A Global Test of the Diversity-Productivity and Diversity-Stability Hypotheses Australia Hautier, Borer, Lind, Seabloom, Anderson, Harpole, Hector, Hillebrand,

A Global Test of the Diversity-Productivity and

Diversity-Stability Hypotheses

Australia

Hautier, Borer, Lind, Seabloom, Anderson, Harpole, Hector, Hillebrand, Mac Dougall, Meyer, Prober, Ries,

Stevens + Anyone interested

Along the natural diversity gradient represented by 39 herbaceous-dominated plant ecosystems on five continents, plant cover at the species and ecosystem organizational levels was stabilized with increasing plant species richness, as indicated by a decrease of the temporal variability. On average, the plots with the highest diversity were about 40% more stable in time than the plots with the lowest diversity.

Fertilization significantly reduced species richness; on average 1.2 species (SEM = 0.4) were lost with fertilization relative to the controls. Along the reduced diversity gradient plant cover at all organizational levels was not related to species richness. The results support the diversity-stability hypothesis that greater diversity leads to greater ecosystem stability in natural systems and demonstrate that fertilization has negative effect on ecosystem stability.

Global Drivers of Loss of Biodiversity with Eutrophication

Hautier, Hector, Anderson, Grace, Seabloom, Borer, Lind, Stevens, Harpole + Anyone interested

Australia

Potential mechanisms explaining the loss of plant diversity with fertilization

• General increase in the strength of competition – aboveground for light and belowground for nutrients (Grime 1973)

• Increase in the strength of aboveground competition for light only (Newman 1973, Tilman 1982)

• Acidification (Silvertown et al. 2006)

• Accumulation of plant litter (Berendse 1999, Foster et al. 1998, Lamb 2008, Tilman 1993)

Using AMOS we made a Structural Equation Modeling to explain the changes in proportional species richness between the control and NPK+ plots. We found that once we controlled for the covariate “elevation”, proportional richness loss is greatest (1) where light reduction is greatest, and (2) where pretreatment evenness was low. Dead biomass had no effect on the change in species richness as well as other mechanisms such as acidification of competition for nutrients.

Elevation Eveness Proportion of total biomass change

Proportion of dead biomass change

Proportion of change in light

Proportion of change in richness

Eurasian herbaceous species are better competitors for nutrients away from home

• 17 species (8 grasses & 9 forbs); 25 sites: 3 UK, 3 EU, 1 China (native sites), 16 US, 1 CA and 1 AU

• 3 years of treatments• Ran separate analyses for introduced sites with multiple study species or just one

species• Analyses: Population level: LMEM with temporal autocorrelations-extracted

parameter estimates; Community level (only included 10 sites that shared 5 or more study species, 5 in native and 5 in introduced): PERMANOVA and tested study species associations (positive, negative and neutral) pre- and post-treatment.

Nutshell: Eurasia herbaceous species are stronger competitors in their away range and consumers likely enable their persistence in native communities.

• Responses (relative cover) to nutrient additions and consumer treatments differ home and away, and depending on life-form (grass or forb)

• Nitrogen and phosphorus increase away; no change home• Fencing: no change or increase (when nutrients added) at away sites; decrease at

home• Grass spp. Interactions at sites with 5 or more study species tended to be positive

at home but negative away. Where forbs had more negative interactions at home but positive away.

Relative importance of deterministic vs. stochastic community assembly increases with increasing productivity (experimental)

Davies, Melbourne, Chase, NCEAS working group, all data contributors and anyone who is interested.

What we see:• By year three, increasing productivity created communities that deviate further from

the null than communities (are more nichey) in which productivity was not increased. • Further, the effect varies with baseline productivity: at low productivity sites,

increasing productivity does not change community structure, while at productive sites, increasing productivity creates communities that are more deterministic.

• Related: increasing productivity reduced beta diversity in some treatments: NPKOther thoughts:• We are hoping that the results will strengthen further in year four.• We plan a second panel of graphs that include some environmental variables like

temperature and rainfall that might help further explain what is going on.• It would make sense to report the effects of nutrient addition on alpha and gamma

diversity -- an option would be to have this paper follow up the paper that reports these results.

