Untitled

What is Ecology?

• Ecology comes from the Greek words "Oios," meaning hours, and "logos," meaning reason
• Ecology studies the natural environment and the relationships between organisms and their surroundings
• Ernst Haeckel coined the term "ecology"
• It involves both external and internal environments, including biotic factors like competition and abiotic factors like temperature
• Abiotic factors include temperature and adaptations to extreme temperatures
• Alexander von Humboldt and Charles Darwin were considered among the first real ecologists

Haeckel

• Haeckel believed in Lamarckian inheritance
• Haeckel proposed the theory of ontogeny recapitulates phylogeny
• Haeckel invented the protista kingdom
• Haeckel was interested in biogeography, or how distributions change over space and time
• He observed that vegetation resembles that of a mountain as one moves away from the poles

Von Humboldt

• Von Humboldt lived in the late 1770s-1800s
• Von Humboldt was interested in development
• Von Humboldt studied gradients in Latin America
• Von Humboldt was interested in biogeography and how distributions change over space and time
• He noticed that vegetation resembles that of a mountain as one moves away from the poles
• He recognized similarities between the Atlantic coasts of South America and Africa

Darwin

• Darwin was interested in competition and species interactions
• Darwin was interested in the population growth of elephants
• Darwin noted that low population rates would become extremely large if there was no carrying capacity

Ecology v. Environmentalism

• Ecology is a science that observes nature, generates hypotheses, tests them, and synthesizes empirically supported hypotheses and theories to make inferences about the scientific world
• Ecology is also considered a political movement aimed at environmental protection
• Environmentalism is considered a concern
• Ecologists must remain objective despite prejudices and biases
• Prejudices are resolved by using the scientific method

Good Science Requirements

• Good science requires data that is of good quality and appropriate for the question
• Good science requires appropriate analysis of the data and appropriate experimental design
• Good science requires reasonable interpretation of results

Scientific Method

• The scientific method involves observation and quantitative description
• The scientific method involves generating and testing falsifiable hypotheses
• The scientific method involves testing specific hypotheses
• The scientific method involves synthesizing empirically supported specific hypotheses and parsimonious inferences into theories and scientific laws
• Not all steps are necessary in every experiment

Observation and Quantitative Description

• A statistical test is done to see if the hypothesis is reasonable when generating a hypothesis
• Empirical data must be gathered through observation to test hypotheses
• A hypothesis cannot be rejected unless there is adequate empirical data
• Quantitative data is desirable because counts and amounts can prevent deception by appearances
• Quantitative data is more amenable to statistical analysis
• Statistics allow for quantifying confidence in the results
• Error bars quantify confidence

Generating Hypotheses

• Hypotheses are educated guesses that remain largely untested
• Example hypothesis: Japanese honeysuckle reduces plant species richness in forests in North Mississippi
• Testing the hypothesis requires doing an experiment, which experiments are costly and difficult
• Empirical data support is sought before conducting an experiment
• Statistical techniques called regression are used to explore data

Regression vs. Correlation

• Regression and correlation are statistical techniques that quantify the strength of the relationship between continuous variables
• Mathematically, regression and correlation are the conceptually same but differ slightly
• Regression examines the relationship between a continuous response variable (y-axis) and a continuous predictor variable
• Regression tests the hypothesis that the predictor variable causes variation in the response variable
• Spurious Correlations are when there is a perceived correlation but its false association with the two variables
• Correlation analyzes two continuous variables where cause and effect is not implied
• With correlation, both variables are assumed to be random, and either variable could be the predictor or response variables
• Correlation looks at the strength of a relationship only
• Correlation doesn't prove causation
• Looking at where sweetgum is grown and obesity rates in America, sweetgum might be assumed to cause increased obesity rates, however this is not true because sweetgum is not ingested

Regression Analysis

• Ecologists are frequently interested in the frequency of one or more species and the values of environmental variables
• Regression analysis can be demonstrated using a scatter plot
• The y-axis represents the response variable
• The x-axis represents the predictor variable
• Species richness is defined as the number of species
• R² ranges between 0 and 1 and tells the amount of variation in the response variable

