Develop a way to characterize and
monitor the well-being of one of the last true frontiers on Earth – the
Amazon Basin rainforest – and devise a set of practical strategies to ensure
its preservation.
Our Goal and Future Completed Mission
is:
Our team mission is to gain an understanding of mathematical
modeling of biological systems as well as an understanding of actual
ecological processes occuring in the Reainforest. We will apply
this knowledge, along with useful information from other groups,
to create two models of specific examples of interactions in the Rainforest.
We will use these models to make recommendations on some vital parameters
necessary to maintain ecological equilibrium.
Weekly work:
September 23 -30
:
I have been reading ecology books to get a general
sense of how ecological systems work. I have
been searching through Vera and the Hayden Library to further my
knowledge of the topic since our group
goal for the week was to get our feet wet in the material so we can
predetermine which system we should consider doing a model of.
Basically I am learning more about
the problem to get plausible and more complicated ways to solve it
.
I contact our mentors and we are going to
meet Marc Mander tentatively on Wendnesday.
Our other mentor Michael Rucker is unavailable
to meet us on Wednesday since he is working near the Hoover Dam at the
moment.
October 1-7 :
Read part of the book: "Pilot Analysis of
Global Ecosystems: Forrest Ecosystem" from E. Matthews, R. Payne, M.
Rohweder and S. Murray. Learned that:
Forrestextent is a basic
measure of condition; if global forrest shrinks, provision of goods
and services from ecosystems will be reduced, in absence of human action
Measures of biological
condition of the world's forest are extremely difficult to develop,
given data limitations and controversy over such concepts as ecosystem
health .
Thespread of "transition
zones" (agriculture practiced at the margins of intact forest), road
construction and use of fire are leading indicators of environmental
change.
The mixed forest/ agriculture zone which are
spreading rapidly at the edges of formerly intact forest, are not
often recorded as forest conversion.
Less than 8% of the global forest area is legally
protected. Legal safeguards appear ineffective against logging, poaching
and other forms of development.
Forest near human settlement or transportation
routes have high concentrations of non-native species, which have
been introduced deliberately or accidentally. Most are benign, but some
invasive plants and insects pest have done extensive damage to both production
and amenity forest.
Moderate estimates of future species extinction
rates in tropical forest range from 1 to 5 percent per decade. However,
this measures have high level of uncertainty.
Forest soils and vegetation store about 40 percent
of all carbon in the terrestrial biosphere.
Read part of the book: "Amazonia"
by Pergan Press. Learned that:
Deforestation far exceeds regrowth.
Tropical metane forest are disappearing faster
than any other tropical forest type.
Ground cover vegetation appears more important
than tree cover in preventing erosion, but erosion rates under shifting
cultivation are ten times higher than in natural forest.
Species-Area Curve serve as Predictors of the
Biodiversity Impacts of Forest Lost.
Precipitation rate of 2-3 meters produces dense
network of streams and rivers.
The plant cover of the Amazon Basin in its actual
appearance depends upon the amount of water in the atmosphere and the
amount and distribution of precipitation during the year.
Decided with the group to
focus on the lowland rainforest and tentatively focus our models on
two topics between: water flow, canopy density, energy flow or nutrient
cycle.
October 8 - 14
:
Search for a species profile of the Amazon to determined
which species to use in our developing models. In the mean time I have
find this databases:
Animal Diversity
- database of animals by Museum of Zoology of the University of
Michigan. Offers good in depth articles of any animal that one searches.
Amazon Plant Database
- A complete database of the known plants of the Amazon listed
in various methods; i.e. common name, botanical name, etc.
Tropical
Database of Biodiversity of Brasil
- Brazilian database of tropical species used by the scientific
and tachnologic community within Brazil. Its purpose is to serve as an
information tool and conservation/ awareness resource
Searching and reading information about lowland rainforrest.
Some links that I have found are:
Contact the Fauna group for information
concerning on general information on how Fauna affects Flora.
