*3.2. Quantitative Real-time PCR of* alp *and* nifH *Genes*

The *alp* and *nifH* genes were quantified in all sweet potato rhizospheres. The values for *alp* gene quantification (copies of *alp* gene/g of rhizosphere soil) were higher than for *nifH* gene (Figure 3). The analyses revealed a statistically significant difference (*p* ≤ 0.05) in the abundance levels of the *alp* genes only in IPB-137 after six months of planting, suggesting that the abundance of the phosphate mineralizing bacterial community present in the tuberous roots of sweet potato is influenced by the genotype and also by the time in IPB-137 (Figure 3A). Similarly, the IPB-137 genotype showed statistically significant differences in the abundance levels of the *nifH* genes between three and six months (t1 × t2) after planting (Figure 3B). Statistically similar *nifH* gene densities (*p* ≤ 0.05) were observed between the t1 and t2 samplings in IPB-149 and IPB-052. When comparing the t2 samplings, no differences in *nifH* gene densities were observed between the different genotypes (Figure 3B).

**Figure 3.** Abundance of bacterial phosphate mineralizing bacteria based on the *alp* gene (**A**) and nitrogen fixing bacteria based on the *nifH* gene (**B**) in the rhizosphere of three different sweet potato genotypes (IPB-149, IPB-137 and IPB-052) sampled after three and six months after planting (t1, \_3M and t2, \_6M, respectively). The error bars indicate standard error. Different letters (a and b) indicate significant differences of means in pairwise comparisons (Tukey test; *p* ≤ 0.05).

#### **4. Discussion**

The use of biofertilizers, including plant growth promoting bacteria, is increasing worldwide as an environmentally-friendly alternative for soil fertility management, sustaining not only sweet potato yields (both in terms of tuber weight and in terms of the levels of starch in sweet potato crops) but also different economically important plants [11], compared with the continuous application of inorganic fertilizers alone [36–39]. The plant positive response to biofertilization stimulates the continuous optimization of crop yields by exploiting beneficial plant-microbe interactions [4]. However, the effect of different biofertilizers on root colonization, nutrient uptake, growth, and yield is not always as effective as expected. Different types of soils found in various countries, as well as plant genotypes and the sampling time (period the crop was maintained in the field) all contribute to different responses in plant yields, including those observed in sweet potato [22,36,37].

It was evident in previous studies conducted by our group that the total bacterial communities associated with different genotypes of sweet potato planted in Brazil varied depending on the plant age and genotype [12,25]. However, the behavior of specific communities such as phosphate mineralizers and nitrogen fixers, which may contribute directly to the plant development, was not taken into account at that time. Indeed, the majority of the studies focusing on phosphate-solubilizing/mineralizing and nitrogen-fixing bacterial communities found in the literature were conducted to screen for beneficial traits that could improve plant growth. Alterations on the structure and abundance of these communities along the plant growth have not been accessed, which are usually restricted to a single genotype and a growth stage of the plant [21,24,40].

In the present study, we demonstrated that the structure of phosphate- mineralizing bacterial community (based on the presence of *alp* genes) varied along the plant growth depending on the genotype considered. The structure of the phosphate mineralizing bacterial community in IPB-137 was mainly different from that in IPB-149. Tarafdar and Jungk [41] observed that both acid and alkaline phosphatase activities of soil were usually increased near the rhizoplane of different plants and such an increase depended on plant species, soil type, and plant age. Although the optimal pH for the activity of the alkaline phosphatase is higher than the pH of the soil where the sweet potatoes were planted, its role remains an important one regardless. For example, both acid and alkaline phosphatase activities near the rhizoplane of maize resulted in alterations of the composition of bacterial communities, as determined by PCR-DGGE [42]. In addition to the influence of the genotype, a statistically difference in phosphate mineralizing bacterial communities in IPB-137 was observed between samplings after three to six months of planting. Changes were observed not only in the structure of this community but also in the abundance of *alp* genes after six months of growth. These changes could be explained by the fact that, at the late growth stage of sweet potato, the roots usually presented stronger nutrient absorption ability due to the increase of the number of root tips, the enlargement of root surface area, and the root volume [43]. Phosphate mineralizers may have contributed to the conversion of insoluble phosphates into available forms for plant via different processes such as acidification, exchange reactions, chelation, and production of gluconic acid [21,44].

Nitrogen-fixing bacteria identified as *Azospirillum* sp. were first isolated from the fibrous roots and storage root peels of sweet potato [45]. From this time on, different studies were carried out on the isolation and identification of nitrogen fixers in sweet potato roots, as important evidence indicated that nonsymbiotic N2 fixation in sweet potato is agronomically significant [40]. Active expressions of the *nifH* gene phylogenetically similar to those of *Bradyrhizobium* spp., *Sinorhizobium* sp., *Azorhizobium* sp., *Bacillus* sp. and *Pelomonas* sp. were found in large quantities in the N2-fixing sweet potato storage tubers planted in different regions [40,46,47]. In our study, the structure and abundance of nitrogen-fixing bacterial communities based on the *nifH* gene were determined and compared between the three genotypes.

The nitrogen-fixing bacterial community in the rhizosphere of IPB-052 seems to be different from those of the other two genotypes. IPB-052 presents a white surface color on the tuberous roots and high starch content. It is quite possible that specific nitrogen-fixing bacteria are being recruited from soil to the rhizosphere. This rhizosphere effect can generate a zone with maximum microbial activity, resulting in a pool of metabolic products that can support plant growth [48]. However, the difference observed on the structure of the nitrogen-fixing bacterial community in IPB-052 does not also appear in the analysis of the *nifH* gene abundance. The *nifH* gene copy numbers appeared to be lower than those of *alp* genes, and the only difference observed among the genotypes studied was between the 3–6 month samplings in IPB-137. This fact suggests that nitrogen-fixing bacteria may only contribute more towards plant development during the early stages of plant growth in this sweet potato genotype.
