	   at	  Buccoo	   Reef.	   In	   the	   early	   1990s,	  mean	   recruitment	   rates	  were	   188	  m-­‐2year-­‐1	   (Laydoo	   1993),	  which	  by	  2007	  was	  reduced	  to	  103	  m-­‐2year-­‐1	  (Mallela	  &	  Crabbe	  2009),	  indicating	  the	  decline	  in	  reproductive	   output	   over	   the	   last	   two	   decades	   and	   following	   the	   2005	   bleaching	   event.	  Consequently,	   it	   is	   likely	  that	  the	  2010	  bleaching	  event	  also	  has	  had	  a	  negative	  impact	  on	  the	  reproductive	  output	  of	  corals,	  leading	  to	  a	  continuously	  low	  juvenile	  community,	  which	  in	  turn	  has	  the	  potential	  to	  slow	  down	  post-­‐disturbance	  recovery.	  	  The	   juveniles	   assessed	   comprised	   primarily	   of	   brooding	   genera,	   particularly	   from	   the	   genus	  Agaricia,	  which	  generally	  contribute	  relatively	  little	  to	  overall	  reef-­‐building	  processes	  (Hughes	  &	  Tanner	  2000).	  While	  juveniles	  form	  the	  largest	  reef-­‐building	  taxa	  on	  Tobago	  were	  very	  rare,	  e.g.	  Montastraea	  and	  Colpophyllia,	  a	  substantial	  number	  of	  Siderastrea	  and	  Diploria	  juveniles	  were	  found	   at	   some	   sites.	  Other	   recent	   studies	   have	   also	   found	  moderate	   juvenile	   abundances	   of	  these	  two	  taxa	  (Miller	  et	  al.	  2000;	  Moulding	  2005;	  Vermeij	  et	  al.	  2011).	  Additionally	  in-­‐situ	  post-­‐settlement	   studies	  have	   found	  Siderastrea	   and	  Diploria	   recruits	   to	  have	  high	   survival	   rates	   in	  their	   post-­‐settlement	   stage	   (Irizarry-­‐soto	  &	  Weil	   2009).	   The	   community	   composition	   did	   not	  vary	   greatly	   across	   sites,	   except	   at	   Black	   Jack	   Hole,	  where	   brooding	   genus	  Porites	   was	  most	  dominant	  while	  Agaricia	  juveniles	  were	  scarce.	  This	  disparity	  can	  be	  explained	  by	  the	  fact	  that	  the	  Agaricia	  adult	  population	  was	  also	  very	  sparse	  at	  this	  site.	  As	  in	  most	  Caribbean	  reefs	  the	  relative	  abundance	  of	  most	   juveniles	  did	  not	  match	  with	  those	  of	  the	  adult	  coral	  community.	  Pioneer	   genera	   like	  Agaricia	   and	  Porites	  were	   over-­‐represented	  while	   large-­‐sized	   genera	   like	  Montastraea	   and	   Colpophyllia	   were	   underrepresented.	   Exceptions	   were	   the	   relative	  abundances	   for	   juveniles	   of	   the	   reef-­‐building	   the	   genera	   Siderastrea	   and	  Diploria,	   which	   did	  match	  the	  adult	  community	  at	  some	  sites.	  	  	  	   18	  The	   juvenile	   community	   composition	   found	   in	   this	   study	   is	   typical	   among	   Caribbean	   reefs	  (Chiappone	   &	   Sullivan	   1996;	   Edmunds	   et	   al.	   1998;	   Irizarry-­‐soto	   &	  Weil	   2009).	   Unlike	   Pacific	  coral	  reefs,	  where	  the	  most	  common	  corals	  also	  have	  the	  most	  abundant	  juveniles	  (Miller	  et	  al.	  2000)	  juvenile	  communities	  in	  this	  region	  tend	  to	  be	  a	  direct	  function	  of	  the	  taxa’s	  different	  life-­‐history	  characteristics	  and	  reproductive	  strategies	  (Bak	  &	  Engel	  1979;	  Miller	  et	  al.	  2000).	  Weedy	  small-­‐size	  brooding	  genera	  such	  as	  Agaricia	  and	  Porites	  tend	  to	  dominate	  in	  terms	  of	   juvenile	  abundance	   as	   they	   are	   very	   fecund	   and	   have	   high	   recruitment	   but	   also	   have	   high	  mortality	  rates.	  Conversely,	  broadcasting	  corals,	  such	  as	  most	  of	  Tobago’s	  key	  framework	  building	  taxa,	  have	   low	   recruitment	   rates	   but	   are	  more	   resistant	   to	   disturbances	   and	   thus	   live	   longer	   and	  grow	   larger	   (Wittenberg	   &	   Hunte	   1992;	   Hughes	   &	   Jackson	   1985).	   The	   key	   reproductive	  difference	  between	   these	   two	  groups	   is	   that	  brooding	   taxa	   reproduce	  on	  a	  monthly	  basis	  by	  undergoing	  self-­‐fertilisation	  and	  releasing	  well-­‐developed	  larvae	  that	  are	  able	  to	  settle	  quickly.	  On	  the	  other	  hand,	  broadcasting	  corals	  only	  spawn	  once	  a	  year	  by	  releasing	  gametes	  that	  need	  to	   undergo	   external	   fertilisation	   to	   form	   larvae,	  which	   then	   take	   over	   a	  week	   to	   settle.	   This	  reproduction	  strategy	  typically	  produces	  fewer	  larvae	  and	  tends	  to	  be	  less	  successful	  at	  settling.	  However,	   once	   settled	   their	   recruits	   and	   juveniles	   are	   more	   resistant	   to	   disturbances	   than	  recruits	  from	  brooding	  taxa	  (Szmant	  1986;	  Ritson-­‐williams	  et	  al.	  2009).	  Even	   though	   low	   recruitment	   of	   massive	   species	   is	   not	   unusual	   among	   Caribbean	   juvenile	  communities,	   juveniles	   of	   the	  Montastraea	   genus	  were	   exceptionally	   rare	   considering	   it	   has	  highest	  parental	  stock	  -­‐-­‐	   this	  genus	  accounts	   for	  >60%	  of	  the	  coral	  cover	  of	  most	  of	  Tobago’s	  reefs	   (Alemu	   I	   &	   Clement	   2014).	   The	   implication	   of	   this	   genus’s	   low	   ability	   to	   recover	   from	  disturbances	   via	   sexual	   reproduction	   and	   recruitment	   is	   that	   it	   could	   lead	   to	   a	   continuous	  decline	  of	  coral	  reefs,	  as	   it	   is	  possibly	  Tobago’s	  most	   important	  reef	   framework-­‐building	  taxa.	  While	  the	   low	  number	  or	   lack	  of	  sexually	  produced	  recruits	  by	  this	  genus	   is	  well	  documented	  across	  Caribbean	  reefs	  (Hughes	  &	  Tanner	  2000;	  Vermeij	  et	  al.	  2011;	  Irizarry-­‐soto	  &	  Weil	  2009),	  the	  exact	   reasons	   remains	  unknown.	  Studies	  on	  some	  Montastrea	   species’	   fecundity	   indicate	  that	   the	   problem	   does	   not	   appear	   to	   lie	   in	   the	   reproductive	   output	   of	   its	   colonies	   but	   the	  processes	   following	   spawning,	   i.e.	   fertilization,	   larval	   settlement	   and/or	   post	   recruitment	  survival	  (Szmant	  1988).	  Ritson-­‐Williams	  et	  al.	  (2009)	  suggest	  that	  since	  the	  Pleistocene	  it’s	  likely	  that	   many	   Caribbean	   framework	   building	   species	   have	   evolved	   with	   low	   sexual	   recruitment	  levels,	   as	   they	   instead	   rely	   on	   the	   production	   of	   recruits	   from	   fragmentation	   or	   budding.	  However,	   this	  asexual	   reproductive	   strategy	  no	   longer	  appears	  viable	  among	  Caribbean	   reefs	  where	  coral	  cover	  has	  dramatically	  declined.	  	  	   19	  While	  sedimentation	  is	  known	  for	  impeding	  reproduction	  and	  recruitment	  processes,	  our	  study	  revealed	  that	  sedimentation	  rates,	  at	  less	  than	  <5	  mg	  cm-­‐2	  d-­‐1,	  were	  low	  on	  most	  Tobago’s	  reef	  sites.	   Though	   composition	   analysis	   indicates	   that	   most	   of	   the	   sediment	   was	   of	   terrigenous	  (land)	  origin,	   the	  daily	   rates	  were	   still	  below	  Rogers’	   (1990)	   threshold	   rate	  of	  10	  mg	  cm-­‐2	  d-­‐1,	  which	   suggests	   that	   Tobago’s	   coral	   communities	   are	   probably	   not	   greatly	   impacted	   by	  sedimentation	   during	   the	   last	   months	   of	   the	   dry	   season.	   Whereas	   at	   Culloden	   West	  sedimentation	  rates	  were	  exceptionally	  high	   in	  comparison	  with	  the	  other	  sites	   (>15	  mg	  cm-­‐2	  day-­‐1),	  juvenile	  abundance	  was	  not	  significantly	  different	  at	  this	  site.	  Additionally,	  the	  high	  rates	  reported	  at	  site	  were	  likely	  so	  high	  due	  to	  high	  current	  driven	  turbulence	  as	  only	  one	  trap	  set	  was	  recovered	  as	  the	  other	  two	  were	  uprooted	  an.	  	  We	   observed	   that	   sediment	   grain	   size	   analysis	   at	   Buccoo	   Reef	   sites,	   which	   had	   the	   lowest	  juvenile	  density	  and	  diversity,	  had	  the	  highest	  proportions	  of	  silt/clay	  (at	  least	  20%	  higher	  than	  at	  the	  other	  sites).	  This	  is	  most	  likely	  a	  result	  of	  these	  sites	  being	  adjacent	  to	  urbanized	  land	  and	  exposure	  to	  point	  and	  non-­‐point	  source	  pollution,	  especially	  considering	  that	  silt/clay,	  due	  to	  its	  lightness,	   is	   the	   land-­‐based	   sediment	  material	  with	   the	   highest	   potential	   to	   drift	   all	   the	  way	  from	   shores	   to	   reefs.	   Though	   both	   coarse	   and	   fine-­‐grained	   sediments	   have	   the	   potential	   to	  harm	   corals,	   studies	   indicate	   that	   fine	   sediment	   tends	   to	   be	   more	   harmful.	   Fine	   sediments	  particularly	  attenuates	   light	   in	  the	  water	  column,	  as	  they	  can	  remain	  suspended	  longer	   in	  the	  water	  column,	  reducing	  photosynthesis	  (Abdullah	  et	  al.	  2011).	  Whereas	  corals	  tend	  to	  remove	  coarser	  grained	  sediment	  more	  readily	  than	  fine	  particles,	  fine	  sediment	  is	  easily	  re-­‐suspended,	  which	  means	  the	  same	  grains	  can	  impact	  the	  coral	  communities	  more	  than	  once	  (Hernandez	  et	  al.	   2009;	  Weber	  et	   al.	   2006).	  Additionally,	   clay	   and	   silt-­‐sized	   sediment	   is	  more	   likely	   to	   carry	  organic	   chemical	   contaminants	   and	   toxins,	   which	   in	   turn	   are	   understood	   to	   increase	   coral	  mortality	  (Fabricius	  et	  al.	  2003).	  	  Even	   though	  we	   found	   that	   sedimentation	   deposition	   was	   low,	   it	   is	   likely	   that	   deposition	   is	  much	   greater	   during	   the	   rainy	   season,	   which	   tends	   to	   occur	   from	   May	   until	   November.	  Considering	   that	  most	   broadcasting	   taxa	   spawn	  within	   this	   time	   frame	   it	   is	   therefore	   crucial	  that	   assessments	   be	   made	   of	   sediment	   deposition	   throughout	   the	   rainy	   season.	   	   Field	   and	  experimental	  studies	  have	  found	  that	  sedimentation,	  and	  the	  associated	  toxins,	  are	  known	  to	  prevent	  fertilization,	  increases	  the	  mortality	  of	  newly	  settled	  recruits	  and	  that	  particles	  covering	  hard	  surfaces	   inhibit	  the	  ability	  of	   larvae	  to	  settle	  (Babcock	  &	  Smith	  2000;	  Torres	  &	  Morelock	  2002;	  Fabricius	  2005).	  	   20	  	  The	  disparities	  of	   sedimentation	   rates	   that	  we	   found	  within	   the	  same	  reef	   system,	   reef	   sites	  and	  even	  trap-­‐sets	   indicate	  that	  sediment	  deposition	  can	  differ	  greatly	  among	  Tobago’s	   reefs	  over	  a	  narrow	  spatial	  scale	  (Fig.1).	  This	  may	  be	  due	  to	  the	  prevalence	  of	  stronger	  currents	  and	  wave	  energy	  at	   these	  reefs	   (Laydoo	  1991).	  However,	   it	   is	  possible	   the	  small-­‐scale	  variation	   in	  measured	   rates	   is	   an	   artefact	   of	   the	   sampling	   method.	   We	   recommend	   that	   further	  sedimentation	   rate	   data	   be	   collected	   to	   produce	   a	   long-­‐term	   sediment	   accumulation	   profile	  that	  covers	  both	  the	  rainy	  and	  the	  dry	  seasons	  (Hernandez	  et	  al.	  2009).	  