Plantings of perennial grasses have been shown to be an effective means to enhance soil qualities for organic production. Similarly, tillage methods can significantly impact production in organic crop production systems. We have previously examined direct effects of these practices on crop yields, profitability, and soil quality for rotations of organic vegetables in a 4-yr study in northern Florida, but less is known about the effects of these treatments on arthropods. We report here on experiments that used large fields of Argentine bahiagrass, Paspalum notatum Flügge (Poaceae) ‘Tifton 9,’ converted to seasonal vegetable rotations of oat/rye, bush beans, soybeans, and broccoli in a nested design using 4 levels (yr) of continuous bahiagrass production prior to vegetable rotations and 2 tillage methods (conventional and strip tillage). During the fourth yr of the study, we conducted pitfall trapping on a subset of plots involving all 8 treatments (4 bahiagrass treatments and 2 tillage treatments) to examine effects on epigeal arthropods. Over 10,000 organisms and 48 species were identified with 36 arthropod species comprising greater than 97% of the collected specimens. Fields with increasing yr in bahiagrass significantly increased the number of carabid beetles, whereas there was a decline in total herbivores. Tillage treatments impacted arthropod abundance with a noted decline in total carabids collected in strip tilled plots. Pest management implications of these treatments are discussed.
Limitations to agricultural production in the Southern Coastal Plain include infertile, compacted soils and soil erosion (West et al. 1997; Endale et al. 2014; Khalilian et al. 2017). High rainfall and temperatures also result in increased pest and disease pressures. Crop rotation is an effective method to mitigate some of these problems, and is an important agricultural practice that provides many bio-based functions (Dogliotti et al. 2004; Kuepper & Gegner 2004; Carter et al. 2009). Perennial grasses, mainly bahiagrass Paspalum notatum Flügge (Poaceae), have been found to impart a number of advantages when used as a rotational crop, although historically they have been considered an invasive weed species (Marois et al. 2002; Katsvairo et al. 2006). Among the merits of bahiagrass is the ability of their roots to penetrate the natural zone of soil compaction to about 1 m (Elkins et al. 1977); 15 to 40 cm is the common depth of compaction for most farmland in the Southeast (Kashirad et al. 1967; Khalilian et al. 2017). Bahiagrass has been reported to increase rooting depth, root area, and biomass of subsequent crops. In addition, water in the soil profile is conserved and rooting depth of row crops may be up to 10 times deeper following bahiagrass sod-based rotation than conventional cropping systems (Elkins et al. 1977). This could result in as little as one-tenth the current water requirements for irrigation and promote more sustainable crop production (Elkins et al. 1977; Katsvairo et al. 2007).
Strip tillage is particularly effective when used in concert with sod-based rotation. Strip tillage minimally impacts soil, because the soil is tilled only in each row where planting directly occurs. This tillage method maximizes macroporosity and carbon storage, and minimizes erosion (Peigne et al. 2007) while still maintaining the integrity of adjacent soils. Moreover, strip tillage works successfully with a wide variety of crops, whereas crops planted in these systems exhibit significantly higher yields, improved crop quality and plant growth, reduced pesticide use, and higher profits (Wiatrak et al. 2004a, b, 2006).
Despite the acceptance of these more sustainable practices in conventional production, only limited research has been conducted under organic certification (Crowder et al. 2010; Sandhu et al. 2010; Smukler et al. 2010), particularly in the Southern Coastal Plain. Typical off-farm inputs used in conventional, non-organically certified production (such as nutrient supply and pest suppression) must be replaced by mechanisms derived from ecosystem structure and function. These processes include internal nutrient fixation and recycling, enhanced soil microbial populations, and biological pest suppression (such as breaking insect, nematode, weed, and pathogen life cycles), while increasing long-term land productivity (Poveda et al. 2006; Crowder et al. 2010).
Our current research in northern Florida has focused on the effects of bahiagrass sod-rotations and strip tillage on vegetable production in organic crop production systems. We have found that nitrogen and phosphorus soil quality increased (Bliss et al. 2016) where previous bahiagrass rotation is practiced; this has led to increased yields and profitability of vegetable crops (Ahmadiana et al. 2016; Bliss et al. 2016). Grasses also may restore soil biological functions (Carter et al. 2009). We also examined the effects of bahiagrass sod-rotations and strip tillage on soil nematodes and documented that many genera of soil nematodes, particularly root knot and reniform nematodes, significantly decreased with increasing yr in bahiagrass production (Andersen et al. 2016).