P Value

• P-value is the probability of being wrong
• A lower value for P the better

R: Simple Regression Coefficient

• R is a weighted standardized slope of the best-fitting line
• A test is done to see if the line is significantly different from zero
• The relationship must be examined to determine of it is positive or negative
• R ranges between -1 and 1

Japanese Honeysuckle Example

• A study determines if non-native Japanese honeysuckle potentially has a negative effect on the number of native plant species in the forests
• Look to see if there is a negative correlation between the number of native species and the abundance of Japanese honeysuckle before the experiment
• Diversity is species richness compared to soil compaction
• r= -.6 (R² =.36)

Reasons to Avoid Assuming Cause and Effect

• Cause may be confused with the effect
• Variables could be opposite
• More species rich communities may be more resistant to invasion by honeysuckle
• Confounding variables and a lack of independence from other predictors can interfere
• The abundance of Japanese honeysuckle may be correlated with another predictor variable that is actually causing variation in species richness

Multiple Regression

• If more than one variable could affect the responuse variable, multiple regression can determine the greatest correlation with resposne variable
• There might be more than one variable that could potentially affect species richness
• Possible examined relationships between the number of Japanese honeysuckle and several environmental variables to see which had the highest correlation with the species richness
• Multiple regression is not multiple simple regressions
• If multiple simple regression is done, R² can improperly add up to more than one
• Multiple regression shows how each variable independently contributes to native species richness and accounts for variation

Simple Regression

• Simple regressions do not account for the common variation explained by the three predictors
• Simple regressions overestimate the overall explained variation by the three variables and misinterpret overall importance
• In the figures, simply regressions gave would assign one thing the highest rank and another the lowest etc
• Multiple regressions would assign different ranks

Multiple Regression

• Multiple regression combines and compares the independent effects of multiple continuous predictor variables on a single-response variable
• When doing multiple regression, several partial regressions are runs
• Partial regression examines residual variation in a response variable against the residual variation in a predictor variable given one or more covariables
• The overlap is subtracted, and partial r² is desired for the sum but not greater than one
• A covariable is a continuous variable whose variation we want to account for
• Residual variation is the unexplained variation in the variable by one or more other variables of interest
• Partial regression of species richness versus Japanese honeysuckle cover given soil compaction can be calculated
• Run two multiple regression analyses to obtain desired information
• Partial regression for each predictor completes the multiple regression
• Only through this process can predictors be described as independent variables

Correlation Limitations

• Correlation doesn't prove causation
• There could be other unaccounted variables
• Variables that explain the most variation in species composition may correlate with unmeasured variables that impact the distribution of species
• Soil compaction may reduce plant species richness or be correlated with past agricultural disturbance, which eliminated native plant species
• Compacted soils may historically have had had deep silt loam soils which were conducive to plant life

Distinguishing b/w Cause and Effect

• Even if all confounding variables could be accounted for and found unimportant you can't necessarily determine the difference between a cause and effect
• There may be a significant, negative partial regression between species richness and honey suckle cover, but it does prove the two are linked and not what is causing eachother

Multiplicative Effects

• Multiple regression assumes additive effects, not multiplicative effects
• A multiple regression approach assumes that the effects of soil compaction and honey suckle cover were additive; what if the effects were multiplicative?
• A what if an experiment is a backseat driver
• What if soil compaction increased the effect of japhoney on native species
• Additive effects: the effects add together when combined
• Ex: low honeysuckle, high compaction + high honeysuckle, low compaction = high honeysuckle, high compaction
• Multiplicative effects: the combined effects are increased compared to their individual effects
• One increases or decreases the other

Soil Compaction

• Soil compaction maybe causes honeysuckle to have greater effects
• The overall combined effect depends on each individual one
• In this case, one is increasing the effect of the other
• Ex: low honeysuckle, high compaction + high honeysuckle, low compaction = high honeysuckle, high compaction