Additionally,	  this	  study	  highlights	  the	  value	  of	  analyzing	  grain	  size	  distribution	  and	  composition	  of	  sediment	  samples,	  as	  this	  allows	  researchers	  to	  distinguish	  the	  origin	  and	  size	  of	  sediment	  particles.	  	  In	  conclusion,	  the	  low	  densities	  of	  sexually	  produced	  juveniles,	  especially	  of	  large	  broadcasting	  species,	   indicate	  that	  post-­‐disturbance	  recovery	  among	  coral	  communities	  will	  probably	  be	  at	  best	   slow	   if	   not	   very	   limited.	   Nonetheless,	   we	   found	   evidence	   that	   Siderastrea	   and	  Diploria	  species	  were	  still	  successfully	  recruiting	  and	  surviving.	  It	  is	  likely	  that	  mass	  bleaching	  events	  are	  reducing	   both	   parental	   stock	   (coral	   cover)	   and	   coral’s	   reproductive	   output,	   leading	   to	   lower	  juvenile	  densities.	  It	  is	  crucial	  that	  other	  disturbances	  that	  further	  impair	  recruitment	  process	  at	  a	  local	  scale,	  such	  as	  terrestrial	  run	  off	  and	  overfishing,	  are	  well	  understood	  and	  managed.	  Even	  though	   our	   study	   indicates	   that	   sedimentation	   rates	   during	   the	   dry	   season	   appear	   to	   be	  minimal,	  it	  is	  essential	  that	  further	  data	  be	  collected	  during	  the	  rainy	  season.	  	  	  	  	   21	  Chapter	  3.	   	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  	  Using	   coral	   size	   distribution	   to	   assess	   the	   recovery	   from	   mass	  bleaching	  in	  the	  southern	  Caribbean	  3.1 Introduction	  Caribbean	  coral	  reefs	  are	  among	  the	  most	  heavily	   impacted	  marine	  ecosystems	  on	  the	  planet	  (Bellwood	  et	  al.	  2004;	  Edmunds	  &	  Elahi	  2007;	  Alvarez-­‐Filip	  et	  al.	  2011).	  Over	  70%	  of	  coral	  cover	  is	  estimated	  to	  have	  been	  lost	  in	  the	  last	  the	  decades	  and	  become	  replaced	  by	  macroalgae	  and	  turf	   dominated	   environments	   (Gardner	   et	   al.	   2003).	   	   Such	   ecological	   phase-­‐shifts	   have	   been	  attributed	   to	   historical	   reduction	   of	   large	   sized	   grazing	   herbivore	   fish	   populations	   due	   to	  overfishing,	  the	  die-­‐off	  of	  grazing	  Diadema	  urchins	  in	  the	  early	  1980s,	  water	  pollution,	  and	  coral	  disease	   outbreaks	   (Mallela	   et	   al.	   2010;	   Weil	   2001;	   Norström	   et	   al.	   2009;	   Hughes	   1994).	   In	  addition,	   over	   the	   last	   few	   decades	   climate	   change	   driven	   mass	   bleaching	   events	   have	  increased	  in	  frequency	  and	  intensity	  exacerbating	  the	  decline	  of	  coral	  communities	  (C.	  M.	  Eakin	  et	  al.	  2010;	  Donner	  et	  al.	  2007).	  	  	  The	  existence	  of	  productive	  and	  attractive	  reefs	  in	  the	  Caribbean	  is	  contingent	  on	  the	  growth	  of	  hard	   corals.	   Therefore,	   there	   is	   a	   strong	  need	   to	  understand	  how	  mass	  bleaching	   events	   are	  impacting	  coral	  reefs	  and	  their	  ability	  to	  recover	  back	  into	  their	  pre-­‐disturbance	  state.	  Studying	  the	   population	   size	   structure	   of	   corals	   can	   provide	   valuable	   demographic	   information	   about	  coral	  communities	  and	  population	  dynamics	  (Smith	  et	  al.	  2005;	  Crabbe	  2009;	  McClanahan	  et	  al.	  2008).	   Coral	   life-­‐history	   processes	   are	   strongly	   related	   to	   colony	   size,	   thus	   data	   on	   taxa’s	  colonies	   size	   frequency	   distribution	   can	   reflect	   corals	   responses	   to	   environmental	   stress	  (Vermeij	  &	  Bak	  2000).	   Though	   labour	  and	   time	   intensive,	   collecting	   coral	   size	   frequency	  data	  can	   provide	   insight	   into	   past	   and	   future	   patterns	   of	   growth	   and	   mortality	   (Bak	   &	  Meesters	  1998;	  Meesters	  et	  al.	  2001;	  McClanahan	  et	  al.	  2008).	  	  Scleractinian	   coral	  populations	   tend	   to	  be	  positively	   skewed	  as	   they	  are	   comprised	  mainly	  of	  small	  to	  medium	  sized	  colonies	  with	  relatively	  fewer	  large	  colonies	  (Babcock	  1991;	  Meesters	  et	  al.	  2001;	  McClanahan	  et	  al.	  2008;	  Adjeroud	  et	  al.	  2007).	  The	  typical	  population	  size	  structure,	  however,	   can	   vary	   among	   different	   taxa	   due	   to	   their	   different	   life	   history	   traits	   such	   as	  fecundity,	   growth	   rates	   and	   susceptibility	   to	   morality	   (Meesters	   et	   al.	   2001).	   Additionally,	  intraspecific	  population	  size	  structure	  can	  vary	  as	  a	  result	  of	  different	  environmental	  conditions	  as	  well	  as	  disturbance	  histories	  (De	  Lins	  Barros	  &	  De	  Oliveira	  Pires	  2006;	  Adjeroud	  et	  al.	  2007).	  	   22	  Research	   on	   Caribbean	   reefs	   have	   found	   that	   coral	   population	   structure	   tends	   to	   become	  negatively	  skewed	   in	  reefs	  with	  poor	  of	  water	  quality,	  as	   large	  colonies	  will	  be	  more	   likely	   to	  survive	   harsh	   reef	   conditions	   than	   smaller	   coral	   and	   conditions	   for	   recruitment	   become	  impaired	  (Bak	  &	  Meesters	  1999;	  Bak	  &	  Meesters	  1998).	  	  Coral	  bleaching,	  a	  paling	  caused	  by	  the	   loss	  of	  the	  symbiotic	  micro-­‐algae	  (Symbiodinium)	   that	  reside	   in	   coral	   tissue,	   leaves	   corals	   in	   energy	   deficit	   and	   consequently	   impairs	   growth	   and	  reproduction,	  making	  them	  vulnerable	  to	  disease	  and	  mortality	  (Ward	  et	  al.	  2000b).	  Loss	  of	  live	  coral	  cover	  due	  to	  complete	  or	  partial	  colony	  mortality	  following	  mass	  coral	  bleaching	  can	  have	  a	   significant	   impact	  on	   the	  population	   size	   structure	  of	  different	   coral	   species	   (Shenkar	  et	  al.	  2005).	  As	  coral’s	  fecundity	  is	  determined	  by	  the	  number	  of	  sexually	  mature	  and	  healthy	  polyps	  on	  a	  colony,	  bleaching	  that	  leads	  to	  the	  decline	  of	  the	  size	  and	  number	  of	  colonies	  can	  impact	  a	  population’s	  reproductive	  output	  and	  thus	  may	  alter	  the	  dominance	  and	  persistence	  of	  certain	  coral	  taxa	  (Hughes	  1984).	  The	  few	  studies	  which	  have	  assessed	  the	  effect	  of	  mass	  bleaching	  on	  coral	  size	  distribution	  have	  found	  that	  many	  taxa’s	  mean	  colony	  size	  declines,	  due	  to	  the	  loss	  of	  large	  sized	  corals	  to	  partial	  or	  complete	  mortality	  resulting	  in	  an	  increase	  in	  small	  sized	  colonies	  (McClanahan	  et	  al.	  2008;	  Crabbe	  2009).	  	  The	  purpose	  of	  this	  study	  was	  to	  examine	  changes	  in	  corals	  population	  size	  structure	  to	  assess	  the	   impact	   of	   climate	   change	   and	   recovery	   among	   coral	   communities	   of	   the	   southern	  Caribbean	   reefs	   of	   Tobago.	   The	   corals	   among	   these	   reefs	  were	   studied	   over	   a	   period	   during	  which	  they	  endured	  a	  mass	  bleaching	  event	  in	  2010.	  	  Tobago’s	  fringing	  reefs	  have	  experienced	  the	   same	   degradation	   patterns	   as	   the	   rest	   of	   the	   region’s	   reefs	   due	   to	   the	   continuous	  overfishing,	   terrestrial	   runoff,	   wastewater	   discharge	   and	   poor	   coastal	   natural	   resource	  management	   (Mallela	  et	  al.	  2010).	  Thus	  these	  reefs	  are	  also	  characterized	  by	   low	  coral	  cover	  dominated	  by	  the	  more	  persistent	  massive	  and	  encrusting	  types	  of	  coral	  species.	  Though	  corals	  were	   known	   to	   have	   bleached	   in	   1998,	   2005	  was	   the	   first	   well	   recorded	   bleaching	   event	   in	  Tobago	   followed	   by	   another	   in	   2010.	   A	   rapid	   bleaching	   response	   assessment	   was	   done	  throughout	  the	  bleaching	  period	  in	  2010	  recorded	  bleaching	  severity	  as	  high	  as	  29-­‐69%	  across	  Tobago’s	  reefs	  coral	  communities	  and	  about	  2-­‐8	  %	  coral	  mortality	   (Alemu	  I	  &	  Clement	  2014).	  Recordings	  of	  disease	  infected	  corals	  (Alemu	  I	  2011)	  additionally	  indicated	  that	  coral	  mortality	  probably	   further	   increased	  after	   the	  bleaching	  event,	   as	  was	   the	  case	   in	  2005	   (Harding	  et	  al.	  2008).	   Little	   is	   known,	   however,	   about	   the	   impact	   of	   bleaching	   events	   on	   the	   population	  dynamics	   of	   Tobago’s	   coral	   communities.	  Most	   knowledge	   about	   the	  health	   and	  disturbance	  history	  of	  Tobago’s	  coral	  reefs	  is	  based	  on	  changes	  in	  percent	  benthic	  cover.	  Only	  one	  study	  has	  	   23	  explored	   the	   impact	   of	   bleaching	   events	   on	   Tobago’s	   coral	   population	   dynamics,	   based	   on	  recruitment	  and	  growth	  modeling	  (Mallela	  &	  Crabbe	  2009).	  	  Here,	  we	  quantified	   the	  population	   size	   structure	  of	   Tobago’s	   dominant	   coral	   species	   before	  the	  bleaching	   episode	   in	   September,	   2010,	   immediately	   after	   the	  bleaching	   ended	   in	  March,	  2011	  to	  assess	  impact	  of	  the	  bleaching-­‐induced	  mortality,	  and	  in	  May,	  2013	  to	  assess	  change	  in	  the	  three	  years	  since	  the	  event.	  Furthermore,	  comparisons	  were	  made	  between	  reef	  systems	  located	  adjacent	   to	  urban	  coastal	   land	  vs.	   rural	   forested	   land	   to	  assess	  how	  coral	  population	  structure	  differs	  generally	  and	  in	  the	  face	  of	  a	  bleaching	  event.	  	  Benthic	  percent	  cover	  data	  was	  also	  re-­‐surveyed	  in	  order	  to	  compare	  if	  any	  recovery	  has	  taken	  place.	  The	  results	  of	  this	  study	  provide	  insight	  into	  the	  ability	  of	  different	  dominant	  Caribbean	  coral	  species	  	  to	  persist	   in	  the	  face	  of	  warming	  ocean	  temperatures.	  3.2 Methods	  Tobago	  is	  the	  smaller	  sister	  island	  of	  the	  nation	  Trinidad	  and	  Tobago,	  part	  of	  the	  Lesser	  Antilles	  island	   arc,	   located	   close	   to	   the	   South	   American	   mainland.	   Due	   to	   Tobago’s	   proximity	   the	  Orinoco	  and	  Amazon	  River	  deltas	  its	  fringing	  coral	  reefs	  systems	  evolved	  under	  the	  influence	  of	  nutrient	   and	   sediment	   rich	   flushes.	   Consequently,	   Tobago’s	   coral	   communities	   are	  characterised	  for	  having	  lower	  species	  diversity	  in	  comparison	  to	  other	  Caribbean	  reefs	  (Moses	  &	  Swart	  2006;	  Potts	  et	   al.	   2004).	   	   Tobago	   is	   a	  hilly	  300	  km2	   island	  of	   volcanic	  origin	   covered	  mostly	  by	  forest	  and	  shrubs	  lands,	  though	  the	  south-­‐western	  part	  of	  the	  island	  has	  undergone	  significant	  urbanisation	  and	  agricultural	  development.	  	  The	   study	   was	   conducted	   on	   three	   major	   reef	   systems:	   Buccoo	   Reef,	   Culloden	   Reef	   and	  Speyside	  (Figure	  1).	  Buccoo	  and	  Culloden	  Reef	  are	  both	  along	  the	  south-­‐western	  coast	  of	   the	  island,	   facing	   the	  Caribbean	  Sea.	  