Epigeal arthropods comprise the bulk of herbivore, predator, and decomposer species in soil and litter ecosystems (Greenstone 2016). Many of these species are important weed seed and pest predators in agricultural systems (Kromp 1999; Holland and Reynolds 2003; Westerman et al. 2008). Carabid beetles (Coleoptera: Carabidae) are diverse, polyphagous ground dwelling insects that comprise the largest single taxonomic group of epigeal arthropods and often are used as indicator species of diversity, ecosystem functioning, and environmental quality (Leslie et al. 2007; Kotze et al. 2011). Other arthropod groups such as rove beetles (Coleoptera: Staphylinidae), spiders, and ants also are natural enemies of many insect pest species that also have been suggested as bioindicator species (Obrist & Duelli 1996; Clark & Samways 1997; Hummel et al. 2002). We report here on part of a larger study investigating the impacts of bahiagrass sod-rotation and strip tillage on soil nutrients and subsequent crop yields where we examined the effects of these 2 cultural practices on the diverse array of epigeal arthropods found in seasonal rotations of vegetables grown organically in northern Florida.
Materials and Methods
An agricultural field planted with bahiagrass in 2009 was used in this study and located at the University of Florida/Institute of Food and Agricultural Sciences (UF/IFAS), North Florida Research and Education Center in Quincy, Florida, USA, at 30.5427833°N, 84.5956833°W. The field was divided into 8 blocks measuring 24.4 m wide × 47.5 m long. Each block was then subdivided into 4 plots (24.4 m wide × 7.3 m long with 6.1 m borders). Each yr, 1 bahiagrass plot per block was randomly selected and brought into vegetable production so that in 2013 (when pitfall sampling began) plots of 0, 1, 2, and 3 yr of continuous bahiagrass production existed within each block. Plots were further divided into 2 subplots 9.1 m long × 7.3 m wide with a 6.1 m border for different tillage treatments.
Cultural practices were based on those used by organic farmers in northern Florida, and specific details are described by Bliss et al. (2016). We note that organic farmers typically use different agronomic practices for cash crops and cover crops. Each winter, beginning in 2011, one-fourth of each bahiagrass plot was randomly selected and converted to a cover crop rotation of oats (Avena sativa L.; Poaceae) cv. ‘Horizon 270' and rye (Secale cereale L.; Poaceae) cv. ‘Horizon 401.' In the spring, the cash crop consisted of bush beans (Phaseolus vulgaris L.; Fabaceae) cv. ‘Valentino' whereas the summer cover crop was soybeans (Glycine max (L.) Merrill; Fabaceae) cv. ‘Hinson Long Juvenile.' Broccoli (Brassica oleracea L.; Brassicaceae) cv. ‘Major’ was used as the cash crop in the fall. Plots that were already in this rotation remained in this rotation. Bush beans and broccoli were the cash crops of interest whereas oats/rye and soybeans served as cover crops to enhance soil characteristics; cultural methodology reflected usage of the crop. Bush beans and broccoli were planted in 8 rows (7.3 m wide × 0.9 m apart) in each subplot with a spacing of 15 cm for bush beans and 23 cm for broccoli.