Search for Interaction
Equations to be used in the Friday meeting where the group
is starting to gear our plan to achieve our two models. Since any system
interaction model has to take into account population growth I have
come across with some population equations that might give a preliminary
building block to our model and then build up from it:
Rate of increase of number of individuals: ((Average
birth rate per capita - Average death rate per capita)
Efficiency of energy Tranmission between species:
E = P / (P + R) ; P = net production, R =
respiratio n
Leslie Model
I will be researching more on the topics
above during the week and in the same time will read ecology books to
learn about interactions and equations to model it
October 22-28
:
Read parts of the book "The Geometry of Ecological
Interactions" by Dieckmann, Law and Metz.Learned about Predator and
Prey models that:
Predator and Prey population models have a very
extensive history but their is a discrepancy between the behavior
of these and that of natural prey and predator populations.
Experimental studies show that equations used in
classical models do indeed give a reasonable qualitative description
of the behavior of predator and prey populations.
Since the models describe the interaction between
Predator and prey in laboratory experiments reasonably well (but
fail to capture the properties of natural populations), and laboratory
and natural prey and predator population are esentially the same, the
spatial scale at which the predator-prey system exist must play a crucial
role in preventing population oscillations in the field.
Spatial scale at which observations of ecological
systems are carried out strongly influences their outcome.
Modeling population dynamics shuld intuitively
start at the level of the individual, at which the organism reproduce,
die, and interact, thus giving rise to population dynamics. At the individual
level, it is appropiate to use a model that describes the whereabouts of
all individual interactions between them. However, this is a very difficult
task.
Descriptions of the dynamics of the global population
can be cast in terms of couple local populations when one abstracts individual-based
descriptions of dynamics in terms of variables that are measurable
at the scale of local populations.
Individual-based Predator-prey model:
The emerging spatial dynamics reflect a system
of coupled local populations where the local populations have a characteristic
spatial extent.
The dynamics of the local populations are close
to the dynamics of nonspatial, or homogeneously mixed, analogue of
the spatial model.
Reduces the study of the global population dynamics
to finding the characteristic spatial scale of the local populations
and studying the dynamics of the coupled set of local populations.
By this model we know the position of where
the predator and prey live of a large lattice determined. Each of
can move in different ways, but if there path coincides the predator
esta the prey and reproducts at that point. If a prey sees a predator
aborts if it has a baby; factors like this affect the model.
Spatial interactions can only influence the
population dynamics if the population densities are not spatially
homogeneous.
A deterministic Model of Two Coupled Local Populations
:
Uses the observation that at the characteristic
spatial scale determinism is maximal ans that the populations behave
as if they are well mixed.
To describe the populations beyond the characteristic
spatial scale, we couple a number of local populations diffusively
by assuming that every individiual has a fixed probability of leaving
its local population and migrating to another.
Fluctuations in the densities in the local
populations have been studied in how they affect the dynamics of the
global population.
Simpliest Predator to Prey Model is the Lotka-Volterra
model
In discrete-entity simulations and the simple
multi-patch models, the occurrence of large amplitude oscillations is
prevented by statistical stabilization.
Discrete entity simulations account for all spatial
scales, from the individual up to the system size.
The multi-patch models only distinguish scales
larger than the local-population scale and are fully deterministic
and phrased in terms of density.
There are 2 main differences between the discrete-entity
simulation and the multi-patch models:
Both the local dynamics and individualmovement
in the discrete-entity simulations are strongly influenced by demographic
stochasticity.
The purely diffusive movement of individuals
in the discrete-entity simulations need not necesarily lead to putely
diffusive movementbetween local populations.
Learned basics of how to use computer programs to modelate
and design our team models by using the programs:
Vensim - modeling tool that allows visualization,
documentation, simulation, analization and optimazation of systems interactions.
It is simple and flexible for the construction of simulation models of
all ranges of difficulty. It allows the a model in construction to be analyzed
on the spot.
StarLogo - a programmable modeling
environment designed to help model and explore the workings of decentralized
systems, such as bird flocks, traffic jams, and market economies.