While	  Culloden	  Reef’s	  bay	  borders	   forested	  hills	  with	  minor	  developments	   of	   gravel	   roads,	   coastal	   lands	   adjacent	   to	   Buccoo	   Reef	   have	   become	   heavily	  urbanized	  and	  wastewaters	  become	  directly	  discharged	  into	  Buccoo’s	  bay.	  Speyside	  features	  a	  large	  network	  of	  fringing	  reefs	  along	  small	  islands	  and	  rocky	  outcrops	  on	  the	  north-­‐eastern	  side	  of	   the	   island	   (Laydoo	   1991).	   These	   reefs	   are	   highly	   valued	   by	   the	   recreational	   diving	  community.	  Speyside	  coastal	  lands	  remain	  relatively	  undeveloped	  comprising	  of	  a	  hilly	  forested	  landscape	  apart	  from	  Speyside	  village	  (a	  fishing	  community)	  and	  two	  medium-­‐sized	  hotels.	  	  	   24	  	  Figure	  7.	  Map	  of	  Tobago	  and	  location	  of	  studied	  reef	  systems	  and	  sites	  Two	   study	   sites	  were	   established	   at	   each	   reef	   system	   (Figure	   7).	  Outer	   Buccoo	   and	  Western	  Buccoo	   represent	   reef	   systems	   exposed	   to	   land	   pollution	   from	   run	   off	   and	   sewage	   waters.	  	  Culloden	   East	   and	   Culloden	   West	   act	   like	   controls	   to	   the	   Buccoo	   sites,	   as	   they	   experience	  similar	   marine	   conditions	   but	   less	   land-­‐based	   pollution.	   Black	   Jack	   Hole	   and	   Angel	   Reef	   at	  Speyside	   represent	   reefs	  exposed	   to	   the	  Atlantic	  Ocean	   that	  are	  also	   less	  affected	  by	  marine	  pollution.	  Surveys	  on	  corals	  colony	  size	  and	  the	  reefs	  benthic	  cover	  were	  collected	  at	  each	  reef	  site	  from	  May	  to	  June	  in	  2013	  using	  SCUBA.	  The	  data	  are	  compared	  to	  previous	  surveys	  carried	  out	   since	   2010	   at	   the	   same	   sites,	   except	   Angel	   Reef,	   gathered	   by	   the	   Trinidad	   and	   Tobago	  Institute	  of	  Marine	  Affairs	  (IMA)	  as	  part	  of	  their	  annual	  monitoring	  programme.	  3.2.1 Benthic	  cover	  survey	  Benthic	  percent	  cover	  was	  estimated	  following	  the	  photo-­‐quadrat	  method	  as	  described	  by	  Hill	  and	   Wilkinson	   (2004).	   Ten	   randomly-­‐placed	   10	   m	   transects	   were	   carried	   out	   at	   each	   site	  between	   the	  depth	  of	  8	  and	  12	  m.	  Along	  each	   transect	  1m2	  non-­‐overlapping	  photo-­‐quadrats	  were	   taken.	   For	   each	   photoquadrat	   benthic	   cover	  was	   identified	   under	   60	   random	  points	   in	  each	   image	  using	  Coral	  Point	  Count	  with	   the	  Excel–extensions	   (Kohler	  &	  Gill	  2006),	   following	  the	   protocol	   developed	  by	   the	   IMA	   (Alemu	   I	  &	  Clement	   2012).	   Surveys	   from	  previous	   years,	  carried	   out	   by	   the	   IMA,	   employed	   similar	  methods,	   however	   they	   used	   permanent	   transects	  	   25	  which	   they	   revisited	   in	  each	   survey.	  Additionally,	   only	   five	   transects	  were	   completed	   in	   their	  2010	  survey.	  	  3.2.2 Colony	  size	  frequency	  survey	  Along	  each	  10m	  transect	   the	   length	  of	  all	   adult	   coral	   colonies	  ≥5	  cm	  were	   recorded,	   that	   lay	  within	   50	   cm	   on	   each	   side	   of	   the	   transect	   tape,	   following	   Done	   et	   al.	   (2010).	   All	   measured	  colonies	  were	   identified	   to	   species	   levels	   expect	   for	   the	   genus	  Agaricia	   due	   to	   identification	  difficulties.	  Colonies	  with	  >50%	  of	  their	  living	  tissue	  within	  the	  belt	  transect	  area	  were	  included.	  Colonies	  divided	  by	  partial	  mortality	  into	  separate	  patches	  of	  living	  tissue	  that	  were	  >3	  cm	  apart	  from	   each	   other	   were	   considered	   separate	   colony	   entities	   and	   were	   measured	   individually	  (McClanahan	  et	   al.	   2008;	  Adjeroud	  et	   al.	   2007;	  Done	  et	   al.	   2010).	   To	   identify	   changes	   in	   the	  coral	   community	   and	   population	   structure	   since	   the	   2010	   bleaching	   event	   we	   used	   similar	  datasets	   collected	   by	   local	   Institute	   of	   Marine	   Affairs	   (IMA)	   in	   September	   2010	   (before	  bleaching	  induced	  coral	  mortality)	  and	  in	  March	  2011	  (after	  bleaching	  induced	  coral	  mortality).	  The	  previous	  surveys	  collected	  length	  and	  width	  of	  each	  living	  colony	  along	  four	  10	  m	  by	  2	  m	  belt	  transects.	  The	  mean	  of	  the	  width	  and	  length	  measurements	  was	  employed	  in	  this	  analysis.	  	  As	   surveys	   in	   2010	   and	   2011	   covered	   a	   total	   of	   80	   m2,	   eight	   transects	   were	   randomly	  subsampled	  from	  the	  10	  transects	  assessed	  in	  2013.	  	  3.2.3 Statistical	  analysis	  All	  data	  analysis	  was	  carried	  out	  in	  in	  R	  version	  2.15.1.	  Colony	  size	  data	  were	  used	  to	  calculate	  mean	  size,	  standard	  deviation,	  standard	  error,	  median,	  skewness	  and	  skewness	  standard	  error	  for	   each	   dominant	   taxon	   per	   site	   and	   year.	   Skewness	   values	   greater	   than	   two	   times	   the	  skewness	  standard	  error	  were	  considered	  to	  be	  significantly	  skewed	  from	  normal	  (McClanahan	  et	   al.	   2008).	   For	   each	   taxon	   percent	   cover	   estimates	   were	   calculated	   per	   site	   following	   the	  index	  developed	  by	  Done	  et	  al	  (2010)	  (which	  was	  based	  on	  Marsh	  et	  al	  1984):	  𝐶 = 100×(𝑁𝜋𝐷?/4)/[(𝑊 + 𝐷)𝐿],	  where	  C	  is	  the	  percentage	  cover,	  N	  is	  the	  number	  of	  colonies,	  D	  is	  the	  mean	  dimension	  of	  the	  N	  colonies,	  W	  is	  the	  width	  of	  the	  belt	  transect	  in	  centimeters	  (100	  cm)	  and	  L	  is	  the	  total	  length	  of	  the	  belt	   transects.	  This	  equation	  makes	   the	  assumption	   that	  all	   colonies	  are	  circular	   in	  shape	  with	  a	  diameter	  equal	  to	  the	  mean	  lateral	  dimension	  recorded	  and	  that	  the	  width	  is	  the	  width	  of	   the	   actual	   belt	   plus	   the	  mean	   diameter	   of	   all	   colonies	   to	   account	   for	   colonies	   extending	  beyond	  the	  100	  cm	  belt	  width.	  Though	  this	  index	  tends	  to	  over-­‐estimate	  true	  percent	  of	  coral	  	   26	  cover,	  the	  index	  provides	  us	  with	  comparable	  values	  to	  assess	  the	  changes	  occurring	  in	  the	  total	  live	  coral	  cover	  per	  species	  and	  sites.	  	  To	  test	  whether	  each	  species’	  mean	  size	  and	  size	  frequency	  distribution	  differed	  between	  the	  years	  and	  between	  the	  sites,	  each	  distribution	  was	  first	  tested	  for	  normality	  using	  Shapiro-­‐Wilk	  test	  and	  for	  homogeneity	  of	  variance	  by	  Levene’s	  test.	  The	  majority	  of	  distributions,	  before	  and	  after	   log-­‐transformation,	   were	   not	   normal	   and	   did	   not	   exhibit	   homoscedasticity.	   Therefore,	  non-­‐parametric	   Kruskal–Wallis	   tests	  were	   	   employed	   to	   test	   for	   the	   significant	   differences	   in	  size	  across	  each	  years	  and	  sites	  (McClanahan	  et	  al.	  2008),	  followed	  by	  post-­‐hoc	  pairwise	  Mann-­‐Whitney	   U-­‐tests	   comparisons,	   and	   	   Kolmogorov-­‐Smirnov	   tests	   to	   test	   for	   differences	   in	   size	  frequency	  distribution	  (Adjeroud	  et	  al.	  2007).	  To	  avoid	  Type	  I	  errors	  across	  multiple	  comparison	  tests,	  critical	  values	  for	  all	  tests	  were	  adjusted	  using	  the	  Bonferroni	  correction,	  resulting	  in	  an	  α-­‐level	  of	  0.0167.	  	  The	  impact	  of	  the	  bleaching	  event	  on	  the	  coral	  community	  of	  each	  site	  over	  the	  three	  years	  was	  visualised	   with	   a	   non-­‐metric	   multidimensional	   scaling	   (NMDS)	   based	   on	   Bray-­‐Curtis	  dissimilarities	  of	  each	  species	  abundances	  per	  site,	  using	  vegan	  package	  version	  2.0-­‐10	  (Borcard	  et	  al.	  2011).	  Changes	  in	  mean	  percent	  live	  coral	  cover	  per	  site	  (based	  on	  photoquadrat	  benthic	  assessment	  data	  using	  number	  of	  transects	  as	  sampling	  units)	  were	  calculated	  using	  a	  one-­‐way	  ANOVA	   followed	  by	   a	   Tukey’s	   honest	   significant	   difference	   test,	   after	   testing	   assumptions	   of	  homogeneity	   of	   variance	   and	   normality	   were	  met	   using	   graphical	   methods.	   Angel	   Reef	   was	  excluded	  from	  all	  historical	  analysis,	  as	  data	  was	  not	  available.	  	  3.3 Results	  3.3.1 Changes	  in	  percent	  coral	  cover	  	  The	  annual	  estimated	  percent	  live	  coral	  cover,	  based	  on	  the	  photo-­‐quadrat	  assessments,	  before	  the	  bleaching	  event	  in	  2010	  across	  the	  five	  sampled	  reefs	  ranged	  from	  15.6	  ±10.7	  to	  28.3	  ±11.0	  (Table	  1).	  After	  the	  2010	  bleaching	  event,	  mean	  coral	  cover	  declined	  by	  ≥35%	  at	  all	  sites	  except	  Blackjack	   Hole,	   but	   this	   decline	   was	   only	   statistically	   significant	   at	   the	   Culloden	   sites	   (Tukey	  HSD,	  <0.05).	  Coral	  cover	  appears	  to	  have	  undergone	  little	  recovery	  by	  between	  2011	  and	  2013;	  no	  significant	  change	  in	  coral	  cover	  was	  determined	  throughout	  this	  recovery	  period.	  Although	  there	  is	  no	  historical	  data	  for	  Angel	  Reef,	  percent	  live	  coral	  cover	  at	  this	  site	  in	  2013	  was	  three	  times	  higher	  than	  at	  all	  other	  sites,	  including	  Black	  Jack	  Hole,	  which	  is	  less	  than	  1km	  away.	  	  	  	   27	  Table	  6.	  Mean	  percent	  cover	  of	  live	  coral	  and	  (±)	  standard	  deviation	  estimated	  at	  each	  site	  and	  year.	  Values	  with	  the	  same	  letter	  subscript	  indicated	  significant	  pairwise	  comparison	  (p<0.05)	  Site	   Year	  (#	  of	  transects)	  2010(5)	   2011(10)	   2012(10)	   2013(10)	  Outer	  Buccoo	   28.3	  ±11.0	   18.1	  ±12.0	   19.8	  ±9.7	   20.2	  ±5.9	  Western	  Buccoo	   20.2	  ±10.6	   12.5	  ±9.2	   13.3	  ±8.7	   13.6	  ±8.4	  Culloden	  East	   24.3	  ±10.8abc	   14.3	  ±5.7a	   13.9	  ±7.3b	   12.2	  ±5.23c	  Culloden	  West	   27.6	  ±5.8abc	   12.5	  ±5.5a	   10.9	  ±4.5b	   16.0	  ±7.6c	  Black	  Jack	  Hole	   15.6	  ±10.7	   17.2	  ±5.8	   	   13.0	  ±7.1	  Angel	  Reef	   	   	   	   46.4	  ±14.6	  	  3.3.2 Changes	  in	  coral	  population	  structure	  and	  community	  composition	  Overall	  3671	  scleractinian	  coral	  colonies	  were	  measured	  among	  which	  27	  species.	  Total	  colony	  abundance	   at	   each	   reef	   site	   after	   the	   bleaching	   event	   only	   changed	   by	   10%	   or	   less	   at	   the	  Buccoo	  sites	  and	  Culloden	  West,	  but	  increased	  by	  23%	  at	  Black	  Jack	  Hole	  and	  decreased	  by	  26%	  at	   Culloden	   East	   (Table	   2).	   However	   from	   2011	   to	   2013,	   the	   number	   of	   colonies	   increased	  markedly	   at	   most	   sites,	   except	   Western	   Buccoo.	   Changes	   in	   species	   richness	   showed	   no	  consistent	  pattern	  of	  change	  across	  sites.	  Table	  7.	  