Detection and projection of climate change impacts on marine fisheries

Transcription

1 Detection and projection of climate change impacts on marine fisheries William W.L. Cheung Changing Ocean Research Unit & Nippon Foundation- Nereus Program, The University i of fbritish i hcolumbia AIMEN Brest France AIMEN, Brest, France 23 August 2013

2 Key focus of this talk Present selected studies of empirical and simulation modelling to understand large-scale pattern and magnitude of climate change impacts on fisheries; Detecting and attributing climate change effects on global fisheries; Projecting changes in distribution of exploited species; Effects of climate change on body size of fishes; Projecting changes in fisheries catch potential.

3 Human impacts on marine ecosystems

4

5

6

7

8

9

10

11

12

13 Limit to global fisheries es production o ion tonne es) Ca atch (mill 90 Determine by environmental and 80 other human drivers (e.g., Cheung et al. 2008; Chassot et al. 2000; Friedland et al. 2012) Catch Effort (GW or watts x 10 9 ) 10 Effective effort* Year Data source: Watson, Cheung et al. (2012) Fish and Fisheries

14 Climate change effects in the ocean? Physical Biological Social/Economics From: Sumaila, Cheung, Lam, Pauly, Herrick (2011) Nature Climate Change

15 Thermal biology of marine fishes and invertebrates From: Pörtner & Farrell (2008) Science Theory predicts that aquatic ectotherms distribute themselves to maximize i their growth performance.

16 Latitude Climate-shifted distribution Invasion Country A Country B Warming Acidification Hypoxia Salinity Local extinction Protected Area Original distribution Depth

17 Rate of change in SST from 1970 to 2010 Rate of changes in SST ( o Cyear -1 )

18 Examples of evidence of climate change effects on fisheries in specific areas From: Poloczanska et al. (2013) Nature Climate Change

19 Does climate change affect tfisheries? i Hypothesis Ocean warming has led to changes in species composition of fisheries catches.

20 Mean Temperature of Catch (MTC) Median preferred temperature = 8 o C MTC = Average preferred temperature weighted by the catch Median preferred temperature = 10 o C Median preferred temperature = 6 o C Cheung, Watson & Pauly (2013) Nature 497:

21 Mean Temperature of Catch

22 Mean Temperature of Catch

23 Attributing climate change effects with generalized additive mixed model Mean Temperature of Catch Fishing effort Large scale oceanographic indices (e.g., NAO, PDO) Warming

24 Changes in catch composition are attributed partly to warming mperature of r -1 ) ge in Mean Tem Catch ( o C year Rate of chang the C Rate of change in SST ( o C year -1 ) All Large Marine Ecosystems (LMEs) - Mean Temperature of Catch increased at a rate of 0.19 o C decade -1 ; - Significantly related to changes in SST; - Robust to alternative species distribution modelling method, indicator of species temperature preference, mis-reporting of catch data. From: Cheung, Watson & Pauly (2013) Nature.

25 Changes in catch composition in the tropics Mean Temperature of the Catch ( o C year -1 ) Mean Temperature of the Catch Year SST anomalie es ( o C) Tropical LMEs only - Increased initially at 0.6 o C decade -1 ; -0.2 Sea Surface Temperature (SST) Stabilized afterward; Hypothesis: initial reduction in representation of sub- tropical species.

26 Hypothesis of changes in catch composition Credit: Pew Charitable Trust; Based on Cheung et al. (2013)

27 Projecting future responses of marine fisheries to climate change: How will distribution of marine fish stocks be affected by climate change by 2050?

28 Dynamic bioclimate envelope model Global climate change projections Ecophysiology and Population dynamics Current species distribution ib ti Probability of occurrence by: Temperature Depth limits Habitats Distance from sea-ice Predicted future species distribution Cheung, Lam, Pauly (2008), Cheung et al. (2011)

29 Physiology Anabolism Catabolism Ocean conditions Temperature Oxygen Acidity Plankton productiv vity and commu unity stru ucture Life history (body size: maximum, maturity; growth) Productivity, mortality Adult Recruitment Juvenile Population dynamics Species distribution Sea ice Salinity Advection Habitat suitability

30 Biogeochemical forcing from Earth System Models NOAA s GFDL ESM2.1 Phytoplankton ecology Recycled nutrients N 2 fixer Small phyto. Protist Filter feeder New nutrients Large phyto. DOM Detritus Oxygen ph Dunne et al (2005, 2007)

31 Example: Small yellow croaker (Larimichthys i i h h polyactis) ) Original (static) distribution Distribution after 50 years (Climate projection from NOAA/GFDL CM 2.1) Relative abundance 0 0 > > > > > > > Low > >High

32 Predicting climate change impacts on marine biodiversity it Relative abundance 0 0 Low > > > > > > > > >High Combining i projected distributional ranges of 1,070 marine species.

