Single species models: Exponential, logistic, and Gompertz growth

 The chapter discusses three key growth models for simple homogeneous
populations, emphasizing the balance between generality and specificity.
 Population dynamics are determined by four processes: birth, death, immigration,
and emigration, with a focus on closed systems.
 The intrinsic rate of growth (r) is defined as the difference between per capita
birth rate (b) and per capita death rate (d).
 The equation for population change is expressed as dN/dt = rN, leading to the
exponential growth model N(t) = N0e^(rt).
 The model predicts exponential growth for positive r, exponential decay for
negative r, and constant population size when births equal deaths.
 Graphs illustrate constant per capita growth rates, increasing population growth
rates, and the resulting population trajectories based on r values.
 The section highlights the importance of understanding these models for
predicting general patterns of population growth.

System equilibrium:

The equilibrium point for E. coli population growth is N*=0, which is unstable if r>0 and
stable if r<0. Under ideal conditions, E. coli doubles approximately every 20 minutes,
indicating a high growth rate.

Problems with the exponential model:

 The equation is represented as ( \frac{dN}{dt} = rN(1 - \frac{N}{K}) ), highlighting

negative density-dependence.
 Density-dependence is influenced by factors such as resource depletion,
disease, and predation, affecting high-density populations more significantly.
 At low population densities, growth approximates exponential growth, while at
high densities, growth slows due to crowding effects.
 The maximum growth rate occurs at half the carrying capacity (K/2),
demonstrating an S-shaped growth trajectory.
 The logistic model was first introduced by P. F. Verhulst in 1838 and later
popularized by R. Pearl in 1920.
 The model has one stable equilibrium point at K and one unstable equilibrium
point at 0, indicating population dynamics' stability.

How to practically determine an equilibrium point is stable or unstable? Assume

N* is an equilibrium point for the autonomous system:

 The equation ( \frac{dN}{dt} = f(N) ) defines population growth, with stable

equilibrium at ( f'(N^) < 0 ) and unstable at ( f'(N^) > 0 ).
 The theta-logistic model describes population growth with the equation ( \frac{dN}
{\theta dt} = rN \left(1 - \frac{N}{K}\right) ), where ( \theta = 1 ) simplifies to the
logistic model.
  r-selected species, like dandelions, focus on high growth rates and produce
many offspring with low survival rates, thriving in unstable environments.
 K-selected species, such as elephants and humans, invest in fewer offspring with
higher survival rates, competing effectively in stable environments.
 The r/K selection theory, coined in 1970, categorizes species based on their
reproductive strategies and adaptations to environmental stability.

C-S-R theory:

 Grime's C-S-R theory categorizes plant strategies into three types: Competitors,
Stress-tolerators, and Ruderals based on stress and disturbance levels.
 Competitors thrive in low stress and low disturbance environments, utilizing
resources efficiently with fast growth and high phenotypic plasticity.
 Stress-tolerators flourish in high stress and low disturbance conditions,
characterized by slow growth rates and long-lived leaves.
 Ruderals are favored in high disturbance but low stress habitats, often being
short-lived annuals that produce many seeds.
 The Gompertz growth curve describes population growth with a decreasing per
capita growth rate, represented by the equation ( \frac{dN}{dt} = r_0 e^{-\alpha t}
N ).
 The Allee effect indicates a decline in individual fitness at low population
densities, with critical thresholds below which populations may face extinction.
 The Allee effect is mathematically represented as ( \frac{dN}{dt} = rN(1 - \frac{K}
{N})(A - 1) ), highlighting positive density dependence.
 The section emphasizes the importance of understanding these ecological
theories for conservation biology and population viability analysis (PVA).

Chapter 2. Harvest models: Sustainable yield and bifurcation

 Ecosystem equilibria can vary with conditions like nutrient loading and fishing
intensity, leading to stable and unstable states.
 The logistic growth model for fish populations includes a harvesting equation,
where equilibrium occurs when growth equals harvest.
 Two equilibrium points exist: a stable equilibrium (N*) and an unstable
equilibrium (N=0), with N* defined as K(1 - qE/r).
 Bifurcation occurs when fishing intensity (qE) approaches the growth rate (r),
leading to potential population collapse.
 The maximum sustainable yield (MSY) is achieved at optimal harvesting effort
(E_MSY = r/2q), maximizing yield.
 An Allee effect model introduces three equilibria, with bifurcation occurring at qE
= 0.1406, indicating a risk of catastrophic collapse.
 The collapse of the Atlantic cod population in the early 1990s exemplifies the
consequences of unsustainable fishing practices.