Stella - allows to represent
key physical, biological and social processes operationally and sanity-check
those representations through simulation.
Model-It
- allows students to easily
build, test, and evaluate qualitative models. We could create models that
represent our take of the system interaction of the Amazon and run simulations
in order to test our models.
Had a tutorial with other members of the group in
order to start learning to use Arc View in order to use its subset Population
Viability Analysis and schedule another tutorial for Wednesday at 3:30.
October 28 - November
5:
Read "Regional application of an ecosystem production
model for studies of biogeochemistry in Brazilian Amazonia" by Potter,Davidson,Klooster,
Neptad, Negreiros and Brooks. This scientific paper talks about a model
of ecosystem producion made in the Amazon, it is a little bit complicated
terminology, but serves as an example of models made for the Amazon
that take into consideration the aystem interaction. It helps our group
by taking into account the steps and process done by this researchers
to model an aspect of the Amazon. I also learned important information to
take into account on our modeling like:
Tropical ecosystems are major locations for biogenic
exchange of greenhouse gases, with the atmosphere, and potentially
emission of reactive trospospheric gases.
Evidence from field measurements supports the hypothesis
that tropical forest are the most important global biogenic source
of N2 O and a major source of soil NO.
Considerable uncertainty remains as how seasonal
rainfall patterns and timing interact with soil and land cover types to
control carbon fluxes and emissions of the major trace gases in the Amazon
region.
Understanding the coupling of ecosystem moisture,
carbon, and nutrient flows is a key to advancements in tropical forest
biogeochemistry, and to biosphere atmosphere coupling.
The movement and storage of rainfall through tropical
plant canopies and soils affects many important ecosystem processes,
including evapotranspiration, net primary production, and soil microbial
activity that can produce or consume atmospheric trace gases.
Changes in the pressumed natural balance between
net primary production and soil heterotrophic respiration is not known
where and how changes most rapidly in the Amazon region.
The interaction of climate variability, plant moisture
use, and land cover change is not well understood with respect to the
sustained productivity of vegetation in converted forest lands.
Direct measurements of whole ecosystem fluxes of
moisture, carbon, and nitrogen can be very difficult and expensive to
gather, especially in remote rainforest locations.
Large-scale field studies of terrestrial biogeochemistry
are often founded on the principle that developement and testing of
spatial simulations models are also required to predict patterns and
changes in carbon and nutrient cycling dynamics, and to scale-up to regional
estimates.
Ecosystem simulations can also help identify the
types of ground-based measurements that are most neede to test new theories
of biogeochemical cycling at intensive field study sites.
The interaction between modelling and field studies
is interactive and frequent.
The ecosystem modelling study done permits a reasonably
detailed geographical analysis that demonstrates the potential interactions
of relatively extensive land cover classes, rainfall gradients, and
soil types in the Amazon during a hypothetical 'average' climate year.
The simulation model done helps formulate hypotheses
that can be tested in tropical field studies, such as the Large Scale
Biosphere-Atmosphere Experiment in Amazonia and to potentially identify
within a geographical context which supporting parameter measurements
need to be included in regional research campaigns in order to improve
future modelling research.
Example of the modelling done:
Solar radiation intercepted by persistent cloud
cover over the Amazon basin is a primary limiting factor to evergreen
forest bet primary production where the dry season typically does not
last for more than one or two months.
Deep rooting and drought tolerance by trees in seasonally
dry evergreen forest of the Amazon maintains primary production during
dry seasons that typically does not last for more than two months.
Under conditions of similar seasonal rainfall and
surfaceirradiance, the vegetation of converted forest lands is less productive
on an annual basis than are relatively undisturbed forests in the Amazon
region, mainly because of lower tolerance to drought of pasture and crop
plants relative to the forest trees they are replacing.
Total soil carbon storage convaries with annual
net primary production across the Amazon basin within broad texture
classes, and hence will change as land cover alters ecosystem production.
