Forest Floor Decomposition Following Hurricane Litter Inputs in Several Puerto Rican Forests

1
Forest Floor Decomposition Following Hurricane
Litter Inputs in Several Puerto Rican Forests

• Rebecca Ostertag
• Frederick N. Scatena
• Whendee L. Silver

2
Hurricanes as Disturbance Factors in Forests

• Redistributes mass and nutrients in a forest
• Damage mostly to foliage, causes massive amounts
  of litter deposition
• Amount of damage and deposition depends on forest
  type

3
Purpose of Study

• Examine litterfall production and disapearance
  from hurricanes and associated pools and dynamics
  of nutrients added through decomposition of
  litter inputs
• Have pre-hurricane data for various ecosystem
  processes (litterfall rates, climate, etc.), so
  can compare to post-hurricane data later
  collected
• Plots span elevation, climate, and species
  composition gradients

4
What They Expected

• 3 main hypotheses
• 1. Hurricane litter inputs best predicted from
  pre-hurricane litterfall rates
• Ex. Sites with high litterfall production will
  have high litterfall inputs, probably due to
  environmental conditions
• 2. Hurricane litter decomposes rapidly and
  effects are only felt for a few months (nutrient
  addition is brief and not sustained)
• 3. High-elevation sites will experience less
  damage, less litterfall inputs but slower
  decomposition rates

5
Study Area

• Luquillo Experimental Forest (LTER)
• Range of conditions and forest types
• Elevation range 300-1000 m a.s.l.
• Rainfall range 2500-4735 mm/yr
• 6 Mixed species to Palm forests sampled
• Moist, Tabonuco, Palm, Colorado, Short Cloud, and
  Tall Cloud
• Area hit by Hurricane Georges in 1998

6
Hurricane Georges
7
Methods

• All sites had preexisting plots and data on
  litterfall and forest floor processes
• Sampling was done 2-3 months before hurricane hit
• Sampling was then done within 2 weeks following
  hurricane landfall using preexisting plots and
  equipment
• Forest floor and litterfall then sampled at
  2,3,4,6,8,10,..52 weeks after hurricane
• Soils sampled every 3 months
• Stand level tree damage quantified using 10 m
  radius plots
• Statistical analyses done on all data (ANOVA,
  decay rates, decay models, etc.)

8
On to the Results
9
Results Initial Inputs of Forest Floor Mass and
Nutrients

• Hurricane Georges litterfall ranged from 0.55 to
  0.93 times annual litterfall, forest floor mass
  increased 1.1 to 2.5 times above perhurricane
  values
• Moist forest had the largest storm litterfall
  inputs and the greatest increase in forest floor
  mass and also higher nutrient content due to
  large mass inputs

10
Types of Litter inputs to the forest floor

• Out of five sites, approximately 50-60 of the
  forest floor mass consisted of leaves
• 30-40 consisted of wood

11
Results Tree Damage

• Defoliation was the most common form of damage,
  accounting for 77 to 86 of all tree damage
• Uprooted trees were common in both moist (5.9)
  and palm forests (5.2)
• Moist Forest had the most structural damage with
  both defoliation and uprooted trees

12
(No Transcript)
13
Results Changes in Forest Floor Mass and
Nutrients over time

• Forest floor mass returned to prehurricane values
  rapidly
• 5-10 months for moist forest
• 4-5 months for tabonuco forest
• 1-2 months for palm forest
• Nutrient Concentrations of the forest floor
  increased and decreased depending on the site and
  nutrient.
• Nitrogen, P, Ca, and Mg concentrations increased
  above initial storm input levels in leaves of
  Moist forest
• Potassium quickly decreased from leaves at all
  Moist, Tabonuco, and Palm forest due to leaching
  losses.

14
Results
15
Discussion

• Moist forest had the greatest proportional
  increase of litterfall, but recovered more
  rapidly.
• Upper-elevation palm forest had smaller increase
  in litterfall, but recovered at about the same
  rate due to a slower decomposition rate.
• They also had higher percentage of leaves and
  fruit which decay faster than wood.

16
Long-term Impacts

• Extent of damage depends on the severity of the
  storm and amount of time since the last storm.
• Extent of recovery depends on the characteristics
  of the ecosystem
• Individuals within the system
• e.g., palms and ferns snap and uproot
  infrequently
• Productivity of the system

17
Decomposition rates not accelerated

• No difference in soil NO3-, NH4, or PO43
  concentrations were evident after 1 ½ 2 years
  after Hurricane Opal.
• Only studies done within first four months of
  hurricanes have shown any increase in soil
  nutrients.
• From this, they determined that if, in fact,
  hurricanes do affect ecosystem processes such as
  nutrient cycling and seedling regeneration, it is
  only within a brief period immediately following
  the event.
• Not surprising because Georges only caused 5-10
  increase in total C in the top 10 cm.

18
Increased litterfall can be an important source
of nutrients over short-term time scales.

• Most (gt 95) of the nutrients from
  hurricane-derived litter are taken up by plants.
• Rapid refoliation (within 3-4 weeks)
• Vigorous understory growth observed in grasses,
  seedlings, and vines.
• desirable microsites.

19
Discussion

• Finally, the take home message
• The long-term impact of a disturbance on
  ecosystem structure and process results from the
  interactions of the initial damage and the
  recovery and repair mechanisms that operate
  during the subsequent phase or reorganization.
• In other words, long-term impact equals how much
  damage is done and how well the system can
  recover from it.

20
(No Transcript)