Large scale patterns of biodiversity

Elements of biodiversity

EcologicalBiomesBioregionsLandscapesEcosystemsHabitatsNichesPopulations

GeneticPopulationsIndividualsChromosomesGenesNucleotides

OrganismalKingdomsPhylaClassOrderFamiliesGeneraSpeciesPopulationsIndividuals

The Ecological NicheGrinnell (1917) – the habitat or environment an organism can occupyElton (1927) – the role a species plays in the communityHutchison (1957) – quantify the abiotic and biotic dimensions of the niche

à an n-dimensional hypervolume

Which axes define the niche?

• abiotic conditions• resources• limiting factors

Condition/Resource Response Curves

Examples: temperature, pH, water

Intensity of Condition or Resource

Per

form

ance Reproduction

Growth

Survival

Survival not possible

Reproduction possible

Diversity and the niche

Elements of biodiversity

EcologicalBiomesBioregionsLandscapesEcosystemsHabitatsNichesPopulations

GeneticPopulationIndividualsChromosomesGenesNucleotides

OrganismalKingdomsPhylaClassOrderFamiliesGeneraSpeciesPopulationsIndividuals

WHY SPECIES RICHNESS?

PracticalExisting informationSurrogacy - it is related to other measures

more species meansmore genetic diversity - genes-popnsmore organismal diversity - ind�s-higher taxamore ecological diversity - niches habitats etc

Wide application

But what is a species and how many are there?

HOW MANY SPECIES ARE THERE?

We don�t know!

3 – 100,000 million species

BEST GUESS ca. 8.75 million Mora et al. 2011

total OceanEukaryoteAnimalia 7,770,000 2,150,000Fungi 611,000 5,320Plantae 298,000 16,600Chromista 27,500 7,400Protozoa 36,400 36,400ProkaryoteArchaea 455 1Bacteria 9,680 1,320

Patterns – three examples

Patterns and the environment

à Hypotheses

Issues and Ideas

Where do we find the most species?What patterns are there?

Where do we find the most species?What patterns are there?

How general is the latitudinal diversity gradient?

Meta-analysis

- a statistical technique to quantify overall effect from multiple studies

- weights studies by the sample size

Meta-analysis

- a statistical technique to quantify overall effect from multiple studies

- weights studies by the sample size

1.5

1.0

0.5

0

-0.5

-1

-1.5

Mea

n E

ffect

�

95%

con

fiden

ce in

terv

al

How general is the latitudinal diversity gradient?

Hillebrand (2004) - 600 gradients- local and regional scale- many organisms- many habitats

(terrestrial, marine, freshwater)- many realms

(New, Australasia, Eurafrica)

Assessed slope of line relating latitude and species richness

local and regional scale

local – small spatial extent, < 1 m – few km2

area depends on organisms- area within which species likely to interact

regional – large spatial extent, many km2, multiple habitats and communities

- area within which speciation and extinction operate- area that provides the pool of species that can over

time colonize a local community

local

local and regionalα-diversity, β-diversity, γ-diversity

How general is the latitudinal diversity gradient?

Hillebrand (2004)

Effect = slope of best line

Low Latitude HighEquator Pole

How general is the latitudinal diversity gradient?

Hillebrand (2004)

Wide range of taxonomic groups -

mammals, birds, fish, copepods

Terrestrial and marine > freshwater

New world > Australasia and Eurafrica

Conclusion - ubiquitous pattern

Patterns

Patterns (INDEPENDENT VARIABLE) and the environment (POSSIBLE EXPLANATORY VARIABLES)

à Hypotheses

Issues and Ideas

Climate – primary productivity (evapotranspiration)

ClimateWallace (1878)- the tropics have suffered fewer harsh events

+ Time - the tropics are older

Area

Why is there a latitudinal gradient in biodiversity?

Four general hypotheses

Null modelsMid-domain effect

Climate hypothesesSpecies – energy model

Historical hypothesesTime + perturbationTime and Area

Evolutionary hypothesesClimate stabilityEvolutionary speed theoryBiotic interactions hypothesis

Null models – the mid-domain effect (Colwell and Hurtt 1994)

Random set of pencil lengths with a constraint

+ bounded domain

Null models – the mid-domain effect (Colwell and Hurtt 1994)

N

S

Random set of species ranges

à ranges overlap more at centre than at limits

à more species in tropics

Climate hypotheses for diversity gradients –species – energy hypothesis

Climate sets limit to richnessclimate in tropics (temperature + precipitation) à higher NPPhigher NPP à more individualsmore individuals à more species with viable pop’ns

