Selecting Coriandrum sativum (Apiaceae) varieties to promote conservation biological control of crop pests in south Florida

1 Introduction

Effective conservation biological control of crop pests depends on attracting and maintaining effective populations of predaceous and parasitic arthropods in targeted cropping systems (Gurr et al. 2017; Landis et al. 2000; Shields et al. 2019). Monoculture cropping systems typically do not provide essential nutritional and other resources, which mitigates against meaningful recruitment and establishment of natural enemies of crop pests in the system (Altieri and Whitcomb 2003; Altieri and Letourneau 1982; Jervis and Kidd 1996). This may be especially problematic when pest populations are beginning to establish in the crop and natural enemy abundance and recruitment is low (Altieri et al. 1982). Adding complimentary non-crop plants to the crop may be an effective management strategy for attracting and maintaining natural enemies in cropping systems (Gurr et al. 2024; Lundgren 2009; van Emden 2003; Wäckers 2005).

This is because these complimentary non-crop plants attract natural enemies (Colazza et al. 2023; Foti et al. 2017) and provide them with critical resources, such as shelter, nectar, alternate prey, and pollen, also referred to as ‘SNAP’ resources (Gurr et al. 2017; Snyder 2019). ‘SNAP’ plants have been shown to promote and support natural enemy recruitment, survivorship, growth, and fecundity in agricultural ecosystems (Bottrell et al. 1998; Cortesero et al. 1998; Gurr et al. 2017, 2024; Jervis and Kidd 1996; Wolcott 1942).

Umbels, the open flat inflorescences that are typical of the plant family Apiaceae are a good SNAP resource because they provide food for many beneficial insects. Umbels consist of numerous florets whose exposed nectar and pollen are fed upon by diverse assemblages of predators, parasitoids and pollinators (Bugg and Wilson 1989; Proctor and Yeo 1972; Wojciechowicz-Zytko 2019). The Apiaceae family includes culinary plants such as carrots, parsley, fennel, dill, and caraway.

Another familiar apiaceous plant, Coriander sativum L., known as coriander and cilantro, is a hardy annual herb that has been used for the past 8,000 years for its pungent foliage, aromatic seeds, and medicinal properties (Diederichsen 1996; Diederichsen and Hammer 2003; Ramadan 2023; Spence 2023). The plant grows to between 0.2 and 1.4 m tall, has a short life cycle, is adaptable to many different soil and growing conditions and is cultivated worldwide (Diederichsen 1996; López et al. 2008). Its rapid life cycle allows it to fit into different growing seasons, making it possible to grow the crop under a wide range of conditions (Kassahun 2020).

The wild progenitor of cultivated coriander is unknown, though a recent study suggests this species originated in the eastern Mediterranean region (Arora et al. 2021). There are three subspecies of coriander, sativum, microcarpum, and indicum, which are further divided into 10 botanical varieties (Arora et al. 2021; Diederichsen and Hammer 2003). According to Diederichsen and Hammer (2003), the subspecies sativum predominates in the Mediterranean area and western Europe while subspecies indica arose in Pakistan and India, and subspecies microcarpum originated in the Caucasus and then spread widely into southeastern Europe, Central Asia and China. The domestication of these various C. sativum varieties across different geographical regions has yielded a wide range of phenotypically diverse landraces that vary in their morphological traits and essential oil composition (Arora et al. 2021; Diederichsen 1996).

In the U.S., commercially available C. sativum seeds are predominately varieties of ‘cilantro’, which is the name also used for the soft, fragrant leaves used in salsa and garnishes that these plants produce. ‘Cilantro’ varieties have been bred from the subspecies sativum and microcarpum and have a prolonged juvenile phase during which a large basal rosette of fragrant leaves is produced (Diederichsen and Hammer 2003; Spence 2023) and may take up to 90 days before flowering (López et al. 2008). On the other hand, C. sativum varieties that are grown to produce the aromatic seeds which are used as a spice or mixed into a masala are known as ‘coriander’ (in the U.S.). These ‘coriander’ varieties are from the subgenus indicum and have been selected to flower quickly (within a few weeks) and profusely and to produce large seeds with highly aromatic, flavorful oils (Diederichsen and Hammer 2003; Spence 2023), which is the harvested commodity. Seeds of ‘coriander’ varieties are difficult to obtain in the U.S. as most of the commercial demand is for ‘cilantro’.

