The Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: Liviidae), is a major economic pest of citrus worldwide (Tiwari et al. 2011; Martini et al. 2014; Monzo & Stansly 2017) and vector of the pathogen, Candidatus Liberibacter asiaticus Jagoueix (Hyphomicrobiales: Rhizobiaceae) (CLas), that causes huanglongbing. The disease affects all commercial citrus varieties accounting for $1 billion in annual losses in Florida, USA, and thousands of lost jobs (Qureshi et al. 2009; Alvarez et al. 2016; Tomaseto et al. 2019; Li et al. 2020). In the US, huanglongbing has reduced production of processed and fresh citrus by approximately 21 and 72%, respectively, from 2007 to 2008 and from 2017 to 2018 (Dala-Paula et al. 2019).

Huanglongbing initially was discovered in South Florida in Aug 2005, and today over 90% of grove acres and over 80% of trees are infected (Zhang et al. 2014; Alvarez et al. 2016). The vector, D. citri, was identified in Florida in 1998 (Zhang et al. 2014; Alvarez et al. 2016). Recommended management practices for huanglongbing include use of disease free stock, inoculum removal by destruction of infected trees, and suppression of D. citri populations with insecticides (Brlansky et al. 2009; Alvarez et al. 2016; Pelz-Stelinski et al. 2017). To date, there is no cure for the disease. Other tactics, including shoot tip grafting, thermotherapy, and antibiotics, have shown modest feasibility as possible huanglongbing therapies under greenhouse conditions, but have not been effective under field conditions (Zhang et al. 2014).

Given the effectiveness of primary and secondary pathogen spread by D. citri, vector control remains a critical component of disease management. Although various vector control strategies have been explored in citrus groves, including cultural, chemical, and biological control, insecticide use has been the most prevalent method employed by growers given the rapid and visible effects on pest mortality (Qureshi & Stansly 2007; Boina & Bloomquist 2015). Although management practices for D. citri vary among Florida citrus growers (Pelz-Stelinski et al. 2017), use of foliar and systemic insecticides is the most common tactic for reducing D. citri populations (Qureshi & Stansly 2009; Boina & Bloomquist 2015). Until recently, growers following standard commercial management practices have made between 8 and 12 insecticide applications per yr, including neonicotinoids, organophosphates, and pyrethroids (Grafton-Cardwell et al. 2013).

Despite the lack of effective organic measures to manage the huanglongbing pathosystem, some growers have continued to produce certified organic citrus. Others have begun to implement low-input (2–4 annual sprays) of conventional insecticides, because the economic toll of huanglongbing has reduced profit (Hodges & Spreen 2012). Furthermore, consequences have emerged following frequent insecticide applications for D. citri management, including insecticide resistance and elimination of beneficial species (Tiwari et al. 2013; Niu et al. 2014; Boina & Bloomquist 2015; Chen et al. 2017).

Another component of D. citri management includes biological control of psyllid populations by natural enemies such as the introduced parasitoids, Tamarixia radiata (Waterston) (Hymenoptera: Eulophidae) and Diaphorencyrtus aligarhensis (Shafee, Alam & Agarwal) (Hymenoptera: Encyrtidae) (Michaud 2002; Qureshi & Stansly 2009). In Florida, T. radiata and D. aligarhensis were imported in classical biological control programs in 1999 and 2000, respectively, but only T. radiata became widely established (Alvarez et al. 2016). In addition to these parasitic wasps, D. citri also are attacked by several species of predators, including coccinellids, syrphid flies, chrysopids, dolichopodids, formicids, and spiders (Aranae) (Michaud 2004; Qureshi & Stansly 2009; Kistner et al. 2016a). Ants can be predators or are involved in mutualistic relationships with hemipterans such as aphids, scales, or psyllids (Calabuig et al. 2015; Amiri-Jami et al. 2017). Some studies have shown that ant exclusion promotes biological control of D. citri; for instance, ant exclusion increased the parasitism of D. citri by T. radiata because of food-for-protection relationships (Navarrete et al. 2013; Tena et al. 2013; Anastasio et al. 2021). Therefore, it is important to understand the nature of the relationship between certain ant species and D. citri (predator-prey versus mutualism), to design effective management practices in citrus agroecosystems.

In Florida, ladybeetles (Coccinellidae) and T. radiata are key natural enemies of D. citri (Michaud 2004; Qureshi & Stansly 2009). Diaphorina citri also is susceptible to fungal pathogens such as Isaria fumosorosea Wize (Hypocreales: Cordycipitaceae) and Beauveria bassiana (Bals.-Criv.) Vuill. (Hypocreales: Cordycipitaceae) (Avery et al. 2009). The effects of biological control on population suppression of D. citri generally have been assumed to be greater within organic than conventionally managed citrus given that these organisms are susceptible to conventional insecticide sprays targeting psyllids (Tiwari et al. 2011). Therefore, it is important to understand the contribution of biocontrol agents to population suppression of D. citri under various management practices.

