Optimizing management of a new invasive species, swede midge, on small-scale organic farms

Objective 1:

Advance understanding of swede midge (SM) pressure and invasion on small-scale organic farms growing Brassicas as it relates to management practices.

1. Intensity and duration of spring emergence.

SM population dynamics are illustrated in figures 14 to 20 and spring emergence trap catch data are generally color-coded in blue, except at Quest produce where they are green.  Dates of first and last trap catches, and peak emergence for each of the spring emergence sites are summarized in Table 8. 

In 2015, first emergence of SM occurred on May-13/15 at 13 out of 15 (= 87%) sites.  At Fellenz, the first SM trap capture occurred two weeks later on May-27, however, this farm had very low outdoor SM pressure in general (Fig. 17), and the later date of the first trap catch may not be due to a micro-climatic effect, but rather a function of the relatively low population.  At Quest, at the low risk site, the first SM capture occurred one week after the first capture at the high risk site.  This difference likely reflected the time delay that it took for SM that emerged from the high risk site (where SM-infested broccoli was grown in fall 2014) to disperse to the low risk site (where brassicas had never been grown).  In our first year of study, first emergence of the overwintering generation of SM consistently emerged in mid-May. In Southern Ontario, Canada, first emergence occurred two weeks later in early June (Marteaux et al. 2015). 

In 8 out of 15 sites (= 53%), spring emergence of SM occurred in bimodal peaks with the first peak occurring for a 1-2 week period during the end of May/early-June, and the second in mid-June (Table 8).  Trap catches were similar among first and second emergence peaks at Canticle (Fig. 14) and Blue Heron Field No. 1 sites (Fig. 19).  The first peak had higher trap catches at Living Acres (Fig. 16) and Blue Heron Field No. 4 sites (Fig. 19), while the second peak had higher trap catches at Quest Produce high risk site (Fig. 15) and one of Blue Heron Field No. 1 sites (Fig. 19).  Bimodal peaks are consistent with the emergence patterns of the overwintering generation of SM in Southern Ontario, Canada (Goodfellow, 2005; Hallett et al., 2009) and in its native parts of Europe (Rogerson, 1963).  In Southern Ontario, the peaks occur in mid-June and early July with the second peak being the larger of the two (Marteaux et al. 2015).  In general, it appears that SM emergence occurs about two weeks earlier in New York than it does in Southern Ontario, Canada.  Bimodal emergence peaks are thought to be a function of two emergence phenotypes of the overwintering cohort, which may be characterized by different diapause durations or post-diapause development. 

                Our trap catch data suggests that SM spring emergence drops off significantly during the second half of July with last trap captures having occurred around Jul-31 2015.  These findings are similar to Southern Ontario, Canada, where 97% of overwintering SM emerge within the first two peaks by the third week in July, while the remaining 3% continue to emerge until early September (Marteaux et al. 2015).  The possible existence of a third late-summer emergence phenotype requires further validation on both sides of the border, because generational overlap after the second peak make it difficult to determine.   Our open-air traps could have captured individuals from either spring emergence or subsequent generations; thus, knowing the definitive end date of spring emergence is warranted.

1. Rate at which SM populations build within a growing season.

With bimodal spring emergence, it is difficult to determine how many generations of SM occurred within the season.  In Southern Ontario, Canada, four generations are known to occur, and it is expected that 4 to 5 generations occur in New York.  At Canticle, the last SM of the season were captured on Oct-15 2015 (Fig. 14), while we still had trap catches on the day that the traps were taken down on Oct-28 2015 at Quest (Fig. 15) and Living Acres (Fig. 16) and on Nov-13 2015 at Fellenz (Fig. 17), Muddy Fingers (Fig. 18) and Blue Heron (Fig. 19 & 20).  In general, SM trap catches remained strong throughout the month of September, and did not drop significantly until after mid-October (Figs. 14 to 20). 

Canticle (Figs. 2 & 14, Table 2). The highest trap catch occurred on Aug-28 2015 in a summer planting of mixed brassicas (Trap 5: 40.4 SM/trap/day) that was 5.7 times as high as the first spring emergence peak (Trap 3) on May-29, 7.1 SM/trap/day).  On this 8 acre farm, SM appeared to not have any trouble moving from the spring emergence sites (Trap 2 & 3) to the summer planting in the middle field (Trap 5).  One month after fall Brussels sprouts were planted in the back field in mid-June, SM were captured in this trap (Fall trial) and the population peaks increased 7.2-fold from 2.8 SM/trap/day on Jul-15 2015 to 20.2 SM/trap/day on Jul-30 and Sep-8 2015.  In the fall planting of mixed kale in the front field (Trap 6), trap catches peaked on Sep-17 at 13.4 SM/trap/day.  SM trap catches remained strong in all of these locations until mid-October. Less than 1.0 SM/trap/day were captured in the open-sided high tunnel where brassica transplants were produced (Trap 1).

