index.fff

\begin{figure}

\centering{

\pandocbounded{\includegraphics[keepaspectratio]{figures/MANOVA_lda.png}}

}

\caption{\label{fig-structure}Linear discriminant analysis (LDA) of
ecological network metrics for six reconstruction approach types. Each
point represents a replicate (n = 400 per approach), and ellipses
indicate 95\% confidence regions for each reconstruction approach.}

\end{figure}
\efloatseparator
 
\begin{figure}

\centering{

\pandocbounded{\includegraphics[keepaspectratio]{figures/combined_ANOVA_time.png}}

}

\caption{\label{fig-anova}Temporal changes in food-web structure and the
relative influence of reconstruction approach and ecological turnover on
network properties. (\textbf{A}) Changes in network metrics through
time, expressed as deviations from pre-extinction conditions
(\(\delta\)). Metrics are faceted and scaled independently to highlight
differences in temporal trajectories among network properties.
(\textbf{B}) Relative influence of reconstruction approach (x-axis) and
time (y-axis) on individual network metrics. The dashed 1:1 line
indicates equal explanatory power. Metrics below the line are more
strongly influenced by reconstruction approach than by temporal
turnover. Bubble size represents the reconstruction approach × time
interaction, and colour indicates disagreement among reconstruction
approaches (mean CV\%), with darker colours indicating greater
divergence in metric estimates.}

\end{figure}
\efloatseparator
 
\begin{figure}

\centering{

\pandocbounded{\includegraphics[keepaspectratio]{figures/beta_div.png}}

}

\caption{\label{fig-beta_div}Pairwise interaction turnover (\(β_{OS}\))
among four food-web reconstruction approaches that infer
species-specific trophic interactions (ADBM, ATN, Body-size Ratio, and
PFWIM). Each cell represents the mean turnover value between a pair of
reconstruction approaches, with darker colours indicating greater
dissimilarity in inferred interactions. High turnover values indicate
strong disagreement in predicted trophic links, whereas lower values
indicate greater similarity.}

\end{figure}
\efloatseparator
 
\begin{figure}

\centering{

\pandocbounded{\includegraphics[keepaspectratio]{figures/kendal_tau.png}}

}

\caption{\label{fig-mad}Heatmaps showing pairwise Kendall rank
correlation coefficients (\(\tau\)) between reconstruction approaches
for each network metric. Each panel corresponds to a different metric
and displays the degree of agreement in extinction-scenario rankings
across reconstruction approaches based on mean absolute differences
(MAD) between observed and predicted network values as well as
differences in species links (TSS). Positive \(\tau\) values (purple)
indicate concordant rankings between reconstruction approaches, whereas
negative \(\tau\) values (sienna) indicate opposing rankings. Paler
colours approaching zero represent little or no agreement.}

\end{figure}