Water Management to Minimize Pesticide Inputs in Cranberry Production (ANE94.21)

Objectives 1 and 2: We studied the impact of LW on plant and pest populations at paired LW/EW sites in 1995 and 1996. We also assessed year-after-LW impacts on plants and diseases. In addition, we separated the plant impact and yield data for the two major cultivars: Early Black and Howes. There appears to be little overall difference in the response of the two cultivars.

Impact on insects: LW reduced the populations of early-season cutworms and cranberry fruitworm (CFW) in the year of the flood compared to those on EW bogs.

In 1995, we monitored early season insects on 6 LW/EW bog pairs. Action thresholds for controlling early-season insect pests were exceeded at all EW bogs but at only 50% of LW bogs. In 1996, the grower-cooperators monitored early-season insects at 7 EW/LW pairs. Based on their sweep counts, they applied 60-70% fewer early-season insecticides to the LW bogs compared to the EW companions.

In 1995, CFW was monitored at 4 EW/LW pairs by berry collection and inspection for eggs. Action threshold was exceeded at all 4 EW sites but at only 1 of 4 LW sites. Five paired EW/LW sites were monitored for the presence of CFW eggs in 1996. Berries were sampled and examined for viable CFW eggs at 3 times: 50% out-of-bloom (time of peak moth flight), 7 days later (time of first spray in current management scheme) and 10 days after the second collection (time of current second spray). At one site no CFW eggs were detected. At the remaining 4 sites, 3 of the LW bogs had lower infestations of CFW eggs than their EW companions (all EW above the action threshold). The very low infestations on the LW bogs would have resulted in no recommended pesticide inputs for all but one of the LW locations. Growers applied no sprays to the LW bogs at 2 of the sites. At 2 others, sprays were applied in the middle of July targeting Sparganothis fruitworm, a pest not controlled by LW.

In both years of the study, LW bogs showed no significant increase in insect damage to fruit despite fewer insecticide applications. In 1995, harvest deductions for 'trash', part of which is shriveled fruit attacked by CFW, were lower on LW bogs (4.6%) than on EW bogs (5.9%) despite a 36% reduction in CFW sprays on the LW bogs. In 1996, growers applied 16% fewer fruitworm sprays to LW bogs than to EW bogs (3.2 vs. 3.8). In many years the number of sprays to the LW bogs would have been even lower. However, 1996 was a bad season for Sparganothis fruitworm - many of the sprays applied to LW bogs were for control of that insect which is NOT controlled by LW. In fact, much of the insect damage found in samples from both LW and EW bogs was from Sparganothis. In 1996, LW bogs had slightly higher (but non-significant) insect damage at harvest (5.51%) compared to EW bogs (4.09%).

Impact on diseases: LW significantly reduced fruit rot incidence allowing growers to significantly reduce fungicide use in the year of late water and in subsequent years with no increase in incidence of upright dieback disease.

In 1995, we compared at-delivery deductions for fruit rot based on grower's delivery slips. Project growers lost less to fruit rot on LW bogs than on EW bogs (6.5% vs. 6.9%) despite applying 47% fewer fungicide sprays. In 1996, at harvest fruit samples were collected and examined from 6 LW/EW bog pairs. LW bogs had less than half the total rot (combined at harvest rot, 30, 60, and 90 day storage rot) of EW bogs despite being treated with 24% fewer fruit rot fungicide sprays (2.5 vs. 3.3). Based on these results, it is now standard practice to reduce fungicide use in the year of LW.

LW changed the populations of the fungal pathogens which cause fruit rot in cranberries. From 1994-96, four pairs of EW/LW beds were examined each year to determine incidence of fungal pathogens in the fruit. Beginning in late July and continuing at weekly intervals until late September, fifty symptomless berries were sampled from five areas in each bed (appendix Table 1). Late water appears to effectively eliminate most fruit rot pathogens, with the exception of Physalospora and Phyllosticta elongata which were able to colonize the berries in the absence of the other fruit rot pathogens.

