ORIGINAL RESEARCH ARTICLE

Fine-scale infestation pattern of Bactrocera dorsalis(Diptera: Tephritidae) in a mango orchard in Central Mozambique

Luis Bota1 & Braine Fabião1& Marc De Meyer2 & Lourenço Manuel3 & Maulid Mwatawala4 & Massimiliano Virgilio2

&

Laura Canhanga3 & Domingos Cugala3

Received: 25 October 2019 /Accepted: 16 April 2020# African Association of Insect Scientists 2020

AbstractThe production ofmango inMozambique is under threat due to fruit fly infestation, and most of the farmers cannot afford the costof control. Spatio-temporal dynamic ofBactrocera dorsaliswas assessed at farm level inManica Province to provide informationfor better calibration of pest management strategies and thereby minimize the use of pesticides and other costs associated to pestcontrol. 64 Chempac Bucket traps baited with torula yeast were monitored weekly at a 10 ha mango orchard. From the speciescaptured on the traps only B. dorsaliswere considered for the study due to its importance on mango. A universal spatio-temporalkriging model was fitted to investigate the spatio-temporal pattern of the fly. Results of the analysis showed high spatialheterogeneity of the fly over time, with the occurrence spreading from the margins of the mango orchard and infesting the wholeorchard during the period of peak of abundances reaching densities of more than 500 fly per trap per week. These results showsthat the flies that infest this particular orchard may come from themarginal area, especially from local varieties which get maturedearlier. The findings of this study provides information to be taken into consideration by farmers about when and where apply thecontrol measures.

Keywords Spatio-temporal . B. dorsalis . Kriging interpolation . Hotspot . Mango orchard

Introduction

Mango (Mangifera indica L.) is considered an important cropin sub-Saharan Africa as a source of income for both ruralpopulation and commercial farmers contributing to the reduc-tion of poverty, as well as providing nutritional supplements(Vayssières et al. 2008). In Mozambique, mango is the fruitcommodity with highest production after banana, largely bysmall scale farmers (FAOSTAT 2015). The occurrence of

fruit flies including Bactrocera dorsalis Hendel is the mostimportant factor hampering the production of the crop inAfrica in general and Mozambique in particular (Ekesi et al.2009; Mwatawala et al. 2006; Cugala 2011). Yield losses dueto fruit fly infestation are reported to be more than 40% with-out control (Ekesi et al. 2016). Some of the control measuresfor this group of pests includes male Anihilation techniques,biological control, orchard sanitation, mass trapping andbaiting (Ekesi and Billah 2007). The costs of control can bevery high, exceeding about $7.5 Million per annum in SouthAfrica’s Western Cape (Barnes 2016) and about $6–10Million per annum in Maghreb region (IAEA 1995). Thiscosts can be reduced or optimized if measures are appliedbefore the population reaches the economic threshold(Kogan 1998). Insect population including fruit flies are spa-tially heterogeneous in their densities (Papadopoulos et al.2003), and this heterogeneity together with temporal changesconstitute an important element in development of manage-ment programs. Spatio-temporal patterns of organisms aredriven by the natural environmental heterogeneity(Kounatidis et al. 2008) and by intrinsic factors of the speciessuch as developmental rate, species behavior or resource use

* Luis Botaluisbota@yahoo.com.br

1 National Fruit Fly Laboratory, Provincial Directorate of Agricultureand Food Security, Po Box 42, Chimoio, Manica Province,Mozambique

2 Royal Museum for Central Africa, Invertebrates Section & JEMU,Leuvensesteenweg 13, B-3080 Tervuren, Belgium

3 Faculty of Agronomy and Forest Engineering, Eduardo MondlaneUniversity, Maputo, Moçambique

4 Department of Crop Science and Horticulture, Sokoine University ofAgriculture, Morogoro, Tanzania

International Journal of Tropical Insect Sciencehttps://doi.org/10.1007/s42690-020-00152-5