Coriandrum sativum plants are highly multibranched and each branch finishes with a compound umbel, which is open for 5- to 7-days (Diederichsen 1996). Many anthophilous insects are attracted to C. sativum and outcrossing from pollinators increases seed setting and fruit yield (Bhowmik et al. 2017; Patil and Pastagia 2016). A disk-like nectary is located at the base of the style (‘stylopodium’) while the stamen protrude above the pinkish-white recurved petals (Diederichsen 1996), making the nectar and pollen accessible to a wide variety of natural enemies (Baggen and Gurr 1998; Patt and Rohrig 2017; Patt et al. 1997a,b; van Rijn and Wäckers 2016; Vattala et al. 2006). The nectar and pollen of C. sativum provides important nutrients to natural enemies of crop pests (Baggen and Gurr 1998; D’Ávila et al. 2016; de Abreu et al. 2017; Patt et al. 2003; Resende et al. 2017; Togni et al. 2016; van Rijn and Wäckers 2016). The odors emitted by coriander flowers are attractive not only to pollinators but to generalist insect predators as well (Lenardis et al. 2017). Generalist insect predators also are attracted to the odors emitted by vegetative C. sativum plants (Salamanca et al. 2015, 2018; Togni et al. 2016). Some predator species preferentially oviposit on coriander plants and use the foliar volatiles to assess habitat quality (Togni et al. 2016). The maturing infructescence can be infested by several aphid species, which are a major pest of coriander when it is grown as a seed crop (Meena et al. 2017). Natural enemy populations develop on aphid-infested umbels (Kant et al. 2019; Kumar et al. 2023; Pareek et al. 2014; Regar et al. 2022; Swami et al. 2018), which then have the potential to migrate to adjacent crops. Studies have shown that the addition of coriander as an insectary plant to cropping systems can enhance biological control of crop pests (Baggen and Gurr 1998; Medeiros et al. 2009; Mena and Gospodarek 2024; Patt et al. 1997b; Resende et al. 2010; van Rijn and Wäckers 2016).

Many citrus groves in Florida have row middles that are bare earth or covered with ruderal vegetation and exotic grasses. Because these plants provide few nutritional resources, the citrus grove environment is not very supportive of predaceous and parasitic insects and is not optimal with respect to promoting biological control of citrus pests. Growing ‘SNAP’ plants along the borders and in the row middles should increase the abundance and diversity of predaceous and parasitic insects in citrus groves, leading to suppression of pest insects. Because studies have established that coriander flowers support a wide variety of predaceous and parasitic insects (Baggen and Gurr 1998; Patt and Rohrig 2017; Patt et al. 1997a,b; van Rijn and Wäckers 2016; Vattala et al. 2006), C. sativum could be an ideal SNAP plant to use in citrus groves. Coriander plants attract and support coccinellids, syrphids, and chrysopids, which prey upon and suppress populations of Asian citrus psyllid (Diaphorina citri Kuwayama; Heteroptera: Psyllidae) (Michaud 2002; Michaud et al. 2004; Qureshi et al. 2009, 2014; Kondo et al. 2015; Hoddle et al. 2022; Qasim et al. 2024), the most serious citrus pest insect in Florida (Halbert and Manjunath 2004; Grafton-Cardwell et al. 2013; Hall et al. 2013). This study represents a first step towards selecting C. sativum accessions that would perform well as insectary plants in Florida citrus groves.