Working title:

N-0

.4-0

.20.

00.

2

Yea

r 1

P K NP NK PK NPK-0

.4-0

.20.

00.

2

Yea

r 2

-0.4

-0.2

0.0

0.2

Yea

r 3

100 300 500 100 300 500 100 300 500 100 300 500 100 300 500 100 300 500 100 300 500

Productivity (live mass in year 1)

Nul

l dev

iatio

n

Green=controls, Black=treatments, x-axis= observed productivity in the first year, pre treatment

Dominant plant species drive the ecosystem response to variation in

resource availability across a precipitation gradient

• Kimberly La Pierre, Dana Blumenthal, Cynthia Brown, Julia Klein, and Melinda Smith

• Submission in Summer 2012

Study Sites

TGP

SGS

MIX

SGS: Shortgrass Steppe (Shortgrass LTER)MIX: Mixed-grass Prairie (Saline Experimental Range)TGP: Tallgrass Prairie (Konza Prairie LTER)

Plant trait responses to chronic nutrient additions

• Kimberly La Pierre and Melinda Smith

• Data collected and analyzed

• Writing in Summer 2012

• Submission in Fall 2012

PC1 (32.77%)

-2 -1 0 1 2 3 4

PC

2 (1

9.2

9%

)

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

green leaves

biomass

SLAthickness

Natural Dominants

Altered Dominants

Control

Nitrogen

Phosphorus

Nitrogen and Phosphorus

short-term

Nutrient additions alter invertebrate community structure and feeding

strategy

• Kimberly La Pierre and Melinda Smith

• Data collected and analyzed

• Writing in Fall/Winter 2012

• Submission in Spring 2012

PN

plant biomass

invertebrate herbivores

K

invertebrate predators

invertebrate parasitoids

TGP

1

3 2 4

Χ2=5.480, df=15, p=0.987, RFI=0.874

0.522

0.311

0.502

(c) Per Capita Rate of Herbivory

TGP MIX SGS

Rat

e of

her

bivo

ry(d

amag

e /

chew

ing

herb

ivor

e)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14ControlNitrogen

a

b

Phylogenetic, functional, and plant trait diversity: Examining potential drivers of

primary production in herbaceous systems

• Network-wide opt-in manuscript

• Data collection and analysis in 2012

• Write in Spring 2012

NutNicheA global test of niche destruction and biodiversity loss

• Lead: Stan Harpole (opt-in)

• Evidence for multiple nutrient limitation

• Addition of nutrients to remove limitation = niche destruction

• Loss of species with greater numbers of added nutrients

• Link to other MS’s:

– Hautier, Grace, et al.

– Fay et al.

1/2

NutNiche•general pattern, but site variation: not all sites show predicted response

•mostly multi/co-limitation (~63%); super-additive (~45%)

•mostly not sampling effect (~80% sites multiple additions > single)

•different nutrients important for different responses:

– max richness effect ≠ max biomass effect

•covariates, not much, so far

•Status: rough draft

•To do:

– composition response—get at species tradeoff assumption?

– using max trt year (responses over time appear consistent but increase with time)

– other?

2/2

N-Gradientfunctional response, critical loads

• Lead: Stan Harpole (opt-in)

• CBGB, CDCR, SEDG N-gradients

– 0,1,5,10 g m-2 yr-1

• General N10 patterns

• Status: need to write up soon!

• NutNet add-ons:

– N-dep collectors

– N1 treatments

1/1

Global EcologySS 20: Global Environmental Challenges Require Global

Ecological Research– MS for Frontiers

• status: some text, outline for ESA panel

• to do: synthesize panel results

– Meeting Report (like White et al. Bio Lett)

– SeSynC proposal

• status: draft

• to do: develop sociology

1/1

Organization/Network

Biosphere 2

EPA

EREN

FunDivEurope, Biodepth-Jena

LTER

NEON

NSF Dimensions of Diversity

Numerical Terradynamic Simulation GroupNutrient Network

Phenology Network

TraitNet, BioMERGE

ZEN seagrass network

GLEON

(Opt-out) Answering one of the 3 original NutNet questions:

Under what conditions do grazers or fertilization control plant diversity and productivity?