The incidence of upright dieback disease did not increase the year after reduced fungicide schedules are used in LW beds. We assessed incidence of upright dieback disease in EW/LW pairs in the year after LW (appendix Tables 2 and 3). Four pairs were assessed in 1995 (LW in 1994) and 1996 (LW in 1995). Over all pairs in both years, there was no difference in upright dieback incidence in EW vs. LW bogs. This was true even for those pairs where fungicide use was decreased on the LW site.

Impact on weeds: Late water floods reduced dewberry (Rubus hispidus) populations. The reduction was observed both by empirical transect observation counts and by tracking the survival of primary crowns and offspring. Bogs receiving a LW flood had an average of 27% fewer viable rooting (offspring) and non-rooting (viable crowns only) shoots compared to the EW companion beds (appendix Table 4). The effect of late water floods on the mean number of runners produced per viable crown was variable. Even so, the net production of viable offshoots from the LW bed was 9-30% (an average of 20%) less than on the EW beds. In addition to reducing the number of viable crowns, LW reduced the vegetative spread of dewberries. Future study is needed to determine if concurrent years of LW (or perhaps one season in between) would further weaken the dewberry population.

LW did not reduce populations of two other key weed pests. Late water floods did not reduce populations of glaucous greenbrier, Smilax glauca. Late water floods did not reduce the germination of dodder (Cuscuta gronovii) seedlings (appendix Table 5).

Impact on cranberry plant growth and flowering: LW bogs had fewer uprights and flowers per unit area then did EW bogs, and uprights tended to be longer. Growth and flowering were compared for LW/EW pairs in 1995 and 1996 (appendix Tables 6 and 7). LW bogs had fewer flowers than did EW bogs; due to both fewer uprights that flowered and fewer flowers per upright. Early Black flowering was much more affected by LW than was flowering in Howes in 1996, being reduced by almost 40%. This severe effect on flowering accounted for the greater impact of LW on Early Black yield (see next section). Despite having 25% fewer flowers than companion EW bogs, LW bogs had less than 10% lower yields in 1996. Increased fruit set may overcome decreased flower number. Vegetative uprights were longer or the same length on LW than on EW bogs despite significantly lower fertilizer use.

Impact on cranberry yield: LW impact on yield may be severe if overall growing conditions are such that crops are generally poor. In 1995 and 1996, MA cranberry crops were reduced due to several weather factors including an abnormally warm winter (1994-1995) and drought conditions (1995). Many bogs had poor crops in 1995 and those that did not, tended to have poor crops in 1996. Since we began studying LW in the early 1990's, 1995 was the only year with a very poor LW (appendix Table 8). In fact, in some cases, bogs that were LW in 1995 did not recover to their previous yield potential until 1997. Based on these outcomes our current recommendation is to avoid LW if the previous winter has been abnormally warm. This phenomenon requires further study.

Compared to EW companion bogs, yield was reduced by 35% on LW bogs in 1995 and by 9% in 1996. Cultivar differences were inconsistent with Howes having poorer LW outcomes in 1995 but Early Black having poorer outcomes in 1996. Yield in 1996 was low on both EW and LW bogs compared to bog histories (21-25% decrease). Residual stress from the 1995 drought and the cold 1995-1996 winter may account for this. LW Early Black bogs seemed to be more affected than EW Early Black bogs. In general, LW had little effect on yield in 1996, certainly the savings realized by using fewer pesticide and fertilizer applications would make up for any minor differences in yield. Yield was not reduced by LW in 1993 or 1994. LW had a large adverse effect on yield in 1995 only.

Objective 3: Management protocols for LW bogs were developed and tested.
IPM monitoring for insects: Based on the comparison of two seasons of data, it appears that IPM scouting techniques for early season cutworms and spanworms and for timing control spray for Sparganothis fruitworm and cranberry girdler should not differ on late water and early water beds. Current standard monitoring techniques should be used on all bogs. In most cases, early-season insecticide sprays for cutworms could be eliminated on LW bogs. However, the potential does exist for movement onto the bog of populations from surrounding EW bogs.

Two approaches were taken regarding the monitoring of CFW: examination of berries for the presence of viable CFW eggs (see Objectives 1 and 2), and developing a protocol for the use of CFW sex pheromone traps as part of the IPM practice for this insect on both LW and EW bogs. While our research has shown that LW successfully eliminates on-bog populations of CFW, we have also shown that there is a low level population off-bog in the woodlands.

