1. Brad deYoung Open Pelagic Ecosystems

2. Roadmap  Ecosystem structure – considerations of the issues and how to think about them  Regime shifts in the ocean – examples of some observed behaviour that we do not quite understand  Variability and modelling of marine ecosystems, some examples

4. Food web structure, and model structures?

5. Nutrient Pool Mass balance models –NPpZzD … Structured population model Simplified life history model or data Top down view more life history driven >> structured models Bottom up view more process driven – represent metabolism, mass- balance Challenge : With a target species focus to couple structured models (even if not IBM) and predators and prey below which may have different model structures or data All models are wrong, but some models are useful. George Box

6. What criteria do we use to simplify our ecosystem food web? Selection of target species A mixture of theory, observation and pragmatism  Functionally important/ecologically significant  Extensive data sets (spatial and temporal)  Choose a food web in which the first PCA contains relatively few species  Concurrence with other relevant data sets  Understanding of life history  Widely distributed/across the basin  Economic and societal importance  Well resolved taxonomy  …

7. Chavez et al. Science (2002) Modelling can be driven by  data exploration (see right)  process exploration  forecast simulations Low frequency changes are more important than we once thought The long-period changes lead to larger spatial scales >> basins Shifts in physical (temperature, mixed layer depth, …) and biological properties (phytoplankton, zooplankton,…)

8. Trophic complexity – maintaining fidelity to life history as it becomes more complex and also more difficult to model Number of state variables Probable number of species Top predators Trophic level Bacteria Detail of resolution Key taxa Predation Feeding ICES - Report of the Study Group on spatial and temporal integration, University of Strathclyde, Glasgow, Scotland, 14-18 June 1993. ICES CM 1993/L:9, (1993).

9. Functional Complexity Trophic level Phytoplankton/ nutrient focus Zooplankto n/fisheries focus Physical Ocean Predators Life History Chemistry Without Without Life History deYoung et al. Science. 2004

10. Zooplankton Focus Fish - myctophids, redfish, herring, blue whiting; Zooplankton - gelatinous, euphausiids Fish - sandlance, capelin, herring, sprat, mackerel, Norway put, blue whiting; Zooplankton - gelatinous, euphausiids Physics and chemistry – high resolution large scale circulation, coupling between global, basin and shelf models Food for zooplankton: Microzooplankton, diatoms, non-diatoms, Phaeocystis Unstructured competitors for structured zooplankton Zooplankton – Structured population representations of key basin distributed species – variously, particularly congeneric Calanus spp., euphausiids First Order Horizontal Structure Top-down predation Challenge lies in coupling the structured and unstructured models and data Coupling with the structured components will likely be one-way

11. Low frequency ‘cycles’ are not likely as linear as they may appear deYoung et al. Prog. Ocgy. ( 2004), TREE 2008  linear shifts, i.e. nothing special happening  abrupt shifts but reversible in principle  non-linear shifts that are not easily reversible  how linear is the fundamental behaviour that we are trying to represent? Anderson et al. TREE 2008

12. DEFINITION OF THE REGIME SHIFT Working definition : a regime shift is a relatively abrupt change between contrasting persistent states in an ecosystem

13. Erosion of resilience Environmental driver Environmental state Erosion of resilience

14. Review of a few examples of regime shifts in pelagic ecosystems Scotian Shelf – driven primarily by fishing, cascading trophic impacts North Sea – combined drivers: natural=biogeographic shift and human=fishing North Pacific – complex natural state change(s) Explore characteristics of the drivers and response of differing examples – time and space scales, trophic structure, predictability

15. Scotian Shelf – Frank et al. 2005 -30% +30%

17. Colour display of 60+ indices for Eastern Scotian Shelf Red – below average Green – above average Grey seals - adults Pelagic fish - #’s Pelagic:demersal #’s Pelagic:demersal wt. Inverts - $$ Pelagics - wt Diatoms Grey seals – pups Pelagics - $$ Greenness Dinoflagellates Fish diversity – richness 3D Seisimic (km2) Gulf Stream position Stratification anomaly Diatom:dinoflagellate Sea level anomaly Volume of CIL source water Inverts – landings Bottom water 150 GRT) Cod – length at age 6 Average weight of fish Community similarity index PCB’s in seal blubber Relative F Pollock – length at age 6 Calanus finmarchicus Groundfish biomass – RV Pelagics – landings Silver hake – length at age Condition – KF Depth of mixed layer Condition – JC Proportion of area – condition RIVSUM Sigma-t in mixed layer Oxygen Wind stress (total) Wind stress (x-direction) Wind stress amplitude SST at Halifax Groundfish - $$ Salinity in mixed layer Ice coverage Wind stress (Tau) Number of oil&gas wells drilled Nitrate Groundfish fish - #’s Shannon diversity index –fish Seismic 2D (km) 1970 1975 1980 1985 1990 1995 2000 Grey seals, pelagic fish abundance, invertebrate landings, fish species richness, phytoplankton Bottom temp., exploitation, groundfish biomass & landings, growth-CHP, avg. fish weight, copepods

18. Top Predators (Piscivores) Forage (fish+inverts) (Plankti-,Detriti-vores) Zooplankton (Herbivores) Phytoplankton (Nutrivores) + - + - Frank et al. 2004/2005 Science et al. Scotian Shelf – top down story

