Isotopes and Animal Movement - CSICebd.csic.es/IsotopeCourse/Conferences/HOBSON3.pdfIsotopes and Animal Movement Seville, 2014 Keith A. Hobson Primary Goals of Migration Research Evolution

Isotopes and Animal Movement Seville, 2014

Keith A. Hobson

Primary Goals of Migration

Research

Evolution and Ecology.

Conservation and Management.

Movement of Contaminants and

Disease.

>3M banded,

20% recovered.

>0.8M banded ,

.01% recovered

Problems with extrinsic markers

Organism often needs to be recovered

Expensive

Body size requirements

Biased to original marked population

Biogeochemical processes result in

isotopic patterns or “isoscapes” …

Isoscapes?

Latitudinal/altitudinal gradients (dD, d13C)

Terrestrial-marine (d13C, δ15N d34S) Inshore-offshore (d13C , d34S, d15N) C-3 vs. C-4, CAM (d13C, δD) Xeric vs. Mesic (d13C, d15N)

Surficial geology (Sr, Pb, others)

First isotope applications

Killingley (1980) – δ18O barnacles

“you are what you swim through”

Killingley and Lutcavage (1983)

δ13C and δ18O

From Schell et al.2002

Bowhead whales

Quillfeldt et al. (2010) BES.

Latest marine

isoscapes ….

McMahon et al (2013)

Limnol. Oceanogr. 58:697-714

Predicted locations using correlates

between fish scale collagen (δ13C) and

SST

Mackenzie et al. (2011), Nature.

Mackenzie et al. (2011), Nature

Two essential components:

1. Transfer function (∆) between isoscape

and tissue.

2. Turnover rate of the tissue.

The isotopic clock and movement

Biome 2

Biome 1

dX

(per

mil)

Time (days)

t2 t2’ t1 t1’

Tissue 2

Tissue 1

Heavy isotopes tend not to

fractionate trophically

Stewart et al. MMS 19:806-818

Promise with

Hg isotopes

Point et al. 2011 Nature Geoscience

Strontium (bedrock model)

Rachel Barnett-Johnson

Modeling Sr isotopes in watersheds:

Other applications do not require

use of discrimination factors

Assignment to “bins”

Assignment to “biomes”

Forensic tracing of African ivory ….

δ13C, δ15N, δ87Sr, δ204Pb

Vogel et al. (1990); van der Merwe et al (1990)

Wintering habitat determines arrival time

on breeding grounds

Marra et al. (Science 1998)

Breakthrough with water isotopes

Latest growing-season δ2Hp

Wassenaar, IAEA.

Altitudinal gradients

are recorded in

hummingbird

feathers:

Hobson et al. Oecologia 136:302-308

The feather isotopes follow large

scale trajectories in precip δD

a

Elevation (m)

0 500 1000 1500 2000 2500 3000 3500

dD

(o/ o

o)

-130

-120

-110

-100

-90

-80

-70

-60

b

d1

3C

(o/ o

o)

-28

-27

-26

-25

-24

-23

-22

a

Expected (Ecuador)

Expected (global)

δDf = -25.6 + E(-0.014) -25 o/oo

δDf = -22 + E(0.0224) -25 o/oo

Ecuador:

Global:

More on transfer functions

involving the water isotopes.

Derived from known-origin tissue transects.

Derived experimentally.

Based on various time series (mean annual,

mean growing season).

Can be year-specific or based on long-term

GNIP records.

Single Top Model:

δ2Hf = Int. + δ2Hp + Migratory Guild +

Foraging Substrate + Guild*Substrate

Explains ~83% of the variance in δ2Hf

Dragonfly wing isoscape

Hobson et al. 2012 MEE 3:766-772

Dragonfly fallout on Maldives

Bats …

Cryan et al. 2004. J. Mammal.

Creating basemaps through direct

spatial sampling.

Two populations, one long distance

journey ….

Previously, tagging was used:

Monarchs can be “grown”

anywhere!

80 elementary schools recruited

throughout the range …..

Transfer function derived

experimentally ….

Hobson et al. 1999 Oecologia 120:397-404

Hobson et al. 1999 Oecologia 120:397-404

Transfer function derived from

basemap ….

Hobson et al. 1999 Oecologia 120:397-404

Docykx et al. Ecol. Applic., in press

Origins: 50% of the population is

produced in the US cornblelt:

Assignment depictions

“Map lookup” approach.

Deriving probability of origin

Likelihoods

Bayesian models

• Constrained with priors and other information.

“Leapfrog” migration revealed ..