Quest Produce (Figs. 3 & 15, Table 3).  At Baker farm, which is 0.7 acre, the broccoli netting trial was set up at the spring emergence site, so the trap catch data illustrates both spring emergence and in-season population dynamics.  Highest trap catches occurred on Sep-17 2015 at 111 and 60 SM/trap/day at the high-risk (trap B2) and low-risk (trap B1) sites, respectively, which were 18.5-fold and 16-fold higher than the highest spring emergence peaks for the high- and low-risk sites, respectively.  There did not appear to be much buildup of SM population during normal growth of broccoli.  After harvest on Jul-15 2015, the broccoli was left in the field for 12 weeks before it was disked up in early October; during this time SM trap catches increased 5-fold and 10-fold at the high- and low-risk sites, respectively.

                At Sexton farm, also 0.7 acre, SM trap catches were low for the duration of the season with maximum trap catches on Sep-30 2015 of 1.0 SM/trap/day in the summer planting of cabbage and broccoli in the field with no history of brassicas (trap S1), and 2.4 SM/trap/day in the fall planting of mixed kale in the field with a history of brassicas (trap S2).

Living Acres (Figs. 4 & 16, Table 4).  At this farm, highest trap catches occurred at the spring emergence site (Trap 1) which peaked at 87.4 and 54.8 SM/trap/day for the first and second emergence phenotypes, respectively.  Despite such high spring emergence on this farm, spring planted kohlrabi (Trap 2) peaked at only 11.8 SM/trap/day on Jun-24.  Trap catches in subsequent summer plantings of cauliflower and Brussels sprouts (Trap 4) and mixed brassicas (Trap 5) in the back had generally less than 10 SM/trap/day until late-August, while the summer planting of mixed brassicas (Trap 3) beside the spring kohlrabi had about 20 SM/trap/day.  SM trap catches doubled at Trap 4 from Jul-2 to Sep-3 2015.  The fall planting of mixed brassicas down from the spring kohlrabi and summer mixed brassicas had the highest trap catches on Sep-30 2015 with 35.6 SM/trap/day, which doubled over the season.  The reason for reduced SM trap catches in 2015 brassica crops compared to the spring emergence site may be because there was a 12 ft fence and a bushy hedge row between the spring emergence site and 2015 brassica crops (Fig. 21).  This scenario appeared similar to the University of Guelph study where 5 ft tall exclusion fencing only temporarily delayed and did not prevent SM infestation.

Fellenz Family Farm (Figs. 5 & 17, Table 5). On this 3 acre farm, spring emergence in the front field was minimal with less than 1.0 SM/trap/day on May-27 2015 (Trap FF).  At most, trap catches in the front field were 4.2 SM/trap/day on Jun-10 2015.  In the fall, no SM were captured in this field (Trap Trial).  From late May until mid-July, the majority of SM captures occurred in high tunnel No. 6, where successions of mixed brassicas were grown, mostly under row cover until it was time to harvest.  Some plantings were not covered in tunnel No. 6, where the highest trap catch on the farm occurred on Jun-17 2015 with 49 SM/trap/day (Trap T6).  Shortly after this, uncovered broccoli raab was harvested and once there were no more brassicas grown in this tunnel, trap catches eventually dropped to zero.  First trap catches inside this high tunnel were similar to timing of first trap catches at spring emergence sites.  Had they been earlier, it could have been an indication of SM over wintering inside the high tunnel.  It is unknown whether SM in tunnel No.6 originated in the field or the tunnel. Trap catches were very low (peak of 6.7 SM/trap/day on Jun-17 2015) in high tunnel No. 2 and dropped to zero when brassicas were no longer grown in the tunnel (Trap T2).  In high tunnel No. 4, which had uncovered Asian brassicas, trap catches were less than 1.0 SM/trap/day. In the back field, which was not planted to brassicas until the fall, had similar SM pressure as the front field (peak of 12 SM/trap/day on Sep-24 2015).  No SM were captured in the brassica transplant greenhouse, which was fully enclosed.  On this small farm, it appeared that their strategy of growing quick-growing brassicas under row cover inside high tunnels (Fig. 22) in combination with delaying planting of brassicas until the fall in fields where brassicas were grown the previous fall to avoid spring emergence (Trap FF & Trial) was successful for managing SM.  