pattern (Fievet et al. 2007). Fruit flies are known to be goodfliers (Kounatidis et al. 2008) and understanding the pattern oftheir spatial distribution over time and related factors is ofgreat importance for better site-management (Castrignanòet al. 2012).Methods for spacio-temporal analysis are groupedinto two: deterministic and geostatistic (Mello et al. 2003).Currently, because of the development of several computersoftware packages, geostatistic methods are much betterknown among which kriging is the one commonly used(Mello et al. 2003). Kriging is an estimating geostatic methodwhich takes into consideration the spatial characteristics ofautocorrelation of regionalized variables (Jakob 2002;Assumpção et al. 2013). In the regionalized variables theremust be a certain spatial continuity, which allows the dataobtained by sampling of certain points to be used in order toparameterize the estimation of points where the value of thevariable is unknown (Berveglieri et al. 2011; Yamamoto2009). However, it may needs many years to fully characterizethe spatial and temporal variability of insect patterns and thendevelop pest monitoring and management strategy(Castrignanò et al. 2012). B. dorsalis is one of the economi-callymost important fruit fly pests in the world due to its direct

and indirect damages (Bo et al. 2014). Many population dy-namic aspects of this fly have been studied during the lastdecade, mainly in Africa (under its junior synonymB. invadens) where it is causing huge damages. Most of thesestudies analyses the temporal pattern of the fly’s population(Vayssières et al. 2009, 2014, 2015; Mwatawala et al. 2006;Mayamba et al. 2014) concluding that B. dorsalis is moreabundant during the raining season, a period that coincideswith the high abundance of its main hosts. Nevertheless, thereare no studies regarding its spatio-temporal dynamics and lit-tle is known about its infestation dynamics within orchards inAfrica. This study aims at providing a first characterization ofthe fine-scale infestation dynamics of B. dorsalis within amango orchard during one mango season in Mozambique.

Material and methods

Study area

Data were collected in a 10 ha mango (Mangifera indica L.)orchard in Mozambique, Manica Province, Vanduzi District

Fig. 1 Map of location of Vandúzi district

Int J Trop Insect Sci

(−18.94o Latitude; 33.18o Longitude; 700 m) (Fig. 1). Twovarieties of mango are produced in the orchard, Tommy(occupying more than 80%) and Keitt (less than 20%).The orchard is bordered by permanent pastures, maize(Zea mays L.), trees of local mango variety and guava(Psidium guajava L.) (Fig. 2). The study area has atropical climate, with an annual average temperature of21.2 °C, and annual rainfall of 1000 to 1020 mm. Thesoils consist mainly of red oxides, are deep and welldrained, with a rolling topography (MAE 2005).

Trapping

Trapping was done following IAEA guidelines (IAEA2013). Twenty-six plots were demarcated in the orchard(13 in the outer sectors of the orchard and 13 in the innersectors). Each plot was composed of six labeled trees. Tworeplicate traps were placed within each plot on differenttrees (Fig. 2). In order to ensure the independence of repli-cates at each sampling date, the two traps were randomlyshifted among the six trees. Three permanent traps were alsoplaced as a control on four external points around the or-chard, making twelve in total. The permanent traps wereplaced to assess the abundances of fruit flies outside theorchard. In total 64 Chempac Bucket traps (©InsectScience) were placed, baited with torula yeast (©InsectScience) (250 ml per trap), diluted in water at the propor-tion of 3 tablets per 1 l of water. The torula solutionwas replaced weekly, after washing the traps. All usedtorula yeast was removed from the orchard to avoidcontamination. The position of each trap was deter-mined with a GPS Garmin etrex 30.

The data were collected during the mango season 2014–2015 (from September 2014 till February 2015).

Species identification

All fruit flies captured in the traps were preserved in 70%ethanol, taken to the Fruit Fly Laboratory of Chimoio andidentified to genus and species level, using the electronickey developed by Virgilio et al. (2014). After identificationto species level, only B. dorsalis was considered for analysis,because of its significance and importance as mango pest(Ekesi et al. 2009; Vayssières et al. 2015). Fruit fly abundancewas expressed according to FTW index (number of adult fliesper trap per week).

Data analysis

A universal spatio-temporal kriging model was fitted to inves-tigate the spatio-temporal pattern of fruit fly. Let Z (s,t) denotethe density of fruit fly in trap s (defined by coordinates latitudeand longitude) at week t. The model assumes that the spatio-temporal process of the prevalence of fruit fly is a stochasticprocess defined by Eq. 1.