Manipulative Experiments

• There are two major challenges when ecologists perform controlled experiments in the field
• Designing an appropriate ecological experiments is hard Normally an appropriate experiment and proper controls and hard to organize

Replicates

• Treat the controlled and experimental group the same except for testing treatment
• Treatment may manipulate other things that you don't want to manipulate
• Adequate replication of the treatment/controls and statistical analysis, is can be hard and difficult to setup
• A "true test" will have more variation of the overall results

Field experiment in Pine Savannahs

• A field experiment will have a problem with control and replication that should be avoided
• A fire simulation is used a fire on a savana
• Design a test which could have too many problems
• Determine the effect of field on flowering on a chosen field or area
• Approach is too simple, find 2 sites and decide if there is a difference

True Replicates

• In an experiment you need lots of replication and treatment
• This may have a significant difference and percentage or "no fire" site and " light night fire" site
• This test assumes fire stimulates the tree and grass to produces
• This is called Pseudoreplication, you did one fire and took samples from it
• A real experiment must be random in is execution

Pseudoreplication

• Treating multiple correlated observations as independent
• Treat all observations as the same "samples within" - are independent sample
• Treat a "site" and take samples
• An observation that randomly assigns an independent " treatment of control" replicates is true
• Each experimental unit is an observation

Issues with Replication

• With the use of the lower figure data shows that fire and flowering percentage results shown slightly overlapping
• This is because that there are two replicates
• This may be an issue that is too small of an experimental amount and statistical results are affected

Experimental Variation

• Pseudoreplication increases "power" but must show is the data valid
• A high number sample count, can cause a false analysis where samples are under looked
• The only effect of this may not be relevant or due to the stimulus, but another factor underlooked

In An Experiment

• If the experimental group is low then pool it
• If variations are great, combine the groups as "psuedoreplactes

Issues with Power of Statistics

• Design experiments by the experimental units and control points
• Meta Analysis; use the combined agreement of all tests that may increase strength in numbers overall

Experiment Limitations

• Complex Systems manipulated for the ecosystem may create new issues or problems

Alternative experiment

• Modelling the data
• Physical situation is not easily seen or data not collected over time

Models

• Models are needed to study something, but it must be something directly being examined
• Regression -- The Line of BEST fit formula will guide experiment
  • Multiple regresses contain data with a cause a reason, but something not to that scope - the slopes can vary on a graph
• Model may be made with different equations, not linear

Model Interpolation

• Predication "values"
• Known and use to "Predick unknow values"
• validate the the high and low "multiple regresses" -Use the "test" to check how may or bad data

Experimental Data

• Predications based directly from the field data
• Extrapolation- data based on what has already has is made from the same data previously extracted

Applications in Science

• Manipulate data ( solve questions) - using "formulation" that test the un observed data already known - the set up with a a set of known variables- then use that to extract other values
• "Solving model" -Manipulate the" what if data that has been setup - Use precollected data to establish an outline for what more is to be seen

Modelling

• Manipulate "The predictor of interest" and that will yield new results
• The is two common categories in "MODEL SCIENCE'
  • an Analytical Model
    • simple model ( " few data variables )
    • many set variables" assumed values that always occur"
    • ONE DATA extraction only !!!

Analytical Limitations

• • Often wrong but can help understand topic more

• Simulation Model

  • COMPLEX ( data value and interperiation of numbers " high data use" -- FEW assumptions, data all collected

Data use

--- large ranges , testing, with high " sensitives set amount" -Make small increments for the the predictor and pre set parameters

Hypothesis testing

• Then create new data and hypothesis
  • use the model for new stability test ( stability that may alter the set "Data test')

Models

• In complex data set there may not be stability with model data -Changes in the equation or solution
  • if stable the " model can continue. if one value goes outside, then the test is no good

MODEL validation

Does the" new set "match the old one? is the main question

• can't find the old values!

• • use the older data set for the new data test

• -- The is the " limitation of the experiment because you made the data that can be wrong or right !!!

