Additional Results and Analyses

Effects of network reconstruction on food web structure

Table S1. Descriptive statistics (mean ± standard deviation) of network metrics by model

In [1]:

library(knitr)
library(tidyverse)

── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.2.1     ✔ readr     2.2.0
✔ forcats   1.0.1     ✔ stringr   1.6.0
✔ ggplot2   4.0.3     ✔ tibble    3.3.1
✔ lubridate 1.9.5     ✔ tidyr     1.3.2
✔ purrr     1.2.2     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

readr::read_csv("tables/Table_S1_descriptive_stats.csv") %>%
kable()

Rows: 6 Columns: 7
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (7): statistic, ADBM, ATN, Body-size ratio, Niche, PFIM, Random

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

statistic	ADBM	ATN	Body-size ratio	Niche	PFIM	Random
Competition index	-0.0322 ±0	-0.00707 ±0	0.0641 ±0	-0.426 ±0	-0.788 ±-1	0.00323 ±0
Connectance	0.297 ±0	0.224 ±0	0.174 ±0	0.163 ±0	0.13 ±0	0.159 ±0
Generality	0.913 ±1	1.07 ±1	0.613 ±1	1.04 ±1	1.6 ±2	0.434 ±0
Max trophic level	3.34 ±3	2.85 ±3	6.16 ±6.2	5.07 ±5.1	3.08 ±3	6 ±6.0016375188
Trophic index	-1 ±-1	-1 ±-1	0.685 ±1	0.193 ±0	-0.15 ±0	0.678 ±1
Vulnerability	0.878 ±1	1.06 ±1	0.707 ±1	0.578 ±1	0.618 ±1	0.435 ±0

Table S2. Canonical discriminant analysis

In [2]:

readr::read_csv("tables/canonical_loadings.csv") %>%
kable()

Rows: 6 Columns: 4
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (1): Metric
dbl (3): Can1, Can2, Can3

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

Metric	Can1	Can2	Can3
connectance	-0.49	-0.61	0.21
trophic_level	0.30	-0.06	0.04
generality	-0.61	0.77	0.07
vulnerability	-0.69	-0.38	0.28
NDTI	0.97	0.20	0.04
NDCI	0.17	-0.95	0.08

Figure S1. Canonical Loadings

Canonical loadings for the first two canonical variates (CV1, CV2) from the canonical discriminant analysis of network metrics. Arrows indicate the contribution of each metric to the multivariate separation among reconstruction models. Colours denote the scale of each metric: Macro (light brown), Meso (brown), Micro (sienna). Metric labels are shown for the most influential variables.

PERMANOVA Variance Partitioning

To quantify the relative contributions of model and temporal turnover to variation in inferred network structure, we conducted permutational multivariate analysis of variance (PERMANOVA). Euclidean distance matrices were calculated from standardised (z-transformed) network metrics. Model and extinction phase (‘time’) were analysed separately to estimate their total contributions to variance, and in combination to assess interaction effects. Significance was assessed using 999 permutations.

Robustness of model effects after temporal centering

To determine whether the dominance of model reflected absolute structural shifts among extinction phases, we repeated the analysis after centering network metrics within each time bin. This procedure removes mean temporal differences while preserving within-phase structural variation. Even after temporal centering, model type explained 84.8% of multivariate variance (R² = 0.848, p < 0.001), exceeding the variance explained in the uncentered analysis. Thus, the strong influence of model identity is not attributable to temporal mean differences, but reflects intrinsic structural divergence among models.

Statistical Drivers of Network Variation

Statistical Robustness and Assumptions

Factorial ANOVA assumptions were validated via residual analysis. Despite significant heteroscedasticity (Levene’s test, p<0.001), the perfectly balanced design (n = 100 per cell) and large sample size (N = 2400) ensure the robustness of the F-test. Visual inspection of Q-Q plots and Residuals-vs-Fitted plots confirmed that the distributions were sufficiently symmetric for parametric analysis.

Figure S2. Temporal Trajectories of Network Structure by Model

Detailed shifts in network properties across the four extinction phases, categorised by organisational scale (Macro, Meso, Micro). Each line represents the mean value for a specific model, with error bars denoting standard error. This figure illustrates the “baseline” differences between models—such as the Niche model’s tendency to over-estimate motif counts—and their divergent responses to species loss.

