SIZE-BASED FOOD WEBS AND ECOSYSTEM FUNCTIONING
Body size plays a significant role in aquatic food webs based on the general rule that larger predators eat smaller preys (Barnes et al., 2010; Cohen et al., 2003; France et al., 1998; Jennings et al., 2002a; Jonsson et al., 2005). Using the regression slope of TP against body mass in Figure 2a, we calculated the predator–prey body mass ratio (PPMR) using Eq. 10(1/slope) (see Jennings et al. 2002). The PPMR was 105.21 for all pooled data and 103.03 when data on gastropods were excluded. These ratios were much higher than the literature average (100.96 ± 0.05SE) for lake invertebrates and comparable to lake ectothermic vertebrates (104.15 ± 0.18SE; Brose et al., 2006). Gastropods exhibited the most dominant biomass abundance in our study communities. They feed on small-sized preys, such as benthic microalgae and microplankton-derived POM, and reach their largest size among macroinvertebrates based on predator resistance due to their armored traits. This explains why PPMR is much higher for the whole community of coastal macroinvertebrates in Lake Biwa. Such predator-resistant taxa are often decoupled from the rest of the food web, as observed in a trophic cul-de-sac (Bishop et al., 2007), which makes the scaling allometry of size-based food webs non-linear (Jennings et al., 2002b).
Although our field research stresses the importance of unraveling BEF relationships in real food webs with multitrophic interactions, such a correlational approach limits our ability to fully understand the causal mechanisms that explain why more diverse communities show higheri TP. There are alternative but not mutually exclusive mechanisms for this. One possible mechanism is that the overall basal energy gained by consumers is greater in more diverse communities, and another is that the trophic transfer efficiency (TTE) is higher in more diverse communities.
Here, approximation allows us to test the former hypothesis. For simplicity, we assume that consumer biomass is trophically transferred at a given rate, irrespective of TPs, within a limited size range (0.03 mg to 1.33 g in dry weight) of benthic macroinvertebrate communities. Generally, TTE is defined as the production rate (biomass multiplied by the turnover rate) for two adjacent TLs (Lindeman, 1942). Based on a theoretical derivation of TTE for communities with a size-specific trophic structure (Andersen et al., 2009), Garcia-Comas et al. (2016) used the prey–predator biomass ratio as a proxy for TTE in aquatic food webs. Considering the small within-lake variations in coastal WT (range = 2.4°C–7.2°C), which is the primary factor affecting metabolic rate, using biomass measurements may be a valid approximation for TTE in our study communities. While setting the TTE at a given rate, we aimed to estimate consumer basal biomass (CBB ), which is defined as a proxy for secondary production needed to support a whole consumer community, which was back-calculated to the primary consumer biomass using the following equation:
CBB = Σ{Bi \(\times\)(1/β )^(TPi -2)} (6)
where CBB is the sum of the primary consumer biomass converted from Bi , the consumer biomass of taxon i with TPi , and β the TTE. For β , we took the minimum (3.7%) and maximum (27.1%) TTE values reported for benthic communities (Jennings et al., 2002b).
Our calculations revealed that CBB significantly increases with increasing biodiversity within the possible TTE range (Figure 4). This means that more diverse communities rely on greater secondary production. Considering that primary productivity did not increase i TP, the underlying mechanism for this may be that an increase in resource use, but not resource availability, allows trophic transfer of more biomass to higher TLs via diverse trophic interactions. This suggests that positive BEF relationships may be more common in many multitrophic systems, where significant biodiversity effects on consumer biomass could not be detected when using i TP as an indicator for ecosystem productivity.
To test another hypothesis, we examined the relationship between TTE andi TP for metacommunities. We estimated the TTE for each local community from slope a of production against TPs for individual taxa (Box S2), according to the allometric scaling empirically formulated by Banse & Mosher (1980) for invertebrates. We estimated the TTE to be 9.77% for a local community, which is within reported literature values for benthic communities, but our estimates were unrealistic for many other communities due to insignificant slopea (Table S6). As aforementioned, a trophic cul-de-sac and allochthonous energy inputs from pelagic food webs may make allometric scaling non-linear and less significant in benthic macroinvertebrate communities, hindering us from testing the VDH under the assumption of size-based food webs.
Benthic macroinvertebrate communities form a part of lake food webs as intermediate consumers linking microalgae and vertebrate predators. The present study was conducted during the winter season when fish predators are absent or inactive in order to minimize their top-down effects on the size and trophic structure of localized food webs. During the productive season, however, coastal benthic food webs are linked with pelagic and adjacent coastal food webs via mobile predators, and eventually, the overall energy flows up to apex predators (e.g., a giant catfish with 3.75 of TP) of the whole lake food web (Okuda et al., 2020). In aquatic food webs, predators also increase their spatial movement scale as TPs increase with increasing body size (McCann et al., 2005; Tucker et al., 2014). Upscaling multitrophic consumer networks that link different ecosystem compartments to the whole ecosystem is a challenging and promising approach to gain better understanding of BEF relationships and achieve improved biodiversity management at the landscape level (Barnes et al., 2018; Eisenhauer et al., 2019; Gounand et al., 2018; Hines et al., 2015; Manning et al., 2019). Future studies should explore how highly compartmentalized and thus heterogeneous coastal food webs integrate mobile predators across spatiotemporal scales, which may allow for more accurate TTE estimates based on a rigid allometric relationship between body mass, biomass, and TP across the entire lake food web that encompasses size ranges from microscopic autotrophs to carnivorous megafauna.