Stochastic modeling of aphid population growth with nonlinear, power-law dynamics

James H. Matis, Thomas R. Kiffe, Timothy I. Matis, Douglass E. Stevenson

Research output: Contribution to journalArticlepeer-review

17 Scopus citations


This paper develops a deterministic and a stochastic population size model based on power-law kinetics for the black-margined pecan aphid. The deterministic model in current use incorporates cumulative-size dependency, but its solution is symmetric. The analogous stochastic model incorporates the prolific reproductive capacity of the aphid. These models are generalized in this paper to include a delayed feedback mechanism for aphid death. Whereas the per capita aphid death rate in the current model is proportional to cumulative size, delayed feedback is implemented by assuming that the per capita rate is proportional to some power of cumulative size, leading to so-called power-law dynamics. The solution to the resulting differential equations model is a left-skewed abundance curve. Such skewness is characteristic of observed aphid data, and the generalized model fits data well. The assumed stochastic model is solved using Kolmogrov equations, and differential equations are given for low order cumulants. Moment closure approximations, which are simple to apply, are shown to give accurate predictions of the two endpoints of practical interest, namely (1) a point estimate of peak aphid count and (2) an interval estimate of final cumulative aphid count. The new models should be widely applicable to other aphid species, as they are based on three fundamental properties of aphid population biology.

Original languageEnglish
Pages (from-to)469-494
Number of pages26
JournalMathematical Biosciences
Issue number2
StatePublished - Aug 2007


  • Birth-death processes
  • Kolmogrov equations
  • Moment closure methods
  • Normal approximation
  • Pecan aphids


Dive into the research topics of 'Stochastic modeling of aphid population growth with nonlinear, power-law dynamics'. Together they form a unique fingerprint.

Cite this