Parameter Estimation for an Allometric Food Web Model

Banks, Harvey Thomas; Banks, John E.; Bommarco, Riccardo; Curtsdotter, Alva; Jönsson, Tomas; Laubmeier, Amanda N.

doi:10.12732/ijpam.v114i1.12

Cited by 5 publications

(17 citation statements)

References 50 publications

(38 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If aphid population growth in the field (in the absence of predation) were lower than in the laboratory, for example, due to decreased plant quality as the crop matures (Leather & Dixon, ), the predation pressure would be overestimated in the model. In fact, the laboratory‐derived growth rates predicted dynamics well early in the season, when plant quality was high, but less so later on (Banks et al., ), suggesting that considering information on temporal plant quality variation would help in capturing aphid dynamics throughout the season.…”

Section: Discussionmentioning

confidence: 99%

“…The number of parameters to estimate was still too large relative to the number of aphid density observations (Banks et al., ), despite the drastic parameter reduction achieved by the allometric assumptions. We show elsewhere that to reduce this source of variation and to better estimate the allometric constants, a higher sampling frequency is needed (Banks et al., ). Additionally, the timing of sampling can be important, as data from different time points can vary in how much information they contribute to the parameter estimation.…”

Section: Discussionmentioning

confidence: 99%

“…For a full description of body mass calculations and data, which were all taken from the literature, see Banks et al. ().…”

Section: Methodsmentioning

confidence: 99%

“…The empirical data were the observed aphid population densities in each field (4–6 data points per field). We assumed an absolute error statistical model for the observations, which was supported by residual analysis (Banks et al., ). Thus, we assumed that the observations were of the form:

N_{1, j, f} = p (t_{j}, q_{0, f}) + ε_{j, f}

where N 1, j,f is the observed aphid population density at sampling occasion j in field f , the function p is the solution of the dynamical system in Equation (and hence Equation ), and ɛ j,f is the observation error.…”

Section: Methodsmentioning

confidence: 99%

“…The resulting node body masses are listed in Table 1. For a full description of body mass calculations and data, which were all taken from the literature, see Banks et al (2016).…”

Section: Body Massesmentioning

confidence: 99%

See 4 more Smart Citations

Ecosystem function in predator–prey food webs—confronting dynamic models with empirical data

Curtsdotter

Banks

et al. 2018

Journal of Animal Ecology

View full text Add to dashboard Cite

Most ecosystem functions and related services involve species interactions across trophic levels, for example, pollination and biological pest control. Despite this, our understanding of ecosystem function in multitrophic communities is poor, and research has been limited to either manipulation in small communities or statistical descriptions in larger ones. Recent advances in food web ecology may allow us to overcome the trade-off between mechanistic insight and ecological realism. Molecular tools now simplify the detection of feeding interactions, and trait-based approaches allow the application of dynamic food web models to real ecosystems. We performed the first test of an allometric food web model's ability to replicate temporally nonaggregated abundance data from the field and to provide mechanistic insight into the function of predation. We aimed to reproduce and explore the drivers of the population dynamics of the aphid herbivore Rhopalosiphum padi observed in ten Swedish barley fields. We used a dynamic food web model, taking observed interactions and abundances of predators and alternative prey as input data, allowing us to examine the role of predation in aphid population control. The inverse problem methods were used for simultaneous model fit optimization and model parameterization. The model captured >70% of the variation in aphid abundance in five of ten fields, supporting the model-embodied hypothesis that body size can be an important determinant of predation in the arthropod community. We further demonstrate how in-depth model analysis can disentangle the likely drivers of function, such as the community's abundance and trait composition. Analysing the variability in model performance revealed knowledge gaps, such as the source of episodic aphid mortality, and general method development needs that, if addressed, would further increase model success and enable stronger inference about ecosystem function. The results demonstrate that confronting dynamic food web models with abundance data from the field is a viable approach to evaluate ecological theory and to aid our understanding of function in real ecosystems. However, to realize the full potential of food web models, in ecosystem function research and beyond, trait-based parameterization must be refined and extended to include more traits than body size.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

“…For a full description of body mass calculations and data, which were all taken from the literature, see Banks et al. ().…”