Soil texture is the dominant controller of N2O:NO
emission ratio over extensive areas of the Amzon basin.
Read "Measuring Water Availability and Uptake in Ecosystem
Studies" in the book "Methods in Ecosystem Science" by Sala, Jackson,
Mooney and Howarth. This chapter is of great help to incorporating to
one of our models the importance and effect of water availability and
how to measure it. I have not finish reading this chapter by time constrains
but in the mean time have learned that:
The 3 most common terms to determine soil water attributes
are mass water content, volumetric water content, and soil water potential.
Soil water content calculated on a mass
basis is defined as ((soil water mass) / (soil dry mass)) or
((wet soil mass - oven-dry soil mass) / (oven dry
soil mass). Mass water percentage is calculated
by multiplying this result by 100%.
Volumetric water content is equal to ((water volume)
/ (bulk soil volume)) or ((Soil water mass / density of water) / (bulk
soil volume)).
Water potential based on the chemical properties of
water is equal to: ((chemical potentisl of water i the system studied
- chemical potential of pure, free water at a reference height, temperature
and atmospheric pressure) / (partial molal volume of water)
There are four ways for meassuring water status on
the environment: gravimetric measurements tecniques, methods for measuring
water potential directly in the environment ( including plant pressure
chamber, thermocouple psychrometry, and the filter paper method), time
domain reflectometry and microwave radiiometry for remotely sensing soil
moisture, a tecnique with considerable promise for ecosystem and landscape
studies.
Simplified water cycle:
The rest will be expanded on next weeks update
Went to the conference by the
CEO of Rocky Mountain Institute and learned ways to develop an economy
and system that reuses and maximizes the production and durability of product
so at the end nature is conserved and greatly reduced the danger people
do on them.
Learned more about modeling in Arc View with the group
of doing a model in Arc View in our group and will continue the expansion
of knowledge on the program in order to do a population viability analysis
of a species that is very adversely affected by change on its
environment as a detector of ecosystem health with a meeting onWednesday
at the Rotch library with a person specialized in modeling programs.
Work on improving the outlook of my website but my 2
hour work did not do any good since my website when seen through the internet
does not reflect the order in which I see when I update and design it.
Plans for next week: Finish reading the chapter in the
book Methods in Ecosystem Science and look for ways to simplify and
understand the complex relationships and conections between key elements
that will be part of our group's final 2 models. Also, try to finish my
problem of how the website looks on the Internet.
November 6 - 13:
Read the book "Handbook of Ecosystem Theories and Management"
edited by S.E.Jorgensen and F. Muller. Concentrated my reading on its chapter
titled "Applications of Ecological Theory and Modelling to Assess Ecosystem
Health" since this is the primary goal of my group's mission. Learned that:
To determine an ecosystem health one must let go
the tempting analogy of ecosystem as an organisim, because it clearly is
not. Rather, one must examine the properties of the complex system that
enables it to persist and evolve in its natural way.
Most regions of the woirld present dysfunctional
ecosystems resulting from the intrusion of human activity. Names a recent
article of Science (Vol. 277, 25 July 1997) that provides evidence that
there are no places left on Earthe that are unaffected by human activity
and that eventually all ecosystems will have to be managed to one extent
or the other.
The global impact of human activity on earth's biotic
and abiotic has transformed between one third and one half of the Earth's
land surface.
A quarter of the Earth's bird species have been driven
to extinction and more than half of the accesible surface fresh water is
being used by humans.
Demonstrated using mathematical modeling the possibility
of ecological restoration of currently degraded land and that it can be demostrated
that the coupling of managment intervention with the power of natural processes
to restore their ecological functions produces new habitats for biodiversity.
Ecological theory and modelling contributes a major
effort to better characterise ecosystem health and informs strategies for
preventive health care of the earth's ecosystems.
Each living organism and its ennviroment form an
individualistic system with the goal of survival and dynamics of systems
with negative feedback control.
The irreversibility of biological processes results
in mortality of individualistic systems whose survivors create an aimless
community organisation.