This paper’s goal is to examine vertebrate herbivory x nutrient supply on •plant diversity •relative abundance (evenness)•net primary production

within the context of site and regional drivers • climate• soils• regional productivity • regional diversity

NUTRIENTS AND HERBIVORES CONTROL PLANT DIVERSITY AND FUNCTION IN A GLOBAL GRASSLAND EXPERIMENT

Results1. Fertilization: species richness declined and live biomass increased (p=0.0001 and

p=0.001, respectively); 2. Fences: no overall effect on plot-scale richness, evenness, or biomass3. Inside fences, NPP declines through time, but this effect is counteracted with

fertilization4. Both Bray-Curtis & Jaccard metrics show that species in plots turnover in response to

both fertilization and fencing (will also test contingency on site diversity)

We will also look at:1. the association between grassland responses to fencing and fertilization among and

within sites (log-ratios)2. the relationship b/w turnover and change in biomass 3. global-scale predictors (contingencies) – best models of change in richness,

evenness, and biomass as a function of soils, climate, and site-level diversity, evenness, NPP (test of Gruner/Bolker model of additivity; Hillebrand contingencies model)

Next steps: re-run with current data; make final figures; hone message; write discussion

What are the biogeographic patterns of species invasions?

Biological Invasions (Seabloom et al.)

• Mainly focuses on observational data • Rejected from PNAS April 2012 – positive reviews but there were concerns about site

selection criteria and land use history (e.g., agriculture and grazing)• Currently preparing for submission to Ecology or Global Change Biology (Other

suggestions welcome)• Main changes will be adding in land use and grazing data from Cini and Suzanne’s

survey, adding in site selection survey data, and removing the bimodal graph.

• Nutrient addition caused a loss of 0.6 native species per year but had no effect on exotic diversity

• Nutrient addition caused exotic cover to increase 6% per year but had no effect on native cover

• Have had first round of reviews by all authors with no major hiccups from anyone

• Still a bit undecided about target journal

Are native and exotic species functionally distinct? Biological Invasions (Seabloom et al.)

What are the effects of fertilization and herbivory on spatial and temporal turnover of species

composition

Damschen , Gruner, Hillebrand, Lind, Wragg, Wright, Yang

Adler, Bakker, Cavender-Barres, Dev, Orrock, Sullivan

Three motivating questions

NF Y1

NF Y3

C Y1

C Y3

Spatial heterogeneity – mean similarity within treated plots relative to mean similarity within control plots

Spatial turnover– mean similarity between treated plots and control plots relative to mean similarity within control plots

Temporal Turnover– mean similarity between treated plots in Y1 and Y3 control plots relative to mean similarity within control plots in Y1 and Y3

t-values (>2 ~significant)

Heterogeneity within treatmentsLog Ratio Fert Fence Fert*FenceJaccard Y2 . . 2.632Bray Y2 . . .Jaccard Y3 . . .Bray Y3 . . .

Spatial Turnover between treatmentsJaccard Y2 . . .Bray Y2 3.298 . .Jaccard Y3 . . .Bray Y3 2.149 . .