The first goal of this study was to screen C. sativum accessions to identify those that would grow vigorously and flower quickly for the purpose of providing nutritional resources to support natural enemies in citrus groves. These selected accessions would then be grown and allowed to open-pollinate for several generations to produce a population adapted for local conditions. The open-pollinated progeny could then be tested for their capacity to attract and support natural enemies of citrus pest insects. The second goal was to co-sow seeds of the open-pollinated ‘coriander’ with seeds of a commercially-available ‘cilantro’ variety and determine the extent to which their flowering and fruiting periods overlapped. The idea here was that coriander plants from the open-pollinated seed should flower rapidly and begin to complete their blooming period as the cilantro plants complete their prolonged juvenile phase and begin to flower. The resultant interplanted population would result in an extended period of floral and prey resource availability for natural enemies versus planting either one by itself.

2 Materials and methods

3 Results

3.1 Coriander accession screening test

Seeds from all accessions germinated and produced plants, although the maximum total percentage of seeds that germinated varied from a low of 20 % to a high of >90 % (Table 2). The accessions that had the highest percentage of seeds that germinated were PI 274291 (91.7 %) and Ames 7546 (88.3 %), which originated from markets in the Indian states of Gujarat and Himachal Pradesh, respectively. The seedlings of all accessions had similar mean health rankings that ranged from 3.9 to 5.0, showing they were in good to excellent health, except for the single Iranian accession, PI 253146, which differed from the others with a health rank of 2.7 (F _(20,209) = 5.644; P < 0.001). The number of days until maximum germination varied from as few as 14 days (PI 674300, which originated in Ethiopia) to as many as 45 days after sowing (PI 674314, which originated in Pakistan).

Table 2:

Germination of Coriandrum sativum accessions received from United States Department of Agriculture North Central Region Plant Introduction Station in Ames, Iowa, USA, in July 2018. Accessions are ranked by maximum percentage of seeds that germinated (n = 90 seeds sown/accession) (column 7). Information is provided about each accession’s geographic origin, including the differences in latitude and elevation between the collection sites and the study site in southeastern Florida (27.3 °N, elevation 10 m), its infraspecific designation (subspecies) (Diederichsen and Hammer 2003), the weight of 100 seeds (HSWT), the maximum percentage of seeds that germinated, and the days to maximum germination.

Accession ID	Region & country of origin	Subspecies	Elevation difference (m)	Latitude difference (°)	HSWT (g)	Max. % germination	Days to max. germination
PI 274291	Himachal Pradesh, India	indicum	1,269	4.7	0.7	91.7	28
Ames 7546	Himachal Pradesh, India	indicum	2,266	4.3	0.6	88.3	38
PI 665270	Sfax, Tunisia	sativum	0	7.1	1.4	75.0	21
PI 664512	Sughd, Tajikistan	microcarpum	462	13.3	0.5	71.7	21
PI 674291	Thimphu Valley, Bhutan	indicum	2,140	0.2	0.5	70.0	24
PI 669960	Bihar, India	indicum	40	−1.8	1.3	68.3	28
PI 665271	Tunisia	sativum	0	9.9	1.0	68.3	28
PI 674314	Buluchistan, Pakistan	indicum	220	1.9	0.7	66.7	45
PI 665268	Médinine, Tunisia	sativum	240	4.9	1.4	66.7	38
Ames 32045	Agadir-Ida-Ou Tanane, Morocco	sativum	64	3.0	1.6	65.0	38
PI 665269	Sidi Bouzid, Tunisia	sativum	365	7.7	1.3	65.0	24
PI 674300	Baluchistan, Pakistan	indicum	1,790	2.2	0.7	65.0	14
Ames 32046	Draa-Tafilalet, Morocco	sativum	999	4.3	1.6	63.3	34
PI 674302	Buluchistan, Pakistan	indicum	180	2.3	0.5	61.7	21
PI 274292	Rajastan, India	indicum	221	−1.0	1.7	60.0	18
PI 669963	Bihar, India	indicum	40	−1.8	0.8	58.3	21
PI 674301	Buluchistan, Pakistan	indicum	2,400	3.1	0.4	48.3	28
PI 193770	Ethiopia	sativum	2,133	−18.2	1.2	46.7	18
PI 274290	Gujarat, India	indicum	199	−3.2	1.5	28.3	24
PI 253146	Isfahan, Iran	microcarpum	1,564	5.1	0.6	26.7	21
PI 193769	Ethiopia	sativum	2,133	−18.2	0.8	20.0	42

