The effects assessment of metals is mainly based on data of single metals on single species, thereby not accounting for effects of metal mixtures or effects of species interactions. Both of these effects were tested in combination, thereby hypothesizing that the sensitivity of a community to synergistic mixture toxicity depends on the correlation of single-species sensitivities among the single metals. Single-metal and metal-mixture effects were tested in full concentration–response experiments (fixed ray of 1:1:3 and 5:1:13 mass ratio Ni:Cu:Zn) on eight single freshwater algal species and 14 algal communities of four species each. The mean correlation of single-species median effect concentrations among the single metals (Ni–Cu, Cu–Zn, and Zn–Ni) for all species in a community ($\bar{r}$) ranged from −0.4 to 0.9 among the communities; most of these (12/14) were positive. Functional endpoints (total biomass) were overall less sensitive than structural endpoints (Bray-Curtis similarity index) for communities with positively correlated single-species sensitivities among the single metals ($\bar{r}>0.33$), suggesting that such correlations indicate functional redundancy under metal-mixture stress. Antagonistic metal-mixture interactions were predominantly found in single species, whereas metal-mixture interactions were antagonistic and surprisingly synergistic for the communities, irrespective of the reference mixture model used (concentration addition or independent action). The mixture interactions close to the carrying capacity (day 7) of communities gradually shifted from antagonism to more noninteractions with increasing correlation of single-species sensitivities among the single metals. Overall, this suggests that functional redundancy under mixed-metal stress comes at the cost of reduced biodiversity and that synergisms can emerge at the community level without any synergisms on the single-species level. Environ Toxicol Chem 2023;42:2666–2683. © 2023 SETAC.