Muddy Fingers (Figs. 6 & 18, Table 6).  Peak trap catches of spring emergence on this farm were 13.1 and 18.6 SM/trap/day on Jun-17 2015 in the small (trap SF1) and new fields (trap NF1), respectively.  During spring emergence (mid-May to mid-July), trap catches in the spring plantings in the small (trap SF2) and big (trap BG2) fields were lower than they were in the spring emergence traps.  In the small field in spring-planted mixed brassicas (trap SF2), trap catches increased 2.4-fold in mid-August compared to mid-July.  However, in the summer planting of broccoli in this field, SM trap catches never built and never exceeded 8 SM/trap/day (trap: summer netting trial – open air).  In the big field in spring broccoli, after main harvest was completed, SM population built 3.9-fold to 45.8 SM/trap/day and 5.7-fold to 66.7 SM/trap/day from Jul-14 2015 to Aug-5 2015 and Aug-26 2015, respectively (trap spring netting trial – open air). Interestingly, in the big field in plantings of Red Russian kale (trap BF2) and Asian brassicas (BF1) adjacent to the spring broccoli, trap catches were less than 5 SM/trap/day.  On this 5.5 acre farm, it was interesting that when spring broccoli was moved into a field that had not had brassicas during the previous fall (big field) that it took until July for the SM that presumably emerged from the small or new fields to find the spring planting of broccoli.  However, the majority of this planting was grown under insect exclusion netting.  It was not until the netting was removed and after main harvest of broccoli heads that that SM population exploded when the crop was left intact in order to harvest secondary side shoots, and then not mowed in a timely fashion after side shoot harvest, because the mower was broken. These results demonstrate the great potential for SM to build tremendously after harvest when the crop is not destroyed.

Blue Heron (Figs. 7, 8, 19 & 20, Table 7). At Blue Heron, three production fields are isolated from each other by wooded areas with 400 to 500 feet between fields.  In field No. 1 (4.4. acres), in the absence of a brassica crop, very high spring emergence trap catches plummeted (Fig. 19 top).  In fall cabbage that was planted in late-July, the peak trap catch was on Sep-15 2015 at 15.6 SM/trap/day. 

                In field No. 4 (12 acres), which also had very high spring emergence trap catches, the SM population built very quickly in a spring planting of mixed cabbage (trap F4T1), where it increased 1.7-fold from Jun-5 2015 (28.2 SM/trap/day) to Jun-25 2015 (48.8 SM/trap/day) and 2.7-fold to Jul-14 2015 (76.5 SM/trap/day) (Fig. 19 bottom).  Summer cabbage that was also planted during spring emergence in this field (trap F4T4) had high trap catches (peak: 52.3 SM/trap/day on Jul-14) and trap catches in fall plantings of Romanesco/cauliflower (trap F4T7), cauliflower (trap F4T4) and broccoli (trap F4T4) all had high trap catches, which peaked at 72.3, 43.7 and 39 SM/trap/day.  Trap catches remained strong throughout October in the fall broccoli.  The SM trap catch data in this field demonstrated how SM populations continue to build across the season as long as there are new plantings of susceptible brassicas available.  An open 12- acre field did not provide enough distance for crop rotation to be effective.  Alternatively, as demonstrated in field No. 1, waiting to plant brassicas until the majority of spring emergence had passed crashed SM population and drastically reduced SM population in fall brassica planting.  Previously, the recommendation for far and wide crop rotation was 3000 feet and 3 years, which is nearly impossible for at-risk small-scale organic brassica farms.  But in this case study, far rotation of an isolated field for a duration just long enough (2.5 months) to avoid the majority of spring emergence (May through mid-July) proved to be effective.

                Field No. 6 (3.7 acres) had the lowest spring emergence (peak of 27.4 SM/trap/day on Jun-10) of the three brassica production fields at Blue Heron (Fig. 20 top).  Romanesco was planted in late-June (trap F6T2) and late-July (trap F6T3).  In the first planting, SM spiked right after planting to 7 SM/trap/day on Jul-8 2015, but then dropped to less than 1 SM/trap/day for the rest of the season.  It is possible that the spike in the SM population in this planting was from the spring emergence population, because it was planted when the SM trap catch was 14 SM/trap/day. However, it could also be from SM-infested transplants, as SM-infested plants were observed in the hardening off area and several of the newly transplanted plants were blind (suspect SM damage).  Why the population dropped off after about a month and stayed down is not understood.  In the later planting, which was planted after the majority of SM had emerged, the population increased from 1.6 SM/trap/day 2.2-fold to 3.5 SM/trap/day and 6-fold to 9.7 SM/trap/day from Jul-22 2015 to Aug-17 2015 and Oct-14 2015, respectively (Fig. 20 top).  SM trap catches remained strong throughout October in this planting. This field provides a good example of how crop rotation is defeated when SM-infested transplants are planted in a “clean” field.

