How can similar species both survive without one species driving the other into extinction?

Spatial Heterogeneity

Most resource partitioning arguments proposed to explain species diversity assume that the environment is heterogeneous. Coexistence is possible because species have different resource requirements and are specialized to succeed on particular patch types. Diversity is maintained through the presence of an array of patch types. As our ability to measure and quantify environmental variation has improved, we have developed clear evidence that the environment is heterogeneous with regard to resources that are critical to plants. To date, however, we have not critically assessed if the degree of heterogeneity recorded by our instruments is representative of the resource environment perceived by the plants themselves (Bazzaz, 1996). Despite this uncertainty, there is now clear evidence that, at least on the broad scale, species do partition themselves along resource axes. Studies that relate species' distributions to multiple environmental factors using either direct or indirect gradient analysis have identified numerous soil-related axes of specialization, e.g., R. H. Whittaker's 1956 work on the vegetation of the Great Smoky Mountains. Research on experimentally produced plant communities also supports the notion that species may be differentiated along soil resource axes (see studies discussed in Bazzaz, 1996).

Although plants themselves can influence their local environment and thus produce intrinsic heterogeneity within the community, much spatial resource heterogeneity is generated by extrinsic factors, the most common of which is disturbance. In fact, disturbance is judged both by change in availability of resources and by perception of that change. Disturbance events create heterogeneity in resource availability at multiple spatial and temporal scales, and these different scales can allow species to partition the environment in more ways. Awareness of the importance of disturbance events for understanding the maintenance of plant species diversity was heightened when Peter Grubb (1977) proposed that regeneration traits of species should provide a critical axis for differentiation, the so-called regeneration niche. Much of the research on partitioning along this regeneration axis has focused on the broad-scale effects of canopy disturbance, addressing differences in conditions and plant responses between gaps and closed forest. Gap creation generates variability in the light environment and has been advanced as a major factor in increasing diversity at the stand level. Broad guilds of plant species that respond similarly to gap formation (light demanding vs shade tolerant) have been identified in numerous forest communities where vertical canopy stratification creates a wide range of light microenvironments (Whitmore, 1989). More recently, sophisticated mathematical techniques have been used to classify both physiological and demographic responses of tree seedlings to light availability more precisely (see SORTIE papers of Steve Pacala, Richard Kobe, and others). Much of this research into species' responses to total light quantity has found considerable overlap in response, and species' partitioning of the light microenvironment is not usually sufficient to explain maintenance of species diversity, particular in communities with high numbers of co-occurring species, such as lowland tropical rain forest.

An additional mechanism for the maintenance of diversity arose from the gap partitioning hypothesis, originally developed by Ricklefs (1977) and applied specifically to the tropical rain forest by Julie Denslow (1980). Here, finer-scale variation in resource availability across the gap–understory continuum was suggested as a possible mechanism for coexistence in multispecies communities. Unfortunately, this hypothesis has not yet been validated in tropical rain forests, where we lack a good understanding of the maintenance of species diversity. Experimental work by Tim Sipe and F. A. Bazzaz in temperate forests in eastern North America, in contrast, provides some support for the gap partitioning hypothesis. In experimentally created gaps at Harvard Forest in central Massachusetts, we found clear micro-environmental differences between gaps of different sizes (especially with regard to light). We compared morphological and physiological responses of three co-occurring maple (Acer) species to this light variation and found clear evidence that the species differed in their response to gap size. Many response variables showed significant differences between large gaps, small gaps, and understory plots, and these differences often varied between species, creating distinct species' preferences for canopy gap environment (Figure 4). We also observed substantial variation in light availability between different parts of canopy gaps, due to seasonal and diurnal trends in solar patterns. This variation, however, was rarely reflected in differences in species' responses to position within the gap.

8.3.2 Resource Partitioning and the Facilitation: Collaboration Concepts

Resource partitioning through competition, resource sharing through CMN and a variation in host responses to microbes (including fungi) and soil community feedbacks (Bever et al., 2010) have so far been largely biased by a plant-centric approach. Also, discussions on low-diversity plant communities mainly focus on competitive dominance, inhibition, or positive feedbacks that plant community development might be based on (ibid.). Belowground biotic complexity was found to drive aboveground dynamics through a soil community feedback (Pendergast et al., 2013). Facilitative interactions, that is, beneficial interactions among different organisms, were placed in focus several years ago (van der Heijden and Horton, 2009) and have raised interest in including a new conceptual model in different areas of ecology (Brooker and Callaway, 2009). Mutualism and facilitation have been subsumed under a common concept called ‘beneficial interactions’ (ibid.), which we here propose to name ‘belowground collaboration’, and includes carbon, water and nutrient transfer among organisms of the same and different taxons and taxonomic groups, thus supporting their survival and influencing sustainable development of the ecosystem.