Historical hypotheses for diversity gradients

ExampleLake Baikal unglaciated lake 580spp benthic invertsGreat Slave Lake glaciated lake 4 spp

Time (Wallace 1878)

CatastrophesEg glaciation

Tropics

Temperate

Historical hypotheses for diversity gradients

Tropical environments are olderàmore time for diversity to accumulate

Tropics have a larger “area” when considering earth’s history

Time and Area Hypothesis

Larger areas = larger popns-----> less extinction

Larger areas = more barriers-------> more speciation

Evolutionary hypotheses for diversity gradients –

The tropicsa “cradle” for the generation of new taxaand/ora “museum” for the preservation of existing biodiversity”àgreater diversification (speciation – extinction)

Evolutionary hypotheses for diversity gradients –

Climate stability hypothesis

Tropics - predictable and mild à greater specialization/speciationà smaller niches à more species

High latitudes - unpredictable and severe à selection for flexibilityà larger niches and less speciationà less species per habitat

evolutionary hypotheses for diversity gradients –

evolutionary speed theory (Rohde 1992)

Evolutionary hypotheses for diversity gradients

Biotic Interactions hypotheses (Schemske 2002)

Biotic interactions (b) are a stronger selective force than abiotic factors (a) in the tropics à higher speciation

Patterns

Patterns (INDEPENDENT VARIABLE) and the environment (POSSIBLE EXPLANATORY VARIABLES)

à Hypotheses

Issues and Ideas

Productivity, history and area, climate stability and temp. covaryOur hypotheses generate similar predictions

Hypotheses

Mid-domain effectSpecies – energy modelTime – perturbationTime and AreaClimate stabilityEvolutionary speed theoryBiotic interactions

Some hypotheses are difficult to differentiate because they make the same predictions about speciation and extinction in the tropics and temperate regions

Spe

ciat

ion

Ext

inct

ion

x xx �x �� �� x � x � x

Issues and Ideas

Currie et al. 2004 – should consider mechanism

Vellend - hypotheses for patterns based on lower level processeswill proliferate

- should consider higher level processesselection = selective loss of species (extinction)ecological drift = chance loss of species (extinction)dispersal = movement of speciesspeciation

AdvancesSpecies – Energy Hypothesis (Currie et al. 2004)

i) higher NPP à more individualsii) more individuals à more species with viable pop’ns

i – true, but weak ii – true, but >> predicted

AdvancesEvolutionary Speed Hypothesis (Rohde 2002)

i) higher Temp à faster molecular evolutionii) faster molecular evolution à more speciation

i – supported eg Oppold et al. 2016 PNAS

ii ltd data, weak relationshipeg Dugo-Cota et al 2015 Global Ecol Biogeogr

# Spp. Latitudehigher nearer equator

lower nearer poles

Advances: phylogenetic approachesCardillo (1999)

So DIVERSIFICATION rate(speciation – extinction) is higher nearer the equator

9/10 pairs of birds clades-more southerly clade more species-rich

10/13 butterfly sister-group comparisons-more southerly clade more species-rich

!

Advances: phylogenetic approachesRoland et al. 2014 - 5020 mammals from 8 orders

Biodiversity and Area

Eg Lesser Antilles19 islands13 - 1510 km2

Species richness increases with island size

Alexander von Humboldt (1807)

Patterns

Lesser Antilles19 islands13 - 1510 km2

Log (Area)

Log

(Num

ber o

f spe

cies

)

Fig 19.8 from Stillings text

Slope Z ~ 0.3

Pattern on 61 islands off the coast of UK

z=0.32

r2 = 0.66

Log Area

Gaston and Blackburn 2000

Log

(Num

ber o

f spe

cies

)

Pattern on 59 Woodland "islands" for birds in UK

z=0.28

Eastern Wood, Surrey

r2 = 0.82

Log Area

Note- the data in these figures are independent

Log

(Num

ber o

f spe

cies

)

(Gaston & Blackburn, 2000)

Other species-area data can be nested

these data are not independent

Species –Area RelationshipsHypotheses/models and Processes

MacArthur and Wilson (1967) - Island biogeography theory

Processes - ?

Species area curves for birds and mammals

Forest mountaintops in the Great Basin of the western USA

Boreal birdsZ=0.165

Boreal mammalsZ=0.326

Q. What is different about birds and mammals?

Islands vs mainland isolation àdifferent islands more different than different areas of mainlandàgreater slope

Dispersal distance greater dispersal ability (of taxonomic group) à greater homogeneity

àtaxa with smaller slope

Latitude gradient in diversity and turnover

Factor mechanism

What could influence the slope (z)?