Table S3. Variance Partitioning of Framework, Time, and Interaction Effects.

Summary of the two-way factorial ANOVA results for all eight metrics. Values represent partial eta-squared (\(\eta^{2}_{p}\)), which quantifies the proportion of variance explained by each factor. The dominance of the ‘Model’ term across all scales confirms that framework choice is the primary determinant of network topology.

In [3]:

readr::read_csv("../tables/ANOVA_Results.csv") %>%
kable()

Rows: 6 Columns: 4
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (1): Metric
dbl (3): Model, Time, Interaction

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

Metric	Model	Time	Interaction
Connectance	0.698	0.040	0.071
Max trophic level	0.097	0.002	0.037
Generality	0.928	0.107	0.297
Vulnerability	0.801	0.199	0.234
Trophic index	0.957	0.010	0.078
Competition index	0.903	0.011	0.414

Figure S3. Model Disagreement (CV%) Across Extinction Phases

Trends in inter-model disagreement, quantified as the Coefficient of Variation (CV%) between framework means. The Y-axis is standardised across panels to facilitate comparison between organisational scales. A characteristic “dip” at the ‘during’ phase in several meso-scale metrics illustrates the structural canalisation effect, where severe species loss forces a temporary convergence in model predictions.

Table S4. Percentage Disagreement Between Frameworks Across Extinction Phases.

Calculated inter-model CV% for each metric at each time step. These data points underpin the bubble sizes in Figure 2 and the trajectories in Figure S3. Note the reduction in CV% for linear chains and omnivory motifs during the peak extinction phase (During).

In [4]:

readr::read_csv("../tables/Model_Agreement_CV.csv") %>%
kable()

Rows: 6 Columns: 7
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (3): statistic, measure, level
dbl (4): Pre-extinction, During extinction, Early extinction, Late extinction

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

statistic	measure	level	Pre-extinction	During extinction	Early extinction	Late extinction
Competition index	SD	Meso	0.3089372	0.4151538	0.3786808	0.2868644
Connectance	CV (%)	Macro	34.7968008	27.3648444	31.8403950	33.8242979
Generality	CV (%)	Micro	50.0442089	35.5929853	43.4943368	44.5496783
Max trophic level	CV (%)	Macro	36.4237528	43.0994262	42.4989620	35.0872664
Trophic index	SD	Meso	0.7707593	0.7617292	0.7712399	0.7631036
Vulnerability	CV (%)	Micro	35.3992011	30.1478011	32.6892257	32.5026065

Figure S4. Pairwise Framework Comparisons (Tukey HSD)

Heatmap showing significant differences between specific pairs of models across each extinction phase. Colours represent the magnitude and direction of the difference (estimate); asterisks (\(*\)) indicate statistical significance (p<0.05). This identifies which specific models drive the high CV values seen in Figure 2.

Reconstruction approach choice influences inferred extinction dynamics

Table S5. Kendall’s Tau Correlation Coefficients

Kendalls Tau (\(\tau\)) values quantifying the amount of agreement between reconstruction approach as to the main driver of extinctions. This is determined based on the mean absolute difference rankings. Significant correlations (p<0.05) are denoted with an asterisk (*). The variability in \(\tau\) values across models for the same metric highlights how reconstruction choice can lead to divergent interpretations of extinction drivers in network structure.

In [5]:

readr::read_csv("tables/kendall_tau_coefficients.csv") %>%
kable()

Rows: 288 Columns: 4
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (3): Model1, Model2, metric
dbl (1): tau

