status on jellyplankton modeling stemmann, l laboratory of oceanography of villefranche sur mer

Download Status on jellyplankton modeling Stemmann, L Laboratory of Oceanography of Villefranche sur Mer

If you can't read please download the document

Upload: thomas-cummings

Post on 18-Jan-2018

217 views

Category:

Documents


0 download

DESCRIPTION

Introduction

TRANSCRIPT

Status on jellyplankton modeling Stemmann, L Laboratory of Oceanography of Villefranche sur Mer Content 1.Introduction 2.Population based modes 3.Individual based models 4.Functionnal type model 5.Bayesien approach 6.ECOSIM/ECOPATH models 7.Back to data Introduction PointB Zooplankton time series study on target species Introduction 3 Jellyfish sps. (Liriope tetraphylla, Solmundella bitentaculata, and Rhopalonema velatum.) 2 Siphonophore sps. (Abylopsis tetragona and Chelophyes appendiculata) 1 Ctenophore sp. (Pleurobrachia rhodopis) 5 copepod sps (C. typicus, T. stylifera, Acartia clausi, Oithona spp., and Oncaea spp.) 3 chaetognath sps (S. enflata, S. setosa and S. minima) Molinero et al., 2005 and 2008 Shift and reorganization of zooplankton community at the end of the 80s (1987 shift year) toward an oligotrophic system. dry years PointB Zooplankton time series study Introduction What about the whole jellyfish, chaetognath and copepod populations? Are the authors predictions confirmed after 10 more years of observations? Molinero et al., Introduction Population matrix model And previous presentation by Tamara Shiganova and Paul Nival Population matrix model Individual based model NPZD models but also the talk of Arthur yesterday NPZD models NPZD model Population dynamics model (number and weight) Probabilistic approach Bayesien : See J. Ruiz yesterday (temp and food) Predicting the distribution of the scyphomedusa Chrysaora quinquecirrha in Chesapeake Bay M. B. Decker1,*, C. W. Brown2, R. R. Hood3, J. E. Purcell4, T. F. Gross5, J. C. Matanoski3,6, R. O. Bannon7, E. M. Setzler-Hamilton8, MEPS 329:99-113(2007) - doi: /meps (temp and salinity) Ecopath with Ecosim Pelagia noctiluca The talk will not focus on large dangerous Scyphozoa but rather on smaller Hydrozoa and their relation with their changing environnement. BACK to data Zooplankton Datasets Results The different groups do not share the same seasonal cycle. Chaetognaths and jellyfish may not be in competition for the same ressource. Chaetognatha (log10(ind.m -3 +1)) Medusae + Siphonophores (log10(ind.m -3 +1)) Copepods (log10(ind.m )) Zooplankton Datasets Results All groups were abundant during the dry years All groups decreased in concentrations. There is a change in the phenology mainly of copepods (lack of second peak). Since chaetognaths are more abundant in summer/autumn and specific predators of copepods, their decrease may have been greater than jellyfish Chaetognatha (log10(ind.m -3 +1)) Medusae + Siphonophores (log10(ind.m -3 +1)) Copepods (log10(ind.m )) Relation with hydro-climate Results 1st PC Environmental datasets (33%): 20,50m T. and S., at.P, Ek.D., Rain, Sun : : Salinity and Temp. Salinity and Mean Temp.autumn Salinity and Temp Salinity All groups (phyto. 2 peaks) All groups copepods and Jellies 1st PC Zooplankton datasets (40%): S.cops., L.cops., Chaeto., Jelly., Dol.,Dec. ~synchronised long-term changes in environment and zooplankton : : All groups (phyto. 2 peaks) Correlation between the winter NAO and 1 st PC of environnemental conditions (r=0.44 / p=0.02) Dry years Conclusion What is missing ? Ecophysiology of several species and mainly adult phase that cannot be cultivated in the lab. - what is the impact on mesozooplankton ? - what is the impact on recycling elements ? - are temp. and food the main forcings ? - what are their main predators ? Spatial distribution in the open sea What could help ? IBM coupled to Lagrangian models Conclusions Previous observations on target species (Molinero et al., 2008) are confirmed at the level of some groups populations (copepods and chaetognaths) but not for others (jellyfish). Trends in target jellyfish species may not be representing the whole community trend. Using the additional data, we observe quasi-decadal oscillations (8-9 years) in the plankton community rather than a regime shift in 1987 as previously proposed (Molinero et al., 2008). Conclusion Conclusions Conclusion Zooplankton changes are closely linked to hydroclimate, and a bottom-up control is suggested (Garcia-Comas, session 64). The winter convection (low T and high S) may be the principal factor to initiate the bloom and zooplankton development in spring. Autumn bloom may be function of salinity stratification. Imaging techniques (here the ZooScan) can be used to monitor changes in the plankton ecosystem. Automatic recognition is reliable only for a few taxonomic groups but deeper taxonomic details can be easily obtained by semi- automatic analysis. The produced datasets are reliable and homogenous (required for long term study). Thanks to : SESAME project, EC Contract No GOCE , funded by the European Commission's Sixth Framework Programme under the priority 'Sustainable Development, Global Change and Ecosystems'. CTD, plankton and coulter data are part of the RADE-HYDRO and SOMLIT programs carried out by the Observatory of Oceanography of Villefranche/mer and the Laboratory of Villefranche sur Mer. Isabelle Palazzoli for providing the meteo data and to Corinne Desnos for the help with the scanning on the ZOOSCAN (http://www.zooscan.com)http://www.zooscan.com Zooplankton Datasets Results Chaetognatha Siphonophora Medusae Decapoda Environmental Datasets Results Water Temp.(20 and 50 m.) Ekman depth Salinity (20 and 50 m.) Rainfall Sampling Site Point B (Villefranche Bay, Ligurian Sea) from 1966 to 2003 Daily sampling with the JB net (330 m mesh) by Vertical hauls (70-0m). Samples are pooled in weekly jars. Material and Method Point B ASLO 500 SAMPLES Time Series TRAINING SET TEST SET Defining groups 1 2 Prediction by the model (Random Forest) Accuracy (confusion matrix) cross-validation Material and Methods Automatic recognition of copepods (ZOOSCAN) COPEPODS 96% recognition 18% contamination 3 (confusion matrix) PREDICTION >10 6 copepods were extracted and sized Material and Methods Automatic recognised Objects (8000) Chaetognatha Medusae Siphonophora (only 1 st bell and no colonies ind.) Doliolida Decapoda Franck Prejger: Gelatinous zooplankton final sorting Semi-Automatic recognition of other groups (ZOOSCAN) Material and Methods Dataset Validation The automatic counting is similar to the visual counting for copepods. Material and Methods Dataset Validation The long term evolution of the three groups are similar as counted independently on the ZOOSCAN or by a taxonomist many years before (S. Dallot, Oliver Beck). Point B time series Results Chaetognaths, medusae and copepods have quite similar long term changes. Total jellyfish abundance deacreases after 1990 and increases slightly after (Medusae + Siphonophores) Introduction