carvalho et al. climate change & algal blooms
TRANSCRIPT
How does climate change affect the response of
cyanobacteria to nutrients?Laurence Carvalho & Stephen ThackerayRita Adrian, Orlane Anneville, Meryem Beklioglu, Hannah Cromie, Seyda Erdogan,Marko Jarvinen, Stephen Maberly, Yvonne McElarney, Jannicke Moe, GiuseppeMorabito, Peeter Nõges, Tiina Nõges, Jessica Richardson, Nico Salmaso, TomShatwell & Helen Woods
Response: Cyanobacteria: mean summer biovolume
Stressors• Nutrient stress: mean Spring TP
• Hydrological stress: summer rainfall
• Temperature stress: mean summer temperature
This study’s perspective
8 countries
26 lakes (min. 10 years data)
705 lake-years
• Generalized Linear Mixed Modelling (GLMM)
• All data transformed (Box-Cox) and centred
• Lake and year included as random effects – slope and
intercept allowed to vary by lake
Analytical Method: GLMM
Stage 1: Examine responses to single stressors
Stage 2: Examine responses to single stressors by lake type
(interaction model & data subset)
Stage 3: Examine responses to two stressors and their interaction
Lake types considered as fixed categorical effects: e.g.
• Trophic Type (oligo-meso, eutrophic)
• Residence Type (short, medium, long)
• Mixing Type (mixed, stratifying)
Response to TP – all lakes
Spring TP(transformed and centred)
Cya
no
bac
teri
a b
iovo
lum
e
(tra
nsf
orm
ed a
nd
cen
tred
)
P<0.01 **
fitted effect to TP
explained 7% of variation
in cyanobacteria
Highly significant effect
Response to TP – by lake
Cyan vs TP
Pearson's r
Fre
quency
-0.5 0.0 0.5
02
46
8
Cyan vs TempSu
Pearson's r
Fre
quency
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
02
46
810
Cyan vs Precip
Pearson's r
Fre
quency
-0.8 -0.4 0.0 0.2 0.4 0.6
02
46
8
Response to TP often weak but generally positive
Depends on TP gradient in time series and where lake sits on gradient
Log Spring TP
Log
Cya
no
bac
teri
a b
iovo
lum
eResponse to TP: by trophic type
No relationship in eutrophic lakes
Highly significant relationship in oligo-mesotrophic lakes
Response to summer rainfall – all lakes
Summer Rainfall(transformed and centred)
Cya
no
bac
teri
a b
iovo
lum
e
(tra
nsf
orm
ed a
nd
cen
tred
)
-4 -2 0 2 4
-2-1
01
2
Mean cyano vs precip, all lakes, transformed and centred
Summer precip
Me
an
cy
an
o b
iov
olu
me
No effect of summer rainfall
-2.0 0.0 1.5
0.0
1.0
Leven
Precip
Mean
cyan
o
-1.5 0.0
0.0
1.0
Muegg
Precip
Mean
cyan
o
-2.0 0.0
1.0
1.4
1.8
Vort
PrecipM
ean
cyan
o-2.0 0.0
-1.6
-1.2
Konn
Precip
Mean
cyan
o
-3 -1
-1.6
-1.2
Lang
Precip
Mean
cyan
o
-2.0 -0.5
-1.0
-0.4
0.2
Lapp
Precip
Mean
cyan
o
-1.0 0.5
-1.6
-1.2
-0.8
Paaj
Precip
Mean
cyan
o
-1.5 0.0
-1.5
-0.5
Pyha
Precip
Mean
cyan
o
-1.5 0.5
-0.2
0.2
0.6
Rusut
Precip
Mean
cyan
o
-1.5 0.5
-1.0
0.0
1.0
Tuus
Precip
Mean
cyan
o-1.5 0.0
-1.0
0.0
Vesi
Precip
Mean
cyan
o
-3.5 -2.0
0.5
1.5
Eymir
Precip
Mean
cyan
o
-3.0 -1.5
0.0
1.0
Mogan
Precip
Mean
cyan
o
-0.2 0.4
-0.4
0.0