Meta-analysis: 794 Species Area Relationships –Result 1. Nested > IndependentResult 2. SAR on islands = SAR on mainlandResult 3 and 4 (see figs)

Slop

e (z

) of

SAR

Drakare et al. 2006

Applying species area curves to conservation

Change Proportion remaining

Log(Area)

Log(

#Spe

cies

)

a A

Log(S)

Log(s)

s/S = caz/cAz

= az/Az

= (a/A)z

WHAT SLOPE OR Z DO YOU USE?estimate from nested SARdepends on taxa – and their dispersal capacity (body size)depends on location (latitude) – due to latitudinal gradient

�…rarity is the attribute of vast numbers of species in all classes…�

Charles Darwin 1859

American moon moth Tiger longwing

Relative abundance of Lepidoptera in Rothamsted, UK

Types of species abundance diagrams

107

106

105

104

103

102

Most abundant least abundant

Geometric series - least equitable

Broken stick - most equitable

Log-normal

Rank

Log 1

0(ab

unda

nce)

Examples of lognormal distributions

lengths of words in phone conversationssize of sand grains on a beachlength of latency period for infectious diseases

Does ecology play any role in species-abundance relationships?

Ecological explanations are based on niche models1. the sequential breakage hypothesis - Sugiwara 1980

The

nich

e sp

ace

+hierarchical community organisation

Lognormal

Species 1 2 3 4

Ecological explanations 2. the broken stick - MacArthur 1957

The niche space

Broken-stick

Colonists arrive Resources are dividedRandomly at one time

Ecological explanations 3. the geometric series - Whittaker 1970

The niche space

+each successive spp takes a fixed fraction

geometric

Species abundance patterns

of many biological communities fit the log-normal dstribution

may be explained by niche apportionment models (see Wikipedia entry)

BUT

Are hypotheses linking niche organisation and species abundance testable?

What two pieces of information would you need?

What predictions can be generated from these pieces of information?

Testing the connection between hierarchical niche organisation and species abundance

base upper

middle 1/3 upper 1/3

inner outer inner outer

Testing the connection between hierarchical niche organisation and species abundance

Dendogram of niche similarity

Corresponding process by whichniche space is split

Sugihara et al 2003 PNAS 100 5246-5251

Testing the connection between hierarchical niche organisation and species abundance

Prediction 1Spp highly nested should be less abundant than spp on undivided branches

Dendogram of niche similarity

Niche splitting

1

2

3

What is the expected abundance distribution?

Prediction 2 Even distributions are associated with even niche dendograms

Species rankRela

tive

abu

ndan

ceTesting the connection between hierarchical niche organisation and species abundance

Testing the connection between hierarchical niche organisation and species abundance

Sugihara et al. 2003DATA: 11 communities with niche dendograms and

species abundance data

Waterfowl, Fish, Amphibians, Lizards, Songbirds

Testing the connection between hierarchical niche organisation and species abundance

Prediction 1Abundance declines with increasing nestedness in 10 of 11

Overall p = 0.012

Testing the connection between hierarchical niche organisation and species abundance

Empirical data support generality of Lognormal distributions

Hierarchical niche models such as sequential breakage hypothesis have some support

Opponents argue1) Species abundance is inappropriate if species differ in mass 2) resources will be divided amongst species from many taxa not only those from a single taxonomic group

So WHY/HOW we have/get lognormal distributions is much debated

Three PatternsLatitudinal gradients in diversitySpecies area relationshipsSpecies abundance distibutions

Many models

With some thought to mechanism null modelsfirst order processes

– Speciation, dispersal, selection and ecological drift

References

Mittelbach (2018) Commnity Ecology – Ch 2.

Worm and Tittensor (2018) A Theory of Global Biodiversity. Princeton University Press

Currie et al. (2004). Predictions and tests of climate-based hypotheses of broad-scale variation in taxonomic richness. Ecol Lett 7: 1121–1134

Schemske and Mittelbach (2017). “Latitudinal Gradients in Species Diversity”: Reflections on Pianka’s 1966 Article and a Look Forward. Am Nat 189: 599-603

Eric Pianka typing his thesis

Selected readings

Latitudinal gradients or Species abundance distributions

Species-area relationships will be revisited during discussions of Neutrality