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

Model1	Model2	tau	metric
ADBM	ADBM	1.0000000	NDCI
ATN	ADBM	0.3076923	NDCI
Body-size ratio	ADBM	0.2307692	NDCI
Niche	ADBM	0.0769231	NDCI
PFIM	ADBM	-0.2051282	NDCI
Random	ADBM	0.4358974	NDCI
ADBM	ATN	0.3076923	NDCI
ATN	ATN	1.0000000	NDCI
Body-size ratio	ATN	0.1025641	NDCI
Niche	ATN	0.1538462	NDCI
PFIM	ATN	-0.0256410	NDCI
Random	ATN	-0.1538462	NDCI
ADBM	Body-size ratio	0.2307692	NDCI
ATN	Body-size ratio	0.1025641	NDCI
Body-size ratio	Body-size ratio	1.0000000	NDCI
Niche	Body-size ratio	0.1282051	NDCI
PFIM	Body-size ratio	0.2564103	NDCI
Random	Body-size ratio	0.1794872	NDCI
ADBM	Niche	0.0769231	NDCI
ATN	Niche	0.1538462	NDCI
Body-size ratio	Niche	0.1282051	NDCI
Niche	Niche	1.0000000	NDCI
PFIM	Niche	-0.1538462	NDCI
Random	Niche	0.1282051	NDCI
ADBM	PFIM	-0.2051282	NDCI
ATN	PFIM	-0.0256410	NDCI
Body-size ratio	PFIM	0.2564103	NDCI
Niche	PFIM	-0.1538462	NDCI
PFIM	PFIM	1.0000000	NDCI
Random	PFIM	-0.0512821	NDCI
ADBM	Random	0.4358974	NDCI
ATN	Random	-0.1538462	NDCI
Body-size ratio	Random	0.1794872	NDCI
Niche	Random	0.1282051	NDCI
PFIM	Random	-0.0512821	NDCI
Random	Random	1.0000000	NDCI
ADBM	ADBM	1.0000000	NDTI
ATN	ADBM	-0.0129035	NDTI
Body-size ratio	ADBM	0.0769231	NDTI
Niche	ADBM	0.0000000	NDTI
PFIM	ADBM	0.1794872	NDTI
Random	ADBM	0.0512821	NDTI
ADBM	ATN	-0.0129035	NDTI
ATN	ATN	1.0000000	NDTI
Body-size ratio	ATN	0.0129035	NDTI
Niche	ATN	0.3225874	NDTI
PFIM	ATN	0.0387105	NDTI
Random	ATN	0.0645175	NDTI
ADBM	Body-size ratio	0.0769231	NDTI
ATN	Body-size ratio	0.0129035	NDTI
Body-size ratio	Body-size ratio	1.0000000	NDTI
Niche	Body-size ratio	0.1025641	NDTI
PFIM	Body-size ratio	-0.2307692	NDTI
Random	Body-size ratio	0.3589744	NDTI
ADBM	Niche	0.0000000	NDTI
ATN	Niche	0.3225874	NDTI
Body-size ratio	Niche	0.1025641	NDTI
Niche	Niche	1.0000000	NDTI
PFIM	Niche	0.2051282	NDTI
Random	Niche	0.0256410	NDTI
ADBM	PFIM	0.1794872	NDTI
ATN	PFIM	0.0387105	NDTI
Body-size ratio	PFIM	-0.2307692	NDTI
Niche	PFIM	0.2051282	NDTI
PFIM	PFIM	1.0000000	NDTI
Random	PFIM	0.0512821	NDTI
ADBM	Random	0.0512821	NDTI
ATN	Random	0.0645175	NDTI
Body-size ratio	Random	0.3589744	NDTI
Niche	Random	0.0256410	NDTI
PFIM	Random	0.0512821	NDTI
Random	Random	1.0000000	NDTI
ADBM	ADBM	1.0000000	connectance
ATN	ADBM	0.6410256	connectance
Body-size ratio	ADBM	0.5128205	connectance
Niche	ADBM	0.2051282	connectance
PFIM	ADBM	-0.0256410	connectance
Random	ADBM	0.5897436	connectance
ADBM	ATN	0.6410256	connectance
ATN	ATN	1.0000000	connectance
Body-size ratio	ATN	0.7692308	connectance
Niche	ATN	0.0512821	connectance
PFIM	ATN	0.1282051	connectance
Random	ATN	0.5897436	connectance
ADBM	Body-size ratio	0.5128205	connectance
ATN	Body-size ratio	0.7692308	connectance
Body-size ratio	Body-size ratio	1.0000000	connectance
Niche	Body-size ratio	-0.0256410	connectance
PFIM	Body-size ratio	0.1538462	connectance
Random	Body-size ratio	0.5641026	connectance
ADBM	Niche	0.2051282	connectance
ATN	Niche	0.0512821	connectance
Body-size ratio	Niche	-0.0256410	connectance
Niche	Niche	1.0000000	connectance
PFIM	Niche	0.0000000	connectance
Random	Niche	0.3076923	connectance
ADBM	PFIM	-0.0256410	connectance
ATN	PFIM	0.1282051	connectance
Body-size ratio	PFIM	0.1538462	connectance
Niche	PFIM	0.0000000	connectance