The sustainability of citrus production in Florida, where huanglongbing is well established, depends on effective management of D. citri. We hypothesized that the impact of natural enemies on populations of D. citri would be lower in citrus groves managed with intermittent application of conventional insecticides than in organic groves with no insecticide sprays, as suggested by Pelz-Stelinski et al. (2017). The objective of this research was to compare the effects of intermittently applied conventional insecticides versus organic management practices on population suppression of D. citri by natural enemies under high huanglongbing incidence. The study was conducted in Florida citrus groves in 2019 by monitoring survival of sentinel D. citri nymphs using exclusion cages, and comparing psyllid and natural enemy abundance between the 2 management practices.

Materials and Methods

Results

ARTHOPOD ABUNDANCE

The abundance of arthropods in groves treated with the 2 management practices are shown in Figure 3. The majority of arthropods belonged to the Coccinellidae, Chrysopidae, Formicidae, Eulophidae, Syrphidae, and Dolichopodidae. Spiders also were observed frequently in both groves. The most abundant arthropods foraging on Valencia trees were ants, averaging (± SE) 7.4 ± 0.68 and 5.36 ± 0.55 individuals per sampling interval in organic and intermittently managed groves, respectively. We found various species of ants including Crematogaster cerasi (Fitch), Pseudomyrmex gracilis (Fabricius), Camponotous floridanus (Buckley), Brachymyrmex obscurior (Forel), Dorymyrmex bureni (Trager), Tetramorium bicarinatum (Nylander), and Monomorium pharaonis (Linnaeus) (all Hymenoptera: Formicidae). There were significantly more ants in organic groves than those managed intermittently with conventional insecticides (P < 0.001). Another commonly encountered group of arthropods was spiders with an average of 3.90 ± 0.28 and 3.51 ± 0.22 individuals counted per sampling interval in organic and intermittently managed groves, respectively. However, there was no significant difference in the abundance of spiders between the 2 management practices (P = 0.820). Similarly, we observed a large number of predatory long-legged flies (Dolicopodidae) which were significantly more abundant in conventional than organic groves (P = 0.002). We recorded 1.90 ± 0.13 predatory flies per sample in intermittently sprayed groves, whereas 1.63 ± 0.15 were observed per sampling date in organic groves. Asian lady beetles, Harmonia axyridis (Pallas) (Coleoptera: Coccinellidae), were the most common generalist predator found in both intermittently sprayed (0.56 ± 0.08 per sampling date) and organically managed (0.26 ± 0.05 per sampling date) groves. There were significantly more lady beetles in intermittently sprayed than organic groves (P < 0.001). Among the groups found, there were few lacewings, Tamarixia (Hymenoptera: Eulophidae), and syrphid flies observed in both types of groves and there were no significant differences among these groups.

Fig. 1.

Average temperature (°C) and relative humidity (%) recorded during surveys at 4 organic (A) and 4 intermittently sprayed (B) citrus grove sites in Florida from Apr to Dec 2019.

Discussion

Our field observations in Florida suggest that insecticide input has an indirect impact on population regulation of D. citri via effects on their natural enemies. As we hypothesized, fewer D. citri survived in our exclusion cage experiments when natural enemies had access to flush with eggs and nymphs; this was evident in organic groves during spring, summer, and fall (Fig. 1). Because there were no insecticides applied during the experimental period in the organic groves, we speculate that the greater psyllid mortality observed in uncaged versus caged treatments was due to mortality inflicted by natural enemies. This result also is consistent with previous investigations employing exclusion cages to measure the impact of biological control (Michaud 2004; Qureshi & Stansly 2009). However, psyllid mortality was not different between caged and uncaged sentinel psyllids in intermittently managed groves that received 2 to 4 annual sprays of neurotoxins, particularly during summer and fall. These results implicate the importance of natural enemies as a source of D. citri population regulation in organic groves, more so than under intermittent use of conventional insecticides.