For cash crops, Organic Nature Safe (Darling Ingredients, Inc., Irving, Texas, USA) (8-5-5) was applied at 135 kg nitrogen per ha before final tillage (within 3 wk of planting). Conventional tillage required that the total sub-plot be tilled with a rotovator. Strip tillage sub-plots used a strip implement (Kelley Manufacturing Company, Tifton, Georgia, USA) leaving the remaining soil between rows undisturbed. Two wk after planting, bush beans and broccoli were fertilized with organically certified sodium nitrate (16-0-0) at a rate of 34 kg nitrogen per ha. After harvest, bush bean and broccoli plant residue was mowed to allow planting of the oats/rye and soybeans. Seeding rates for cover crops was 70 kg per ha for oats plus 50 kg per ha for rye; soybeans were seeded at 112 kg per ha. Fertilization and tillage were used only for cash crops, because the purpose of cover crops was to increase soil carbon without the expense required to maximize yields for cash crops. Similar to most organic growers, pesticides were provided only on an as needed basis. These were provided twice with Aza-Direct (Gowan Co., Yuma, Arizona, USA) being applied to bush beans in Apr (2.37 L per ha; 28 g ai azadirachtin per ha) and Entrust (147 mL per ha; 35 g ai spinosad per ha; Dow AgroSciences Canada, Inc., Calgary, Alberta, Canada) applied to broccoli in Sep (Table 1). Planting dates were late Mar to early Apr (bush beans), Jun (soybeans), Sep (broccoli), and Dec to early Jan (oats and rye).
Pitfall trapping was conducted throughout 2013 at times when agricultural activity was minimal, but dates were chosen to bracket significant production events (Table 1). Sixteen paired sub-plots were randomly selected (pairing tillage methods) and were sampled multiple times within each rotation. Four pitfall traps were used in each sub-plot per sampling date (64 traps total per date). Collections were completed on 15 and 19 Feb; 3 and 24 May; 11 Jun; 5, 19, and 26 Aug; and 8 Oct. The requirement of using a certified organic preservative (see composition later in this section) in the traps prohibited trapping for longer than a wk due to decomposition of captured organisms. Pitfall traps were centered in the fourth or fifth row of each sub-plot and placed 1.8 m apart. Traps were placed at the same location for each collection period.
Significant events in study plots of organic vegetable production during 2013.
Ground surface pitfall traps were constructed as follows: the center of a 17.8 cm brown plastic dinner plate was cut out such that a 455 mL Solo cup (Solo Cup Co., Lake Forest, Illinois, USA) would nest in the hole with the top of the cup resting flush with the bottom of the inverted plate. Fine soil was glued to the exposed plate surface to help camouflage the pitfall trap and roughen the plastic surface. The unit was buried in the soil such that the top of the pitfall trap was approximately even with the surface of the ground. A 20.3 cm diam × 2.5 cm high metal wire mesh (12 × 25 mm) ring was placed over the top of each pitfall trap. A second brown plastic plate (25.4 cm diam) was placed atop the metal ring. This plate was secured with soil and served as a protective cover for the pitfall trap beneath. The design helped prevent irrigation and rain water from entering and flooding pitfall traps, and also prevented capture of frogs, shrews, and other small vertebrates. All trap cups were half-filled with an organically certified killing agent consisting of 50:50 Safer Soap (Woodstream Corp., Lancaster, Pennsylvania, USA) and a saturated salt water solution. Unfortunately, this formulation precluded the evaluation of collembolan species and some lepidopterans that became too decomposed to identify. Pitfall traps were emptied and reset after each 4-d sampling date. Samples were cleaned, sorted, and then stored in glass vials containing 95% ethanol solution and later identified in the laboratory.
Most organisms were identified to species, although in some cases only to genus. In a few cases, degradation of specimens in pitfalls only allowed identification to higher taxonomic levels. Data were sorted and analyzed separately by crop rotation (oats/rye, bush bean, soybean, broccoli). Homogeneity of variance was examined by Barlett's test (SAS 2009), and about one-third of the species exhibited heterogeneity of variance. Thus, analyses were performed nonparametrically using Friedman's test. With this procedure, trap catches for each species within a rotation were ranked. Larger taxonomic groups (family, order) and trophic level (predators, omnivores, herbivores) also were analyzed in a similar fashion. In the case of trophic level, species were grouped by the predominant method by feeding behavior (herbivorous, predacious, omnivorous). Nonparametric analyses were completed using SAS (2009). Differences were considered significant at P ≤ 0.05. All voucher specimens were deposited in the insect collection at North Florida Research and Education Center-Quincy in Quincy, Florida, USA.