PFIM	PFIM	1.0000000	connectance
Random	PFIM	0.2307692	connectance
ADBM	Random	0.5897436	connectance
ATN	Random	0.5897436	connectance
Body-size ratio	Random	0.5641026	connectance
Niche	Random	0.3076923	connectance
PFIM	Random	0.2307692	connectance
Random	Random	1.0000000	connectance
ADBM	ADBM	1.0000000	generality
ATN	ADBM	0.2820513	generality
Body-size ratio	ADBM	0.0769231	generality
Niche	ADBM	0.2307692	generality
PFIM	ADBM	0.3846154	generality
Random	ADBM	0.0256410	generality
ADBM	ATN	0.2820513	generality
ATN	ATN	1.0000000	generality
Body-size ratio	ATN	0.4358974	generality
Niche	ATN	0.3333333	generality
PFIM	ATN	0.3846154	generality
Random	ATN	-0.1794872	generality
ADBM	Body-size ratio	0.0769231	generality
ATN	Body-size ratio	0.4358974	generality
Body-size ratio	Body-size ratio	1.0000000	generality
Niche	Body-size ratio	0.0769231	generality
PFIM	Body-size ratio	0.1282051	generality
Random	Body-size ratio	-0.0256410	generality
ADBM	Niche	0.2307692	generality
ATN	Niche	0.3333333	generality
Body-size ratio	Niche	0.0769231	generality
Niche	Niche	1.0000000	generality
PFIM	Niche	0.3333333	generality
Random	Niche	0.1794872	generality
ADBM	PFIM	0.3846154	generality
ATN	PFIM	0.3846154	generality
Body-size ratio	PFIM	0.1282051	generality
Niche	PFIM	0.3333333	generality
PFIM	PFIM	1.0000000	generality
Random	PFIM	-0.1282051	generality
ADBM	Random	0.0256410	generality
ATN	Random	-0.1794872	generality
Body-size ratio	Random	-0.0256410	generality
Niche	Random	0.1794872	generality
PFIM	Random	-0.1282051	generality
Random	Random	1.0000000	generality
ADBM	ADBM	1.0000000	trophic_level
ATN	ADBM	-0.2820513	trophic_level
Body-size ratio	ADBM	-0.1538462	trophic_level
Niche	ADBM	0.0256410	trophic_level
PFIM	ADBM	-0.1282051	trophic_level
Random	ADBM	-0.0512821	trophic_level
ADBM	ATN	-0.2820513	trophic_level
ATN	ATN	1.0000000	trophic_level
Body-size ratio	ATN	0.4615385	trophic_level
Niche	ATN	0.2307692	trophic_level
PFIM	ATN	-0.1794872	trophic_level
Random	ATN	-0.1538462	trophic_level
ADBM	Body-size ratio	-0.1538462	trophic_level
ATN	Body-size ratio	0.4615385	trophic_level
Body-size ratio	Body-size ratio	1.0000000	trophic_level
Niche	Body-size ratio	0.0000000	trophic_level
PFIM	Body-size ratio	0.1025641	trophic_level
Random	Body-size ratio	-0.0256410	trophic_level
ADBM	Niche	0.0256410	trophic_level
ATN	Niche	0.2307692	trophic_level
Body-size ratio	Niche	0.0000000	trophic_level
Niche	Niche	1.0000000	trophic_level
PFIM	Niche	-0.4358974	trophic_level
Random	Niche	-0.1538462	trophic_level
ADBM	PFIM	-0.1282051	trophic_level
ATN	PFIM	-0.1794872	trophic_level
Body-size ratio	PFIM	0.1025641	trophic_level
Niche	PFIM	-0.4358974	trophic_level
PFIM	PFIM	1.0000000	trophic_level
Random	PFIM	0.4102564	trophic_level
ADBM	Random	-0.0512821	trophic_level
ATN	Random	-0.1538462	trophic_level
Body-size ratio	Random	-0.0256410	trophic_level
Niche	Random	-0.1538462	trophic_level
PFIM	Random	0.4102564	trophic_level
Random	Random	1.0000000	trophic_level
ADBM	ADBM	1.0000000	vulnerability
ATN	ADBM	0.2307692	vulnerability
Body-size ratio	ADBM	0.2307692	vulnerability
Niche	ADBM	-0.3846154	vulnerability
PFIM	ADBM	0.0000000	vulnerability
Random	ADBM	0.4615385	vulnerability
ADBM	ATN	0.2307692	vulnerability
ATN	ATN	1.0000000	vulnerability