                Traps were set up in three high tunnels where brassica transplants were produced at Blue Heron and although the odd SM was captured in these traps, captures were generally less than 1.0 SM/trap/day (Fig. 20 bottom).  Slightly higher trap catches occurred at the hardening off site (peak of 7 SM/trap/day on Jul-8 2015) and low catches remained through September.  The hardening off area was located just outside of the high tunnels.  Even though brassica transplant production was seemingly isolated from brassica field production and separated by about 500 feet at Blue Heron, the fact that SM-infested transplants were observed both in the hardening off area and in newly transplanted plants in the field is an indication that production of clean transplants on this farm warrants further attention.           

iii) Relative susceptibility or SM-preference of different Brassica plant types.

In three side-by-side comparisons, red cabbage had higher levels of SM incidence and significantly more unmarketable plants than green cabbage.  At Living Acres under high SM pressure, red cabbage had 55% SM incidence with 25% unmarketable, compared to 35% SM incidence with 12.5% unmarketable in green cabbage (Table 9).  At Blue Heron, under very high SM pressure in Field No. 4, red cabbage had 100% incidence with 72% unmarketable, and 97% incidence with 60% unmarketable in the spring and summer plantings, respectively, which were significantly higher than green cabbage (spring: 40% incidence/15% incidence; summer: 52% incidence/3% unmarketable) (Table 10).  At Blue Heron under low SM pressure in Field No. 1, red cabbage had only 25 and 48% SM incidence with only 2% and 6% unmarketable in two fall plantings of red cabbage (Table 11). This was the field where the growers waited until the majority of the spring emergence had passed before planting a brassica crop.  Compared to field No. 4 where brassicas were planted during spring emergence in spring and summer, wide rotation in field No. 1 increased marketability by 54 to 70%.  At Canticle, in a mid-season evaluation, there were no differences between red and green cabbage where SM infestation was only 8.3% (Table 12).  

                Highest levels of SM infestations occurred in broccoli which ranged from 55 to 100% SM incidence with 23 to 60% unmarketable heads under very high SM pressure at Blue Heron Fields No. 4 and 6 (Table 11 & 10) and at Canticle (Table 12).  Also under very high SM pressure in Blue Heron Field No. 4, Romanesco and cauliflower were equally as infested as broccoli with 70 to 98% SM incidence and 35 to 58% unmarketable heads.  In the same field, SM infestation and damage was similar in cone cabbage as it was in green cabbage (Table 10).  At Living Acres under high SM pressure, Brussels sprouts and cauliflower had 83% and 70% infestation with 20% and 23% unmarketable plants, respectively, in a mid-season evaluation, while Chinese cabbage remained clean.  In both plantings on this farm, the cabbage, Brussels sprouts and cauliflower had higher SM infestation than winterbor and dinosaur kale.  There were no significant differences in SM infestation between these two types of kale, although numerically, winterbore (15%) had more than dinosaur (0%) in the summer planting, while dinosaur (49%) had more than wintebore (23%) in the fall planting (Table 9).  At Canticle, highest levels of SM infestation also occurred in Red Russian (summer - 75%; fall – 82%) and White Russian kale (75%), which were significantly higher than other kale types.  In the summer planting, purple, winterbor and dinosaur kale had 30%, 20% and 5% SM infestation, respectively.  In the fall planting, rainbow lacinato kale had 47% SM incidence while winterbor, red curly and dinosaur kale had no SM damage (Table 12).  At Quest, both Red and White Russian kale had 100% SM infestation, with Red Russian having significantly more unmarketable plants (54%) than White Russian (30%).  Winterbor (17.5%), Lacinato (16%) and Red curly kale (3%) all had significantly much less SM infestation and less severe damage (Table 13).  At Muddy Fingers, there were no significantly differences in SM infestation among green and red Brussels sprouts, and peacock, dinosaur, and winterbore kales, although wintebor kale numerically had the lowest infestation (3%) compared to the others (range; 30 to 55%) (Table 14).   Although no data was recorded, in the big field in a fall planting right beside the broccoli trial where SM pressure was very high, no damage was detected in various turnips, and Chinese cabbage, which suggests that these crops may not be preferred or are tolerant to SM.

                Based on our first year of observation, broccoli and Red Russian kale appear to be the most preferred/susceptible brassica crops to SM, although SM results in more unmarketable damage in broccoli, because only the heads are harvested, which have a zero tolerance for SM damage.  In kale, the leaves are harvested, and when not all of the leaves are damaged, there are still marketable portions on an SM-damaged plant.  Cabbage, especially red cabbage, can be as preferred/susceptible to SM as broccoli when it is attacked prior to head formation.  After head formation, its tolerance to SM increases substantially as SM does not prefer it over a plant where it can easily access the growing point.  Curly types and dinosaur kale, turnips and especially Chinese cabbage appeared to tolerate very high SM pressure in this project.