Facilitation has been discussed as paving the way for ecological speciation, enhancing local adaptation, and allowing long-term adaptation to environmental extremes in marginal habitats and in contact zones (Liancourt et al., 2012). Also, a supporting role of the existing diversity both above- and belowground has been found to promote persistence of organisms which, according to current research methods, have been named ‘low-quality fungal partners’ (Hart et al., 2013). Therefore, the distribution of different forms of mycorrhiza in the world’s biomes (Figure 8.2) is subject to change and will determine future distribution, adaptation and survival of large terrestrial biomes.

3.3 Testing the Niche Partitioning Theory with NGS-Based Trophic Networks

The niche partitioning theory is central to our understanding of biodiversity. The term niche partitioning refers to the process by which natural selection drives competing species into different patterns of resource use or different niches (Hector and Hooper, 2002; MacArhur, 1958). This differentiation of ecological niches reduces competition and promotes co-existence between species (Chesson, 2000; Levine and HilleRisLambers, 2009). NGS techniques promise to bring significant changes in our understanding of niche partitioning because they provide information about the entire diet range of a species, while also highlighting new and unexpected trophic links. For instance, Ibanez et al. (2013) carried out a trait-based estimation of the trophic niche width of four species of grasshoppers through choice experiments and NGS study of their diet. They showed that observed trophic niche breadth in generalist herbivorous insects depends on both species-specific food preferences and habitat diversity, and that it is not an intrinsic property of the species as usually considered in theoretical studies. Kartzinel et al. (2015) also used NGS techniques, in combination with stable isotope analyses, to investigate trophic niche partitioning among sympatric large mammalian herbivores in Kenya. They observed unsuspected fine-scale resource partitioning even between species within the same trophic guild. Consequently, through this study, NGS methods illuminated mechanisms behind the large diversity of co-existing herbivorous mammal species observed in African savannas.

Here, we used NGS techniques to resolve the trophic interactions in a community of carabid beetles inhabiting European arable landscapes. These beetles significantly contribute to the biological control of pests (Kromp, 1999) but their contribution is unpredictable given their broad diet spectrum including alternative preys such as other natural pest enemies. We used NGS approach in order to investigate changes in carabid diet among 14 common arable species by analyzing the prey DNA contained in their guts. The carabid beetles were sampled in six arable fields in two different agroecosystems in Brittany, France. To cover the whole carabid diet spectrum, four barcode genes were combined, including mitochondrial 16S and COI for characterizing animal prey (Bienert et al., 2012; De Barba et al., 2014) and the chloroplast trnL for detecting consumed plant species (Taberlet et al., 2007). We obtained bipartite ecological networks showing an extensive degree of niche overlapping between carabid species (Fig. 3), confirming their generalist feeding behaviour. The use of NGS also allowed us for the first time to quantify carabid community contribution to ecosystem services (the consumption of several animal and plant pest species) and dis-services (the consumption of other service-providing organisms) (Fig. 4). These findings suggest that important ecological functions as pest regulation and intraguild predation may not be associated with the identity of particular species but are more likely an emergent property of the community probably modulated by extrinsic environmental factors. Further investigations are required for see how important agronomical features such as the cropping system or management intensity impact the structure of the trophic web.

Ontogeny

Intraspecific resource partitioning due to ontogenetic niche shifts occurs in many species. This is probably most obvious when age groups live in different habitats and use different types of food. Examples of such ontogenetic niche shifts include particle-feeding amphibian larvae that turn into carnivorous adult amphibians, immature stages of aquatic insects that turn into adult terrestrial insects, planktonic marine invertebrate larvae that settle down to be sessile adults. Other species live in the same general habitat but use distinct food types depending on size or age class, as for example, fish where the young are gape-limited planktivores or herbivores and the adults are predators. There might also be large ontogenetic changes in diet within the same general food type as for example for predators that start to feed on small prey species but as they grow they can and will include larger prey types in their diet.

Differences in diet over the ontogeny of an animal can have different reasons. Search and handling efficiencies can change as animals grow and gape-limitation eases. Other animals change their diet with growth because predation risk declines with size, allowing them to use previously risky habitats. Still other species exhibit a fundamental change in habitat use as a result of life history changes or dispersal (e.g., marine invertebrates with planktonic larvae that settle to become sessile adults).

III Resource Allocation Patterns of Clonal Species in Homogeneous Growing Conditions

h. Nocturnal vs. Diurnal Feeding

The basis for resource partitioning along the 24-hour temporal axis is based largely on visual and other sensory modalities. These systems are beyond the scope of this review, but we mention this major axis of ecological radiation in reef fishes, in part because the implications of feeding in low light at night, with respect to the functional morphology of the prey capture apparatus, have not been explored. We note that nocturnal reef fish tend to be either predators of large, elusive prey, mobile benthic prey, or zooplankton. Very few nocturnal fish predators feed on hard-shelled prey (e.g., Diodon) or firmly attached prey.

