Organic fertilization influences nematode diversity and maturity index in coffee tree plantations using an agroforestry system

This work was carried out at a rural property within the municipality of Araponga, Zona da Mata in Minas Gerais state, Brazil (20º 38´ 39.76´´ S, 42º 3´ 0.27´´W), located at an altitude of 1,315 m. The region’s climate is mesothermal and annual rainfall varies from 1,280 mm to 1,800 mm. The terrain is mountainous; declivity is between 20 and 45%, with prevalence of latosols. This rural property had three coffee plantations (Coffea arabica L. cv. Red Catuaí) using the agroforestry system and was close to an Atlantic rainforest fragment.

In the 1980s, this rural property was used for corn and bean cultivation and cattle raising. At that time, burning practices were frequent. The coffee plantations began in 1999. Chemical fertilizers were applied to the coffee tree plantations between 1999 and 2001. After that, chemical fertilizers were gradually replaced by organic fertilizers. In 2003, other species of trees were incorporated inside the coffee plantations, and since then, the transition from solely coffee plantations to AFS began. Farmers planted trees, especially fruit trees, randomly grown within the coffee plantations. Other management changes also occurred: farmers stopped weeding spontaneous growing plants, started tilling the soil, and stopped using pesticides. Currently, two coffee plantations are certified as organic by BCS Öko-Garantie, a German certification organization. One of the plantations is fertilized using poultry litter and the other one with cow manure. In a third plantation, a natural farming system (only plant residues are added to the tilled soil) was adopted and the crop has been certified by the Shumei Japanese organization of natural farmers.

The experimental design was carried out in coffee plantations using three different organic fertilizer treatments (natural fertilizer (NF), poultry litter (PL), cow manure (CM)) and in a fragment forest (FF). The 25,000 square meters of Atlantic forest fragment is considered as a remnant of the forest and does not present recent signs of anthropogenic disturbances, except for activities involving removal of fallen branches that were used for fertilizing natural coffee plantations.

The naturally fertilized coffee plantation was comprised of an area of 8,000 square meters, with 2.3 × 1.2 meters spacing between coffee trees. In this system, coffee trees were last pruned in 2011. Such plantations are mostly cultivated together with the following plant species: banana (Musa sp.); capoeira-branca (Solanum argenteum); royal palm (Archontophoenix cunninghamiana) and pinto peanut (Arachis pintoi). Plant residues already present in coffee plantations are used as fertilizers, including banana pseudostem, capoeira-branca trunks, pinto peanuts, and leaf litter from the nearby forest fragment. Fertilization is performed by placing the plant material in a circle on the soil around the coffee tree, and the coverage is based on the size of the crown. This is performed once in November and again 45 days later. In total, approximately 70 tonnes ha^-1 of residues are used per year in the plantations.

The coffee plantation fertilized with poultry litter comprised an area of 4,000 square meters, with 1.5 × 3.0 m plant spacing. Fertilization was performed by placing the fertilizer on the soil in a circle around the coffee tree, and coverage was based on the size of the crown. This was carried out once in December and again after 45 days after the first application. A total of 44.5 tonnes ha^-1 of composted poultry litter, obtained from neighboring farms, was used per year in this plantation. To compost poultry litter, this material was assembled in layers interspersed with manure and coffee bean peels from the intended plantation. The material was moistened at least once a week and turned in every 30 days. The entire process takes 90 days. The predominant tree species in the area are “guatambu” (Aspidosperma polyneurum); “capoeira-branca” (Solanum argenteum) and banana (Musa sp.). Except for the banana trees that are planted around the coffee tree plantations as a physical wind barrier, the other species were randomly distributed in the plantations.

The coffee tree plantation fertilized using composted cow manure comprised an area of 15,000 square meters, with 3.0 × 1.5 m plant spacing. Fertilization was performed by placing the fertilizer on the soil in a circle around the coffee tree, and coverage was based on the size of the crown. This was performed once in December and again after 45 days after the first application. In total, approximately 13.2 tonnes ha^-1 of composted cow manure from the rural property was used per year in the plantation. For composting, the manure was heaped in a pile, moistened every week, and turned every 30 days. The entire process takes 90 days. In this area, the main tree species were avocado (Persea gratissima) and banana (Musa sp.).

Physical analyses of soil total porosity, microporosity, macroporosity, soil moisture, and chemical analyses of soil macronutrients, organic matter, C:N ratio, and pH from the four areas evaluated were carried out. In each experimental plot were samples were collected from the soil layer at a depth of 0–10 cm with the aid of a volumetric ring. The methods used in the physical and chemical soil analyses were according to Donagema et al. (2011). The collection of soil samples for physical and chemical analyses was simultaneous with the collection of samples for nematological analyses.

Soil samples for estimation of nematode populations were collected from the three coffee plantations and the Atlantic forest fragment. For each fertilization treatment, five coffee trees were randomly selected and the soil at the base of the plant was sampled in a zigzag pattern. In each area, five soil samples were collected at a depth of 0–20 cm in September 2015 using a volumetric ring with a 2.5-cm diameter. A total of 12 samples were collected and homogenized as a mixed sample for each coffee tree selected (Huising et al., 2008) (Figure 1). Samples were then transferred into plastic bags, sealed, and labeled for subsequent extraction and identification of nematodes.

