Biology subjects

Lajaaiti, I.

Publications and source records attributed to Lajaaiti, I..

3 recordsLinked to original sources

bioRxiv2024

Unpacking sublinear growth: diversity, stability and coexistence

How can many species coexist in natural ecosystems remains a fundamental question in ecology. Theory suggests that competition for space and resources should maintain the number of coexisting species far below the staggering diversity commonly found in nature. A recent model finds that, when sublinear growth rates of species are coupled with competition, species diversity can stabilize community dynamics. This, in turn, is suggested to explain the coexistence of many species in natural ecosystems. In this brief note we clarify why the sublinear growth (SG) model does not solve the long standing paradox of species coexistence. This is because in the SG model coexistence emerges from an unrealistic property, in which species per-capita growth rate diverges at low abundance, preventing species from ever going extinct. When infinite growth at low abundance is reconciled with more realistic assumptions, the SG model recovers the expected paradox: increasing diversity leads to competitive exclusion and species extinctions.

ecology↗

bioRxiv2024

EcologicalNetworksDynamics.jl: A Julia package to simulate the temporal dynamics of complex ecological networks

O_LISpecies interactions play a crucial role in shaping biodiversity, species coexistence, population dynamics, community stability and ecosystem functioning. Our understanding of the role of the diversity of species interactions driving these species, community and ecosystem features is limited because current approaches often focus only on trophic interactions. This is why a new modelling framework that includes a greater diversity of interactions between species is crucially needed. C_LIO_LIWe developed a modular, user-friendly, and extensible Julia package that delivers the core functionality of the bio-energetic food web model. Moreover, it embeds several ecological interaction types alongside the capacity to manipulate external drivers of ecological dynamics like temperature. These new features represent important processes known to influence biodiversity, coexistence, functioning and stability in natural communities. Specifically, they include: a) an explicit multiple nutrient intake model for producers, b) competition among producers, c) temperature dependence implemented via the Boltzmann-Arhennius rule, and d) the ability to model several non-trophic interactions including competition for space, plant facilitation, predator interference and refuge provisioning. C_LIO_LIThe inclusion of the various features provides users with the ability to ask questions about multiple simultaneous processes and stressor impacts, and thus develop theory relevant to real world scenarios facing complex ecological communities in the Anthropocene. It will allow researchers to quantify the relative importance of different mechanisms to stability and functioning of complex communities. C_LIO_LIThe package was build for theoreticians seeking to explore the effects of different types of species interactions on the dynamics of complex ecological communities, but also for empiricists seeking to confront their empirical findings with theoretical expectations. The package provides a straightforward framework to model explicitly complex ecological communities or provide tools to generate those communities from few parameters. C_LI

ecology↗

bioRxiv2023

A Comparison of Deep Learning Architectures for Inferring Parameters of Diversification Models from Extant Phylogenies

AO_SCPLOWBSTRACTC_SCPLOWTo infer the processes that gave rise to past speciation and extinction rates across taxa, space and time, we often formulate hypotheses in the form of stochastic diversification models and estimate their parameters from extant phylogenies using Maximum Likelihood or Bayesian inference. Unfortunately, however, likelihoods can easily become intractable, limiting our ability to consider more complicated diversification processes. Recently, it has been proposed that deep learning (DL) could be used in this case as a likelihood-free inference technique. Here, we explore this idea in more detail, with a particular focus on understanding the ideal network architecture and data representation for using DL in phylogenetic inference. We evaluate the performance of different neural network architectures (DNN, CNN, RNN, GNN) and phylogeny representations (summary statistics, Lineage Through Time or LTT, phylogeny encoding and phylogeny graph) for inferring rates of the Constant Rate Birth-Death (CRBD) and the Binary State Speciation and Extinction (BISSE) models. We find that deep learning methods can reach similar or even higher accuracy than Maximum Likelihood Estimation, provided that network architectures and phylogeny representations are appropriately tuned to the respective model. For example, for the CRBD model we find that CNNs and RNNs fed with LTTs outperform other combinations of network architecture and phylogeny representation, presumably because the LTT is a sufficient and therefore less redundant statistic for homogenous BD models. For the more complex BiSSE model, however, it was necessary to feed the network with both topology and tip states information to reach acceptable performance. Overall, our results suggest that deep learning provides a promising alternative for phylogenetic inference, but that data representation and architecture have strong effects on the inferential performance.

ecology↗