1. Rate at which SM populations drop after harvested crop is destructed

In 2015, we had three sites where we recorded mowing events in relation to SM trap catches.  At Canticle on the new farm, a relatively low population slowly built up to a moderate-high level (9 SM/trap/day) when it was mowed, after which the population plummeted to 0.7 and then 0.0 SM/day/day for the next two weeks (Fig. 14).  At Fellenz, in high tunnel No. 6, an uncovered broccoli Rabb crop built SM pressure up to 40 SM/trap/day at its peak, which crashed to 11.4 and 8.2 SM/trap/day after the crop was harvested and plant residue disked up (Fig. 17). At Muddy Fingers in the big field, a very high population was mowed at its peak when the trap catch was 67 SM/trap/day (trap BG2), which appeared to crash the population to 1.9 SM/trap/day. The week after, the population increased to 19 SM/trap/day, but after that, trap catches were less than 2 SM/trap/day for the rest of the season (Fig. 18).  These results are very similar to our previous experiences with SM monitoring where for example, a population of 11 SM/trap/day crashed to less than 1.0 SM/trap/day within two weeks after mowing (Hoepting, 2006). At Blue Heron in Field No. 4, the spring planting of mixed cabbage (trap F4T1) was finally disked under four weeks after an already prolonged 6 weeks of harvest just when the SM population had dropped from 76.5 SM/trap/day to 22 SM/trap/day (Fig. 19).  After mowing, SM increased again to 29 SM/trap/day before dropping to 10 or less for the rest of the season.  During this time, SM trap catches were very high in the adjacent fall plantings of Romanesco and cauliflower.  In this case, it appeared that the SM had used up their resources in the spring cabbage and had moved on their own to the new plantings. Clearly, post-harvest crop destruct via disking or chopping/flail mowing were extremely effective in reducing SM pressure at these sites. Theoretically, during post-harvest crop destruct, SM larvae are destroyed as the growing points are destroyed, or younger larvae within intact growing points die as the plant tissue dies.  SM larvae that are almost ready to pupate in intact growing points may still drop to the soil to pupate, but when they emerge they have to spend more of their 3-day lifespan looking for a mate and a brassica host to lay their eggs, which may overall reduce the rate of SM population buildup at the site/farm.

Objective 2.

Optimize implementation of newly developed disruption tactics including insect exclusion netting and garlic oil repellant to manage SM in Brassicas on small-scale organic farms.

1. Compare insect exclusion netting used on ground previously cropped to SM-infested crop to ground rotated away from brassicas for several years.

Quest Spring Broccoli Netting Trial, 2015 (Table 15).  In the open air traps over bare ground, the high-risk site which was located in the exact location where an SM-infested crop was grown the previous fall in 2014, had 2.6-fold higher total SM trap catches than the low-risk site, which was about 70 ft away on ground that had never been cropped to brassicas. The trap catches at the high-risk site (May-29: 5.2 SM/trap/day; Jun-19: 13.1 SM/trap/day) were 4.7- and 4.3-times higher on May-29 and Jun-19, respectively, than they were at the low-risk site (May-29: 1.1 SM/trap/day); Jun-19: 3 SM/trap/day) (data not shown).  By mid-July, trap catches at both sites were similar (11.4 SM/trap/day).  The 2.6-fold difference in trap catches between sites is likely due to the slight delay in SM finding the broccoli planted in the low risk site that emerged at the high-risk site.  Once the first generation established itself at the new site, the second generation increased substantially.

                At the high-risk site, the trap on bare ground under insect exclusion netting captured only 28% (= 164 SM/12 weeks) of what was captured in open air.  However, 40% more SM (= 229 SM/12 weeks) were captured under insect exclusion netting with straw mulch.  This result suggests that straw mulch does not prevent SM from emerging from the soil beneath it.  It also may indicate that if SM emerge from the soil into insect exclusion netting that SM may build to higher levels, theoretically because the enclosure would enhance the ability of female SM adults to find a mate and suitable host on which to lay eggs.  With the small sample size of plants in this demonstration (i.e. 20) and such high SM pressure, level of SM infestation and unmarketable heads due to SM damage was the same among all of the treatments (89 to 95%).

                At the high-risk site, under insect exclusion netting, total SM trap catches were 26% lower (= 110 SM/12 weeks) with black plastic mulch than they were with bare ground, and SM damage was only 33% infested/unmarketable heads.  Theoretically, plastic mulch may prevent SM from emerging from the soil.  In this trial, bare ground from which SM could have emerged, was exposed along the edges of the netting as the netting was 1 ft wider than the plastic mulch, and through the transplant holes.  Slightly less than 10% of the plants in this treatment showed signs of heat stress (e.g. yellow or uneven beading).  At the low-risk site, 35% and 17% of the broccoli had heat stress in this treatment and insect exclusion netting over straw mulch, respectively.  Also at this site, the broccoli grown under insect exclusion netting had delayed maturity on bare ground compared to plastic.  These results suggest that insect exclusion netting, especially when used in combination with mulch may modify the microclimate, which may have positive or negative effects, such as advanced maturity or heat stress, respectively.