Figure 1:

Extraction of nematodes was performed using the flotation method with a sucrose solution (Jenkins, 1964). Soil samples (100 cc) were first passed through a 2-mm mesh sieve to remove stones and roots. Sieved soils were then placed in a beaker containing 1,000 mL of water, and suspensions were manually stirred for a minute. Then, stirring was interrupted for 20 s so that thicker particles could settle and the supernatants were then transferred to a 400-mesh sieve. The retained material was then washed and transferred to Falcon tubes (50 mL) and centrifuged for 5 min at 1,008 × g, and the supernatants were discarded. Then, a sucrose solution (460 g of sucrose per liter of water) was added to the tubes, and the mixture was centrifuged for one minute at 112 × g. Suspensions were then passed through a 400-mesh sieve and washed under running water to remove the sucrose. Nematodes retained by the sieve were then transferred to Falcon tubes (50 mL) containing 10 mL of water. The tubes were kept in an incubator for 5 min at a temperature of 64°C, resulting in the death of the nematodes; however, the nematodes were not deformed by this process. Then, they were fixed by the addition of 1 mL of formaldehyde solution and stored at 4°C in a refrigerator.

Suspensions containing nematodes were pipetted into flat glass Petri dishes (60 diameter × 15-mm height) that were then placed under a binocular stereoscope (Bel Photonics®, model SZ) to count the nematodes. Nematodes were picked up using a needle and transferred to microscope slides (Precision Glass®, 26 × 76 mm) containing a drop of water + formaldehyde at a ratio of 10:1.

A double layer of matt varnish (Acrilex®) was applied to each slide to fix the coverslips in place (Paiva et al., 2006). To seal the space between the slide and coverslip, a double layer of colorless nail polish (Risqué®) was then applied. Nematodes were observed using an optical microscope (40 × objective) (Motic®-A210 model) and identified to the genus level using standard keys (Andrássy, 2005; Chaves et al., 1995; Jairajpuri and Wasim, 1992; Siddiqi, 2000), and classified into five trophic groups based on feeding habits: bacterivores, plant-parasites, fungivores, omnivores, and predators.

The effect of fertilization on the nematode communities was assessed by calculating total abundance, trophic abundance, and relative abundance. Trophic abundance is the abundance of nematodes within the different trophic groups. Relative abundance was calculated as the percentage value of the number of genera that belong to a particular taxon or trophic group in relation to the total number of nematodes present in a sample using the following equation: p _i  = (n/N) × 100, wherein n was the number of nematodes from each genus and N was the total number of nematodes in each sample (Cares and Huang, 2008).

To compare the diversity and dominance between the areas, the Shannon (H´) and Simpson (Ds) indexes were calculated. The Shannon index applies equal weight to rare and abundant genus (Shannon, 1948) using the equation H´ = −∑ p _i × LN (p _i ), wherein p _i is the relative abundance and LN is the logarithmic function. The Simpson index measures the probability of two randomly selected nematodes from a sample belonging to the same species. This index is calculated using the equation Ds = ∑ (p _i )², wherein p _i was the relative abundance of nematodes in a sample (Simpson, 1949).

The specific indexes for nematodes that were developed by Bongers (1990), Maturity (MI), Maturity 2–5 (MI 2–5), and Plant Parasites (PPI) indexes, and by Ferris (2010), Basal (BI), enrichment (EI), and structure (SI) indexes were calculated to obtain the environmental conditions of the areas studied.

To calculate the maturity indexes, all soil nematodes except plant-parasitic ones were considered. MI 2–5 was based on the calculations for the nematode groups that had a c-p value between 2 and 5; therefore, plant-parasitic and nematodes with a c-p value of 1 are excluded. The Maturity Index is considered a measure of environmental disturbance, and low MI values indicate disturbed and enriched environments while high MI values indicate stable environments (Bongers, 1990). Due to the functionality of plant-parasitic nematodes in the agroecosystem, this group is used to calculate the level of anthropic disturbance using PPI (Bongers, 1990). The equation used to calculate MI, MI 2–5 and PPI is ∑[v(i) × pi], where v(i) corresponds to the c-p value of the nematode family and pi the relative abundance of that family in the sample (Bongers, 1990).

The Basal Index (BI), Enrichment Index (EI), and Structure Index (SI) are based on the importance of functional guilds of nematodes as indicators and are descriptors of food web conditions. Basal food webs are those which are diminished due to stress, scarce availability of resources, contamination, or other harsh environmental conditions. Nematodes present in these food webs are represented by bacterial and fungal feeding taxa from the c-p2 class of the MI. Enriched food webs are those with high availability of resources due to the occurrence of a disturbance event. Opportunistic bacterial feeding nematodes from the c-p1 class are predominant in these food webs. Fungal feeding nematodes from the c-p2 class might increase when more complex resources, such as with a higher C:N ratio, becomes available, or when fungal feeding activity is enhanced in detriment to bacterial feeding activity. Structured food webs are those with more resources available. Nematode taxa from the c-p3 class are present in less structured food webs, while structure in the community will be greater when nematode taxa from the c-p4 and c-p5 classes are present. These indexes were obtained through the NINJA platform (Sieriebriennikov et al., 2014) and to characterize the conditions of the studied areas, the data were plotted according to Ferris (2010).

The values obtained for the attributes from the physical and chemical soil analyses were submitted to analysis of variance (one-way ANOVA), and when the results were statistically significant, they were compared using the Tukey test (p < 0.05). The chemical attributes for phosphorus (P) and potassium (K) did not follow a normal distribution, therefore the averages obtained were submitted to non-parametric Kruskal–Wallis analysis, and compared using Dunn’s method (p < 0.05).

The calculated abundances, diversity, and dominance measures, and nematode indexes were submitted to analysis of variance, and when the results were statistically significant, they were compared using the Tukey test (p < 0.05). Genera-related abundance did not follow a normal distribution; therefore, the non-parametric Dunnett’s method was used in this case. Statistical analysis was performed using the SigmaPlot® 12.0 (Systat Software, Inc.) software.