Chapter 4: Discrete-time models

  The simplest discrete growth model is represented by the equation (N_{t+1} =
R_0 N_t), where (R_0) is the net reproduction rate.
 When (R_0 > 1), populations exhibit geometric growth, while (R_0 < 1) indicates
geometric decay.
 The logistic difference equation is given by (N_{t+1} = (1 + r)N_t - \frac{r}
{K}N_t^2), showing density-dependent growth.
 The Beverton-Holt model is defined by (N_{t+1} = \frac{R_0 N_t}{1 + (R_0 - 1) \
frac{N_t}{K}}), with different density-dependence characteristics compared to the
logistic model.
 The logistic model can lead to overcompensation, while the Beverton-Holt model
demonstrates compensatory dynamics.
 Stability analysis shows that (N^* = 0) is unstable if (R_0 > 1) and (N^* = K) is
stable if (R_0 < 1).
 The section emphasizes the importance of understanding density dependence in
population dynamics and its implications for ecological stability.
 The logistic map illustrates how changes in parameters can lead to complex
behaviors and chaos in population dynamics.

Logistic map

The growth rate α is defined as 1+r, with x_t+1 = αx_t(1-x_t).

Chapter 5. Models of species interactions

 Species interactions can be harmful, beneficial, or neutral, with four main types:
competition, predation, mutualism, and commensalism.
 Competition negatively impacts both species involved, while predation benefits
the predator and harms the prey.
 Mutualism benefits both species, and commensalism benefits one while the other
remains unaffected.
 The Lotka-Volterra competition model illustrates how two species compete for
resources, affecting their growth rates.
 Coexistence of species occurs when interspecific competition is weaker than
intraspecific competition, as shown by Gause's Principle of Competitive
Exclusion.
 The Lotka-Volterra predation model describes the dynamics between prey and
predator populations, highlighting their cyclical nature.
 Harvesting can increase prey populations while decreasing predator populations,
demonstrating the impact of human intervention on ecological balance.
 The Rosenzweig-MacArthur model incorporates density dependence and shows
stable equilibria in predator-prey interactions, emphasizing the complexity of
ecological relationships.

Mitosis versus Meiosis

Sexual reproduction enhances genetic diversity by producing unique gametes through
meiosis, resulting in haploid (n) cells. The human genome serves as a historical
narrative, with genotype influencing phenotype, while the central dogma describes the
flow of genetic information from DNA to protein.

Each combination of three bases forms a codon that codes for an amino acid.
Since there are four bases, there are 64 possible amino acids (4^3=64). Only the
following twenty, however, code for protein synthesis

Locus defines gene location; alleles are gene variants; allele frequency indicates
relative allele proportions.

Genetic diversity is just allele diversity!

 Genetic diversity (GD) is essential for populations to adapt to environmental

changes and is measured by allele frequency.
 In a study of African lions, 20 out of 26 loci were monomorphic, while 6 showed
variation, indicating low genetic diversity.
 The average number of alleles per locus (allelic diversity) was calculated to be
1.27.
 The proportion of polymorphic loci was 6/26, or 23%.
 The average heterozygosity per individual lion was 7.1%, indicating low genetic
variation.
 Heterozygosity at the ADA locus was 0.564, while the overall heterozygosity
ranged from 0.015 to 0.038 for allozymes.
 The percentage of heterozygous loci in African lions is comparable to that of
identical twins in humans, at 3.8%.

Hardy-Weinberg Equilibrium

 The Hardy-Weinberg principle states that allele frequencies in a population

remain constant in the absence of evolutionary influences.
 Allele A has a frequency of p, while allele a has a frequency of q, where p + q =
1.
 The frequencies of genotypes are represented as P(AA) = p², P(Aa) = 2pq, and
P(aa) = q².
 In North America, the frequency of the PKU allele is approximately 0.01, leading
to a PKU incidence of 1 in 10,000 newborns.
 Two populations with the same allele frequencies can reach the same genotype
distribution after one generation of random mating.
 A study on one-horned rhinoceros showed a χ² value of 0.048, indicating no
significant deviation from Hardy-Weinberg equilibrium.
 In a sample of 38 Polish brown hares, the frequencies of three alleles need to be
estimated to test for H-W equilibrium.
 In a flower color study, 375 white and 125 pink flowers were counted, allowing for
the calculation of allele frequencies and heterozygote frequency.
There are four major evolutionary forces that drive species to evolve. In other
words, these forces will break down HWE and change the allele frequencies in a
population.

The frequency of brown coloration genes in a beetle population increased from 25% to
29%, while green genes decreased, potentially due to mutation, migration, genetic drift,
or natural selection. Genetic drift can significantly impact small populations, leading to
rapid loss of genetic diversity.