                At the low risk site, broccoli grown on bare ground in open air had 67% infested/unmarketable heads, while the broccoli under insect exclusion netting had 0% SM, representing 100% control.  These results suggest that as long as insect exclusion netting is used on ground that has not been cropped to SM-infested brassicas that it can be extremely effective for controlling SM even if it is used in a field that has high SM pressure.  Although 31 SM/12 weeks were captured under the netting in the bare ground treatment, this did not occur until after Jul-7 when it was noticed that the end of netting had come lose.  Thus, ensuring that the edges of insect exclusion netting are secured to the ground so that wind can’t blow them open or weeds lift them up is important in ensuring that SM do not enter from the outside into the netting enclosure.

                Very high numbers of imported cabbage worm butterflies and larvae and feeding damage was observed inside the insect exclusion netting, especially in the straw mulch treatment at both sites.  This illustrates the importance need to start with clean transplants; if transplants are infested with worms, the problem will only get worse under the netting.  It was also observed that the bare ground and straw mulch treatments had the most weeds under the netting, while weeds only occurred along the edges with the plastic mulch.  Thus, weed control is an added benefit of using mulch in combination with insect exclusion netting, so that the netting does not have to be lifted for weeding operations, which may risk entry of SM.

2. Evaluate insect exclusion netting in combination with different mulches.

Muddy Fingers Spring Broccoli Netting Trial (Table 16).  At Muddy Fingers, this insect exclusion netting trial was set up in the big field on ground that had not had brassicas grown in at least a year, but other areas of the field had SM-infested brassicas the previous year.  In the two open-air traps on bare ground, SM pressure was relatively low for the 8-week duration of the trial.  In the Diplomat broccoli variety, 59 SM/8 weeks (= ~ 1.1 SM/trap/day) resulted in 76.5% SM infestation, while the Aracdia variety had 19 SM/8 weeks (= ~0.3 SM/trap/day) and significantly less SM infestation (19%).  With only one replication, it is not possible to elucidate whether this is a differential varietal response to SM, or whether it is a consequence of the location of the trap.  The open-air trap in Aracdia was tucked in between three netted treatments and the side of a high tunnel (Fig. 12).  None of the broccoli plants were infested with SM in any of the netting treatments, except the one with black plastic mulch which had only 1% SM infestation, representing 99 to 100% control of SM with insect exclusion netting.  Even though no SM were captured in the trap in the netting + black plastic mulch treatment, SM must have gotten under the netting either from the outside through the netting, or it emerged from the soil into the netting.

                Insect exclusion netting also significantly reduced moderate levels of damage from both flea beetle and imported cabbageworm to a minor level.  In the Diplomat variety, compared to the broccoli grown on bare ground in the open air, plant height at harvest was taller by 36% (= 9 inches)  in the treatments where insect exclusion netting was over bare ground and black plastic mulch, and by 16% (= 4 inches) and 21% (= 5 inches) over landscape fabric and hay mulch, respectively.  We also observed that the broccoli under netting matured earlier than the broccoli grown in the open air, and the broccoli grown under insect exclusion netting with black plastic mulch matured earlier than the broccoli with hay mulch (Fig. 23).  There was generally no difference in head weight at harvest and all treatments with insect exclusion netting yielded twice as much as the open air broccoli due to the excellent SM control that it provided.  Insect exclusion netting with black plastic mulch yielded the highest at 64 lb/100 ft of bed.

Muddy Fingers Late Summer Broccoli Netting Trial (Table 17).  This insect exclusion netting trial was set up in the small field on ground that had not been cropped to brassicas in about 5 years, but in other areas of the field, SM-infested brassicas were grown the previous year and in 2015.  Trap catches in the open air treatments were 107, 121 and 332 SM/8.7 weeks (= 1.8, 2.0 and 5.5 SM/trap/day) with black plastic mulch, landscape fabric and hay mulch, respectively.  Five or less SM/8.7 weeks were captured under insect exclusion netting with black plastic mulch and with landscape fabric, while 21 SM/8.7 weeks were captured with hay mulch, with captures occurring for the duration of the project.  Either there was a tear in the insect exclusion netting from the beginning or the SM were emerging from the soil; perhaps the forth bed (hay mulch treatment) overlapped with ground that was planted to SM-infested brassicas in 2014.  Despite the odd SM under the netting, SM damage in this treatment was only 6% infested (4% unmarketable).  In the other insect exclusion netting treatments, SM infestation was 0%.  In the open-air plots, SM infestation ranged from 92 to 100% infestation with 50 to 85% unmarketable. In this trial, the insect exclusion netting provided 94 to 100% control of SM. 

