The mean number of flies per trap per week increased to a peak of 4.6 on 10 September 2012 to a peak of 11.3 on 29 August 2013. Infested fruit was first found in 2013 on 24 June in Pergine , and on 19 July in Sant’Orsola . First fruit infestation in Pergine in 2012 was determined on 24 June , compared to 28 July 2013 .In order to compare environmental differences between all of the regions in this study, we illustrate degree-day accumulation for 2013 . When comparing differences between seasons, we found basic differences in accumulation for Salem and Pergine during 2012 and 2013. Of the three US locations, accumulation was initially the greatest in 2013 in Wilmington, followed by Parlier and then Salem . However, accumulation of degree-days increased at a higher rate in Parlier and exceeded the accumulation in Wilmington by 3 March. The accumulation in Parlier was greatest for the remainder of the season and was closely followed by Wilmington. Salem accumulation was the least of the three regions presented in the US. In Salem, accumulation followed a similar pattern in 2012 compared to 2013, but the total number of degree-day degree-days was less in 2012 compared to 2013 . In Italy during 2013, the accumulation was greatest in Riva del Garda, followed by Pergine, and then Sant’Orsola . In Pergine, accumulation followed a broadly similar pattern in 2012 compared to 2013. Greater early-season degree-day accumulation in 2012 allowed a higher season-long total in 2012 compared to 2013 .In this study, hydroponic nft channel we demonstrated how temperature-dependent fecundity and survival data could be integrated into a matrix population model to describe relative D. suzukii population increase and age structure according to environmental conditions in four environmentally-distinct fruit production regions.
We found that the environment had major effects on how populations of D. suzukii behaved over a season in the different regions and that the population trends had implications for management. We also found that the different environments affected population stage-structure, and that stage structure also related to management of this pest. To see if independent measures would support population predictions, we used trap and fruit infestation data as well as degree-day estimates for comparison. In general, we found some corroboration of population trends with trap data, and to a limited extent with fruit infestation data. We found that degree day accumulations did not reflect population predictions, and had limited capabilities to predict phenology or voltinism of this pest. The trap counts in our data are from either treated or untreated crops or crops that are unsuitable for D. suzukii population buildup. We realize that data from the traps placed in our study in some cases may, aside from other shortcomings, not provide an accurate early reflection of D. suzukii population levels. The environment had important implications for when populations of D. suzukii were a threat for crops. When comparing predicted population trends from Wilmington, Parlier and Salem, it is apparent that in cooler regions such as Salem, early-ripening fruits would escape D. suzukii attack because early season temperatures are unsuitable for early population increase. In warmer regions such as Parlier and Wilmington, management of D. suzukii should begin as soon as susceptible fruits start to ripen, as favorability of early-season temperatures mean that populations of D. suzukii are high at the beginning of the season. In interiorareas of Califtornia and lower elevations of Italy, there is a midseason decrease in pest pressure as temperatures become very hot and less suitable to D. suzukii. This was true to a lesser extent in North Carolina and Riva del Garda in 2013, where mid-season populations declined only slightly.
An implication of these predicted population trends is that management of D. suzukii during these periods could be less important relative to earlier and later periods when populations peak. Clear differences in stage-specific population structure were found between Wilmington, Parlier and Salem. Temperatures appeared to be better suited for all lifte stage activities and survival in Wilmington throughout the calendar year, followed by Parlier and then Salem. Stability of population structure was highest in Wilmington, followed by Parlier and then Salem. The period characterized by the highest stability in population structure generally coincides with the period when fruit is harvested in some regions. This suggests that that there can be consistent pressure on the crop during harvest, which is a period when allowable pest management activity may be restricted. Stage-specific population structure was generally characterized by a small percentage of adults compared to immature stages in the D. suzukii population. This may explain why traps are such a poor indicator of fruit infestation. For pest management, the only lifte stage targeted currently is the adult stage. Unfortunately, the implications of the predicted age structure are that only a small percentage of the population is affected. This could help explain why frequent spray intervals are required to minimize crop damage from D. suzukii. Unless immature stages can be specifically targeted in future management, it will likely remain challenging to manage this pest with scheduled spray intervals in a way that breaks the lifte cycle of the insect. While we were able to corroborate population projections with independent trap and fruit infestation data, we do not consider these data to be reliable or validating of early D. suzukii pest pressure. Previous literature indicates that action thresholds for fruit infestation demand fruit samples in far greater quantities of the numbers of fruit collected in the current study. In order to get a 5% error rate for a sample containing 0.5% infested fruit, at least 600 fruits need to be collected and searched externally and internally through dissection.