With respect to the amount of time between sowing and flowering, the accessions could be divided roughly into three groups (Figure 1, Table 3). The early flowering group consisted of accessions that flowered soon after sowing and included accessions PI 274292 and PI 274290, which originated from markets in the Indian states of Rajasthan and Gujarat, respectively. The plants from these two accessions began flowering 48 days after sowing and reached peak flowering 60–65 days after sowing, with >70 % of the plants flowering. The mid-flowering group consisted of accessions that began flowering 56–83 days after sowing and reached peak flowering approximately 85–90 days after sowing, with >70 % of the plants flowering. This group included accessions PI 193770, which originated in Ethiopia, and PI 669960, and PI 669963, which originated from the Indian state of Bihar. The late flowering group consisted of accessions that began flowering at approximately 93 days after sowing and reached peak flowering 108–111 days after sowing, with <70 % of the plants flowering. This group included accessions PI 674302 and PI 665271, which originated from the Pakistani state of Buluchistan and northern Tunisia, respectively.

Figure 1:

Percentage of Coriandrum sativum plants in each accession in flower during the 3 weeks of the peak blooming period at the study site in southeastern Florida. Only accessions in which >50 % of the plants flowered are shown.

Table 3:

Flowering levels of Coriandrum sativum accessions received from United States Department of Agriculture North Central Region Plant Introduction Station in Ames, Iowa, USA, in July 2018. Accessions are ranked by the mean percentage of plants flowering (column 6) during the peak flowering period. Information also is provided for each accession about the weight of 100 seeds (HSWT), the mean weight (g) of the seeds obtained from the open-pollinated plants (OP F1) and the number of days to initial flowering. Each accession’s geographic origin, including the differences in latitude and elevation between the collection sites and the study site in southeastern Florida (27.3 °N, elevation 10 m), its infraspecific designation (subspecies) (Diederichsen and Hammer 2003) also are listed.

Accession ID	Region & country of origin	Subspecies	Elevation difference (m)	Latitude difference (°)	HSWT (g)	Mean % plants flowering	Mean weight (g) of seed obtained from OP F1	Days to initial flowering
PI 274292	Rajastan, India	indicum	221	−1.0	1.7	0.73	0.59	49
PI 274290	Gujarat, India	indicum	199	−3.2	1.5	0.67	0.51	49
PI 193770	Ethiopia	sativum	2,133	−18.2	1.2	0.62	0.31	87
PI 669960	Bihar, India	indicum	40	−1.8	1.3	0.55	0.28	69
PI 669963	Bihar, India	indicum	40	−1.8	0.8	0.54	0.32	83
PI 674314	Buluchistan, Pakistan	indicum	220	1.9	0.7	0.48	0.26	59
PI 665271	Tunisia	sativum	0	9.9	1.0	0.43	0.35	93
PI 665270	Sfax, Tunisia	sativum	0	7.1	1.4	0.42	0.3	83
PI 665269	Sidi Bouzid, Tunisia	sativum	365	7.7	1.3	0.40	0.32	83
Ames 7546	Himachal Pradesh, India	indicum	2,266	4.3	0.58	0.37	0.16	83
PI 674302	Buluchistan, Pakistan	indicum	180	2.3	0.53	0.37	0.3	93
PI 193769	Ethiopia	sativum	2,133	−18.2	0.8	0.36	0.34	87
PI 674300	Baluchistan, Pakistan	indicum	1,790	2.2	0.7	0.35	0.21	56
PI 664512	Sughd, Tajikistan	microcarp.	462	13.3	0.5	0.34	0.26	77
PI 665268	Médinine, Tunisia	sativum	240	4.9	1.4	0.32	0.4	93
PI 674301	Buluchistan, Pakistan	indicum	2,400	3.1	0.4	0.31	0.28	97
Ames 32046	Draa-Tafilalet, Morocco	sativum	999	4.3	1.6	0.28	0.43	83
Ames 32045	Agadir-Ida-Ou Tanane, Morocco	sativum	64	3	1.6	0.27	0.39	97
PI 253146	Isfahan, Iran	microcarp.	1,564	5.1	0.6	0.26	0.22	93
PI 274291	Himachal Pradesh, India	indicum	1,269	4.7	0.7	0.26	0.22	93
PI 674291	Thimphu Valley, Bhutan	indicum	2,140	0.2	0.5	0.00	NA	NA

