Weigang Helene C
Department of Mathematics and Statistics, University of Helsinki, P.O. Box 68, FI-00014 Helsinki, Finland.
J Theor Biol. 2017 Aug 7;426:140-151. doi: 10.1016/j.jtbi.2017.05.020. Epub 2017 May 18.
The three phases of dispersal - emigration, transfer and immigration - are affecting each other and the former and latter decisions may depend on patch types. Despite the inevitable fact of the complexity of the dispersal process, patch-type dependencies of dispersal decisions modelled as emigration and immigration are usually missing in theoretical dispersal models. Here, I investigate the coevolution of patch-type dependent emigration and patch-type dependent immigration in an extended Hamilton-May model. The dispersing population inhabits a landscape structured into many patches of two types and disperses during a continuous-time season. The trait under consideration is a four dimensional vector consisting of two values for emigration probability from the patches and two values for immigration probability into the patches of each type. Using the adaptive dynamics approach I show that four qualitatively different dispersal strategies may evolve in different parameter regions, including a counterintuitive strategy, where patches of one type are fully dispersed from (emigration probability is one) but individuals nevertheless always immigrate into them during the dispersal season (immigration probability is one). I present examples of evolutionary branching in a wide parameter range, when the patches with high local death rate during the dispersal season guarantee a high expected disperser output. I find that two dispersal strategies can coexist after evolutionary branching: a strategy with full immigration only into the patches with high expected disperser output coexists with a strategy that immigrates into any patch. Stochastic simulations agree with the numerical predictions. Since evolutionary branching is also found when immigration evolves alone, the present study is adding coevolutionary constraints on the emigration traits and hence finds that the coevolution of a higher dimensional trait sometimes hinders evolutionary diversification.
扩散的三个阶段——迁出、转移和迁入——相互影响，前后决策可能取决于斑块类型。尽管扩散过程的复杂性是不可避免的事实，但在理论扩散模型中，通常缺少将扩散决策建模为迁出和迁入时与斑块类型的相关性。在此，我在扩展的汉密尔顿 - 梅模型中研究与斑块类型相关的迁出和与斑块类型相关的迁入的共同进化。扩散种群栖息在由两种类型的许多斑块构成的景观中，并在连续时间季节内进行扩散。所考虑的性状是一个四维向量，由从斑块迁出的概率的两个值和迁入每种类型斑块的概率的两个值组成。使用自适应动力学方法，我表明在不同参数区域可能会进化出四种定性不同的扩散策略，包括一种违反直觉的策略，即从一种类型的斑块完全迁出（迁出概率为1），但个体在扩散季节仍总是迁入这些斑块（迁入概率为1）。我给出了在广泛参数范围内进化分支的例子，此时在扩散季节具有高局部死亡率的斑块保证了高预期扩散输出。我发现进化分支后两种扩散策略可以共存：仅完全迁入具有高预期扩散输出斑块的策略与迁入任何斑块的策略共存。随机模拟与数值预测一致。由于仅当迁入单独进化时也发现了进化分支，本研究在迁出性状上增加了共同进化约束，因此发现高维性状的共同进化有时会阻碍进化多样化。
