Hello Susan, apologies for the late response .. Regional modelling, I would say ReSAKSS (P.Guthiga@cgiar.org) particularly in relation to trade and East & Central Africa. Global modelling mostly falls to the Global Futures group (D.Enahoro@cgiar.org). We also are interested in sub-continental regional work but it is not one of our strengths.

Thanks Dolapo. We need to put our heads together on this.

Thanks Dolapo. We need to put our heads together on this.

Indeed, this would be very good, for a number of reasons. Perhaps worthy of a ‘Big Idea’ entry?

Will depend on the scale of interest. Value chain and regional to global modelling rests naturally with PTVC, environmental, farm household and maybe production modeling more with LSE?

Would take the scenario work actually beyond livestock related ones, you cannot look at livestock related scenarios without looking at the other land uses. With impact, the VC modelling, the farming system mapping, production and household modelling skills in house at ILRI we are in a good position to become the ‘go to’ place. However, we are not good (enough?) at getting the best out of the combined multi-scale and multi-level strengths of our approaches.

Community of practice would be fab, also to better explore the synergies across the approaches. Analyses and (foresight) modelling work is very compartmentalized at the moment at ILRI, thereby reducing our impact. Other important discussion point: many people agree foresight and impact assessment modelling is important, but we are not good at positioning ourselves, and making clear how exactly our work is important and how it can help to create impact.

Thank you Iain. We could maybe start with inventorying who is using what models at what scales and to what purpose, within the broad theme of integrated livestock modelling. That way, we can start to understand the entry points and outputs for the various models and how synergies can be explored.

Thanks Randall for your comment. This is an important discussion to have and needs to be tabled – on the level of the PIs and maybe later with ILRI’s research management/leadership.

Hello Tim and thank you for your comment. Some of the livestock quantification work has been captured in the paper with Msangi and others. We probably need to do a better job with communicating our work! The paper should be available on Mahider or at: http://www.sciencedirect.com/science/article/pii/S0306919214001444. That work outlines changes suggested to the specification of livestock production in IMPACT and presents region-specific global projections to 2030 based on the new specifications. Follow-on work incorporates these changes into a revised IMPACT model framework that allows for commodity value chain modelling and single-country analyses. This is still under development. We are in complete agreement on the need for robust herd/flock models to inform the integrated global modelling platform. These will need, as you suggest, to be livestock production-system specific, as is the current characterization of IMPACT, but will need to be developed outside of the model (similar to the way DSSAT currently informs crop modelling in IMPACT). Outputs from the simulation models will thus quantify ‘yield gaps’ and the extents to which specific interventions change these, and can then be properly and systematically synthesized to provide the global databases needed for IMPACT. The current model can incorporate costs and benefits associated with livestock interventions.