The accessions in the early flowering group had a prolonged fruiting period that lasted until 91 days after sowing, at which point the plants senesced (Figure 2). Nearly all of the plants in this group produced seed. The accessions in the mid-flowering group also had an extended fruiting period that lasted until over 120 days past sowing. In the late flowering group, < 60 % of the plants produced fruit, with fruit production extending beyond 125 days past sowing. The two accessions from the first flowering group, PI 274292 and PI 274290, also produced seeds with the highest mean weight (Table 3).

Figure 2:

Percentage of Coriandrum sativum plants in each accession in fruit during the three weeks of the peak fruiting period at the study site in southeastern Florida. The accessions with the highest percentage of plants that fruited are shown.

An overall positive relationship was observed between the weight of the parental seed of each accession (hundred seed weight (HSWT)) and the corresponding mean percentage of parental plants in flower (r(19) = 0.8231, P < 0.001) (Figure 3A). As well, the accessions that had higher percentage of plants in flower per observation produced the largest seeds (r(18) = 0.8754, P < 0.001) (Figure 3B).

Figure 3:

Relationship between seed size and percentage of plants flowering in each Coriandrum sativum accession at the study site in southeastern Florida. A. Relationship between weight of seeds from each accession (the weight of 100 seeds = HSWT), and mean percentage of plants flowering/week; and B. mean percentage of plants in flower/week and weight of resultant seeds (OP F1).

3.2 Co-planting coriander and cilantro test

Seed from both the cilantro (variety Santo) and open-pollinated coriander accessions produced healthy plants that flowered and produced fruit (Figure 4). The tallest umbels of the coriander variety were approximately 55 cm tall while those of the cilantro variety grew to a height of approximately 105 cm (Figure 5). At 31 days after sowing, plants of both varieties were in the juvenile stage (Figure 4). By 36 days, some of the coriander plants were producing floral shoots while the cilantro plants remained in the juvenile condition. The coriander plants began to flower at approximately 45 days and were in full flower by 54 days, while stem elongation had just begun in some of the cilantro plants. The cilantro began to flower at approximately 65 days, overlapping with the coriander flowers. By 76 days, the coriander plants had completed flowering and had mature fruit while the cilantro plants were still producing flower buds and flowers and were beginning to produce fruit. By 91 days, half of the coriander plants had senesced while the cilantro plants remained in full bloom. At 115 days, all the coriander plants had senesced while the cilantro plants continued to produce flowers and fruit, with some plants just beginning to senesce.

Figure 4:

Phenology of container-grown coriander and cilantro varieties of Coriandrum sativum, spring and summer 2022 at the study site in southeastern Florida, showing the extended flowering period obtained when the two varieties are sown together. The orange arrows show the onset and conclusion of flowering the period of the coriander variety and the green arrow indicates the onset of flowering in the cilantro variety.

Figure 5:

Height of tallest umbels observed in container-grown coriander and cilantro varieties of Coriandrum sativum, spring and summer 2022 at the study site in southeastern Florida.

4 Discussion

Our results demonstrate the feasibility of conducting small-scale screening tests to identify C. sativum accessions that grow vigorously and flower quickly in south Florida for the purpose of providing habitat for natural enemies of crop pests. The use of latitude as a guide to narrow the selection of accessions to test worked well overall, although there was a surprising exception, i.e., the Ethiopian accession whose origins were the most different in terms of latitude and elevation from those of the study. Comparisons of the germination success and flowering and fruiting periods of C. sativum plants grown from 21 accessions received from USDA-NCRPIS revealed differences in the levels of germination, fruiting, and flowering among them when they were planted in southeastern Florida. For example, two accessions, PI 274290 and PI 274292, flowered much sooner than the others, with peak flowering occurring 55–65 days after sowing (Figure 1). In comparison, peak flowering occurred 20 days later for accession PI 669960 and not for another 40–60 days later for other accessions tested. In other studies where several C. sativum accessions were co-planted, differences were observed among the accessions in numerous agronomic, morphological, and biochemical traits (Angelini et al. 1997; Berkomah et al. 2022; López et al. 2008).

