A comparative analysis of the structure and behavior of three gap models at sites in northeastern China

Three gap models, KOPIDE, NEWCOP, and ForClim, were compared with respect to their structure and behavior at four sites along an elevational gradient on Changbai Mt., northeastern China, under current climate and six climate change scenarios. This study intends to compare the three gap models under identical conditions, using a standardized simulation protocol. The three models were originally developed with different backgrounds and for different purposes. While they are relatively similar in the level of structural detail they include, they still differ in many respects regarding the assumptions that are made for representing specific ecological processes. The simulations showed that none of the three gap models provides satisfactory results in all situations; each gap model has strong and weak points in its behavior. While all models are fairly successful in simulating the composition of dominant species along the gradient under current climatic conditions, their projections under a set of hypothetical scenarios of climatic change diverge rather strongly. The analysis of these simulation results shows that several problem areas need to be addressed before any of the models can be used for a reliable impact assessment. Recommendations for improvements of the models are made, including the formulation of temperature and drought effects on tree establishment and tree growth, the size of the species pool, the appropriate choice of patch size and disturbance regimes, and allometric relationships. When aiming to use gap models under new environmental conditions, we propose to carefully reconsider their formulations based on our knowledge of the relevant processes in the region under concern, instead of using the models in an 'as-is' mode.