Body-size ratio	ATN	0.0769231	vulnerability
Niche	ATN	0.0769231	vulnerability
PFIM	ATN	-0.1025641	vulnerability
Random	ATN	0.0512821	vulnerability
ADBM	Body-size ratio	0.2307692	vulnerability
ATN	Body-size ratio	0.0769231	vulnerability
Body-size ratio	Body-size ratio	1.0000000	vulnerability
Niche	Body-size ratio	0.1794872	vulnerability
PFIM	Body-size ratio	0.1025641	vulnerability
Random	Body-size ratio	0.3076923	vulnerability
ADBM	Niche	-0.3846154	vulnerability
ATN	Niche	0.0769231	vulnerability
Body-size ratio	Niche	0.1794872	vulnerability
Niche	Niche	1.0000000	vulnerability
PFIM	Niche	0.3076923	vulnerability
Random	Niche	-0.3589744	vulnerability
ADBM	PFIM	0.0000000	vulnerability
ATN	PFIM	-0.1025641	vulnerability
Body-size ratio	PFIM	0.1025641	vulnerability
Niche	PFIM	0.3076923	vulnerability
PFIM	PFIM	1.0000000	vulnerability
Random	PFIM	0.2820513	vulnerability
ADBM	Random	0.4615385	vulnerability
ATN	Random	0.0512821	vulnerability
Body-size ratio	Random	0.3076923	vulnerability
Niche	Random	-0.3589744	vulnerability
PFIM	Random	0.2820513	vulnerability
Random	Random	1.0000000	vulnerability
ADBM	ADBM	1.0000000	Link
Body-size ratio	ADBM	0.3589744	Link
ATN	ADBM	0.4871795	Link
Niche	ADBM	-0.2051282	Link
PFIM	ADBM	0.1794872	Link
Random	ADBM	0.3589744	Link
ADBM	Body-size ratio	0.3589744	Link
Body-size ratio	Body-size ratio	1.0000000	Link
ATN	Body-size ratio	0.5641026	Link
Niche	Body-size ratio	-0.1282051	Link
PFIM	Body-size ratio	0.3076923	Link
Random	Body-size ratio	0.0256410	Link
ADBM	ATN	0.4871795	Link
Body-size ratio	ATN	0.5641026	Link
ATN	ATN	1.0000000	Link
Niche	ATN	0.0000000	Link
PFIM	ATN	0.1282051	Link
Random	ATN	0.1025641	Link
ADBM	Niche	-0.2051282	Link
Body-size ratio	Niche	-0.1282051	Link
ATN	Niche	0.0000000	Link
Niche	Niche	1.0000000	Link
PFIM	Niche	-0.3589744	Link
Random	Niche	0.0256410	Link
ADBM	PFIM	0.1794872	Link
Body-size ratio	PFIM	0.3076923	Link
ATN	PFIM	0.1282051	Link
Niche	PFIM	-0.3589744	Link
PFIM	PFIM	1.0000000	Link
Random	PFIM	0.3076923	Link
ADBM	Random	0.3589744	Link
Body-size ratio	Random	0.0256410	Link
ATN	Random	0.1025641	Link
Niche	Random	0.0256410	Link
PFIM	Random	0.3076923	Link
Random	Random	1.0000000	Link
ADBM	ADBM	1.0000000	Node
Body-size ratio	ADBM	0.1282051	Node
ATN	ADBM	0.3076923	Node
Niche	ADBM	0.5641026	Node
PFIM	ADBM	0.6153846	Node
Random	ADBM	0.5128205	Node
ADBM	Body-size ratio	0.1282051	Node
Body-size ratio	Body-size ratio	1.0000000	Node
ATN	Body-size ratio	0.7692308	Node
Niche	Body-size ratio	0.5641026	Node
PFIM	Body-size ratio	0.2564103	Node
Random	Body-size ratio	0.5641026	Node
ADBM	ATN	0.3076923	Node
Body-size ratio	ATN	0.7692308	Node
ATN	ATN	1.0000000	Node
Niche	ATN	0.7435897	Node
PFIM	ATN	0.4871795	Node
Random	ATN	0.7435897	Node
ADBM	Niche	0.5641026	Node
Body-size ratio	Niche	0.5641026	Node
ATN	Niche	0.7435897	Node
Niche	Niche	1.0000000	Node
PFIM	Niche	0.3333333	Node
Random	Niche	0.7435897	Node
ADBM	PFIM	0.6153846	Node
Body-size ratio	PFIM	0.2564103	Node
ATN	PFIM	0.4871795	Node
Niche	PFIM	0.3333333	Node
PFIM	PFIM	1.0000000	Node
Random	PFIM	0.5897436	Node
ADBM	Random	0.5128205	Node
Body-size ratio	Random	0.5641026	Node
ATN	Random	0.7435897	Node
Niche	Random	0.7435897	Node
PFIM	Random	0.5897436	Node
Random	Random	1.0000000	Node