Considering only the accessions where >70 % of the plants flowered (Figure 1, Table 3), some similarities regarding geographic origin were noted. The accessions that flowered first, PI 274290 and PI 274292, originated in northwestern India, from the states of Gujarat (24.1 °N) and Rajasthan (26.2 °N), which have latitudes similar to that of the study site in southeastern Florida (27.3 °N) (Table 1). Both collection locations and the study site are situated below 250 m in elevation. Of the second flowering group, accessions PI 669960 and PI 669963 were collected in Samastipur, a city in northern India located at 25.5 °N at an elevation below 250 m. However, the other accession in this group, PI 193370, originated in Ethiopia at 9.1 °N and at an elevation above 2,100 m. In an evaluation of 62 C. sativum accessions conducted in Ames, Iowa, USA, the Samastipur accession PI 669960 was in first group of accessions to flower while PI669963 was in the second group (López et al. 2008). Similarity in latitude and elevation of the accessions’ geographic origin relative to the study site may have factored into the rapid umbel production observed here. As well, these four Indian accessions would be classified as C. sativum subspecies indicum, which typically has a short juvenile stage and flowers quickly (Diederichsen and Hammer 2003; López et al. 2008) and would be expected to grow as observed.

All of the accessions produced fruit, with fruit production in general lasting for several weeks after the peak blooming period (Figure 2). In certain management situations, growers may want their insectary plants to reseed to reduce costs and effort associated with re-sowing. For example, a quick-flowering insectary plant could be sown so that it provides habitat for natural enemies before pest populations develop in the crop. The results here suggest that these coriander accessions would readily reseed themselves. Interestingly, there was a strong correlation between the weight of the parental seed and the mean percentage of plants flowering in a particular accession; and in a similar way, the percentage of plants flowering and the weight of the open pollinated (OP) F1 seeds were strongly correlated (Figure 3, Table 3). In C. sativum, seed mass and shape and floral shoot production have high heritability (Arif et al. 2014; Bhandari and Gupta 1991, 1993; Kumar et al. 2022; López et al. 2008; Nair et al. 2012; Verma et al. 2018), and our results align with those findings. Our results also indicate that seed weight is an important trait to consider for varietal selection for conservation biological control purposes because it is predictive of flowering outcome.

The accession screening test was conducted in the cooler part of the Florida growing season. It is possible that different results would have been obtained if the test was replicated in a warmer period of the growing season. However, we have observed that C. sativum grows poorly when temperatures are consistently above 30 °C during the hottest part of the growing season in south Florida and so it is likely to be unsuitable as an insectary crop for the summer.

The two accessions, PI 274290 and PI 274292, which the screening test identified as optimal in terms of rapidity of growth and flowering, were used as parental plants to produce an open-pollinated coriander hybrid adapted for south Florida growing conditions. When these hybrid coriander seeds were co-sown with a cilantro variety, an extended flowering period could be generated, relative to the flowering period of each alone (Figure 4). The results showed that planting either variety alone resulted in flower production for 30 days while co-planting the two varieties extended flower production by approximately 35 days (Figure 4). These results showed that co-planting quick-flowering coriander varieties with bolting-resistant cilantro varieties could provide natural enemy habitat for an additional month without the need to stagger-plant the two varieties to obtain a similar outcome. While it is unknown how the resultant population of C. sativum subspecies sativa × indicum hybrids would sort with respect to the time to onset of flowering, it is likely that bimodal flowering times would persist among the progeny for several years.