Kelly et al. (Oecologia 2002)

Other isotopic delineations of

population structure …

Rubenstein et al. (Science 2002)

Bearhop et al. Science 2005

Categorical assignment

Gen

eti

cd

ista

nc

e:

FS

T/(

1-F

ST)

Geographic distance (km)M

igra

tory

dista

nce (k

m)

Origins of Woodpigeons killed in

France

Long-range migrant (winter in southern range of Europe)

Medium-range migrant (winter in France)

sedentary

Several distinct populations with

specific migratory traits

dD precip

-200 -150 -100 -50 0 50 100

Pro

babili

ty D

ensity

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

Probability Density0.00 - 0.100.11 - 0.200.21 - 0.300.31 - 0.400.41 - 0.500.51 - 0.600.61 - 0.700.71 - 0.800.81 - 0.900.91 - 1.00

Probability Density0.00 - 0.100.11 - 0.200.21 - 0.300.31 - 0.400.41 - 0.500.51 - 0.600.61 - 0.700.71 - 0.800.81 - 0.900.91 - 1.00

Probability Density0.00 - 0.100.11 - 0.200.21 - 0.300.31 - 0.400.41 - 0.500.51 - 0.600.61 - 0.700.71 - 0.800.81 - 0.900.91 - 1.00

A)

B)

Conditional Probability

0.000000 - 0.000001

0.000002 - 0.000002

0.000003 - 0.000003

0.000004 - 0.000015

0.000016 - 0.000064

0.000065 - 0.000070

0.000071 - 0.000625

0.000626 - 0.000685

0.000686 - 0.005397

0.005398 - 0.006830

A

B

C

D

E

F

G

H

I

J

Conditional Probability

0.000000 - 0.000001

0.000002 - 0.000002

0.000003 - 0.000003

0.000004 - 0.000015

0.000016 - 0.000064

0.000065 - 0.000070

0.000071 - 0.000625

0.000626 - 0.000685

0.000686 - 0.005397

0.005398 - 0.006830

Conditional Probability

0.000000 - 0.000001

0.000002 - 0.000002

0.000003 - 0.000003

0.000004 - 0.000015

0.000016 - 0.000064

0.000065 - 0.000070

0.000071 - 0.000625

0.000626 - 0.000685

0.000686 - 0.005397

0.005398 - 0.006830

A

B

C

D

E

F

G

H

I

J

Origins of Woodcock taken in

Spain

Applying prior probabilities

Royle and Rubenstein

Combining genetics and δD

Boulet et al. (Ornithol Monogr)

Modeling isoscapes and origins

using multiple isotopes

Assumes a multivariate normal distribution

All isotopes are orthogonal

Africa has a rich history of stable

isotope applications

Major overwintering site for

Palearctic-Aftrotopical migrants

Plant isotope distribution models

Plant isotope map

based on predicted C3

vs. C4 plants

Still and Powell (2010)

Feather δ13C isoscape

A feather δ15N isoscape Based on Craine et al. (2009, New Phytologist)

Modeled MAT, MAP with foliar δ15N

Modeled MAT, MAP with foliar and soil

δ15N

Feather δ15N isoscape

Feather δ2H isoscape

Combining δ13C, δ15N, δ2H feather

isoscapes

Pied Flycatcher

Collared Flycatcher

Data from Hjernquist

A multi-isotope (δ2H,δ13C,δ15N)

probability surface

Some other considerations and

future directions….

Several controlled studies for δ15N,

δ13C but few for δ2H, δ18O ….

δ15N, δ13C …

δ2H, δ18O …

Using a wind tunnel and isotopic

dietary shifts to mimic migration

Max Planck Institute for Ornithology

Advances in water isoscapes

Multiple sources of O and H

Pietsch et al. 2011 PLoS ONE

Pietsch et al. 2011 PLoS ONE

Pietsch et al. 2011 PLoS ONE

Oxygen?

20-35%

evaporative loss

Human Hair

Ehleringer et al. (PNAS 2008)

Meteoric relationship preserved in

Monarch Butterflies

Manitoba

Indiana

Florida

y = 7.951x - 123.948

R2 = 0.936

-140.0

-130.0

-120.0

-110.0

-100.0

-90.0

-80.0

-70.0

-60.0

-2.5 -1.5 -0.5 0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5

dO18

smow(Pyrolysis)

dD

sm

ow(P

yro

lysis

)

Feather 18

O (per mil)

6 8 10 12 14 16 18 20 22 24

Feath

er

2H

(per

mil)

-140

-120

-100

-80

-60

-40

-20

Y = -145.7 + 4.8X

r2 = 0.34

Trace element/heavy isotope

Compound specific

Mass spectrometry