33 Temperature preference ranges of exploited fishes and invertebrates Mea an inter-qu uartile rang ge ( o C) Polar spp. Tropical spp. Source: Cheung, Watson & Pauly (2013) Median temperaturet preference ( o C)

34 Comparing temperature preference from distribution range with laboratory-determined tolerance limits 55 N = 28 spp. imits ( o C) 45 perature li Environm mental tem Lower tolerance Upper tolerance In vitro determined dtemperature tolerance range ( o C)

35 Species invasion rate (N=1,070 spp.) (2050 relative to 2000 with RCP8.5: +4 o C warming) New version from Cheung et al. (2009) Fish and Fisheries

36 Species local extinction (2050 relative to 2000 with RCP8.5: +4 o C warming)

37 Shift in latitudinal centroid of 1077 spp. of fishes and invertebrates in the world ocean from Pole Distanc ce shifte ed (km) Median rate = 35 km decade km 65 km decade -1 decade -1 Equator

38 Examples of evidence of climate change effects on fisheries in specific areas 42 km decade -1 From: Poloczanska et al. (2013) Nature Climate Change

39 Comparing with observed range shifts Bering Sea and around the UK and Ireland Rate of range shift from (data from Meuter and Litzow 2007) Abundance change with temperature from (data from Simpson et al. 2011)

40 Projected range shifts and abundance changes from hindcasts significantly agree with observation Driven by reanalysis-forced simulation from GFDL ESM2.1 de -1 ) t (km deca (km decade-1) Pole eward Poleward latitu l atitudinal shift Bering Sea: Rates of range shift Model hindcasts Observations Model Median Observations Percentile Predicted change in abundance Predicted change mass over water over temp. es in relative biom emeprature ( o C -1 ) UK and Ireland: Abundance changes t y = x R² = Observed changes in abundance over water temperature ( o C -1 ) Observed change in abundance over temp. Source: Cheung, Sarmiento, Froelicher, Lam, Palomares, Watson, Pauly (2012)

41 Comparing observed and predicted Mean Temperature of Catch in NE Pacific EB Bering Sea Gulf of falaska Obs. Models California Current

42 How will climate change affect maximum body size of fishes by 2050?

43 Maximum body size and temperature: Atlantic cod (Taylor 1959)

44 Changes in von Bertalanffy growth parameters (L and K) of haddock (Melanogrammus aeglefinus) inthenorthsea that is significantly correlated with SST Source: Baudron et al. (2011)

45 Oxygen and capacity limited thermal tolerance Current condition Warmer, more acidic or less oxygenated ocean Growth Linking aerobic scope and growth function; Effects on natural mortality, maturity, fecundity and recruitment. Source: Pauly (2010); Cheung, Dunne, Sarmiento, Pauly (2011)

46 Linking ocean conditions to growth functions Growth rate Anabolism Catabolism dw/dt = H W - k W a H = g [O 2 ] e -j/t k = h e -i/t

47 Projected decreases in maximum body size of fishes Credit: Pew Environmental Trust; Based on Cheung et al. (2013) Nature Climate Change

48 Predicted changes in maximum body weight by 2050 relative to 2000 (+4 o C warming by 2100) Average individual-level decrease in maximum body size (W ) is about 8-12%; Assemblage-averaged W is projected to decrease by 14 24% from 2001 to Source: Cheung et al. (2013) Nature Climate Change

49 Comparing model results with data Model predictions Consistent trends between model projection and data e.g. cod and haddock; However, model results are more conservative. Source: Cheung et al. NCC (2013)

50 Comparing model results with data Source: Cheung, Pauly, Sarmiento (2013) ICES J. Mar. Sci.; Forster et al. (2012) PNAS

51 How will climate change affect global fisheries catch potential by 2050?

52 Factors affecting maximum catch potential A Net primary production Species distribution B C Growth & body size Catch potential

53 Empirical model predicting maximum catch potential Predicted catch (tonne es, log) N = 1,000, P < 0.001, R 2 = 0.70 Fit the model to observed maximum m catch potential of 1,000 spp of marine fishes and invertebrates; Include species from krill to tuna and sharks; The model has high explanatory power and agrees with expectations from theory. -2 Pg Observed catch (tonnes, log) Cheung et al. (2008) Mar. Ecol. Prog. Ser. 365:

54 Projected change in catch potential by 2055 Tropical regions are projected to suffer from losses while high latitude regions are projected to gain in catch potential. Source: Cheung, Lam, Kearney, Sarmiento, Watson and Pauly (2010) Global Change Biology

55 Projected decadal mean changes in marine primary production from 1860s to 2090s

56 By 2050, warming and de-oxygenation combined are predicted to decrease catch potential Baseline This analysis considered 778 exploited species. Source: Cheung, Sarmiento, Frölicher, et al. (in prep.)