Total	  number	  of	  coral	  colony	  and	  species	  (in	  parenthesis)	  recorded	  per	  site	  	  Sites	   2010	  (Pre-­‐bleaching)	   2011	  (Post-­‐bleaching)	   2013	   %	  change	  between	  	  	  	  2010-­‐11	  	  	  	  	  	  	  2011-­‐13	  Outer	  Buccoo	   192	  (16)	   209	  (12)	   381	  (14)	   9%	   82%	  Western	  Buccoo	   264	  (11)	   248	  (14)	   230	  (15)	   -­‐6%	   -­‐7%	  Culloden	  East	   186	  (16)	   138	  (14)	   246	  (15)	   -­‐26%	   78%	  Culloden	  West	   99	  (10)	   109	  (13)	   328	  (16)	   10%	   201%	  Black	  Jack	  Hole	   200	  (10)	   245	  (14)	   292	  (11)	   23%	   19%	  Angel	  Reef	   	   	   304	  (14)	   	   	  	  The	  most	  dominant	  coral	  species	   in	  2010,	  based	  on	  abundance	  among	  the	  five	  surveyed	  sites	  were:	   Montastraea	   faveolata,	   Diploria	   strigosa,	   Siderastrea	   siderea,	   Agaricia	   spp.,	   Porites	  astreoides,	   Colpophyllia	   natas,	   Montastraea	   cavernosa,	   and	   Diploria	   labyrinthiformis.	   M.	  faveolata	  was	  by	   far	   the	  most	  dominant	   species	   at	   all	   sites	   in	   terms	  of	  percent	   cover	   and	   in	  many	   cases	   also	   dominated	   in	   colony	   abundance.	   Most	   of	   these	   abundant	   species	   were	  massive	  and	  encrusting	  coral	  types,	  and	  belong	  to	  the	  Faviidae	  family	  with	  the	  exception	  of	  S.	  siderea,	  P.	  astreoides	  and	  Agaricia	  spp..	  Species	  with	  large	  populations	  at	  all	  sites	  (≥12	  colonies)	  included	  M.	  faveolata,	  S.	  siderea	  and	  D.	  Strigosa	  (except	  at	  Black	  Jack	  Hole).	  M.	  cavernosa	  was	  common	  only	  at	  Culloden	  sites	  and	  Black	  Jack	  Hole,	  while	  C.	  natas	  and	  D.	  labyrinthiformis	  were	  	   28	  only	  abundant	  enough	  at	  the	  Buccoo	  sites.	  Weedy	  species	   like	  Agaricia	  spp.	   frequented	  at	  all	  sites	   except	   Black	   Jack	   Hole,	   where	   P.	   astreoides	   was	  more	   common	   despite	   being	   sparsely	  present	   at	   the	   other	   sites	   (see	   Appendix	   Table	   A1	   for	   summary	   statistics	   of	   each	   dominant	  species	  per	  site).	  	  The	   changes	   among	   each	   dominant	   species	   over	   time	   varied	   greatly	   per	   site	   in	   terms	   of	  abundance	  and	  percent	  cover	  (determined	  from	  size	  frequency	  data).	  We	  generally	  found	  that	  after	   the	  bleaching	  event,	   the	  majority	  of	   coral	   	  populations	  at	  each	   site	   saw	  decline	   in	   total	  percent	   cover	   as	   determined	   from	   the	   size	   frequency	   data.	   While	   colony	   abundance	   also	  declined	   among	   most	   population,	   many	   remained	   relatively	   unchanged	   or	   increased	   in	  abundance,	  even	  among	  some	  populations	  that	  saw	  a	  decline	  in	   	  percent	  cover.	  For	  example,	  the	  abundance	  of	  colonies	  for	  M.	  faveolata	  at	  Black	  Jack	  Hole	  and	  C.	  natans	  at	  Outer	  Buccoo	  increased	   after	   the	   bleaching	   event	   but	   their	   percent	   cover	   almost	   halved.	   However,	   most	  Agaricia	  spp.	  and	  P.	  astreoides	  populations	  did	  not	  decline	  in	  abundance	  nor	  percent	  cover.	  Before	  the	  bleaching	  event,	  the	  coral	  populations	  at	  each	  site	  were	  mostly	  dominated	  by	  small	  to	  medium	  sized	  colonies,	  as	  the	  skewness	  of	  most	  size	  distributions	  was	  significantly	  positive	  or	  slightly	  positive	  (Appendix	  Table	  A1).	   	  The	  few	  normally	  skewed	  distributions	  compromised	  of	  the	  populations	  of	  P.	  astreoides,	  C.	  natas,	  D.	  labyrinthiformis,	  and	  S.	  siderea	  at	  Culloden	  sites	  and	  Western	  Buccoo.	  After	  the	  bleaching	  event,	  we	  found	  that	  skewness	  became	  slightly	   less	  positive	  or	  tended	  towards	  normality	  among	  some	  species	  (Appendix	  Table	  A1).	  However	  some	  other	  coral	  populations,	  like	  that	  of	  M.	  faveolata	  at	  most	  sites,	  became	  more	  positively	  skewed	  after	  the	  bleaching.	  At	  each	  site	  we	  found	  that	  the	  size	  frequency	  distributions	  for	  the	  majority	  of	  species	  did	  not	  significantly	  change	  (KS-­‐test,	  P<0.016)	  between	  2010	  and	  2011	  (Appendix	  Table	  A2).	  	  Significant	  change	  was	  found	  among	  three	  species:	  Agaricia	  spp.	  at	  Outer	  Buccoo,	  S.	  siderea	  at	  Western	  Buccoo	  and	  M.	  faveolata	  at	  Black	  Jack	  Hole.	  These	  populations	  showed	  a	  decline	   in	  the	   large	  size	  classes	  and	  an	  increase	  in	  the	  smaller	  size	  classes	  (Figure	  8).	  Mean	  colony	  size	  (diameter)	  decreased	   among	   some	   species	   and	   sites	   by	   2011	   (Figure	   9),	   especially	   among	   S.	   siderea,	  Agaricia	  spp.,	  P.	  astreoides	  and	  C.	  natans.	  However	  among	  most	  species	  mean	  colony	  size	  did	  not	  significantly	  differ	  (Mann-­‐Whitney	  test,	  P<0.016)	  after	  the	  bleaching	  event.	  The	  only	  colony	  mean	  sizes	  that	  did	  significantly	  declined	  were	  for	  Agaricia	  spp.	  at	  Outer	  Buccoo,	  S.	  siderea	  at	  Western	  Buccoo	  and	  M.	  faveolata	  at	  Outer	  Buccoo	  and	  Black	  Jack	  Hole	  (Figure	  9).	  	   29	  	  	  	  Figure	  8.	  Size	  frequency	  distributions	  of	  coral	  taxa	  at	  sites	  with	  significant	  differences	  between	  years,	  as	  determined	  by	  the	  Kolmogorov-­‐Smirnov	  test	  (*)	  	  	  	   30	  	  	  	  	  Figure	  9.	  Boxplot	  and	  mean	  size	  (white	  filled	  dots)	  per	  species	  per	  site	  indicating	  changing	  trends	  in	  colony	  size	  between	  each	  year.	  Letter	   in	  each	  plot	   indicate	   if	  there	  was	  a	  significant	  difference	  (Mann-­‐Whitney	  U	  test,	  P<	  0.016)	  between	  2010-­‐2011	  (A),	  2010-­‐2013	  (B),	  2011-­‐2013	  (C).	  	  	  Between	  2011	   and	   2013	  we	  observed	   that	   colony	   abundance	   among	  most	   coral	   populations	  had	   increased,	   though	   quite	   a	   few	   remained	   unchanged.	   While,	   percent	   cover	   did	   increase	  among	  some	  populations,	   in	  many	  cases	   it	   remained	  unchanged	  and	   in	  some	   it	  declined.	  We	  noted	  that	  abundance	  of	  P.	  astreoides	  and	  Agaricia	  spp.	  more	  than	  doubled	  by	  2013	  at	  most	  sites	   and	   their	   percent	   cover	   also	   increased	   or	   stayed	   the	   same.	   On	   the	   other	   hand	   colony	  	  	  	  *2010-­‐11	  	  	  2010-­‐13	  *2010-­‐11	  	  	  2011-­‐13	  	  	  2010-­‐13	  *2010-­‐11	  	  	  2010-­‐13	  *2010-­‐11	  	  *2011-­‐10	  	  	  2010-­‐13	  *2010-­‐13	  	  	  2011-­‐13	  *2010-­‐13	  	  *2010-­‐13	  	  *2011-­‐13	  	  *2010-­‐13	  	  	  2011-­‐13	  	   31	  abundance	  and	  cover	  among	  C.	  natas,	  D.	  labyrinthiformis	  and	  S.	  siderea	  populations	  remained	  relatively	   stable	  or	  declined.	  We	  also	  noted	  Black	   Jack	  Hole	  most	  of	   its	  dominant	   species,	  M.	  faveolata,	  M	  cavernosa	  and	  S.	  siderea,	  which	  lost	  ≥45%	  of	  its	  percent	  cover	  by	  2011,	  showed	  no	  recovery	  by	  2013	  and	  M.	  cavernosa	  had	  declined	  further.	  By	   2013	  most	   species	   population	   distributions	   became	  more	   positively	   skewed	   than	   in	   2011	  and	   2010,	   indicating	   a	   greater	   proportion	   of	   smaller	   sized	   colonies	   with	   some	   exceptions	  (Appendix	  Table	  A1).	  Additionally,	  a	  visible	  rise	  of	  smaller	  size	  classes	  was	  noted	  among	  the	  size	  frequency	  distribution	  of	  many	  species	  (Figure	  8).	  However,	  we	  found	  that	  for	  the	  majority	  of	  species	  no	  significant	  difference	  was	  detected	  among	  the	  size	  frequency	  distribution	  and	  mean	  colony	   size	   between	   2011	   and	   2013.	   The	   only	   significant	   changes	   were	   found	   among	   the	  populations	   of	   M.	   faveolata,	   S.	   siderea	   and	   P.	   astreoides	   at	   Black	   Jack	   Hole,	   as	   well	   as	   for	  Agaricia	   spp.	   at	   Culloden	  West	   and	  D.	   strigosa	   at	   Culloden	   East	   (Appendix,	   Table	   A2).	   All	   of	  these	  populations	  by	  2013	  showed	  an	  increase	  in	  the	  smaller	  size	  classes	  large	  size	  classes	  and	  some	  a	  decline	   in	  the	   larger	  size	  classes	   (Figure	  8).	  More	  populations	  experienced	   	  significant	  changes	   in	   their	   size	   frequency	   distributions	   and	  mean	   colony	   size	   from	   before	   bleaching	   in	  2010	  to	  2013,	  than	  from	  after	  bleaching	  in	  2011	  to	  2013	  (Appendix,	  Table	  A2).	  	  	  Figure	  10.	  Non-­‐metric	  multidimensional	  scaling	  using	  Bray-­‐Curtis	  dissimilarities	  plot	  of	   the	  qualitative	  changes	  among	   the	   coral	   communities	   at	   each	   of	   the	   sites	   per	   year	   (named	   and	   colour	   coded)	   across	   Tobago.	  NMDS	  stress	  =	  0.133.	  	  	   32	  In	   terms	   the	   community	   species	   composition	   at	   each	   site,	   we	   found	   the	   species	   that	   were	  dominant	   in	   2010	   continued	   to	   be	   dominant	   by	   2013.	   The	   nMDS	   analysis,	   based	   on	   colony	  abundances	   of	   all	   species	   present	   per	   site,	   illustrates	   (Figure	   4)	   how	   the	   community	  composition	   among	   most	   sites	   remained	   relatively	   similar	   across	   the	   three	   assessed	   years,	  especially	  among	  the	  Culloden	  Sites.	   	  The	  greatest	  community	  composition	  change	  from	  2010	  to	  2013	  occurred	  at	  Black	  Jack	  Hole	  possibly	  due	  to	  an	  increase	  in	  P.	  astreoides	  and	  decline	  of	  M.	   cavernosa.	   The	   community	   composition	   shift	   at	   the	   Buccoo	   sites	   by	   2013	   also	   stand	   out,	  which	   was	   most	   likely	   due	   to	   an	   increase	   in	   abundance	   of	   Agaricia	   spp.	   Intraspecific	  comparisons	   of	   populations	   between	   the	   different	   reefs	   sites	   also	   showed	   that	   species	   size	  frequency	   distributions	   did	   not	   significantly	   differ	   between	   most	   sites.	   The	   few	   significant	  differences	  were	  mostly	  found	  in	  2013	  (Appendix	  Table	  3).	  	  3.4 Discussion	  Even	   though	   severe	   mass	   bleaching	   events	   are	   often	   followed	   by	   extensive	   loss	   of	   coral	  mortality	   (Hughes	   et	   al.	   2003;	   Baker	   et	   al.	   2008),	   the	   overall	   impact	   of	   this	   bleaching	   event	  appears	   to	  have	  been	   low	  for	  most	  of	   the	  assessed	  coral	  populations.	  This	  may	  be	  related	  to	  the	   fact	   that	   coral	   cover	   at	   these	   reef	   systems	   was	   already	   low,	   having	   been	   affected	   by	   a	  historical	  loss	  of	  herbivores,	  water	  pollution,	  and	  more	  recently	  from	  the	  2005	  bleaching	  event	  and	   subsequent	   disease	   outbreaks.	   Thus	   it	   is	   likely	   that	   by	   the	   time	  of	   the	   2010	   event	   coral	  assemblages	  were	  already	  narrowed	  down	  to	  more	  resistant	  species,	  like	  the	  massive	  species	  of	  Montastraea,	   Diploria,	   Colpophyllia	   and	   Siderastrea	   that	   have	   thick	   tissues	   and	   large	   inter-­‐corallite	  spacing	  	  (Baker	  et	  al.	  2008;	  Ritson-­‐williams	  et	  al.	  2009).	  	  