Z s; tð Þ ¼ ηþ ε s; tð Þ ð1ÞWhere η is a constant mean and ε is the spatio - temporalcorrelated stochastic component with zero mean and covari-ance structure given by Σ(θ) (Heuvelink and Griffith 2010).To estimate the spatio-temporal covariance structure of ε(Σ(θ)), we assumed that the variance of ε is constant and thatthe covariance between two positions (s, t) and (s + h, t + u)only depends on the separation distance (h, u), where h is the

Fig. 2 Location of the plots/traps (Source: Google earth 2016)

Int J Trop Insect Sci

Euclidean spatial distance (metres) and u is the distance intime (weeks). We assumed ε to be stationary and spatiallyisotropic. The spatio-temporal covariances are usually de-scribed using a spatio-temporal variogram (γ), which mea-sures the spatio-temporal dependence between data separatedin the spatio-temporal domain using the distance vector (h, u)defined as follows:

γ h; uð Þ ¼ 1

2E ε s; tð Þ−ε

�sþ h; t þ u

�h i2ð2Þ

Where γ (h,u) denotes the semivariance of ε and E denotes themathematical expectation. In this study, we use an inseparablemodel called “Sum-metric” model. This assumes that thespatio-temporal variogram consists of three stationary and in-dependent components:

γ h; uð Þ ¼ γs hð Þ þ γt uð Þ þ γjoint

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffih2 þ k:μð Þ2

q� �ð3Þ

Where γs(h) and γt(u) are purely spatial and temporalvariograms respectively, γjoint(.) is a joint variogram and k isa real coefficient. The Sum-metric model can be seen as asurface with six parameters: two parameters for eachvariogram (sill and range) and a joint spatio-temporal silland nugget. According to Hu et al. (2015) the parameters ofeach variograms are used in spatio-temporal kriging to com-pute the best linear unbiased predictor for any space-time

point where Z was not observed. The universal kriging inthe spatio-temporal domain is defined as:

ε s0; t0ð Þ ¼ c0Tc−1ε ð4ÞWhere c is the n × n variance-covariance matrix of the resid-uals at the n observation space-time points, as derived fromthe spatio-temporal variogram, c0 is a vector of covariancesbetween the residuals at the observation and prediction points,T denotes matrix transpose, and ε is a vector of residuals at then observation points. The final prediction of density of fruit flyZ at a point (s0, t0) is defined as follow:

Z s0; t0ð Þ ¼ ηþ ε s0; t0ð Þ ð5Þ

All analysis were carried out using R software (RDevelopment Core Team 2011) package spacetime(Pebesma 2012).

Results

The spatio-temporal distribution of the density of B. dorsalisfrom 4th September 2014 up to 27th February 2015 is repre-sented in Fig. 3. It shows that the prevalence of fruit fly variesspatially and temporally during the period of the study. Theprevalence of fruit fly trends to increase from low prevalence

2014−09−04 2014−09−11 2014−09−18 2014−09−25 2014−10−02 2014−10−09

2014−10−16 2014−10−23 2014−10−30 2014−11−06 2014−11−13 2014−11−20

2014−11−27 2014−12−04 2014−12−11 2014−12−18 2014−12−26 2015−01−02

2015−01−09 2015−01−16 2015−01−23 2015−01−30 2015−02−06 2015−02−13

2015−02−20 2015−02−27

0

100

200

300

400

500

600

700

800

Fig. 3 Spatio-temporal distribution of fruit fly from September 2014 up to February 2015 in mango orchard

Int J Trop Insect Sci

in September, October and November (less than 100 individ-uals/trap/week) until high densities in December, January andFebruary (more than 100 individuals/trap/week).

Figures 4 and 5 show respectively the empirical spatio-temporal variogram and the sum metric variogram for theprevalence of the fruit fly in mango orchard during the periodof the study. The high variability at both spatial and temporaldimension in both sample variogram and sum metric modelindicates that spatio-temporal correlation is present when

analysing the density of fruit fly in mango orchard fromSeptember 2014 up to February 2015.

Table 1 summarizes the parameters estimates of the Summetric variogram model. All variogram components weremodelled using exponential function. The results show that therange of spatial dependency and temporal dependency is 1.4 kmand 93 days, respectively. This indicates that the prevalence offruit fly in mango orchard is spatially auto correlated up to adistance of 1.4 km, which means that there is a similarity in theoccurrence of fruit for distance less than 1.4 km. If any regioninto the mango orchard present high density of fruit fly, it is alsolikely to observe high densities in distances less than 1.4 km. Inthe other hand, the temporal auto correlation of 93 days indicatesthat the prevalence of fruit fly has similarity over the time. Valuesof fruit fly density trend to be similar or related for a period up to93 days (3 months). In fact, this is observed over the first3 months of the period of the study (September – November)where the fruit fly density trends to be low (Fig. 3). High densi-ties of fruit fly are verified fromDecember to February (period of3 months), although with highest densities in January.

