Lindsay Wieland

1
Seed Dispersal Estimating Dispersal Kernels and
Seed Shadows

• Lindsay Wieland
• November 10, 2009

2
Outline

• Seed Dispersal
• Fruits and Frugivores
• Seed Shadow v. Dispersal Kernel
• Estimating Dispersal Kernels
• Conceptual Model
• Elaeocarpus grandis Case Study
• Toucan-generated Dispersal Model
• Spatially Explicit Model

3
Introduction

• In tropical forests, seed dispersal processes are
  dominated by vertebrate dispersers and can
  directly involve individuals belonging to
  hundreds of species
• Dispersal and resultant seed shadows may
  influence key processes, such as colonization,
  population persistence and community structure

4
Dispersal Types

• Gravity (Gravichory)
• Wind (Anemochory)
• Water (Hydrochory)
• Ballistic (Autochory)
• Animal (Zoochory)
• Epizoochory (transported externally)
• Inadvertent (cached)
• Myrmechory (ant-dispersed)
• Endozoochory (through the digestive tract)

5
Fruits

• Nutritive fleshy arils,
• pericarp or pulp
• Chemical attractant
• Colors
• Fruit size varies from 0.01g to 40g
• In Neotropical forests, 50 - 90 of the canopy
  trees bear fruits adapted for animal dispersal,
  while close to 100 of the shrubs and sub-canopy
  trees produce fleshy fruits
• In Paleotropical forests 35 40 canopy trees
  bear fruits and 70-80 of shrubs

6
Frugivores

• High variety of tropical bird, bat, ant, mammal
  dispersers
• Frugivore size ranges from 10g tyrannid
  flycatchers to African Elephant
• Enormous differences in fruit and frugivore
  scales imply an enormous potential range of
  phenomena

7
Seed Shadow v. Dispersal Kernel

• Seed shadow the spatial distribution of seeds
  dispersed from a single plant
• Dispersal kernel frequency distribution of the
  dispersal distance within a crop or an entire
  population

8
Frugivores

• Movement and behavior directly impacts seed
  shadows/kernels
• Frugivores that remain for long periods in
  fruiting tree will drop most seeds beneath parent
  tree, whereas short-term visitors to fruiting
  trees will disperse most seeds at sites away from
  parent tree
• Results in spatial variability in
• seed shadows, which can have
• consequences for seed and
• seedling survival and population
• demographics

9
Outline

• Dispersal types
• Fruits and Frugivores
• Seed Shadow v. Dispersal Kernel
• Estimating Dispersal Kernels
• Conceptual Model
• Elaeocarpus grandis Case Study
• Toucan-generated Dispersal Model
• Spatially Explicit Model

10
Estimating Dispersal Kernels

• Elaeocarpus grandis (Elaeocarpaceae)
• Rainforest canopy tree
• Medium(25.6mm diameter)blue fruits
• 3-5 seeds per fruit
• Highly sculpted, thick, bony endocarp
• Native to Australia

11
Conceptual Model
12
Functional Groups
13
Functional Groups

• Incorporates ecological redundancy by treating
  species that provide similar services as a single
  class
• To define disperser functional groups
• Measure quantity of fruit handled
• Measure quality of fruit handling
• Quantify diversity of species handled

14
Case Study Elaeocarpus grandis

• Functional Groups
• 65 vertebrate dispersers grouped into 15
  functional classes
• Wide-ranging slow-gut
• Wide-ranging rapid-gut
• Wide-ranging large fruit
• Terrestrial within-forest
• Mega-terrestrial
• Predatory Rodents
• Large within-forest
• Small within-forest
• Etc

15
Spatial and Temporal Variation
16
Spatial and Temporal Variation

• Generalization for patterns of seed dispersal is
  not simple and cannot be done in short-term
  studies in particular locations
• Variability in
• Disperser distribution and abundance
• Climatic variation across years and locations
• Spatial variation in biotic and abiotic
  conditions
• Size and timing of fruit crops
• Animal behavior and movement

17
Case Study Elaeocarpus grandis

• Pied currawongs (Strepera graculina) exhibit
  greater displacement rates in fragmented
  landscapes than in continuous forest

Strepera graculina

• Complete sampling under each treatment
  condition
• Kernel estimates for fragments and continuous
  forest separately

Dennis Westcott 2007
18
Relative Contributions to Dispersal
19
Relative Contributions to Dispersal

• Sampling strategy must account for these
  different contributions
• Observation of trees (day and night)
• Measure removal rates of fruits and seeds placed
  on forest floor
• Measure fruit production in canopy and fruit
  fallen to ground
• Which dispersers are relevant to which plants?
• What proportion of crop is removed by different
  vectors?
• What is the manner in which seeds are handled?

20
Case Study Elaeocarpus grandis

• Which dispersers are relevant to which plants?

• Large E. grandis fruits imply large-bodied
  dispersers

21
Case Study Elaeocarpus grandis

• What proportion of crop is removed by different
  vectors?

