Pentatomoids as Vectors of Plant Pathogens 631
for becoming infected at a rate δ, in a frequency-dependent manner, as in model 13.1. We assume that
infected hosts in the population serve as sources for subsequent infections, making the rate of new infec-
tions dependent on the density of infected hosts, I. Although no vector population dynamics are included
in model 13.2, the parameter β could be modeled as a function of vector dynamics in future extensions of
the model. Results from model 13.2 indicate that spread of facultative pathogens is affected by pentato-
mid herbivory rates and transmission rates in qualitatively similar ways (Figure 13.3). Moreover, these
dynamics are similar to those of directly transmitted diseases (Brauer et al. 2008).
In summary, as little work has been done on the epidemiology of pentatomid-borne pathogens, we
developed a series of models to describe the most relevant features of hemipteran-pathogen relationships
and their effects on pathogen spread. We hypothesize that the epidemiology of pathogens with symbiotic
or facultative relationships with pentatomids are fundamentally different. As with other groups of symbi-
otic pathogens (Madden et al. 2000), we predict that vector-recovery rates and transovarial transmission
of pathogens are both important determinants of pathogen spread. Given this, future work should inves-
tigate if pathogens with persistent-propagative relationships with pentatomid vectors can be vertically
transmitted and, if so, at what rates.
For the sake of model generality and tractability, we have ignored many important features of vector
biology in our models. Nonetheless, they provide useful hypotheses that hopefully will motivate future
empirical and theoretical work.
13.8 Future Research Directions
It is clear that pentatomids are important vectors of pathogens. What still is in its infancy is work focused
on sources of pathogen acquisition and differentiating between damage due solely to insect feeding and
vector-borne disease. These factors are significant and should be incorporated into insect thresholds to
adjust appropriate detection levels accordingly. Toews et al. (2010) reported the interrelationship between
cotton losses due to stink bug pressure and boll disease. Based on 10 years of data, the number of bales
attributed to loss due to stink bugs did not correlate consistently with the number of stink bugs detected
in a given year. In other words, some seasons reported relatively low insect numbers and higher than
expected bale losses and vice versa. This disproportion was hypothesized to be due to higher losses
occurring in years that the insects were harboring and transmitting pathogens. This concept was sup-
ported by a separate study that compared the number of diseased bolls collected in plots that were
protected from sucking pests by repetitive insecticide applications with diseased boll numbers from
unprotected plots (Medrano et al. 2015). The incidence of disease was significantly lower in insecticide-
protected plots. These findings warrant future research to identify phytopathogen reservoirs (i.e., weeds,
water, soil, or other crops planted in adjacent fields). Further, are there particular weather situations that
are conducive to elevate pathogen prevalence? Adding the pathogen location information to a farmscape
scenario may assist in developing pathogen-avoidance strategies and adjusting thresholds during seasons
that pathogens are expected at higher rates in the environment.
It has been shown that a spectrum of pathogens acquired and retained by pentatomids are, in turn,
transmitted, causing plant host infection. Future work should be based on aseptic or at least sanitary
dissections of insect tissues for accurate detection of the targeted phytopathogens. Careful analysis will
expand our knowledge and ability to recognize which insect tissues the pathogens invade and inhabit.
Identification of colonization patterns also will assist in learning whether a pathogen resides transiently
or persistently. Further, information of this type could provide a baseline to determine whether the
insects acquire the pathogens horizontally or vertically.
Future work towards definitively demonstrating the vector potential of pentatomids should be consid-
ered. Currently, there are many more associations of vector activities in affected crops than fulfillment of
Koch’s postulates. One manner to address this is the formation of collaborations between entomologists
and plant pathologists. Knowledge gained could be incorporated and assist in further development of
predictive epidemic disease models, such as those described in the preceding section, to increase their
robustness.