A total of 10,192 specimens belonging to 48 taxonomic groups (most identified to species) were collected from the pitfall traps (Table 2). For the more common groups (individuals collected > 5), 33 of the 34 species were arthropods with the exception of slugs (Soleolifera; 236 specimens). The most commonly collected species of arthropods were: Gryllus pennsylanicus Burmeister (Orthopera: Grillidae) (2,497 individuals), Solenopsis (Hymenoptera: Myrmicinae) (2,177 individuals), Blattella asahinai Mizukobo (Blattodea: Ectobiidae) (1,184 individuals), Forficula auricularia L. (Demaptera: Forficulidae) (859 individuals), and Heteroderes amplicollis Gyllenhal (Coleoptera: Elateridae) (724 individuals). Species diversity was highest for Coleoptera with 11 species being collected, including 5 species of carabid beetles. Other common predatory beetles collected included Staphylinids. Herbivorous beetles in collections included H. amplicollis, Phyllophaga latifrons LeConte (Coleoptera: Scarabaeidae), Sitophilus (Coleoptera: Curculionidae), Lobiopa insularis Castelnau (Coleoptera: Nitidulidae), and another unidentified species of sap beetle (Nitidulidae).
Pit fall trap collections varied greatly with time of yr. However, we did not compare arthropod abundance in crops between seasons in the experimental design because it would be impossible to separate effects of season from those of the crop host, which also varied seasonally. Our analyses were sorted by crop-season because most arthropod species have seasonal fluctuations. Cultural practices applied to cash crops also were different than that of cover crops, including applications of organic insecticide during each cash crop rotation that were applied on an as needed basis. However, we do note that, in general, insect abundance (particularly herbivores) was lower in cold weather mo when oats/rye and broccoli were the crops available to arthropods (Table 2). Only 5 of the 22 species categorized as herbivorous or omnivorous were captured during both cold weather rotations, whereas 8 of the 12 predator/parasitoid species were captured during all crop rotations.
Mean(± SE) abundance per 10 pitfall traps for common (total caught > 5) arthropods sorted by crop species (tillage methods combined).
Nearly half (44%) of the common species (n > 5 total individuals captured) were affected by the number of yr in continual bahiagrass or tillage treatment during at least 1 of the crop rotations (Tables 3–5). Effects varied with time of yr and crop. Predatory beetle species, particularly carabids, most often were significantly increased by increasing yr in bahiagrass production (Table 3). Platynus decentis Say (Coleoptera: Carabidae) abundance also increased significantly with increasing continuous yr in bahiagrass and conventional tillage systems throughout the winter with oats/rye rotation. The carabid Harpalus pennsylvanicus (Coleoptera: Carabidae) similarly showed significantly higher trap catches with increasing yr in bahiagrass with broccoli. Another carabid species, Scarites subterraneus (Coleoptera: Carabidae), showed significantly increased trap catches in strip tillage plots in winter (oats/rye rotation). Trap catches of Galertia bicolor (Coleoptera: Carabidae) significantly increased with yr in bahiagrass production in the bush bean rotation, and catches of Tetracha floridana (Coleoptera: Carabidae) significantly increased in plots with conventional tillage compared with strip tillage in soybeans. In contrast, the wolf spider, Trochosa tericola (Thorell) (Araneae: Lycosidae), showed significantly decreased captures with number of yr in continuous bahiagrass production. The most abundant predacious arthropod species (Solenopsis spp.) had over 5 times greater capture rate in pitfall traps from strip tilled sub-plots than conventionally tilled plantings of oats and rye.
Effects of bahiagrass treatments on herbivorous or omnivorous species were much less consistent than on predacious species. Sixty-seven percent of the predacious arthropod species were impacted by these treatments at some time during the crop rotations compared with omnivores (40%) and herbivores (30%). Sap feeding beetles (L. insularis and other Nitidulidae) increased significantly with increasing yr in bahiagrass with summer bush bean rotation. Asian cockroaches (B. asahinai) were one of the most frequently captured insects, and increased significantly with increasing yr of bahiagrass and strip tillage in the soybean rotation. In contrast, G. pennsylvanicus significantly decreased with yr in bahiagrass with bush bean and soybean rotations. Captures of Musca domestica L. (Dipera: Muscidae) decreased with yr of bahiagrass in the soybean rotation, whereas captures of Miridae increased with increasing yr of continuous bahiagrass production in oats/rye.