Flows of energy and materials are undirectional and
have the dynamics of systems with positive feedback control.
States of organisation of forrested landscapes determine
baskets of benefits that satisfy human desires.
Simulation modelling and ecological modelling can
be used to:
Establish the "norms" for operating characteristics
for ecosystems, catchment areas, landscapes and regional environments.
Ascertain what abnormalities might develop with
various stresses or combination of stresses.
Identify causal relationships in the evnt of
unexpected deviations.
Ecological theory provides the foundation upon which
to asses ecological condition.
Combined with modelling and quantification of the
operating characteristics of ecosystems, the practitioner of ecosystem health
is better able to assess existing situations as to evidence for pathology
and better able to establish whether managment practices have been effective
in providing remediation to damaged systems.
One must be highly selective as to the significant
parameters and variables that ought to be monitored (practical limitations
bear heavily on what is realistic in this domain), and one needs to establish
for those selected indicators, the normal operating ranges of healthy ecosystems.
One needs to pay attention to the risks od misdiagnosis
through reliance on too few variables, on the role of human values in establishing
the operating characteristics, and on the potential for the transmission
of ecosystem pathology far from the site origin.
Also read the chapter of the above mention book titled:
"Ecological Modelling: Sytems Analysis and Simulation"; after close reading
and comprehension I learned that:
The concept that understanding the bahaviour of system
components does not guarantee an understanding of the behaviour of the system.
The bahaviour of an individual component can be understood
only within the context of the system of which it is a part.
To develop predictive ecosystem theories, models
are an effective way of dealing with the complexity generated by the interaction
among ecosystem parts.
Ecological systems analysis and simulation, or ecological
modelling, provides a powerfuyl approach for integrating our understanding
of the diverse parts of ecosystems ina rigorous manner.
Systems anlysis is both a philosophical approach
and a collection of techniques, including simulation, developed explicitly
to address problems dealing with complex systems. It emphasises an holistic
approach to problem solving and the uses of mathematical models to identify
and simulate important characteristics of complex systems.
A mathematical model is a set of equations that describes
the interrelationships among objects. By solving the equations comprising
a mathematical model we can mimic or simulate, the dynamic behavior of the
system.
The goal of the systems approach within the context
of ecosystem research is to provide a useful perspective on complex systems
that promotes good research and development of predictive ecosystem theories.
In relation to the physcial or biological sciences,
a system is an organised collection of interrelated physical components characterised
by a boundary and functional unity.
A system is any collection of 'communicating' materials
and processes characterised by many recirpocal cause and effect pathways.
A collection of interacting objects can be viewed as a system.
Systems analysis can be defined as the application
of the scientific method to problems involving complex systems. The essence
of systems analysis is on a braid problem solving strategy rather than in
a collection of quantitative techniques.
A model is a formal description of the essential
elements of a problem. They can be classified as:
Empirical versus mechanistic: Both developed
primarily to describe and summarise a set of relationships, without regard
for appropiate representation of processes or mechanisms that operate in
the real system. Their goal is prediction, not explanation.
Deterministic versus stochastic: Deterministic
models contains no ramdom variables and its predictions under a specific
set of conditions are always exactly the same.
Stochastic models are the opposite; one or more ramdom variables and predictions
not always the same under a set of specific conditions.
Simulation versus analytical: Models that can
be solved in closed form mathematically are analytical models.Models tha
have no general analytic solution must be solved numrically using a specified
set of arithmetic operations for each particular situation the model can represent,
which are named simulation models. Many ecological models are simulation
ones.Simulation models are composed of a series
of arithmetic and logical operations that together represent the structure
and behaviour of the system-of-interest.
Simulation is the process of using a model to mimic,
or trace through step by step, the behaviour of the system we are studying.
There are four fundamental phases in the process
of developing and using a system of models:
Conceptual model formation: Based on the clear
objectives of the modelling project, one abstracts from the real system
those components that must be considered to address our questions. By including
this concepts within our model and excluding all others, we bind the system
of interest. Next we categorise model components depending on their specific
roles in describing the system structure and identify specific relationships
among components that generate system dynamics. Then we formally represent
the resulting conceptual model and finally, describe expected patterns of
model behaviour.