Fig. 4 Spatio-Temporal sample variogram for prevalence of fruit fly

Fig. 5 Sum metric variogram model for prevalence of fruit fly in mangoorchard from September 2014 up to February 2015

Table 1 Parameters estimates for sum metric Variogram model

Components Parameters

Nugget Sill Range

Spatial 1135.95 1941.29 1461.86

Time 1.00 809.63 93.41

Space-time 1282.51 11,667.60 1685.63

Int J Trop Insect Sci

The spatio-temporal kriging in the mango orchard fromSeptember 2014 to February 2015 is presented in Fig. 6.The result showed that the density of fruit fly in mango or-chard trends to be lower than 100 individuals per trap duringthe first 12 weeks (from 04th September 2014 – 20thNovember 2014). From the 13th week up to the 18th weekthe fruit fly prevalence trends to increase from low values upto 200 individuals per trap. High densities of fruit are observedfrom the 19th week up to 22ndweek with the highest densitiesfrom 09th January to 23rd January. The occurrence of fruit flyspreads from the border of the mango orchard and infests thewhole orchard during this period reaching densities more than500 individuals per trap. After the 23rd week up to 26th weekthe prevalence of fruit fly trends to decrease substantially.

Discussion

This study showed that B. dorsaliswas present in the orchard,with fluctuating abundances, across the entire sampling peri-od. Densities of B. dorsalis were low from September toNovember and high from December to February. As alreadyobserved in other studies (Mwatawala et al. 2006) abundancevariations ofB. dorsalis are related to the phenology of its hostplant. In this study area, the maturity of mango occurs fromlate November to February. Similar results were reported byMayamba et al. (2014), where the peak periods of B. dorsalistrap catches corresponded to the harvesting period when man-go fruits were at physiological maturity to ripeness, while the

lowest trap catches corresponded to periods of fruit set whenfruits were very small and immature.

The spatio-temporal diagrams from kriging interpola-tion showed a highly variable distribution of B. dorsaliswithin the orchard thus suggesting associated heterogene-ity of biotic or abiotic factors throughout the orchard,affecting the relative abundance of the fly. It is knownthat agricultural systems are heterogeneous systems con-taining different arrangements of soils, habitat, microcli-mate, and plant communities and productivity (Sciarrettaand Trematerra 2014). These environmental variabilityproved to be the key that drives the spatial and temporalpattern of organisms (Kounatidis et al. 2008). Israely et al.(2005), reported on a study conducted in Israel that thespatial heterogeneity demonstrated in their spatio-temporal analysis was affected by the phenology of dif-ferent hosts within or outside the orchard, the differencein the period of maturation, fruit host abundance and dis-tance between them.

The dispersion pattern of B. dorsalis during the studyprogressed from the borders throughout the entire orchard.This may suggest that fruit flies move from the orchard tothe neighborhoods when ripe mangoes are not available with-in the orchard and move back to the orchard when host ispresent again. In the neighborhood of the orchard local varie-ties of mango are present and their maturation occurs 1 monthbefore varieties of mango produced within the orchard. Wecan speculate that B. dorsalis relies on external mango trees,as a stepping-stone before moving to the orchard when fruitsare present. Majacunene (2014) reported in Manica Provincethat alternative hosts serves as a refuge for fruit flies duringperiods of no availability of fruits in the orchards, and laterbecame a source of infestation. Deguine et al. (2012) reportedthat fruit flies spend more time outside the orchard roosting,feeding and for protection against natural enemy, going to theorchard only for oviposition activity. So far there is no recordof studies related to the spatio-temporal distribution ofB. dorsalis at such a fine scale. However, Balagawi et al.(2014) reported that, at the beginning on the fruiting season,the abundance of Bactrocera tryoniwas higher at the edges ofa strawberry orchard and it was gradually increasing withinthe orchard as the strawberry season progressed.Papadopoulos et al. (2003) reported a spatio-temporal hetero-geneity of the female population of C. capitata related to thephenology of host trees and to sequential availability of ripe orsemi ripe fruits in the orchard. Alemany et al. (2006) reportedthe existence of hotspots of high densities of C. capitataaround and outside citrus orchards, from where thepopulation dispersed throughout the orchard. Similarly,Wahab et al. (2006) using kriging interpolation identifiedhotspots of C. capitata population in the orchard borderswhich progressively extended to the rest of the orchard.Kriging analysis seems to be an extremely useful complement