Temporal turnover between treatments

Jaccard Y2 2.498 . .Bray Y2 . .Jaccard Y3 - -Bray Y3 . . 2.32

To do

• Re-run with this year’s data– Are lack of effects in Y3 due to limited sample size

• Test important question: What controls between site variability

• Predictor variables:– Site fertility (Y0 mean productivity)– Site sensitivity (Log response of productivity to trts)– Species Pool – Evenness, site α

Grassland Soil Stoichiometry at the Global Scale

• Currently working on a manuscript draft explaining the abundance and stoichiometry of soil organic matter in relation to climate and vegetation using NutNet observational data

The distribution of soil stoichiometric ratios is an order of magnitude larger than previously reported Redfield-like ratios for soil; this is largely due to large variation in phosphorous as reflected in C:P and N:P ratios

C and N are highly correlated with little relationship between C and P and N and P despite a previous meta-analysis reporting these relationships

• In addition, we see strong relationships between climate (temperature and precipitation) with SOM and very weak relationships between productivity or diversity

• These different factors may make it difficult to apply cornerstones of ecological stoichiometry like the Redfield ratio to grassland soils at the global scale

species response to treatments: competition-defense

vgrowth-defense

32

Eric LindElizabeth BorerEric SeabloomPeter AdlerJonathan D. BakkerDana BlumenthalMick CrawleyKendi DaviesJennifer FirnDan Gruner

Stan HarpoleYann HautierHelmut HillebrandJohannes KnopsBrett MelbourneBrent MortensenAnita C. RischMartin SchuetzCarly StevensPeter Wragg

33

34

Effects of primary production and producer diversity on consumer biomass

Lind, Borer, Kay, Wolkovich, Wright, Gruner, Yang, LaPierre, others….

Does primary productivity predict "secondary productivity"?

Native-Exotic Richness Relationship (NERR)

• 32 Nutnet sites: ‘native’ grasslands from four regions (Australia, Central, Intermountain, and Pacific.

• Four sections:

• Scale-dependent relationships between native and exotic richness (four scales: subplot 1m2, plot, site, and region).

• Drivers of NR and ER: independent or interactive? [Jenn Firn]

• Scale-dependent heterogeneity (Davies et al. 2005, 2007; Melbourne 2006): the impacts of CV on the NERR slope.

• Species associations: are species pairs more positively or negatively associated with each other, than expected by chance (Fridley et la. 2004, 2007)? [Joe Bennett]

Australia

Pacific

Central

Intermountain

f

Native richness (fine-scale 1 m-2)

Exoti

c ric

hnes

s (fi

ne-s

cale

1 m

-2) fr

eque

ncy

NS

Neg

ative

slo

pes

Posi

tive

slop

es

Fixed effects

F values (df as subscript), P value

INT_rich F1, 460=0.68, P< 0.50

MAP F1, 8=1.34, P< 0.30

MAT F1, 8=6.14, P< 0.04

MAP_VAR F1, 8=13.90, P< 0.006

Temp_VAR F1, 8=0.09, P< 0.80

Int_rich MAP_VAR F1, 460=3.15, P< 0.08

MAP MAT F1, 8=6.80, P< 0.04

MAP_VAR Temp_Var F1, 8=3.55, P< 1.0

Plot-level native richness [random intercepts due to site (σ

2 site), and block within site (σ 2

block). Full model AICc =2261; simplified model with StepAIC

function AICc=2247]

Fixed effects

F values (df as subscript), P value

Nat_rich F1, 456=3.21, P< 0.08MAP F1, 12=0.93, P< 0.40MAP_VAR F1, 12=5.68, P< 0.04MAT F1, 12=5.47, P< 0.04Temp_VAR F1, 12=3.48, P< 0.09Nat_rich MAP F1, 456=0.00006, P< 1.0Nat_rich MAP_VAR F1, 456=12.94, P< 0.0004

Nat_rich Temp_VAR F1, 456=0.07, P< 0.80MAP MAT F1, 12=7.31, P< 0.02Nat_rich x MAP x MAT F1, 456=13.74, P< 0.0002

Plot-level introduced richness {random intercepts due to region (σ

2 region), site within region

(σ 2 site), and block within site within region (σ

2 plot). Full model

AICc =1593; simplified model with StepAIC function AICc=1584

Fixed effects

F values (df as subscript), P value

MAP F1, 21=0.91, P< 0.40

MAT F1, 21=2.72, P< 0.20

Temp_VAR F1, 21=12.75, P< 0.002

MAP x MAT F1, 21=2.05, P< 0.20

MAP Temp_VAR F1, 21=3.98, P< 0.06

Site-level native richness[random intercepts due to region (σ

2 region), and site within block (σ 2 site). Full

model AICc =273; simplified model with StepAIC function AICc=256]