Key points

 Genetic diversity is measured by heterozygosity, and small populations are

particularly vulnerable to genetic drift and inbreeding.
 The Hardy-Weinberg Equilibrium (HWE) states that genetic change occurs when
its assumptions are violated, leading to evolution.
 Genetic drift causes random changes in allele frequencies, resulting in loss of
genetic variation and increased homozygosity.
 In a population of size N=1, allele fixation can occur, permanently losing one
allele.
 For a progeny population size of N=200, the expected allele frequency interval
can be calculated using variance formulas.
 The Mauritius kestrel experienced a bottleneck of N=2, leading to a decrease in
heterozygosity from 0.231 to 0.099 after generations.
 The expected heterozygosity after three generations for the kestrel is
approximately 0.097, closely matching observed values.
 The decline in heterozygosity per generation is given by the formula ∆h = 1/(2N),
emphasizing the impact of small population sizes.

Summary on the Effect of Genetic Drift

 Genetic drift reduces genetic diversity due to sampling error, particularly in small
populations, leading to increased inbreeding.
 The inbreeding coefficient (ft) indicates the increase in homozygosity over
generations, with significant effects in small populations.
 A population bottleneck occurs when environmental events sharply reduce
population size, increasing the risk of allele loss.
 The expected number of alleles remaining after a bottleneck can be calculated
using allele frequencies and population size.
 Effective population size (Ne) is crucial for conservation, with inbreeding effective
size (Nef) and variance effective size (Nev) being key metrics.
 The 50/500 Rule suggests populations with Nef under 50 are at immediate
extinction risk, while those with Nev under 500 face long-term risks.
 The Eastern fence lizard population example shows Nef of 68 and Nev of 70,
indicating significant inbreeding and drift effects.
 Conservation strategies must consider demographic stochasticity and genetic
drift to maintain viable populations over time.
Which one is more diverse?

 Species diversity is determined by species richness (S) and evenness of

individuals among species.
 The text outlines various field methods for collecting biodiversity data, including
quadrat sampling and trapping.
 Species richness (S) is defined as the number of species in an area, with S being
5 in the provided examples.
 Two widely used indices for measuring biodiversity are Simpson’s index (D) and
Shannon’s index (H).
 The data table shows species abundance, with the highest abundance being
7599 for species ALSEBL.

The Simpson index:

 The Gini-Simpson index and Shannon index are used to measure biodiversity,
with Simpson's index focusing on dominant species and Shannon's on rare
species.
 In a community with 5 species, the Simpson index calculated is D' = 0.425, while
the Shannon index is H' = 1.145.
 A hypothetical scenario shows that a plague reducing species from 30 million to
15 million results in a 4% drop in Shannon diversity and negligible change in
Simpson index.
 Hill's numbers provide a way to express diversity indices in terms of equivalent
species, with D0 representing species richness, D1 as the exponential Shannon
index, and D2 as the Simpson index.
 An example from a 50 ha plot with 211 tree/shrub species shows a Simpson
index of D = 0.971, translating to an effective number of species of 34.483.
 The Shannon index for the same plot is H' = 4.231, equating to 68.786 effective
species.
 The text emphasizes that comparing diversity indices is only valid when
converted to equivalent communities with even abundance.
 Overall, diversity indices reveal different aspects of biodiversity, highlighting the
importance of considering species abundance distribution in assessments.

How many species are there on the Earth?

There are approximately 1.8 million known species on Earth, including 960,000 insects
and 270,000 plants. The total includes 5,000 viruses, 4,000 bacteria, and 4,500
mammals.

5 ~ 10 millions

 Terry Erwin estimated 30 million insect species in tropical rainforests based on

beetle diversity from a single tree species.
  He found over 600 beetle species on L. seemannii, leading to an estimate of 600
insect species specialized to that tree.
 Erwin assumed two-thirds of insect species reside in the canopy, resulting in a
total of 600 species per tree.
 With approximately 50,000 tree species in tropical forests, this leads to an
estimate of 30 million insect species.
 The text also introduces the coupon collector's problem, highlighting the
challenges in estimating species richness through sampling.

Species-individual Curves

 The species-individual and species-area curves help estimate species richness,

determine minimum sample sizes, and compare community characteristics.
 The cumulative species-individual curves can be modeled using logarithmic,
power, and logistic models, with predictions of species richness varying by
model.
 In a 25 ha plot of Bei-Shan-Zu Nature Reserve, 223,381 individual stems were
recorded, belonging to 192 species.
 The predicted number of species based on different models from 100 samples of
50 trees each were 203.3, 375.5, and 228.3.
 Increasing sample size generally improves species estimation, and there is no
universally superior model for species richness estimation.
 The analysis emphasizes the importance of sampling effort in accurately
estimating species diversity in ecological studies.