To date, results with the CFW traps have been too inconsistent for us to recommend their use for deciding to spray or for timing sprays for CFW. However, they remain an excellent research tool for studying the movement of these key insect pests.

It is important for a LW grower to continue scouting for CFW eggs even if none are found soon after fruit set. Populations may move back in from alternate hosts, the uplands, and adjacent EW bogs. In 1996, we looked at the movement of CFW moths using a mark recapture system. We also used the experimental pheromone traps to monitor moth flight differences between cranberry and blueberry stands. Male moths were highly mobile - some flew straight to the woods upon release. Movement of male moths was concentrated around the edges of the bog. Female moths also moved between woods and bog, although less so than the male moths. High moth mobility makes it likely that CFW can move onto LW bogs from surrounding bogs and uplands.

Moths invaded from external population reservoirs in blueberry stands. Some cranberry sites have commercial blueberry stands adjacent to the bogs. We placed 2 traps in blueberry stands and 2 traps in cranberry at 4 different sites and monitored them once a week. Commercial blueberry stands definitely supported a population of CFW. A grower should not assume that CFW pressure has been eliminated by a LW treatment if there is a stand of blueberry nearby.

Insect management: We recommend reduced insecticide use on LW bogs as is indicated by a thorough IPM scouting program. In all years of this project, growers have been able to reduce use of insecticides for early-season cutworms and especially for CFW with no economic loss. See also IPM section above.

Disease management: The number and rate of fungicide applications for fruit rot disease can be reduced during the first and second years after LW is used in addition to reductions (or elimination) in the year of LW. Plots were set up in 1995 and 1996 to examine whether fruit rot could be controlled with fewer than the standard 3 fungicide applications in the year after late water. In addition, lowest and highest label rates for the fungicide were compared (appendix Tables 9 and 10). In both years, fungal fruit rots were controlled by fewer than 3 sprays at the lowest label rate. Plots were set up in 1996 to examine whether fruit rot could be controlled with reduced fungicides during the second year after LW (LW in 1994) with fewer fungicide applications. Fungicide was applied 1, 2, or 3 times at a mid-range rate (appendix Table 11). The data indicate that although a full fungicide schedule and fungicide rate are not required during the second year after late water, eliminating fungicides led to an increased incidence of fruit rot.

Weed management: We studied how LW fits into a management protocol for controlling two key weed pests: dewberry and dodder.

Dewberries: Even though late water floods reduce crown and runner fecundity, this single cultural practice is not sufficient to provide adequate economic control of this pest. Dewberry management must be multifaceted; the grower must be consistent and persistent in the application of any integrated program that targets dewberry.

Follow-up treatments with glyphosate may be beneficial following LW. Results were variable but some trends were observed. Plants clipped with an herbicide-applicator produced less runners per crown than the control at one out of two sites. Hand-wiping was much more labor-intensive than the clipper technique and did not show a significant reduction on the number of runners produced by Rubus spp. in the year of LW. Treatment with either technique increased crown mortality between 7-27% compared to untreated plants. Accidental removal of flags at the grower sites prevented re-evaluation of the treatments in 1997. Further study is needed to verify the efficacy of these techniques following LW and to determine if they are more effective after LW compared to EW.

Dodder: Dichlobenil has been used on cranberry bogs as a pre-emergent herbicide to control the parasitic weed dodder. We studied the effectiveness of dichlobenil when used in conjunction with LW (appendix Table 12). When 100 lb/A was applied prior to the flood, herbicidal activity remained after the flood. This was not the case for the 70 lb pre-flood rate. Pre-flood applications of dichlobenil at rates lower than 100 lb/A may not effectively control any weeds unless they are combined with low rates applied post-flood. Dichlobenil applied in one season may not exceed 100 lb/A.

On bogs where dodder is a major problem, LW use may not be indicated. This is based on poor longevity of pre-flood casoron and dangers associated with post-flood applications (phytotoxicity), along with the fact that LW does not suppress dodder germination.