                In this trial, we also observed that the broccoli under netting with black plastic mulch matured earlier than the other treatments.  Of plants under netting, those with hay mulch appeared larger and healthier than those with landscape fabric.  A few plants with aphids (3%) were observed under the netting with landscape fabric and hay mulch, and minor incidence of heat stress was observed in the plants grown under netting with black plastic mulch and hay mulch.

Canticle Fall Brussels Sprouts Netting Trial (Table 18).  This insect exclusion netting trial was located in the back field on ground that had not been cropped to brassicas in at least 2 years.  The only open-air trap (on bare ground) captured 597 SM/15.3 weeks (= 5.6 SM/trap/day), while only 8 and 2 SM/trap/15.3 weeks were captured under the netting treatments.  In the open-air treatments, SM infestation was 60% and 42.5% in bare ground and with black plastic mulch, respectively.  Insect exclusion netting significantly reduced SM incidence to 3.6% and 2.5% with bare ground and black plastic, respectively, representing 94% control of SM with insect exclusion netting.  In none of the treatments was SM damage more than minor. 

Interestingly, the netting treatments resulted in significantly taller plants, smaller sprouts and no blind sprouts compared to the open-air treatments.  Brussels sprouts plants were similar in height under netting with bare ground (28 inch) and black plastic mulch (27 mulch), which were 42% (=8 inches) and 28% (=6.3 inches) taller than these treatments in open air (Fig. 24).  In open air, the plants were significantly 2.7 inches taller on bare ground than on black plastic mulch.  There were no significant differences in the number of sprouts per plant between the open air and netting treatments, although numerically the plants under netting had more sprouts.  The plants grown on bare ground (diameter: OA – 1.8 inch; IEN – 1.5 inch) had significantly larger sprouts than those grown on black plastic mulch (diameter: OA – 1.2 inch; IEN – 0.8 inch), and in both cases, those in the open air were significantly larger than those under netting.  There were several blind sprouts (no sprout formation in leaf axil, Fig. 25) that did not appear to be caused by SM in the plants grown in open air (BG – 4.3%; plastic – 2.7%), that did not occur under the netting.  It was also observed that the quality of the sprouts under netting was much better than those in open air, which had a lot of Alternaria leaf spot (Fig. 24).  Although we were not able to get yield data from this trial, the grower reported that the Brussels sprouts continued to grow and produced nice high quality sprouts after the netting was removed on Oct-15 2015 when we made our harvest evaluation.

In this trial, we set up temperature sensors in the bare ground treatments and it appeared that the netting had a moderating effect on the temperature within the netting compared to the ambient temperature (Fig. 26).  During the heat of summer when the average weekly average temperatures were in the 70s °F (Jul-22 to Aug-19), the average weekly maximum temperatures inside the netting were 2 to 18 °F cooler than the outside.  Then, when the average weekly temperatures were less than 70 °F (Aug-26 to Nov-5), the average maximum weekly temperatures were 1 to 6 °F warmer inside the netting than outside.  Although growing degree days (GDD, base 45 °F) accumulated slower at first, by the time of our harvest assessment on Oct-15 2015, inside the netting had accumulated 136 more GDD than outside.  After the netting was removed on Oct-15-2015, the treatment originally under netting had accumulated only 23 GDD more than the open air by Nov-5-2015.

Fellenz Fall Kohlrabi Netting Trial (Table 19).  SM pressure was extremely low in this trial with only 1 and 4 SM/trap/8 weeks in the open-air traps, so it was not surprising that none of the kohlrabi had any SM damage.   However, the insect exclusion netting significantly reduced infestation and damage from imported cabbageworm (ICW).  The open-air treatments had 99% and 96% ICW infestation with minor damage, while the insect exclusion netting with black plastic mulch had only 2.5% ICW infestation, which was significantly lower than insect exclusion netting with hay mulch (22%). 

                Temperature sensors were set up in all of the treatments in this study. Overall, cumulated GDD (base 45 °F) were practically identical among treatments for the first 6 weeks of the trial until early October when the open air treatments began to accumulate more GDD (Fig. 27).  At the end of the trial, the two open-air treatments had accumulated similar total GDD, which were 53 and 43 higher than straw mulch and black plastic under netting, respectively.  The open-air black plastic mulch treatment had consistently the highest weekly maximum temperature throughout the duration of the trial, which was about 6 to 20 °F higher than black plastic mulch with insect exclusion netting, which also had the lowest maximum temperatures of all the treatments.  The second highest weekly maximum temperatures were in the hay mulch with netting treatment, which were 3 to 12 GDD higher than hay mulch in open air.  Although these differences in maximum temperature did not affect GDD accumulation in this fall trial, they demonstrated the different effects on air temperature of insect exclusion netting when used with different mulch types; compared to mulch in open-air, netting reduced and increased maximum temperatures with black plastic and hay, respectively.  Clearly, insect exclusion netting and mulch type affect the micro-climate, which could be either detrimental or beneficial to the crop, and knowing these effects should be known when a mulch and netting combination is selected for SM protection during the different seasonal planting times.  For example, netting in combination with hay mulch should be avoided for summer plantings to avoid excessive damage caused by heat stress.