Constraining connectance decreases observed differences between models

It has been argued that connectance constrains emerging network properties (Poisot and Gravel 2014; Strydom et al. 2021). To assess if constraining the connectance of all networks would result in similar network structures we actively constrained the connectance of all networks by randomly downsampling them to the target connectance i.e., the PFIM, ATN, ADBM, and Body-size ratio models had links randomly prunes until they reached the target connectance. Note that in some instances the initial connectance of the network may have already been below the target connectance and in these instances the network was used ‘as is’. We repeated both the multivariate and temporal trajectory analyses for these connectance constrained networks.

Figure S5. Constraining connectance decreases observed differences in network structure

Although there is still a significant effect of model type on network structure (Pillai’s trace = 1.64, approximate \(F_{30,10080}\) = 164.5, \(p<\) 0.001) we can see that the visual clustering is much less dramatic than for the unconstrained networks (Figure 1 in the main text). However we do not see connectance as the main driver of differences in network structure in the mulitvariate space – note the connectance is more strongly correlated with LDA axis 2 than with LDA axis 1 indicating that although it ‘reduces’ the differences in network structure between models it is not the main driver in the observed differences.

Figure S6. Connectance increases the signal of time

When we look at the temporal changes in network structure there is qualitatively more agreement with the changes in network structure over time than for the unconstrained networks (Figure 2A in the main text) i.e., we often see most models indicating similar changes in structure over time, although the magnitude of these changes remain different. Note that here we made no assumptions as to the expected changes in connectance over the extinction time and so there is a ‘random’ connectance signal across time - thus we can assume that the signal of temporal changes in network structure here still retain some sort of time \(\times\) model signal. Indeed when we look at the two-way factorial ANOVA (panel B) we can see that model type is still the dominant determinant of network structure, however we see a greater source of variance coming from time (notably for vulnerability time becomes the dominant driver of variance).

Overall these results suggest that the specification of network connectance will impact the resulting overall network structure, however we still get distinct model-specific signature. This result is two-fold. Firstly it showcases that for any model where connectance is one of the parameters in the model there is a greater risk of enforcing differences or similarities in network structure depending on how connectance is specified – e.g., in this instance any assumptions we make on expected change in connectance over time will drive at least in part any inferences we make about changes in network structure for both Niche and Random networks. Secondly, despite connectance-driven constraints we still see model-specific differences in structure which re-enforces that these models are still encoding fundamentally different processes that determine where links are present within the network.

References

Poisot, Timothée, and Dominique Gravel. 2014. “When Is an Ecological Network Complex? Connectance Drives Degree Distribution and Emerging Network Properties.” PeerJ 2 (February): e251. https://doi.org/10.7717/peerj.251.

Strydom, Tanya, Michael D. Catchen, Francis Banville, et al. 2021. “A Roadmap Towards Predicting Species Interaction Networks (Across Space and Time).” Philosophical Transactions of the Royal Society B: Biological Sciences 376 (1837): 20210063. https://doi.org/10.1098/rstb.2021.0063.