Fig. 6 Spatio-temporal kriging using sum metric model for fruit fly inmango orchard from September 2014 to February 2015

Int J Trop Insect Sci

to the FTD index. Information about the local fine-scale dis-tribution of target pests could enable a more calibrated imple-mentation of control measures (i.e. on the most suitable or-chard areas), thus reducing the economic and environmentalcosts of pest control (Alemany et al. 2006).

With respect to pest management, the high populationof B. dorsalis along the border of the mango orchardearlier in the season would justify concentration of controlmeasures at the borders during this period. However, sim-ilar studies should be repeated across multiple seasons forbetter consistency of the conclusion since environmentalfactors can vary from one season to another. Similarly,these results cannot be extrapolated to other farms dueto the difference of biotic and abiotic conditions between.Nevertheless, the results of this study provide a first quan-titative information on the fine-scale distribution ofB. dorsalis and have the potential of assisting in the man-agement of B. dorsalis by providing tools for the timelyimplementation of control measures in the areas of higherabundance of flies, before the population grow up anddisperse to other parts of the orchard.

Conclusions

Our results demonstrate that the distribution of B. dorsaliswere heterogeneous on the orchard, with the population build-ing up from the border to the internal region of the orchard.The population peak were observed in January with densitiesof more than 500 flies per trap per week.

Acknowledgments We thank the Royal Museum for Central Africa forthe funding of the study, project S1TNZ_MOZ_IPM, to EduardoMondlane University and Sokoine University of Agriculture throughthe Framework Agreement with the Belgian Development Cooperation.And thanks to the technicians from the National Fruit Fly Laboratory fortheir technical assistance and help during the field work.

Authors’ contributions Not applicable.

Funding information This study was funded by the BelgianDevelopment Cooperation through the Framework Agreement with theRoyal Museum for Central Africa (Belgium) under project “Spatio-tem-poral population dynamics of fruit fly populations and optimization ofIPM program in Manica Province, Mozambique” (Project Number: S1_TNZ_MOZ_IPM).

Compliance with ethical standards

Conflict of interest The authors declare that they have no conflict ofinterest.

Availability of data and material Not applicable.

Code availability Not applicable.

References

Alemany A, Miranda MA, Alonso R, Martín Escorza C (2006) Changesin the spatial and temporal population density of the Mediterraneanfruit fly (Diptera: Tephritidae) in a citrus orchard. Span J Agric Res4(2):161–166

Assumpção HCP, Ucha JM, Hadlich GM (2013) Uso da Krigagem paravariáveis ambientais nomunicípio deMadre deDeus –Bahia. AnaisXVI Simpósio Brasileiro de Sensoriamento Remoto - SBSR, Foz doIguaçu, PR, Brasil, 13 a 18 de abril de 2013, INPE

Balagawi S, Jackson K, Clarke AR (2014) Resting sites, edge effects anddispersion of a polyphagous Bactrocera fruit fly within crops ofdifferent architecture. J Appl Entomol 138:510–518

Barnes BN (2016) Sterile insect technique for fruit fly (SIT) – the southAfrican experience. In: Ekesi S, Mohamed SA, De Meyer M (eds)Fruit fly research and development in Africa – towards a sustainablemanagement strategy to improve horticulture. Springer,The Netherlands (in press)

Berveglieri A, Junior MM, Piteri MA (2011) Aplicação do interpoladorkrigagem em modelos digitais de terrenos, comparando-se os casosde isotropia e anisotropia. Anais XV Simpósio Brasileiro deSensoriamento Remoto - SBSR, Curitiba, PR, Brasil, 30 de abril a05 de maio de 2011, INPE p 722

Bo W, Ahmad S, Dammalage T, Sto Tomas U, Wornoayporn V, HaqUL, Cáceres UL, Vreysen MJB, Hendrichs J, Schutze MK (2014)Mating compatibility between Bactrocera invadens and Bactroceradorsalis (Diptera: Tephritidae). Journal of Economic Entomology107(2): 623Ð 6629; DOI: https://doi.org/10.1603/EC13514