Quantitative model specification: Using the conceptual
model as a template for this quantitative development, we describe the rules
governing the flow of materials in the model using mathematical equations.
We choose a general qualitative structure of the model that lends itself
well to description of complex models, because it facilitates decomposition
of complex iterrelations into simpler cause-effect pathways. Then we develop
the specific equations that collectively comprise the model. Finally, we
translate our mathematical equations to a computer, executing the baseline
simulation, and formally presenting model equations.
Model evaluation: Evaluation of the model in
terms of its relative usefulness for the specific purpose.We should examine
a broad array of qualitative as well as quantitative aspects of model structure
and behaviour. A sensitivity analysis must be performed, which provides valuable
insight into the functioning of the model and also suggests that the level
of confidence we should have in model predictions.
Model use:Assess if
the model meets the objectives that were identified at the begginning of
the modeliing task.The general scheme for model use follows exactly the same
steps involved in addresing a question through expeimentation in the real
world.
Formal exposure to system dynamics should be an integral
part of the training of all ecologist interedted in the development of predictive
theories of ecosystem dynamics.
Model of fluxes, storage and transformations of substances
through ecosystems, watersheds and ladnscapes:
Read in the same book the chapter named: "Ecosystems
as Functional Entities" and learned that:
Mathematical functions are variable quantities, dependent
on one another.
System anaytical functions are subsets or the set
of systemic relations.
Ecosystem analytical functions are observer defined
esembles of ecological processes. They can, for example, be distinguished
into the basic process classes of water processes, energy processes, matter
processes, or community processes.
Socio-economic functions of ecological entities are
natural processes which operate as essentials for human activities.
Example of an energetic flows and storages as parts
of ecosystem function:
November 14 - 21:
Read the book Applied Population Ecology by H. R. Akçakaya,
M. Burgman and L. Ginzburg and learned about population viability analysis
that:
Population viability analysis (PVA) is a process of
identifying the threats faced by a species and evaluating the likelihood
that it will persist for a given time into the future.
Population viability analysis is often oriented towards
the conservation and management of rare and threatened species, with the
goal of applying the principles of population ecology to improve their chances
of survival. Threatened species management has two broad objectives. The
short term objective is to minimize the risk of extinction. The longer term
objective is to promote conditions in which species retain their potential
for evolutionary change without intensive management. Within this context,
PVA may be used to address three aspects of threatened species management:
Planning research and data collection. PVA may reveal
that population viability is insensitive to particular parameters. Research
may be guided by targeting factors that may have an important impact on extinction
probabilities or on the rank order of management options.
Assessing vulnerability. Together with cultural
priorities, economic imperatives and taxonomic uniqueness, PVA may be used
to set policy and priorities for allocating scarce conservation resources.
Ranking management options. PVA may be used to predict
the likely response of species to reintroduction, captive breeding, prescribed
burning, weed control, habitat rehabilitation, or different designs for nature
reserves or corridor networks.
Components of Population Viability Analysis:
The most appropriate model structure for a population
viability analysis depends on the availability of data, the essential features
of the ecology of the species or metapopulation, and the kinds of questions
that the managers of the population need to answer.
We need to estimate the model parameters with field
studies (and sometimes experiments). The kind of parameters that need to
be estimated will depend on the model structure, and the type of data already
available.
For most PVA studies, this is the limiting step, because
data are often insufficient. However, if a decision will be made no matter
what, it is better if the decision-maker has some input from a PVA, even
if the data are not perfect. If a parameter is not known very well, then
a range of numbers can be used for that parameter instead of a single number.
Building a model is a method of combining the existing
information into predictions about the persistence of species under different
assumptions of environmental conditions and under different conservation
and management options. When building a model, it is important to keep a
list of assumptions made.
