Sugarcane (Saccharum sp.) is a major agricultural crop in Florida and is planted by vegetative methods in 15 - 30 cm deep furrows. Land preparation for sugarcane planting in Florida normally involves at least 3 - 4 disk passes followed by furrow preparation. Growers also cultivate between rows during the first few months of growth to reduce compaction and increase tillering. Reduced tillage or no-tillage may play an important role in reducing soil erosion (Gebhardt et al. 1985, Science 230: 625 - 630), production costs (Al-Kaisi and Yin 2004, Soil Till. Res. 78: 91 - 101), and consumption of fossil fuels (Phillips et al. 1980, Science 208: 1108 - 1113). However, reduced or no-tillage practices may affect arthropod ground predators by altering soil microclimate, altering weed occurrence which may be alternate hosts, and/or direct arthropod mortality via crushing, desiccation, or exposure to predators such as birds. This study was conducted to determine short- and long-term effects of no-tillage, minimum and conventional preplanting and postplanting tillage levels on population of arthropod ground predators in Florida sugarcane.
Field testing was conducted at the Everglades Research and Education Center, Belle Glade, FL. The field tested was chemically fallowed for 7 - 8 months before tillage tests, meaning there was no cultivation and weeds were controlled with herbicides. There were 3 preplanting tillage treatments: conventional, minimum, and no-tillage. In conventional tillage, a field area was disked 3 times followed by furrow preparation. In minimum tillage, the field area was disked once followed by furrow preparation. In no-tillage there was no disking and involved only furrow preparation before planting. The sugarcane variety CP88 - 1762, which is the dominant variety planted in Florida, was planted on 24 January 2013 in all the plots. There were two postplanting interrow cultivations (mid-April and late-June) in conventional tillage, one in minimum tillage (mid-April) and none in no-tillage plots.
The experimental design was a randomized complete block design with 6 replications. Plot size was 10 m × 4 rows (6 m). Populations of arthropod ground predators were measured with pitfall traps. Pitfall traps have been used extensively for studies on surface dwellers, and their use is well discussed by Southwood and Henderson (2000, Blackwell Science, Malden, MA). More specifically, pitfall traps have been used to measure populations of predaceous arthropods in Louisiana sugarcane fields (Hensley et al. 1961, J. Econ. Entomol. 54: 146 - 149; Negm and Hensley 1969, J. Econ. Entomol. 62: 1008 - 1013; Reagan et al. 1972, Environ. Entomol. 1: 588 - 591) and Florida sugarcane fields (Cherry 2003, Florida Entomol. 86: 49 - 52). Each pitfall trap consisted of a 9-cm diam plastic cup containing 100 ml of ethylene glycol. A 5-cm deep plastic collar was also cut from the 9-cm plastic cups. The top of this collar was taped in the middle of a 26-cm diam paper plate with its center removed. This collar was then inserted into the pitfall trap and the plate loosely covered with soil. This arrangement prevented soil subsidence around the trap rim thus allowing arthropods easy access to the trap. A small metal roof was also placed above each trap to prevent rainfall from filling traps.
Two pitfall traps were located in the middle of each plot in each of the 2 center rows of sugarcane. After 2 wks, cups were taken to a laboratory where arthropods were filtered out and then counted including microscopic examination of small specimens. Major predator groups counted were ants (Formicidae), earwigs (Dermaptera), ground beetles (Carabidae), rove beetles (Staphylinidae), spiders (Arachnida), and centipedes (Chilopoda). More specifically, imported fire ants (Solenopsis invicta Buren) and wolf spiders (Lycosidae) were counted because these were easy to identify and are 2 major predators found in Florida sugarcane fields (Cherry 2003).
Effects of the preplanting tillage treatments were assessed by collecting pitfall trap samples on 26 March, soon after plant cane emergence. To measure the short-term effects of the 2 postplanting cultivations, pitfall trap samples were collected 1 wk after each cultivation, and long-term effects were measured by collecting trap samples in December, 6 months after the last cultivation.
Data were analyzed by using Proc Mixed model in SAS (SAS Institute 2003, Cary, NC) with repeated measures statement. Tillage treatments were considered as fixed effects, and replications as random effects. Different sampling dates were considered to be the repeated parameter. Compound symmetry was used as covariance structure. Least square means were separated by using Tukey's HSD test with significance at α =0.05.
Our results show that sampling date had significant effects on all arthropod catches in traps (Table 1). However, tillage did not have a significant effect on any arthropod catches. Date and tillage interactions were significant only with wolf spider occurrence in this limited evaluation for one crop season.
In general, imported fire ants, earwigs, and ground beetle catches increased from the first sampling date (26 March) to the second sampling date (8 May) and then decreased on the third (8 July) and fourth (18 December) sampling dates (Table 2). During the first 3 sampling dates, earwigs per trap (8.3 - 25/trap) were high compared with other arthropods. However, these levels were reduced significantly during the last sampling date (0.1/trap). Cherry (2003) reported that earwigs were 6% of total predator catches in Florida sugarcane fields. In this study, earwigs were the most abundant predator caught in traps for unknown reasons. More imported fire ants were collected in traps than all other ants combined which is consistent with the earlier findings of ants in commercial Florida sugarcane fields (Cherry 2003).
Mean* number of ground predators (averaged across all tillage levels) at different sampling dates
Overall, catches of ants not including imported fire ants were quite low (0 - 0.7/trap) in all sampling dates with highest (0.7/trap) number counted on the last sampling date. Reagan et al. (1972) reported that no ants other than the imported fire ants were collected in a Louisiana sugarcane field and that other ants disappear from sugarcane fields wherever the fire ants occur. However, Cherry (2003) reported that the red imported fire ant was clearly the dominant ant species in Florida sugarcane fields, but 8 other species also were noted. Number of rove beetles per trap was highest during first sampling date (2.8/trap) and decreased with time. Number of wolf spiders caught increased significantly from 0.8/trap on the first sampling date to 3.5/trap on the third sampling date, but decreased to 0.5/trap on the last sampling date. In interaction of sampling date with tillage, conventional tillage had significantly greater number of wolf spiders (5.3/trap) compared with minimum (3.6/trap) and no-tillage (1.7/trap) on the third sampling date. Catches of other spiders were low initially and increased significantly on the last sampling date. Cherry (2003) reported that wolf spiders were the dominant spider family in pitfall traps in Florida sugarcane fields. This is consistent with this study in which wolf spider catches were greater than all other spiders combined.
Most importantly, our data indicate that there were no significant short- or long-term effects of tillage on catches of any arthropod ground predators in this study (Table 1). However, there are reports in the literature which showed positive effects of no-tillage on ground predators. For example, continuous no-tillage was reported to increase ground beetles (House and All 1981, Environ. Entomol. 10: 194 - 196) and spiders (Blumberg and Crossley 1983, Agro-Ecosystems 8: 247 - 253) in sorghum. Our initial data show that if Florida sugarcane growers switch from conventional to no tillage there should be no negative impact on population of arthropod ground predators. Further long-term research needs to be conducted to validate the results.
The authors thank Alvin Wilson, Jairo Sanchez, Richard Bryant, and Calvin Howard, EREC, Belle Glade for their assistance in data collection.