We	   found	   that	   following	   the	   heat	   stress	   in	   2010	   coral	   populations	   showed	   signs	   of	   having	  experienced	  some	  bleaching-­‐induced	  mortality.	  Many	  species	  experienced	  a	  decline	   in	  colony	  abundance;	  percent	  cover	  and	  mean	  colony	  size	  by	  2011,	  symptomatic	  of	  corals	  having	  suffered	  complete	  mortality	  and/or	  partial	  mortality.	  Nonetheless,	  almost	  all	  populations’	  size	  frequency	  distributions	   and	   colony	   mean	   size	   did	   not	   significantly	   differ	   between	   2010	   and	   2011,	  indicating	  that	  the	  bleaching	  event	  did	  not	  have	  a	  major	  impact	  on	  the	  population	  size	  structure	  of	   most	   species..	   The	   composition	   of	   the	   dominant	   coral	   community	   also	   remained	   largely	  unchanged	   following	   the	   bleaching	   event.	   Additionally,	   mean	   percent	   cover,	   based	   on	  photoquadrat	   transects,	   only	   declined	   significantly	   at	   Culloden	   sites,	   whilst	   no	   significant	  changed	  was	  detected	  at	   the	  other	   three	  sites.	  However,	   the	   lack	  of	  statistical	   significance	   in	  the	  decline	  at	  the	  other	  sites	  	  may	  be	  an	  artefact	  of	  the	  sampling	  method;	  given	  that	  coral	  cover	  	   33	  is	   low	  across	  Tobago	  and	   that	   several	   species	  are	   rare	  at	   some	  sites,	   the	   individual	   transects	  may	  be	  too	  short	  to	  be	  representative	  of	  coral	  cover	  or	  population	  size	  structure.	  	  By	  2013,	  most	  species	  population	  size	  structure	  did	  not	  significantly	  differ,	  suggesting	  that	  most	  populations	  remained	  unchanged	  throughout	  the	  two	  years	  following	  the	  bleaching	  event.	  We	  did	  observe	   that	   in	  many	  populations,	   though	   colony	  abundance	   increased	  by	  2013,	   this	   rise	  was	   not	   always	   accompanied	   by	   an	   increase	   in	   percent	   cover,	   indicating	   that	   perhaps	   coral	  colonies	  may	  have	  been	  affected	  by	  further	  fragmentation	  in	  the	  years	  following	  the	  bleaching.	  Additionally,	  the	  skewness	  of	  most	  populations	  in	  2013	  became	  more	  positive	  and	  there	  was	  a	  noticeable	  increase	  in	  the	  smallest	  size	  class	  (5-­‐15cm)	  among	  many	  species.	  The	  rise	  of	  smaller	  size	  classes	  could	  suggest	  that	  there	  has	  been	  a	  input	  of	  juvenile	  colonies	  that	  have	  grown	  large	  enough	   (<4cm)	   in	   the	   past	   two	   years	   becoming	   part	   of	   adult	   demographic,	   which	  would	   be	  indicative	   of	   the	   first	   steps	   of	   post-­‐bleaching	   recovery	   taking	   place.	   However,	   findings	   from	  Chapter	   2	   indicate	   that	   recruits/juveniles	   densities	   were	   very	   low	   throughout	   Tobago	   and	  mainly	  compromised	  of	  brooding	  genera	  Agaricia	  or	  Porites,	  which	  means	  this	  was	  unlikely	  the	  case	  for	  most	  other	  species.	  It	  is	  possible	  that	  the	  rise	  in	  small	  sized	  colonies	  between	  2011	  and	  2013	   among	   the	   massive	   species	   may	   have	   instead	   resulted	   from	   colonies	   undergoing	  fragmentation	   due	   to	   partial	   mortality	   driven	   by	   other	   disturbances,	   such	   as	   predation,	  sedimentation	   or	   disease.	   Disease	   could	   be	   the	   most	   likely	   explanation,	   given	   that	   the	   IMA	  recorded	  a	  rise	  in	  disease	  infected	  corals	  after	  the	  bleaching	  event,	  and	  after	  the	  2011	  survey	  used	  in	  this	  study	  (Alemu	  I	  2011).	  Many	  studies	  have	  found	  that	  disease	  outbreaks	  tend	  to	  be	  the	  primary	  cause	  of	  post-­‐bleaching	  coral	  mortality,	  whilst	  few	  corals	  actually	  die	  of	  bleaching	  itself	  (Miller	  et	  al.	  2006;	  Wilkinson	  &	  Souter	  2008;	  Brandt	  &	  McManus	  2009)	  Though	  the	  response	  to	  the	  2010	  bleaching	  event	  varied	  greatly	  among	  and	  within	  species	  and	  sites,	   some	   noticeable	   patterns	   of	   change	   were	   detected	   across	   all	   sites.	   S.	   siderea,	   one	   of	  Tobago’s	   main	   reef-­‐building	   species,	   was	   most	   impacted	   by	   the	   2010	   bleaching	   event.	   The	  decline	   of	   the	  mean	   colony	   size	   continued	   after	   the	   bleaching	   event,	   likely	   due	   to	   the	   post-­‐bleaching	  fragmentation	  discussed	  above.	  Other	  Caribbean	  studies	  have	  reported	  S.	  siderea	  to	  be	   among	   the	   species	   most	   susceptible	   to	   bleaching,	   bleaching	   induced	   mortality	   (van	  Hooidonk	  et	  al.	  2012;	  Oxenford	  et	  al.	  2007)	  and	  becoming	   infected	  by	  diseases	   that	   typically	  occur	  following	  bleaching	  (Gochfeld	  et	  al.	  2006).	  	  Significant	  population	  changes	  were	  also	  noted	  among	  the	  weedy	  species	  Agaricia	  spp.	  and	  P.	  astreoides	   (abundant	   only	   at	   Black	   Jack	   Hole).	   While	   their	   populations	   remained	   relatively	  unchanged	  following	  the	  bleaching	  event	  (2011),	  by	  2013	  abundance	  of	  both	  corals	  more	  than	  	   34	  doubled	   in	   size	   and	   increased	   in	   percent	   cover	   with	   a	   conspicuous	   rise	   in	   the	   smaller	   size	  classes.	   This	   rise	   in	   smaller	   size	   colonies	  most	   likely	   came	   from	   an	   input	   of	   juveniles.	   These	  brooding	   species	   tend	   be	   have	   high	   recruitment	   and	   post-­‐settlement	   survival	   rates	   as	   their	  recruits	   initially	   grow	   quickly	   in	   comparison	   to	   other	   taxa	   (Arnold	   &	   Steneck	   2011).	   This	   is	  supported	   by	   our	   finding	   in	   Chapter	   2,	   as	   the	   genera	   Agaricia	   and	   Porites	   compromised	   of	  between	  30-­‐80%	  of	   all	   juveniles	   at	   the	   sites	   at	  which	   they	  were	   present.	   Thus,	   these	   results	  indicate	   that	   both	   Agaricia	   spp.	   and	   P.	   astreoides	   appear	   to	   favor	   post-­‐disturbance	   reef	  conditions	  highlighting	  their	  niche	  as	  opportunist	  and	  weedy	  species	  (Hughes	  &	  Jackson	  1985).	  	  Our	  results	  also	  revealed	  that	   	  population	  size	  structures	  between	  the	  assessed	  reef	  sites	  did	  not	  significantly	  differ,	  despite	  Buccoo	  reefs	  being	  adjacent	  to	  urbanized	  land	  and	  its	  sites	  were	  likely	   to	   be	  more	   environmentally	   stressed	   due	   to	   poor	  water	   quality	   (Lapointe	   et	   al.	   2010).	  Species’	   population	   size	   structure	   at	   all	   sites	   tended	   to	   be	   positively	   skewed	   as	   the	  majority	  were	  dominated	  by	  small	   to	  medium	  sized	  colonies	  throughout	  the	  three	  years.	  The	  only	  site	  that	   set	   itself	   apart	   was	   Black	   Jack	   Hole,	   which	   out	   of	   all	   sites	   most	   of	   its	   dominant	   coral	  species’	   population	   size	   structures	   became	   altered	   following	   the	   bleaching	   event.	   The	  populations	  of	  M.	  faveolata,	  M.	  cavernosa	  and	  S.	  siderea,	  saw	  a	  decline	  in	  mean	  colony	  size	  and	  percent	  cover	  after	   the	  bleaching	  by	  2011	  and	   in	   the	  two	  years	   following.	  This	   indicated	  that	  the	   coral	   populations	   likely	   suffered	   from	   bleaching-­‐induced	   partial	   and	   complete	  mortality,	  and	  the	  surviving	  individuals	  may	  have	  succumbed	  after	  the	  bleaching	  event.	  Correspondingly,	  Alemu	  I	  and	  Clement	  (2014),	  who	  carried	  out	  the	  bleaching	  assessment	  at	  the	  time	  of	  the	  2010	  heat	   stress,	   found	   that	   reefs	   in	   Speyside	   had	   the	   highest	   bleaching	   and	   mortality	   response	  among	  the	  three	  reef	  systems	  studied	  (same	  ones	  as	  in	  this	  study).	  	  Taken	  together,	  the	  results	  indicate	  that	  acute	  disturbances,	  such	  as	  bleaching	  events,	  can	  lead	  to	  the	  decline	  of	  the	  mean	  colony	  size	  of	  species’	  populations	  due	  to	  the	  dominance	  of	  smaller	  sized	  colonies	  as	  a	  result	  of	  fragmentation.	  Similar	  findings	  were	  established	  among	  Kenya	  reefs	  based	   on	   the	   impact	   of	   a	   bleaching	   event	   in	   2005	   using	   the	   similar	   definitions	   of	   an	  independent	   coral	   colony	   (McClanahan	   et	   al.	   2008).	   Coral	   colonies	   that	   have	   experienced	  partial	   mortality	   need	   to	   prioritize	   their	   energy	   sources	   to	   recover	   from	   lesions	   and	   often	  postpone	   reproduction	   (Hughes	   et	   al.	   2003).	   Therefore,	   considering	   that	   smaller	   colonies	  produce	   fewer	   gametes	   and	   are	   also	  more	   susceptible	   to	   other	   disturbances	   like	   disease	   or	  sedimentation	   (Graham	  &	   van	  Woesik	   2013),	   a	   reduction	   of	  mean	   colony	   size	   can	   lead	   to	   a	  reduction	   in	   fecundity	   and	   overall	   resilience	   to	   disturbance.	   This	   will	   slow	   down	   post-­‐disturbance	  recovery	  as	  it	  declines	  the	  chances	  newly	  dead	  corals	  becoming	  rapidly	  recolonized	  	   35	  by	   coral	   recruits	   and	   instead	  enabling	   the	   invasive	  establishment	  algae	  and	   sponges	   (Connell	  1997).	  Thus,	  overall	  acute	  disturbances	  that	  drive	  fragmentation	  such	  as	  bleaching,	  pose	  a	  large	  threat	  Tobago’s	  fragile	  communities.	  In	   conclusion,	   from	  assessing	   the	   changes	   in	   population	   size	   structure	   among	   Tobago’s	  most	  dominant	  coral	  species,	  we	  found	  that	  bleaching	  events	  caused	  by	  heat	  stress,	  such	  as	  in	  2010,	  act	  as	  chronic	  disturbances	  as	  they	  lead	  to	  the	  slow	  shrinkage	  of	  Tobago’s	  coral	  communities.	  We	  found	  many	  populations	  of	  Tobago’s	  coral	  communities	  resisted	  become	  heavily	  impacted	  by	   the	   heat	   stress.	  Nevertheless,	  we	   also	   found	   evidence	   of	   some	  population’s	  mean	   colony	  size	  decreasing,	  including	  most	  S.	  siderea	  populations	  and	  those	  of	  Black	  Jack	  Hole’s	  dominant	  species.	  We	  also	  found	  that	  the	  majority	  of	  population	  that	  experienced	  post-­‐bleaching	  decline	  instead	   of	   recovering,	   the	   mean	   colony	   size	   continued	   to	   the	   declined	   in	   the	   two	   years	  following	   the	   bleaching	   event.	   Considering	   that	   heat	   stress	   events	   are	   predicted	   to	   be	  more	  frequent	   and	   intense	   in	   the	   next	   20	   to	   30	   years	   (Donner	   et	   al.	   2007;	   Hoegh-­‐Guldberg	   et	   al.	  2007),	   it	   is	   likely	   that	   the	  population	  size	  structure	  of	  Tobago’s	  coral	   community	  will	  become	  dominated	  by	  smaller	  sized	  colonies	  leading	  to	  the	  shrinkage	  of	  populations’	  mean	  colony	  size	  and	  overall	  size.	  