The structure of the model and the questions addressed
usually determine how the results will be presented. In most cases, the model
will include random variation (stochasticity), which means that the results
must be presented in probabilistic terms, i.e., in terms of risks, probabilities
or likelihoods.
Often, the model must be run many times, with different
combinations of the low and high values of each parameter to make sure that
all uncertainty in parameter values is accounted for. This provides a way
to measure the sensitivity of results to each parameter. Sensitivity analysis
is useful for determining which parameters need to be estimated more carefully.
A minimum viable population (MVP) is one that meets
"the minimum conditions for the long-term persistence and adaptation of a
species or population in a given place" (Soulé 1987a: 1). It is theoretically
sufficiently large to protect against extinctions caused by harmful and unpredictable
genetic, demographic or environmental factors over a given period of time
(generally expressed in hundreds of years). Determination of a generic, rule-of-thumb
MVP size has been the subject of considerable effort and debate and, based
solely on long-term conservation of genetic diversity, should be considered
equivalent to an effective population size (Ne) of several hundred
Ne is defined as the size of an ideal population which
maintains the same genetic diversity as the real population , and is equivalent
to the number of breeding animals per generation. It is a function of social
organization and population demographics, and as a standardized measure permits
comparison between species, between populations of the same species, and
between the same population at different times. It is usually only a fraction
of actual population size (N), because not all animals in the population
are breeders. The smaller the ratio of Ne to N, the greater the chance for
genetic drift and the greater the level of inbreeding characterizing a species.
A population size sufficient to mitigate environmental
and catastrophic uncertainty (habitat change, or an epidemic or natural disaster)
should be considerably larger than one required only to conserve genetic
diversity . Taking these factors into account, viable populations are thus
expected to be of the order of several thousand individuals.
MVP size decreases sharply if the population is
not completely isolated, but maintains even a low rate of genetic migration
from other populations. Simulation models indicate that immigration of new
animals can substantially reduce a population's extinction risk.
Read "Predictive accuracy of population viability analysis
in conservation biology" by Barry w. Brook, Julian J. O'Grady, Andrew P.
Chapman, Mark A. Burgman, H. resit Akcakaya and Richard Frankham, and
learned that:
Population Viability Analysis is widely applied in conservation
biology to predict extinction risks for threatened species and to compare
alternative options for their management. It can also be used as a basis
for listing species as endangered under World Conservation Union criteria.
There is considerable scepticism regarding the predictive
accuracy of population viability analysis. Some factors that prove this are:
The confidence intervals are so wide that the analysis
provides little or no information about the magnitude of extinction probabilities.
The data is inconclusive and unssupportived many
times.
Risk estimates for individual species are very imprecise.
PVA is a sufficient accurate tool for categorizing and
managing endangered species, its results provide strong empirical justification
for its use of categorizing the vulnerability of endangered species and evaluating
options for their recovery.
November 21- 28:
Between the whole process of rearranging teams, I managed
to read and learned more about population viability analysis in order to
finish our models. I read "Emerging issues in Population Viability Analysis"
by J. Michael Reed, L. Scott Mills, John B. Dunning Jr, Eric S. Menges, Kevin
S. McKelvey, Robert Frye, Steven R. Beissinger, Marie-Charlotte Anstett and
Phillip Miller and learned that:
There is no single process that constitutes PVA, but
all approaches have in common an assessment of a population's risk of extinction
or its projected population growth either under current conditions or expected
from proposed management.
As model sophistication increases, and software programs
that facilitate PVA without the need for modeling expertise become more available,
there is greater potential for the misuse of models and increased confusion
over interpreting their results.
PVA is a powerful tool in conservation biology for comparing
alternative research plans and relative extinction risks among species. However
it should be used with caution for many reasons:
Because PVA is a model, its validity depends on
the appropriatness of the model's structure and data quality.
Results should be presented with appropiate assesment
of confidence.
Model construction and results should subject to
external review
Model structure, input, and results should be treated
as hypothesis to be tested
PVA should not be used to detemine minimun population
sizes or the specific probability of reaching extinction.