Fixed effects

F values (df as subscript), P value

NAT_rich F1, 14=1.83, P< 0.20

MAP F1, 14=0.70, P< 0.50

MAT F1, 14=2.86, P< 0.20

MAP_VAR F1, 14=8.5, P< 0.01

Temp_VAR F1, 14=0.80, P< 0.40

MAP x MAT F1, 14=4.80, P< 0.05

Nat_rich Temp_VAR F1, 14=2.47, P< 0.15

Nat_rich MAP F1, 14=0.24, P< 0.70

Nat_rich MAT F1, 14=0.05, P< 0.90

Nat_rich MAP x MAT F1, 14=0.61, P< 0.50

Nat_rich MAP x Temp_VAR F1, 14=2.98, P< 0.20

Site-level introduced richness[random intercepts due to region (σ

2 region), and site within region (σ 2

site). Full model AICc =236; simplified model with StepAIC function

AICc=227]

Prop

ortio

n of

sig

nific

ant c

orre

latio

ns

****

**

A. Site-level species associations - natives vs. exotics

B. Plot-level species associations - natives vs. exotics

** contrasts the proportion of significant asstns that are positive vs those that are negative

A global study of below-ground allocation patterns in grasslands

• Proportional root allocation should decline when growth is limited by above-ground resources (e.g. light), and increase when growth is limited by below-ground resources (e.g. water and nutrients).

• Extracted roots from 21 sites (8 more expected this summer)

• Plan to do a full analysis and write much of a paper this week (hope that the additional sites do not change the main results substantially)

Specific hypotheses• Proportional root allocation will decline as total

production increases– With increasing number of nutrients added– Inside herbivore exclosures (high litter accumulation

leading to light limitation)

• Flexibility in allocation will decline with decreasing soil moisture availability (interaction between nutrient addition and site-level mean rainfall)

Manuscript 1: Strong abiotic controls over seed removal in a continent-wide study

Status: analyses completed, ms drafted, submit after current round of comment by co-authors (Ecology is target journal)

400 600 800

0.0

0.2

0.4

0.6

0.8

1.0

Annual Evapotranspiration (mm)

400 600 800 1000 1200

0.0

0.2

0.4

0.6

0.8

1.0

Mean Annual Precipitation (mm)

Pro

po

rtio

n o

f S

ee

ds

Re

mo

ved

Authors: Orrock, Brudvig, Firn, MacDougall, Yang, Melbourne, Baker, Bar-Massada, Borer, Crawley, Damschen, Davies, Gruner, Kay, Lind, McCulley, Seabloom.

Manuscript 2: Contingency in consumer-mediated invasion

Status: introduction drafted, analyses in progress. Pending completion of analyses, send discussion draft to co-authors by September, submit ms by November

Objective: examine how the role of consumers in affecting exotic plant abundance varies as a function of abiotic constraints (e.g. temperature, precipitation)

Authors: Orrock, Firn, Bakker, Blumenthal, Borer, Brown, Brudvig, Buckley, Chu, Cleland, Cottingham, Crawley, Damschen, Davies, Firn, Frater, Gruner, Kay, Kirkman, Klein, Knops, LaPierre, Leakey, Li, Lind, MacDougall, McCulley, Melbourne, Moore, Morgan, Nelson, Prober, Seabloom, Stevens, Wolkovich, Wright, Yang

Manuscript 3: Biofuel potential of three semi-natural grasslands

Objective: examine the energy produced by plant biomass derived from grasslands in CA, MO, and SC

Authors: Orrock, Watling, Brudvig, Damschen, Borer, Seabloom, Baker

Status: analyses need to be re-visited, ms drafted, revise discussion, send to co-authors, then submit (Ecological Applications is target journal)