57 By 2050, warming and de-oxygenation combined are predicted to decrease catch potential Baseline F = 2M This analysis considered 778 exploited species. Source: Cheung, Sarmiento, Frölicher, et al. (in prep.)

58 By 2050, warming and de-oxygenation combined are predicted to decrease catch potential Baseline F = 2M Temp. only This analysis considered 778 exploited species. Source: Cheung, Sarmiento, Frölicher, et al. (in prep.)

59 By 2050, warming and de-oxygenation combined are predicted to decrease catch potential Baseline F = 2M Temp. only Temp. + O 2 This analysis considered 778 exploited species. Source: Cheung, Sarmiento, Frölicher, et al. (in prep.)

60 By 2050, warming and de-oxygenation combined are predicted to decrease catch potential Baseline F = 2M Temp. only Temp. + O 2 Temp./O 2 /F=M This analysis considered 778 exploited species. Source: Cheung, Sarmiento, Frölicher, et al. (in prep.)

61 By 2050, warming and de-oxygenation combined are predicted to decrease catch potential Baseline F = 2M Temp. only Temp. + O 2 Temp./O 2 /F=M Temp./O 2 /F=2M This analysis considered 778 exploited species. Source: Cheung, Sarmiento, Frölicher, et al. (in prep.)

62 Incorporating trophic interactions into Dynamic Bioclimate Envelop Model Abundance (log) Energy demand from fishes Size class (log) From: Fernandes, Cheung, Jennings, Grant (2013) Global Change Biology

63 Comparing projections with data in N. Atlantic

64 What are the socio-economic implications?

65 Linking to economic and social impacts Global l climate change projections Predicted future species distribution Species composition in each EEZ Catch potential & landings (t) Ex-vessel price Gear type of each species composition Unit ($/tonne) variable cost Total variable Landed ($/tonne) fishing cost ($) Values ($) Economic rents in each EEZ

66 Percentage change in landed values in each EEZ in the 2050s Source: Lam et al. (submitted)

67 Change in undernourishment due to climate change Source: Sumaila et al. (in prep.)

68 Vulnerability Index Implications on local economies Country Vulnerability Index Rank Cote d'ivoire Ghana Sierra Leone Nigeria Togo Liberia Senegal Guinea Mauritania Guinea- Bissau Benin Gambia Cape Verde Vulnerability of national economies to the impact of climate change on fisheries in WA countries under high range GHG emission scenario (SRES A1B) From: Lam, Cheung, Swartz and Sumaila (2012) African J. of Marine Sci.

69 Uncertainties Physical and lower-trophic level models; Species adaptation to environmental changes (on-going PhD project); Multi-model comparisons and ensembles (regionally: North Sea - Jones et al. 2013a, b, NE Pacific - Cheung et al. submitted; globally: on-going project).

70 Summary Long-term changes in ocean temperature has led to changes in species composition of catch since the 1970s; In higher latitude regions, increasing dominance of warmer water species in catch; In tropical regions, further warming may impact the production of tropical species which have now become the dominant species in catch; Shrinking of fishes may further exacerbate the impact; On-going and future works to further address uncertainties and model skills testing.

71 Acknowledgement Nippon Foundation-Nereus Program; National Geographic Society; Natural Sciences and Engineering Research Council of Canada; Pew Charitable Trust;

72 By Tom Tales: published in Washington Post: 20 May 2013

73 Group discussion (A)

74 Questions for discussion What are the major challenges in developing models to assess climate change impacts on living marine resources? What are the possible ways to address these challenges?

75 Predicting species distributions Current ( ) distributional ranges of over 1,000 species of marine fishes and invertebrates are predicted from: Attributes: Depth limits; Latitudinal limits; Associated habitats; Relative abundance > Known range > boundary. 0 0 > > > > > > Low >High Small yellow croaker Close, Cheung, Hodgson, Lam, Watson, Pauly (2006) Fisheries Centre Research Report 14(4)

76 Projected changes in net primary productivity Source: Redrawn using data presented in Steinacher et al. (2010) Biogeosciences.

77 Comparison with MTC calculated from survey data - E.g., in North Sea; - No significant difference in the rate of change of MTC (ANCOVA).

78 Calculating Example: median Small preferred yellow croaker temperature (Larimichthys i i polyactis) ) Example: Small yellow croaker (Larimichthys y polyactis) ) Original (static) distribution ( ) Probability of occurrence by water temperature Relative abundance 0 0 > > > > > > > Low > >High Source: Cheung, Lam & Pauly (2008)