Statistics and mean abundance of predatory epigeal arthropods captured in pitfall traps as a function of yr in bahiagrass and tillage method. Data and statistics are presented only for species and crops that had significant yr in bahiagrass or tillage effects. CT = conventional tillage; ST = strip tillage.
When analyzed by higher-level taxonomic or trophic grouping, carabid beetles showed the most consistent treatment effects. The number of continuous yr in bahiagrass production significantly enhanced carabid trap catches in bush beans (P < 0.036; df = 3,1,187), broccoli (P < 0.035; df = 3,1,59), and oats/rye (P < 0.001; df = 3,1,123); only catches in the soybean rotation showed no effects (Fig. 1). Conventional tillage also increased total carabids in cold weather mo compared with strip tillage in oats/rye (P < 0.043; df = 3,1,123) and broccoli (P < 0.40; df = 3,1, 159; Fig. 1). Total predatory beetles yielded somewhat similar results where significant increases in trap catch was observed with increasing continuous yr of bahiagrass production in bush beans (P < 0.003; df = 3,1,187) and oats/rye (P < 0.002; df = 3,1,123) rotations. Trap catches also were higher significantly with conventional tillage compared with strip tillage in the broccoli rotation (P < 0.010; df = 3,1,59; Fig. 2). Analyses of other groups of predators, total predators, or total predators minus predatory beetles showed no treatment effects. Analyses of total herbivores also showed significant decreases in trap catches in oats/rye (P < 0.026; df = 3,1,123) and soybean rotations (P < 0.040; df = 3,1,187) with increasing yr in bahiagrass although tillage practice had no significant effects. Analyses of herbivores by taxonomic order showed that Orthoptera decreased with increasing yr in bahiagrass in bush bean (P < 0.003; df = [3,1,187]) and soybean rotations (P < 0.050; df = [3,1,187]). Total Coleoptera captured were significantly higher in conventionally tilled plots than strip tilled plots in the broccoli rotation (P < 0.012; df = [3,1,59]).
The most consistent, and perhaps most significant, treatment effects were the number of yr in continuous bahiagrass production increasing captures of all carabid species. Also conventional tillage (as opposed to strip tillage) significantly increased captures of carabid beetles in some crop rotations. The mechanisms of how treatments impact carabids and other species were not addressed in this study. However, this study was conducted as part of a larger one that offers some possible explanations. The larger study investigated the effects of bahiagrass and tillage on crop yields, nutritional status, and soil parameters (Bliss et al. 2016). Increasing yr in bahiagrass cultivation altered soil nutrition and characteristics resulting in significantly increased crop yields, total nitrogen content of vegetation, and total biomass within the individual plots. Increased vegetative biomass suggested more groundcover for carabids, which has been shown to provide more suitable overwintering refuge for these beetles (MacLeod et al. 2004); however, seasonal effects of grasses may vary with carabid species (Labruyere et al. 2016). Although winters in northern Florida are relatively mild, the strong effects of bahiagrass treatments on carabid trap catch during colder weather (oats/rye and broccoli rotations) are consistent with benefits in overwintering.
Herbivorous epigeal arthropods captured in pitfall traps (mean catch per trap) that were significantly affected by treatments. CT = conventional tillage; ST = strip tillage.
Omnivorous species (arthropods and others) captured in pitfall traps (mean catch/trap) that were significantly affected by treatments. CT=Conventional tillage; ST=strip tillage.
We found 7 of the 23 arthropod species categorized as herbivorous or omnivorous were significantly affected by bahiagrass treatment with only 3 species increasing with increased yr of bahiagrass production. Our tangential study (Bliss et al. 2016) previously showed that increasing yr of bahiagrass production resulted in higher plot biomass and higher crop nitrogen content. Thus, it is surprising that these plots with increased plant quality and quantity did not result in increased trap catches of phytophagous insects. Increased predation, particularly by carabids, may have contributed to this lack of increased abundance of herbivores and omnivores. The only herbivorous insects that increased significantly in the summer rotation were 2 species of Nitidulidae. As a rule, sap feeding insects are more likely to be nutrient limited (Brodbeck & Strong 1987). Moreover, members of this family have been shown to respond strongly to changes in diet, as well as having a high reproductive potential (Ellis et al. 2002; Stuhl 2017). Our results with Nitidulidae may parallel those of Holland (1998) and Collins et al. (2002), where aphids (who are often also nutrient limited with a high reproductive potential) may be reduced by carabids and other polyphagous predators until their reproductive capacity outstrips the predator's ability to effectively reduce their numbers.
