Natural Selection in a Model Ocean

1
Natural Selection in a Model Ocean Mick Follows,
Scott Grant, Stephanie Dutkiewicz, Penny
Chisholm MIT
2
Ocean productivity regulates distribution and
storage of nutrients and carbon biological pumps
3
Composition and functional characteristics of
pelagic ecosystem vary in space and time...
coccolithophores CaCO3 structural material
diatoms Si structural material
diazotrophs fix nitrogen
picoplankton
4
...affecting efficiency/quality of export e.g.
recycling microbial loop vs. exporting diatom
blooms
5
Biogeography What are the dynamics underlying
provinces?
(Longhurst)
6
Johnson et al., (2006) Prochlorococcus ecotypes
along AMT section
7
Models of the Marine Ecosystem Volterra (1928),
Cushing (1935) Riley (1946)
8
Nutrient conservation NPZ models... e.g. Fasham
et al. (1990)
9
recent biogeochemical models begin to represent
functional diversity in the ecosystem
(e.g. Moore et al., 2002 Gregg et
al., 2002 Chai et al. 2002 Dutkiewicz et al.,
2005)
10
Multiple functional groups of phytoplankton
simplified example...
Functional group characteristics imposed by
parameter values
11
Prochlorococcus ecotypes
(Johnson et al., 2006)
12
AMT observations Johnson et al.
(2006) From modeling point of view,
reveals... More complexity functional diversity
within species More simplicity well defined
functional differences between otherwise very
closely related organisms
13

• Simplify modeling approach by introducing
  explicit natural selection
• Many possible functional groups (10's 100's)
• Nutrient conservation (physical principle)
• Natural selection (ecological principle)
• Generic phytoplankton
• assign functions randomly
• choose sensitivities randomly within prescribed
  ranges

14
Multiple functional groups generalized
system... Parameter values assigned
with some randomness Successful functional groups
determined by competition
15
Random assignment of functional properties
(trade-offs?)
16
sub-tropical 1-dimensional model seasonal
cycle initially 100 functional groups
phyto (log scale) temp PAR
nutrients
17
Ensemble averages
phyto nutrients
18
max growth rate Kpo4 Kno3
Kpar
Kinhib Npref Topt
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
Why do only a handful of functional groups
persist in each case?

• Reflects number of potentially limiting resources
  (Tilman, 1977)
• Also sensitive to physical environment, e.g.
  scales of turbulent variation (Tozzi et al., 2004)

Tilman (1977)
23

• Applying principle of competition simplifies
  model construction
• Level of diversity emerges, not imposed
• Self-selects functional groups according to
  physical conditions and nutrient availability
• Do plausible biological regimes and ecotypes
  emerge?

24
Johnson et al., (2006) Prochlorococcus ecotypes
along AMT section
25

• global circulation model
• 30 functional groups of phytoplankton
• 2 grazers
• nutrients NO3, NH4, NO2, PO4, Si, Fe
• phytoplankton functions and parameter values set
  by random process
• ensemble approach

26
Single ensemble member (Iseed 5007) annual mean
surface phyto (uM P) after 5 yrs
27
annual mean phyto (P), 0-120m
(Iseed 5007)
28
annual mean nutrients, 0-120m
(Iseed 5007)
29
Prochlorococcus
Synechococcus
obs (log)
model (log) model (linear)
30
Observed Modeled
NO3 NH4 NO2
31
Johnson et al., (2006)
observed
modeled
32
Outlook Natural selection approach appropriate
for modeling ocean ecosystems and biogeochemical
cycles Enables focus on underlying dynamics of
model, not tuning of parameter values Dynamic
ecosystem approach can adapt to different
climate/nutrient environments Ensemble approach
provides statistical viewpoint (c.f. adaptive
approach?) Prochlorococcus ecotype observations
provide well defined system can model help
interpret the observed structures?
33
(No Transcript)
34
Single ensemble member (Iseed 17656) annual mean
surface phyto (uM P) after 5 yrs
35
annual mean phyto(P), 0-120m (Iseed
17656)
36
annual mean nutrients, 0-120m
(Iseed 17656)
37
Prochlorococcus Diversity within species...
38

• Productivity of the oceans controlled by
• Availability of nutrients (light, phosphorus,
  nitrogen iron...)
• Significant role for wind-driven, upper ocean
  circulation

39
... and quality of sinking particulate
material association of organic
carbon with CaCO3 and opal, gt2000m
Klaas and Archer (2002)