As	  this	  will	  impact	  the	  reproductive	  output	  of	  coral	  communities,	  species	  that	  are	  naturally	  more	   fecund	   such	  as	  brooding	   species	  Agaricia	   spp.	   and	  P.	  astreoides	   are	   likely	  going	  to	  become	  more	  dominant.	  	  	   36	  Chapter	  4.	   Conclusion	  In	  this	  study	  I	  assessed	  the	  recovery	  of	  coral	  communities	  of	  three	  representative	  reef	  systems	  around	   the	   southern	   Caribbean	   island	   of	   Tobago	   following	   the	   2010	   mass	   bleaching.	   To	  understand	  which	  coral	  taxa	  are	  producing	  sexual	  recruits	  that	  can	  grow	  successfully	  surviving	  on	   among	   Tobago’s	   reefs	   in	   2013	   I	   assessed	   the	   abundance	   and	   composition	   of	   the	   juvenile	  community	  at	  each	  site	  (Chapter	  2).	  As	  coral	  growth	  and	  regeneration	  is	  affected	  by	  the	  impact	  of	  sedimentation,	  deposition	  rates	  and	  the	  composition	  of	  sediment	  deposited	  at	  each	  reef	  site	  were	   also	   evaluated	   (Chapter	   2).	   Finally,	   to	   assess	   the	   impact	   and	   recovery	   of	   the	   bleaching	  event,	  I	  examined	  the	  changes	  in	  the	  size	  population	  structure	  of	  Tobago’s	  dominant	  coral	  taxa	  using	  data	  collected	  before	  the	  bleaching	  episode	  in	  2010,	  immediately	  after	  bleaching	  ended	  in	  2011,	  and	  two	  and	  a	  half	  years	  after	  the	  bleaching	  event	  in	  2013	  (Chapter	  3).	  	  The	  results	  suggest	  very	  slow	  recovery	  of	  coral	  cover	  through	  substrate	  re-­‐colonization	  via	  the	  sexual	  production	  of	  recruits	  since	  the	  bleaching	  event.	  Similarly	  to	  other	  Caribbean	  reefs,	  we	  found	   low	   juvenile	  densities	  across	  most	  of	   the	  Tobago’s	   reefs.	  The	   juvenile	  coral	  community	  was	  dominated	  by	  weedy	  brooding	  taxa,	  while	  broadcasting	  corals	  represented	  the	  minority	  of	  juveniles	   found.	   Juveniles	   of	   the	   key	   framework	   building	   coral	   taxa	   in	   Tobago,	  Diploria	   and	  Siderastrea,	  were	  however	  present,	  but	  rare.	  The	  disparity	  between	  brooding	  vs.	  broadcasting	  juvenile	   abundances	   is	   consequence	   of	   the	   different	   life-­‐histories	   among	   these	   two	   types	   of	  corals.	  Nevertheless,	   the	  overall	   low	  abundance	  of	   juveniles	   is	  a	  sign	   that	   the	  vitality	  of	  coral	  populations	   across	   Tobago’s	   reefs	   and/or	   their	   environmental	   conditions	   have	   become	   so	  altered	  that	  they	  are	  no	  longer	  suitable	  for	  recruitment	  processes.	  The	   sedimentation	   assessment	   found	   that	   between	   May	   and	   June	   of	   2013,	   sedimentation	  deposition	   rates	  were	   below	   levels	   that	   tend	   to	   subject	   coral	   communities	   to	   stress	   at	  most	  sites.	  The	  exception	  was	  the	  Culloden	  West	  site	  where	  rates	  were	  three	  times	  higher	   than	  at	  the	   other	   sites;	   these	   higher	   sedimentation	   rates	   were	   likely	   caused	   by	   strong	   currents	  occurring	  at	   that	   site.	  A	   larger	   fraction	  of	   the	   sediment	  deposited	  at	   all	   reefs	   sites	  was	   likely	  coming	  from	  land,	  as	  most	  of	  it	  was	  composed	  of	  terrigenous	  material.	  Consequently,	  it	  is	  likely	  that	   sedimentation	   rates	   to	   be	   significantly	   higher	   during	   the	   rainy	   season,	   as	   terrestrial	  leaching	  and	  runoff	  from	  land	  into	  coastal	  zones	  tends	  to	  be	  much	  higher.	  Thus,	  it	  is	  likely	  that	  corals	  communities	  experience	  seasonal	  sedimentation	  stress,	  especially	  those	  near	  urbanized	  landscapes.	  Though	  I	  found	  sedimentation	  rates	  did	  not	  differ	  between	  reefs	  sites	  near	  urban	  or	   rural	   land,	   there	   was	   more	   silt/clay	   sized	   material	   at	   rural	   sites.	   Additionally	   the	   highest	  	   37	  juvenile	   density	   and	   diversity	   of	   juvenile	   corals	   were	   only	   at	   reefs	   adjacent	   to	   undeveloped	  land,	  whilst	  sites	  at	  Buccoo	  reef	  had	  the	  lowest	  density.	   I	  highly	  recommend	  that	  a	   long-­‐term	  sediment	   accumulation	  assessment	  be	  done	  among	  Tobago’s	   reefs	   that	   cover	  both	   the	   rainy	  and	  the	  dry	  seasons.	  Especially	  considering	  that	  most	  broadcasting	  taxa	  spawn	  during	  the	  rainy	  season	   as	   released	   gametes,	   larvae	   and	   recruits	   are	   particularly	   susceptible	   to	   becoming	  damaged	  by	  sedimentation	  (Ritson-­‐williams	  et	  al.	  2009).	  	  From	  examining	  changes	  in	  the	  population	  size	  structure	  before	  and	  after	  the	  bleaching	  event	  we	   found	  no	  significant	  change	   in	   the	  population	  structure	  of	   the	  majority	  of	  dominant	  coral	  species	   after	   undergoing	   severe	   heat	   stress	   in	   the	   late	   summer	   of	   2010.	   However,	   those	  populations	   that	  did	  significantly	  change,	  apart	   from	  experiencing	  declines	   in	   their	   total	  coral	  cover,	   had	   reduced	   mean	   colony	   size	   due	   to	   complete	   or	   partial	   mortality	   shrinking	   or	  fragmenting	   larger	   sized	   colonies	   into	   smaller	   ones.	   For	   example,	   this	   was	   the	   case	   for	   S.	  siderea,	  which	  was	   found	   to	  be	   among	   the	  most	   sensitive	   taxa	   to	   bleaching	  mortality	   across	  Tobago’s	  reefs,	  as	  well	  as	  for	  key	  reef	  building	  taxa	  (M.	  faveolata,	  M.	  cavernosa	  and	  S.	  siderea)	  at	  Black	  Jack	  Hole.	  This	  indicates	  that	  in	  the	  event	  of	  bleaching	  being	  more	  severe	  in	  the	  future,	  it	  may	  lead	  to	  coral	  populations’	  mean	  colony	  size	  declining	  due	  to	  extensive	  fragmentation	  and	  complete	  mortality	  among	  primary	  reef	  building	  coral	  species.	  	  In	   the	   two	  years	   following	  bleaching	  event	  at	  most	  sites	  no	  change	  was	  observed	  among	  the	  population	   size	   structure	   for	   the	  majority	   of	   coral	   species.	   	   This	   is	   not	   surprising,	   given	   slow	  coral	  growth	  rates;	  changes	  should	  again	  be	  examined	  over	  a	  longer	  time	  period,	  such	  as	  5	  to	  8	  years.	   Nonetheless,	   we	   did	   find	   that	   many	   taxa	   by	   2013	   had	   experience	   an	   increase	   in	   the	  abundance	   of	   small	   sized	   colonies,	   without	   a	  major	   change	   in	   coral	   cover,	   and	   leading	   to	   a	  significant	  decline	  in	  colony	  size	  among	  a	  few	  taxa.	  This	  increase	  in	  smaller	  sized	  colonies	  could	  have	   been	   caused	   by	   colonies	   experiencing	   post-­‐bleaching	   fragmentation	   due	   to	   other	  secondary	  disturbance	  such	  as	  disease.	  A	  major	  implication	  of	  coral	  population	  becoming	  more	  positively	  skewed	  (i.e.	  dominated	  by	  smaller	  sized	  corals)	  is	  a	  decreases	  the	  fecundity	  of	  coral	  populations.	  This,	  in	  turn,	  can	  affect	  the	  long-­‐term	  recovery	  of	  the	  reefs	  as	  it	  reduces	  the	  larvae	  and	  recruitment	  output	  slowing	  down	  coral	  populations	  overall	  regeneration.	  Among	  the	  only	  taxa	   that	  may	  have	  been	  starting	   to	   recover	  were	  Agaricia	   spp.	   and	  P.	  astreoides,	  which	  had	  high	  number	  of	  juveniles	  at	  the	  reef	  sites	  where	  they	  were	  present.	  Many	  Caribbean	  reefs’	  coral	  communities	  over	  the	  last	  decades	  have	  become	  narrowed	  down	  to	  the	  most	  resistant	  species.	  Though	  these	  communities	  can	  avoid	  major	  mortality	   following	  mass	   bleaching,	   they	   still	   become	   weakened	   and	   can	   experience	   enough	   individual	   full	   and	  	   38	  partial	   mortality	   to	   significantly	   change	   the	   population	   size	   structure.	   Thus,	   mass	   bleaching	  events	  greatly	  threaten	  the	  future	  persistence	  and	  health	  of	  Caribbean	  coral	  reefs	  like	  Tobago.	  This	  study	  indicates	  that	  across	  Tobago’s	  different	  reef	  sites,	  the	  bleaching	  disturbance	  can	  lead	  to	  a	  dominance	  of	   smaller	   size	   coral	   colonies,	  which	   could	  negatively	  affect	   the	   reproductive	  output.	  This	  could	  further	  decrease	  these	  coral	  communities	  already	  low	  ability	  to	  develop	  coral	  juveniles.	  Given	  the	  evidence	  that	  recruitment	  processes	  are	  already	  stunted	  among	  Tobago’s	  coral	  communities,	  as	  has	  been	  observed	  for	  other	  Caribbean	  reefs,	   it	   is	  essential	   to	   improve	  the	   current	   health	   of	   coral	   communities	   to	   increase	   the	   chances	   of	   successful	   recruitment	  taking	  place.	  Thus,	  it	  is	  paramount	  that	  local	  coastal	  ecosystem	  conservation	  and	  management	  efforts	   strive	   towards	   increasing	   herbivory	   and	   reducing	   sedimentation	   and	   nutrient	  enrichment.	  	  	  	   39	  Bibliography	  Abdullah,	  A.L.	  et	  al.,	  2011.	  Sediment	  Fallout	  Rates	  in	  Tanjung	  Rhu	  Coral	  Reefs.	  Kajiian	  Malaysia,	  29(2),	  pp.1–29.	  Adjeroud,	  M.	  et	  al.,	  2007.	  Small-­‐scale	  variability	  in	  the	  size	  structure	  of	  scleractinian	  corals	  around	  Moorea,	  French	  Polynesia:	  patterns	  across	  depths	  and	  locations.	  Hydrobiologia,	  589(1),	  pp.117–126.	  Alemu	  I,	  J.B.,	  2011.	  A	  Summary	  Report	  on	  the	  Impacts	  of	  the	  2010	  Mass	  Coral	  Bleaching	  Event	  in	  Tobago,	  Port	  of	  Spain,	  Trinidad	  and	  Tobago.	  Alemu	  I,	  J.B.	  &	  Clement,	  Y.,	  2014.	  Mass	  coral	  bleaching	  in	  2010	  in	  the	  southern	  Caribbean.	  PloS	  one,	  9(1),	  p.e83829.	  Alemu	  I,	  J.B.	  &	  Clement,	  Y.,	  2012.	  Status	  of	  the	  coral	  reefs	  of	  Tobago-­‐	  2012	  and	  summary	  report	  2010-­‐2012,	  Alvarez-­‐Filip,	  L.	  et	  al.,	  2011.	  Drivers	  of	  region-­‐wide	  declines	  in	  architectural	  complexity	  on	  Caribbean	  reefs.	  Coral	  Reefs,	  30(4),	  pp.1051–1060.	  Arnold,	  S.N.	  &	  Steneck,	  R.S.,	  2011.	  Settling	  into	  an	  increasingly	  hostile	  world:	  the	  rapidly	  closing	  “recruitment	  window”	  for	  corals.	  PloS	  one,	  6(12),	  p.e28681.	  Babcock,	  R.,	  1991.	  Comparative	  Demography	  of	  Three	  Species	  of	  Scleractinian	  Corals	  Using	  Age-­‐	  and	  Size-­‐	  Dependent	  Classifications.	  Ecological	  Society	  of	  America,	  61(3),	  pp.225–244.	  Babcock,	  R.	  &	  Smith,	  L.,	  2000.	  Effects	  of	  sedimentation	  on	  coral	  settlement	  and	  survivorship,	  Proceedings	  9th	  International	  Coral	  Reef	  Symposium,	  Bali,	  Indonesia.	  Bak,	  R.P.M.	  &	  Engel,	  M.S.,	  1979.	  Distribution,	  abundance	  and	  survival	  of	  juvenile	  hermatypic	  corals	  (Scleractinia)	  and	  the	  importance	  of	  life	  history	  strategies	  in	  the	  parent	  coral	  community.	  Marine	  Biology,	  54(4),	  pp.341–352.	  Bak,	  R.P.M.	  &	  Meesters,	  E.H.,	  1998.	  Coral	  population	  structure :	  the	  hidden	  information	  of	  colony	  size-­‐frequency	  distributions.	  