An alternative strategy for extending the flowering period of C. sativum plantings would be to co-sow the early-flowering Gujarat or Rajasthan accessions, PI 274290 and PI 274292, with the mid-flowering Bihar accessions, PI 669960 and PI 669963. We anticipate that co-sowing these accessions would result in sequential flowering similar to that obtained from co-planting coriander and cilantro varieties. As these accessions belong to subspecies indicum, the manner in which the resultant hybrids sort with respect to time to onset of flowering might be different than in hybrid populations generated from crossing subspecies sativum and indicum. However, additional experimentation is needed to confirm whether sequential flowering would occur when the different accessions are co-planted.

Other studies have shown that different mixtures or growth stages of C. sativum can influence recruitment of natural enemies. For example, Lenardis et al. (2017) examined three different C. sativum accessions sown in the early and late part of the growing season and evaluated the abundance and diversity of pollinators and natural enemies in the umbels of each accession. They also found distinctive flowering peaks among the accessions, with one accession flowering sooner after sowing than the other two and found that beneficial insect composition and abundance differed among accessions. They attributed these differences, in part, to differences in the floral fragrance profiles among the different accessions. Because the floral volatile profile of each accession attracted a different suite of insects, they suggested that combining different genotypes of C. sativum would enhance overall beneficial insect abundance and diversity in the planting. In other plant species, plots that contained a mixture of genotypes which varied with respect to traits such as flowering phenology, chemical composition of plant tissue, and odor signals, had increased diversity in arthropod species assemblages relative to those plots or patches with fewer genotypes (Bálint et al. 2016; Burkle et al. 2013; Johnson and Agrawal 2005). Further testing is needed to determine whether the floral fragrance and other traits are different among the accessions tested here and whether growing these accessions in combination would have a synergistic effect on promoting the occurrence of natural enemies (Patt et al. 2020).

In a mixture study of C. sativum with a different crop, tomatoes (Solanum lycopersicum L.; Solanaceae), Togni et al. (2016) demonstrated that coccinellids were attracted to the volatiles emitted by C. sativum and that they used these volatiles to assess habitat quality. They found that C. sativum functioned as an oviposition site for lady beetles, and that larval beetles hatched on C. sativum readily dispersed to adjacent crop plants. They suggest that intercropping C. sativum with crop plants will attract lady beetles that oviposit in C. sativum before the arrival of aphids in the crop, thus enhancing biological control of the aphids. Further work is needed to determine whether the foliar volatiles emitted by the different coriander accessions tested here also would have a differing attractiveness to natural enemies of citrus pests.

Patt et al. (2020) found that proportional mixing of insectary plants could synergize the effects of various traits that promote natural enemy abundance and diversity. For example, as shown here and by Lenardis et al. (2017), co-sowing different C. sativum varieties resulted in the production of flowers for a relatively long period. The prolonged availability of both food resources and chemical signaling provided over time by co-planting each accession are expected to result in a strong foraging and colonization response by beneficial insects.

Although not measured as part of this study, observations indicated that aphid abundance can be very high on the umbels and shoots of both the open-pollinated coriander variety and cilantro variety Santo grown in planters at the study site (Patt, unpublished data). This resulted in recruitment of natural enemies, and the development of their immatures (Figure 6) (Patt, unpublished data). Further studies are planned to determine whether proportional mixing of different C. sativum accessions can lead to mixtures that optimally attract and support natural enemies and whether this results in a suppression of pest insects in citrus groves.

Figure 6:

The arrows point to two coccinellid pupae attached to the infructescence of an open-pollinated hybrid variety of Coriandrum sativum spp. indicum. The plant was previously infested with aphids.

In addition to C. sativum, other annual plant species will need to be utilized to provide natural enemies with SNAP resources across the entire citrus growing season in Florida. Examples are established insectary plants, such as buckwheat (Fagopyrum esculentum Moench; Polygonaceae), mustards (Brassica spp. Brassicacea), and sorghum (Sorghum bicolor (L.) Moench) (Bowie et al. 1995; Wäckers and Van Rijn 2012; Harris-Schultz et al. 2022). Growing plants that reliably provide SNAP resources for natural enemies will be key to implementing successful conservation biological control against citrus pests.