Allocation and Reproductive Strategies

Biomechanical limitations restrict niche partitioning involving body form. However, plants with different forms may vary the allocation or timing of energetic output, thus increasing the potential for coexistence. Separate species attain their “adult” size at different times by altering growth rates. Unless the total amount of energy available to the individual plant changes, such changes in growth rates must be accompanied by a trade-off in other activities requiring energy. For some species, changes in growth rate are compensated for by changes in wood density: Fast-growing arborescent species generally produce wood of a much lower density than do slower growing species (e.g., balsa versus mahogany; Enquist et al., 1999). This has obvious implications for relative mechanical stability, susceptibility to disease, and species longevity. Similarly, an energetic trade-off exists with regard to the ability to tolerate herbivory, parasitism, and disease (Crawley, 1997). By investing energy in the formation of secondary chemicals, plants can often avoid enemies, albeit at the cost of slower growth and delayed sexual maturity.

Perhaps the area in which different allocation strategies have the strongest effect is within reproduction. Plants can alter both the proportion of resources used in reproduction and the amount of resources allocated to individual disseminules. Such adaptations are often linked to community structure. Large seeds, for example, tend to occur in forest communities, whereas smaller seeds are more common in open or disturbed habitats (Crawley, 1997). Variation in reproductive strategy provides plants with additional ways to partition environmental opportunities and coexist.

The fossil record suggests that reproductive adaptations have permitted the ecological exploration of previously poorly colonized terrestrial environments, starting with the invasion of land, the evolution of the seed, and finally the suite of adaptations involved in the origin of angiosperms (Niklas, 1997). Early land plants possessed pteridophytic reproduction, requiring freestanding water. With evolution of advanced forms of gymnospermy, this dependence was lost, but at the cost of vegetative reproduction, because all nonangiospermous seed-bearing plants were apparently either shrubs or trees capable of only limited vegetative reproduction.

The evolution of the seed habit, specialized dispersal mechanisms, and the capacity for seed dormancy opened up many new habitats in the late Paleozoic and Mesozoic. Together, specialized dispersal and seed dormancy permitted further niche diversification within plant communities. The morphology of dispersal units through the Late Cretaceous suggests that seed dispersal was primarily abiotic. The evolution of large seeds in many angiosperm lineages during the early Cenozoic indicates the increasing importance of animal dispersal. Additionally, the ability of seeds to remain dormant has many important ecological implications, potentially enhancing local coexistence by enabling potential competitors to partition niches in time and space. Furthermore, dormancy enables plants to “wait” until specific environmental conditions are present, such as a rare rain in the desert or a high nutrient defecation following passage through an animal's digestive tract. Dormancy holds the potential for greater dispersal distances and the formation of seed banks, enhancing the geographic ranges of species and their capacity to deal with potentially protracted inclement environmental conditions. Dormancy, however, may also increase the probability of seed predation.

It is only with the angiosperms that asexual vegetative reproduction and the seed habit are found to commonly co-occur within an individual with either an herbaceous or woody growth habit. The possession of rapid sexual life cycles, the potential for an herbaceous habit, and in many cases the potential for vegetative reproduction predisposed angiosperms to outdiversify and potentially outcompete other lineages in multiple ways in response to environmental changes especially in the Cenozoic. These traits allowed for further invasion of new habitats, such as desert and arctic/alpine environments.

Manipulative Field Experiments

Recognizing the limitation of observational studies of niche partitioning as evidence for competition, ecologists turned to manipulative field experiments, removing species and monitoring changes in the abundance of others. Reviews of such experiments show that when species are suspected to compete (e.g., because of overlap in resource requirements and habitat), they often do compete. For instance, Hairston noted that in the Great Smoky Mountains, two species of Plethodon salamanders had altitudinal ranges that were nearly mutually exclusive, with only a narrow range of overlap. In contrast, in the Balsam Mountains, altitudinal overlap was extensive. Hairston hypothesized that this was due to stronger competition in the Smokies, and using reciprocal transplants and removals, he demonstrated that competition (due to aggression) was indeed much stronger in the narrow overlap zone in the Smokies.

Unfortunately, few field experiments have been directly tied to mechanistic models of competition. Moreover, such experiments tend to focus on species that already coexist, at least to a degree; removals then assess the magnitude of competition, given coexistence, rather than coexistence or exclusion per se. One study that directly addressed coexistence was performed by Bengtsson, who studied three species in the zooplankton genus Daphnia (which compete exploitatively for algal food) on the coast of the Baltic Sea. Observational studies suggested that usually just one, or more infrequently two, species was present in any given pool. Bengtsson added all possible combinations of the three species to artificial pools and found that there were no extinctions in the single-species pools. Yet high extinction rates occurred in the two- and three-species sets. This study directly demonstrates the importance of competition in constraining coexistence in natural conditions.