Figure 1.
Food web diagram of the Venice lagoon with 27 nodes or funtional groups.
Colors of flows depict different fishing target (artisanal fisheries in blue, and clam fishery in red) and non-target species (for clam harvesting, in green). Modified from Pranovi et al. [102].
Figure 2.
Distribution of food web model used in this study.
Models are divided into Eastern Pacific (black), Western Atlantic (red), Eastern Atlantic (blue), North and Baltic Seas (orange), Mediterranean Sea (green), Indian Ocean (purple) and Eastern Pacific (pink). Each model is numbered on the graph according to its number in Table S1, where details and references to each model area are given.
Table 1.
Ecological and fisheries related indicators used in this comparison.
Table 2.
Significance between ecosystems indicators when grouping data by ecosystem traits.
Figure 3.
Boxplot of significant differences of food web ecological indicators by ecosystem type.
The smallest observation (minimum), lower quartile, median, upper quartile, largest observation (maximum) and outliers are indicated. The boxes are drawn with widths proportional to the square-roots of the number of observations in each class. TST = total systems throughput (t.km−2.year−1), TB community = total biomass of the community (t.km−2), and mean EE = mean ecotrophic efficiency (proportion).
Table 3.
Significance of individual indicator differences between levels of ecosystems traits.
Figure 4.
Boxplot of significant differences of food web ecological indicators by depth class.
The smallest observation (sample minimum), lower quartile, median, upper quartile, largest observation (sample maximum) and outliers are indicated. The boxes are drawn with widths proportional to the square-roots of the number of observations in each class. FD/TST = flow to detritus/total systems throughput (proportion), R/TST = respiration/total systems throughput (proportion), TEm = mean transfer efficiency (%) and mean EE = mean ecotrophic efficiency (proportion).
Figure 5.
Boxplot of significant differences of food web ecological by size class.
The smallest observation (sample minimum), lower quartile, median, upper quartile, largest observation (sample maximum) and outliers are indicated. The boxes are drawn with widths proportional to the square-roots of the number of observations in each class. Mean EE = mean ecotrophic efficiency (proportion), Ex/TST = export/total systems throughput (proportion) and FCI = Finn cycling index.
Figure 6.
Boxplot of significant differences of food web ecological by ocean basin.
The smallest observation (sample minimum), lower quartile, median, upper quartile, largest observation (sample maximum) and outliers are indicated. The boxes are drawn with widths proportional to the square-roots of the number of observations in each class. R/TST = respiration/total systems throughput, Ex/TST = export/total systems throughput (proportion), mean TE = mean transfer efficiency, FD/TST = flow to detritus/total systems throughput (proportion), MTLco = mean trophic level of the community, Q/TST = consumption/total systems throughput (proportion), SOI = systems omnivory index, IFO = Internal flow overhead (%), A/C = relative ascendency (%), O/C = relative overhead (%).
Figure 7.
Key ecological roles of functional groups of marine food web models.
a) Keystone (KS ≥0; black circles) and dominant groups (KD ≥0.7; grey circles), respectively, in terms of absolute overall effect () for each food web. Open dots represent non key functional groups. b) The 105 trophic groups franking first in terms of absolute overall effect within each food web model (Fig. 2, Table S1). The figure shows the trophic level (TL) vs. the fraction of top-down effect (td%). Groups identified as keystones are represented in black symbols and dominant groups are reported in grey symbols, respectively, whereas open circles represent non key functional groups. Groups are highlighted for both keystones and dominant: birds (star within square), marine mammals (triangles), sharks and rays (squares), top-predators (romboid), primary producers (crossed squares), other groups (circles). Large squares with error bars identify mean+/−SD for all keystones and dominants identified in the 105 models.
Figure 8.
Proportion of key functional groups by food web ecological traits and by exploitation.
Graphs report proportion of Keystone (KS ≥0) and dominant groups (KD ≥0.7) by a) ecosystem size, b) depth, c) ecosystem type, d) ocean basin and e) fishing category. Only traits showing significantly different patterns (on the basis of total Chi-squared) are reported (***, ** respectively p<0.01, p<0.05). Main contribution to Chi-squared are highlighted by asterisk (see also Table 4).
Table 4.
Chi-squared obtained from the Correspondence Analysis between traits (fishing, size, ecosystem type, latitude, longitude, depth and year) and key functional groups.
Figure 9.
Boxplot of significant differences of food web ecological traits by exploitation and of fishing indicators.
The smallest observation (sample minimum), lower quartile, median, upper quartile, largest observation (sample maximum) and outliers are indicated. The boxes are drawn with widths proportional to the square-roots of the number of observations in each class. PP/P = primary production/total production (proportion), meanEE = mean ecotrophic efficiency (proportion), TBco = total biomass of the community (t.km−2), Mean TLc = mean trophic level of the catch, Lindex = loss in production index, and Psust = probability of being sustainably fished.