Section: Methodsmentioning

confidence: 99%

N_{1, j, f} = p (t_{j}, q_{0, f}) + ε_{j, f}

Section: Methodsmentioning

confidence: 99%

“…The resulting node body masses are listed in Table 1. For a full description of body mass calculations and data, which were all taken from the literature, see Banks et al (2016).…”

Section: Body Massesmentioning

confidence: 99%

See 3 more Smart Citations

Ecosystem function in predator–prey food webs—confronting dynamic models with empirical data

Curtsdotter

Banks

et al. 2018

Journal of Animal Ecology

View full text Add to dashboard Cite

show abstract

Partitioning time series to improve process-based models with machine learning

Boussange

Aceituno²,

Pellissier

2022

Preprint

View full text Add to dashboard Cite

Ecosystems are involved in global biogeochemical cycles that regulate climate and provide essential services to human societies. Mechanistic models are required to describe ecosystem dynamics and anticipate their response to anthropogenic pressure, but their adoption has been limited in practice because of issues with parameter identification and because of model inaccuracies. While observations could be used to directly estimate parameters and improve models, model nonlinearities as well as shallow, incomplete and noisy datasets complicate this process. Here, we propose a machine learning (ML) framework relying on a mini-batch method combined with automatic differentiation and state-of-the-art optimizers. By splitting the data into mini-batches with a short time horizon, we show both analytically and numerically that the mini-batch method regularizes the learning problem. When combined with the proposed numerical implementation, the resulting ML framework can efficiently learn the parameter of complex dynamical models and is a workhorse for model selection. We evaluate the performance of the ML framework in recovering the dynamics of a simulated food-web. We show that it can efficiently learn from noisy, incomplete and independent time series, accurately estimating the model parameters and providing reliable short-term forecasts. We further show that the ML framework can provide statistical support for the true generating model among several candidates. In summary, the proposed ML framework can efficiently learn from data and elucidate mechanistic pathways to improve our understanding and predictions of ecosystem dynamics.

show abstract

From a local descriptive to a generic predictive model of cereal aphid regulation by predators

Stell,

Bommarco,

Laubmeier

et al. 2024

Journal of Animal Ecology

View full text Add to dashboard Cite

The temporal dynamics of insect populations in agroecosystems are influenced by numerous biotic and abiotic interactions, including trophic interactions in complex food webs. Predicting the regulation of herbivorous insect pests by arthropod predators and parasitoids would allow for rendering crop production less dependent on chemical pesticides. Curtsdotter et al. (2019) developed a food‐web model simulating the influences of naturally occurring arthropod predators on aphid population dynamics in cereal crop fields. The use of an allometric hypothesis based on the relative body masses of the prey and various predator guilds reduced the number of estimated parameters to just five, albeit field‐specific. Here, we extend this model and test its applicability and predictive capacity. We first parameterized the original model with a dataset with the dynamic arthropod community compositions in 54 fields in six regions in France. We then integrated three additional biological functions to the model: parasitism, aphid carrying capacity and suboptimal high temperatures that reduce aphid growth rates. We developed a multi‐field calibration approach to estimate a single set of generic allometric parameters for a given group of fields, which would increase model generality needed for predictions. The original and revised models, when using field‐specific parameterization, achieved quantitatively good fits to observed aphid population dynamics for 59% and 53% of the fields, respectively, with pseudo‐R2 up to 0.99. But the multi‐field calibration showed that increased model generality came at the cost of reduced model reliability (goodness‐of‐fit). Our study highlights the need to further improve our understanding of how body size and other traits affect trophic interactions in food webs. It also points up the need to acquire high‐resolution data to use this type of modelling approach. We propose that a hypothesis‐driven strategy of model improvement based on the integration of additional biological functions and additional functional traits beyond body size (e.g., predator space search or prey defences) into the food‐web matrix can improve model reliability.

show abstract

Parameter Estimation for an Allometric Food Web Model

Cited by 5 publications

References 50 publications

Ecosystem function in predator–prey food webs—confronting dynamic models with empirical data

Ecosystem function in predator–prey food webs—confronting dynamic models with empirical data

Partitioning time series to improve process-based models with machine learning

From a local descriptive to a generic predictive model of cereal aphid regulation by predators

Contact Info

Product

Resources

About