Marine	  Ecology	  Progress	  Series,	  162(Sebens	  1987),	  pp.301–306.	  Bak,	  R.P.M.	  &	  Meesters,	  E.H.,	  1999.	  Population	  Structure	  as	  a	  Response	  of	  Coral	  Communities	  to	  Global	  Change.	  Integrative	  and	  Comparative	  Biology,	  39(1),	  pp.56–65.	  Baker,	  A.C.,	  Glynn,	  P.W.	  &	  Riegl,	  B.,	  2008.	  Climate	  change	  and	  coral	  reef	  bleaching:	  An	  ecological	  assessment	  of	  long-­‐term	  impacts,	  recovery	  trends	  and	  future	  outlook.	  Estuarine,	  Coastal	  and	  Shelf	  Science,	  80(4),	  pp.435–471.	  	   40	  Begin,	  C.,	  2012.	  Land	  Use	  and	  Sedimentation	  Impacts	  on	  Coral	  Reefs	  in	  the	  Eastern	  Caribbean.	  Simon	  Fraser	  University.	  Bellwood,	  D.R.	  et	  al.,	  2004.	  Confronting	  the	  coral	  reef	  crisis.	  Nature,	  429(6994),	  pp.827–33.	  Birrell,	  C.L.,	  McCook,	  L.J.	  &	  Willis,	  B.L.,	  2005.	  Effects	  of	  algal	  turfs	  and	  sediment	  on	  coral	  settlement.	  Marine	  pollution	  bulletin,	  51(1-­‐4),	  pp.408–14.	  Borcard,	  D.,	  Gillet,	  F.	  &	  Legendre,	  P.,	  2011.	  Numerical	  Ecology	  with	  R,	  London:	  Springer	  Science+Business	  Media.	  Brandt,	  M.E.	  &	  McManus,	  J.W.,	  2009.	  Disease	  incidence	  is	  related	  to	  bleaching	  extent	  in	  reef-­‐building	  corals.	  Ecology,	  90(10),	  pp.2859–67.	  Burke,	  L.,	  Spandling,	  M.	  &	  Perry,	  A.,	  2011.	  Reefs	  at	  Risk:	  Revisited	  W.	  R.	  Institute,	  ed.,	  Washington	  DC,	  USA.	  Burt,	  J.,	  Bartholomew,	  a.	  &	  Usseglio,	  P.,	  2008.	  Recovery	  of	  corals	  a	  decade	  after	  a	  bleaching	  event	  in	  Dubai,	  United	  Arab	  Emirates.	  Marine	  Biology,	  154(1),	  pp.27–36.	  Chiappone,	  M.	  &	  Sullivan,	  K.M.,	  1996.	  Distribution,	  abundance	  and	  species	  composition	  of	  juvenile	  scleractinian	  corals	  in	  the	  Florida	  Reef	  Tract.	  Bulletin	  of	  Marine	  Science,	  58(2),	  pp.555–569.	  Connell,	  J.H.,	  1997.	  Disturbance	  and	  recovery	  of	  coral	  assemblages.	  Coral	  Reefs,	  16,	  pp.S101–S113.	  Crabbe,	  M.J.C.,	  2009.	  Scleractinian	  coral	  population	  size	  structures	  and	  growth	  rates	  indicate	  coral	  resilience	  on	  the	  fringing	  reefs	  of	  North	  Jamaica.	  Marine	  environmental	  research,	  67(4-­‐5),	  pp.189–98.	  Done,	  T.J.	  et	  al.,	  2010.	  Coral	  growth	  on	  three	  reefs:	  development	  of	  recovery	  benchmarks	  using	  a	  space	  for	  time	  approach.	  Coral	  Reefs,	  29(4),	  pp.815–833.	  Donner,	  S.D.,	  Knutson,	  T.R.	  &	  Oppenheimer,	  M.,	  2007.	  Model-­‐based	  assessment	  of	  the	  role	  of	  human-­‐induced	  climate	  change	  in	  the	  2005	  Caribbean	  coral	  bleaching	  event.	  PNAS,	  104(13),	  pp.5483–5488.	  Eakin,	  C.M.	  et	  al.,	  2010.	  Caribbean	  corals	  in	  crisis:	  record	  thermal	  stress,	  bleaching,	  and	  mortality	  in	  2005.	  PloS	  one,	  5(11),	  p.e13969.	  Eakin,	  M.	  et	  al.,	  2010.	  Caribbean	  corals	  in	  crisis:	  record	  thermal	  stress,	  bleaching,	  and	  mortality	  in	  2005.	  PloS	  one,	  5(11),	  p.e13969.	  	   41	  Edmunds,	  P.J.	  et	  al.,	  1998.	  Photographic	  Versus	  Visual	  Census	  Techniques	  for	  the	  Quantification	  of	  Juvenile	  Corals.	  Bulletin	  of	  Marine	  Science,	  62(3),	  pp.937–946.	  Edmunds,	  P.J.	  &	  Elahi,	  R.,	  2007.	  The	  demographics	  of	  a	  15-­‐year	  decline	  in	  cover	  of	  the	  caribbean	  reef	  coral	  Montastraea	  annularis.	  Ecological	  Monographs,	  77(1),	  pp.3–18.	  Fabricius,	  K.E.,	  2005.	  Effects	  of	  terrestrial	  runoff	  on	  the	  ecology	  of	  corals	  and	  coral	  reefs:	  review	  and	  synthesis.	  Marine	  pollution	  bulletin,	  50(2),	  pp.125–46.	  Fabricius,	  K.E.	  et	  al.,	  2003.	  Effects	  of	  transparent	  exopolymer	  particles	  and	  muddy	  terrigenous	  sediments	  on	  the	  survival	  of	  hard	  coral	  recruits.	  Estuarine,	  Coastal	  and	  Shelf	  Science,	  57(4),	  pp.613–621.	  Gardner,	  T.	  a	  et	  al.,	  2003.	  Long-­‐term	  region-­‐wide	  declines	  in	  Caribbean	  corals.	  Science	  (New	  York,	  N.Y.),	  301(5635),	  pp.958–60.	  Gilmour,	  J.P.	  et	  al.,	  2013.	  Recovery	  of	  an	  isolated	  coral	  reef	  system	  following	  severe	  disturbance.	  Science	  (New	  York,	  N.Y.),	  340(6128),	  pp.69–71.	  Gochfeld,	  D.J.,	  Olson,	  J.B.	  &	  Slattery,	  M.,	  2006.	  Colony	  versus	  population	  variation	  in	  susceptibility	  and	  resistance	  to	  dark	  spot	  syndrome	  in	  the	  Caribbean	  coral	  Siderastrea	  siderea.	  Diseases	  of	  aquatic	  organisms,	  69(1),	  pp.53–65.	  Graham,	  J.E.	  &	  van	  Woesik,	  R.,	  2013.	  The	  effects	  of	  partial	  mortality	  on	  the	  fecundity	  of	  three	  common	  Caribbean	  corals.	  Marine	  Biology,	  160(10),	  pp.2561–2565.	  Harding,	  S.	  et	  al.,	  2008.	  Continued	  degradation	  of	  Tobago	  ’	  s	  coral	  reefs	  linked	  to	  the	  prevalence	  of	  coral	  disease	  following	  the	  2005	  mass	  coral	  bleaching	  event	  .,	  Ft.	  Lauderdale,	  Florida.	  Hernandez,	  R.	  et	  al.,	  2009.	  Spatial	  and	  temporal	  patterns	  in	  reef	  sediment	  accumulation	  and	  composition	  ,	  southwestern	  insular	  shelf	  of	  Puerto	  Rico.	  Caribbean	  Journal	  of	  Science,	  45(2),	  pp.138–150.	  Hoegh-­‐Guldberg,	  O.	  et	  al.,	  2007.	  Coral	  reefs	  under	  rapid	  climate	  change	  and	  ocean	  acidification.	  Science	  (New	  York,	  N.Y.),	  318(5857),	  pp.1737–42.	  Van	  Hooidonk,	  R.J.	  et	  al.,	  2012.	  Coral	  bleaching	  at	  Little	  Cayman,	  Cayman	  Islands	  2009.	  Estuarine,	  Coastal	  and	  Shelf	  Science,	  106,	  pp.80–84.	  Hughes,	  T.P.,	  1994.	  Catastrophes,	  PhaseShifts,	  and	  Large-­‐Scale	  Degradation	  of	  a	  Caribbean	  Coral	  Reef.	  Science,	  265(5178),	  pp.1547–1551.	  Hughes,	  T.P.	  et	  al.,	  2003.	  Climate	  change,	  human	  impacts,	  and	  the	  resilience	  of	  coral	  reefs.	  Science	  (New	  York,	  N.Y.),	  301(5635),	  pp.929–33.	  	   42	  Hughes,	  T.P.,	  1984.	  Population	  Dynamics	  Based	  on	  Individual	  Size	  Rather	  than	  Age :	  A	  General	  Model	  with	  a	  Reef	  Coral	  Example.	  The	  American	  Naturalist,	  123(6),	  pp.778–795.	  Hughes,	  T.P.	  &	  Jackson,	  J.B.C.,	  1985.	  Population	  Dynamics	  and	  Life	  Histories	  of	  Foliaceous	  Corals.	  Ecological	  Society	  of	  America,	  55(2),	  pp.141–166.	  Hughes,	  T.P.	  &	  Tanner,	  J.E.,	  2000.	  RECRUITMENT	  FAILURE	  ,	  LIFE	  HISTORIES	  ,	  AND	  LONG-­‐TERM	  DECLINE	  OF	  CARIBBEAN	  CORALS.	  Ecology,	  81(8),	  pp.2250–2263.	  Huitric,	  M.	  &	  Mcfield,	  M.,	  2000.	  Effects	  of	  multiple	  disturbances	  on	  hard	  coral	  recruits	  in	  Glovers	  Reef	  Atoll	  ’	  s	  lagoon	  ,	  Belize.	  ,	  pp.1–20.	  Irizarry-­‐soto,	  E.	  &	  Weil,	  E.,	  2009.	  Spatial	  and	  temporal	  variability	  in	  juvenile	  coral	  densities	  ,	  survivorship	  and	  recruitment	  in	  La	  Parguera	  ,	  southwestern	  Puerto	  Rico.	  Caribbean	  Journal	  of	  Science,	  45(2),	  pp.269–281.	  Jackson,	  J.	  et	  al.,	  2014.	  Status	  and	  Trends	  of	  Caribbean	  Coral	  Reefs :	  1970-­‐2012,	  Gland,	  Switzerland.	  Kohler,	  K.E.	  &	  Gill,	  S.M.,	  2006.	  Coral	  Point	  Count	  with	  Excel	  extensions	  (CPCe):	  A	  Visual	  Basic	  program	  for	  the	  determination	  of	  coral	  and	  substrate	  coverage	  using	  random	  point	  count	  methodology.	  Computers	  &	  Geosciences,	  32(9),	  pp.1259–1269.	  Lapointe,	  B.E.	  et	  al.,	  2010.	  Land-­‐based	  nutrient	  enrichment	  of	  the	  Buccoo	  Reef	  Complex	  and	  fringing	  coral	  reefs	  of	  Tobago,	  West	  Indies.	  Marine	  pollution	  bulletin,	  60(3),	  pp.334–43.	  Laydoo,	  R.,	  1991.	  A	  guide	  to	  the	  Coral	  Reef	  of	  Tobago	  First	  Edit.,	  Port	  of	  Spain,	  Trinidad	  and	  Tobago:	  Institute	  of	  Marine	  Affairs.	  De	  Lins	  Barros,	  M.M.	  &	  De	  Oliveira	  Pires,	  D.,	  2006.	  Colony	  size-­‐frequency	  distributions	  among	  different	  Populations	  of	  the	  scleractinan	  coral	  Siderastrea	  stellata	  in	  Southwestern	  Atlantic:	  implications	  for	  life	  history	  patterns.	  Brazilian	  Journal	  of	  Oceanography,	  54(4),	  pp.213–223.	  Mallela,	  J.	  &	  Crabbe,	  M.J.C.,	  2009.	  Hurricanes	  and	  coral	  bleaching	  linked	  to	  changes	  in	  coral	  recruitment	  in	  Tobago.	  Marine	  environmental	  research,	  68(4),	  pp.158–62.	  Mallela,	  J.,	  Parkinson,	  R.	  &	  Day,	  O.,	  2010.	  An	  assessment	  of	  coral	  reefs	  in	  Tobago.	  Caribbean	  Journal	  of	  Science,	  46(1),	  pp.83–87.	  Manfrino,	  C.	  et	  al.,	  2013.	  A	  positive	  trajectory	  for	  corals	  at	  little	  cayman	  island.	  PloS	  one,	  8(10),	  p.e75432.	  	   43	  McClanahan,	  T.R.,	  2000.	  Bleaching	  Damage	  and	  Recovery	  Potential	  of	  Maldivian	  Coral	  Reefs.	  Marine	  Pollution	  Bulletin,	  40(7),	  pp.587–597.	  McClanahan,	  T.R.,	  Ateweberhan,	  M.	  &	  Omukoto,	  J.,	  2008.	  Long-­‐term	  changes	  in	  coral	  colony	  size	  distributions	  on	  Kenyan	  reefs	  under	  different	  management	  regimes	  and	  across	  the	  1998	  bleaching	  event.	  Marine	  Biology,	  153(5),	  pp.755–768.	  Meesters,	  E.H.	  et	  al.,	  2001.	  Colony	  size-­‐frequency	  distributions	  of	  scleractinian	  coral	  populations :	  spatial	  and	  interspecific	  variation.	  Marine	  Ecology	  Progress	  Series,	  209,	  pp.43–54.	  Miller,	  J.	  et	  al.,	  2006.	  Coral	  bleaching	  and	  disease	  combine	  to	  cause	  extensive	  mortality	  on	  reefs	  in	  US	  Virgin	  Islands.	  Coral	  Reefs,	  25(3),	  pp.418–418.	  Miller,	  M.W.,	  Weil,	  E.	  &	  Szmant,	  a.	  M.,	  2000.	  Coral	  recruitment	  and	  juvenile	  mortality	  as	  structuring	  factors	  for	  reef	  benthic	  communities	  in	  Biscayne	  National	  Park,	  USA.	  Coral	  Reefs,	  19(2),	  pp.115–123.	  Moulding,	  A.L.,	  2005.	  Coral	  recruitment	  patterns	  in	  the	  Florida	  Keys.	  Revista	  de	  biología	  tropical,	  53	  Suppl	  1(May),	  pp.75–82.	  Mumby,	  P.,	  1999.	  Bleaching	  and	  hurricane	  disturbances	  to	  populations	  of	  coral	  recruits	  in	  Belize.	  Marine	  Ecology	  Progress	  Series,	  190,	  pp.27–35.	  Norström,	  A.	  et	  al.,	  2009.	  Alternative	  states	  on	  coral	  reefs:	  beyond	  coral–macroalgal	  phase	  shifts.	  Marine	  Ecology	  Progress	  Series,	  376(Hatcher	  1984),	  pp.295–306.	  Obura,	  D.,	  2005.	  Resilience	  and	  climate	  change:	  lessons	  from	  coral	  reefs	  and	  bleaching	  in	  the	  Western	  Indian	  Ocean.	  Estuarine,	  Coastal	  and	  Shelf	  Science,	  63(3),	  pp.353–372.	  Oxenford,	  H.	  a.	  et	  al.,	  2007.	  Quantitative	  observations	  of	  a	  major	  coral	  bleaching	  event	  in	  Barbados,	  Southeastern	  Caribbean.	  Climatic	  Change,	  87(3-­‐4),	  pp.435–449.	  Parkinson,	  R.,	  2010.	  Coral	  Reef	  Monitoring	  in	  Tobago:	  Findings	  from	  January	  2007	  to	  February	  2010,	  Port	  of	  Spain,	  Trinidad	  and	  Tobago.	  Pearson,	  R.,	  1981.	  Recovery	  and	  Recolonization	  of	  Coral	  Reefs.	  Marine	  Ecology	  Progress	  Series,	  4,	  pp.105–122.	  Ritson-­‐williams,	  R.	  et	  al.,	  2009.	  New	  Perspectives	  on	  Ecological	  Mechanisms	  Affecting	  Coral	  Recruitment	  on	  Reefs.	  Smithson	  Contrib	  Mar	  Sci,	  38(April),	  pp.437–457.	  