Effects of tillage practice appeared more variable than the persistence of bahiagrass habitat. Total captures of carabid beetles decreased with strip tillage in cold weather rotations (oats/rye and broccoli). These results contradict those of Kromp (1999) who reviewed numerous studies and concluded reduced tillage benefited carabid populations; however, this review also listed multiple exceptions dependent on study and species. Similarly, Hatten et al. (2007) established differential responses to tillage practice by different carabid species within one study. Holland and Reynolds (2003) documented differential response to identical tillage treatments on a yr-to-yr basis. In our study, trap abundance of the most common predator, Solenopsis, was > 5-fold greater in strip tillage than conventional tillage in cold weather (oats/rye rotation), perhaps because ant nests were more susceptible to disturbance by tillage in cold weather, similar to the findings of Landis et al. (2000) and Pereira et al. (2010). We also found 2 of the 23 arthropod species categorized as herbivorous or omnivorous were significantly impacted by tillage treatment.
Cover crops, as well as grasses, have been shown to increase carabid overwintering success (Rivers et al. 2017). We documented the persistence of carabids throughout the yr, particularly when compared with the seasonal nature of other species such as herbivorous insects. Our use of winter cover crops, as well as collecting from plots with previous yr of continuous bahiagrass cultivation, may contribute to those results. Conversely, most non-predacious species investigated in our study showed strong seasonal population fluctuations. Reasons for such fluctuations could be due to normal seasonal fluctuations for individual species or the result of host-specific responses by plant-eating insects to different crops planted throughout the crop rotations. Specifically, seasonal differences in abundance also may have been influenced by host associations with the diversity of weed species within plots. One should note that the study followed a protocol typical for organic production pest control where organic insecticides were applied on an as needed basis. Aza-Direct was applied on 23 Apr while Entrust (spinosad) was applied on 30 Sep. These treatments could have contributed to temporal variation in insect populations. Although identifying mechanisms to explain seasonal variability were not investigated, there was a clear trend regarding feeding type; 67% of all predatory species were found in all rotations, as opposed to 30% of omnivorous species and 15% of herbivorous species.
Carabid beetles provide valuable ecosystem services as pest and seed predators (reviewed by Lövei & Sunderland 1996, and Kromp 1999). The presence of beneficial insects, such as carabids and arachnids, may lengthen periods between pest outbreaks. Interestingly, in the United Kingdom and Europe, “beetle banks” have been created (Thomas et al. 2002; MacLeod et al. 2004), where habitats are managed specifically for the maintenance of carabid populations; most often these are perennial grasses planted and maintained adjacent to or within crops. The efficacy of beetle banks in suppressing pest populations is well documented (Chiverton 1987; Collins et al. 2002). Grassland corridors (Do et al. 2017) and the presence of grassland on crop margins (Thomas et al. 2002) also have been investigated as management tools to encourage the ecological services of carabids.
We believe the value of carabid beetles, and other generalist predators, is enhanced in organic crop production systems where conventional methods of pest control are limited. These organic systems have been shown to maintain generally higher populations of carabids in crops and adjacent habitats than do conventional systems (Purtauf et al. 2005; Lundgren et al. 2006; Caprio et al. 2015), although anomalies exist (Rondon et al. 2013). Organic crop production systems also may enhance carabid emigration from adjacent habitats which may be of great benefit given the relatively low dispersal abilities of many carabid species (Lövei & Sunderland 1996; Do et al. 2017). Our results suggest that a further understanding of how bahiagrass enhances the abundance of carabid populations and the resultant pest management potential of these treatments merit further investigation in organic and conventional crop production systems.
We thank Victoria Ospina for technical assistance. Our thanks also to Baron Riddle and Charlie Riddle for collection of data and preliminary statistical analyses. This research was partially funded by grant No. 2010-51106-21866 from the USDA NIFA-Organic Transition Program to P. C. Andersen, R. F. Mizell, III, J. J. Marois, D. L. Wright, and others.