The definition of PVA should be restricted to development
of formal quantitative models.
One should focus more research to determine how pervasive
density dependence feedback is across species.
The most appropriate use of PVA may be for comparing
the relative effects of potential managment actions on population growth
or persistence.
PVA is:
one of the most powerful and pervasive tools in conservation
biology.
an assesessment of the risk of reaching some threshold,
such as extinction, or the projected growth for a population, either under
current conditions or those predicted for proposed management.
important to review and assess as a tool in conservation
biology.
used increasingly in policy development and management
planning.
popular by its models' ability to provide apparently
precise results.
PVA should not be used superficially.
Early viability assessment focused on estimating minimum
viable population size, a concept that has long been recognized.
Despite recognition that population viability is determined
by a combination of factors, the first viability analyses focused on subsets
of factors.
Increasingly sophisticated models are required if one
wishes to account for strong interspecific interactions, such as mutualism.
The primary factor affecting viability of many rare
species is loss of habitat.
Populations can be affected by the actual loss of habitat
from a region, by changes in the suitability of remaining habitat patches,
or by landscape factors such as isolation or connectedness of the habitat
fragments remaining after habitat loss.
If the spatial distribution of habitat potentially affects
the viability of a study population, than a PVA developed for this population
must explicitly deal with changes in habitat quality and quantity across
space.
When a species is limited by the amount of suitable
habitat present within a region, but is unaffected by the distribution of
habitat patches, then a researcher can conduct a PVA by tracking the total
amount of habitat present in the landscape.
Models must integrate space to a greater degree than
by simply tracxking total amounts of habitat.
Forest management plans that could change the spatial
distribution of suitable habitat could therefore have dramatic effects on
an endangered animal.
Models that incorporate the exact spatial and temporal
location of objects into their structure are said to be spatiallly explicit.
Individual-based simulation models assign to individuals
habitat-specific demographics traits based on where individuals are, and
they move individuals based on specific dispersal rules. This models average
across individuals to gain population statistics, such as time to extinction
or population size at specific moments in a simulation period.
Spatially explicit models have been used to address
at least two kinds of conservation questions that are simulation of populations
on hypothetical landscapes and the examination of potential effects of a
specific landscape change proposed for a specific, real-world landscape.
The most valuable uses of spatial PVA models may be
in the identification of extreme population responses to landscape change
and possibly to rank landscape-change scenarios and their potential to affect
target populations.
Spatially explicit models add the requirements of severe
unique kinds of data: the distribution and quality of habitat in the real
world, local habitat-specific demography, and an idea of dispersal
patterns and movement rules.
Spatially explicit models must specify the habitat needs
of the population so that the model can determine how individuals are distributed
across the landscape at each time step.
Spatially explicit models require information on how
organisms disperse across complex landscapes.
Modeling studies can assess the effects of estimation
errors for various dispersal parameters on the performance of spatial PVAs.
Spatial explicit models may be greatly susceptible to
error propagation and should be used with caution.
Sensitivity analysis can complement the predictions
that arise from PVA by providing constructive insights into factors that
most affect population growth or quasi-extinction probability.
There are several approaches for conducting sensitivity
analysis, ranging from analytical sensitivity and elasticity analysis.
Sensitivity analysis can benefit researchers by identifying
factors whose estimation is most critical for population-level studies. It
can include spatial dynamics, thereby evaluating the relative impact of within
population processes on metapopulation persistence or growth.
Although sensitivity analysis helps identify actions
that can be taken after traditional PVA has identified a problem, it is a
critical to account for not only the extent to which equal changes in different
vital rates affect population growth or extinction, but also the amount that
different rates could change.
Last update: somewhere in the 20th century at 6:20
PM ET (daylight savings time)
December 9, 2002
"Better to have a mediocre website with good content than
a flashy website with the font color only sticking out"
In other words I prefer doing research and spending my
time in more useful things for the improvement of the team work than
making my site artistically tasteful.
