From soil to the stem interior. Microbial habitats in Myrtillocactus

Two cacti in the same landscape

In 2012, Citlali Fonseca-García and colleagues sampled natural populations of two species at El Magueyal and San Felipe in the Mexican state of Guanajuato. The sites were approximately 130 km apart. Collections took place at the ends of the dry and rainy seasons, in May and October respectively. Three healthy plants of each species were selected at every site and sampling date. The study therefore involved 24 plants. This arrangement allowed the researchers to consider host identity, location, season and the sampled plant compartment as separate influences on the recovered microbial material [Fonseca-García et al. 2016].

Both cacti grew under comparable climatic conditions and in sandy loam soils. Their local coexistence was important to the comparison. When two species occupy distant regions, differences in their microbiomes can reflect both plant characteristics and different pools of microorganisms in the surroundings. Here the hosts encountered locally similar source material while retaining their own structure and physiology. The researchers could therefore examine how much of their microbial communities was shared and identify the compartments in which a clearer association with cactus species appeared [Fonseca-García et al. 2016].

Six sampling compartments

The design began with soil collected one metre from a plant and loose soil within 10 cm of its roots. Another sample represented the rhizosphere, the microorganisms associated with soil adhering firmly to the root. Very little soil remained attached in these dry, sandy sites. Including loose root-zone soil preserved an intermediate sampling category between the surroundings and the root itself. These distinctions describe how material was actually collected and provide a basis for comparing the gradual differentiation of communities in the vicinity of the plant [Fonseca-García et al. 2016].

Above ground, the researchers sampled stem surfaces, termed the phyllosphere in the paper. In these cacti the term referred to the green stem surface. Material from the root interior and the stem interior was prepared separately, representing two parts of the endosphere. There were six sampling categories altogether. The design captured both above-ground versus below-ground position and the distinction between surfaces and internal tissues. Similarity between the two cacti could thus be assessed in corresponding compartments while keeping the environment inside each organ separate. These boundaries were important to the interpretation of the results [Fonseca-García et al. 2016].

What the DNA revealed

Genetic material was used to detect bacteria, archaea and fungi. A region of the 16S rRNA gene was sequenced for the first two groups, and the ITS2 region for fungi. Similar reads were grouped into operational taxonomic units, abbreviated as OTUs. Such a unit organises sequences according to a defined similarity threshold. In this study, the short marker sequences allowed taxonomic assignments down to genus level at best. The results therefore described communities recognised through DNA markers. Strains grouped in one unit could still differ in biological properties that the marker could not distinguish [Fonseca-García et al. 2016].

Six samples were obtained from each plant, giving 144 DNA preparations. Before sequencing, equal amounts of DNA from three plants representing the same species, site, season and compartment were combined. The final analysis comprised 48 pooled samples. Preliminary comparisons of material from inside the plants had indicated limited variation among replicates, and pooling reduced the cost of the work. Its consequence was an averaging of information from individual plants. The community associated with each combination of conditions could be described more directly than the variation among the three plants contributing to that sample [Fonseca-García et al. 2016].

Before comparing diversity, the researchers standardised the number of reads randomly drawn from each sample. This reduced the possibility that more deeply sequenced material would appear richer simply because more DNA fragments had been read. The resulting picture still depends on the selected markers, sequence-clustering procedure and sampling depth. Repeating the study can therefore compare corresponding compartments using the same procedure. Measuring microbial biomass and activity is a separate task. Those measurements would help connect community structure with the intensity of processes occurring inside tissues or on their surfaces [Fonseca-García et al. 2016].

The compartment structures the community

The clearest division in microbial composition occurred among compartments. Soil samples resembled one another, while communities associated with plant surfaces and interiors formed different groups. Host species, season and site had smaller effects that depended on the compartment being examined. This pattern supports an interpretation known as habitat filtering. A particular part of a plant sustains a subset of the microbial pool available in its surroundings. Conditions in that compartment are therefore associated with which sequences are subsequently detected and the proportions in which they occur [Fonseca-García et al. 2016].

Most units detected around roots, on stems and inside tissues also occurred in soil samples. Their shared presence indicates a potential connection between the environmental pool and plant-associated microorganisms. DNA comparisons alone, however, leave the route of movement unresolved. Similar reads could reach several compartments through different pathways, and the study did not track individual cells over time. The authors considered soil an important reservoir and emphasised the role of successive habitats in structuring communities. Establishing the direction of colonisation would require a separate experiment using identifiable microbial strains [Fonseca-García et al. 2016].

