Row Cover Use Methods for Cucurbit Pest and Pollinator Management

2015 Trials

Squash yield data for the 2015 trials is shown in Table 1.  Fruit production during the first 2-week harvest period was primarily in the plots that were not covered or those with covers removed at first flowering.  At the Shawnee and Bixby locations the first harvest period yields were significantly greater in these treatments as compared to the DCR treatment and the treatments with brief cover removal and replacement (Table 1). 

During the second 2-week harvest period yields did not differ significantly across treatments for the Atoka and Shawnee locations.  However, there was a numerical trend at each location of greater yields in the BCR treatment compared to LCR treatment.  Similarly, for the Bixby location yields were substantially and significantly lower with the LCR treatment as compared other treatments. 

During the final 2-week harvest period the responses across treatments differed for the 3 locations.  At the Atoka location only the untreated plots had appreciable fruit yields.  At Shawnee yields were comparable across treatments and were numerically comparable to the yields obtained in the second harvest period.  For the Bixby location yields were comparable for all treatments except the untreated plots where fruit numbers were significantly lower.  This is explained by the fact that many of these plants had died. 

2016 trials

Yield data for the 2016 trials is presented in Table 2.  For the Atoka and Shawnee locations fruit production during the first 2-week harvest periods was primarily in the plots that were not covered or those with covers removed at first flowering (Table 2). At Bixby there were comparable yields across all treatments except for the untreated plots. The untreated plots had significantly lower yields than where the cover was removed completely at first flowering.  An explanation for the fruit production in plots with delayed cover removal and brief uncovering is that covers were removed or damaged by high winds and plots remained uncovered for several days until replacement cover material could be obtained. The low yields in the untreated plots are explained by the fact that plants in some of these plots were weak from the start for undetermined reasons. 

At the Atoka location the second 2-week harvest period yields were comparable across all treatments except for the delayed cover removal treatment.  Here, yields were numerically about one half of the others.  At Shawnee yields were also comparable across the same treatments except for the DCR.  In this case, however, yields were substantially and significantly greater where cover removal was delayed 2 weeks after first flowering.  At Bixby yields were comparable across the untreated, the delayed cover removal, and the BCR treatments.  Compared to these, yields were greater for the treatment uncovered at 50% flowering and the LBR treatment.  Because of the previously discussed cover removal by the wind the explanation of these differences would be rather speculative.  It is of interest that for the treatments with brief cover removal yields were numerically double for the 3-hour cover removal period compared to the 1.5-hour removal. 

For the final harvest period in 2016 all treatments at Atoka were still producing fruit and yields were similar across treatments. At Shawnee appreciable yields were still being obtained during this period.  Substantially greater numerical differences were found in the delayed cover removal treatment and the treatments with brief cover removal compared to the untreated and where covers were removed completely at 50% flowering. At Bixby there were no significant differences among treatments at the final harvest period but the numerical trend of yield values followed a pattern very similar to that of the second 2-week harvest period.

Looking at the results of squash yields across locations and trials it can be seen that covering squash with row covers is an effective means of preventing the pollination that is needed for fruit set. Squash will not set fruit if pollinators are not present (Layton, 2015). This was exemplified in the first harvest period at each location in at least one year of the trials.  There were exceptions for Atoka in 2015 and at Bixby in 2016.  In Atoka, a yield of 4.3 fruits per plot was obtained in the SBR treatment. Looking at the data for individual harvests was pooled to produce the 2-week harvest period values (data not shown), this fruit production came in at the final days of the 2-week period.  At Bixby, as previously mentioned, a wind storm removed or damaged the covers and exposed the plants to pollinator access, which explains the appreciable early yields in all treatments. 

Looking at the squash fruit yields of the second 2-week harvest interval, there was appreciable fruit production in all treatments at all locations and for both years.  For the Atoka location in both years, and Shawnee in 2015, there were no significant differences across treatments. The delayed uncovering treatment produced significantly greater yields that were double those of the other treatments at Shawnee in 2016 but showed no advantage at Bixby in either year. The two treatments with brief uncovering and recovering did not differ significantly from each other in any individual trial and there were not any consistent trends of the duration of uncovering having any advantage over the other. 

For the final harvest period the only significant differences that were detected were found between the insecticide treated control and other treatments. At Atoka in 2015 the yields were greatest in this treatment but lowest at Bixby. At Atoka insecticide application was given substantial attention as squash bugs were abundant in the trial. At Bixby, however, insect pest populations were relatively low and less attention was given to application of insecticides to this treatment. Apparently the plants were attacked by squash bugs which resulted on their decline. Looking across the various treatments that employed covers there is no indication that any of the treatments had a consistent effect on the squash yields obtained in the latter part of the production period. 

The results of these two years of trials suggest that the squash yield responses to the treatments that we evaluated are rather erratic. In certain situations, certain treatments were advantageous as compared to others, an example being at Shawnee in 2016 for the second harvest period of the DCR treatment where yields were at least two times greater as compared to the other treatments.  Based on observations made over the course of this study there are several factors that may have an influence on the effectiveness of a given treatment.  One is that row covers have an influence on plant growth.  Plants that were under the covers were clearly larger than those not covered.  Another observation was that the pollinator complexes varied across the trial locations and years.  Each of metallic bees (a sweat bee), squash bees, bumble bees and honey bees were prevalent pollinators in at least one trial.  It is conceivable that differences in the flower visitation habits of these could have influenced the yields obtained in certain treatments.  For example, the SBR and LBR treatments would likely have been visited more when honeybees, which move rapidly from flower to flower, were the primary pollinator, than when the pollinators were squash bees, which tended to spend much more time in a given flower. 

Over the two years of the trials the primary insect pest that was observed was the squash bug Anasa tristis (De Geer).  This insect is a major pest of squash in the region and due to its activity infested plants are weakened and fruit production and quality are reduced (McLeod).  The attractiveness of squash to this insect was observed repeatedly during these trials when the insect was found walking and ovipositing egg masses on the external surface of the covers that prevented the insect from accessing squash plants.