There was less variation in the density of natural enemy populations in groves sprayed intermittently with conventional insecticides than those managed organically. During our visual observations, ants appeared to be the most frequently observed beneficial arthropods followed by spiders, long-legged predatory flies, and lady beetles in both types of groves. Ants can have mutualistic or antagonistic relationships with hemipterans such as aphids, scales, or psyllids (Amiri-Jami et al. 2017). Certain species of ants have a mutualistic relationship with D. citri and provide psyllids protection against natural enemies and in return receive benefit from the carbohydrate-rich honeydew secreted by them (Milosavljević et al. 2017). For example, the Argentine ant, Linepithema humile (Mayr) (Hymenoptera: Formicidae), provides D. citri with significant protection against predation through their mutualistic relationship (Milosavljević et al. 2017). Exclusion of Argentine ants from citrus increases parasitism of D. citri by Tamarixia by 90% compared to areas where ant populations are not manipulated (Milosavljević et al. 2018). Similarly, the presence of ants in aphid colonies reduces predation by syrphid flies (Amiri-Jami et al. 2017) and parasitism by many aphidiine wasps. There are several species of ants that tend and collect honeydew from D. citri nymphs, including Crematogaster ashmeadi Mayr (Hymenoptera: Formicidae), D. bureni, B. obscurior, and C. floridanus (Michaud 2004; Chong et al. 2010; Navarrete et al. 2013). Other species, including D. bureni, C. floridanus, and P. gracilis, have been observed carrying D. citri nymphs in the field (Michaud 2004; Peña et al. 2008; Monzo et al. 2014). The large numbers of ants found in organic groves during our observations may explain in part the observed composition of the natural enemy community; however, further investigations are necessary to understand the contributing role of each ant species in the interaction between psyllids and their natural enemies. The mutualistic relationship between ants and D. citri might affect the success of biological control given that ants can disrupt natural enemy activity directly. Thus, future categorization of ant species occurring in Florida citrus as mutualistic or predatory could benefit the design of management practices accordingly.

Fig. 2.

Mean numbers of surviving sentinel Diaphorina citri on caged and uncaged flush in organic and intermittently sprayed citrus groves in Florida during spring, summer, and fall. Bars indicated by different letters indicate significant differences (P < 0.05) among treatments.

Another dominant group of natural enemies observed here were spiders. The most common spiders that have been documented to consume D. citri are ghost spiders, jumping spiders, and yellow sac spiders (Michaud 2004; Qureshi & Stansly 2009; Kistner et al. 2016b). We often observed these species in the current investigation. Although we did not observe spiders directly consuming D. citri, some psyllids were found ensnared within their webs. Dolichopodids, also known as predatory long-legged flies, feed on adult D. citri (Chong et al. 2010; Cicero et al. 2017) and were the third most frequently observed arthropods group in our study. More dolichopodids were found in intermittently sprayed than organic groves. Similarly, lady beetles are the important guild of predators that are known to contribute significantly to the mortality of D. citri (Michaud 2004; Pluke et al. 2005; Milosavljević et al. 2018) and we observed more lady beetles in intermittently sprayed than organic groves. Although coccinellid and dolichopodid abundance was higher in intermittently sprayed than organic groves, we did not observe a difference in survival of D. citri between caged and uncaged psyllids in groves treated with conventional insecticides, which suggests an effect of insecticide sprays on natural enemy activity. It is well known that the susceptibility of natural enemies to insecticides varies depending on their mode of exposure, such as through direct contact, uptake from residues on contaminated plant surfaces, and ingestion of prey or floral resources that were in contact with insecticides (El-Wakeil et al. 2013). Although insecticides cause direct mortality of natural enemies, they also have negative indirect effects on feeding and foraging behavior of natural enemies, and thus may have interfered with predation or prey consumption (Cloyd 2012). Such indirect effects could hinder population regulations of herbivores. Both direct and indirect effects of insecticides should be considered when measuring the impact on natural enemy activity, and disentangling these 2 complementary factors requires further investigation.

Fig. 3.

Mean number of arthropods observed per 2-min visual observation on trees in organic and intermittently sprayed citrus groves in Florida (Mar to Dec 2019). Significant differences (*) are defined at P < 0.05 level.

Fig. 4.

Mean number of flushes sampled per tree in organic and intermittently sprayed citrus groves in Florida recorded over time. Significant differences (*) are defined at P < 0.05 level.

Fig. 5.

Mean number of adult Diaphorina citri observed per tree in organic and intermittently sprayed citrus groves in Florida (May to Dec 2019). Significant differences (*) are defined at P < 0.05 level.

Another key natural enemy of D. citri is T. radiata, which was imported into Florida as part of a classical biological control effort. In addition to parasitism by female T. radiata, adult wasps, particularly females, are known to kill a portion of psyllid nymphs by host feeding (Michaud 2004). The difficulties associated with T .radiata establishment in certain target regions in Florida are perhaps related to indirect competition caused by other predators, such as Coccinellidae and Dolichopodidae. For example, Michaud (2004) found that T. radiata suffer mortality (64–95%) due to predation of parasitized nymphs by lady beetles in central Florida. Also, dolichopodids compete for resources with T. radiata and eat them if encountered during normal host searching/feeding behavior (Cicero et al. 2017). In our study, T. radiata rarely were observed in groves under both management regimes. Other predators such as lacewings and syrphid flies also contribute to the mortality of D. citri (Michaud 2004; Qureshi & Stansly 2009) and are known to account for up to 86% mortality of immature D. citri as revealed by field studies conducted in California (Kistner et al. 2016b). Throughout the duration of our investigation, the abundance of lacewings and syrphid flies was not affected by management regime based on our survey data.