19. North Pacific regime shift – Hare and Mantua (2000)

20. Physical forcing – air temperature - but there are dozens, and dozens of other such time series

22. The technique of Hare and Mantua has been criticized as being subject to false positives – taking the normalized variance anomalies of many different time series with red spectra can lead to ‘apparent’ shifts The lack of sufficient clear data is one problem The time series are too short The regimes are likely never completely in equilibrium Many different possible states are likely Anderson et al. Reviewed the different approaches, and confirm the basic result of Hare and Mantua

23. North Sea regime shift – a mixture of biogeography, environmental change and fishing Mean number of calanoid species per CPR sample 5862667074788286909498 1 2 3 4 5 6 7 8 9 10 11 12 1 1.5 2 2.5 3 3.5 4 4.5 5 Years MONTHSMONTHS Line in black: warm-temperate species Line in red: temperate species Before 1980After 1980

24. Gadoid species (cod) SST NHT anomalies plankton change Flatfish salinity Westerly wind plankton change

25. Beaugrand & Ibanez (in press, MEPS) Beaugrand G (2004) Progress in Oceanography

26. Beaugrand & Ibanez (in press, MEPS) Beaugrand G (2004) Progress in Oceanography

27. Long-term changes in the abundance of two key species in the North Sea Percentage of C. helgolandicus Reid et al. (2003)

28. Consequences of plankton changes on higher trophic level Mismatch between the timing of calanus prey and larval cod Abundance of C. finmarchicusAbundance of C. helgolandicus Beaugrand, et al. (2003) Nature. Vol. 426. 661-664.

29. But there is also a significant influence of fishing – how much??

30. Meteorological/oceanographic forcing Ocean circulation Biogeographic shift Ocean conditions Ecosystem status and function Fishing North Sea - dynamics

32. ICES Report on Ocean Climate 2006. Prepared by the Working Group on Oceanic Hydrography Sarah L. Hughes and N. Penny Holliday, Editors. ICES cooperative research report no. 289 special issue September 2007 (from Figure 4). From the NOAA Optimum Interpolation SSTv2 dataset, provided by the NOAA-CIRES Climate Diagnostics Center, USA. The anomaly is calculated with respect to normal conditions for the period 1971–2000. The data are produced on a one-degree grid from a combination of satellite and in situ temperature data. Regions with ice over for >50% of the averaging period are left blank. Seasonal sea surface temperature anomalies over the North Atlantic for 2006

33. Expected Result: Major impact for marine exploited resources and biogeo- chemical processes (e.g. sequestration of CO 2 by the ocean). Biological consequences expected under climatic warming Or changes in water mass structure. Changes in the range and spatial distribution of species. Shifts in the location of biogeographical boundaries, provinces, and biomes. Change in the phenology of species (e.g. earlier reproductive season). Modification in dominance (e.g. a key species can be replaced by another one). Change in diversity. Change in other key functional attributes for marine ecosystems. Change in structure and dynamics of ecosystem with possible regime shifts.

34. Warm-water species have extended their distribution northwards by more than 10° of latitude, while cold-water species have decreased in number and extension. ( Beaugrand, G. ICES Journal of Marine Science, 62: 333-338 (2005) Long-term changes in the mean number of species per assemblage based on three periods: 1958-1981, 1982-1999, and 2000-2002.

35. Calanus finmarchicus in the North Atlantic - open ocean, deep and shallow, spreads out onto shelf, for some species some evidence for genetic separation, copepods key organisms for food web, coupled with circulation

37. Diapause depth – how deep do they go? Heath et al. (2004)

38. Calanus in the Labrador Sea  Population achieves maximum growth rate when emergence is 1 month prior to the spring bloom  Timing of population peaks is closely matched observations Phytoplankton Temperature Non-diapausing individuals Biological model – with a lot of detail on copepod development and growth – in this case the numerical organism eats satellite (SeaWifs) chlorphyll data The physical model represents the seasonal circulation in the Labrador Sea and the organisms are carried around in it In the vertical the zooplankton behaviour determines their position Diapausing individuals Tittensor et al. Fish. Ocgy. 2004

39. Advection  Latitudinally dependent emergence, starting in South (March) and later to the North (May)  only start out in water > 1000m deep (none on the shelf)  results in a peak in the centre of the Labrador Sea  some Calanus move up onto the shelf  Locally sustainable population January JulyNovember May Tittensor et al. Fish. Ocgy. 2004

40. Model design for the North Atlantic Calanus problem – Heath, Speirs, Gurney et al. (2005)

41. PhotoperiodLow foodH2 Development at depth Low foodH1 EntryExit Use the model to test different hypotheses of diapause – a process for which we have no direct process model

42. Surface Copepodites Diapausers Newly surfaced overwinterers No diapausers in spring Sharp drop at awakening H1 H3 OWS Mike - hypothesis test

43. Long-term spatial structure and advection through the basin Year 1 Year 3 Year 6 Speirs et al. Fish. Ocgy. (2005)  Preliminary conclusion is hat biology dominates over circulation  Requires some ‘adjusting’ different parts of the basin  Is able to reproduce the population dynamics at the basin scale – for the first time.

44. If god had consulted me before embarking on the creation, I would have suggested something simpler. Alfonso of Castile (15 th century)