0.4
Van1
Precip
Mean
cyan
o
-0.2 0.4
-0.5
0.5
Van2
Precip
Mean
cyan
o
-1.0 0.5
-1.5
-0.5
0.5Gjer
Precip
Mean
cyan
o
-1.5 0.5-1
.00.0
1.0
Kolb
Precip
Mean
cyan
o
-1.5 0.0
-2.0
-1.0
0.0
Mjos
Precip
Mean
cyan
o
-1.0 0.5
-0.4
0.0
Garda
Precip
Mean
cyan
o
0 2
1.4
1.6
Neagh
Precip
Mean
cyan
o
-1.0 1.0 2.5
-20
12
Nbas
Precip
Mean
cyan
o
-1.0 1.0 2.5
-20
12
Sbas
Precip
Mean
cyan
o
-1 1 3
0.2
0.8
1.4
Esth
Precip
Mean
cyan
o
-1 1 3
-0.4
0.2
0.8
Blel
Precip
Mean
cyan
o
-0.5 1.0 2.5
-0.6
0.0
0.6
Genev
Mean
cyan
o
-1.0 0.0
-1.0
0.0
Magg
Mean
cyan
o
Short residence time type
Cyan vs TP
Pearson's r
Fre
quency
-0.5 0.0 0.5
02
46
8
Cyan vs TempSu
Pearson's r
Fre
quency
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
02
46
810
Cyan vs Precip
Pearson's r
Fre
quency
-0.8 -0.4 0.0 0.2 0.4 0.60
24
68
Response to rainfall weak
Response to rainfall – by lake
Response to rainfall – short residence lakes
-3 -2 -1 0 1 2
-2-1
01
23
Mean cyano vs precip, all lakes, transformed and centred
Summer precip
Me
an
cy
an
o b
iov
olu
me
fitted effect to rainfall
explains 15% of the total
variation in cyanobacteria
Summer rainfall(transformed and centred)
Cya
no
bac
teri
a b
iovo
lum
e (t
ran
sfo
rmed
an
d c
entr
ed)
P<0.028 *
Significant negative effect
-1.0 -0.5 0.0 0.5 1.0 1.5
-2-1
01
23
TP
Pre
cip
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
TP and Rainfall – short res. lakes
Significant negative rainfall effect but no interaction with TP
Low cyanobacteria
High cyanobacteria
Sum
mer
rai
nfa
ll(t
ran
sfo
rmed
an
d c
entr
ed)
Spring TP(transformed and centred)
No significant effect
Explains <1% of variation in data
Response to temperature: all lakes
Summer Temperature(transformed and centred)
Cya
no
bac
teri
a b
iovo
lum
e (t
ran
sfo
rmed
& c
entr
ed)
Response to temperature: by lake
weak and varied response in individual lakes
Cyan vs TP
Pearson's r
Fre
quency
-0.5 0.0 0.5
02
46
8
Cyan vs TempSu
Pearson's r
Fre
quency
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
02
46
810
Cyan vs Precip
Pearson's r
Fre
quency
-0.8 -0.4 0.0 0.2 0.4 0.6
02
46
8
Little difference in mean but generally higher values in hot years
Response to temperature: by summer type
TP and Temperature – Interaction
significant positive TP effect except at high temperatures(antagonistic)
Sum
mer
Tem
pe
ratu
re(t
ran
sfo
rmed
an
d c
entr
ed)
Spring TP(transformed and centred)
Summary
Difficult to generalise across all lakes how cyanobacteria
respond to stressors acting alone or in combination
-1.0 -0.5 0.0 0.5 1.0 1.5
-2-1
01
23
TP
Pre
cip
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
Lake typology
adds
predicatability
Stress gradient
of study alters
perspective
Canada
UK
Species-specific
responses
Laurence Carvalho
Freshwater Ecology Group
@LacLaurence
MARS Project: Managing Aquatic
ecosystems and water resources under
multiple stress
Funded by the EU FP7, contract no. 603378
www.mars-project.eu/