• High proportion of E. grandis dispersed by volant
  dispersers

22
Case Study Elaeocarpus grandis

• What is the manner in which seeds are handled?

• Some cached seeds, only 2 survival for seeds
  handled by rodents

23
Seed Retention Time
24
Seed Retention Time

• Estimate retention time for swallowing and
  defecation/regurgitation by wild animals in
  captivity

25
Case Study Elaeocarpus grandis

• 10 disperser functional groups (that consume E.
  grandis)
• 17 species of fruit (in same functional group as
  E. grandis)
• 1707 medium-sized few-seeded fruits passed
  through 90 feeding trials
• Range of retention time 1min-28hrs
• Effects of gut passage time on germinability
  varies widely

26
Disperser Movement Patterns
27
Disperser Movement Patterns

• Direct observation or telemetry
• Mark-recapture omits short-term patterns of
  individual movement between captures

Dennis Westcott 2007
28
Case Study Elaeocarpus grandis

• Used continuous radio-telemetry to triangulate
  location at 5 min intervals
• No attempt to approach or sight the animal to
  minimize observer effects

Dennis Westcott 2007
29
Estimating Dispersal Kernel for a Disperser
30
Case Study Elaeocarpus grandis

• Seed Retention Time Disperser Movement Patterns

Dennis Westcott 2007
31
Estimating Total Dispersal Kernels
32
Total Dispersal Kernels

• Incorporates contribution of all dispersers
  relevant to focal plant individual, species or
  functional type
• Scale the y-axis of each disperser kernel
  relative to other dispersers according to
  proportion of total crop removed

• Case Study Elaeocarpus grandis
• 77 of seeds dispersed lt100m
• 3 dispersed beyond 400m

33
Case Study Elaeocarpus grandis

34
Case Study Elaeocarpus grandis

• Some functional groups dominate close dispersal,
  others dominate long-distance dispersal, or
  contribute across the entire range
• Functional groups differ in contribution of crop
  removal
• Large bodied species and
• abundant species provide
• greatest proportion of dispersal

35
Post-primary Dispersal Processes
36
Post-primary Dispersal Processes

• Secondary dispersal
• Seed predation
• Mortality
• Mortality Kernel relationship between dispersal
  distance and mortality

• Recruitment Kernel Total Dispersal Kernel -
  Mortality Kernel

37
Predicted Recruitment Kernel
38
Outline

• Dispersal types
• Fruits and Frugivores
• Seed Shadow v. Dispersal Kernel
• Estimating Dispersal Kernels
• Conceptual Model
• Elaeocarpus grandis Case Study
• Toucan-generated Dispersal Model
• Spatially Explicit Model

39
Using Toucan-generated Dispersal Models to
Estimate Seed Dispersal in Amazonian Ecuador

• Estimate seed dispersal distances for a
  Neotropical tree, Virola flexuosa
  (Myristicaceae), based solely on toucan movements
  and seed retention times
• Present a spatially explicit model, which
  realistically outlines the dispersion patterns
  generated by toucans

Ramphastos tucanus
40
Study Site and Species

• Yasuní Biosphere Reserve, Ecuador
• Terra firme, lowland rainforest,
• gt3000mm rain per year
• V. flexuosa is a dioecious, shade-intolerant
• species widespread in South America
• Toucans (Ramphastidae) are important seed
  dispersers throughout the Neotropics
• Many-banded Araçari (Pteroglossus pluricinctus)
• White-throated Toucan (Ramphastos tucanus)
• Channel-billed Toucan (R. vitellinus)
• Differ in size, diet, movement patterns, and seed
  dispersal ecology

R. vitellinus
41
(No Transcript)
42
Estimating Seed Shadows Field Methods

• Fruit Removal
• Tree watches during fruiting season (0600
  -1000, 8 reps/tree)
• Crop Size
• Seed traps under focal trees (5 crown area)
• Seed Retention
• Wild toucans- marked V. flexuosa seeds with
  thread, placed in papaya and bird gel, continuous
  observation (0600-1800)
• St. Louis Zoo - V. flexuosa seeds placed in
  papaya and grapes, continuous observation
  (0800-1700)
• Movement Patterns
• Radio-tracking (2001-2005), 15 min intervals
• Several tracking days to follow individuals for
  detailed movement and location data

43
Results Fruit Removal and Seed Retention

• 400 observation hours, 13 individual fruiting
  trees
• Toucans represent 64.3 of visits and remove gt52
  dispersed seeds from V. flexuosa
  (Pteroglossus12.2, Ramphastos 39.8)
• Average seed retention time 30min
• Most seeds (95) were regurgitated
• Pteroglossus Ramphastos

44
Results Movement Patterns

• Radio-tracked individuals 3-25 days
• Distance traveled per movement bout ranged from 0
  to gt2000m
• Time between observations vary (15min-165min due
  to signal loss)
• Average movement distance in 30min
• Pteroglossus 528m
• Ramphastos 348m
• Proportion of movements lt100m
• Pteroglossus16
• Ramphastos 28
• Home range estimates
• Pteroglossus191ha
• Ramphastos 86ha