Fertilizer management: On LW bogs studied in 1995, growers reduced N use by about 70%, perhaps contributing to the low yield on the 1995 LW study bogs. Those same bogs recovered some in 1996 but continued to have yields 20% lower than their previous 5 year averages. Clearly such drastic reductions in N fertilizer are not warranted despite the fact that plant growth appeared normal. In 1996, growers reduced N use on LW by only 23% compared to the EW companion bogs and yield differences between LW and EW bogs were much less apparent. Previous research has shown that N use may be reduced by 30-35% in the LW year without impact on yield in the year of LW or in the following year.

In order to assess N rates for LW bogs in the year of LW and the following year, we evaluated nitrogen rates from 0 to 30 lb/A on LW bogs in 1995 and 1996 (follow-up in 1996 and 1997). Adjacent EW bogs were included in the study for comparison. Results for the year of LW and the 1996 follow-up were presented in previous progress reports. Follow-up results from 1997 are shown in Table 13 in the appendix. In all years of this study, yield differences among the N protocols were not significant. However, based on trends in the data, tentative recommendations can be made as follows:

EW bogs___20-30 lbs N/A
LW bogs___0-20 lbs N/A
LW bogs year after LW___20-30 lbs N/A (return to EW schedule)

Variety differences were minimal but based on previous research, Howes would fall at the upper end of the N rate range.

Objective 4: We continue to investigate the possibility of using LW more often than one year in three. We received anecdotal reports that on rare occasions, LW was held more often than one year in three. However, no grower could provide record to substantiate the outcome. Further, we were unable to find or convince any grower to attempt consecutive or every two year LW for this project. At this time, we hesitate to recommend more frequent LW use and are more concerned with studying ways to ensure good outcomes of LW use, i. e. no vine overgrowth or loss of yield.

Objective 5: Water quality was studied in 10 LW floods in 1995 and 11 LW floods in 1996. Turbidity, oxygenation, temperature: Oxygen was above the 5 ppm critical level (plant health) in the upper levels of the flood on all collection dates (appendix Figure 1) but sometimes was lower at plant canopy level. Decreased oxygen at bog level may account for control of CFW and other insects. However, severe oxygen depletion may also result in decreased yield. Water turbidity was high at the time of flooding (turbulence) but during the flood was generally low, 5 NTU or less. Turbidity increase late in the flood indicates the presence of algae in the flood water and may trigger early flood removal. Temperatures in the upper levels of the flood (appendix Figure 2) and at vine level (appendix Figure 3) were often near or above 600F (the level considered necessary to control CFW) and a few times above 650F (the danger point for plant injury). As we expected based on water quality factors, CFW control was good. Based on these data, we recommend visual monitoring for algal growth and flood temperature monitoring if ambient temperatures are high. Ideally, temperatures in LW should be elevated (60-65F) to kill CFW but higher temperatures may lead to adverse plant effects. Conversely, if temperatures remain low in the flood due to inflow (stream or pumping up flood level), insect control may be compromised.

Nutrients in LW: Nutrients do not appear to be dispersing into the LW flood in significant amounts. We analyzed water from LW bogs for dissolved phosphate in 1995 and 1996 and for nitrate in 1996. At one bog on one date 0.14 ppm NO3 was detected, otherwise levels were generally not detectable with an occasional detection of 0.05 ppm or less. PO4 data for 1996 are shown in appendix Figure 4. All 1996 collections contained less than 1 ppm PO4, most less than 0.2 ppm or not detectable. In 1995, most samples contained nondetectable levels or 0.1 ppm PO4, the highest level detected was 0.5 ppm.

Dichlobenil in LW: Initial residues of dichlobenil in LW floods were highest on bogs that received 100 lb/A of the herbicide prior to the flood. There was some indication that application at least 7 days prior to the flood with adequate incorporation due to rainfall decreased the movement of the material into the flood. The importance of preventing loss into the flood water is obvious. However, the major issue is efficacy rather than environmental as most of the initial residue has decayed by the time the flood is released (appendix Figures 5 and 6).

Objective 6: See 'D. Economic Analysis' below.

Objective 7: see 'B. Dissemination of Findings' below.