Taxonomic discovery curves

 The cumulative number of plant species discovered increased significantly from

1750 to 2000, with a total of 25,000 species documented.
 Key historical events influencing species collection in China include the arrival of
Matteo Ricci in 1583 and the end of the Qing Dynasty in 1911.
 Various models for predicting undiscovered species include negative
exponential, logistic, and nonlinear models, emphasizing the importance of
taxonomic effort.
 Notable studies referenced include Joppa et al. (2011), which found that
biodiversity hotspots contain most undiscovered plant species.

k = Tt (a + bΔSt) Our model, better describing discovery efficiency

 The first-order Jackknife estimator (Jack1) estimates species richness using

quadrat data, yielding an observed richness of approximately 19 species.
 The second-order Jackknife estimator (Jack2) improves upon Jack1, estimating
species richness at 21.1 species based on the same data.
 The Bootstrap estimator incorporates the proportion of quadrats containing each
species, but its variance is complex and referenced in Smith & van Belle (1984).
  The Chao1 estimator, designed for abundance data, estimates species richness
at 126.5, with a variance of 89.7.
 Chao1 is particularly effective for datasets with many rare species, while Chao2
adapts it for quadrat presence-absence data.
 The Chao3 estimator combines aspects of Chao1 and Chao2, yielding a similar
estimate of 126.3 species.
 The estimators utilize parameters such as unique species counts (r1) and
doubletons (r2) to refine richness estimates.
 Overall, these estimators provide valuable tools for assessing biodiversity in
ecological studies.

Comparison of the estimators

Palmer (1990, 1991) evaluated eight estimators for species richness in 30 plots (0.1 ha
each) using 40 quadrats per plot. He found Jack2 to be the most accurate estimator,
while the Monod model had the highest correlation with true richness. Bootstrap was
deemed good by Palmer but poor by Colwell & Coddington.

Species-individual (area) based methods vs non-parametric methods

Area-based methods overestimate richness, while non-parametric methods

underestimate it, limiting landscape extrapolation.

Upscaling biodiversity: estimating the species-area relationship from small

samples

 The study tested 19 techniques for biodiversity upscaling using the GB

Countryside Survey 1999, covering 2,326 focal plant taxa.
 Upscaled species richness estimates varied significantly, ranging from 62 to
11,593 across different methods and datasets.
 The Shen and He model provided estimates within 10% of the true species
richness, while Hui's Occupancy Rank Curve had an average error of <20%.
 A hybrid method combining the Shen and He model with the Sizling model
achieved a mean relative error of 15.5% in predicting the species-area
relationship.

Some recent literature

The text discusses various methodologies for quantifying biodiversity, including species
richness estimators and the species-area relationship, highlighting significant studies
from 2001 to 2018. Key findings include the recovery of tropical rainforest ecosystems
after logging, emphasizing the importance of accurate biodiversity assessment.

Rarefaction method
  Park A has more tree species (s1) than Park B (s2) if s1 > s2, indicating higher
diversity.
 If Park A is larger but has fewer species than Park B, it suggests Park B is more
diverse.
 The study of a Malaysian rainforest shows that species density decreases with
larger grain sizes, affecting diversity measurements.
 The rarefaction method allows for comparing species diversity by sampling
smaller sizes from larger communities.
 The total number of trees in the Malaysian plot is 335,356, belonging to 814
species, highlighting significant biodiversity.
 Sampling scale profoundly impacts species diversity, with smaller scales yielding
higher species counts per unit area.

R implementation

 The hypergeometric distribution calculates the probability of sampling a specific

number of individuals from a population with known species abundances.
 For example, with a total abundance of 50 and species A having an abundance
of 10, the probabilities for sampling 0 to 8 individuals of species A are provided.
 Rarefaction methods estimate the expected number of species in a sample size,
assuming random sampling from a larger community.
 The expected number of species when drawing 30 individuals from a total of 42
in a sample of Yukon rodents is calculated using rarefaction.
 Research indicates that the impact of logging on forest diversity can be
temporary, with recovery starting as early as 8 years post-logging.
 However, some studies suggest that logging impacts can be long-lasting, with
irreversible effects observed over 150 years, especially when invasive species
dominate.
 The section emphasizes that rarefaction assumes random distribution of species
and that logging impacts should be assessed based on land base rather than
abundance.
 A coverage-based rarefaction method developed by Chao and Jost in 2012
offers a more appropriate comparison for ecosystems with differing diversity
levels.