Harnessing uncertainty to approximate mechanistic models of interspecific interactions

Adam Thomas Clark, Claudia Neuhauser

Research output: Contribution to journalArticle

Abstract

Because the Lotka–Volterra competitive equations posit no specific competitive mechanisms, they are exceedingly general, and can theoretically approximate any underlying mechanism of competition near equilibrium. In practice, however, these models rarely generate accurate predictions in diverse communities. We propose that this difference between theory and practice may be caused by how uncertainty propagates through Lotka–Volterra systems. In approximating mechanistic relationships with Lotka–Volterra models, associations among parameters are lost, and small variation can correspond to large and unrealistic changes in predictions. We demonstrate that constraining Lotka–Volterra models using correlations among parameters expected from hypothesized underlying mechanisms can reintroduce some of the underlying structure imposed by those mechanisms, thereby improving model predictions by both reducing bias and increasing precision. Our results suggest that this hybrid approach may combine some of the generality of phenomenological models with the broader applicability and meaningful interpretability of mechanistic approaches. These methods could be useful in poorly understood systems for identifying important coexistence mechanisms, or for making more accurate predictions.

Original languageEnglish (US)
Pages (from-to)35-44
Number of pages10
JournalTheoretical Population Biology
Volume123
DOIs
StatePublished - Sep 1 2018

Fingerprint

interspecific interaction
mechanistic models
uncertainty
prediction
coexistence

Keywords

  • Interspecific competition
  • Interspecific tradeoff
  • Lotka–Volterra competitive equations
  • Model abstraction
  • Model uncertainty
  • Process noise

PubMed: MeSH publication types

  • Journal Article
  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, Non-P.H.S.

Cite this

Harnessing uncertainty to approximate mechanistic models of interspecific interactions. / Clark, Adam Thomas; Neuhauser, Claudia.

In: Theoretical Population Biology, Vol. 123, 01.09.2018, p. 35-44.

Research output: Contribution to journalArticle

@article{0c92633230ae4b07a40ad18d648af0bd,
title = "Harnessing uncertainty to approximate mechanistic models of interspecific interactions",
abstract = "Because the Lotka–Volterra competitive equations posit no specific competitive mechanisms, they are exceedingly general, and can theoretically approximate any underlying mechanism of competition near equilibrium. In practice, however, these models rarely generate accurate predictions in diverse communities. We propose that this difference between theory and practice may be caused by how uncertainty propagates through Lotka–Volterra systems. In approximating mechanistic relationships with Lotka–Volterra models, associations among parameters are lost, and small variation can correspond to large and unrealistic changes in predictions. We demonstrate that constraining Lotka–Volterra models using correlations among parameters expected from hypothesized underlying mechanisms can reintroduce some of the underlying structure imposed by those mechanisms, thereby improving model predictions by both reducing bias and increasing precision. Our results suggest that this hybrid approach may combine some of the generality of phenomenological models with the broader applicability and meaningful interpretability of mechanistic approaches. These methods could be useful in poorly understood systems for identifying important coexistence mechanisms, or for making more accurate predictions.",
keywords = "Interspecific competition, Interspecific tradeoff, Lotka–Volterra competitive equations, Model abstraction, Model uncertainty, Process noise",
author = "Clark, {Adam Thomas} and Claudia Neuhauser",
year = "2018",
month = "9",
day = "1",
doi = "10.1016/j.tpb.2018.05.002",
language = "English (US)",
volume = "123",
pages = "35--44",
journal = "Theoretical Population Biology",
issn = "0040-5809",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Harnessing uncertainty to approximate mechanistic models of interspecific interactions

AU - Clark, Adam Thomas

AU - Neuhauser, Claudia

PY - 2018/9/1

Y1 - 2018/9/1

N2 - Because the Lotka–Volterra competitive equations posit no specific competitive mechanisms, they are exceedingly general, and can theoretically approximate any underlying mechanism of competition near equilibrium. In practice, however, these models rarely generate accurate predictions in diverse communities. We propose that this difference between theory and practice may be caused by how uncertainty propagates through Lotka–Volterra systems. In approximating mechanistic relationships with Lotka–Volterra models, associations among parameters are lost, and small variation can correspond to large and unrealistic changes in predictions. We demonstrate that constraining Lotka–Volterra models using correlations among parameters expected from hypothesized underlying mechanisms can reintroduce some of the underlying structure imposed by those mechanisms, thereby improving model predictions by both reducing bias and increasing precision. Our results suggest that this hybrid approach may combine some of the generality of phenomenological models with the broader applicability and meaningful interpretability of mechanistic approaches. These methods could be useful in poorly understood systems for identifying important coexistence mechanisms, or for making more accurate predictions.

AB - Because the Lotka–Volterra competitive equations posit no specific competitive mechanisms, they are exceedingly general, and can theoretically approximate any underlying mechanism of competition near equilibrium. In practice, however, these models rarely generate accurate predictions in diverse communities. We propose that this difference between theory and practice may be caused by how uncertainty propagates through Lotka–Volterra systems. In approximating mechanistic relationships with Lotka–Volterra models, associations among parameters are lost, and small variation can correspond to large and unrealistic changes in predictions. We demonstrate that constraining Lotka–Volterra models using correlations among parameters expected from hypothesized underlying mechanisms can reintroduce some of the underlying structure imposed by those mechanisms, thereby improving model predictions by both reducing bias and increasing precision. Our results suggest that this hybrid approach may combine some of the generality of phenomenological models with the broader applicability and meaningful interpretability of mechanistic approaches. These methods could be useful in poorly understood systems for identifying important coexistence mechanisms, or for making more accurate predictions.

KW - Interspecific competition

KW - Interspecific tradeoff

KW - Lotka–Volterra competitive equations

KW - Model abstraction

KW - Model uncertainty

KW - Process noise

UR - http://www.scopus.com/inward/record.url?scp=85048105243&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048105243&partnerID=8YFLogxK

U2 - 10.1016/j.tpb.2018.05.002

DO - 10.1016/j.tpb.2018.05.002

M3 - Article

VL - 123

SP - 35

EP - 44

JO - Theoretical Population Biology

JF - Theoretical Population Biology

SN - 0040-5809

ER -