Often coprozoic, they are in general most common in organic-rich environments of low oxygen content (Margulis et al., 1989). Kinetoplastids are closely related to a much better known group, the euglenids. Euglenids are commonly characterized by the green photosynthetic protist Euglena; however, most organisms in this group are heterotrophic, and there is reason to believe that the euglenid acquisition of plastids was a relatively recent event involving the incorporation of a green algal symbiont or chloroplast (Gibbs, 1981; Whatley, 1981). Therefore, for the purposes of this argument, undue weight should not be accorded to the euglenid plastid. Kinetoplastids and euglenids both provide a perspective on early mitochondria-bearing eukaryotes as aerobic, bacteriophagous heterotrophs capable of engulfing potential symbionts and able to thrive at relatively low PO2. Such organisms are plausible candidates for the types of eukaryotes that might have radiated during the period when PO2 stood at 1 to 2% PAL. Eukaryotes may have gained ecological prominence as micropredators and scavengers well before they became important as photoautotrophs.
Other sequenced organisms that branch earlier than the main algal limbs include amebomastigotes (common soil and water organisms that live as flagellates under low-nutrient conditions but as amebas in nutrient-rich environments); entamebas (often parasitic, some without mitochondria, others with bacterial symbionts); and cellular slime molds.
The "crown" of the eukaryote tree is studded with photosynthetic members (Sogin et al., 1989). As argued by Cavalier-Smith (1987), there is no ultrastructural reason why organisms capable of engulfing mitochondrial precursors could not also have incorporated cyanobacteria. The diversity of photosynthetic eukaryotes certainly indicates that once plastid acquisition became feasible, a number of protistan lineages acquired them. The barrier to protoplastid acquisition could have been environmental, and the relatively late rise of PO2 from 1 to 2% PAL to >15% PAL provides a plausible explanation for the observed phylogeny. Until this increase in PO2, the availability of NO3- may have been severely limited, giving ecological advantage to free-living, nitrogen-fixing cyanobacteria. Once oxygen levels increased, however, nitrogenase activity was inhibited (Towe, 1985) and odd nitrogen availability increased by more than an order of magnitude. Eukaryotic phytoplankton and benthos could radiate to become important parts of most surficial ecosystems, except for stressed environments such as the upper intertidal zone of restricted seaways—environments well represented in the Proterozoic fossil records. Surprisingly, increasing paleontological data suggest that the "big bang" of higher eukaryotic evolution did not occur until 1200 to 1000 Ma (Knoll, 1992b). Earlier algae, including those that formed the 2100 Ma fossils, apparently belonged to extinct lineages.
Given the requirement that three independent criteria must be satisfied, we cannot unequivocally accept the Cloud hypothesis as it relates to early Proterozoic evolution. The geochemical and paleontological record is improving rapidly. It now appears to fit well with the molecular phylogenetic record, and it is at least consistent with paleontological observation. A modified Cloud model provides the best framework for the available biological and geochemical data.
In the Cloud model, the other principal period of linked environmental and biological evolution is the end of the Proterozoic Eon, when further increases in PO2 are thought to have allowed the evolution of macroscopic animals. Here the relative strengths of the three lines of evidence are reversed. The paleontological data are quite extensive; it is the geochemical evidence that is consistent and suggestive rather than compelling. Details of latest Proterozoic Earth history are presented in Knoll (1992a) and Derry et al. (1992); therefore, only a brief synopsis is provided here.
The radiation of macroscopic animals was the cardinal evolutionary event of the Neoproterozoic Era. In 1968, Cloud argued that no unequivocal animal remains are present in rocks older than the great Varangian (ca. 610 to 590 Ma) ice age, and in the ensuing 20 years, a great deal of detailed stratigraphic research has strengthened this conclusion. Hofmann et al. (1990) have reported small, simple disks of probable metazoan origin in immediately subVaranger strata from northwestern Canada, but macroscopic animal remains and traces are otherwise conspicuously absent from pre-Varanger successions. On six continents, large and diverse, but structurally simple, animals—the so-called Ediacaran fauna (Figure 1.4C)—first appear in strata that lie above Varangian glaciogenic rocks (Runnegar, 1982a; Glaessner, 1984). Metazoan trace fossils have a parallel history of appearance and diversification (Crimes, 1987).
It is important to note the terms "macroscopic" and "large." Ancestral microscopic metazoans may well have evolved significantly earlier, but left no fossil record. Indeed, it is biologically appealing to posit some sort of metazoan prehistory. The significant point, however, is