Working Title: What limits productivity in grasslands worldwide?Co-Leads: Phil Fay, Suzanne Prober, John Knops (opt-out paper)

Hypotheses:1.Nitrogen is a globally significant limitation on productivity at most sites2.Nitrogen response of Productivity is independent of climate3.Phosphorus limitation prominent on old, weathered soils4.Additional nutrient (co-)limitation in some cases/places

ANPP: 40 sites

Site

frue

.ch

trel

.us

hall.

ushe

ro.u

kaz

i.cn

bogo

ng.a

ucb

gb.u

sco

wi.c

ase

reng

.tzty

so.u

sko

nz.u

ssp

in.u

sel

liot.u

ste

mpl

e.us

unc.

ushn

vr.u

sro

ok.u

ksm

ith.u

sbu

rraw

an.

sier

.us

sedg

.us

mcl

a.us

bnch

.us

salin

e.us

valm

.ch

hopl

.us

look

.us

lanc

aste

rcd

pt.u

sba

rta.

uski

ny.a

usa

ge.u

sbl

dr.u

ssa

va.u

ssh

ps.u

scd

cr.u

sm

tca.

ausg

s.us

hart

.us

sevi

.us

AN

PP

(g

m-2

s-1

y-1

)

0

200

400

600

800

1000

1200

1400

1600

1800

Mean all yearsyear 1year 2year 3

Nutrient main effects 40 sites, all avail. years

Nutrients added

C N P NP K NK PK NPK

AN

PP

(g

m-2

s-1

y-1

)

0

100

200

300

400

500

600

ab ab

c

cc

ab

d

Nutrient response to rainfall and temperatureMAT.N Chi P=0.003MAP.N Chi P<0.001MAP.P Chi P<0.001MAT.K Chi P= 0.009

Random model: site_code + site_code.year_trt + site_code.block + site_code.year_trt.block + site_code.plot + plot.syVariance structures: differing variance between sites; differing variance between pots within sites each year

• Currently: Fine-tuning hypotheses and statistical models• Investigating covariance of nutrient effects with climate/grazing intensity

variables.• Need to investigate soil property effects.

• Currently: Fine-tuning hypotheses and statistical models• Investigating covariance of nutrient effects with climate/grazing intensity

variables.• Need to investigate soil property effects.

Site

AN

PP

(g

m-2

s-1

)

0

200

400

600

800

1000

1200

1400

Control+N

22 Sites with N-main effect significant

20 - 84% increase

Site

hero

.uk

hall.

ussp

in.u

sbu

rraw

an.

unc.

usro

ok.u

khn

vr.u

sbn

ch.u

ssm

ith.u

sla

ncas

ter

sier

.us

look

.us

valm

.ch

salin

e.us

hopl

.us

sage

.us

bldr

.us

shps

.us

cdcr

.us

hart.

ussa

va.u

sm

tca.

au

AN

PP

(g

m-2

s-1

)

0

200

400

600

800

1000

1200

Control+P+K+PK

22 Sites withP or PxK effect significant

P, P x K: -28 to +184%

Nutrient response to rainfall

MAP*N*P P=0.005, adjusted for grazing coefficientRandom model: site_code + site_code.year_trt + site_code.block + site_code.year_trt.block + site_code.plot + plot.syVariance structures: differing variance between sites; differing variance between pots within sites each year

History of Grazing & Cultivation Survey• Surveys completed for 39 sites• Information from survey

– Whether there is grazing currently & when started– Whether or not & when site was cultivated– Management activities w/in plots– What herbivores present & relative abundance in

evolutionary, ecological, recent historic times– Timing of influential climatic events– What is the level of current and recent historic grazing

(nil, low, medium, high)

Evolutionary History of Vertebrate Grazing and Plant Invasions

• What is the role played by evolutionary histories of grazing on the relative success of native and introduced plant species?

• Is recovery from exotic invasion with grazing exclusion influenced by the evolutionary history of grazing?