Castrignanò A, Boccaccio L, Cohen Y, Nestel D, Kounatidis I,Papadopoulos NT, De Benedetto D, Mavragani-Tsipidou P (2012)Spatio-temporal population dynamics and area-wide delineation ofBactrocera oleaemonitoring zones using multi-variate geostatistics.Precis Agric 13:421–441

Cugala DR (2011) Management and mitigation measures for alien inva-sive fruit fly (Bactrocera invadens) in Mozambique. Terminal state-ment prepared for the government of Mozambique and theDepartment of Plant Protection by the food and agricultureOrganization of the United Nations. Food and AgricultureOrganization of the United Nations (FAO)

Deguine J-P, Atiama-Nurbel T, Douraguia E, Chiroleu F, Quilici S(2012) Species diversity within a community of the curcurbit fruitflies Bactrocera cucurbitae, Dacus ciliatus, and Dacus demmereziroosting in corn borders near cucurbit production areas of ReunionIsland. J Insect Sci 12:32–480. https://doi.org/10.1051/fruits/2014033

Ekesi S, Billah MK (2007) A field guide to the management of econom-ically important Tephritid fruit flies in Africa. ICIPE Science Press,Nairobi

Ekesi S, BillahMK, Nderitu PW, Lux S, Rwomushana I (2009) Evidencefor competitive displacement of Ceratitis cosyra by the invasivefruit fly Bactrocera invadens (Diptera: Tephritidae) on mango andmechanisms contributing to the displacement. J Econ Entomol102(3):981–991

Ekesi S, De Meyer M,Mohamed SA, Virgilio M, Borgemeister C (2016)Taxonomy, ecology, and management of native and exotic fruit flyspecies in Africa. Annu Rev Entomol 61:219–238

FAOSTAT (2015) FAO statistics division. http://www.fao.org/faostat/en/#data/QC. Accessed 5 Dec 2015

Fievet V, Dedryver CA, Plantegenest M, Simon JC, Outreman Y (2007)Aphid colony turn-over influences the spatial distribution of thegrain aphid Sitobion avenae over the wheat growing season. AgricFor Entomol 9:125–134

Heuvelink G, Griffith D (2010) Space-time geostatistics for geography: acase study of radiation monitoring across parts of Germany.Geography Anal 42:161–179

Int J Trop Insect Sci

Hu Y, Li R, Bergquist R, Lynn H, Gao F, Wang Q (2015) Spatio-temporal transmission and environmental determinants ofSchistosomiasis Japonica in Anhui Province, China. PLoS NeglTrop Dis 9(2):e0003470. https://doi.org/10.1371/journal.pntd.0003470

IAEA (1995) Economic evaluation of damage caused by, and methods ofcontrol of the Mediterranean fruit fly in the Maghreb. An analysiscovering three control options, including the sterile insect technique.Report of an expert group organized by the joint FAO/IAEADivision of Nuclear Techniques in Food and Agriculture

IAEA (2013) Trapping guidelines for area-wide fruit fly programmes.(Vienna, Austria: International Atomic Energy Agency. Insect PestControl Section, Joint FAO/IAEA Division of Nuclear Techniquesin FOOD and Agriculture), TG/FFP-2013

Israely N, Ziv Y, Oman SD (2005) Spatio-temporal distribution patternsof Mediterranean fruit Fly (Diptera: Tephritidae) in the central re-gion of Israel. Entomol Soc Am 98:77–84

Jakob AAE (2002) A Krigagem como Método de Análise de DadosDemográficos. XIII Encontro da Associação Brasileira de EstudosPopulacionais, realizado em Ouro Preto, Minas Gerais, Brasil de 4 a8 de novembro de 2002

KoganM (1998) Integrated pest management: historical perspectives andcontemporary developments. Annu Rev Entomol 43:243–270

Kounatidis I, Papadopoulos NT, Mavragani-Tsipidou P, Cohen Y,Tertivanidis K, Nomikou M, Nestel D (2008) Effect of elevationon spatio-temporal patterns of olive fly (Bactrocera oleae) popula-tions in northern Greece. J Appl Entomol 132:722–733