3. Determine whether garlic essential oil will work in the field as a SM repellent and effectively reduce SM damage.

In our two trials, 6 to 8 weekly applications of 1% essential garlic oil applied to spring broccoli failed to control SM damage when SM infestation in the untreated was 48% and 97% (Table 20).  Sometimes, we observed the smell of garlic on the treated plants the following week after application, but in general, June 2015 was fairly wet.  Perhaps, the garlic oil was easily washed off with rainfall, especially since it readily beaded on the waxy brassica broccoli leaves. 

Objective 3.

To increase awareness of SM and knowledge of its management practices among at-risk small-scale organic Brassica growers.

See outreach and publication section.

Economic Analysis

Prior to this project, our grower cooperators reported 20 to 100% crop losses due to SM in broccoli, cauliflower, kohlrabi, red cabbage and certain varieties of kale in certain plantings in certain years.  They quoted losses due to SM as costing them: i) up to an over $1,000 per year per farm, ii) $1,200 to $2,400 per broccoli planting, and iii) $4,000 in 2000 broccoli plants. Additionally, two growers cited abandonment of broccoli production, and two reported lost markets due to SM damage.

Crop rotation.  At Blue Heron, in fields that had heavily SM-infested brassica during the previous fall, and consequent, high spring emergence, waiting to plant red cabbage until after the majority of spring emergence was completed resulted in a 69% increase in marketable yield.  In Field No. 4, red cabbage that was planted during spring SM emergence (Fig. 19) had only 28.3% marketable heads at harvest (Table 10).  In Field No. 1, red cabbage that was planted after the majority of spring SM emergence was completed (Fig. 19) had 98% marketable heads at harvest (Table 11).  Using USDA 2008 Census averages, by avoiding staggering crop losses, crop rotation resulted in $4472 per acre in increased revenue.   

Insect exclusion netting. With insect exclusion netting in our trials resulting in 92 to 100% reduction in unmarketable broccoli heads at harvest, according to averages derived from 2008 USDA Census data, this would result in prevention of crop losses to the tune of $3,130 to $4,316 per acre.

Since the spring broccoli netting trial at Muddy Fingers was the only trial from which we collected yield data, we will use it to conduct an economic analysis (Table 21).  Excellent SM control provided by insect exclusion netting resulted in significantly doubling the marketable yield.  With a selling price of $3.00 per lb, the netting resulted in an increase in sales of $78.90 to $116.70 per 100-ft bed.  Since hoops, stakes and clamps are reusable for several additional uses, they were not included as part of the input costs.  Only the cost of netting, mulch, and labor to set up netting and mulch, and hand weeding were included in the input costs.  Because insect exclusion netting is expensive and the major expense of this SM management strategy (Table 22), the net profit (sales – input costs) was in the negative by $21.84 to $74.21 per 100-ft bed.  To make the use of insect exclusion netting profitable, the netting would have to be reused.  In this trial, the only treatment that would be more profitable than the untreated (bare ground open-air; net: $52.50/100 ft bed) when the netting is reused twice would be the netting + black plastic mulch treatment (net $66.96/100 ft bed).  This treatment had one of the highest yields, did not require any labor for hand weeding and used an inexpensive mulch.  If the netting could be used three times, all treatments except netting over bare ground profited more than the untreated, with netting over black plastic mulch being the most profitable at $95.96/100-ft bed, $43.46 more than the untreated.  The values of: i) resuming a once lucrative broccoli crop that had to be abandoned due to SM; ii) fulfilling a contract with confidence; iii) keeping customers and iv) gaining customers due to a reliable and diverse offering that was previously not possible due to SM may have a much higher value that of just the crop itself.  Additionally, the added protection of Imported cabbageworm and flea beetles, and advanced maturity and increased quality in general may gain a higher price for product or gain access to an earlier market, which could further increase the net profit of using insect exclusion netting to manage SM.  

                It is feasible to reuse insect exclusion netting.  The most likely place for tearing the netting is where it is secured to the ground, especially if is secured with soil or sod, because weeds grow through the netting and it tears very easily at this location when it is removed from the hoops.  Care must also be taken to not make holes in the netting when the clamps are removed from the hoops.  Another strategy to save on the cost of netting could be to buy wider netting and use it to cover two adjacent hoops simultaneously; generally, when product is purchased in bulk it is cheaper.  The shorter the period of time that the netting is used per planting, generally the better shape it is in.  Muddy Fingers uses the same netting for both spring and late-summer broccoli plantings successfully, and has reported using the same netting for up to four plantings.