Interestingly, despite the lower abundance of natural enemies in organic groves, we found fewer D. citri in these areas compared to those areas managed with intermittent applications of conventional insecticides throughout the yr (Fig. 5). Populations of adult D. citri peaked during May, Jun, and Jul in conventional groves, while similar peaks were not observed in organic counterparts. Another factor explaining higher psyllid populations in intermittently sprayed than organic groves observed here might be that the level of insecticide input used was insufficient to adequately reduce D. citri populations as similarly observed in Pelz-Stelinski et al. (2017), yet sufficient to disrupt normal activity of natural enemies. More D. citri found in intermittently managed conventional groves also could indicate some level of insecticide resistance among populations of D. citri in these groves. Diaphorina citri in Florida exhibit resistance to various insecticides including imidacloprid, chlorpyriphos, thiamethoxam, malathion, and fenpropathrin (Tiwari et al. 2011; Chen & Stelinski 2017; Chen et al. 2018).

An additional factor that could have affected the D. citri population is the availability of flush. Feather flush is the only site where D. citri oviposit and nymphs develop (Pluke et al. 2008). Monthly inspections of flush density revealed that flush was available to D. citri at similar times and levels in both the organic and intermittently managed blocks. Therefore, we believe adult psyllids had roughly equivalent resources for egg laying and development in both types of groves investigated here. Although the amount of flush available to D. citri was higher in organic than conventional blocks during Nov and Dec, the opposite trend in D. citri population density was observed during this period. Similarly, studies have shown that abiotic environmental factors, such as temperature and humidity, can affect the flight activity and distribution of D. citri populations (Martini & Stelinski 2017; Martini et al. 2018; Wang et al. 2019). In the current investigation, we did not find significant differences in microclimate variables between organic and intermittently sprayed sites, thus we assume that weather conditions did not influence psyllid populations here.

Diaphorina citri nymphs are suitable prey for a wide range of generalist predators. Although our arthropod survey did not reveal a clear difference in abundance of key natural enemy groups between the 2 management practices, our exclusion cage experiments indicated an overall greater effect of natural enemies on mortality of D. citri in organic than intermittently managed groves. It is possible that cryptic or nocturnal predators could have contributed to the difference in predation observed between treatments as measured by the exclusion cage experiments. For example, adult lacewings are nocturnal, and larvae of lacewings and syrphid flies are cryptic as well as active mostly during dusk and dawn (Hagen et al. 1999; Canard et al. 2001; Keulder & Van den Berg 2013), which may have been overlooked during our visual observations. In future studies, the sampling of natural enemies could be improved by use of additional methods such as sticky traps or vacuums, in addition to visual observations in order to better estimate natural enemy diversity and abundance (Martini et al. 2015). Also, more detailed investigations elucidating the impact of ants and their interaction with generalist predators on biological control of D. citri could further improve management practices for this insect vector (Milosavljević et al. 2018).

The current study suggests that intermittent applications of insecticide sprays for management of D. citri could sufficiently disrupt activity of natural enemy populations to reduce their impact on population regulation of D. citri, even if populations of natural enemies are not eliminated entirely. In contrast, undisturbed populations of natural enemies in organically managed groves in Florida appeared to maintain populations of D. citri at levels lower than observed in nearby conventional groves treated with only 2 to 4 annual sprays. While our work confirms earlier findings from Pelz-Stelinski et al. (2017) regarding the risk of D. citri spread in citrus grove receiving intermittent insecticide sprays, further confirmation of our results among a greater sample size of organic groves in other regions is warranted. Reduction of D. citri populations can improve yield under conditions of near 100% huanglongbing infection, and a threshold has been established beyond which overspraying for D. citri reduces profit (Monzo & Stansly 2017). Recently, a physiological mechanism in citrus has been proposed that may explain why management of D. citri improves yield, even in areas where all trees express huanglongbing. Although inoculation frequency is not related to pathogen load in the phloem, prolonged feeding by D. citri stunts plant growth and decreases salicylic acid-dependent immune response in citrus (Ibanez et al. 2019). Thus, tree defenses against the CLas causal pathogen of huanglongbing appear compromised when populations of D. citri are not adequately managed. Although biological control is unlikely to impact D. citri populations sufficiently to curtail pathogen transmission, it could impact yield indirectly by contributing to population suppression of D. citri below a measurable action threshold when most or all trees are already infected as occurs in Florida.