Feedback

• Negative ( change the value by something effecting another parameter)

  • High CO2 (increases temp) this happens to increase plants that use Co2, this will decrease CO2

• Positive (change the parameter to effect an extreme result)

  • increase CO2 - increases temp- and increases erosion- this will increase CO2

• ---These can make "cause or effect " relations impossible

Are Thresholds present?

• Inaccurate test can be made by the the value set in the test Dramatic increase can occur with a low value- results in limited values that are not present in test

Empiricism

• test are repeated
• test can have Meta-analysis

Rationalization

• Know what one is looking for!
• Prioritize observations based on theory

Plant STRATEGY

• STRESS data - PLANTS with low nutrients can fix or store those nutrients, slow growth

RUdelals Plants

• Adapt to high disturbance, rapid life

• COMPETITORS

• low disturbance and stable growing

• Not possible to high STess and Disturbence

• Plant STRATEGY is to "extract data" and not have an environment where it would do well

• Violate certain assumptions

Carnivoiry

• Not a stress factor on " the plant" it can produce energy in the food already found in the soil

Regional Biomas Studies

• Study what makes and affects the plants in their systems
• Temp has effects!!
• Sun strength varies the temp, higher at the equater
• Sun hits the earth at various angles with heat as" angle increases, so does heat that is radiated
• ** Co2 cycles and Gases
• Greenhouse effects - natural process - radiation from space is trapped, and warms the planet

Gasses

• Greenhouse Gas

• -- Seafloor gasses and Volcanoes

• • Volcanoes release co2 underwater increasing ratios over time

• Ocean data and Temp

• Sea Spreading causes temp to fluctuate -- As areas shift or move they can create areas where gases and trapped or released

Lava and Ash

• Volcanoes underground increase global climate
• Eruptions limit heat that is not able to reach earth

Mountains

• Higher in elevation reduces temp
• Sun heats a level zone then expands out- high elevation may never get heated
• Atmosphere density decreasing is due to the sensible Heat
• Sensible heat relates to the decrease in density of are in the atmosphere that results in lack of heat
  • Mountains are cold " snow and such" even if they are in an area that is warm! That zone is not directly impacted

Global warming

• changes with the planets cycles in Space
• Sun cycles of radiation output ---sun Spots cause extra energy ---black areas in these sun flares are cooler; Orange is Hotter

VOLCANOES

• add particles that block the entry of heat

Latitude and Temp

• warm air tends to rise, in an area where cold air sinks

Adiebetic Laps

• limited heat transfer, warm does not need help

Condensation

• precipitation that occurs that the temp increases
• Warm can hold more then cold, lots of the aquater

Air

• Cool air is dry

Hadley Cell's

• Air has pull with earth
• Warm air will escape, then return

HIGH AND DRY

• areas that are dry 30 deg from Earth, -- Deserts!

• -- sun increase the heat, increases the pressure in rain

• Polar Ice melting ( dry cold ) high pressure - winds that have limited effects based on earth cycle

Coriolis Effect

• Jet streams

Trade Winds

• pick up moisture on a cycle along the ocean surface

Regional Climate Effects

• North america ( East Coast --influenced by air masses and water zones -Jets move Easterly

Pacific Wind Effects

• Water in the west " stacks'' water on islands, pushing it in all directions creating Tsunami 's
• --Low is rain and storms in the arears where pressure is limited and or high

Ocean Shadows

• Rain shadow mountain ranges or an obstacle causes a mountain effect
• • Hot air that is warm or is in rain shadow on the edge- if warm causes storms Biomes
• Defined through the " Vegetation" as a growing mass Whackers, Biomass
  • temp and percipitation - rain effects can do it on the area around

WHACKERS

• temp and rains on the biomass - High temp high rain ( tropical rain)

  - temperate can produce, desert, grass, forests " temperate peramiters"
  

• Climate change ( causes change and effects across " earth and all parts)

Transpiration

• Heat can effect soil
• --soils can effect life that can thrive - so all is clay or too thin for holding