Given these limitations and the limited number of fruit that were collected in this study, fruit infestation could have happened earlier than observed. Erratic trap catches in Wilmington may have reflected yeast activity in monitoring traps used there. In the cases where we had trap and fruit infestation data, the first trap counts were observed during approximately the same period as when the first infested fruit was found. These findings illustrate the fact that traps cannot be seen as an early warning tool. Our predictions of stage structure in populations of D. suzukii illustrate the challenge of discerning distinct generations or predicting important lifte events in any useful way using a traditional degree-day model. Degree-day accumulations also did not directly reflect pest pressure as predicted by the population model, nor did the degree-day trajectories capture the subtlety in population fluctuations. When comparing estimations of population levels between years at Salem, there were clear differences in risk. Temperatures were more suitable for population buildup in Salem during the early portion of the season in 2013 compared to 2012. In Salem during May 2013, D. suzukii population projections were ten-fold higher compared to 2012, indicating the greater potential of crop losses for early-ripening crops. Assessing degree-day accumulations alone, differences between 2012 and 2013 were small and provided limited insight into the far higher early season risk in 2013 compared to 2012. For the annual comparison in Pergine, temperatures were within the optimal range for longer periods during 2013 compared to 2012, and D. suzukii population pressure was projected to be higher. In Pergine, the differences in population are not clearly reflected by degree day estimations, because lower degree-day accumulation was measured in 2013 compared to 2012. For all production regions, D. suzukii population estimation has application for use as a virtual laboratory where ‘what-if’ statements can be raised and answered prior to management action, by simulating population changes as typically achieved during pesticide intervention, Wolbachia infection, or biological control. Age-structured population models can be used to simulate mortality on specific lift stages to predict how different management strategies affect D. suzukii pressure. These strategies can then be validated by field implementation. Population estimation in this study was not aimed at simulating the behavior of individuals or populations of D. suzukii as found in other more complex models for insects of medical importance. Winter survival, availability of suitable host medium, nutrient sources, nft growing system humidity and suitable host plant environments were not taken into consideration in this study. These factors can have strong effects population densities. It is clear that behavior of D. suzukii is important. The migration of flies to track favorable environmental conditions and host suitability was also not taken into consideration when making these estimates. The mechanisms of D. suzukii thermal extreme tolerance are not well documented and need further investigation in order to determine potential adaptation or behavioral mitigation to temperature extremes. Existing literature on Drosophila indicates that mechanisms of thermal tolerance may be influenced by the gene expression of heat shock proteins. These mechanisms have not been studied in D. suzukii but data from future studies on the influence of these factors would benefit population models such as presented here. Other refinements of our model could potentially account for meta population dynamics, host availability, and overwintering survival, and density dependence.
Like all models, ours necessarily makes assumptions and presents a simplified representation of complex ecosystems, ignoring some factors that may influence D. suzukii population levels. However, temperature is clearly one of the most important factors for D. suzukii population growth. This model has clear application for predicting relative pressure from D. suzukii in crops, and can be used as a temperature-related and physiology based comparative risk tool for pending larval infestation. Further application of the population projections would be to extend them into the future based on weather forecasts. Validation of this model will require controlled experiments on D. suzukii to test hypotheses about survival and fecundity based on model output in response to environmental conditions. For example, caged populations of a known size could be subjected to temperature simulations over intervals of time to assess population trajectories and age structures for comparison with model predictions. Additional validation studies could also include passive or live traps to capture flies over the season to be reared in field cages for assessment of survival and reproductions. We believe that the model described here is useful to approximate population levels, to more clearly define the age structure of populations, and to provide additional information that may aid in decision-making for D. suzukii. Given the many complexities in predicting populations, we argue that absolute precision is not necessary to identify effective management interventions or to improve understanding of this pest. We recognize the limitations of our projection model but believe that it represents a novel technique and a potentially powerful tool for management and research on D. suzukii and other damaging insects.The Coffea genus of the Rubiaceae family includes about 100 species , native to Africa, Madagascar, the Mascarene Islands, and Indomalaysia . The seeds, or “coffee beans,” of these small shrubs or trees are roasted, ground, and brewed to make coffee beverages and products . Two slippery seeds are typically found inside each drupe, commonly called a cherry, the fruit produced by the Coffea tree. Trees reach up to 7m tall, with fragrant white flowers that have both male and female sex organs . Of all Coffea species, only Coffea arabica L. is tetraploid and self-fertile . Coffee trees grow best in areas with rich soil, mild temperatures, frequent rain, and shaded sun . About five years of growth after planting is needed to reach full fruit production and trees can live up to 100 years, although they are most productive between the ages of 7 to 20 years old. The average coffee tree produces around 10 pounds of coffee cherries per year, equivalent to 2 pounds of green coffee beans . Commercial coffee production relies mainly on the two species Coffea arabica L. and Coffea canephora Pierre ex Froehner , with C. arabica boasting the better cup quality. In regards to the world supply of coffee, the Arabica coffee plant makes up 75% while the Robusta plant contributes 25% . Global coffee production in the 2015-16 year was estimated to be a staggering 143.4 million bags . In 2015, Arabica coffee plants contributed 84.3 million bags while Robustas contributed 59.1 million bags. Following the increased demand and production of coffee is the swift consumption of the beverage; global coffee consumption reached up to 150.2 million bags in the 2014 calendar year . According to FAOSTAT , the top five producers of coffee were Brazil , Vietnam , Indonesia , Columbia , and India , with Ethiopia at seventh place .