A diverse surface and a selective interior

Diversity was assessed with the Shannon index, which takes into account both the number of detected units and their relative representation. Soil had the highest values for bacteria and archaea. Among plant-associated samples, stem surfaces were more diverse than the rhizosphere. This was a notable feature of the system because exposed stems experience radiation and temperature fluctuations. The stem interior was among the compartments with the lowest diversity. The index describes community structure, so a higher value reflects a different distribution of detected units rather than automatically implying a larger number of cells [Fonseca-García et al. 2016].

Fungi followed a different pattern. Their diversity was high in soils and the rhizosphere, and lower on stem surfaces and inside roots and stems. This difference shows that the single term microbiome encompasses organisms responding differently to the same habitat boundaries. Host identity was particularly evident in the fungal community of the stem interior. Shared units represented a small part of the detected set there and a much smaller share of reads than in most other compartments. The interior of above-ground tissue therefore retained a clearer association with the particular cactus species [Fonseca-García et al. 2016].

Shared microorganisms in different proportions

In both cacti, a substantial proportion of bacterial and archaeal reads belonged to shared units, even where many rarer units were detected in only one host. Distinguishing the number of units from their relative abundance changes how the result is read. Two communities can contain long lists of different rare components while remaining dominated by the same groups. The authors therefore also examined the smaller subset of most abundant units. Comparing their identities and proportions revealed finer differences between hosts within communities that were broadly similar [Fonseca-García et al. 2016].

Among fungi in the stem interior, reads assigned to Henningsomyces distinguished Myrtillocactus geometrizans, whereas reads assigned to Prathoda were associated mainly with Opuntia robusta. Both cacti also contained shared units, including members of Cladosporium and Beauveria. These names indicate the relationships of the recognised sequences. The functions of these particular fungi in living stems remained to be established. The authors proposed isolating them and testing host colonisation and effects on plants. Presence in healthy tissue identifies a location, but does not by itself establish a mutually beneficial association [Fonseca-García et al. 2016].

From community description to functional tests

Stem surfaces also contained cyanobacteria, including reads assigned to Microcoleus and Phormidium. The authors considered whether their presence might help other microorganisms withstand dry periods and strong radiation. They drew on cyanobacterial properties known from other habitats, including the formation of cell aggregates within an enclosing matrix. For the sampled cacti this remained a hypothesis. Measurements of water retention, nitrogen transfer or the survival of other microorganisms would be needed to connect the presence of these groups with a specific mechanism operating on the stem surface [Fonseca-García et al. 2016].

Fonseca-García and co-authors combined their community map with a separate culture-based stage. They isolated 17 bacterial strains from disinfected seedlings grown in vitro from disinfected seeds of the two cacti. Culture assays tested traits including auxin production, phosphate solubilisation and drought tolerance. The results described strain traits in the laboratory. Their connection to cactus development then required testing in the plant [Fonseca-García et al. 2016].

Mascot-Gómez and co-authors used a different sampling unit in their seed microbiome study of four species, Echinocactus platyacanthus, Ferocactus latispinus, F. pilosus and Stenocereus queretaroensis. For each host, they pooled seeds from fruits of at least ten plants and then separated microbes from the seed surface and interior. These pooled samples described the community associated with a seed batch. The earlier six-habitat map instead compared compartments of adult plants in two seasons. Shared bacterial genera in these studies identify potential sources of colonists, while tracing transmission requires recognising the same strains at successive stages of the host’s life [Fonseca-García et al. 2016] [Mascot-Gómez et al. 2021].

The seed study also included an experiment on E. platyacanthus. Seeds bearing their natural microbial community germinated better than disinfected seeds, and reinoculating disinfected seeds with the recovered community or selected isolates did not restore that result. Disinfection also removed surface mucilage, so the difference involved both biological and physical seed properties. At the same time, an Enterobacter isolate, one of the Aspergillus isolates and the recovered community reduced fungal colonisation. Together, the studies show how community composition, traits of a cultured strain and the response of a whole seed are tested separately. Each measurement contributes different information about possible microbiome functions [Mascot-Gómez et al. 2021] [Fonseca-García et al. 2016].