Pteroglossus
Ramphastos
45
Results Dispersal Models

• Pteroglossus 84 of seeds
• Ramphastos 72 of seeds
• Both dispersal kernels have long tails
• Pteroglossus Ramphastos

gt100m from parent
46
Using Toucan-generated Dispersal Models to
Estimate Seed Dispersal in Amazonian Ecuador

• Estimate seed dispersal distances for a
  Neotropical tree, Virola flexuosa
  (Myristicaceae), based solely on toucan movements
  and seed retention times
• Present a spatially explicit model, which
  realistically outlines the dispersion patterns
  generated by toucans

Ramphastos tucanus
47
Estimating Seed Shadows Probability of Seed
Deposition

• Following Murray (1988) and Holbrook Smith
  (2000), estimated toucan-generated dispersal
  kernel using seed retention times and movement
  data
• pd (adt bt) where
• Pprobability of a seed being deposited at a
  particular distance category (d) from the parent
  tree,
• a probability of a bird being within a
  particular distance category (d) in time interval
  (t)
• b probability of a seed being passed in that
  time interval (t)

48
Estimating Seed Shadows Spatially Explicit
Models

• Probabilities of seed deposition (P) for each
  toucan species were combined with V. flexuosa
  fruit removal data to more realistically estimate
  seed shadow
• Nx (pxm rm) where
• N number of seeds predicted to fall at a
  particular location (x)
• p probability of seed deposition at varying
  distances (x) from each female tree (m)
• r number of fruit removed at each female tree

49
Results Dispersal Models

• Significant differences between distributions are
  result of movement patterns
• (Cramér-von Mises ?2.461, P0.001)
• Higher density seed-fall in South- East where
  more fecund trees located
• V. flexuosa produce fruit every 2-3 years,
  therefore spatial depiction is restricted to the
  time period of this study (2001-2005)

Pteroglossus
Ramphastos
50
Effectiveness of Toucan Dispersal

• Very effective dispersers at Tiputini and may
  help decrease density-dependent seed/seedling
  mortality by transporting seeds away from parent
  plant
• What is an effective dispersal distance?
• Howe et al. (1985) found 44-fold increase in seed
  survival when moved gt45m from parent Virola tree
• Holbrook Loiselle (2007) found greater numbers
  of larger seedlings beyond 40m suggesting
  differential seedling survival
• Long distance dispersal
• Toucans can fly up to 3665m (Pteroglossus maximum
  distance)
• Facilitate gene flow, colonization of new sites,
  and forest regeneration
• Reduces kin competition

51
Shape and Scale of Dispersal

• Patterns of behavior (movement, seed retention,
  foraging behavior) and plant parameters (crop
  size) can significantly impact the shape and
  scale of dispersal kernels and patchy nature of
  seed shadows
• Seed dispersal studies require integrating
  processes across a wide range of scales
• 84-ha may not encompass movement patterns of all
  Virola dispersers
• Does not cover entire home range of Pteroglossus
• Does not incorporate seed rain from trees outside
  the 84ha area

52
Conclusions

• Many factors interact to determine the density
  and dispersion patterns of plant populations.
  Seed dispersal is only the first step in the
  process.
• Due to great diversity of animal dispersers in
  tropical forests, understanding and predicting
  seed dispersal patterns can be difficult
• Dispersal and resultant seed shadows may
  influence key processes, such as colonization,
  population persistence and community structure

53
Works Cited

• Clark, C.J. , J.R. Poulsen and V.T.Parker. 2001.
  The role of arboreal seed dispersal groups on the
  seed rain of a lowland tropical forest.
  Biotropica 33 606-620.
• Dennis, A. J., and D. A.Westcott. 2007.
  Estimating Dispersal Kernels Produced by a
  Diverse Community of Vertebrates. Pp. 201-228 in
  A. J. Dennis, E. W. Schupp, R. Green, and D. W.
  Westcott, editors, Seed Dispersal Theory and its
  Applications in a Changing World. CABI
  Publishing, Wallingford, Oxfordshire, UK. 
• Fleming, T.H., R. Breitwisch and G.H. Whitesides.
  1987. Patterns of Tropical Vertebrate Frugivore
  Diversity. Ann. Rev. Ecol. Syst. 18 91-109.
• Holbrook, K. M., and B. A. Loiselle. 2007. Using
  toucan-generated seed shadows to estimate seed
  dispersal in Amazonia Ecuador. Pp. 300-321 in A.
  J. Dennis, E. W. Schupp, R. Green, and D. W.
  Westcott, editors, Seed Dispersal Theory and its
  Applications in a Changing World. CABI
  Publishing, Wallingford, Oxfordshire, UK.
• Howe, H.F. and J. Smallwood. 1982. Ecology of
  seed dispersal. Ann. Rev. Ecol. Syst. 1 3
  201-28.
• Nathan, R. and H. C. Muller-Landau. 2000.
  Spatial patterns of seed dispersal, their
  determinants and consequences for recruitment.
  Trends in Ecology and Evolution 15278-285.