Shenkar,	  N.,	  Fine,	  M.	  &	  Loya,	  Y.,	  2005.	  Size	  matters :	  bleaching	  dynamics	  of	  the	  coral	  Oculina	  patagonica.	  Marine	  Ecology	  Progress	  Series,	  294(2001),	  pp.181–188.	  	   44	  Smith,	  L.D.	  et	  al.,	  2005.	  A	  demographic	  approach	  to	  monitoring	  the	  health	  of	  coral	  reefs.	  Marine	  pollution	  bulletin,	  51(1-­‐4),	  pp.399–407.	  Szmant,	  A.M.,	  1986.	  Coral	  Reefs	  Reproductive	  ecology	  of	  Caribbean	  reef	  corals.	  Coral	  Reefs,	  5,	  pp.43–53.	  Szmant,	  A.M.,	  1988.	  Sexual	  reproduction	  by	  the	  Caribbean	  reef	  corals	  Mon	  tastrea	  ann	  ularis	  and	  M	  .	  ca	  vernosa.	  Szmant,	  A.M.	  &	  Gassman,	  N.J.,	  1990.	  The	  effects	  of	  prolonged	  “bleaching”	  on	  the	  tissue	  biomass	  and	  reproduction	  of	  the	  reef	  coral	  Montastrea	  annularis.	  Coral	  Reefs,	  8,	  pp.217–224.	  Tamelander,	  J.,	  2002.	  Recruitment	  Following	  Coral	  Event	  Mortality	  a	  Mass.	  Royal	  Swedish	  Academy	  of	  Science,	  31(7),	  pp.551–557.	  Torres,	  J.L.	  &	  Morelock,	  J.,	  2002.	  Effect	  of	  Terrigenous	  Sediment	  Influx	  on	  Coral	  Cover	  and	  Linear	  Extension	  Rates	  of	  Three	  Caribbean	  Massive	  Coral	  Species.	  Caribbean	  Journal	  of	  Science,	  38(3),	  pp.222–229.	  Vermeij,	  M.J.	  a.	  et	  al.,	  2011.	  Juvenile	  Coral	  Abundance	  Has	  Decreased	  by	  More	  Than	  50%	  in	  Only	  Three	  Decades	  on	  a	  Small	  Caribbean	  Island.	  Diversity,	  3(4),	  pp.296–307.	  Vermeij,	  M.J.	  a.	  &	  Bak,	  R.P.M.,	  2000.	  Inferring	  demographic	  processes	  from	  population	  size	  structure	  in	  corals.	  In	  9th	  International	  Coral	  Reef	  Symposium.	  Bali,	  Indonesia,	  pp.	  1–5.	  Ward,	  S.,	  Harrison,	  P.	  &	  Hoegh-­‐Guldberg,	  O.,	  2000a.	  Coral	  bleaching	  reduces	  reproduction	  of	  scleractinian	  corals	  and	  increases	  susceptibility	  to	  future	  stress	  .	  In	  Proceedings	  9th	  International	  Coral	  Reef	  Symposium,	  Bali,	  Indonesia.	  pp.	  23–27.	  Ward,	  S.,	  Harrison,	  P.	  &	  Hoegh-­‐Guldberg,	  O.,	  2000b.	  Coral	  bleaching	  reduces	  reproduction	  of	  scleractinian	  corals	  and	  increases	  susceptibility	  to	  future	  stress	  .	  In	  Proceedings	  9th	  International	  Coral	  Reef	  Symposium,	  Bali,	  Indonesia.	  pp.	  23–27.	  Weber,	  M.,	  Lott,	  C.	  &	  Fabricius,	  K.E.,	  2006.	  Sedimentation	  stress	  in	  a	  scleractinian	  coral	  exposed	  to	  terrestrial	  and	  marine	  sediments	  with	  contrasting	  physical,	  organic	  and	  geochemical	  properties.	  Journal	  of	  Experimental	  Marine	  Biology	  and	  Ecology,	  336(1),	  pp.18–32.	  Weil,	  E.,	  2001.	  Caribbean	  Coral	  Reef	  Diseases	  ,	  Status	  and	  Research	  Needs.	  In	  Workshop:	  Priorities	  for	  Caribbean	  Coral	  Reef	  Research.	  Miami,	  pp.	  1–9.	  Wilkinson,	  C.	  &	  Souter,	  D.,	  2008.	  Status	  of	  Caribbean	  coral	  reefs	  after	  bleaching	  and	  hurricanes	  in	  2005	  C.	  Wilkinson	  &	  D.	  Souter,	  eds.,	  Townsville:	  Global	  Coral	  Reef	  Monitoring	  Network,	  and	  Reef	  and	  Rainforest	  Research	  Centre.	  	   45	  Wittenberg,	  M.	  &	  Hunte,	  W.,	  1992.	  Effects	  of	  eutrophication	  and	  sedimentation	  on	  juvenile	  corals.	  Marine	  Biology,	  138,	  pp.131–138.	   	   46	  Appendices	  Table	  8.	  Summary	  of	  the	  colony	  abundance	  and	  size	  data	  collected	  for	  each	  dominant	  species	  present	  at	  each	  reef	  sites,	  which	  includes:	  number	  of	  colonies	  (n),	  percent	  cover	  (%),	  mean	  size	  (mean),	  standard	  deviation	  (sd),	  and	  skewness	  (sk).	  If	  skewness	  was	  significantly	  positive	  skewness	  values	  are	  underlined.	  	  	  M.	  faveolata	   	   	   	   	   	   	   	   S.	  siderea	   	   	   	   	  site	   year	   n	   %cover	   mean	   sd	   sk	   	   site	   year	   n	   %cover	   	  size	   sd	   sk	  AR	   2013	   83	   4.8	   27.3	   22.7	   0.7	   	   AR	   2013	   29	   0.7	   16.8	   13.2	   1.5	  BJH	   2010	   51	   6.6	   43.4	   43.6	   1.6	   	   BJH	   2010	   70	   4.4	   28.6	   23	   1.8	  BJH	   2011	   98	   4	   22.5	   23.3	   3.2	   	   BJH	   2011	   52	   2.7	   25.8	   21.3	   1.9	  BJH	   2013	   99	   3.4	   20.4	   26.1	   4.8	   	   BJH	   2013	   82	   2.2	   17.9	   16.2	   2.9	  CE	   2010	   29	   2.8	   36.7	   42.6	   2.4	   	   CE	   2010	   22	   1.3	   27.8	   18.9	   1.1	  CE	   2011	   13	   0.6	   24.4	   30.8	   2.5	   	   CE	   2011	   12	   0.2	   12	   6.1	   0.4	  CE	   2013	   52	   4.2	   33.2	   50.5	   5.5	   	   CE	   2013	   18	   0.6	   20.5	   13.5	   0.5	  CW	   2010	   24	   1.3	   26.9	   25.5	   2.4	   	   CW	   2010	   11	   0.4	   21.7	   24.7	   2.5	  CW	   2011	   6	   0.6	   38.6	   17.3	   1.5	   	   CW	   2011	   14	   0.9	   28.2	   17.9	   1.1	  CW	   2013	   48	   9.5	   56.2	   65.6	   3.3	   	   CW	   2013	   16	   0.8	   24.9	   13	   0.3	  OB	   2010	   46	   4.4	   36.7	   34.1	   2.6	   	   OB	   2010	   35	   3.5	   37.4	   27	   2.1	  OB	   2011	   102	   7.9	   32.3	   46.8	   4.2	   	   OB	   2011	   21	   1.3	   28.2	   25.3	   2.6	  OB	   2013	   121	   12.1	   37.4	   41.1	   2.5	   	   OB	   2013	   49	   2.5	   25.4	   20.1	   1.6	  WB	   2010	   105	   10.5	   37.4	   28.8	   1.8	   	   WB	   2010	   40	   4.4	   39.5	   25.2	   1.1	  WB	   2011	   59	   4.2	   30.7	   24.8	   2.1	   	   WB	   2011	   36	   2	   27.1	   21	   1.2	  WB	   2013	   59	   10.7	   53.1	   56.7	   2	   	   WB	   2013	   23	   0.6	   17.9	   12.6	   1.5	  C.	  natans	   	   	   	   	   	   	   D.	  labyrinthiformis	   	  site	   year	   n	   %cover	   mean	   sd	   sk	   	   site	   year	   n	   %cover	   mean	   sd	   sk	  OB	   2010	   14	   3.1	   60.5	   49.3	   1.5	   	   OB	   2010	   21	   1.2	   26.6	   20.5	   1.8	  OB	   2011	   17	   1.8	   38.6	   26.3	   1.1	   	   OB	   2011	   9	   0.6	   30.7	   16.2	   0.4	  OB	   2013	   22	   2.5	   40.1	   41.6	   3.5	   	   OB	   2013	   6	   0.3	   26.7	   15.5	   0.7	  WB	   2010	   34	   7.5	   59.8	   61.3	   1.7	   	   WB	   2010	   13	   0.7	   25.3	   20.7	   2.4	  WB	   2011	   16	   2.3	   46.5	   43.5	   0.9	   	   WB	   2011	   16	   1	   29.1	   17.6	   0.6	  WB	   2013	   22	   3.6	   50.1	   57.4	   1.6	   	   WB	   2013	   13	   0.4	   18.4	   7.2	   1.7	  Agaricia	  spp.	  	   	  	   	  	   	  	   	  	   	  	   	   D.	  strigosa	   	  	   	  	   	  	   	  	   	  	  site	   year	   n	   %cover	   mean	   sd	   sk	   	   site	   year	   n	   %cover	   mean	  	   sd	   sk	  AR	   2013	   67	   0.5	   9.5	   6.4	   2.2	   	   BJH	   2013	   14	   0.1	   10.3	   15.9	   3.6	  BJH	   2010	   17	   0.1	   6.5	   2.5	   1.4	   	   CE	   2010	   47	   1.6	   20.2	   12.4	   2.6	  BJH	   2013	   1	   0	   4	   NA	   NA	   	   CE	   2011	   29	   0.6	   16.1	   6	   0.8	  CE	   2010	   23	   0.1	   7.7	   4.4	   2.4	   	   CE	   2013	   60	   1.2	   15.1	   22.5	   6.9	  CE	   2011	   19	   0.1	   7.9	   4.1	   1.4	   	   CW	   2010	   17	   0.2	   11.2	   3.8	   0.8	  	   47	  CE	   2013	   31	   0.2	   7.7	   4.1	   3.1	   	   CW	   2011	   33	   0.7	   15.2	   8.8	   0.6	  CW	   2010	   8	   0.1	   9.4	   4.5	   0.8	   	   CW	   2013	   121	   3.4	   18.3	   14	   3	  CW	   2011	   16	   0.1	   9.1	   2.6	   0.1	   	   OB	   2010	   26	   0.8	   18.9	   14.6	   3.2	  CW	   2013	   52	   0.2	   7.1	   4.7	   3.8	   	   OB	   2011	   12	   0.5	   22	   16.5	   1.4	  OB	   2010	   26	   0.3	   11	   9.5	   3.2	   	   OB	   2013	   17	   0.7	   23.1	   18.7	   2.2	  OB	   2011	   28	   0.1	   6.6	   3.3	   2	   	   WB	   2010	   28	   0.6	   16.1	   6.4	   0.6	  OB	   2013	   122	   0.6	   7.6	   6.2	   3.7	   	   WB	   2011	   47	   0.8	   13.8	   8.8	   1.2	  WB	   2010	   22	   0.2	   10.7	   6.6	   0.9	   	   WB	   2013	   39	   1	   17.2	   10.9	   1.4	  WB	   2011	   43	   0.2	   7.3	   3	   1.9	   	   M.	  cavernosa	   	  	   	  	   	  	   	  	   	  	  WB	   2013	   50	   0.2	   6.8	   2.8	   1.3	   	   site	   year	   n	   %cover	   mean	   sd	   sk	  P.	  astreoides	   	  	   	  	   	  	   	  	   	  	   	   BJH	   2010	   25	   1.2	   25	   21.3	   2.4	  sites	   year	   n	   %cover	   mean	   sd	   sk	   	   BJH	   2011	   13	   0.4	   19.3	   18.2	   1.9	  AR	   2013	   32	   0.6	   14.2	   8.7	   2	   	   BJH	   2013	   4	   0.1	   15.8	   5.9	   -­‐0.7	  BJH	   2010	   24	   0.3	   12.3	   4.1	   0.5	   	   CE	   2010	   28	   1.2	   23.6	   16.3	   1.3	  BJH	   2011	   33	   0.4	   12	   5.4	   0.4	   	   CE	   2011	   32	   1.2	   21.8	   11.8	   1.1	  BJH	   2013	   65	   0.5	   9.1	   3.6	   0.8	   	   CE	   2013	   31	   1.7	   26.7	   18.8	   1.3	  CE	   2010	   10	   0.2	   16.2	   3.4	   -­‐2.2	   	   CW	   2010	   27	   1.3	   24.4	   21.2	   2	  M.	  mirabilis	   	  	   	  	   	  	   	  	   	  	   	   CW	   2011	   22	   0.8	   20.7	   12.1	   1.2	  sites	   year	   n	   %cover	   mean	   sd	   sk	   	   CW	   2013	   23	   1.7	   31	   45.2	   3.3	  AR	   2013	   46	   18.9	   88.9	   41.5	   2.7	   	   	   	   	   	   	   	   	  	  Table	  9.	  Significant	  comparison	  of	  size	  frequency	  distributions	  and	  colony	  size	  between	  years	   (2010,	  2011	  and	  2013)	  determined	  using	  Kolmogorov-­‐Smirnov	  test	  (KS)	  and	  Kruskal-­‐Walis	  (KW)	  test	  respectively.	  Species	   Site	   year	   KS	  P-­‐value	   KW	  P-­‐value	  Agaricia	  spp.	  	   Outer	  Buccoo	   2010-­‐2011	   0.00	   0.00	  M.	  faveolata	   Outer	  Buccoo	   2010-­‐2011	   NA	   0.01	  M.	  faveolata	   Black	  Jack	  Hole	   2010-­‐2011	   0.01	   0.00	  S.	  siderea	   Western	  Buccoo	   2010-­‐2011	   NA	   0.00	  S.	  siderea	   Culloden	  East	   2010-­‐2011	   0.01	   0.00	  Agaricia	  spp.	  	   Culloden	  West	   2011-­‐2013	   0.00	   NA	  D.	  strigosa	   Culloden	  East	   2011-­‐2013	   0.01	   0.00	  M.	  faveolata	   Black	  Jack	  Hole	   2011-­‐2013	   0.01	   0.04	  P.	  astreoides	   Black	  Jack	  Hole	   2011-­‐2013	   0.00	   0.01	  S.	  siderea	   Black	  Jack	  Hole	   2011-­‐2013	   0.00	   0.00	  	   48	  Agaricia	  spp.	  	   Outer	  Buccoo	   2010-­‐2013	   0.00	   0.00	  D.	  strigosa	   Culloden	  East	   2010-­‐2013	   0.00	   0.00	  D.	  strigosa	   Culloden	  West	   2010-­‐2013	   0.01	   NA	  M.	  faveolata	   Culloden	  West	   2010-­‐2013	   0.00	   0.00	  M.	  faveolata	   Black	  Jack	  Hole	   2010-­‐2013	   0.00	   0.00	  P.	  astreoides	   Black	  Jack	  Hole	   2010-­‐2013	   0.01	   0.00	  S.	  siderea	   Black	  Jack	  Hole	   2010-­‐2013	   0.00	   0.00	  S.	  siderea	   Outer	  Buccoo	   2010-­‐2013	   0.01	   0.01	  S.	  siderea	   Western	  Buccoo	   2010-­‐2013	   0.00	   0.00	  	  Table	  10.	  Significant	  comparison	  size	  frequency	  distributions	  between	  reef	  sites	  determined	  using	  Kolmogorov-­‐Smirnov	  test	   (KS)	   test	   respectively.	  Abbreviated	  site	  codes:	   	  OB	   is	  Outer	  Buccoo,	  WB	   is	  Western	  Buccoo,	  CE	   is	  Culloden	  East,	  CW	  is	  Culloden	  West,	  and	  BJH	  is	  Black	  Jack	  Species	   Year	   KS	  	  P-­‐value	  >0.0167	  D.	  strigosa	   2010	   CE-­‐CW	  M.	  faveolata	   2011	   OB-­‐BJH	  M.	  faveolata	   2011	   WB-­‐BJH	  S.	  siderea	   2011	   CE-­‐CW	  S.	  siderea	   2011	   WB-­‐BJH	  D.	  strigosa	   2013	   CE-­‐OB	  M.	  faveolata	   2013	   WB-­‐BJH	  M.	  faveolata	   2013	   WB-­‐OB	  M.	  faveolata	   2013	   CW-­‐EC	  	  	  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items