MAE (2005) Perfil do Distrito de Manica. http://www.ine.gov.mz/documentos/outros-documentos/perfil-do-distrito-de-manica-manica.pdf/view

Majacunene AP (2014) Avaliação do Impacto da Estratégia de ManeioIntegrado na População da Mosca Invasiva da Fruta, Bactrocerainvadens, na província de Manica. Tese de Mestrado, UEM-FAEF

Mayamba A, Nankinga CK, Isabirye B, Akol AM (2014) Seasonal pop-ulation fluctuations of Bactrocera invadens (Diptera: Tephritidae) inrelation to mango phenology in the Lake Victoria crescent Uganda.Fruits J 69:473–480

Mello CR, Lima JM, Silva AM,Mello JM, OliveiraMS (2003) Krigageme inverso do quadrado da distância Para interpolação dos parâmetrosda Equação de chuvas intensas. Rev Bras Ciênc Solo 27:925–933

Mwatawala MW, De Meyer M, Makundi RH, Maerere AP (2006)Seasonality and host utilization of the invasive fruit fly,Bactrocera invadens (Diptera: Tephritidae) in Central Tanzania. JAppl Entomol 130(9–10):530–537

Papadopoulos NT, Katsoyannos BI, Nestel D (2003) Spatial autocorrela-tion analysis of a Ceratitis capitata (Diptera: Tephritidae) adult pop-ulation in a mixed deciduous fruit orchard in northern Greece.Environ Entomol 32(2): 319 326

Pebesma E (2012) Spacetime: Spatio-Temporal Data in R. J Stat Softw51(7)

R Development Core Team (2011) R: a language and environment forstatistical computing. R Foundation for Statistical Computing,Vienna, Austria. ISBN 3–900051–07-0, URL http://www.R-project.org

Sciarretta A, Trematerra P (2014) Geostatistical tools for the study ofinsect spatial distribution: practical implications in the integratedmanagement of orchard and vineyard pests. Plant Prot Sci 50:97–110

Vayssières JF, Korie S, Temple L, Boueyi SP (2008) The mango tree incentral and northern Benin: cultivar inventory, yield assessment,infested stages and loss due to fruit flies (Diptera Tephritidae).Fruits 63:335–348

Vayssières JF, Korie S, Ayegnon D (2009) Correlation of fruit fly(Diptera Tephritidae) infestation of major mango cultivars inBorgou (Benin) with abiotic and biotic factors and assessment ofdamage. Crop Prot 28:477–488. https://doi.org/10.1016/j.cropro.2009.01.010

Vayssières JF, Sinzogan A, Adandonon A, Rey JY, Dieng EO, CamaraK, Sangaré M, Ouedraogo S, Hala N, Sidibé A, Keita Y, GogovorG, Korie S, Coulibaly O, Kikissagbe C, Toussou A, Bilah M, BineyK, Nobime O, Diatta P, N’dépo R, Noussourou M, Traoré L,Saizonou S, TamoM (2014) Annual population dynamics of mangofruit flies (Diptera: Tephritidae) in West Africa: socio-economicsaspects, host phenology and implications for management. Fruits J69:207–222

Vayssières JF, De Meyer M, Ouagoussounon M, Sinzogan A,Adandonon A, Korie S, Wargui R, Anato F, Houngbo H, DidierC, De Bon H, Goergen G (2015) Seasonal abundance of mango fruitflies (Diptera: Tephritidae) and ecological implications for theirmanagement in mango and cashew orchards in Benin (Centre &North). J Econ Entomol 108:2213–2230. https://doi.org/10.1093/jee/tov143

Virgilio M, White I, De Meyer M (2014) A set of multi-entry identifica-tion keys to African frugivorous flies (Diptera, Tephritidae).Zookeys 428:97–108

Wahab MK, Shazly MA, Deeb MA, Ghiyaty MF (2006) Spatio-termporal monitoring of mediterranean fruit fly ceratitis capititapresence by precision agriculture techniques. J Biol Eng:1178–1186

Yamamoto JK (2009) Cálculo de Mapas de Probabilidade Diretamentedos Pesos daKrigagemOrdinária. Geol. USP, Sér. cient., São Paulo,Revista do Instituto de Geociências. v. 10, n. 1, p. 3–14,Março 2010

Publisher’s note Springer Nature remains neutral with regard to jurisdic-tional claims in published maps and institutional affiliations.

Int J Trop Insect Sci